Jan2009-Journal of Porphyrin and Pthcyns

Volum

e 13 Num

ber 1 Pages 1-174

Journal of Porphyrins and P

hthalocyanines 2009 S

PP

MICA (P) 207/01/2009

2009 - ISSN 1088-4246

Volu

me

13 /

Num

ber

1 /

Pag

es 1

-174 January 2009

CONTENTS

Journal of Porphyrins and PhthalocyaninesJ. Porphyrins Phthalocyanines 2009; 13: 1–174

pp. 1–13Synthesis of all-cis and all-trans tetrakis (phenylviny-lene)phthalocyaninesAlexander Efi mov*, Essi Sariola and Helge Lemmetyinen

We have synthesized a series of tetrakis(arylvinylene)phthalocyanines from the corresponding phthalonitriles. According to 1H and 19F NMR data, the cis- or trans-conformations of the starting materials retain du ring condensation, thus phthalocyanines formed are in all-cis or all-trans form. The cis-trans photoi-somerization occurs easily for phthalonitriles, while phthalocyanines retain their conformations under UV beam.

See Katsuhiko Ariga*, Jonathan P. Hill*, Yutaka Wakayama, Misaho Akada, Esther Barrena and Dimas G. de Oteyza (pp 22–34)

We now can align, organize, and watch porphyrins in addition to traditional actions such as designing and synthesizing. This cover displays 2D molecular arrays of por-phyrin families, phenol- and quinone-substituted tetrapyrrole units, including phase transitions (left), molecules at 2D phase boundaries (middle), and hydrogen-bonding networks (right). Molecular level control of arrayed structures of functional mole-cules should reveal new aspects of nanotechnology.

pp. 14–21Photoinduced electron transfer in supramolecu larassemblies involving saddle-distorted porphy rinsand phthalocyaninesTakahiko Kojima*, Tatsuaki Nakanishi, Tatsuhiko Honda and Shunichi Fukuzumi*

Saddle-distorted dodecaphenylporphyrin (H2DPP) can undergo facile pro-tonation to give a stable diprotonated species, H4DPP2+. H4DPP2+ can be a building block to construct novel supramolecular assemblies with hydroqui-none derivatives for “porphyrin nanochannels” and with saddle-distorted Zn(II)-phthalocyanine complexes. In those supramo lecules, H4DPP2+ acts as an electron acceptor in the photoinduced electron transfer.

About the Cover

Articles

CONTENTS

J. Porphyrins Phthalocyanines 2009; 13: 1–174

pp. 35–40Interaction of nitric oxide with Ru(II) complexes of deuteroporphyrindimethylester derivativesRudy Martin, Roberto Cao*, Ana M. Esteva and Franz-Peter Montforts*

A new ruthenium(II) porphyrin disulphide derivative, [Ru(Pds)CO)], was pre-pared from ruthenium(II)(carbonyl) deuteroporphyrin(IX) and cystamine. The interaction of the [Ru(Pds)(CO)] complex with nitric oxide in solution and im-mobilized on gold was studied spectrophotometrically as well as by XPS and cyclovoltametry.

pp. 41–50Structural determination of the 2,10-phytadienyl subs-tituent in the 17-propionate of bacteriochlorophyll-bfrom Halorhodospira halochlorisTadashi Mizoguchi*, Megumi Isaji, Jiro Harada, Kazuyuki Watabe and Hitoshi Tamiaki*

Bacteriochlorophyll-b, having a unique 2,10-E,E-phytadienyl group as the 17-propionate ester was isolated from a thermophilic purple photosynthetic bacterium, and the structure of the ester was unambiguously determined by 1H/13C NMR techniques.

N N

N N

O

Mg

O

CO2CH3

17

O O

2 10

pp. 51–59Synthesis of porphyrin-carbohydrate conjugates using “click” chemistry and their preliminary evaluation in human HEp2 cellsErhong Hao, Timothy J. Jensen and M. Graça H. Vicente*

We describe the syntheses of a series of four new porphyrin-carbohydrate con-jugates containing either one or four galactose or lactose moieties linked via triazole units to a meso-phenyl group of a TPP or TBP macrocycle. The most effi ciently accumulated within human carcinoma HEp2 cells of all the conju-gates was the TBP-galactose, localizing preferentially in the lysosomes. The TPP-galactose and -lactose conjugates were found mainly in the cell ER and endosomes.

N

N

N

NPd

N NN

OOH

H

HOH

H

HHO H

HONN

NO

HO

H

HHO

H

HOHH

OH

NNN

O

HOH

HOH

H

HOHH

HO

NNN

O

HOH

HOH

H

HOHH

HO

pp. 22–34New aspects of porphyrins and related com pounds: self-assembled structures in two-di mensionalmolecular arraysKatsuhiko Ariga*, Jonathan P. Hill*, Yutaka Wakayama, Misaho Akada, Esther Barrena and Dimas G. de Oteyza

In this review, recent research on porphyrin assemblies, including two-dimensio nal porphyrin arrays, is described with emphasis on phenol- and quinone- substituted tetrapyrrole units. A series of research aimed at developing strategies for preparation of porphyrin molecular arrays, where several novel aspects of molecular arrays, including phase transitions, ordered 2-D phase boundaries, and hydrogen-bonding networks, are introduced.

CONTENTS


pp. 70–76Discotic liquid crystals of transition metal com plexes40: unique double-clearing behavior of a se ries of novel discotic metallomesogen based on bis (phthalo-cyaninato)europium(III) complexesHidetomo Mukai, Miho Yokokawa, Kazuaki Hatsusaka and Kazuchika Ohta*

It was revealed for the fi rst time in the phthalocyanine based on metallomeso-gens that the central metal had large infl uence on their clearing points and the phase transition behavior.

M

NNN

N

NN

NNN

N

N NNN

NN

ORRO

OR

OR

OR

RORO

OROR

RORO OR

OR

RORO

OR

pp. 77–83Spectral fi ngerprinting of porphyrins for distributed chemical sensingDaniel Filippini*, Emanuela Gatto, Adriano Alimelli, Muhamad Ali Malik, Corrado Di Natale, Roberto Paolesse, Arnaldo D’Amico and Ingemar Lundström

Recent progress in spectral fi ngerprinting of fl uorescent indica-tors u sing distributed instrumentation based on consumer electronic devices is reviewed. In particular, the evaluation of disposable assays using a com puter screen photo-assisted technique (CSPT) is discussed. Sample iden tifi cation and optimization strategies are analyzed as well as the un derlying theoretical background for polychromatic spectral fi ngerprin ting.

pp. 84–91Electrical transduction in phthalocyanine-based gas sensors: from classical chemiresistors to new functional structuresMarcel Bouvet*, Vicente Parra, Clémentine Locatelli and Hui Xiong

Based on different types of phthalocyanine molecules, the elaboration of va rious devices is described. It is shown how a multimodal detection can be pro fi table to answer the increase demand for selective sensors.

pp. 60–69Structures and properties of hemiquinone-substituted oxoporphyrinogensJonathan P. Hill*, Katsuhiko Ariga and Francis D’Souza

The structure and physical properties of a series of N-substituted hemiquinone-substituted oxoporphyrinogens is presented and discus-sed. Structures, electrochemical properties and use of the compounds as supramolecular tectons are described. Also presented are proper-ties related to guest binding and photophysical properties of oligo-chro mophoric host-guest complexes involving oxoporphyrinogen N-substituted with porphyrins and appropriately substituted fullerene guest electron acceptors.

CONTENTS


pp. 99–106Ground- and excited-state interactions of 2,7,12,17-te-traphenylporphycene with model target biomolecules for type-I photodynamic therapyNoemí Rubio, Víctor Martínez-Junza, Jordi Estruga, José I. Borrell, Margarita Mora, M. Lluïsa Sagristá and Santi Nonell*

The interaction between 2,7,12,17-tetraphenylporphycene and different biomo-lecules including guanosine, 7,8-dihydro-8-oxoguanosine, human serum albu-min, gramicidin A, phosphatidylethanolamine, and lipid A have been studied as a means to assess the potential role of type-I interactions in photodynamic therapy with this photosensitizer.

pp. 107–113Investigating different strategies towards the preparation of chiral and achiral bis-picket fence corrolesMartin Bröring*, Markus Funk and Carsten Milsmann

Several pathways and strategies have been thought and evaluated for the pre-paration of superstructured and chiral corrole ligands. Palladium catalyzed ami dations of o-bromophenyl substituents at the stage of the completed cor ro le macrocycle is the key to a successful synthesis of such sterically encumbe red ligands.

N N

NNH H

H

NH

O

NH

O

HN

O

HN

O

MeO

OMe

OMe

H H

H H

pp. 114–121Bisporphyrin connected by tetrathiafulvaleneKazuya Ogawa* and Yasunori Nagatsuka

A new porphyrin-tetrathiafulvalene composite, where two porphyrins are brid-ged by tetrathiafulvalene using acetylene bonds was synthesized. Spec tros copic and electrochemical properties were investigated. The effective two-photon absorption cross section values were measured by using a nanosecond open aperture Z-scan method.

S

S

S

SN

N N

N

Ar

Ar

Zn

COOEtAr =

N

N N

N

Ar

Ar

Zn

pp. 92–98Solvent-induced supramolecular assemblies of crown-substituted ruthenium phthalocyaninate: morphology of assemblies and non-linear optical propertiesAntonina D. Grishina, Yulia G. Gorbunova*, Victor I. Zolotarevsky, Larisa Ya. Pereshivko, Yulia Yu. Enakieva, Tatiana V. Krivenko, Vladimir Savelyev, Anatoly V. Vannikov and Aslan Yu. Tsivadze

The supramolecular wires over 600 nm in length based on ruthenium(II) tetra-15-crown-5-phthalocyaninate with triethylenediamine molecules have been observed by use of atomic force microscopy. The third order nonlinear optical characteristics of complexes in tetrachloroethane solution were studied by z-scanning method. Molecular polarizability of the complex is about 4.5 × 10-32

esu and increases by factor of 3.6 when the individual molecule assembles into a supramolecular aggregate.

CONTENTS


pp. 136–152Addition of carbenes derived from aryldiazoacetates to arenes using chloro(tetraphenylporphyrinato)iron as catalystHarun M. Mbuvi and L. Keith Woo*

Chloroiron(III) tetraphenylporphyrin, Fe(TPP)Cl, is an active catalyst for the cycloaddition of carbenes obtained from para-substituted methyl 2-phenyldi-azoacetates to arenes to form rapidly equilibrating mixtures of norcaradiene-cycloheptatriene valence isomers in yields over 70%. This one-pot process is used with benzene, chlorobenzene, anisole, p-xylene, p-chlorotoluene (etc.) as substrates.

pp. 153–160Synthesis and photophysical behavior of axially substituted phthalocyanine, tetrabenzotriazaporphyrin and triazatetrabenzcorrole phosphorous complexesEdith M. Antunes and Tebello Nyokong*

Metallophthalocyanines (MPcs), including MPc analogues such as tetrabenzotri-azaporphyrin (TBTAPs) and triazatetrabenzcorroles (TBCs) are effective as photosensitizers depending on the nature of the central metal and axial ligands. The synthesis and photophysics of phosphorous Pc, TBTAP and TBC com-plexes is reported.

pp. 161–165A fi rst ABAC phthalocyanineFabienne Dumoulin*, Yunus Zorlu, M. Menaf Ayhan, Catherine Hirel, Ümit Isci and Vefa Ahsen*

The derouting of the selective synthesis of crosswise phthalocyanines (Pcs) ap plied to a statistical precursors mixture leads to a fi rst member of a new fa mi ly of Pcs bearing three different substituents: the ABAC phthalocyanines. By playing on precursors’ relative ratio, yields and selectivity of the method have been optimized.

N

HN N

N

N

NH

NN

A

B

A

CABAC

phthalocyanine

pp. 122–135Synthesis, structures, and properties of BCOD-fused por phyrins and benzoporphyrinsHiroki Uoyama, Takahiro Takiue, Kazuyuki Tominaga, Noboru Ono and Hidemitsu Uno*

Bicyclo[2.2.2]octadiene (BCOD)-fused porphyrins with no other subs-tituents were prepared by the [2+2] and [3+1] porphyrin syntheses from ethanodihydroisoindole derivatives in fairly good yields. Ther mal retro-Diels-Alder reactions of BCOD-fused porphyrins gave the cor-responding benzoporphyrins with no substituent in quantitative yields.

CONTENTS


pp. 166–170Synthesis of an amide-linked chlorin-anthraquinone dyad and its structural and spectroscopic propertiesThorsten Könekamp, Tobias Borrmann and Franz-Peter Montforts*

An enantiomerically pure chlorin-anthraquinone dyad was synthesized as a model compound for investigation of light-induced electron transfer. The con-formation and electron/energy transfer was investigated by NMR experiments, PM3 calculations and fl uorescence measurements.

AUTHOR INDEX (cumulative)

AAhsen, Vefa 161Akada, Misaho 22Alimelli, Adriano 77Antunes, Edith M. 153Ariga, Katsuhiko 22, 60Ayhan, M. Menaf 161

BBarrena, Esther 22Borrell, José I. 99Borrmann, Tobias 166Bouvet, Marcel 84Bröring, Martin 107

CCao, Roberto 35

DD’Amico, Arnaldo 77D’Souza, Francis 60de Oteyza, Dimas G. 22Di Natale, Corrado 77Dumoulin, Fabienne 161

EEfi mov, Alexander 1Enakieva, Yulia Yu. 92Esteva, Ana M. 35Estruga, Jordi 99

FFilippini, Daniel 77Fukuzumi, Shunichi 14Funk, Markus 107

GGatto, Emanuela 77Gorbunova, Yulia G. 92Grishina, Antonina D. 92

HHao, Erhong 51Harada, Jiro 41

Hatsusaka, Kazuaki 70Hill, Jonathan P. 22, 60Hirel, Catherine 161Honda, Tatsuhiko 14

IIsaji, Megumi 41Isci, Ümit 161

JJensen, Timothy J. 51

KKojima, Takahiko 14Könekamp, Thorsten 166Krivenko, Tatiana V. 92

LLemmetyinen, Helge 1Locatelli, Clémentine 84Lundström, Ingemar 77

MMalik, Muhamad Ali 77Martin, Rudy 35Martínez-Junza, Víctor 99Mbuvi, Harun M. 136Milsmann, Carsten 107Mizoguchi, Tadashi 41Montforts, Franz-Peter 35, 166Mora, Margarita 99Mukai, Hidetomo 70

NNagatsuka, Yasunori 114Nakanishi, Tatsuaki 14Nonell, Santi 99Nyokong, Tebello 152

OOgawa, Kazuya 114Ohta, Kazuchika 70Ono, Noboru 122

PPaolesse, Roberto 77Parra, Vicente 84Pereshivko, Larisa Ya. 92

RRubio, Noemí 99

SSagristá, M. Lluïsa 99Sariola, Essi 1Savelyev, Vladimir 92

TTakiue, Takahiro 122Tamiaki, Hitoshi 41Tominaga, Kazuyuki 122Tsivadze, Aslan Yu. 92

UUno, Hidemitsu 122Uoyama, Hiroki 122

VVannikov, Anatoly V. 92Vicente, M. Graça H. 51

WWakayama, Yutaka 22Watabe, Kazuyuki 41Woo, L. Keith 136

XXiong, Hui 84

YYokokawa, Miho 70

ZZolotarevsky, Victor I. 92Zorlu, Yunus 161


JPP Volume 13 - Number 1 - Pages 1–174

Aacetylene 114amidation 107anthraquinone 166asymmetric phthalocyanine 161atomic force microscopy 92

Bbacteriochlorophyll-b 41benzoporphyrin 122bicyclo[2.2.2]octadiene 122Büchner reaction 136

Ccarbene 136carbohydrate 51catalysis 136chemical sensing 77chiral ligands 107chlorin 166cis-trans isomerization 1click chemistry 51computer screen illumination 77corrole 107crown-ether 92crystal structure 122

Ddiazo reagent 136diiminoisoindoline 161diode 84discotic liquid crystal 70donor-acceptor dyad 166double-clearing 70

Eelectrochemistry 60electronic spectra 122

F19F NMR 1fl uorescence quantum yield 153

Ggold 35

gramicidin 99guanosine 99

Hheterojunction 84human serum albumin 99

Iimmobilization 35iron porphyrin 136

Llanthanoid phthalocyanine 70light-induced electron transfer 166lipid A 99

Mmetallomesogen 70molecular array 22molecular material 84molecular semiconductor 84

Nnitric oxide 35non-linear optical properties 92nucleosides 99

Ooligochromophore 60optical fi ngerprinting 77optical sensing 77oxoporphyrinogen 22, 60

Ppalladium 107PDT 51phospholipids 99phosphorous phthalocyanines 153phosphorous tetrabenzotriaza-

porphyrins 153phosphorous triazatetrabenz-

corroles 153photodynamic therapy 99photoinduced electron transfer 14photorefractive properties 92photosynthesis 41

phthalocyanine 1, 84, 92phthalonitrile 1picket fence corrole 107polluting gas 84porphycenes 99porphyrins 22, 51, 60, 77, 114propionate 41

Rresistor 84retro-Diels-Alder reaction 122reverse-phase HPLC 41ruthenium 92ruthenium porphyrin 35

Ssaddle-distorted phthalocyanine 14saddle-distorted porphyrin 14sandwich-type 70selectivity 161self-assembly 22sensor 35, 84singlet oxygen 99supramolecular assembly 14supramolecular ensembles 92

Tthermophilic purple bacteria 41transducer 84transistor 84trichloroisoindolenine 161triplet lifetime 153triplet yields 153TTF 114two-dimensional 22two-photon absorption 114type-I mechanism 99

Vvalence isomer 136

Zz-scanning 92

KEYWORD INDEX (cumulative)


JPP Volume 13 - Number 1 - Pages 1–174


Copyright © 2009 World Scientific Publishing CompanyCopyright © 2009 World Scientific Publishing Company

INTRODUCTION

Phthalocyanines have been under intensive study for many decades due to their high absorption of visible light, chemical stability, and fl exible preparation meth-ods. The practical applications range from the dyes and pigments industry to TFT displays, lasers, and recently, organic solar cells (OSCs) and organic light-emitting diodes (OLEDs) [1–9].

The side substituents play a crucial role in overall func-tionality of the phthalocyanines, as they determine solu-bility, crystal structure, and (to some extent) the redox potentials. The most commonly studied compounds are the macrocycles with alkyl, alkoxy, and aryloxy groups in either the α- or β-position of the phthalo ring. It should be noted here that phthalocyanines with extended π-conjugation are of special interest in OSC applications as they could benefi t from wider absorption in the visible region and improved electron-donating properties. Such large conjugated π-electron substituents are known, but they often require extra “ballast” of a long alkyl chain to improve the solubility, and therefore, to simplify the synthesis and analysis of the macrocycles [10–24].

To the best of our knowledge, there are only a few examples of arylvinylene-substituted phthalocyanines, synthesized by Torres and Yamazaki [14, 19, 25]. They describe either mono- or octa-substituted compounds, with trans confi guration of the double bonds and inspired us to further develop the approach and prepare the phtha-locyanines with four phenylvinylene groups, located at the β-positions of the macrocyle.

RESULTS AND DISCUSSION

Phthalonitriles are commonly used building blocks for phthalocyanine synthesis. We started out with 4- methyl-phthalonitrile 1 (Fig. 1), which was brominated with NBS in the presence of benzoyl peroxide. Bromomethylphtha-lonitrile 2 was then reacted with triphenylphosphine in dimethylformamide (DMF), and triphenylphosphonium bromide 3 was obtained with 67% overall yield. In the fi rst experiments, the intermediate 2 was separated, subjected to fl ash chromatography, and analyzed by 1HNMR. The bromination degree was around 70%. Extend-ing the reaction duration from 16 up to 24 h did not pro-vide signifi cant improvement in the yield. The attempts to separate bromomethylphthalonitrile 2 from the start-ing compound 1 were unsuccessful, due to the very close mobilities of the compounds on silica. However, using

Synthesis of all-cis and all-trans tetrakis(phenylvinylene)-phthalocyanines

Alexander Efi mov*, Essi Sariola and Helge Lemmetyinen

Department of Chemistry and Bioengineering, Tampere University of Technology, P. O. Box 541, 33101, Tampere, Finland

Received 25 September 2008

Accepted 1 November 2008

ABSTRACT: We have synthesized a series of tetrakis(arylvinylene)phthalocyanines from the corresponding phthalonitriles. Substitution of the para-fl uorine in the pentafl uorphenyl ring with an alkoxy group from the solvent occurs during the synthesis. This substitution can be suppressed by the addition of zinc chloride or tin chloride to the reaction mixture. According to 1H and 19F NMR data, the cis- or trans-conformations of the starting materials are retained during condensation; thus the phthalocyanines formed are in all-cis or all-trans form. Cis-trans photoisomerization occurs easily for phthalonitriles, while phthalocyanines retain their conformations under UV beam.

KEYWORDS: phthalocyanine, phthalonitrile, 19F NMR, cis-trans isomerization.

*Correspondence to: Alexander Efi mov, email: alexandre.efi [email protected] , fax :+358 3-3115-2108, tel: +358 3-3115-3640

Copyright © 2009 World Scientific Publishing Company J. Porphyrins Phthalocyanines 2009; 13: 2–13

2 A. EFIMOV ET AL.

Fig. 1. Structures of the compounds


SYNTHESIS OF ALL-CIS AND ALL-TRANS TETRAKIS(PHENYLVINYLENE) PHTHALOCYANINES 3

the mixture of 1 and 2 without further purifi cation did not harm the yield of phosphonium bromide 3. Wash-ing with diethyl ether effi ciently removes all the impuri-ties from the salt 3, as was demonstrated by 1H NMR and ESI-TOF MS. The mass spectra were measured in both positive and negative modes. The molecule loses the bromine when ionized in positive polarity; thus, the molecular ion with m/z 403.1413 was observed. Two sig-nals were observed in negative mode. The one with m/z483.0221 corresponds to [M - H]- molecular ion, while the signal with m/z 560.9366 is due to the adduct of the molecule with another bromine. The isotopic pattern of those signals matches the elemental model perfectly.

Di-tert-butylphenylvinylene phthalonitriles

Arylvinylene phthalonitriles were synthesized from the phosphonium salt 3 and the corresponding benzalde-hyde by the Wittig reaction. Similarly to the procedures described in the literature [26, 27], the reagents were dis-solved in dimethyl sulfoxide (DMSO), sodium methylate was added as base, and the reaction mixture was stirred at an elevated temperature for several hours. After the usual work-up, the reaction product was isolated as a glassy orange viscous substance and was analyzed by NMR. The proton spectra contained the signals of both cis and trans vinylene protons. TLC analysis did not reveal any suffi cient difference in mobility of the isomers. The spot of the compound was quite uniform and smooth, and the only sign of the presence of cis-trans isomers was visible under UV light, as the head of the spot was darker while its tail was fl uorescent blue. Moreover, column chroma-tography did not provide a good separation. The best results were achieved with gradient elution hexane:ethyl acetate. Even then, the separated fractions were the mix-tures of cis- and trans-isomers, as revealed by NMR anal ysis. However, when the mixture was extracted with n-hexane, only part of the material was dissolved and a white precipitate remained in the solution. The pre-cipitate was fi ltered and analyzed by TLC along with a hexane solution sample. The Rf for both spots were similar (approx. 0.7 in chloroform:hexane, 1:1), with the “hexane” fraction being slightly faster, but the color of the spots under UV light was different. The “hexane” spot was dark, while the “precipitate” spot was a bright fl uorescent blue. After 2–3 min in the open air, the dark spot also became fl uorescent blue.

The white precipitate was extracted two times more with hexane, and the combined hexane solution was evaporated to yield a transparent tar, slowly solidifying into fl ower-shaped crystals. Both products were addition-ally purifi ed by fl ash chromatography and analyzed by 1HNMR. According to the spectral data, the 4-[(Z)-2-(3,5- di-tert-butylphenyl)vinyl]phthalonitrile 4 and 4-[(E)-2-(3,5- di-tert-butylphenyl)vinyl]phthalonitrile 5 were success fullyseparated from each other. The 1H spectra of 4 shows two doublets of vinylene protons at 6.91 ppm and 6.48 ppm,

with a coupling constant of 12.2 Hz, which is typical for cis-confi guration of the double bond. For compound 5, the doublets of vinylene protons are shifted to the lower fi eld (7.33 ppm and 7.05 ppm, respectively), and their cou-pling constant is 16.4 Hz, indicating that the double bond is in trans confi guration. No cis-trans isomerization was observed for NMR samples, and the purity of the isomers was estimated as > 95%. The optical spectra are in good agreement with the NMR data. The absorption maximum for compound 4 is 333 nm, while for molecule 5, it is 343 nm, due to its extended π-conjugation. Cis-transisomerization occurs rapidly in solution upon UV irra-diation (excitation wavelength of 257 nm). After a few minutes, the absorption spectrum of either compound changes to a cross between two isomers, predominantly being trans. However, the isomerization did not occur when phthalonitrile 4 was heated at refl ux in 1-pentanol in a glass fl ask under ambient light. Even after three hours of boiling, no sign of trans-isomer was observed in the solution, thus the double bond confi guration should not be affected during further phthalocyanine synthesis.

Surprisingly, the yield of cis-compound 4 was twice that of trans-compound 5, despite the apparent sterical hindrance (34% and 16%, respectively). The chromato-graphic mobility of the isomers is also diffi cult to explain. Usually, the trans-compounds are faster on normal phase sorbents, but in the present case the cis-isomer is defi -nitely more mobile, although the difference is not huge.

Pentafl uorophenylvinylene phthalonitriles

The synthesis of pentafl uorophenyl-substituted phtha-lonitriles was carried out by the Wittig reaction in DMSO. The reaction was carried out at room temperature to avoid possible substitution of the para-fl uorine atom and the reaction time was prolonged up to 72 h to maximize the yield. Two main products were visible on TLC. This time the Rf values were signifi cantly different (0.86 and 0.72, respectively, in hexane:ethyl acetate, 1:1), which allowed the separation of the products by fl ash chromatography. The gradient elution with hexane:ethyl acetate yielded three fractions with good resolution between peaks. The fastest one was identifi ed as starting pentafl uorobenz-aldehyde. The 1H NMR spectra of the second product demonstrated two doublets of the double bond at 7.00 ppm and 6.74 ppm with coupling constant of 16.8 Hz, which corresponds to trans-compound 6. NMR spectra of the third product showed the signals of the double bond shifted to the higher fi eld (6.52 ppm and 6.20 ppm, respectively), and their coupling constant was 12.1 Hz, indicating the cis-confi guration of molecule 7.

The 19F NMR spectra of the two compounds gave additional proofs for cis-trans assignment. Measure-ments were done in ODCB-d4 using trifl uoromethylben-zene as internal standard. The reference peak was set to -63.73 ppm and the spectra were calibrated accordingly. Although no concentration dependence of chemical shifts


4 A. EFIMOV ET AL.

was observed, the samples taken for comparison were prepared with the same concentrations of the analytes. The spectra of both compounds show three peaks with relative integrals 2:1:2, corresponding to ortho-, para-,and meta-fl uorines, respectively [28, 29]. The ortho-sig-nals are located in the lowest fi eld, whereas meta-fl uorines give the signal in the highest fi eld. For trans-compound6, the signals are located at -142.35, -154.59, and -163.01 ppm (Table 1). For the cis-compound 7, the signals are shifted to lower fi eld and appear at -138.78, -154.34, and -162.06 ppm. Evidently, the ortho-fl uorines are the most affected by the neighboring double-bond confi guration, while the meta- and para-atoms are less infl uenced. The fl uorine chemical shifts are solvent dependent. Thus, in CDCl3 cis-pentafl uorophenylvinylenephthalonitrile 7 gives the signals at -138.85, -153.09, and -161.34 ppm, which are slightly downfi eld-shifted compared to those mea-sured in ODCB-d4. In any case, the difference between ortho-fl uorines of cis- and trans-isomers is signifi cant enough to serve as a good tool for structure evaluation.

The absorption maxima of 6 and 7 differ by 10 nm (297 nm and 286 nm, respectively). When illuminated with UV light in solution under ambient air, both com-pounds undergo fast decomposition without any sign of cis-trans isomerization. In fact, after 30 min of irradia-tion, no starting material was found in the solution. How-ever, the compounds have a good shelf-life in solid form and are very stable in solutions when protected from UV light.

Di-tert-butylphenylvinylene-substituted phthalocyanines

Di-tert-butylphenylvinylene phthalonitriles 4 and 5were converted into the corresponding phthalocyanines by both the “Li method” and the “DBU method”. Initial-ly, the phthalonitriles were heated at refl ux in 1-pentanol without the addition of a base in order to check that the starting compound does not isomerize under reaction con-ditions. As no change in absorption was observed after 2 h, we concluded that the isomer retains its confi guration;

therefore, the all-cis and all-trans compounds can be prepared.

After the lithium shots were dissolved in hot 1- pentanol, a portion of the phthalonitrile was added, and the solu-tion was heated at refl ux for some time; a 2 h reaction was suffi cient for the synthesis of cis-phthalocyanine 8.Prolonging the reaction time did not increase the yield of the target compound. The synthesis of trans-phthalocy-anine 9 required 5 h to complete. The reaction mixture was diluted with hexane after cooling, the lithium base was removed by washing with water, and the product was purifi ed by fl ash chromatography. To our great surprise, the yield of cis-phthalocyanine 8 was 26%, whereas for trans-phthalocyanine 9, the yield was only 11%. This result is somewhat in contradiction with our estimations, as compound 8 is more sterically hindered.

The same difference in reactivity was observed when the Zn(II) phthalocyanines 10 and 11 were synthesized by the “DBU method”. For cis-compound 10, the reac-tion time of 7 h was suffi cient, whereas the trans-product11 required 16 h of heating to maximize the yield. After solvent removal, the phthalocyanines were purifi ed by fl ash chromatography. The yield of cis-product 10 (12%) was twice as high as that of the trans-product 11 (7%), indicating that a phthalonitrile with Z-arylvinylene sub-stituent is more favorable for the formation of Zn(II) phthalocyanines. The difference in reactivity could be explained by their solubility. In fact, phthalocyanines 8 and 10 were much more soluble in common organic solvents. Regarding the chromatographic mobility, the cis-compounds were always faster than the trans ones, with about 30% difference in Rf and retention times.

Copper complexes can be synthesized by the “DBU method”, but the yield of the product depends on the salt used. 4-[(E)-2-(3,5-di-tert-butylphenyl)vinyl]phtha-lonitrile 5 does not tetramerize into the phthalocyanine when heated at refl ux with CuCl2 and DBU. On the other hand, the reaction with Cu(OAc)2 and 4-[(Z)-2-(3,5- di-tert-butylphenyl)vinyl]phthalonitrile 4 does work and produces the corresponding phthalocyanine 12 with 43% yield. The structure of compound 12 was proved by ESI-TOF MS.

Assignment of the cis-trans confi guration of the phthalocyanines was done by combination of NMR and UV-vis spectroscopy. NMR data were quite easy to inter-pret for the compound 8 (Fig. 2). The symmetric multip-let around 8.7 ppm is due to the neighbouring α,β-phthaloprotons. The multiplet around 7.8 ppm was attributed to the isolated α-phthalo-protons. The hydrogens of aryl rings give the signal at 7.3 ppm, and fi nally the vinylene protons give a signal at 7 ppm. This last multiplet con-sists of 3 pairs of doublets with a coupling constant of 12 Hz which, therefore, gives evidence for an all-cis con-fi guration of the vinylene bridges. COSY and decoupling experiments were done to prove the above-mentioned assignment. Thus, strong coupling was observed between the signals at 8.7 ppm and 7.8 ppm, whereas no coupling

Table 1. 19F NMR shifts

Compound o-F, ppm p-F, ppm m-F, ppm

6 -142.35 -154.59 -163.01

7 -138.78 -154.34 -162.06

13 -140.63 - -158.51

14 -140.78 - -158.63

15 -145.76 - -160.27

16 -145.95 - -160.14

17 -138.25 -156.51 -162.74

18 -142.83 -155.96 -163.48



with other groups was detected for the multiplets at 7.3 ppm and 7 ppm.

In contrast, the NMR spectra of phthalocyanines 9, 10,and 11 were poorly resolved in either deuterochloroform or deutero o-dichlorobenzene. No coupling constants can be measured reliably from those spectra. However, a few conclusions can still be drawn. Comparing the spec-tra of 8 and 9, we can assume the all-trans nature of the vinylene substituents in the latter. Its broad signal of the vinylene protons is shifted more to the low fi eld, as was observed for the corresponding phthalonitriles. Another evident fact is that zinc complexes are aggregated more strongly than the corresponding free bases. Comparing the spectra of 9 and 11, one can see a few separated broad signals in the aromatic region for the free base, trans-phthalocyanine, whereas for its Zn complex, the proton signals form one very broad peak.

The all-cis nature of phthalocyanines 8 and 10 along with the all-trans confi guration of the molecules 9, 11becomes evident from their absorption and emission spectra. The all-trans compound should have a larger π-conjugated system including arylvinylene peripheral groups; therefore, it should absorb and emit light at lon-ger wavelengths. The all-cis compound, in turn, should be suffi ciently non-planar, with a distorted conjugation between the central core and the side group. In fact, the absorption maxima of the cis-compounds are 7–10 nm blue- shifted compare to those of trans-molecules (Table 2, Fig. 3). The same effect is observed in steady-state fl uo-rescence spectra, where cis-compounds 8 and 10 emit at 5–10 nm shorter wavelengths. In contrast with the phtha-lonitriles, the phthalocyanines do not undergo cis-trans

isomerization upon UV irradiation. The chloroform solutions of 8 and 10 were illuminated with UV light (λmax 334 nm). No changes of absorption maxima posi-tion were observed with time, but a dramatic bleaching of

Fig. 2. 1H NMR spectra of phthalocyanine 8

Fig. 3. Absorption spectra of di-tert-butylphenyl-substituted phthalocyanines 8–12 in CHCl3

Table 2. Absorption and emission maxima of di-tert-butylphe-nyl-substituted phthalocyanines 8–12

Compound Absorption λmax, nm Emission λmax, nm

8 722, 686, 667, 626 811, 728

9 732, 698, 641, 636 822, 738

10 703, 633 713

11 710, 639 718

12 701, 632 712


6 A. EFIMOV ET AL.

the Q-bands region occurred after 30 min. The nature of this bleaching should be a photoreaction with oxygen in solutions, since the phthalocyanines 8–12 are stable in deoxygenated solutions, as well as in solid form, under UV irradiation or direct sunlight.

Pentafl uorophenylvinylene-substituted phthalocyanines

The pentafl uorophenyl group is always an interesting addition to a chromophore core, due to its strongly elec-tronegative, yet unpolar, character. The molecules with a pentafl uorophenyl fragment often demonstrate some unique photochemical and self-assembling properties [28–30]. As a drawback, the reactivity of this interesting synthon usually differs from the rest of similar aromat-ics, and in particular, side reactions at the para-fl uorine position are possible. We have mentioned earlier that, the Wittig synthesis with pentafl uorobenzaldehyde smoothly produced a pair of trans- and cis-pentafl uorophenylvinyle-nephthalonitriles 6 and 7. As the para-fl uorine remained intact in both cases, we proceeded with the synthesis of pentafl uorophenyl-substituted phthalocyanines.

The synthesis by the “Li method” was done in 1-butanol. Lithium shots were fi rst dissolved in a hot solvent, after which cis-phthalonitrile 7 was added. The solution rap-idly became green. The reaction was heated at refl ux for 4 h, and the green band of a main product 13 was sepa-rated by fl ash chromatography. Surprisingly, the Rf of the compound on TLC was very high (0.9 in CHCl3/hexane, 1:1). The solubility was also remarkably high. Despite this fact, the 1H NMR spectra in CDCl3 contained quite broad peaks of the phthalo-protons, showing evidence of aggregation. In addition to signals of aromatic protons, the spectra contained extra multiplets at 4.3, 2.0, and 1 ppm. The multiplets were much better resolved and were attributed to alkoxy-substitution at the molecule’s periphery, i.e. at the pentafl uorophenyl rings. Changing the solvent to deutero o-dichlorobenzene did not destroy the aggregates. The chemical shifts changed, but the sig-nals’ shape remained the same. The 19F NMR spectra (ODCB-d4, referenced to trifl uormethylbenzene) revealed only two fl uorine peaks at -140.63 and -158.51 ppm, thus supporting the idea that all four para-fl uorines were replaced with butoxy groups (Table 1). Unfortunately, we were unable to measure the ESI-TOF mass-spectra for this product. Neither of the solvents traditionally used for ESI infusion worked in that case; they simply did not give any signal of a reasonable mass. To obtain more proof of the structure, the compound was treated with zinc acetate in chloroform in the presence of DBU. After 20 h of stirring at room temperature, product 14was purifi ed by fl ash chromatography. To measure the ESI-TOF mass-spectra, the compound was dissolved in acetone and infused at 15 μl.min-1 rate. The compound does not give a reasonable peak in positive polarity, but in negative polarity, the adduct with chlorine ion [M + Cl]-

was observed. Comparing the experimental m/z value (1595.3140) with the calculated one (1595.3149), one can conclude that the structure was assigned correctly and that phthalocyanine 14 carries four 4-butoxy-2,3,5,6-tetrafl uorophenyl groups. The 1H NMR and 19F NMR spectra were measured in ODCB-d4. Only slight upfi eld shift was observed for the proton signals, compared to the free base molecule 13. The chemical shifts of the fl uorine atoms showed practically no difference between the free base and the zinc complex.

The conversion of the trans-phthalonitrile 6 by the “Li method” led to similar results. The reaction proceeded rapidly over 4 h. The main green product 15 was puri-fi ed by fl ash chromatography. Good solubility and high mobility were noticeable on TLC, although not as high as for compound 13. Ionization of the free base was not pos-sible with the ESI method, independently of the infusion solvent and the polarity used. Similarly to the procedure described above, phthalocyanine 15 was converted into zinc complex by treatment with zinc acetate in the pres-ence of DBU in chloroform at room temperature. Com-plex 16 gave a good signal in ESI-TOF mass-spectra. The infusion was undertaken in acetone, in negative polarity mode, and the intense peak corresponding to the adduct of 16 with chlorine ion [M + Cl]- was detected. The m/zvalue (1595.3237 measured, 1595.3149 calculated) and the isotopic pattern of the peak matched the proposed structure perfectly, thus proving that the substitution of the para-fl uorines by butoxy groups had occurred. The 1HNMR spectra were measured in CDCl3 and ODCB-d4. In either case, the peaks were very broad, particularly those in the aromatic region. However, the signals of butoxy groups were evident, although not as sharp as for the ciscompounds 13 and 14. The 19F NMR spectra were mea-sured in ODCB-d4 with trifl uoromethylbenzene as refer-ence (Table 1). For both free base 15 and zinc complex 16, only two peaks were visible. The chemical shifts were virtually independent of the macrocore central atom and were found around -145 ppm and -160 ppm. The peaks were much broader than those of cis-compounds. Con-sidering the NMR data, we can assume that the phtha-locyanines with trans-vinylene substituents are strongly aggregated, due to their planarity.

While the problem of basic structure of the com-pounds 13, 14, and 15, 16 was solved, the question about the vinylene bridges’ confi guration remained un-answered. The shape of the 1H NMR signals suggested that compounds 13 and 14 are in cis-confi guration as they are less aggregated, while the molecules 15 and 16 are in trans-confi guration. The absorption maxima of 15, 16were red-shifted by 10 nm, and their emission maxima were shifted even more, up to 20 nm (Fig. 4, Table 3). Such a red shift, along with the unusual shape of the Q-bands, was attributed to extensive aggregation and the trans-confi guration of the molecules, similarly to what was described for compounds 9 and 11. The direct evidence for all-cis and all-trans confi guration of the



phthalocyanines was gained from the 19F NMR. Compar-ing the spectra of 13 and 15, it was noticed that only two signals were present in each spectrum (-140.63, -158.51 ppm for 13; -145.76, -160.27 ppm for 15), and the dif-ference in their positions was as big as 5 ppm. We have mentioned before that the chemical shifts of the ortho-fl uorines differ signifi cantly for cis-phthalonitrile 7 and trans-phthalonitrile 6, being upfi eld-shifted for the latter. In addition, the signals of cis- and trans-molecules are not overlapping, thus the possible mixtures would always be visible. Therefore, we can assert that phthalocyanines 13 and 14 exist in all-cis confi guration, while the phtha-locyanines 15 and 16 are the all-trans molecules.

After the fi rst experiments, the question arose whether is it possible to retain the para-fl uorines during the macrocyclization. The conventional synthesis by the “DBU method” usually provides milder conditions,

therefore, we decided to try this method. The 4-[(E)-2-(pentafl uorophenyl)vinyl]phthalonitrile 6 was heated at refl ux in 1-butanol for 4 h with DBU and zinc acetate. After the usual work-up and chromatographic purifi cation, the reaction product was analyzed by NMR. The signals of alkoxy protons were clearly visible in 1H NMR spectra. In addition, the 19F NMR spectra demonstrated fi ve sig-nals at -142.17, -143.94, -154.50, -158.26, -162.54 ppm, with relative integrals 2:5:1:5:2. We assumed that the signals at -142.17, -154.50, and -162.54 ppm belong to an intact pentafl uorophenyl group, while the signals at -143.94 ppm and -158.26 ppm are due to the tetrafl uoro-monobutoxyphenyl fragment. It was clear that the para-fl uorines are partly replaced with butoxy groups, and the reaction product is a mixture of several differently substituted phthalocyanines. Partial substitution was con-fi rmed by MS data. When measured in negative polarity, the spectra demonstrated three peaks at 1379.0420 m/z,1433.1095 m/z, and 1487.1757 m/z. The fi rst one corre-sponds to [M + Cl]- ion, and the other two with 54 Daltons difference, corresponds to butoxy groups replacing one or two fl uorines. It should be noted that products with three or four butoxy groups were not observed. The “DBU method” with 4-[(Z)-2-(pentafl uorophenyl)vinyl]phtha-lonitrile 7 and zinc acetate gave similar results. The para-fl uorines were partly substituted during a 4 h reaction.

Recalling that the “DBU method” is sensitive to the salt used, we tried to replace zinc acetate with zinc chloride. The 4-[(Z)-2-(pentafl uorophenyl)vinyl]phthalonitrile 7,zinc chloride, and DBU were dissolved in 1-butanol and heated at refl ux. After 1 h, there was no sign of phthalocy-anine formation and the solution remained light- yellow. To promote the reaction, zinc acetate was added, and heating was continued. The solution turned green after a few minutes, and the reaction was heated at refl ux over-night to maximize the yield of the phthalocyanine. After the usual work-up and chromatographic purifi cation, the reaction product was analyzed by NMR and MS. The only suitable conditions for correct MS measurements were negative polarity and infusion of the analyte as acetone solution. To our great surprise, the ESI-TOF mass- spectra contained one main peak with m/z = 1379.0521, cor-responding to the [M + Cl]- ion of phthalocyanine 17with four pentafl uorophenyl rings untouched (calcu-lated m/z = 1379.0471). The isotopic pattern of the peak matched the molecular composition. The 19F NMR spectra were measured in ODCB-d4. Three peaks at -138.25 ppm, -156.51 ppm, and -162.74 ppm were observed (Table 1). Their relative intensities were 2:1:2, thus agreeing with MS data and giving more evidence that the para-fl uorines are not replaced by butoxy tails. The 1H NMR was mea-sured in CDCl3 and in ODCB-d4. The peaks were sur-prisingly well-resolved in both cases. The signals from α- and β-phthalo protons could be distinguished, along with vinylene protons (Fig. 5). Practically no sign of alkoxy-protons was found, which was in good agreement with MS and 19F-NMR data. Based on NMR and MS data,

Table 3. Absorption and emission maxima of fl uoro-substituted phthalocyanines 13–18

Compound Absorption λmax, nm Emission λmax, nm

13 716, 680, 660, 619 756, 720

14 694, 625 775, 703

15 727, 692, 662, 640 774, 732

16 724, 689 816, 732

17 695, 628 773, 702

18 707, 637 797, 715

Fig. 4. Absorption spectra of fl uoro-substituted phthalocyanines 13, 15 (top) and 14, 16–18 (bottom) in CHCl3


8 A. EFIMOV ET AL.

we concluded that the product is zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(pentafl uorophenyl))vinyl]phtha-locyanine 17. The yield of the compound after purifi ca-tion was 37%.

The same conditions were applied to prepare phtha-locyanine 18. The 4-[(E)-2-(pentafl uorophenyl)vinyl]phthalonitrile 6 was heated at refl ux with zinc chloride and DBU in 1-butanol. No phthalocyanine formation was visible after 1 h. Zinc acetate was added to the reaction mixture, and heating was continued for 16 h. As in the previous case, it was only possible to measure the ESI mass-spectra in negative polarity and by infusion in ace-tone solution. The main peak of the spectra with m/z = 1379.0526 corresponded to the [M + Cl]- ion. Another small peak with m/z = 1433.1219 was visible and it cor-responds to a mono-butoxy substituted phthalocyanine. The MS data were supported by 19F NMR spectra. The main three signals at -142.46 ppm, -155.07 ppm, and -163.19 ppm were accompanied by two small signals at -144.69 ppm and -158.91 ppm. The relative integrals ratio was 2:0.6:1:0.6:2, which corresponded roughly to a 3:1 proportion of the phthalocyanine 18 and its mono-butoxy derivative. The next synthesis was undertaken in the same way as above, but this time, SnCl2 was used instead of ZnCl2. Phthalonitrile 6 did not form a phthalocyanine when heated with tin chloride in the presence of DBU. However, when zinc acetate was added to the mixture, the phthalocyanine formation proceeded rapidly and was over

in a few hours. A drawback is that a large amount of insol-uble precipitate forms in the reaction; thus, the yield of the target phthalocyanine 18 was only 4%. Surprisingly, no tin phthalocyanine complex was found. One of the expla-nations that we can propose is that the tin complex is very poorly soluble in organic solvent, and precipitates during the reaction work-up. Nevertheless, the mass-spectra of the product contained only one peak at m/z = 1379.0358. The 19F NMR spectra demonstrated just three signals at -142.83 ppm, -155.96 ppm, and -163.48 ppm (Table 1), and no alkoxy protons were observed in 1H NMR (Fig. 5). The question of why chloride salts suppress the substitu-tion of para-fl uorines with alkoxy groups remains open and will be the subject of separate studies.

The assignment of the vinylene bridges’ confi gura-tion was based on 19F NMR data. Comparing the chemi-cal shifts of phthalocyanines 17, 18 and corresponding phthalonitriles 7, 6, one can see that the main difference between cis- and trans-compounds appears in the signals of ortho-fl uorines, while meta- and para-fl uorines give the signals at approximately the same values. The signals of ortho-fl uorines of cis-phthalonitrile 7 and its derivative 17are located at -138.78 ppm and -138.25 ppm, respectively. The signals of ortho-fl uorines of trans-phthalonitrile 6and its derivative 18 are located at -142.35 and -142.83, respectively. The signals of the cis- and trans-compoundsare not overlapping; thus since only one peak of ortho-fl uorines is visible for phthalocyanines 17 and 18, we can

Fig. 5. 1H NMR spectra of fl uoro-substituted phthalocyanines 13–18



assert that molecule 17 is in all-cis confi guration, and molecule 18 is in all-trans confi guration of the vinylene bridges. This conclusion is supported by optical spec-troscopy data (Table 3, Fig. 4). The Q-bands absorption maxima are located at 695 nm and 707 nm for phthalo-cyanines 17 and 18, respectively. The absorption maxima do not change when the degassed solution is illuminated with UV light, thus no cis-trans isomerization occurs for phthalocyanines 17 and 18. The emission maximum of compound 17 is 773 nm, while compound 18 emits at 797 nm, due to a larger conjugation of the all-trans isomer.

EXPERIMENTAL

The solvents of HPLC grade were purchased from VWR International. The reagents were purchased from Sigma-Aldrich and TCI Europe. The chemicals were used without further purifi cation unless otherwise mentioned.

The 1H and 19F NMR spectra were run on a 300 MHz Varian Mercury spectrometer. The mass-spectra were measured with Waters LCT Premier XE ESI-TOF mass spectrometer. The sample of the analyte was dissolved in an appropriate solvent at concentration ca 0.01 mg.mL-1

and infused at a rate of 15 μL.min-1. The reference solu-tion of Leucine Enkephaline (50 pg.mL-1) was infused simultaneously. The experimental spectra were centered and TOF lock-mass correction was applied.

(3,4-dicyanobenzyl)triphenylphosphonium bro-mide (3). A suspension of 4-methylphthalonitrile (1) (1.4 g, 10 mmol), N-bromosuccinimide (2.07 g, 12 mmol), and benzoyl peroxide (19 mg, 0.08 mmol) in carbon tet-rachloride (12 mL) was heated at refl ux for 16 h. The orange reaction mixture was cooled down, and the pre-cipitate was fi ltered off. The clean organic solution was evaporated to dryness to yield 2.4 g of 4-bromometh-ylphthalonitrile (2) (2:1 mixture with 4-methylphthalo-nitrile). The solid residue without further purifi cation was dissolved in DMF (30 mL), after which triphenylphos-phine (1.9 mg, 7.26 mmol) was added, and the reaction mixture was stirred for 60 h at 90 °C under argon fl ow. The solvent was removed under reduced pressure, and a tar-like residue was extracted with diethyl ether (30 mL). After 30 min, the solid residue formed was fi ltered, washed with diethyl ether (2 × 50 mL), and dried in air to yield 2.15 g (67%) of (3,4-dicyanobenzyl)triphenylphospho-nium bromide (3). 1H NMR (300 MHz; CDCl3; Me4Si):δH, ppm 8.03–7.49 (18H, m, ArH), 6.07 (2H, d, J = 15.4 Hz, CH2). MS (ESI-TOF): m/z ES+ 403.1413 [M - Br]+.Calcd. for C27H20N2P 403.1364.; ES- 483.0221 [M-H]-,560.9366 [M + Br]-. Calcd. for C27H19N2PBr 483.0453, C27H20N2PBr2 560.9731.

4-[(Z)-2-(3,5-di-tert-butylphenyl)vinyl]phthalo-nitrile (4) and 4-[(E)-2-(3,5-di-tert-butylphenyl)vinyl] phthalonitrile (5). (3,4-dicyanobenzyl)triphenylphos-phonium bromide (3) (980 mg, 2.05 mmol) and 3,5- di-tert-butylbenzaldehyde (440 mg, 2 mmol) were dissolved in DMSO (30 mL) in a screw-cap vial. The vial was

warmed up to 60 °C and fl ushed with argon, after which NaOMe (136 mg, 2.5 mmol) was added in one portion. The vial was sealed and the reaction mixture was stirred at 80 °C for 14 h. The hot solution was poured into 150 mL of water. The suspension formed was extracted with diethyl ether (3 × 70 mL). The combined ether fraction was washed with water (3 × 100 mL) and evaporated to yield a glassy orange solid. The solid was added to 15 mL of hexane and stirred in an ultrasound bath for 5 min. The organic solution was separated, and the solid residue was washed in the same way with hexane, twice more. The solid residue was purifi ed by fl ash chromatography on Silica 60 (eluent CHCl3) to yield 104 mg (16%) of 4-[(E)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (5)as a white powder. UV-vis (1-pentanol): λmax, nm (ε, M-1.cm-1) 254 (24113), 343 (34502). 1H NMR (300 MHz; CDCl3; Me4Si): δH, ppm 7.92 (1H, d, J = 1.7 Hz, 3-phtha-loH), 7.82 (1H, dd, J1 = 8.2 Hz, J2 = 1.6 Hz, 5-phthaloH),7.76 (1H, d, J = 8.2 Hz, 6-phthaloH), 7.45 (1H, t, J =1.7 Hz, 4-ArH), 7.38 (2H, d, J = 1.7 Hz, 2,6-ArH), 7.33 (1H, d, J = 16.4 Hz, CHCH), 7.05 (1H, d, J = 16.4 Hz, CHCH), 1.37 (18H, s, C(CH3)3). MS (ESI-TOF): m/z342.1753 [M]+. Calcd. for C24H26N2 342.2096. All three hexane extracts were combined, the solvent was removed under reduced pressure, and the residue was purifi ed by fl ash chromatography (Silica 60, eluent CHCl3) to yield 235 mg (34%) of 4-[(Z)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (4) as slowly solidifying white crystals. UV-vis (1-pentanol): λmax, nm (ε, M-1.cm-1) 244 (38017), 333 (26830). 1H NMR (300 MHz; CDCl3; Me4Si): δH,ppm 7.68 (1H, d, J = 1.5 Hz, 3-phthaloH), 7.63 (1H, d, J = 8.2 Hz, 6-phthaloH), 7.58 (1H, dd, J1 = 8.2 Hz, J2 =1.5 Hz, 5-phthaloH), 7.34 (1H, t, J = 1.8 Hz, 4-ArH), 7.01 (2H, d, J = 1.8 Hz, 2,6-ArH), 6.91 (1H, d, J = 12.2 Hz, CHCH), 6.48 (1H, d, J = 12.2 Hz, CHCH), 1.24 (18H, s, C(CH3)3). MS (ESI-TOF): m/z 342.1694 [M]+. Calcd. for C24H26N2 342.2096.

4-[(E)-2-(pentafluorophenyl)vinyl]phthalonitrile (6) and 4-[(Z)-2-(pentafl uorophenyl)vinyl]phthalo-nitrile (7). (3,4-dicyanobenzyl)triphenylphosphonium bromide (3) (987 mg, 2.04 mmol) and pentafl uorobenz-aldehyde (400 mg, 2.04 mmol) were dissolved in DMSO (30 mL) in a screw-cap vial. The vial was fl ushed with argon and NaOMe (134 mg, 2.5 mmol) was added in one portion. The vial was sealed and the reaction mixture was stirred at rt for 72 h. The solution was poured into brine (150 mL) and extracted with diethyl ether (150 mL). The ether phase was separated, washed with water (3 × 100 mL), and evaporated under reduced pressure to yield a tar-like residue. The tar was dissolved in CCl4 (20 mL), fi ltered, purifi ed by fl ash chromatography on CombiFlash (40 g Silica 60 column, elution with hexane:ethyl acetate 9:1 for 1 min, then gradually changed up to 7:3 in 9 min (elution speed 40 mL.min-1) to yield three fractions: frac-tion 1 tR = 4 min, fraction 2 tR = 6.5 min, and fraction 3 tR = 9 min. Fraction 1 was identifi ed by TLC as unreacted pentafl uorobenzaldehyde. Fraction 2 was evaporated


10 A. EFIMOV ET AL.

to yield 160 mg (25%) of 4-[(E)-2-(pentafl uorophenyl)vinyl]phthalonitrile (6). UV-vis (CHCl3): λmax, nm 314. 1H NMR (300 MHz; ODCB-d4): δH, ppm 7.34 (1H, d, J = 1.5 Hz, 3-phthaloH), 7.29 (1H, dd, J1 = 8.2 Hz, J2 =1.6 Hz, 5-phthaloH), 7.23 (1H, d, J = 8.2 Hz, 6-phthaloH),7.00 (1H, d, J = 16.8 Hz, CHCH), 6.74 (1H, d, J = 16.8 Hz, CHCH). 19F NMR (300 MHz; ODCB-d4; trifl uorom-ethylbenzene): δH, ppm -142.35 (2F, m, o-ArF), -154.59 (1F, m, p-ArF), -163.01 (2F, m, m-ArF). MS (ESI-TOF): m/z 343.0103 [M + Na]+, 663.0371 [2M + Na]+, (calcd. for C16H5N2F5Na 343.0271, C32H10N4F10Na 663.0643). Fraction 3 was evaporated to yield 158 mg (25%) of 4-[(Z)-2-(pentafl uorophenyl)vinyl]phthalonitrile (7). UV-vis (CHCl3): λmax, nm 288. 1H NMR (300 MHz; ODCB-d4): δH, ppm 7.14 (1H, d, J = 8.2 Hz, 6-phthaloH), 7.12 (1H, d, J = 1.5Hz, 3-phthaloH), 7.02 (1H, dd, J1 = 8.2 Hz, J2 = 1.6 Hz, 5-phthaloH), 6.52 (1H, d, J = 12.1 Hz, CHCH), 6.20 (1H, dd, J1 = 12.1 Hz, J2 = 1.2 Hz CHCH).19F NMR (300 MHz; ODCB-d4; trifl uoromethylbenzene): δH, ppm -138.78 (2F, m, o-ArF), -154.34 (1F, m, p-ArF),-162.06 (2F, m, m-ArF). MS (ESI-TOF): m/z 343.0097 [M + Na]+, 663.0339 [2M + Na]+. Calcd. for C16H5N2F5Na343.0271, C32H10N4F10Na 663.0643.

2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (8). The freshly cut shots of lithium (25 mg, 3.6 mmol) were dissolved in 1-pentanol (15 mL) by heating at refl ux under argon fl ow. 4-[(Z)-2-(3,5-di-tert-butylphenyl)vinyl] phthalonitrile (4)(58 mg, 0.17 mmol) was added in one portion and the reaction mixture was heated at refl ux under argon fl ow for 2 h. The dark green solution was cooled down, poured into water (90 mL), and diluted with hexane (15 mL). The organic phase was separated and washed with water (3 mL × 100 mL). The dark-green organic solution was evaporated to dryness under reduced pressure. The residue was dissolved in hexane and subjected to fl ash chroma-tography (Silica 100, eluent chloroform:hexane 1:1). The main green fraction was collected and evaporated to yield 15 mg (26%) of the target 2,6(7),10(11),14(15) tetrakis[4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (8). UV-vis (CHCl3): λmax, nm (ε, M-1.cm-1) 349 (78788), 626 (41655), 667 (63722), 686 (142960), 722 (165780). 1H NMR (300 MHz; CDCl3): δH, ppm 8.86–8.47 (8H, m, phthaloH), 8.00–7.79 (4H, m, phthaloH), 7.44–7.28 (12H, m, ArH), 7.16–6.89 (8H, m, CHCH), 1.26–1.12 (72H, m, C(CH3)3), -2.05- -2.74 (2H, br, NH). MS (ESI-TOF): m/z1370.6248 [M]+. Calcd. for C96H106N8 1370.8540.

2,6(7),10(11),14(15)-Tetrakis [4-((E)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (9). The freshly cut shots of lithium (25 mg, 3.6 mmol) were dissolved in 1-pentanol (17 mL) by heating at refl ux under argon fl ow. 4-[(E)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (5)(100 mg, 0.29 mmol) was added in one portion, after which the reaction mixture was heated at refl ux under argon fl ow for 5 h. The dark green solution was cooled down and washed with water (3 mL × 15 mL). The organic phase was collected, diluted with toluene (20 mL), and

the solvent was evaporated to dryness under reduced pressure. The dry green residue was dissolved in hexane and subjected to fl ash chromatography (Silica 100, eluent chloroform:hexane, 2:1). The main green fraction was collected and evaporated to yield 11 mg (11%) of the tar-get 2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (9). UV-vis (CHCl3):λmax, nm (ε, M-1.cm-1) 342 (106970), 636 (41740), 641 (63080), 698 (154800), 732 (168100). 1H NMR (300 MHz; CDCl3/CD3OD, 20/1): δH, ppm 8.66–6.52 (32H, br m, phthaloH, ArH), 1.69–0.77 (72H, br m, C(CH3)3), NHnot resolved. MS (ESI-TOF): m/z 1370.9213 [M]+. Calcd. for C96H106N8 1370.8540.

Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (10). 4-[(Z)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (4) (100 mg, 0.29 mmol), DBU (0.056 mL, 0.4 mmol), and Zn(OAc)2·2H2O (50 mg, 0.23 mmol) were heated at refl ux in 1-pentanol (20 mL) for 7 h. The dark green solution was cooled down and evaporated to dryness under reduced pressure. The solid residue was dissolved in hexane and subjected to a fl ash chromatography (Silica 60, eluent chloroform). The main green fraction was collected and evaporated to yield 13 mg (12%) of the target zinc(II) 2,6(7),10(11),14(15) tetrakis [4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (10).UV-vis (CHCl3): λmax, nm (ε, M-1.cm-1) 300 (111430), 354 (103502), 633 (34691), 703 (181800). 1H NMR (300 MHz; CDCl3): δH, ppm 9.34–6.25 (32H, br m, phthaloH,ArH, CHCH), 1.47–0.68 (72H, m, C(CH3)3). MS (ESI-TOF): m/z 1433.7767 [M + H]+. Calcd. for C96H105N8Zn1433.7754.

Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (11). 4-[(E)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (5) (50 mg, 0.15 mmol), DBU (0.028 ml, 0.2 mmol), and Zn(OAc)2·2H2O (27 mg, 0.12 mmol) were heated at refl ux in 1-pentanol (20 mL) for 16 h. The green solution was cooled down and evaporated to dryness under reduced pressure. The solid residue was dissolved in dichlo-romethane and subjected to fl ash chromatography (Silica 60, eluent dichloromethane, then dichloromethane:ethyl acetate 2:3). The main green fraction was collected and evaporated to yield 3.9 mg (7%) of the target zinc(II) 2,6(7),10(11),14(15) tetrakis [4-((E)-2-(3,5-di-tert-bu-tylphenyl))vinyl]phthalocyanine (11). UV-vis (CHCl3):λmax, nm (ε, M-1.cm-1) 314 (107754), 639 (32619), 710 (179770). 1H NMR (300 MHz; CDCl3): δH, ppm 9.02–6.12 (32H, br, phthaloH, ArH, CHCH), 1.87–1.19 (72H, m, C(CH3)3). MS (ESI-TOF): m/z 1433.7821 [M + H]+.Calcd. for C96H105N8Zn 1433.7754.

Copper(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (12). 4-[(Z)-2-(3,5-di-tert-butylphenyl)vinyl]phthalonitrile (4) (50 mg, 0.156 mmol), DBU (0.022 mL, 0.156 mmol), and CuCl2·2H2O (8.8 mg, 0.05 mmol) were heated at refl ux in 1-pentanol (10 mL) for 18 h under drying tube.



The dark green solution was cooled down and evaporat-ed to dryness under reduced pressure. The solid residue was dissolved in chloroform and subjected to fl ash chromatography (Silica 100, eluent chloroform:hexane, 1:1). The main green fraction was collected and evap-orated to yield 24 mg (43%) of the target copper(II) 2,6(7),10(11),14(15)- tetrakis [4-((Z)-2-(3,5-di-tert-butylphenyl))vinyl]phthalocyanine (12). UV-vis (CHCl3):λmax, nm (ε, M-1.cm-1) 347 (39700), 632 (20600), 701 (178400). MS (ESI-TOF): m/z 1431.7856 [M]+. Calcd. for C96H104N8Cu 1431.7679.

2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(4-butoxy- 2,3,5,6-tetrafl uorophenyl))vinyl]phthalocyanine (13). The freshly cut shots of lithium (27 mg, 3.9 mmol) were dissolved in 1-butanol (15 ml) by heating at refl ux under argon fl ow. 4-[(Z)-2-(pentafl uorophenyl)vinyl]phthalo-nitrile 7 (49 mg, 0.15 mmol) was added in one portion, after which the reaction mixture was heated at refl ux under argon fl ow for 4 h. The dark green solution was cooled down and evaporated to dryness under reduced pressure. The dry residue was dissolved in acetone (3 mL) and subjected to a fl ash chromatography (Silica 60, eluent toluene). The main green fraction was collected and evap-orated to yield 12 mg (21%) of the 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(4-butoxy-2,3,5,6-tetrafl uorophenyl))vinyl]phthalocyanine (13). UV-vis (CHCl3): λmax, nm (ε,M-1.cm-1) 298 (40100), 357 (50020), 619 (21515), 660 (36760), 680 (110000), 716 (123800). 1H NMR (300 MHz; CDCl3): δH, ppm 8.43–7.76 (8H, br, phthaloH),7.72–7.36 (4H, br, phthaloH), 6.84–6.40 (4H, br, phtha-loH), 4.49–4.02 (8H, m, OCH2C3H7), 2.01–1.70 (8H, m, OCH2CH2C2H5), 1.09–0.80 (12H, m, OC3H7CH3), NHnot resolved. 19F NMR (300 MHz; ODCB-d4; trifl uorom-ethylbenzene): δH, ppm -140.63 (2F, m, o-ArF), -158.51 (2F, m, m-ArF).

Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(4-butoxy-2,3,5,6-tetrafl uorophenyl))vinyl]phthalocya-nine (14). 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(4-butoxy-2,3,5,6-tetrafl uorophenyl))vinyl]phthalocyanine (13) (6 mg, 0.0047 mmol) and Zn(OAc)2·2H2O (2 mg, 0.009 mmol) were dissolved in chloroform (3 mL). DBU (0.003 mL, 0.02 mmol) was added, and the reaction mix-ture was stirred for 20 h. The solvent was removed under reduced pressure, the dry residue was dissolved in chloro-form (2 ml) and purifi ed by fl ash chromatography (Silica 60, gradient elution from chloroform to chloroform:ethyl acetate 4:1 in 8 min, fl ow rate 15 mL.min-1, tR = 2 min) to yield quantitatively zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-(4-butoxy-2,3,5,6-tetrafluorophenyl))vinyl]phthalocyanine (14). 1H NMR (300 MHz; ODCB-d4):δH, ppm 9.12–8.51 (6H, br, phthaloH), 8.05–7.69 (4H, br, phthaloH), 7.62–7.36 (4H, br, phthaloH, CHCH), 6.67–6.56 (2H, br, CHCH), 4.38–3.91 (8H, m, OCH2C3H7),1.77–1.48 (8H, m, OCH2CH2C2H5), 1.46–1.27 (8H, m, OC2H5CH2CH3), 0.94–0.55 (12H, m, OC3H7CH3).

19FNMR (300 MHz; ODCB-d4; trifl uoromethylbenzene): δH, ppm -140.78 (2F, s, o-ArF), -158.63 (2F, s, m-ArF).

MS (ESI-TOF): m/z ES-1595.3140 [M + Cl]-. Calcd. for C80H56N8O4ZnCl 1595.3149.

2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(4-butoxy- 2,3,5,6-tetrafl uorophenyl))vinyl]phthalocyanine (15). The freshly cut shots of lithium (32 mg, 4.6 mmol) were dissolved in 1-butanol (15 mL) by heating at refl ux under argon fl ow. 4-[(E)-2-(pentafl uorophenyl)vinyl]phthalo-nitrile 6 (52 mg, 0.16 mmol) was added in one portion, after which the reaction mixture was heated at refl ux under argon fl ow for 3 h. The dark green solution was cooled down and evaporated to dryness under reduced pressure. The dry residue was dissolved in chloroform (15 mL) and the solution was fi ltered and subjected to fl ash chromatography (Silica 100, eluent chloroform). The main green fraction was collected and evaporated to yield 14 mg (23%) of the 2,6(7),10(11),14(15)- tetrakis [4-((E)-2-(4-butoxy-2,3,5,6-tetrafluorophenyl))vinyl]phthalocyanine (15). UV-vis (CHCl3): λmax, nm (ε, M-1.cm-1) 335 (66300), 640 (32400), 662 (36300), 692 (48800), 727 (45000). 1H NMR (300 MHz; CDCl3): δH,ppm 7.68–5.07 (12H, br, phthaloH), 4.46–3.32 (8H, br m, OCH2C3H7), 1.90–1.33 (16H, br m, OCH2CH2CH2CH3),1.17–0.44 (12H, br, OC3H7CH3), NH not resolved. 19FNMR (300 MHz; ODCB-d4; trifl uoromethylbenzene): δH,ppm -145.76 (2F, br, o-ArF), -160.27 (2F, br, m-ArF).

Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((E)-2- (4-butoxy-2,3,5,6-tetrafluorophenyl))vinyl]phtha-locyanine (16). 2,6(7),10(11),14(15)-tetrakis [4-((E)- 2-(4-butoxy-2,3,5,6-tetrafl uorophenyl))vinyl] phthalocy-anine (15) (6.4 mg, 0.005 mmol) and Zn(OAc)2·2H2O(2 mg, 0.009 mmol) were dissolved in chloroform (3 mL). DBU (0.003 mL, 0.02 mmol) was added and the reaction mixture was stirred for 20 h. The solvent was removed under reduced pressure, after which the dry residue was dissolved in chloroform (2 mL) and purifi ed by fl ash chromatography (Silica 60, gradient elution from chloro-form to chloroform:ethyl acetate 4:1 in 8 min, fl ow rate 15 mL.min-1, tR = 2.5 min) to yield quantitatively zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(4-butoxy-2,3,5,6-tetrafl uorophenyl))vinyl]phthalocyanine (16). 1H NMR (300 MHz; ODCB-d4): δH, ppm 9.0–5.3 (br, phthaloH,CHCH), 4.4–3.4 (br, OCH2C3H7), 1.9–0.3 (br m, OCH2

C3H7).19F NMR (300 MHz; ODCB-d4; trifl uoromethyl-

benzene): δH, ppm -145.95 (2F, br, o-ArF), -160.14 (2F, m, m-ArF). MS (ESI-TOF): m/z ES- 1595.3237 [M + Cl]-

(calcd. for C80H56N8O4ZnCl 1595.3149).Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((Z)-2-

(pentafl uorophenyl))vinyl]phthalocyanine (17). 4-[(Z)-2-(Pentafl uorophenyl)vinyl]phthalonitrile (7) (60 mg, 0.18 mmol) and ZnCl2 (14 mg, 0.1 mmol) were dissolved in 1-butanol (15 mL). DBU (0.025 ml, 0.18 mmol) was added in one portion and the reaction mixture was heated at refl ux under a drying tube for 30 min. At this point, heating was removed, the light green solution was cooled down, and a portion of Zn(OAc)2 (17.7 mg, 0.08 mmol) was added to the reaction mixture. The solution was further heated at refl ux for 20 h. The solvent was removed under


12 A. EFIMOV ET AL.

reduced pressure, the dry residue was dissolved in dichlo-romethane (5 mL), and subjected to fl ash chromatography (Silica 60, eluent dichloromethane, then chloroform). The main green-blue fraction was collected and evaporated to yield 23 mg (37%) of the 2,6(7),10(11),14(15)-tetrakis[4-((Z)-2-(pentafluorophenyl))vinyl]phthalocyanine (17). UV-vis (CHCl3): λmax, nm (ε, M-1.cm-1) 363 (55507), 628 (22403), 695 (143450). 1H NMR (300 MHz; ODCB-d4): δH, ppm 9.07–8.86 (4H, m, phthaloH),7.76–7.62 (2H, m, phthaloH), 7.50–7.36 (2H, m, phtha-loH), 6.78–6.67 (4H, m, phthaloH), 6.53–6.24 (6H, m, phthaloH, CHCH), 5.87–5.03 (4H, br, CHCH). 19F NMR (300 MHz; ODCB-d4; trifl uoromethylbenzene): δH, ppm -138.25 (2F, m, o-ArF), -156.51 (1F, m, p-ArF), -162.74 (2F, m, m-ArF). MS (ESI-TOF): m/z ES-1379.0521 [M + Cl]-. Calcd. for C64H20N8F20ZnCl 1379.0471.

Zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(pentafl uorophenyl))vinyl]phthalocyanine (18). 4-[(E)-2-(pentafl uorophenyl)vinyl]phthalonitrile (6) (49 mg, 0.15 mmol) , DBU (0.021 ml, 0.15 mmol), and SnCl2·2H2O(14 mg, 0.08 mmol) were dissolved in 1-butanol (15 mL), a portion of Zn(OAc)2 (17.5 mg, 0.08 mmol) was added , and the solution was further heated at refl ux for 18 h. The green solution was cooled down, after which the solvent was removed under reduced pressure. The dry residue was dissolved in chloroform (5 mL) and subjected to fl ash chromatography (Silica 100, eluent chloroform). The main green-blue fraction was collected and evaporated to yield 4 mg (8%) of zinc(II) 2,6(7),10(11),14(15)-tetrakis [4-((E)-2-(pentafl uorophenyl))vinyl]phthalocyanine (18).UV-vis (CHCl3): λmax, nm (ε, M-1.cm-1) 312 (98995), 637 (29166), 707 (120980). 1H NMR (300 MHz; ODCB-d4): δH, ppm 7.61–6.37 (m, phthaloH, CHCH). 19F NMR (300 MHz; ODCB-d4; trifl uoromethylbenzene): δH, ppm -142.83 (2F, m, o-ArF), -155.96 (1F, m, p-ArF), -163.48 (2F, m, m-ArF). MS (ESI-TOF): m/z ES-1379.0358 [M + Cl]-. Calcd. for C64H20N8F20ZnCl 1379.0471.

CONCLUSION

We have synthesized a series of tetrakis(arylvinylene)phthalocyanines from the corresponding phthalonitriles. According to 1H and 19F NMR data, the cis- or trans- conformations of the starting materials are retained during condensation, thus the phthalocyanines formed are in all-cis or all-trans form. The cis-trans photoi-somerization occurs easily for phthalonitriles, while phthalo cyanines retain their conformations under a UV beam. The extended π-conjugation is more pronounced for all-trans phthalocyanines, while for all-cis phthalocy-anines, the arylvinylene substituents are not conjugated with the macrocore. However, the all-cis molecules might be particularly interesting due to their improved solubility and unusual shielding of the macrocycle plane by arylvi-nylene groups. 19F NMR is an especially valuable tool to monitor the reactions and conformations of the com-pounds with a pentafl uorophenyl fragment, which are

diffi cult to analyze by conventional methods. The alkoxy substitution of the para-fl uorines in the pentafl uorphenyl ring can be suppressed by the addition of zinc chloride or tin chloride to the reaction mixture.

Acknowledgements

This work is supported by the Academy of Finland and TEKES (Finnish Funding Agency for Technology and Innovation).

Supporting information

NMR spectra of the phthalonitriles 4–7 and phthalo-cyanines 8–11 and 13–18 are given in the supplementary material. This material is available at http://www.world scinet.com/jpp/jpp.shtml

REFERENCES

Gregory BW, Vaknin D, Gray JD, Ocko BM, Stoeve 1.P, Cotton TM and Struve WS. J. Phys. Chem. B1997; 101: 2006–2019.Yoshida Y, Nakamura M, Tanaka S, Hiromitsu I, 2.Fujita Y and Katsumio Y. Synth. Met. 2006; 156:1213–1217.De la Escosura A, Martinez-Diaz MV, Guldi DM 3.and Torres T. J. Am. Chem. Soc. 2006; 128: 4112–4118.Imahori H and Sakata Y. 4. Adv. Mater. 1997; 9: 537–546.Peumans P and Forrest SR. 5. Appl. Phys. Lett. 2001; 79: 126.Loi MA, Denk P, Hoppe H, Neugebauer H, Winder 6.C, Meissner D, Brabec C, Sariciftci NS, Gouloumis A, Vazquez P and Torres T. J. Mater. Chem. 2003; 13: 700–704.Barker CA, Zeng X, Bettington S, Batsanov AS, 7.Bryce MR and Beeby A. Chem. – Eur. J. 2007; 13:6710–6717.Gacho EH, Imai H, Tsunashima R, Naito T, Inabe 8.T and Kobayashi N. Inorg. Chem. 2006; 45: 4170–4176.Calvete MJF, Dini D, Flom SR, Hanack M, Pong 9.RGS and Shirk JS. Eur. J. Org. Chem. 2005; 16:3499–3509.Jin S, Cheng G, Chen GZ and Ji Z. 10. J. Porphyrins Phthalocyanines 2005; 9: 32–39.Maya EM, Snow AW, Shirk JS, Flom SR, Pong 11.RGS and Callahan JH. J. Porphyrins Phthalocya-nines 2002; 6: 463–475.De la Escosura A, Claessens CG, Ledoux-Rak I, 12.Zyss J, Martinez-Diaz MV and Torres T. J. Porphy-rins Phthalocyanines 2005; 9: 788–793.Vagin SI, Frickenschmidt A, Kammerer B and 13.Hanack M. Chem. – Eur. J. 2005; 11: 6568–6573.De la Escosura A, Martinez-Diaz MV, Barbera J and 14.Torres T. J. Org. Chem. 2008; 73: 1475–1480.



Cammidge AN and Gopee H.15. Chem. – Eur. J. 2006; 12: 8609–8613.Gouloumis A, Gonzalez-Rodriguez D, Vazquez 16.P, Torres T, Liu S, Echegoyen L, Ramey J, Hug GL and Guldi DM. J. Am. Chem. Soc. 2006; 128:12674–12684.Lee C-H, Sooksimuang T and Mandal BK. 17. J. Por-phyrins Phthalocyanines 2006; 10: 135–139.Donyagina VF and Luk”yanets EA. 18. Russ. J. Gen. Chem. 2005; 75: 795–799.De la Escosura A, Martinez-Diaz MV, Thordarson P, 19.Rowan AE, Nolte RJM and Torres T. J. Am. Chem. Soc. 2003; 125: 12300–12308.Vagin S and Hanack M. 20. Eur. J. Org. Chem. 2003; 14: 2661–2669.Kimura M, Sakaguchi A, Ohta K, Hanabusa K, Shirai H 21.and Kobayashi N. Inorg. Chem. 2003; 42: 2821–2823.Sooksimuang T and Mandal BK. 22. J. Org. Chem.2003; 68: 652–655.Mandal BK and Sooksimuang T. 23. J. Porphyrins Phthalocyanines 2002; 6: 66–72.

Gouloumis A, Vazquez P, Torres T, Liu S-G and 24.Echegoyen L. Chem. Commun. (Cambridge, UK)2001; 5: 399–400.Yamazaki Y. Jpn. Kokai Tokyo Koho 1997; 15 pp. 25.CODEN: JKXXAF JP 09031350 A 19970204 Heisei.Kaunisto K, Vuorinen T, Vahasalo H, Chukharev V, 26.Tkachenko NV, Efi mov A, Tolkki A, Lehtivuori H and Lemmetyinen H. J. Phys. Chem. C 2008; 112:10256–10265.Sereno L, Silber JJ, Otero L, del Valle Bohorquez 27.M, Moore AL, Moore TA and Gust D. J. Phys. Chem. 1996; 100: 814–821.Yatsunyk L and Walker FA. 28. Inorg. Chim. Acta 2002; 337: 266–274.Renak ML, Bartholomew GP, Wang S, Ricatto P, 29.Lachicotte RJ and Bazan GC. J. Am. Chem. Soc.1999; 121: 7787–7799.Efi mov AV, Anikin M, Tkachenko NV, Mironov AF 30.and Lemmetyinen H. Chem. Phys. Lett. 1998; 289:572–578.

Copyright of the works in this Journal is vested with World Scientific Publishing. Thearticle is allowed for individual use only and may not be copied, further disseminated, orhosted on any other third party website or repository without the copyright holder’swritten permission.



INTRODUCTION

Photoinduced energy transfer and electron transfer, the key steps in photosynthesis, have attracted signifi -cant attention from scientists who endeavor to develop artifi cial photosynthetic systems [1–6]. The importance and complexity of energy transfer and electron transfer processes in photosynthesis have prompted design and preparation of a number of covalently-linked, multi- porphyrin systems to mimic photosynthetic energy trans-fer and electron transfer [7–12]. The synthetic diffi culty

has precluded rapid development of such multi-porphyrin systems as artifi cial photosynthetic and molecular electro-nic materials [13–15]. In contrast to the covalently-linked, multi-porphyrin systems, self-assembly of porphyrins with non-covalent bonding provides a much simpler way to construct multi-porphyrin ensembles which can act as effi cient light-harvesting systems [16–22]. On the other hand, phthalocyanines (Pcs) have also received signifi -cant attention, in view of their magnifi cent absorption in the 600–700 nm region (leading to a blue-coloring ability), their robust nature, and high potential as elec-tron donors [23, 24]. Porphyrins and phthalocyanines are two-dimensional π-systems, being planar in most cases. However, introduction of many bulky substituents to porphyrin and phthalocyanine maclocycles results in distortion of the macrocycle π-plane to provide a curved

Photoinduced electron transfer in supramolecular

assemblies involving saddle-distorted porphyrins and

phthalocyanines

Takahiko Kojima* , Tatsuaki Nakanishi, Tatsuhiko Honda and Shunichi Fukuzumi*

Department of Material and Life Science, Graduate School of Engineering, Osaka University, SORST, Japan Science and Technology Agency (JST), Suita, Osaka 565-0871, Japan

Received 25 September 2008Accepted 22 October 2008

ABSTRACT: Unique supramolecular assemblies are constructed based on a saddle-distorted non-planar porphyrin, dodecaphenylporphyrin (H2DPP), and its metal complexes. The saddle distortion facilitates protonation of pyrrole nitrogens to afford a stable diprotonated porphyrin, which can act as an electron acceptor. A diprotonated hydrochloride salt of a saddle-distorted dodecaphenylporphyrin ([H4DPP]Cl2)forms a nano-sized channel structure called a “porphyrin nanochannel”. Electron-donating molecules such as hydroquinones are included as guest molecules in the porphyrin nanochannel. Photoinduced electron transfer from the guest molecules to the singlet state of H4DPP2+ occurs, producing H4DPP+•

and cation radicals of the guest molecules. The saddle distortion also results in higher Lewis acidity at a metal center to maintain axial coordination of ligands, due to poor overlap of the lone-pair orbitals with the d orbitals of the metal center. By taking advantage of saddle distortion of both H4DPP2+ and zinc phthalocyanine, 1,4,8,11,15,18,22,25-octaphenylphthalocyanine (ZnOPPc), a discrete supramolecular assembly composed of H4DPP2+ and ZnOPPc, is obtained using 4-pyridinecarboxylate (4-PyCOO-)that connects the two components by hydrogen bonding and coordination bonding, respectively. Photoexcitation of the assembly results in effi cient electron transfer from ZnOPPc to H4DPP2+ in the supramolecular complex.

KEYWORDS: supramolecular assembly, photoinduced electron transfer, saddle-distorted porphyrin, saddle-distorted phthalocyanine.

SPP full member in good standing

*Correspondence to: Takahiko Kojima, email: [email protected] and Shunichi Fukuzumi, [email protected], fax: +81 6-6879-7370, tel: +81 6-6879-7368


PHOTOINDUCED ELECTRON TRANSFER IN PORPHYRINS AND PHTHALOCYANINES 15

surface [25–27] for example, dodecaphenylporphyrin (H2DPP) [28] and octaphenylphthalocyanine (H2OPPc)exhibit saddle-distorted structures [29]. Such a curved surface can be utilized to construct novel self-assembled, three-dimensional structures that cannot be realized by using planar π-systems. In addition, distortion of the macrocycle π-plane is expected to change the acid-base and redox properties of porphyrins and phthalocyanines.

In this mini-review, we describe recent development on construction of novel self-assembled, three-dimensional structures of distorted macrocycles and photoinduced electron transfer in supramolecular complexes composed of distorted porphyrins and/or phthalocyanines.

Porphyrin nanochannels derived from self-assembly of saddle-distorted and protonated porphyrins

A porphyrin nanochannel (PNC) can be formed as a result of self-assembly of the saddle-distorted hydro-chloride salt of dodecaphenylporphyrin ([H4DPP]Cl2) by recrystallization from a mixture of chloroform (CHCl3)and acetonitrile (CH3CN) (Scheme 1) [30, 31]. This struc -tural motif is available regardless of the presence or absence of guest molecules. In the presence of guest mol-ecules such as hydroquinone derivatives, PNC can exhibit molecular inclusion of the guest molecules into the cavity. In this inclusion process, interaction is certain between the electron-defi cient porphyrin dications as electron acceptors and electron-rich hydroquinone derivatives as electron donors, to construct PNC-guest supramolecular assemblies as unique donor-acceptor organic materials (vide infra) [31]. Thus, these materials are formed via molecular recognition, depending on electronic charac-teristics and steric requirements of the guest molecules.

Recrystallization of the chloride salt of doubly protonated H4DPP2+ from CHCl3 with vapor diffu-sion of CH3CN gave crystals of a PNC, including

water molecules and CHCl3, formulated as [H4DPP]Cl2•0.5((H2O)4)•0.5(CHCl3)•(CH3CN)2 [PNC-water; Fig. 1(a)] [31]. Upon mixing of hydroquinone deriva-tives and the PNC-water, guest exchange from the water molecules to a series of guest molecules took place to give guest- including PNCs, denoted as [H4DPP]Cl2•(guest)•(CH3CN)2, in the course of recrystallization. We can recognize the {[H4DPP]Cl2•(CH3CN)2} unit as a skeleton unit. Among the skeleton units, a guest-inclusion cavity is provided by a pair of [H4DPP]Cl2. The nano-channel structure arose from self-assembly of [H4DPP]Cl2 as a building block and crystallized in a monoclinic space group of C2/c with the center of symmetry at the midpoint of two of the water molecules [31]. Two chan-nels run into the directions of [1 1 0] and [1 -1 0] in the crystal with an angle of ca 70° between them. The porphyrin dication units underwent intermolecular π–π interactions (3.44 ~ 3.59 Å) among the peripheral phenyl groups, mainly in the direction of the crystallographic c-axis [31]. The [H4DPP]2+ ion exhibited larger saddle distortion compared with that of [Mo(DPP)(O)(H2O)]+ in the porphyrin nanotube [26, 27] and the size of the chan-nel was estimated as 1.0 nm × 0.7 nm, smaller than that (1.0 nm × 1.4 nm) of the porphyrin nanotube [26, 27]. The chloride anions are hydrogen-bonded to the N-H protons that are oriented to the upper and lower sides of the mean porphyrin plane as shown in Fig. 1(b) [31]. The distances of N-H•••Cl- were determined to be 3.19 ~ 3.23 Å.

The crystal structure of the porphyrin nanochannel including hydroquinone (PNC-H2Q) was determined as shown in Fig. 2 [30, 31]. In the PNC-H2Q, the hydroqui-none molecule was included due to intermolecular π–πinteractions with H4DPP2+. Hydrogen bonding was observed for the hydroxyl groups of hydroquinone with the nitrogen atoms of the CH3CN molecules in the skel-eton units, at a distance of 2.85 Å, with the confi guration

[H4DPP]Cl2

PorphyrinNanochannel (PNC)

Self-assembly

In the presenceor absence ofguest molecules

Scheme 1. Schematic description of formation of porphyrin nanochannel


16 T. KOJIMA ET AL.

Fig. 1. Crystal structures of PNC-water (a) and an ORTEP drawing of the [H4DPP]Cl2 unit. (b) The H4DPP2+ unit is described with capped sticks in the crystal packing. Carbon, gray; oxygen, red; nitrogen, blue; chloride, light green

Fig. 2. Crystal structure of PNC-H2Q. The H4DPP2+ unit is described with capped sticks in the crystal packing. Carbon, gray; oxy-gen, red; nitrogen, blue; chloride, light green.

(a)

(b)



of the two hydrogens of the hydroxyl groups fi xed in a trans confi guration [31].

Due to the characteristics of PNC-guest supramo-lecular assemblies as donor-acceptor organic materials, photoexcitation of the porphyrin moiety can give rise to photoinduced electron transfer from the electron- donating guest molecules to the electron-accepting dicationic [H4DPP]2+ moiety in the solid state [31]. Under photoirradiation ( > 340 nm), PNC-hydroquinone deriv-atives showed the ESR spectra assigned to those of cation radicals of the corresponding guest molecules without

deprotonation [31]. This is represented by the ESR spectrum of H2Q

+• in PNC-H2Q under photoirradiation, depicted in Fig. 3 [31]. The spectrum was well simulated by hyperfi ne coupling constants due to the hydrogen nuclei attaching to the aromatic ring (1.6 G and 3.6 G) and the hydrogen nucleus of the OH group (4.6 G), indicating that the H2Q molecule maintains the trans confi guration without deprotonation [31].

In order to gain mechanistic insights into the photo-induced electron transfer in PNC-guests, femtosecond laser flash photolysis was performed in the solid state in KBr pellets [31]. As a blank test, a KBr pellet without samples did not show any absorption in the course of laser flash photolysis at 355 nm. In contrast, the tran-sient absorption spectra of PNC-H2Q are observed as shown in Fig. 4 [31]. The transient absorption spectrum observed at 1.5 ps, which has an absorption maximum at 580 nm, with bleaching at 500 nm, is assigned to the singlet excited state, 1(H4DPP2+)* [31]. The decay of absorption at 620 nm due to 1(H4DPP2+)* is accompa-nied by a rise in absorption at 530 nm. The transient absorption spectrum at 25 ps may be assigned to H4DPP+• produced by photoinduced electron transfer from H2Q to 1(H4DPP2+)* (vide infra). The rate constant of the photoinduced electron transfer is determined as 2.1 × 1011 s-1 at 298 K [31]. At 2000 ps (2 ns), the tran-sient absorption spectrum is again changed to that due to the triplet excited state of the dication, 3(H4DPP2+)*[31]. As for the absorption derived from H2Q

+•, it has been reported to exhibit the absorption maximum at 423 nm [32]. Thus, this absorption does not perturb the absorption spectra of the porphyrin-derived species described above.

We could establish the energy diagrams of PNC-guest assemblies on the basis of spectroscopic and elec-trochemical measurements on each component in solution as shown in Fig. 5 [31]. The energies of elec-tron-transfer states of PNC-H2QF4 and PNC-H2QCl4 are higher than that of 1(H4DPP2+)*; however, those assem-blies can undergo photoinduced electron transfer to

form their electron-transfer states involving H2QF4

+• and H2QCl4+•. This is probably due

to non-covalent interactions such as OH/πinteraction between the OH groups of the hydroquinone derivatives and π-electronsof the [H4DPP]2+ units to lower the oxida-tion potentials of those guest molecules. The photochemical processes in PNC-guest are summarized in Fig. 6 [31]. It has been proposed that the H4DPP+• produced by photoinduced electron transfer undergoes back electron transfer (charge recombina-tion; CR) to generate 3(H4DPP2+)* and H2Q, in addition to disproportionation viaelectron-hopping in the solid state, to give non-radical products and H2Q

+• as observed by ESR spectroscopy [31].

Fig. 3. ESR spectrum of H2Q+• generated in the single crystals

of PNC-H2Q under photoirradiation ( > 340 nm); measured at room temperature. Its computer-simulated spectrum is given as a lower trace

H

H H

H

O

OH

H

+•

4.6 (4.6)

1.6 (1.5)

3.6 (2.7)

HΔ msl = 1.7 G

g = 2.0036

16 G

Exp

Sim

Fig. 4. Femtosecond laser fl ash photolysis of PNC-H2Q with excitation at 355 nm in a KBr pellet: transient absorption spectra; at 1.5 ps (black), 25 ps (dark grey), and 2000 ps (pale grey)

500 600 700 800

0

4

–4

–8

103

ΔAbs

Wavelength,nm

1.5 ps25 ps

2000 ps

Fig. 5. Energy diagrams of photoexcitation of H4DPP2+ and the electron- transfer states of the various PNC-guest assemblies

1(Por2+)*-G

3(Por2+)*-GISC

1.72 eV

1.48 eV

Por+•-G+• (Electron-Transfer State)

G = o-H2Q

1.70 eV

G = p-H2Q

1.55 eV

G = H2QF4 G = H2QCl4

1.83 eV 1.85 eV

[H4DPP]Cl2-Guest (Por2+-G) [H4DPP]Cl2-Guest (Por2+-G)


18 T. KOJIMA ET AL.

Supramolecular assembly composed of ZnOPPc and H4DPP2+

Porphyrins exhibit strong Soret-bands around 400–450 nm, whereas phthalocyanines show strong Q-bands around 700–800 nm. Thus, the combination of those two π-systems can cover nearly the whole range of the visible region of sunlight and can be a useful strategy for the development of photofunctional materials for effi cient light-energy conversion. However, covalently linked porphyrin-phthalocyanine heterodyads usually exhibit fast energy transfer from the singlet excited state of the porphyrin moiety to give the singlet excited state of the phthalocyanine part, which decays without form-ing a charge-separated state [33]. In order to develop a novel electron-transfer system consisting of both por-phyrin and phthalocyanine, we took advantage of the hydrogen bonding of [H4DPP]2+ as an electron acceptor with an anionic saddle-distorted Zn(II)-phthalocyanine complex as an electron donor involving 4-pyridine car-boxylate (4-PyCOO–) as an axial ligand to give rise to a supramolecular donor-acceptor heterodyad [34]. As a saddle- distorted phthalocyanine complex, we chose a Zn(II) complex of 1,4,8,11,15,18,22,25-octaphenylphtha-locyanine (Zn(OPPc)) [34].

The reaction of a supramolecular assembly [H4DPP](4-PyCOO)2 with Zn(OPPc) was performed in toluene, followed by crystallization with vapor diffusion of hexane into the reaction mixture as described in Scheme 2 [34]. The 4- PyCOO- in [H4DPP](4-PyCOO)2 coor-dinates to the Zn(II) center in Zn(OPPc) through the pyridine nitrogen atom to form an unprecedented dis-crete supramolecular assembly, [(H4DPP){Zn(OPPc)(4- PyCOO)}2] (1) [34]. The crystal structure of 1 is depicted in Fig. 7 [34]. In the crystal structure, each of the two [Zn(OPPc)(4-PyCOO)] complexes forms two-point hydrogen bonding with protonated pyrroles in the saddle-distorted [H4DPP]2+ on both sides of the porphy-rin mean plane. The structure was maintained in solution since the 1H DOSY spectrum of 1 in CDCl3 allowed us to estimate its molecular volume to be 4849 Å3, in good agreement with that calculated on the basis of the crystal structure (5096 Å3) [34].

The photoinduced electron transfer in the supramo-lecular complex 1 was examined by femtosecond laser fl ash photolysis of a PhCN solution of 1 excited at 410 nm to observe transient absorption spectra as given in Fig. 8 [34]. At 4 ps after laser illumination, we could observe a spectrum due to the singlet excited states of both the [H4DPP]2+ and Zn(OPPc) moieties, with the rise of absorp-tion at 547 nm and 900 nm. At 150 ps, new absorption bands raised at 547 nm and 1030 nm. The absorption at 547 nm was assigned to that derived from H4DPP+• and that at 1030 nm was ascribed to the overlapped absorption of Zn(OPPc+•) and H4DPP+•, mainly due to Zn(OPPc+•),in comparison with reference spectra of both species

H4DPP2+

1[H4DPP2+]*

3[H4DPP2+]*

H2QH4DPP•+

H4DPP + H4DPP2+

PET

h

Disproportionation via

Electron HoppingkCS

kCR

kDISx2

H2Q

H2Q•+

H2Q•+

H2Q

ET State

Fig. 6. Summary of photochemical processes in PNC-H2Q in the solid state

HN

NNH

N N

N

N

N

N

N

N

N

Ph

Ph

Ph Ph

Ph

Ph

Ph

PhPhPh

Ph

Ph

Ph Ph

Ph

Ph

PhPh

Ph

Ph

N

OHO

Zn

H2DPP ZnOPPc4-PyCOOH

[H4DPP](4-PyCOO)2

{[H4DPP][Zn(OPPc)(4-PyCOO)]2} (1)

Scheme 2. Formation of 1Fig. 7. Crystal structure of 1. Carbon, grey; oxygen, red; nitrogen, blue; zinc, purple



[34]. Based on the rise of the absorption at 1030 nm, we were able to determine the apparent fi rst-order rate constant of 1.6 × 1010 s-1 [34]. This process was ascribed to the intra-supramolecular electron transfer from the Zn(OPPc) moiety to the [H4DPP]2+ one to form the electron-transfer state of 1, [{Zn(OPPc+•)(4- PyCOO)}(H4DPP+•){Zn(OPPc)(4-PyCOO)}] (1-ET). Prior to this electron transfer, we could observe very fast decay of the absorption at the rate of 4.2 × 1012 s-1 (2.4 ps) [34]. We assigned this decay to energy transfer from photoexcited 1(H4DPP2+)* to Zn(OPPc) to give 1(Zn(OPPc))*.The rate constant of the intra-supramolecular energy transfer was comparable to that of intramolecular energy transfer observed for a covalently linked ZnPor-ZnPc dyad [33]. After 150 ps, the absorption at 1030 nm decays to afford another fi rst-order rate constant to be 1.5 × 109 s-1 (Fig. 8) [34]. We assigned this process to intra- supramolecular back electron transfer from Zn(OPPc+•) to H4DPP+• to recover the ground state of 1. Therefore, the lifetime of the electron-transfer state (1-ET) was determined to be 667 ps, which was much longer than that (85 ps) observed in a covalently

linked ZnPc-SnPor dyad reported by Odobel and co-workers [35].

Together with electrochemical and spectroscopic mea-surements, we were able to establish an energy diagram as described in Fig. 9 [34]. Photoexcitation at 410 nm gives rise to the generation of the singlet excited states of both [H4DPP]2+ and Zn(OPPc) units (1(H4DPP2+)*

and 1(Zn(OPPc))*) in 1. The singlet excited state of the [H4DPP]2+ unit at the energy level of 1.66 eV undergoes fast intra-supramolecular energy transfer to give that of the ZnPc moiety at the energy level of 1.53 eV. This energy transfer is followed by intra-supramolecular electron transfer from 1(Zn(OPPc))* to [H4DPP]2+ to give the intra-supramolecular electron-transfer state 1-ET at the energy level of 0.83 eV in 63 ps. The ET state exhibits the lifetime of 667 ps, going back to the ground state.

CONCLUSION AND OUTLOOK

In this mini-review, we have described the utilization of a diprotonated porphyrin dication with saddle distor-tion as an electron acceptor in photoinduced electron transfer. The saddle distortion can facilitate the protona-tion of the pyrrole nitrogens and elevate the reduction potential of the porphyrin ring. The hydrochloride salt of H2DPP, [H4DPP]Cl2, forms PNC in the crystal as a result of its self-assembly. In the course of the forma-tion of PNC, the guest inclusion can be achieved with molecular recognition in terms of the size and character-istics of the guest molecules. The inclusion occurs only for intrinsically electron-donating molecules in accept-able sizes for the cavity. It has been demonstrated that the saddle-distortion of porphyrins and phthalocyanines cooperatively stabilizes the supramolecular assembly by virtue of the facile protonation of a distorted porphyrin to form strong hydrogen bonding, and also by the enhanced Lewis acidity of Zn center in a distorted phthalocyanine to strengthen the axial coordination bond. In both cases, photoexcitation of the assemblies affords the singlet excited states of the [H4DPP]2+ moiety, 1(H4DPP2+)*. In the case of PNC guests, 1(H4DPP2+)* undergoes electron transfer from the guest molecules in the nanochannels to give electron-transfer state in the solid state. On the other hand, in the case of the porphyrin-phthalocyanine supramolecule 1, the photoexcited 1(H4DPP2+)* per-forms energy transfer to generate the singlet excited state of the Zn(OPPc) moiety, which clearly gives rise to the generation of an unprecedented intra-supramolecular, electron-transfer state (1-ET) involving cation radicals of both one-electron-reduced diprotonated porphyrin and one-electron-oxidized Zn(OPPc). The electron-transfer state 1-ET can generate via the singlet excited states of both the [H4DPP]2+ and the Zn(OPPc) moieties. The strategy using distorted porphyrins and phthalocya-nines described here can be applied to the construction of novel supramolecular assemblies consisting of por-phyrins and phthalocyanines to develop photofunctional

600 800 1000 1200

2

4

0

–2

102

ΔAbs

Wavelength (nm)

4 ps150 ps

3000 ps

Fig. 8. Femtosecond laser fl ash photolysis of 1 in PhCN: tran-sient absorption spectra; at 4 ps (black), 150 ps (dark grey), and 3000 ps (pale grey)

Pc-Por2+-Pc (1)

Pc+ -Por+ -Pc (1-ET)

1Pc*-Por2+-Pc

1.53 eV

0.83 eV

Pc-1[Por2+]*-Pc

1.66 eV

667 ps

63 ps

2.4 ps

Fig. 9. Energy diagram of photodynamics of 1 in PhCN


20 T. KOJIMA ET AL.

materials, which can utilize a wide range of visible light effectively.

Acknowledgements

The authors gratefully acknowledge the contributions of collaborators and co-workers mentioned in the refer-ences. The authors also acknowledge support of their study by Grants-in-Aid and a Global COE program, “The Global Education and Research Center for Bio- Environmental Chemistry” from the Ministry of Edu-cation, Culture, Sports, Science and Technology, Japan and Japanese Society of Promotion of Science. T. K. appreciates support from Japan Society of Science and Technology through “the potentiality verifi cation stage” program.

REFERENCES

Blankenship RE (Ed.). 1. Molecular Mechanisms of Photosynthesis, Blackwell Science, 2002.Gust D, Moore TA and Moore AL. In 2. Electron Transfer in Chemistry, Balzani V (Ed.), Vol. 3, Wiley-VCH: Weinheim, 2001; 272–336.a) Fukuzumi S and Guldi DM. In 3. Electron Transfer in Chemistry, Balzani V (Ed.), Vol. 2, Wiley-VCH: Weinheim, 2001; 270–337. b) Fukuzumi S. Org. Biomol. Chem. 2003; 1: 609–620. Fukuzumi S. 4. Bull. Chem. Soc. Jpn. 2006; 79:177–195.Fukuzumi S. 5. Phys. Chem. Chem. Phys. 2008; 10:2283–2297.a) Guldi DM. 6. Chem. Soc. Rev. 2002; 31: 22–36. b) Guldi DM, Rahman GMA, Zerbetto F and Prato M. Acc. Chem. Res. 2005; 38: 871–878. c) Guldi DM. J. Phys. Chem. B. 2005; 109: 11432–11441.a) Tsuda A and Osuka A. 7. Science 2001; 293: 79–82. b) Cho HS, Jeong DH, Cho S, Kim D, Matsuzaki Y, Tanaka K, Tsuda A and Osuka A. J. Am. Chem. Soc. 2002; 124: 14642–14654.a) Imahori H, Guldi DM, Tamaki K, Yoshida Y, Luo 8.C, Sakata Y and Fukuzumi S. J. Am. Chem. Soc.2001; 123: 6617–6628. b) Guldi DM, Imahori H, Tamaki K, Kashiwagi Y, Yamada H, Sakata Y and Fukuzumi, S. J. Phys. Chem. A 2004; 108: 541–548. c) Imahori H, Sekiguchi Y, Kashiwagi Y, Sato T, Araki Y, Ito O, Yamada H and Fukuzumi S. Chem.–Eur. J. 2004; 10: 3184–3196.Fukuzumi S and Kashiwagi Y. 9. J. Porphyrins Phtha-locyanines 2007; 11: 368–374.Choi M-S, Aida T, Luo H, Araki Y and Ito O. 10. Angew. Chem. Int. Ed. 2003; 42: 4060–4063.Kelley RF, Shin WS, Rybtchinski B, Sinks LE and 11.Wasielewski MR. J. Am. Chem. Soc. 2007; 129:3173–3181.Kodis G, Terazono Y, Liddell PA, Andréasson J, 12.Garg X, Hambourger M, Moore TA, Moore AL

and Gust D. J. Am. Chem. Soc. 2006; 128: 1818–1827.Imahori H and Fukuzumi S. 13. Adv. Funct. Mater.2004; 14: 525–536.Fukuzumi S and Kojima T. 14. J. Mater. Chem. 2008;18: 1427–1439.Fukuzumi S. In 15. Functional Organic Materials, Wiley-VCH: Weinheim, 2007; 465–510.Hasobe T, Imahori H, Kamat PV, Ahn TK, Kim SK, 16.Kim D, Fujimoto A, Hirakawa T and Fukuzumi S. J. Am. Chem. Soc. 2005; 127: 1216–1226.Hasobe T, Fukuzumi S, Hattori S and Kamat PV. 17.Chem. Asian J. 2007; 2: 265–272.Saito K, Qiu H, Troiani V, Solladié N, Sakata T, 18.Mori H, Ohama M and Fukuzumi S. J. Phys. Chem. C 2007; 111: 1194–1199.Herranz MA, Ehli C, Campidelli S, Gutierrez M, Hug 19.GL, Ohkubo K, Fukuzumi S, Prato M, Martin N and Guldi DM. J. Am. Chem. Soc. 2008; 130: 66–73.a) D’Souza F, Chitta R, Gadde S, Shafi qul Islam 20.DM, Schumacher AL, Zandler ME, Araki Y and Ito O. J. Phys. Chem. B. 2006; 110: 25240–25250. b) D’Souza F, Chitta R, Gadde S, McCarty AL, Karr PA, Zandler ME, Sandanayaka ASD, Araki Y and Ito O. J. Phys. Chem. B. 2006; 110: 5905–5913.D’Souza F and Ito O. 21. Coord. Chem. Rev. 2005; 249:1410–1422.Tanaka M, Ohkubo K, Gros CP, Guilard R and 22. Fukuzumi S. J. Am. Chem. Soc. 2006; 128: 14625–14633.Leznoff CC and Lever ABP (Eds.). 23. Phthalocya-nines: Properties and Applications, Wiley-VCH: Weinheim, 1989, 1993, 1996; Vols. 1–4.Rodríguez-Morgade MS, de la Torre G and Torres 24.T. In The Porphyrin Handbook, Kadish KM, Smith KM and Guilard R (Eds.), Academic Press: San Diego, 2003.Yokoyama A, Kojima T and Fukuzumi S. 25. Chem.Commun. 2007; 3997–3999.Harada R, Matsuda Y, Okawa H and Kojima T. 26.Angew. Chem. Int. Ed. 2004; 43: 1825–1828; Angew. Chem. 2004; 116: 1861–1864.Kojima T, Harada R, Nakanishi T, Kaneko K and 27.Fukuzumi S. Chem. Mater. 2007; 19: 51–58.Medforth CJ, Senge MO, Smith KM, Sparks LD 28.and Shelnutt JA. J. Am. Chem. Soc. 1992; 114:9859–9869.a) Kobayashi N, Fukuda T, Ueno K and Ogino H. 29.J. Am. Chem. Soc. 2001; 123: 10740–10741. b) Fukuda T, Homma S and Kobayashi N. Chem.–Eur. J. 2005; 11: 5205–5216.Harada R and Kojima T. 30. Chem. Commun. 2005;716–718.Kojima T, Nakanishi T, Harada R, Yamauchi S, 31. Ohkubo K and Fukuzumi S. Chem.–Eur. J. 2007; 13: 8714–8725.



Tripathi GNR. 32. J. Am. Chem. Soc. 1998; 120: 5134–5135.Soares ARM, Martínez-Díaz MV, Bruckner A, 33.Pereira AMVM, Tomé JPC, Alonso CMA, Faustino MAF, Neves MGPMS, Tomé AC, Silva AMS, Cavaleiro JAS, Torres T and Guldi DM. Org. Lett.2007; 9: 1557–1560.

Kojima T, Honda T, Ohkubo K, Shiro M, Kusukawa 34.T, Fukuda T, Kobayashi N and Fukuzumi S. Angew. Chem. Int. Ed. 2008; 47: 6712–6716.Fortage J, Göransson E, Blart E, Becker H-C, Ham-35.marström L and Odobel F. Chem. Commun. 2007; 4629–4631.




INTRODUCTION

It is widely appreciated that porphyrins are useful func-tional elements in naturally-occurring and artifi cial sys-tems. A vast knowledge of their photonic and electronic properties as well as in the synthesis and mo difi cation of porphyrin derivatives has been accumulated [1]. There-fore, they make good candidates for incorporation into advanced scientifi c and technological applications. Use of porphyrins in nano-organized systems is undoubtedly an attractive research target. For example, photoinduced electron transfer using a supramolecular ferrocene-zinc porphyrin-py ri dyl naphthalenediimide triad was recent-ly reported by Fukuzumi and Kashiwagi [2]. Recently,

related su pra molecular nanosystems have been intensive-ly in ves tigated [3].

The importance of porphyrin organization is also well recognized in biological systems. For example, porphy-rins and related molecules have essential ro les in many biological systems such as in photosynthe tic complexes where porphyrins form sophisticated assemblies with proteins and other dyes. The anten na complex of pho-tosynthetic bacteria consists of a core light-harvesting antenna (LH1) and a periphe ral light-harvesting antenna (LH2). Excitation energy migrates within the wheel-like arrays of chlorophylls in the LH1 and LH2 complexes and is fi nally funneled into the chlorophyll dimer (spe-cial pair) at the photosynthetic reaction center. Such self-assembled systems contribute to the collection of light energy, which is of some interest as one potential mecha-nism for solving global energy problems [4].

Some functionalized porphyrins, such as por phy rin-crown ether conjugates, are useful for artifi cial light-harvesting

New aspects of porphyrins and related compounds: self-

assembled structures in two-dimensional molecular arrays

Katsuhiko Ariga*a, Jonathan P. Hill*a , Yutaka Wakayamab, Misaho Akadaa,

Esther Barrenac and Dimas G. de Oteyzac

a World Premier International (WPI) Research Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba 305-0044, Japanb Advanced Electric Materials Center, National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba 305-0044, Japanc Max-Planck Institute for Metals Research, D-75069 Stuttgart, Germany

Received 25 August 2008Accepted 7 September 2008

ABSTRACT: The advanced state of development of molecular design and synthetic chemistry of por-phyrins and related molecules makes these compounds good candidates for technological appli cations,in which well characterized and designed structures and properties are required. In particu lar, 2-dimen-sional molecular level control of porphyrin array structures should reveal new aspects of nanotechno-logy. In this review, recent research on porphyrin assemblies, including 2-dimensio nal porphyrin arrays, is described with emphasis on phenol- and quinone-substituted tetra pyrrole units. A series of research aimed at developing strategies for preparation of porphyrin molecular arrays, where several novel aspects of molecular arrays, including phase transitions, ordered 2-D phase boundaries, and hydrogen-bonding networks, are introduced.

KEYWORDS: porphyrin, self-assembly, two-dimensional, molecular array, oxoporphyrinogen.


*Correspondence to: Katsuhiko Ariga, email: [email protected], fax: +81 29-860-4832 and Jonathan P. Hill, email: [email protected], fax: +81 29-860-4832


SELF-ASSEMBLED STRUCTURES IN TWO-DIMENSIONAL MOLECULAR ARRAYS 23

systems, as recently reported by Solladié et al. [5]. Addi-tionally, develop ment of me tho do lo gies to create designer porphyrin assem blies has be come important since these strat-egies can lead to por phyrin arrays similar to the natural light-har vesting systems but immobilized at an appropriate so lid substrate such as an electrode [6]. The Lang muir-Blodgett (LB) [7] method and layer-by-layer (LbL) [8] assembly are well-known strategies for preparation of well-ordered, layered molecular assemblies. For example, Torimoto, Takagi and coworkers re por ted sequential LbL structures of tetrakis(4-car bo xyphenyl) por phyrin and 1,3,6,8-pyrene-tetra sul fo nic acid with the aid of layered double hydroxide nano sheets, where photo -induced energy transfer between these two organic components was investiga ted [9]. Al-though some examples of two-dimensional mole cular pat-terning based on the LB technique have been reported [10], these two methodologies (LB and LbL) are not useful for regulation of molecular arrangement at substrates in two dimensions. Other methods such as self-assembled mono-layer (SAM) techniques and specifi c adsorption to surfaces are necessary for preparation of nanostructured porphyrin arrays [11]. For example, Arima et al. immobilized a sub-mono layer of the chromophore zinc(II) 5,10,15,20-tetra-kis (4-tert-butyl phenyl) porphyrin by means of an aro matic thiol mono layer, 4-amino thio phenol SAM, on the Au(111) surface, where gold clusters bound to the molecules were moved under an electric fi eld produ ced by the tip of a scan-ning tunneling microscope (STM) [12].

Self-assembly of porphyrins and related molecules has become one of the dominant concepts in development of porphyrin functional systems [13], some of which have been found to mimic the excellent pro perties seen in na-turally occurring systems. In par ti cu lar, molecular level control of porphyrin arrays in two dimensions is a rapidly developing fi eld in na no technology. In this review, recent research on por phy rin assemblies, including 2-dimensio-nal porphyrin arrays, is described with emphasis on phe-nol- and quinone-substituted tetrapyrrole units. A series of research aimed at developing strategies for prepara tion of porphyrin molecular arrays, where several no vel aspects of molecular arrays, including phase transitions, ordered 2-D phase boundaries, and hydrogen-bonding networks, are introduced.

RECENT STUDIES ON PORPHYRIN ASSEMBLIES IN THREE DIMEN SIONS

In this section, recent research on porphyrin as sem-blies are briefl y summarized. Further examples can be found in recent reviews of the related research areas [14].

Before describing assemblies of porphyrins, porphy-rin arrays based on covalent linkages are introdu ced. The strategy of covalent linkage can be regar ded as a method to prepare stable porphyrin clus ters, whereas molecular assemblies are not always sta ble under harsh conditions and sometimes disso ciate in to their individual compo-nents. Osuka and co wor kers have synthesized many

excellent examples of covalently-linked porphyrin arrays [15]. For exam ple, they synthesized several kinds of por-phyrin oligo mers of varying discrete lengths up to a 128-mer, and their specifi c photo electronic pro perties were investigated [16]. Very recently, the same research group suc ceeded in the synthesis of longer 1,3-pheny lene brid-ged porphyrin arrays and their intra mo le cular cyclizationreactions to large nano meter-scale por phy rin wheels [17]. Investigation on photo elec tronic pro perties of porphyrin oligomers and con ju gates as well as their syntheses have been reported by Anderson, Albunsson and coworkers [18], Tsuji, Tamao and co workers [19] and Cramariuc et al. [20]. Por phy rin po lymers appended with den dritic moi eties have also been used as species possessing light-har ves ting pro perties [21].

Although the covalent approach provides elegant oli-goporphyrinic conjugates, it requires a conside ra ble syn-thetic effort. Therefore, cluster formation of por phyrin structures through self-assembly of simple molecularunits is a more practical approach. Because porphyrin compounds can coordinate a metal cation at their core, molecular association through coordi na tion is often used for formation of porphyrin assem blies. Coordination complexes have fi xed geometries, a feature which can be used for construction of por phy rin assemblies of de-sired geometries and mor pho lo gies. Coordination-based porphyrin assemblies ha ve been recently summarized in excellent reviews by Kobuke [22] and Alessio and cowor-kers [23]. Kobuke and coworkers developed a supramole-cular approach for the construction of porphyrin arrays, prin ci pally using coordination between central metal and imida zole groups attached at the meso-posi tion [24]. A series of conjugated zinc porphyrin oligo mers for med stable ladder complexes with linear biden tate li gands, such as 1,4-diaza bi cyclo [2.2.2] octane (DABCO) and 4,4’- bi pyri dyl, as reported by Anderson and coworkers [25]. Similarly, DABCO-induced self-assem bly of a triporphy-rin double-decker cage was de mons tra ted by Ballester, Hunter and coworkers [26]. Rowan and coworkers self-assembled disk-shaped he xa kis (porphyrinato) benzene molecules into linear ar rays through metal coordination [27]. Hupp and co workers recently reported on the forma-tion of well-de fi ned, prism-shaped assemblies that were similarly pre pa red via coordinative self-assembly with the tri ethy nyl pyridyl benzene ligand [28].

Various interactions other than coordination have been used for porphyrin assemblies. D’Souza, Ito and cowor-kers prepared a special pair donor consisting of a cofacial porphyrin dimer through potassium ion induced dime-rization of meso-(benzo-[15]crown-5)-porphyrinatozinc, which subsequently self-assem bled with functionalized fullerenes by axial coordina tion and crown ether-alkyl ammonium cation com ple xa tion to form the donor- acceptor pairs [29]. Kobuke and coworkers reported self-assemblies consisting of ace tylene-linked bisporphyrins with a 4-nitrophenyl ethy nyl substituent and with a pheny-lethynyl substi tuent [30]. This assembly may be attributed


24 K. ARIGA ET AL.

to the par ti cipation of the nitro group in the delocalized electronic system. Hydrogen bonding is a powerful tool for preparation of defi ned porphyrin assem blies. Effi cient energy transfer from a cytosine-tethered zinc-por phyrin to a guanine-linked, free-base porphyrin congener, which are specifi cally paired through hy dro gen bonding, was demonstrated in pioneering work by Sessler and cowor-kers [31]. Recently, Nishide and coworkers reported the selective formation of a cy clic tetra mer from a metallo-porphyrin with two self-complementary quadruple hydrogen-bon ding units [32]. Sada and coworkers repor-ted the cons truction of a double-stranded helix genera-ted by the twis ting of a polymeric supramolecular ladder which relies on the stacking of the bridging porphyrins at the in ner ladder region, and on the complementary hy dro gen bonds formed at the polymeric back bone [33]. Gulino et al. reported porphyrin assemblies of the te tra ca tionic meso-tetra kis (N-methyl-4-pyridyl)porphy-rin and its metallo-derivatives with the o cta-ani onic form of 5,11,17,23-tetrasulfonato-25,26,27,28-tetra kis-(hydroxy carbonyl methoxy) calix[4] arene via elec tro-static inclusion complexation, where the sto ichio metry and porphyrin sequence can be easily tuned [34]. Fla-migni et al. investigated a [2] catenane assembly contai-ning a zinc(II) porphyrin, a gold(III) porphyrin, and two free phenan throline binding sites, and the corresponding copper(I) phenanthroline complex [35]. There are some unique examples in re cent researches. For example, Safaei et al. investi ga ted interaction of N,N′,N′′,N′′′-tetra methyl tetra-2,3-py ridino porphyra zinato nickel(II), and N,N′,N′′,N′′′-te tra methyl tetra-2,3-pyridino porphy-razinato iron(II), with calf thymus DNA [36]. Gedan-ken and co wor kers reported that nickel phthalo cyanine synthe si zed in an ionic liquid (1-butyl-3-methyl imi da-zo lium tetra fl uo ro borate) results in one-dimensio nal structures [37].

Further extension of self-assembly lead to materializa-tion of porphyrin derivatives. For example, Aoyama andcoworkers pioneered the construction of su pra molecular network solids stabilized by hy drogen bon ding between porphyrin-bis-resor cinol units [38]. Harada and Kojima developed the dica tion of the saddle-distorted dode ca-phenyl por phyrin as a buil ding block for novel nanochan-nel materials [39]. They al so reported preparation of porphyrin nano tube and inclusion of Mo-Oxo clusters [40]. Porphy rin-con taining gels were also reported. Liu and co wor kers reported gel-formation and supramole-cular chira lity induction [41]. Shinkai, Takeuchi and co workers reported a series of porphyrin-based gels [42]. Shelnutt and coworkers reported the synthesis of discrete free-standing porphyrin nanosheets using a re preci pi ta-tion method, in which an ethanolic solution of Sn(IV) 5-(4-pyridyl)-10,15,20-tri phenyl por phyrin dichlo ri de was simply injected into deionized water at room tem perature under vigorous stirring [43]. They also de monstrated that phase-transfer reactions could be used to generate por-phyrin nanofi ber bundles by the pha se-transfer ionic

self-assembly of water-soluble por phyrins with water-insoluble porphyrins [44]. Gong et al. reported the fi rst synthesis and charac terization of porphyrin nanoparticles in 20–200 nm diameter [45]. Charvet et al. reported the self-assem bly of a non-lipid type -electronic amphi-phile consis ting of a zinc porphyrin-fullerene dyad into uniformly-sized microvesicles in aqueous media [46].

RECENT STUDIES ON PORPHYRIN ASSEMBLIES IN TWO-DIMENSIONS

Unlike self-assembled structures in three-dimensional aggregates and solid materials, molecular arrays formed at solid interfaces provide us with opportunities to connect porphyrin functions with man-made devices. Characteri-zation of molecular ar rays on surfaces requires high reso-lution analysis, where scan ning probe microscopies (SPM) such as scan ning tunnel microscopy (STM) and atomic force mi cros copy (AFM) play crucial roles in identifying po sition and ordering of molecules on the surface. For these studies, shape-defi ned molecules inclu ding porphy-rin, phthalocyanine, and fullerenes can be good candida-tes of component molecules because they have predicta-ble shapes. Therefore, observa tion of mole cular arrays of these molecules have been inves tigated intensively, and have been summarized in re cent reviews [47]. In parti-cular, technical progress in molecular level observation using STM enables us to characterize porphyrin-based assembled structures with molecular level precision. In the following part, re cent examples of porphyrin assem-blies having precise molecular arrangements are briefl y summarized.

Otsuki et al. recently reported that monolayer ar rays of various structures can be obtained by as sem bling series of meso-tetra-substituted porphy rins con taining octadecyloxy and carboxyl (or pyridyl) groups on HOPG surface at the liquid/solid inter face [48]. They also re-ported surface patterns com posed of double-decker spe-cies [49]. Spillmann and cowor kers presented a detailed investigation of the ad sorp tion and dynamics of C60 and C70 fullerenes hosted in a self-assembled, two-dimensio-nal, nano po rous porphy rin network on a solid Ag surface [50]. Wakayama and coworkers reported adsorption- in du ced chiral con formations of a porphyrin molecule on a Cu sur face, resulting in the formation of twin su per-struc tures [51]. Elemans and coworkers fi rst repor ted an STM image of meso-tetra dode cyl por phy rin im mo bi lized on Au(111) [52a], and they recently demons trated STM observation of giant disk-like porphy rin ar rays, namely, porphyrin dode camers at liquid-solid in terfaces, where the distance between the por phy rin molecules on the surface can be precisely con trol led [52b]. Feringa and cowor-kers observed a self-or ga nized monolayer of meso-te tra dode cyl por phy rin coordinated to a Au(111) surface [53]. Auwärter and co workers reported controlled me-talation of self-as sembled porphyrin nanoarrays in two dimensions [54]. Itaya et al. proposed a unique strategy



for con trol of electrochemical reactions with the aid of a metal lo octa ethyl por phyrin on a Au(111) electrode, ba-sed on detailed STM study [55]. Although a fulle rene (C60)-ferrocene dyad directly attached to the Au(111) elec-trode gave a poorly defi ned electrochemical response, co-assembling with the porphyrin resul ted in a well-defi ned electrochemical reaction of the ferrocene group. The well-defi ned electrochemical response of the ferrocenyl group was due to the control of orientation of the C60-ferrocene dyad mole cules. The same research group recently reported mole cular images of supramolecular nanobelt arrays consisting of a cobalt(II) “picket-fence” porphyrin, cobalt(II) 5,10,15,20-tetra kis (R,R,R,R-2-piv-ala- mido phenyl) por phyrin, on Au surfaces [56]. Elemans and cowor kers reported the single-molecule imaging of oxida tion catalysis by monitoring, with STM, individual manganese porphyrin catalysts, in real time, at a li quid-solid interface [57]. In pioneering work, Yokoyama and coworkers used site-specifi c hydrogen bon ding between porphyrin components in 2 dimensions [58]. Two kinds of tetra phenyl porphyrin that contain two cyano groups in different relative positions (cis-isomer and trans-isomer)were used. Cyano groups introduced at the porphyrin phenyl substituents can dimerize through hydrogen-bond formation. Tetra meric cyclic domains were formed preferentially in the case of the cis-isomer, while linear molecular wires were found in STM observations of the trans-isomer. The same research group recently reported the multi la yer thin fi lm growth of carboxyphenyl-substi-tuted porphyrins on Au(111) [59].

From the viewpoint of bottom-up fabrication, forma-tion of molecular arrays with multiple components is an important step towards development of more highly inte-grated systems. For example, Bonifazi et al. summarizedaddressable, multi-component, molecular arrays from C60

and porphyrin derivatives [60]. Yoshimoto et al. proposed a guide for desig ning two-dimensional alternate arrays consisting of bi components from solution phase, where the forma tion of supramolecularly patterned fullerenes on the two-dimensional bimolecular structure consisting of porphyrin and phthalocyanine molecules was inves-tigated on Au(111) and Au(100) electrode surfaces by using electrochemical scanning tunneling microscopy [61]. Very recently, Barrena, Wakayama and co workers used a sophisticated approach for two-dimensional supra-molecular self-assembly of binary organic monolayers through vacuum deposition of bi-indenoperylene and fl uorinated copper-phthalo cya nine that exhibit p-typeand n-type conduction, res pectively [62]. As summarized in Fig. 1A, mixing patterns of these components change depending on both substrates (Au(111) or Cu(111)) and mixing ratio. Di-indenoperylene and fl uorinated copper-phthalocyanine forms a neat alternate arrangement on the Au(111) surface as seen in the enlarged image (Fig. 1B(a)). Formation of an entropically unfavorable re gular arrangement must be accompanied by some en thalpicgain. AM1 semi-empirical computations re veal that one

of the metastable structures contai ning a well-developed hydrogen bond array between F- and H- atoms (Fig. 1B(b)) closely matches the ex perimentally determined arran-gement. Because these calculations did not include contributions from the substrate, molecular interactions between two components should be the major factor in determining the structures on Au(111). In sharp contrast, molecular patterns of di-indenoperylene and fl uorinated copper- phthalocyanine on Cu(111) differ markedly from that on Au(111) (Fig. 1C(a)). The proposed hydrogen bonding pattern (Fig. 1C(b)) cannot be regenerated by the substrate-free computation. This suggests that interaction of these components with Cu(111) cannot be ignored. A series of investigations using varied mixing ratios revealed that an ordered array was only formed at particular ratios, where appropriate molecule-substrate and molecule–molecule interactions are satisfi ed. This phenomenon can be regarded as “molecular alloy”.

ASSEMBLY OF OXO PORPHYRINO GEN

We have been recently developing systematic approa-ches to modulate structures of porphyrin molecular arrays using the oxoporphyrinogen unit. As shown in Fig. 2A, this particular unit provides several op portunities to modify its structure. Rotational freedom between central porphyrin core and phenyl rings is permitted in its redu-ced form. However, oxidation to its quinone form results in a fi xed planar conformation. Hydroxy groups at the phenyl substituents can potentially form hydrogen bonds with neighbo ring molecules although this interaction depends on ortho-alkyl substituent size and orientation. Bulky tert-butyl groups strongly hinder hydrogen bon-ding between hydroxyl groups, while small methyl groups permit hydrogen bond formation. These features of the oxoporphyrinogen unit provide certain variations in two-dimensional molecular arrays. In addition, N-substitu-tion at the porphyrin core leads to out-of-plane modifi ca-tion. In this section, several recent attempts in our group to control two-dimensional oxo porphyrinogen arrays are introduced.

Fundamental properties

Fundamental properties of oxoporphyrinogen deriva-tives including spectroscopic and electrochemical cha-racteristics have been investigated by Hill and cowor-kers [63]. In particular, characterization of N-substi-tuted oxoporphyrinogen derivatives (Fig. 2B) such as N-naphthyl-2-ylmethyl [64] and N-py ren-1-ylmethyl derivatives [65] was recently perfor med. Electrochemi-cal and photochemical properties of bis(zinc-porphyrin)-oxoporphyrinogen derivative (Fig. 2B) as a novel donor-acceptor pair were also in vestigated [66]. The modifi ed species were used as twisted two-faced porphyrin hosts for bispyridyl ful lerenes [67]. Such basic research is not limited to the oxoporphyrinogen families. Tautomerism


26 K. ARIGA ET AL.

of oxocorrologens (Fig. 2C) has also been recently investiga ted [68].

Phase transition of two-dimensional array

In order to exploit potential host-guest proper ties for sen-sor applications, immobilization of the oxo por phyrinogen

structures on a solid surface is necessary. Formation of well-ordered molecular arrays is essential for their use in many advanced nanotechnological applications. In the remaining sections, our efforts to prepare two-dimensional molecular arrays of oxoporphyrinogen and some phenol-substituted por phyrins are briefl y introduced.

Fig. 1. (A) Two-dimensional molecular patterns prepared from various mixtures of di-indenoperylene and fl uorinated copper-phthalocyanine on Au(111) and Cu(111). (B) Molecular pattern of equimolar mixture of di-indenoperylene and fl uorinated cop-per-phthalocyanine on Au(111): (a) STM image; (b) plausible hydrogen bonding. (C) Molecular pattern of equimolar mixture of di-indenoperylene and fl uorinated copper-phthalocyanine on Cu(111): (a) STM image; (b) plausible hydrogen bonding



In the intial section, phase transition of the two-dimen-sional molecular arrays of the oxoporphyri no gen in its reduced form (tetrakis(3,5-di-t-butyl-4-hy droxyphenyl)porphyrin, TDtBHPP) and fi xation of the array through oxidation (oxidized form, Ox (TDtBHPP)) are described (see Fig. 2A for their struc tures) [69]. The reduced spe-cies TDtBHPP has struc tural characteristics similar to conventional te tra-substituted porphyrins such as tetra-phenyl por phy rin (TPP) derivatives, containing an essen-tially pla nar te tra pyrrole macrocycle subtending an angle of 60º with the planes of its phenyl substituents, while the 2-electron oxidized and formally non-aromatic deri va tive Ox(TDtBHPP) has dihedral angles between the planes of the pyrrole groups and least squares plane of the macro-cycle even and alternating at ±48°. In the latter form, the meso-substituents are approxima te ly co planar with the macrocyclic least squares plane and they are fi xed in this confi guration by the conjuga ted -electron system.

As seen in Fig. 3A, TDtBHPP at submonolayer co ve-rage, which was deposited on Cu(111) under ul tra high vacuum, comprises surface-mobile hexago nal ly-packed domain islands interspersed in a two-di men sional gas

phase at lower temperature. Upon increa sing the tempe-rature to ambient, the hexagonally-pa cked structure under-goes a transition to the square pa cked grid motif. Because this phase transition oc curs on a timescale of seconds, observation of the phase transition becomes possible by scanning tunne ling mi croscopy (STM). The hexagonal packing at lower tem perature is observed where the dihedral angle is low. This molecular conformation with an increased co planarity between porphyrin and meso-substi tuents is stabilized by the presence of electron- donating hydroxyl groups at the periphery of TDtBHPP. The disc-like shape of the planar conformations of TDtB-HPP favors a hexagonal arrangement for optimization of intermolecular van der Waals contacts. On the other hand, the square packing at ambient temperature is associated with a molecular conformation having a large dihedral angle between the porphyrin macrocyclic plane and the phenyl substituents. Non-planar conformations optimize the van der Waals contacts between tert-butyl groups by formation of the square grid structure. The structure at the contacts between molecules with non-planar confor-mations is mirro red in the crystal structure of TDtBHPP,

Fig. 2. (A) Tetrakis(3,5-di-t-butyl-4-hydroxyphenyl)porphyrin (TDtBHPP) and its oxidized form (Ox(TDtBHPP)). (B) Various N-substituted oxoporphyrinogens. (C) Oxo cor ro logen


28 K. ARIGA ET AL.

resulting in linear arrays of the porphyrin. The grid struc-ture observed by STM is an extension of these linear arrays into two dimensions and is made possible by nearly orthogonal substitution of the porphyrin with identical groups. From the viewpoint of the conformational ener-gy of an individual molecule, square-packing is prefera-ble. However, van der Waals contacts among neighbo-ring molecules and those between molecule and surface are reduced when compared with those in the hexagonal packing domains. In contrast, the stabi lity of the hexa-gonal packing arises from increased intermolecular and surface-molecule contacts of the co-planar conforma-tion, although this comes at the ex pense of an increased conformational energy of the distorted molecular confor-mation. Coexistence of the hexagonal and square domains indicates that both phases are in fact metastable and the energy bar rier bet ween these domains is relatively small. Sequen tial ob servation elucidated that the difference in the conformational energies between the two domains is actually small enough to allow alternative packing. Tran-sition from hexagonal phase to square phase is highly cooperative, where a domino-like effect in this transition structure took place.

Oxidation of these molecules to Ox(TDtBHPP) dras-tically changes the situation, where hexagonal packingsare maintained even at ambient temperature (Fig. 3B). The lack of mobility in the oxidized form leads to a much stronger molecule-surface interaction. The increased coplanarity of the oxidized porphyrin increases the multiplicity of contacts between tert-butyl substituents and the metal surface allowing for mation of structures that are stable and immobile even at ambient temperatu-res, similar to the case of the planar conformations of the non-oxidi zed porphyrin. There can be no possible com-peting conformations of the molecule, and this is the main factor responsible for improvement of the stability

and immobility of the adsorbed structures since all of the tert-butyl groups can remain in contact with the surface. This is meaningful control in two-dimensional molecu-lar array. Tuning of the oxidation state of the adsorbed molecule can determine stability of the molecular array, which permits a new level of control over self-assemblies at metallic interfaces in compounds containing appro-priate substituents.

Conformational adaptation at phase boundary

Detailed STM analyses revealed an interesting me-chanism for molecular packing at the phase boundary between two 2-dimensional patterns. The term “lattice mismatch” is more commonly used in solid state che-mistry especially in epitaxially-grown thin fi lms of inor-ganic materials. We found a similar mismatch between lattice parameters of two-dimensional molecular arrays of TDtBHPP on Cu(111) surface [70]. In STM images, the molecules in the hexagonally-packed phase are com-posed of eight bright spots, corresponding to the tert-butyl groups. These molecules adopt a conformation in which the phenyl substituent groups and tetrapyrrole macro-cycle approach co planarity (upper domain in Fig. 4A). This phase was neighbored by a two-dimensional phase containing molecules of differing conformation. Rotation of the phenyl substituents to a low dihedral angle, in res ponse to surface adsorption, causes distortion of the te tra pyrrole core to a saddle shape (lower domain in Fig. 4A). A notable feature emerging from this is a dark line (between the two parallel protrusions) and this can be used to observe the relative orientation of the molecules within a surface-adsorbed structure. In this domain, the TDtBHPP molecules are arranged in a distorted square packing. There is an additional but slight irregularity in the structure, which may be due to the small size of this two-dimensional domain. Orientation of the molecules

Fig. 3. STM images of TDtBHPP at submonolayer coverage on Cu(111) surface with phase transition from hexagonal phase to grid phase upon temperature increase. (B) STM image of submonolayer of Ox(TDtBHPP). Reprint with permission from Hill JP et al.Chem. Commun. 2006; 2320–2322. © 2006, Royal Society of Chemistry



with respect to the ma crocyclic saddling (and resulting dark lines) is uniform within the unusual assembly with the “saddle line” coinciding with a mirror plane of the deformed TDtBHPP molecule. This uniformity is a direct result of the conformation of the meso-phenyl substi-tuents and is due to steric factors.

Surprising features were revealed by further ins pection of the phase boundary structure. There is an apparent “mixed” conformation of the molecules at the periphery of the assembly (Fig. 4B). For those molecules, phenyl substituents at the interior of the phase domain are likely to have similar dihedral angles (i.e. ca 45º) while those at the exterior appear to have much smaller dihedral angles between them and the porphyrin macrocycle. One signifi -cant feature of the substituent mixed conformation of the porphy rin molecule is that it further permits its assem-blies to exist in contact with other supramolecular assem-blies (of differing domain structure) and this is illustrated in Fig. 4B together with space-fi lling representations of the three differing molecular conformations. The boun-dary region is characterized by a highly ordered struc-ture, which is due solely to the existence of the mi xed molecular conformation of TDtBHPP. In the present case, a lattice match between two non-iden tical phases is not achieved by an interlock but by the creation of a confor-mationally unsymmetrical mo le cule, which exists at the interface between two dif fering domain structures.

Hydrogen-bonded network array

In the results described in the above sections, hydro-gen bonding is obstructed by the presence of tert-butyl

sub stituents so that the interaction does not affect the structure in the crystal or surface-adsor bed structures. As the next target, structural possibili ties of allowing hydrogen bonding were investigated u sing porphyrin substituted with 3,5-dimethyl-4-hydro xy phenyl groups, (5,10,15,20-tetrakis(3,5-dimethyl-4-hy dro xyphenyl)porphyrin) [71].

When the methyl-substituted analogue is adsor bed atthe Cu(111) surface, the molecules present a topo graphy that can be interpreted, with reference to the molecular structure (from X-ray crystallography), as shown in Fig. 5A. The relative positioning of methyl groups observed in the STM images can be used to infer the conforma-tion of the methyl groups that are in contact with the surface. The methyl groups in contact with the surface are likely to be positioned at lattice vertices. Intermole-cular interactions occur at the meso substituents and are presumed to be hydro gen bonds because of the strength of these interactions present in the crystal structure. At submonola yer coverage, there is a pronounced tendency for the molecules to form a trimer unit through interac-tions at the meso substituents. In addition, more than one possible trimer unit structure is possible because of the asymmetry of the molecular conformation so that self- assembly of this unit into extended structures is interrup-ted. Both macrocyclic and ring-open forms of the trimer are commonly observed, as shown in Fig. 5B. The trime-ric units represent a potential sub unit of larger structures especially in macrocyclic form, which could be a vertex of an extended hexago nal structure. In fact, a hexagonal structure could occa sionally be observed. This represents

Fig. 4. (A) STM image of boundary region of two different domains of TDtBHPP on Cu(111) surface: molecules in coplanar shape (upper domain); molecules with saddle shape (lower domain). (B) STM image and models of TDtBHPP molecule at upper domain (a), boundary region (b), and lower domain (c). Reprint with permission from Hill JP et al. Phys. Chem. Chem. Phys. 2006; 8:5034–5037. © 2006, Royal Society of Chemistry


30 K. ARIGA ET AL.

the sim plest ca se for a hexagonal structure built from the trimer having C3 symmetry and is a porous honeycomb-type structure as shown in Fig. 5C, where it can be seen as a Kagomé structure containing void areas within its lattice. Larger areas of apparently unoccupied subs tratehave an intermediate contrast when com pa red to the areasbetween molecular junctions or un occupied areas contai-ned within the Kagomé lattice. This intermediate contrast is due to a substrate-bound “gas phase” of molecules that have not been incorporated into the self-assembly. An absence of gas-phase mo lecules in the enclosed areas by Kagomé lattice was indicated by the high contrast of the STM image in those regions.

FUTURE PERSPECTIVES

In this microreview, we briefl y introduced recent pro-gress of research on porphyrin assemblies, especially

molecular arrays on a two-dimensional surface. Because porphyrins, phthalocyanines, and their derivatives have attractive functions, a considerable number of papers on this subject are available. In addition, organic chemistry for porphyrin derivatives is well-developed, and gives us opportunities to modi fy them in the most appropriate fashion according to application. As seen in the examples of oxoporphyrinogen molecular arrays, molecular struc-tures and conformations are crucial keys for controlling molecular arrangement. Such fi ne nanostructures can be regulated by appropriate molecular design of the por-phyrin derivatives. Therefore, porphyrins are appropriate structural units with which we can pursue the science andtechnology of self-assembly.

Figure 6 summarizes strategies to construct functional materials from molecular units. Assembling processes from molecules to materials can be classifi ed into three phases; (i) molecule to nano; (ii) nano to micro; (iii) micro

Fig. 5. (A) Formulae and structure in crystal of 5,10,15,20-tetrakis (3,5-dimethyl-4-hydroxyphenyl)porphyrin. (B) STM image and model of cycric trimer (a) and open trimer (b). (C) STM image and model of its Kagomé lattice. Reprint with permission from HillJP et al. J. Phys. Chem. C 2007; 111: 16174–16180. © 2007, American Chemical Society



to macro (bulk). In the phase from molecule to nano, high precision regulation of molecular arrangement is a key issue [72]. Forma tion of microstructures from nano-level self-assem bly phe nomena and designed-assembly pro-cesses plays another important role [73]. Several tech-niques such as structure transcription and template syn-theses are often used in order to fabricate bulk functional structures from microscopic molecular assemblies [74]. As illustrated by some of the examples in this microreview, porphyrins and their derivatives make a si gni fi cant contribution to all these areas. In addition, me thodologies to bridge the molecular world and bulk systems has received much attention because controlling molecular phenomena by easily accessible for ces and mechanisms should lead to development of molecular technologies [75]. Advanced concepts such as molecular manipulation and molecular machines are contained in this category. Porphyrin derivatives also contribute to these concepts, where the accumula ted synthetic knowledge allows us to construct well-desi gned porphyrin structures appro-priate for machi nes and manipulators [76]. Therefore, porphyrins are key materials in such bottom-up technolo-gies and will ine vitably be developed further in the near future.

Acknowledgements

This research was supported by Grant-in-Aid for Scientifi c Research on Priority Areas “Chemistry of Coordination Space” and a Grant-in-Aid for Science Research in a Priority Area “Super-Hierarchical Struc-tures” from Ministry of Education, Science, Sports and Culture, Japan, World Premier International Re search Center Initiative (WPI Initiative) on Mate rials Nanoar-chitronics, from Ministry of Education, Science, Sports and Culture, Japan, and Grants-in-Aid for Scientifi c Research (B) from Japan Society for the Promotion of Science.

REFERENCES

a) Gryko DT, Fox JP and Goldberg DP. 1. J. Por phyrins Phthalocyanines 2004; 8: 1091–1105. b) Srinivasan A and Furuta H. Acc. Chem. Res. 2005; 38: 10–20. c) Elemans JAAW, Van Hameren R, Nolte RJM and Rowan AE. Adv. Mater. 2006; 18: 1251–1266. d) Xie Y, Hill JP, Charvet R and Ariga K. J. Nanosci. Nanotech nol. 2007; 7: 2969–2993.Fukuzumi S and Kashiwagi, Y.2. J. Porphyrins Phthalocyanines 2007; 11: 368–374.a) Guldi DM. 3. Chem. Soc. Rev. 2002; 31; 22–36. b) Boyd PDW and Reed CA. Acc. Chem. Res. 2005; 38: 235–242. c) Imahori H. J. Mater. Chem. 2007; 17: 31–41. d) Eu S, Hayashi S, Urneyama T, Ma-tano Y, Araki Y and Imahori H. J. Phys. Chem. C2008; 112: 4396–4405.a) Gust D, Moore TA and Moore AL. 4. Acc. Chem. Res. 2001; 34: 40–48. b) Choi M-S, Yamazaki T, Yamazaki I and Aida T. Angew. Chem. Int. Ed.2004; 43: 150–158. c) Imahori H. J. Phys. Chem. B2004; 108: 6130–6143.Solladie N, Rein R and Walther M.5. J. Porphyrins Phthalocyanines 2007; 11: 375–382.Umeyama T and Imahori H. 6. Photosynth. Res. 2006; 87: 63–71.a) Cha X, Ariga K, Onda M and Kunitake T. 7. J. Am. Chem. Soc. 1995; 117: 11833–11838. b) Cha X, Ariga K and Kunitake T. J. Am. Chem. Soc. 1996; 118: 9545–9551. c) Ariga K, Kunitake T and Furuta H. J. Chem. Soc., Perkin Trans. 2 1996: 667–672. d) Marchi-Artzner V, Artzner F, Karthaus O, Shi-momura M, Ariga K, Kunitake T and Lehn J-M. Langmuir 1998; 14: 5164–5171. e) Ariga K and Kunitake T. Acc. Chem. Res. 1998; 31: 371–378. f) Ariga K, Kamino A, Cha X and Kunitake T. Lang-muir 1999; 15: 3875–3885. g) Ariga K, NakanishiT, Hill JP, Shirai M, Okuno M, Abe T and Kikuchi

Fig. 6. Various strategies for material fabrication from mo lecules


32 K. ARIGA ET AL.

J. J. Am. Chem. Soc. 2005; 127: 12074–12080. h) Wang T and Liu M. Soft Matter 2008; 4: 775–783. i) Ariga K, Michinobu T, Nakanishi T and Hill JP. Curr. Opin. Colloid Interface Sci. 2008; 13: 23–30.a) Lvov Y, Ariga K and Kunitake T. 8. Chem. Lett.1994; 2323–2326. b) Lvov Y, Ariga K, Ichinose I and Kunitake T. J. Am. Chem. Soc. 1995; 117:6117–6123. c) Onda M, Lvov Y, Ariga K and Ku-nitake T. Biotechnol. Bioeng. 1996; 51: 163–167. d) Onda M, Lvov Y, Ariga K and Kunitake T. J. Fer-ment. Bioeng. 1996; 82: 502–506. e) Lvov Y, Ariga K, Ichinose I and Kunitake T. Langmuir 1996; 12:3038–3044. f) Lvov Y, Ariga K, Onda M, Ichinose I and Kunitake T. Langmuir 1997; 13: 6195–6203. g) Ariga K, Lvov Y and Kunitake T. J. Am. Chem. Soc. 1997; 119: 2224–2231. h) Caruso F, Furlong DN, Ariga K, Ichinose I and Kunitake T. Langmuir1998; 14: 4559–4565. i) Lvov Y, Onda M, Ariga K and Kunitake T. J. Biomater. Sci., Polym. Ed. 1998; 9: 345–355. j) Onda M, Ariga K and Kunitake T. J. Biosci. Bioeng. 1999; 87: 69–75. k) Lvov Y, Ariga K, Onda M, Ichinose I and Kunitake T. Colloid Surf. A, Physicochem. Eng. Asp. 1999; 146: 337–346. l) Katagiri K, Hamasaki R, Ariga K and Kikuchi J. J. Am. Chem. Soc. 2002; 124: 7892–7893. m) Katagiri K, Hamasaki R, Ariga K and Kikuchi J. Langmuir2002; 18: 6709–6711. n) Ariga K, Hill JP and Ji Q. Phys. Chem. Chem. Phys. 2007; 9: 2319–2340. o) Ariga K, Hill JP and Ji Q. Macromol. Biosci. 2008; 8: 981–990.Yui T, Kameyama T, Sasaki T, Torimoto T and 9. Takagi K. J. Porphyrins Phthalocyanines 2007; 11:428–433.a) Oishi Y, Torii Y, Kuramori M, Suehiro K, Ariga K, 10.Taguchi K, Kamino A and Kunitake T. Chem. Lett.1996: 411–412. b) Koyano H, Yoshihara K, Ariga K, Kunitake T, Oishi Y, Kawano O, Kuramori M and Suehiro K. Chem. Commun. 1996; 1769–1770. c) Oishi Y, Kato T, Kuramori M, Suehiro K, Ariga K, Kamino A, Koyano H and Kunitake T. Chem.Commun. 1997; 1357–1358. d) Oishi Y, Torii Y, Kato T, Kuramori M, Suehiro K, Ariga K, Taguchi K, Kamino A, Koyano H and Kunitake T. Langmuir1997; 13: 519–524. e) Ariga K, Kikuchi J, Naito M, Koyama E and Yamada N. Langmuir 2000; 16:4929–4939. f) Ariga K. J. Nanosci. Nanotechnol.2004; 4: 23–34. g) Oishi Y, Kato T, Narita T, Ariga K and Kunitake T. Langmuir 2008; 24: 1682–1685. h) Ariga K, Hill JP and Wakayama Y. Phys. Stat. Sol. A 2008; 205: 1249–1257. i) Ariga K. J. Photo-polym. Sci. Technol. 2008; 21: 553–558.a) Offord DA, Sachs SB, Ennis MS, Eberspacher 11.TA, Griffi n JH, Chidsey CED and Collman JP. J. Am. Chem. Soc. 1998; 120: 4478–4487. b) Krie-gisch V and Lambert C. Supramol. Dye Chem. 2005; 358; 257–313. c) Konishi T, Ikeda A and Shinkai S. Tetrahedron 2005; 61: 4881–4899.

Arima V, Blyth RLR, Matino F, Chiodo L, Della 12.Sala T, Thompson J, Regier T, Del Sole R, Mele G, Vasapollo G, Cingolani R and Rinaldi R. Small2008; 4: 497–506.a) Satake A and Kobuke Y. 13. Tetrahedron 2005; 61:13–41. b) Ariga K, Nakanishi T and Michinobu T. J. Nanosci. Nanotechnol. 2006; 6: 2278–2301. c) Ariga K, Nakanishi T and Hill JP. Curr. Opin. Col-loid Interface Sci. 2007; 12: 106–120. d) Ariga K, Hill JP, Lee MV, Vinu A, Charvet R and Acharya S. Sci. Technol. Adv. Mater. 2008; 9: 014109.a) Balaban TS. 14. Acc. Chem. Res. 2005; 38: 612–623. b) Maeda C, Kamada T, Aratani N and Osuka A. Coord. Chem. Rev. 2007; 251: 2743–2752. c) Chitta R and D’Souza F. J. Mater. Chem. 2008; 18: 1440–1471.a) Kim D and Osuka A. 15. Acc. Chem. Res. 2004; 37:735–745. b) Nakamura Y, Aratani N and Osuka A. Chem. Soc. Rev. 2007; 36: 831–845.a) Aratani N, Osuka A, Kim YH, Jeong D and Kim 16.D. Angew. Chem. Int. Ed. 2000; 39:1458–1462. b) Tsuda A and Osuka A. Science 2001; 293: 79–82.Hori T, Peng X, Aratani N, Takagi A, Matsumoto T, 17.Kawai T, Yoon ZS, Yoon M-C, Yang J, Kim D and Osuka A. Chem. Eur. J. 2008; 14: 582–595.Winters MU, Kärnbratt J, Eng M, Wilson CJ, An-18.derson HL and Albinsson B. J. Phys. Chem. C 2007; 111: 7192–7199.Sasaki M, Shibano Y, Tsuji H, Araki Y, Tamao K and 19.Ito O. J. Phys. Chem. A 2007; 111: 2973–2979.Cramariuc O, Hukka TI, Rantala TT and Lemmetyi-20.nen H. J. Phys. Chem. A 2006; 110: 12470–12476.Jiang D-L and Aida T. 21. Nature 1997; 388: 454–456.Kobuke Y. 22. Eur. J. Inorg. Chem. 2006; 2333–2351.Iengo E, Zangrando E and Alessio E. 23. Acc. Chem. Res. 2006; 39: 841–851.a) Ogawa K, Zhang T, Yoshihara K and Kobuke Y. 24.J. Am. Chem. Soc. 2002; 124: 22–23. b) Ogawa K, Ohashi A, Kobuke Y, Kamada K and Ohta K. J. Am. Chem. Soc. 2003; 125: 13356–13357. c) Ikeda C, Satake A and Kobuke Y. Org. Lett. 2003; 5: 4935–4938. d) Kuramochi Y, Satake A and Kobuke Y. J. Am. Chem. Soc. 2004; 126: 8668–8669.a) Taylor PN and Anderson HL. 25. J. Am. Chem. Soc. 1999; 121: 11538–11545. b) Anderson HL. Chem.Commun. 1999; 2323–2330. c) Screen TEO. Thor-ne JRG, Denning RG, Bucknall DG and Anderson HL J. Mater. Chem. 2003; 13: 2796–2808.a) Camara-Campos A, Hunter CA and Tomas S. 26.Proc. Natl. Acad. Sci. USA 2006; 103: 3034–3038. b) Hunter CA and Tomas S. J. Am. Chem. Soc. 2006; 128: 8975–8979. c) Ballester P, Oliva AI, Costa A, Deya PM, Frontera A, Gomila RM and Hunter CA. J. Am. Chem. Soc. 2006; 128: 5560–5569.Elemans JAAW, Lensen MC, Gerritsen JW, van 27.Kempen H, Speller S, Nolte RJM and Rowan AE.Adv. Mater. 2003; 15: 2070–2073.



Lee SJ, Mulfort KL, O’Donnell JL, Zuo X, Goshe 28.AJ, Wesson PJ, Nguyen ST, Hupp JT and Tiede DM. Chem. Commun. 2006: 4581–4583.D’Souza F, Chitta R, Gadde S, Rogers LM, Karr 29.PA, Zandler ME, Sandanayaka ASD, Araki Y and Ito O. Chem. Eur. J. 2007; 13: 916–922.Ogawa K, Hara C and Kobuke Y.30. J. Porphyrins Phthalocyanines 2007; 11: 359–367.Sessler JL, Wang B and Harriman A. 31. J. Am. Chem. Soc. 1995; 117: 704–714.Ohkawa H, Takayama A, Nakajima S and Nishide 32.H. Org. Lett. 2006; 8: 2225–2228.Sugimoto T, Suzuki T, Shinkai S and Sada K. 33. J. Am. Chem. Soc. 2007; 129: 270–271.Gulino FG, Lauceri R, Frish L, Evan-Salem T, 34.Cohen Y, De Zorzi R, Geremia S, Di Costanzo L, Randaccio L, Sciotto D and Purrello R. Chem. Eur. J. 2006; 12: 2722–2729.Flamigni L, Talarico AM, Chambron J-C, Heitz V, 35.Linke M, Fujita N and Sauvage J-P. Chem. Eur. J.2004; 10: 2689–2699.Safaei E, Ranjbar B and Hasani L. 36. J. Porphyrins Phthalocyanines 2007; 11: 805–814.Jacob DS, Mallenahalli S, Gedanken A, Solovyov 37.LA, Xenogiannopoulou E, Iliopoulos K and Couris S. J. Porphyrins Phthalocyanines 2007; 11: 713–718.a) Kobayashi K, Koyanagi M, Endo K, Masuda H 38.and Aoyama Y. Chem. Eur. J. 1998; 4: 417–424. b) Matsuura K, Ariga K, Endo K, Aoyama Y and Oka-hata Y. Chem. Eur. J. 2000; 6: 1750–1756.Harada R and Kojima T. 39. Chem. Commun. 2005;716–718.Kojima T, Harada R, Nakanishi T, Kaneko K and 40.Fukuzumi S. Chem. Mater. 2007; 19: 51–58.Li Y, Wang T and Liu M. 41. Soft Matter 2007; 3: 1312–1317.a) Ishi-I T, Jung J H and Shinkai S. 42. J. Mater. Chem.2000; 10: 2238–2240. b) Tamaru S, Uchino S, Ta-keuchi M, Ikeda M, Hatano T and Shinkai S. Te-trahedron Lett. 2002; 43: 3751–3755. c) Kishida T, Fujita N, Sada K and Shinkai S. Chem. Lett. 2004;33: 1002–1003.Wang Z, Li Z, Medforth C J and Shelnutt JA. 43. J. Am. Chem. Soc. 2007; 129: 2440–2441.Wang Z, Ho KJ, Medforth CJ and Shelnutt JA. 44. Adv. Mater. 2006; 18: 2557–2560.a) Gong X, Milic T, Xu C, Batteas JD and Drain 45.CM. J. Am. Chem. Soc. 2002; 124: 14290–14291. b) Drain CM, Smeureanu G, Patel S, Gong X, Garnod J and Arijeloyea J. New J. Chem. 2006; 30:1834–1843.Charvet R, Jiang D-L and Aida T. 46. Chem. Commun.2004; 2664–2665.a) Yoshimoto S and Itaya K. 47. J. Porphyrins Phtha-locyanines 2007; 11: 313–333. b) Yoshimoto S. Bull. Chem. Soc. Jpn 2006; 79: 1167–1190.

Otsuki J, Nagamine E, Kondo T, Iwasaki K, Asa-48.kawa M and Miyake K. J. Am. Chem. Soc. 2005;127: 10400–10405.Otsuki J, Kawaguchi S, Yamakawa T, Asakawa M 49.and Miyake K. Langmuir 2006; 22: 5708–5715.Kiebele A, Bonifazi D, Cheng F, Stöhr M, Diede-50.rich F, Jung T and Spillmann H. ChemPhysChem2006; 7: 1462–1470.Sekiguchi T, Wakayama Y, Yokoyama S, Kamikado 51.T and Mashiko S. Thin Solid Films 2004; 464/465:393–397.a) Hulsken B, van Hameren R, Thordarson P, Ger-52.ritsen JW, Nolte RJM, Rowan AE, Crossley MJ, Elemans JAAW and Speller S. Jpn. J. Appl. Phys.2006; 45: 1953–1955. b) Lensen MC, Elemans JAAW, van Dingenen SJT, Gerritsen JW, Speller S, Rowan AE and Nolte RJM. Chem. Eur. J. 2007; 13:7948–7956.Katsonis N, Vicario J, Kudernac T, Visser J, Pollard 53.MM and Feringa BL. J. Am. Chem. Soc. 2006; 128:15537–15541.Auwärter W, Alexander Weber-Bargioni A, Brink 54.S, Riemann A, Schiffrin A, Ruben M and Barth JV. ChemPhysChem 2007; 8: 250–254.Yoshimoto S, Saito A, Tsutsumi E, D’Souza F, 55.Ito O and Itaya K. Langmuir 2004; 20: 11046–11052.Yoshimoto S, Sato K, Sugawara S, Chen Y, Ito O, 56.Sawaguchi T, Niwa O and Itaya K. Langmuir 2007;23: 809–816.Hulsken B, Van Hameren R, Gerritsen JW, Khoury 57.T, Thordarson P, Crossley MJ, Rowan AE, Nolte RJM, Elemans JAAW and Speller S. Nat. Nano-technol. 2007; 2: 285–289.Yokoyama T, Yokoyama S, Kamikado T, Okuno Y 58.and Mashiko S. Nature 2001; 413: 619–621.Nishiyama F, Yokoyama T, Kamikado T, Yokoya-59.ma S and Mashiko S. Appl. Phys. Lett. 2006; 88:253113.Bonifazi D, Kiebele A, Stöhr M, Cheng F, Jung T, 60.Diederich F and Spillmann H. Adv. Funct. Mater. 2007; 17: 1051–1062.Yoshimoto S, Honda Y, Ito O and Itaya K. 61. J. Am. Chem. Soc. 2008; 130: 1085–1092.Barrena E, de Oteyza DG, Dosch H and Wakayama 62.Y. ChemPhysChem 2007; 8: 1915–1918.a) Hill JP, Hewitt IJ, Anson CE, Powell AK, McCarty 63.AL, Karr PA, Zandler ME and D’Souza F. J. Org. Chem. 2004; 69: 5861–5869. b) Schumacher AL, Hill JP, Ariga K and D’Souza F. Elec tro chem. Com-mun. 2007; 9: 2751–2754.Hill JP, Schmitt W, McCarty AL, Ariga K 64.and D’Souza F. Eur. J. Org. Chem. 2005; 2893–2902.Hill JP, Ariga K, Schumacher AL, Karr PA and 65.D’Souza F. J. Porphyrins Phthalocyanines 2007;11: 390–396.


34 K. ARIGA ET AL.

Hill JP, Sandanayaka ASD, McCarty AL, Karr PA, 66.Zandler ME, Charvet R, Ariga K, Araki Y, Ito O and D’Souza F. Eur. J. Org. Chem. 2006; 595–603.Xie Y, Hill JP, Schumacher AL, Sandanayaka ASD, 67.Araki Y, Karr PA, Labuta J, D’Souza F, Ito O, Anson CE, Powell AK and Ariga K. J. Phys. Chem. C 2008; 112: 10559–10572.Xie Y, Hill JP, Schumacher AL, Karr PA, D’Souza 68.F, Anson CE, Powell AK and Ariga K. Chem. Eur. J. 2007; 13: 9824–9833.a) Hill JP, Wakayama Y, Schmitt W, Tsuruoka T, 69.Nakanishi T, Zandler ML, McCarty AL, D’Souza F, Milgrom LR and Ariga K. Chem. Commun. 2006; 2320–2322. b) Wakayama Y, Hill JP and Ariga K. Surf. Sci. 2007; 601: 3984–3987.Hill JP, Wakayam Y and Ariga K.70. Phys. Chem. Chem. Phys. 2006; 8: 5034–5037.Hill JP, Wakayama Y, Akada M and Ariga K. 71. J. Phys. Chem. C 2007; 111: 16174–16180.a) Nakanishi T, Miyashita N,72. Michinobu T, Wakaya-ma Y, Tsuruoka T, Ariga K and Kurth DG. J. Am. Chem. Soc. 2006; 128: 6328–6329. b) Nakanishi T, Takahashi H, Michinobu T, Takeuchi M, Teranishi T and Ariga K. Colloid Surf. A Physicochem. Eng. Asp. 2008; 321: 99–105.a) Nakanishi T, Schmitt W,73. Michinobu T, Kurth DG and Ariga K. Chem. Commun. 2005; 5982–5984. b) Nakanishi T, Ariga K, Michinobu T, Yoshida K, Takahashi H, Teranishi T, Möhwald H and Kurth DG. Small 2007; 3: 2019–2023. c) Nakanishi T, Michi-nobu T, Yoshida K, Shirahata N, Ariga K, Möhwald H and Kurth DG. Adv. Mater. 2008; 20: 443–446. d) Ji Q, Miyahara M, Hill JP, Acharya S, Vinu A, Yoon SB, Yu JS, Sakamoto K and Ariga K. J. Am. Chem. Soc. 2008; 130: 2376–2377. e) Ariga K, Vinu A, Ji Q, Ohmori O, Hill JP, Acharya S, Koike J and Shiratori S. Angew. Chem. Int. Ed. 2008; 47: 7254–7257. f) Ji Q, Acharya S, Hill JP, Richards GJ and Ariga K. Adv. Mater. 2008; 20: 4027–4032.

a) Ariga K. 74. Chem. Rec. 2004; 3: 297–307. b) Vinu A, Ariga K, Mori T, Nakanishi T, Hishita S, Golberg D and Bando Y. Adv. Mater. 2005; 17: 1648–1652. c) Vinu A, Terrones M, Golberg D, Hishita S, Ariga K and Mori T. Chem. Mater. 2005; 17: 5887–5890. d) Vinu A, Miyahara M, Sivamurugan V, Mori T and Ariga K. J. Mater. Chem. 2005; 15: 5122–5127. e) Jin H, Liu Z, Ohsuna T, Terasaki O, Inoue Y, Sakamoto K, Nakanishi T, Ariga K and Che S. Adv. Mater. 2006; 18: 593–596. f) Vinu A, Miyahara M and Ariga K. J. Nanosci. Nanotechnol. 2006; 6:1510–1532. g) Michinobu T, Nakanishi T, Hill JP, Funabashi M and Ariga K. J. Am. Chem. Soc. 2006; 128: 10384–10385. h) Vinu A, Mori T and Ariga K. Sci. Technol. Adv. Mater. 2006; 7: 753–771. i) Ariga K, Vinu A, Miyahara M, Hill JP and Mori T. J. Am. Chem. Soc. 2007; 129: 11022–11023. j) Ariga K, Vinu A, Hill JP and Mori T. Coord. Chem. Rev. 2007; 251: 2562–2591. k) Hill JP, Alam S, Ariga K, Anson CE and Powell AK. Chem. Commun. 2008; 383–385. l) Acharya S, Gautam UK, Sasaki T, Bando Y, Golan Y and Ariga K. J. Am. Chem. Soc. 2008; 130:4594–4595. m) Hill JP, Palza H, Alam S, Ariga K, Schumacher A, D’Souza F, Anson C and Powell A. Inorg. Chem. 2008; 47: 8306–8314.a) Ariga K, Terasaka Y, Sakai D, Tsuji H and Kikuchi 75.J. J. Am. Chem. Soc. 2000; 122: 7835–7836. b) Ariga K, Nakanishi T, Terasaka Y, Tsuji H, Sakai D and Kukuchi J. Langmuir 2005; 21: 976–981. c) Ariga K, Nakanishi T, Hill JP, Terasaka Y, Sakai D and Kikuchi J. Soft Matter 2005; 1: 132–137. d) Ariga K, Nakanishi T and Hill JP. Soft Matter 2006; 2:465–477. e) Michinobu T, Shinoda S, Nakanishi T, Hill JP, Fujii K, Player TN, Tsukube H and Ariga K. J. Am. Chem. Soc. 2006; 128: 14478–14479.a) Kottas GS, Clarke LI, Horinek D and Michl J. 76.Chem. Rev. 2005; 105; 1281–1376. b) Kay ER, Leigh DA and Zerbetto F. Angew. Chem. Int. Ed.2007; 46: 72–191.




INTRODUCTION

The role of nitric oxide (NO) in mammalian physiolog-ical processes is well established. The importance of NO as a messenger in the cardiovascular, immune, and ner-vous systems has been described in the literature [1–3].

Ruthenium complexes with different types of ligands are well-known NO scavengers [4]. Several types of ruthenium nitrosyl complexes with porphyrins have been reported [5–9, 11]. These types of systems have served to develop NO sensors. Fluorescence and electron para-magnetic resonance (EPR) spectroscopy, amperometry, chemiluminescence and cyclic voltammetry are some of the techniques employed for the detection of NO [6].

Tsutsui et al. reported the fi rst ruthenium(II) complexes with natural porphyrins [carbonylmesopo-rphyrin and mesoporphyrin(IX)] [7, 8]. In our groups,

deuteroporphyrin(IX) dimethylester (DmDP) obtained from the red blood pigment heme has been used as the starting material for several synthetic modifi ca-tions, biological and chemical applications. Different derivatives of DmDP and the corresponding metallodeu-teroporphyrins have been reported [10]. Iron, cobalt and nickel porphyrin disulphide derivative complexes self-assembled on gold electrodes were electrochemically characterized [11].

In the present work, we report on the preparation of a new ruthenium(II) porphyrin disulphide derivative (Fig. 1) and its self-assembly on a gold electrode which can be applied to the electrochemical detection of NO.

EXPERIMENTAL

General

Reagents. All chemicals used were of analytical grade (Aldrich) and were used without further purifi cation. The solvents were dried before use. All reactions were

Interaction of nitric oxide with Ru(II) complexes of deuteroporphyrindimethylester derivatives

Rudy Martina, Roberto Cao*a, Ana M. Estevab and Franz-Peter Montforts*c

a Laboratory of Bioinorganic Chemistry, Faculty of Chemistry, University of Havana, La Havana 10400, Cubab Department of Analytical Chemistry, Faculty of Chemistry, University of Havana, La Havana 10400, Cubac Institute of Organic Chemistry, University of Bremen, Leobener Strasse NW 2, 28359, Bremen, Germany

Received 4 September 2008Accepted 24 September 2008

ABSTRACT: A new ruthenium(II) porphyrin disulphide derivative, [Ru(Pds)(CO)], was obtained from ruthenium(II)(carbonyl)deuteroporphyrin(IX), [Ru(DPdc)(CO)] and cystamine. The interaction of this complex with nitric oxide was studied spectrophotometrically and a bathochromic shift of the charge transfer band and considerable change in the α and β bands of the complex were observed. According to the IR spectrum, the product of this interaction is [Ru(DmDP)(NO+)(NO2

-)]. [Ru(Pds)(CO)] was then self-assembled on polycrystalline gold and characterized by X-ray photoelectron spectroscopy. [Ru(Pds)(CO)] was also self-assembled on gold electrode beads and its interaction with nitric oxide in aqueous solution was studied by cyclic voltammetry. A shift in the ruthenium redox process and a new irreversible cathodic peak at -0.59 V were observed, both indicating coordination of NO.

KEYWORDS: ruthenium porphyrin, nitric oxide, sensor, immobilization, gold.


*Correspondence to: Roberto Cao, email: [email protected], fax: +537 873-3502, tel: +537 879-2145 and Franz-Peter Montforts, email: [email protected], fax: +49 0421-218-3569, tel: +49 0421-218-3569


36 R. MARTIN ET AL.

carried out under argon. Deuteroporphyrin(IX) dimethy-lester (DmDP) was isolated and purifi ed as reported in the literature [9]. The X-ray photoelectron spectroscopic investigations (XPS) were performed on self-assembled monolayers (SAM) on gold. The SAMs were prepared on glass slides with a chromium layer coated with polycrys-talline gold. Prior to the self-assembling procedure the slides were evaporated.

Equipment. Electronic spectra were recorded with a Varian Cary spectrophotometer interfaced with a micro-computer for data acquisition. IR spectra were recorded with a Perkin-Elmer Paragon 500 FT-IR spectrometer with samples prepared as KBr tablets. 1H NMR spectra were obtained with a Bruker DPX-200 and AVANCE DRX-600 MHz spectrometer at 300 K; all chemical shifts were referenced to TMS lock signal. 30 mg of samples were dissolved in CD2Cl2 (0.5 mL) + d5-pyridine (5 μL). MS were obtained with a Finnigan MAT 8200 spectrometer [EI (70 eV) and ESI]. The X-ray photoelectron spectra were obtained on Surface Science Instruments X-probe and M-probe spectrometers using monochromatic Al-Kα

X-ray sources (hν) 1486.6 eV. The BE scales for the monolayers on gold were referenced by setting the Au4f7/2

BE to 84.0 eV.

Electrochemical measurements were carried out with a Yanaco P-900 apparatus coupled to a Graphtec WX1000 X-Y recorder using a standard three-holder cell. Gold beads, Pt and Ag/AgCl (0.3 M KCl) were used as work-ing, counter and reference electrodes, respectively. The supporting electrolyte was 1 M Na2SO4. All solutions were previously deoxygenated with argon.

Preparation of gold electrodes

The gold electrode was prepared by annealing the tip of a gold wire (99.99%) that had been cleaned overnight with nitric acid (68%). The formed gold bead was elec-trochemically pretreated in HClO4 (0.1 M) at 2.35 V until gold(III) oxide formation (red color) was observed, in

order to guarantee the formation of a Au{111} surface. The electrode was then cleaned with HCl (0.1 M) and thoroughly rinsed with doubly distilled water [12]. The effective area of the electrode (3.8 mm2) was calculated by the Randles-Sevcik equations using K3[Fe(CN)]6 as a model electroactive species.

Preparation of carbonyl[13,17-bis-(2-methoxy-carbonylethyl)-2,7,12,18-tetramethylporphyrinato] ru-thenium(II), [Ru(DmDP)(CO)]. Deuteroporphyrin(IX) dimethylester (DmDP) (100 mg, 0.18 mmol) was dis-solved in benzene (100 mL) and dodecacarbonyltriruthe-nium, (Ru3CO12), (115 mg, 0.18 mmol) was added. The mixture was stirred and refl uxed under argon for 48 h and then fi ltered and concentrated to dryness. The dark orange crude product was purifi ed chromatographically on alumina II-III with dichloromethane (CH2Cl2) to elute Ru3CO12 and DmDP and then with CH2Cl2:MeOH (15:1) to elute [Ru(DmDP)(CO)]. Yield: 63.5 mg (0.095 mmol, 53%). TLC (Alox, CH2Cl2:EtOAc, 10:1): Rf = 0.42. UV-vis (CH2Cl2): λmax, nm (ε, M-1·cm-1) 306 (17000), 390 (180000), 514 (14000), 546 (26000). IR (KBr): ν, cm-1 2920 (w, CH), 2854 (w, CH), 1931 (s, C≡O),1736 (m, C=O), 1438 (w, δ(CH2)), 1323 (w, δ(CH3)),1269 (m), 1246 (m), 1160 (w), 1125 (m), 1032 (w), 849 (w), 751 (w). 1H NMR (200 MHz, CHCl3): δ, ppm 3.27 (m, 4H, J = 7.56 Hz CH2 α to carbonyl), 3.50 (s, 3H, CH3), 3.54 (s, 3H, CH3), 3.56 (s, 3H, CH3), 3.58 (s, 3H, CH3), 3.62 (s, 3H, CH3), 3.64 (s, 3H, CH3), 4.31 (m, 4H, CH2 β to carbonyl), 8.84 (s, 1H, 8-H), 8.89 (s, 1H, 3-H), 9.71 (s, 1H, 10-H), 9.81 (s, 1H, 5-H), 9.84 (s, 1H, 15-H), 10.00 (s, 1H, 20-H). MS (ESI, CH2Cl2:MeOH) positive: m/z 689 [M + Na+], negative: m/z 701 [M + Cl-].

Preparation of carbonyl[13,17-bis(propanoic acid)-2,7,12,18-tetramethylporphyrinato]ruthenium(II), [ruthenium(II)(carbonyl)deuteroporphyrin(IX)], [Ru(DPdc)(CO)]. [Ru(DmDP)(CO)] (40 mg, 0.06 mmol) was dissolved in dry THF (25 mL). A 5 M KOH solution (45 mL) was added and the mixture was stirred and refl uxed under an argon atmosphere for 18 h.

Fig. 1. Preparation of ruthenium(II) porphyrin disulphide derivative [Ru(Pds)(CO)]

NH N

N NH

CH3

CH3

CH3

CH3

CO2CH

3CO

2CH

3

N N

N N

CH3

CH3

CH3

CH3

Ru(II)

CO

NH OO NH

S S

1. Ru3CO

12, benzene, reflux (53%)

2. KOH/H2O/THF, reflux (96%)

3. a) ClCO2iBu, NEt

3, THF, -15 °C

b) cystamine, THF, rt (41%)


INTERACTION OF NITRIC OXIDE WITH RU(II) COMPLEXES 37

After cooling, the organic phase was separated. The aque-ous phase was mixed with pH 4 phosphate buffer solution (50 mL) and neutralized (pH 3–4) with 5 M HCl. The resulting dark solution was extracted with CH2Cl2. The resulting organic solution was evaporated and dried in vacuo overnight. Yield: 36.74 mg (0.058 mmol, 96%). TLC (silica gel, CH2Cl2:MeOH, 5:1): Rf = 0.41. UV-vis (CH2Cl2): λmax, nm (ε, M-1·cm-1) 306 (10000), 390 (95000), 514 (6800), 546 (11000). IR (KBr): ν, cm-1 3478 (m, OH), 2923 (s, CH), 2854 (s, CH), 1921 (s, C≡O), 1732 (m, C=O), 1463 (w, δ(CH2)), 1323 (w), 1367 (w), 1261 (m), 1098 (w), 1032 (w), 800 (w). 1H NMR: (200 MHz, CHCl3 + 5 μL d5-pyridine): δ, ppm 3.27 (m, 4H, CH2 αto carbonyl), 3.50 (s, 3H, CH3), 3.54 (s, 3H, CH3), 3.62 (s, 6H, CH3), 4.36 (m, 4H, CH2 β to carbonyl), 8.84 (s, 1H, 8-H), 8.89 (s, 1H, 3-H), 9.71 (s, 1H, 10-H), 9.81 (s, 1H, 5-H), 9.84 (s, 1H, 15-H), 10.00 (s, 1H, 20-H). MS (ESI, CH2Cl2:MeOH) positive: m/z 610 [M-C≡O]+, nega-tive: 637 [M]-.

Preparation of ruthenium(II) carbonyl porphyrin disulphide derivative, [Ru(Pds)(CO)]. Ru(DPdc)(CO)(40 mg, 0.054 mmol) was dissolved under argon in dry THF (25 mL) and freshly distilled triethylamine (0.6 mL) was added. The resulting dispersion was stirred under cooling (-15 °C) and a solution (0.136 mL) of isobutyl-chloroformate (2.5 mL, 0.054 mmol) in THF (10 mL) was added. The reaction mixture was stirred for 2 h and controlled by TLC. The temperature was increased to -10 °C and a 0.034 M solution of cystamine in THF (1.58 mL, 0.054 mmol) was slowly added dropwise over 1 h. The reaction mixture (under argon) was stirred at room temperature for 4 h. The resulting reaction mixture was extracted with CH2Cl2 and washed with saturated solutions of NaHCO3 and NaCl, and fi nally with water. The organic phase was evaporated and the crude brown solid was purifi ed by column chromatography (silica, CH2Cl2:MeOH, 100:1). Yield: 16.7 mg (0.022 mmol, 41%), mp: 296 °C. TLC (silica gel, CH2Cl2 :MeOH,10:1): Rf = 0.59. UV-vis (CH2Cl2): λmax, nm (ε, M-1·cm-1)314 (13800), 394 (150000), 516 (10500), 546 (14400). IR (KBr): ν, cm-1 3396 (m, NH), 2919 (s, CH), 2855 (s, CH), 1933 (s, C≡O), 1645 (m, C=O, amide), 1538 (m, δ(NH)), 1436 (w, δ(CH2)), 1246 (m), 1128 (w), 1031 (w), 1032 (w), 849 (w). 1H NMR (600 MHz, CD2Cl2 + 5 μL d5-pyridine): δ, ppm 2.34–2.38 (m, 4H, CH2 cys), 2.89–3.08 (m, 4H, CH2 cys), 3.07–3.17 (m, 4H, CH2 βto carbonyl), 3.20 (s, 3H, CH3), 3.26 (s, 3H, CH3), 3.37 (s, 3H, CH3), 3.42 (s, 3H, CH3), 4.03 (q, 2H, β-CH2 α to carbonyl), 4.22 (q, 2H, α-CH2 α to carbonyl), 8.27 (t, 2H, NH), 8.66 (s, 1H, 8-H), 8.67 (s, 1H, 3-H), 9.54 (s, 1H, 10-H), 9.59 (s, 1H, 5-H), 9.66 (s, 1H, 20-H), 10.12 (s, 1H, 15-H). MS (ESI, CH2Cl2: MeOH) positive: m/z777 [M + Na]+, negative: m/z 789 [M + Cl-].

Synthesis of {(nitrosyl)(nitrite)[13,17-bis-(2-methoxycarbonylethyl)-2,7,12,18-tetramethylpor-phyrinato] ruthenium(II)}, [Ru(DmDP)(NO)(NO2)].Nitric oxide (NO) was obtained by reaction of nitric acid

(5 M) and copper under an argon atmosphere using the Schlenk technique at 293 K. The gaseous products of this reaction were passed through NaOH (12 M) traps in order to remove undesirable products (nitrogen oxides other than NO). The generated NO was bubbled through the reaction fl ask containing Ru(DmDP)(CO) (10 mg) dissolved in CH2Cl2 (10 mL) for 20 min. The reaction fl ask was sealed and stirred for 2 h at room tempera-ture. The dark green solution was evaporated and dried under vacuum overnight. UV-vis (CH2Cl2): λmax, nm (ε, M-1·cm-1) 345 (sh), 390 (180000), 514 (14000), 546 (26000). IR (KBr): ν, cm-1 2948 (w, CH), 2880 (w, CH), 1854 (s, N≡O), 1731 (m, C=O), 1436 (w, δ(CH2)), 1384 (s, νas (NO2

-)), 1323 (w, δ(CH3)), 1271 (m), 1246 (m), 1169 (w), 1131 (m), 1040 (w), 849 (w), 725 (w). 1HNMR (200 MHz, CHCl3): δ, ppm 3.35 (m, 4H, J = 7.56 Hz CH2 α to carbonyl), 3.56 (s, 3H, CH3), 3.66 (s, 3H, CH3), 3.69 (s, 3H, CH3), 3.72 (s, 3H, CH3), 3.75 (s, 3H, CH3), 3.79 (s, 3H, CH3), 4.50 (m, 4H, CH2 β to carbo-nyl), 9.15 (s, 1H, 8-H), 9.21 (s, 1H, 3-H), 10.43 (s, 1H, 10-H), 10.47 (s, 1H, 5-H), 10.52 (s, 1H, 15-H), 10.57 (s, 1H, 20-H). MS (ESI, CH2Cl2:MeOH) positive: m/z668 [M + NO+].


For the preparation of Ru(DmDP)CO three different solvents were studied, giving 14% yield (THF), 23% (decaline) and 57% (benzene). The latter two solvents gave the best yields for the complexation which required a reaction time of 48 h.

The UV-vis spectrum of Ru(DmDP)(CO) in CH2Cl2

is in agreement with that reported for ruthenium(II) mes-oporphyrin complexes [7]. The conversion of the four Q-bands into α and β bands at 548 (log ε = 4.41) and 516 nm (log ε = 4.14), respectively, and the presence of a new band at 306 nm (log ε = 4.13), corresponding to a charge transfer, confi rm the complexation. The insertion of ruthenium into DmDP is also confi rmed by the pres-ence of an IR band at 1931 cm-1, corresponding to νC≡O

of the carbonyl group coordinated to the ruthenium metal center. The ESI (+) mass spectrum shows the expected molecular peak at m/z 689 corresponding to Ru(DmDP)(CO) + Na+.

Ru(DmDP)(CO) was reacted with NO for the substitu-tion of CO. The loss of the νCO band at 1931 cm-1 and the presence of a new band at 1854 cm-1 in the IR spectrum confi rmed the interchange of CO against NO. The posi-tion of the νNO band at 1854 cm-1 indicates a linear coor-dination of NO+ [4, 5]. A new band at 1384 cm-1 can be assigned to the asymmetric stretching (νas) of coordinated nitrite-N anion formed by a process described in the liter-ature [13]. Therefore, the presence of both bands, at 1854 and 1384 cm-1 indicates the formation of Ru(DmDP)(NO+)(NO2

-). A peak registered at m/z 713 in the ESI positive mass spectrum confi rms the formation of this complex.


38 R. MARTIN ET AL.

[Ru(Pds)(CO)] was prepared in order to use the disulfi de group to self-assemble the complex on gold surfaces. In order to confi rm that Ru(Pds)(CO) coordi-nates NO like Ru(DmDP)(CO), a dichloromethane solu-tion (0.05 mM) of the complex was studied spectrophoto-metrically after addition of 20 μL aliquots of NO solution (0.03 M) as depicted in Fig. 2. A red-shift of the charge transfer band to 345 nm was observed and appeared as a shoulder of the Soret-band. In addition, the Q-bands at 514 nm and 546 nm merged into a single broad band with a maximum at about 554 nm, a process which included the formation of three isosbestic points. This result is similar to that reported by Lorkovic and Ford for a syn-thetic porphyrin [14].

The self-assembly of Ru(Pds)(CO) on gold was per-formed by immersion (24 h) of freshly prepared poly-crystalline Au slides in a Ru(Pds)(CO) CH2Cl2 solution (1 mM) under an argon atmosphere. The slides were thoroughly rinsed with CH2Cl2, water and ethanol, dried in a gentle argon fl ow and characterized by X-ray pho-toelectron spectroscopy (XPS). The XPS measurements of the self-assembled monolayers (SAM) of Ru(Pds)(CO) showed a peak at 281.13 eV corresponding to Ru3d5/2.

Figure 3a depicts the XPS spectrum in the N(1s) region. Two peaks at 398.43 and 399.13 eV were observed, indicating the presence of two different nitrogen species. The peak at 398.43 eV is assigned to porphyrin nitro-gen atoms while the other peak corresponds to the amide nitrogen atoms [15]. The ratio of the peaks is 2:1, a result which is in accordance with the composition of nitrogens in Ru(Pds)(CO). A similar ratio was obtained from the XPS spectrum for O(1s) region (Fig. 3c).

The XPS spectrum of the sulfur (2p) region (Fig. 3b) show three peaks at 161.12, 162.01 and 163.36 eV. The peaks at 162.01 eV and 163.36 eV correspond to S2p3/2 and S2p1/2, respectively, indicating the self-assembled thiolate

groups are formed by breakdown of the disulfi de bond [12, 16]. The ratio of both peaks was 2:1, as expected.

The self-assembly of [Ru(Pds)(CO)] on gold beads was also characterized electrochemically. Previously,

Fig. 2. Titration of [Ru(Pds)(CO)] (0.05 mM) with 20 μL aliquots of NO (saturated in CHCl3)

Fig. 3. XPS spectrum of [Ru(Pds)(CO)] self-assembled on Au: (a) nitrogen (1s) region, (b) sulfur (2p) region and (c) oxygen (1s) region

396397398399400401

Inte

nsity n

orm

alize

d

E, eV

398.43

399.13

a

159160161162163164165166

Inte

nsity N

orm

alized

E, eV

162.01

163.36

161.12

b

528530532534536538

Inte

nsity N

orm

ala

ized

E, eV

533.55

531.81

c


INTERACTION OF NITRIC OXIDE WITH RU(II) COMPLEXES 39

a [Ru(Pds)(CO)] solution in CH3CN was studied. A typical cyclic voltammogram of this system (10-3 M) is depicted in Fig. 4. The voltammogram shows two peaks at 0.4 V and 0.68 V (ΔEp = 0.072 V), corresponding to the ruthenium redox pair which behave in a highly revers-ible manner, similar to that reported for a ruthenium(II) tetraphenylporphyrin (ΔEp = 0.084 V).

The cyclic voltammogram of [Ru(Pds)(CO)] self- assembled on a gold bead is shown in Fig. 5. The redox cou-ple of coordinated ruthenium seen at E1/2 = 0.32 V (ΔEp = 0.045 V) was found for the porphyrin ring. Comparison of theΔEp value of the self-assembled ruthenium complex with that of the ruthenium complex in CH3CN solution shows that the immobilized redox system is more reversible. This observation is typical for a surface-confi ned, redox system requiring no diffusion to the electrode surface [15].

The electrochemical reductive desorption of self- assembled [Ru(Pds)(CO)] was performed in order to

estimate the stability and compactness of the formed SAM. The cyclic voltammogram of reductive desorp-tion of the monolayer (0.5 M KOH, 0.1 V.s-1) is depicted in Fig. 6. The intense and irreversible cathodic peak at -0.92 V, corresponding to the reductive desorption of the complex, shows evidence for a stable and compact self-assembled Ru(Pds)(CO) monolayer.

The surface coverage of self-assembled [Ru(Pds)(CO)] was determined by integration of the desorption peak to give a value of 0.22 nmol.cm-2, similar to that reported for immobilized cobalt and iron porphyrin disulfi des [11, 17].

Once the [Ru(Pds)(CO)] SAM on Au{111} beads was characterized, this system was studied as a NO electro-chemical sensor. For this purpose, electrode beads modi-fi ed with [Ru(Pds)(CO)] SAM were introduced into a sealed standard three-cell holder containing a Na2SO4

solution (1 M). Gaseous NO was bubbled through this

Fig. 4. Cyclic voltammogram of bulk solution (1 mM) of [Ru(Pds)(CO)] in CH3 CN (–), bare gold (3.8 mm2) (---) (0.1 V.s-1)

Fig. 5. Cyclic voltammogram of [Ru(Pds)(CO)] SAM (–), bare gold bead (3.8 mm2) (---) in Na2SO4 (1 M) (0.05 V.s-1)

Fig. 6. Reductive desorption of [Ru(Pds)(CO)] self-assembled on gold bead in KOH (o.s.M) (3.8 mm2, 0.1 V.s-1)

Fig. 7. Cyclic voltammograms of [Ru(Pds)(CO)] self-assem-bled on gold bead (----). Bare gold electrode (·····) and [Ru(Pds)(CO)] self-assembled on gold bead (–) both bubbled with NO for 30 min (0 °C, Na2SO4 1M, 0.5 V.s-1). Inset: Zoom of that 0.0–0.6 V region


40 R. MARTIN ET AL.

solution for 30 min at 0 °C after which the NO fl ow was interrupted and the cyclic voltammogram then recorded (Fig. 7). The ruthenium redox pair was observed at E1/2 = 0.32 V, corresponding to a shift from 0.250 V. The observed shift is due to the electronic effect introduced by the coordination of NO, as reported for similar sys-tems [18, 19]. An additional irreversible cathodic peak at -0.59 V indicated the reduction of coordinated NO:

[Ru(Pds)(NO)(NO2)] + e- → [Ru(Pds)(NO)(NO2)]-

This peak was recorded within the range given for the reduction of NO in similar ruthenium nitrosyl com-pounds [19].

Acknowledgements

We thank the Deutsche Forschungsgemeinschaft (DFG) for fi nancial support of the project MO 274/24-1; “Synthesis, immobilization and electrochemical charac-terization of metalloporphyrinoids for catalysis, sensor-ing and light-induced electron transfer”. We also thank the University of Havana (Cuba) and the University of Bremen (Germany) for supporting the collaboration and Prof. Dr. Luis A. Montero Cabrera for his assistance in the preparation of the manuscript.

REFERENCES

Furchgott RF. 1. Angew. Chem. Int. Ed. Engl. 1999;38: 1870–1880.Ignarro LJ. 2. Angew. Chem. Int. Ed. Engl. 1999; 38:1882–1892.Palmer RM, Ferrige AG and Moncada S. 3. Nature1987; 327: 524–526.Ortiz M, Penabad A, Diaz A, Cao R, Otero A, An-4.tinnolo A and Lara A. Eur. J. Inorg. Chem. 2005;15: 3135–3140.

Ford PC and Laverman LE. 5. Coord. Chem. Rev. 2005; 249: 391–403.Lim MH and Lippard SJ. 6. Inorg. Chem. 2004; 43:6366–6370.Tsutsui M, Ostfeld D and Hoffman LM. 7. J. Am. Chem. Soc. 1971; 93: 1820–1823.Srivastava TS, Hoffman LM and Tsutsui M. 8. J. Am. Chem. Soc. 1972; 94: 1385–1386.Smith KM. (Ed.). 9. Porphyrins and Metalloporphy-rins, Elsevier: Amsterdam 1975.Wedel M, Walter A and Montforts FP. 10. Eur. J. Org. Chem. 2001; 9: 1681–1687. Cordas CM, Viana AS, Leupold S, Montforts FP 11.and Abrantes LM. Electrochem. Com. 2003; 5:36–41.Cao R. Jr, Díaz A, Cao R, Otero A, Cea R, Rodríguez-12.Arguelles MC and Serra C. J. Am. Chem. Soc. 2007; 129: 6927–6930. Miranda KM, Bu X, Lorkovic I and Ford PC. 13. Inorg. Chem. 1997; 36: 4838–4848.Lorkovic IM and Ford PC. 14. Inorg. Chem. 1999; 38:1467–1473.Offord DA, Sachs SB, Ennis MS, Eberspacher TA, 15.Griffi ng JH, Chidsey CED and Collman JP. J. Am. Chem. Soc. 1998; 120: 4478–4487.Love JC, Estroff LA, Kriebel JK, Nuzzo RG and 16.Whitesides GM. Chem. Rev. 2005; 105: 1103–1170.Viana AS, Leupold S, Montforts FP and Abrantes 17.LM. Electrochim Acta. 2005; 50: 2807–2813.Mori V, Toledo JC, Silva HAS, Franco DW and 18. Bertotti M. J. Electroanal. Chem. 2003; 547:9–15.Kadish KM, Adamian VA, Caemelbecke EV, 19.Tan Z, Tagliatesta P, Bianco P, Boschi T, Yi GB, Khan MA and Richter-Addo GB. Inorg. Chem.1996; 35: 1343–1348.




INTRODUCTION

Almost all naturally occurring chlorophyll (Chl) pigments have a propionate-type ester group at the 17- position [References 1–3 for comprehensive reviews and references cited therein]. The substituent is not directly attached to the conjugated π-system of Chl molecules and does not affect their electronic-absorption proper-ties. Therefore, it has attracted less attention than other

peripheral substituents, although it constitutes about 30% of the weight of a molecule. The hydrophobic propionate-type ester group is believed to play an important role in stabilizing complexation with peptides in photosynthetic apparatus. Moreover, it serves as an anchor to lock in (bacterio)chlorophyll (B)Chl molecules at the appropri-ate position in pigment-protein complexes.

Recent progress in crystallization of pigment-protein complexes enables us to determine the structure of apo-proteins as well as that of photosynthetic pigments at an atomic level [4]. However, the propionate-type ester group often showed low density of the diffraction in X-ray analysis, primarily due to its fl exibility. As a result, the structure and/or the detailed binding mechanism to photosynthetic peptides remained unclear.

Structural determination of the Δ2,10-phytadienylsubstituent in the 17-propionate of bacteriochlorophyll-bfrom Halorhodospira halochloris

Tadashi Mizoguchi*, Megumi Isaji, Jiro Harada, Kazuyuki Watabe and Hitoshi Tamiaki*

Department of Bioscience and Biotechnology, Faculty of Science and Engineering, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan

Received 7 September 2008Accepted 24 November 2008

ABSTRACT: Bacteriochlorophyll(BChl)-b, having a unique phytadienyl group in the propionate-type residue at the 17-position instead of a usual phytyl group, was isolated from the thermophilic purple photosynthetic bacterium Halorhodospira (Hlr.) halochloris. The structure of the propionate, especially for the positions of two C=C double bonds, was unambiguously determined to be C2=C3 and C10=C11 (Δ2,10-phytadienyl) by means of NMR spectroscopy. To confi rm the molecular structure by reverse-phase (RP) HPLC, two types of chlorophyll (Chl) derivatives, 3-acetyl-Chl-a, having different phytadienyl groups to the 17-propionate were prepared: one had a Δ2,10-phytadienyl group prepared by isomerization of the structurally determined BChl-b, and the other had a Δ2,14-phytadienyl by oxidation of BChl-a from the other purple bacterium Rhodopseudomonas sp. Rits as reported previously. HPLC analyses of these derivatives showed their distinct retention times under reverse-phase conditions; the Δ2,14-phytadienyl-type derivative was eluted more slowly than the Δ2,10-phytadienyl type. The results clearly indicated that the positions of two C=C double bonds in the ester group affected RP-HPLC elution, which directly refl ected the hydrophobicity in a molecule. RP-HPLC analyses thus serve as an aid for structural determination of (B)Chl molecules esterifi ed with various long hydrocarbon chains in photosynthetic organisms, and also enable estimation of their hydrophobicity and hydrophilicity.

KEYWORDS: bacteriochlorophyll-b, propionate, thermophilic purple bacteria, reverse-phase HPLC, photosynthesis.


*Correspondence to: Tadashi Mizoguchi, email: [email protected], fax: +81 77-561-2659, tel: +81 77-561-4959 and Hitoshi Tamiaki, email: [email protected], fax: +81 77-561-2659, tel: +81 77-561-2765


42 T. MIZOGUCHI ET AL.

In our previous study on the identifi cation of 17- propionate ester groups in BChls-a, we reported the pres-ence of unique BChls-a having Δ2,10,14-phytatrienyl and Δ2,14-phytadienyl groups in purple photosynthetic bacteria, and proposed the sequential hydrogenation of a geranylgeranyl (GG) to a phytyl (Phy) group at C6=C7 → C10=C11 → C14=C15 as shown in Scheme 1 [5]. There are two types of nomenclature for the 17-propi-onate in (B)Chl molecules. One is based on unsaturation of fully saturated “phytanyl”, the other on saturation of “geranylgeranyl” possessing four double bonds. In the present paper, the former nomenclature is used as in Δ2,10-phytadienyl (= 6,7,14,15-tetrahydrogeranylgera-nyl), except for generally accepted geranylgeranyl (= Δ2,6,10,14-phytatetraenyl) and phytyl (= Δ2-phytaenyl).

In addition to the above esters, the presence of other (B)Chls having different propionate ester groups has been reported in some photosynthetic bacteria [2]. For exam-ple, Scheer and his co-workers reported the presence of a Δ2,10-phytadienyl group (see the middle of Fig. 1b) in BChl-b isolated from a thermophilic purple bacterium, Halorhodospira (Hlr.) halochloris (previously named Ectothiorhodospira halochloris) [6]. The ester group was proposed on the basis of GC-MS analyses of its cleaved alcohol and the derivatives. It is necessary to determine unambiguously the structure of the Δ2,10-phytadienyl substituent in BChl-b to extend our knowledge of a variety of the propionate-type ester group in chlorophyl-lous pigments. Moreover, Kobayashi et al. reported the presence of Δ2,6-phytadienylated Chl-a isolated from a thermophilic green bacterium, Chlorobium (Chl.) tepidum(see also the top of Fig. 1b) [7]. Interestingly, both bacte-ria grew under thermophilic conditions, ~45 °C.

In this study, we isolated BChl-b which has a unique Δ2,10-phytadienyl group in the 17-propionate ester group, from Hlr. halochloris and determined its structure

unambiguously by its NMR and MS spectra, although the stereochemistry of the C7-chiral center in the ester group could not be determined. Furthermore, we demonstrated the effect of the positions of two C=C double bonds in phytadienyl-type ester groups upon reverse-phase (RP)-HPLC elution, since the order of elution directly ref-lected the hydrophobic properties of (B)Chl molecules. The properties should play an important role in binding (B)Chl molecules to photosynthetic apparatus produced under various culturing conditions, since preferential binding of BChls-a having several 17-propionate ester groups to light-harvesting complexes cultured under dif-ferent conditions has already been demonstrated [8].

EXPERIMENTAL

Bacterial strains and growth conditions

Hlr. halochloris DSM 1059, a thermophilic purple photosynthetic bacterium, was purchased from the Deutsche Sammlung von Mikroorganismen und Zellkul-turen GmbH (DSMZ; Braunschweig, Germany) [9, 10]. This bacterium was cultured anaerobically at 40–45 °C with an incandescent lamp in 2000 mL screw-capped

3 7 11 15XO2 6 10 14

3 7 11 15XO2 6 10 14

3 7 11 15XO2 6 10 14

3 7 11 15XO2 6 10 14

GG(Δ2,6,10,14-Phytatetraenyl)

Δ2,10,14-Phytatrienyl

Δ2,14-Phytadienyl

Phy(Δ2-Phytaenyl)

16

16

16

16

Scheme 1. Stepwise reduction of a geranylgeranyl (GG) to a phytyl (Phy) group found in biosynthesis of BChl-a. XOH represents bacteriochlorophyllide-a or diphosphate

Fig. 1. (a) Molecular structure of BChl-b and (b) partial struc-ture of phytadienyl-type substituents found in photosynthetic bacteria. The replacement of the central magnesium by two pro-tons gives the corresponding BPhe-b

N N

N N

O

Mg

O

CO2CH3

17

173

O OR

(a)

(b)

172-COO

2 3 6 1011 14157

172-COO

2 3 1415

A B

CD

E

172-COO

2 3 6 7

7

8

132

18

4 5 8 91213

116

31 71 111 151


STRUCTURAL DETERMINATION OF THE Δ2,10-PHYTADIENYL SUBSTITUENT 43

bottles with liquid 253 medium as described in the DSMZ instructions, and was used for the isolation of BChl-b, hav-ing a Δ2,10-phytadienyl substituent as the 17-propionate. The purple photosynthetic bacterium Blastochloris (Blc.)viridis (previously named Rhodopseudomonas viridis)DSM 133 was provided by Prof. K. Inoue (Kanagawa University, Japan), and was used for the authentic BChl-bhaving a phytyl group as the 17-propionate [11, 12]. Blc.viridis was grown under anaerobic-light conditions at 30 °C in a liquid malate-basal salt medium as described in the literature [13]. Rhodopseudomonas sp. strain Rits from our culture collection was cultured anaerobically at 30 °C in the liquid PYS medium [8], and was used for the preparation of another Chl derivative which has a Δ2,14-phytadienyl substituent.

Preparation of membranes from Blc. viridis

The harvested cells of Blc. viridis resuspended in 50 mM phosphate buffer (pH 7.0) were broken twice by French Press at 1200 kgf.cm-2. The unbroken debris was then sedimented by centrifugation at 11000 g for 30 min at 4 °C. The resulting supernatant, including membranes, was used for the absorption measurement.

Determination of the compositions of BChls-b having several propionate ester groups in the cells of purple bacteria

BChls-b were extracted from the cultured wet cells of Hlr. halochloris and Blc. viridis as follows. A mixture of acetone and methanol (9:1 (v/v)) was added to the above harvested cells, and mixed using a vibrator. A mix-ture of petroleum ether and diethyl ether (1:1 (v/v)) and then distilled water were added to transfer the BChls-bcomponent to the ether layer for collection. The ether containing BChl-b components was evaporated to dry-ness by a stream of nitrogen. The extract thus obtained was dissolved in methanol for LC-MS analysis. The LC-MS of the extracted BChls-b was performed using a Shimadzu LCMS-2010EV system (Shimadzu, Kyoto, Japan) comprising a liquid chromatograph (SCL-10Avp system controller, LC-10ADvp pump and SPD-M10Avp photodiode-array detector) and a quadrupole mass spec-trometer equipped with an atmospheric pressure chemi-cal ionization (APCI) probe as described previously [5]. HPLC was performed using reverse-phase chromatogra-phy under the following conditions: column, Cosmosil 5C18-AR-II (3.0 mm × 150 mm, Nacalai Tesque, Kyoto); elu ent, 5.0% (v/v) H2O in methanol; fl ow rate, 0.5 mL.min-1; detection wavelength, 790 nm and 680 nm.

Preparation and structural determination of BPhe-b having a Δ2,10-phytadienyl group from Hlr. halochloris

The pigment components were extracted from the wet cells of Hlr. halochloris as mentioned above. To obtain

more stable, metal-free bacteriopheophytin(BPhe)-bhaving a Δ2,10-phytadienyl group as the 17-propionate, aqueous HCl (~10 M, ~20 mL) was added to the result-ing ether solution as mentioned in the above section (~20 mL). The mixture was shaken vigorously, washed with distilled water three times, and evaporated, after which the residue was purifi ed by the following normal-phase HPLC to give pure BPhe-b: column, Cosmosil 5SL-II (6.0 mm × 250 mm); eluent, 20% (v/v) acetone in hexane; fl ow rate, 1.75 mL.min-1 and detection wavelength, 780 nm and 680 nm. BPhe-b having a Δ2,10-phytadienyl group was isolated in ~5 mg. The 600 MHz 1H NMR and 150 MHz 13C NMR spectra of BPhe-b were recorded in chloroform-d (Merck, Darmstadt, Germany) using a JEOL ECA-600 NMR spectrometer (JEOL, Akishima, Japan). A set of assignments of 1H signals was obtained by correlation spectroscopy (COSY) and rotating frame Overhauser enhancement spectroscopy (ROESY; τm = 400 msec). A set of assignments of 13C signals was obtained using distortionless enhancement polarization transfer (DEPT) and heteronuclear multiple-bond corre-lation (HMBC; nJCH = 12 Hz) spectroscopies.

Preparation of 3-acetyl-Chls-a having different phytadienyl groups and their HPLC analysis

As reported earlier, BChl-a having a Δ2,14-phytadienyl group was extracted from the cells of Rhodopseudomonas sp. Rits [5]. The C7–C8 single bond at ring-B in BChl-a was oxidized with 2,3-dichloro-5,6-dicyano-p-benzoquinone(DDQ) in acetone to afford 3-acetyl-Chl-a, according to the reported procedures [14]. Briefl y, DDQ in acetone (~8 mM) was added, with stirring, to an acetone solution (10 mL) of BChl-a possessing various esters (16 mg). After the red-most absorption band was blue-shifted (770 → 678 nm in acetone), 2-propanol was added to the reac-tion mixture. The crude product was purifi ed and separat-ed by RP-HPLC under the following conditions, yielding 3-acetyl-Chl-a having a Δ2,14-phytadienyl group: column, Cosmosil 5C18-AR-II (4.6 mm × 150 mm); eluent, 7.5% (v/v) H2O in methanol; fl ow rate, 1.0 mL.min-1; detection wavelength, 440 nm. Isomerization of BChl-b to 3-acetyl-Chl-a was performed based on the reported procedures [15, 16]. Briefl y, BChl-b in an aqueous acetone solution containing HCl (8 × 10-5 M) ([BChl-b] = 3.0 × 10-6 M; 30 mL) was stirred for 50 min in the dark, in air: λmax = 795 → 675 nm in acetone. After extraction of the pigment components with toluene, the mixture was evaporated and purifi ed by HPLC, yielding 3-acetyl-Chl-a having a Δ2,10-phytadienyl group. The resulting 3-acetyl-Chls-awere analyzed by RP-HPLC using a method similar to that used for the above- mentioned BChl-b, except that 7.5% (v/v) H2O in methanol was used as an eluent.

General methods

Visible absorption and circular dichroism (CD) spectra were measured with a Hitachi U-3500 (Hitachi, Ltd.,



Tokyo, Japan) and a JASCO J-720W spectrophotometer (JASCO, Ltd., Hachioji, Japan), respectively. Optical density was about 1.0/10 mm at the Qy absorption maxi-mum for spectroscopic measurements. Chemical shifts of proton and carbon-13 signals were represented using internal references, CHCl3 (δ = 7.26 ppm) and 13CDCl3

(δ = 77.00 ppm).

Characterization of compounds

BChl-b having a phytyl group. UV-vis (acetone): λmax, nm 796 (relative intensity, 1.00), 582 (0.28), 408 (0.85) and 370 (0.97). MS (APCI): m/z found: 909.4, calcd. for C55H72N4O6Mg: MH+, 909.53.

BPhe-b having a Δ2,10-phytadienyl group. UV-vis (acetone): λmax, nm 777 (relative intensity, 0.49), 528 (0.23), 399 (0.85) and 368 (1.00). 1H NMR (chloro-form-d): δ, ppm 8.98 (1H, s, 5-H), 8.81 (1H, s, 10-H), 8.41 (1H, s, 20-H), 7.03 (1H, q, J = 6.8 Hz, 81-H), 6.10 (1H, s, 132-H), 5.14 (1H, q, J = 7.6 Hz, 7-H), 4.28 (1H, dq, J = 1.4, 6.9 Hz, 18-H), 4.01 (1H, m, 17-H), 3.86 (3H, s, 135-H3), 3.48 (3H, s, 21-H3), 3.47 (3H, s, 121-H3), 3.15 (3H, s, 31-H3) 2.2–2.6 (4H, m, 171- and 172-H2), 2.30 (3H, d, J = 6.9 Hz, 81-H3), 1.85 (3H, d, J = 7.6 Hz, 71-H3),1.71 (3H, d, J = 6.9 Hz, 181-H3), 0.48 (1H, s, NH), -0.96 (1H, s, NH) for bacteriochlorin skeleton and 5.17 (1H, t, J = 6.8 Hz, 2-H), 5.05 (1H, t, J = 6.9 Hz, 10-H), 4.48 (2H, m, 1-H2), 1.85–1.95 (6H, 4-, 9-, and 12-H2), 1.59 (3H, s, 31-H3), 1.54 (3H, s, 111-H3), 1.49 (1H, m, 15-H), 1.20–1.40 (8H, 5-, 7-, 8-, and 13-H2), 1.10 (2H, m, 14-H2), 1.07 (2H, m, 6-H2), 0.84 (6H, d, J = 6.8 Hz, 151- and 16-H3), 0.81 (3H, d, J = 6.2 Hz, 71-H3) for Δ2,10-phytadienyl. 13C NMR (chloroform-d): δ, ppm 199.03 (C31), 189.05 (C131), 172.92 (C173), 169.50 (C133), 168.86 (C6), 158.60 (C16), 155.84 (C9), 148.27 (C14), 144.52 (C8), 139.18/123.06 (C11/C12, interchangeable), 138.11/136.48/136.04 (C1/C2/C3/C19, two of them overlapped each other), 133.19 (C4), 128.94 (C13), 118.57 (C81), 107.77 (C15), 97.03 (C5), 95.93 (C10), 95.26 (C20), 64.35 (C132),52.82 (C135), 50.73 (C17), 49.58 (C18), 46.27 (C7), 33.32 (C32), 31.12 (C172), 29.85 (C171), 22.87 (C181),21.29 (C71), 15.30 (C82), 13.42 (C21), 11.60 (C121)for bacteriochlorin skeleton and 142.85 (C3), 135.00 (C11), 124.53 (C10), 117.70 (C2), 61.48 (C1), 39.88 (C12), 39.77 (C4), 38.56 (C14), 37.02 (C8), 36.52 (C6), 32.21 (C7), 27.83 (C15), 25.68 (C13), 25.36 (C9), 24.93 (C5), 22.62 (C151 and C16), 19.51 (C71),16.27 (C31), 15.79 (C111) for Δ2,10-phytadienyl. MS (APCI): m/z found: 885.8, calcd. for C55H72N4O6: MH+,885.55.

3-acetyl-Chl-a having a Δ2,14-phytadienyl group.UV-vis (acetone): λmax, nm 676 (relative intensity, 0.85), 437 (1.00) and 389 (0.64). MS (APCI): m/z found: 907.60, calcd. for C55H70N4O6Mg: MH+, 907.51.

3-acetyl-Chl-a having a Δ2,10-phytadienyl group.UV-vis (acetone): λmax, nm 676 (relative intensity, 0.85),

437 (1.00) and 389 (0.64). MS (APCI): m/z found: 907.80, calcd. for C55H70N4O6Mg: MH+, 907.51.


Molecular structure of BChl-b and a variety of the 17-propionate substituent

Figure 1a shows the molecular structure of BChl-b.Replacement of the central magnesium by two protons gives the corresponding metal-free BPhe-b. BChl-b is found in some anoxygenic, purple photosynthetic bac-teria and is characterized by the partially saturated bac-teriochlorin π-system as well as the unique ethylidene group at the 8-position on ring-B [1, 3, 17]. Based on these structural features, the BChl-b molecule exhibits a far red-shifted Qy-absorption band up to the near infrared region as shown in Fig. 2: ~800 nm for in vitro isolated BChl-b in acetone (the dotted line) and ~1000 nm for in vivo BChl-b in an integral membrane protein from Blc.viridis in an aqueous buffer (thin solid line). Due to the presence of the ethylidene group, BChl-b was reported to be quite unstable [15, 16]. As a result, isomerization of BChl-b to a chlorin-type derivative, 3-acetyl-Chl-a, easily occurred as shown at the bottom of Scheme 2 [15, 16].

Scheme 1 depicts a biosynthetic pathway from a geranylgeranyl (Δ2,6,10,14-phytatetraenyl) to a phy-tyl (Δ2-phytaenyl) group for BChl-a [5, 18, 19]. Here, XOH represents bacteriochlorophyllide-a or diphosphate (replacement of R1 in BChl-a of Scheme 2 by a hydrogen atom gives bacteriochlorophyllide-a). Hydrogenation of one (two) of four C=C bonds in GG gives phytatrienyl (phytadienyl). Due to the stepwise reduction, a variety of the propionates were produced, although the sequential hydrogenation step, C6=C7 → C10=C11 → C14=C15, was proposed based on the structural determination of a Δ2,10,14-phytatrienyl and a Δ2,14-phytadienyl group in BChl-a as shown in Scheme 1 [5]. Figure 1b shows other phytadienyl groups found in photosynthetic bacteria: the presence of a Δ2,6-phytadienyl group in Chl-a from the thermophilic green bacterium Chl. tepidum (the top structure in Fig. 1b) was unambiguously identifi ed [7]. On the other hand, the presence of a Δ2,10-phytadienyl group in BChl-b from the thermophilic purple bacterium, Hlr. halochloris (the middle structure in Fig. 1b) was reported [6]. The ester was tentatively assigned on the basis of GC-MS analyses of its cleaved alcohol and the derivatives, but the positions of two C=C double bonds have not yet been fully confi rmed.

HPLC profi les of the extracts from purple photosynthetic bacteria containing BChl-b

Figure 3 shows the representative HPLC profi les monitored at 790 nm for the extracts from the cells of Hlr. halochloris cultured at ~45 °C (a) and Blc. viridis at ~30 °C (b). Several BChl-b components were detected as



the pigment absorbed a 790-nm light; these were termed peaks #1 – #3 in trace (a) and peaks #1′ – #4′ in trace (b) of Fig. 3. According to previous reports on the accumu-lation of several BChl molecules in purple bacteria [5, 20, 21], these three or four peaks are ascribed to BChl-bmolecules having different 17-propionate ester groups as shown in Scheme 1. Based on the stepwise reduction from a geranylgeranyl to a phytyl group, BChls-b #1(#1′)to #3(#3′) were tentatively assigned as BChls-b esterifi ed

with GG, phytatrienyl and phytadienyl in the order of elution and BChl-b #4′ in Fig. 3(b) as Phy.

First, we confi rmed the assignment of these different BChls-b by APCI-LCMS analysis. Table 1 lists the results of mass spectrometry of BChls-b #1 to #4′. The molecu-lar and fragment ions can be used to identify each BChl-bcomponent. BChl-b #1(#1′), #2(#2′), #3(#3′) and #4′gave peaks at m/z = 903.3 (903.5), 905.3 (905.5), 907.3

N N

N N

O

Mg

O

CO2CH3O OR1

Rhodopseudomonassp. Rits

Hlr. halochloris DSM1059

N N

N N

O

Mg

O

CO2CH3O OR2

N N

N N

O

Mg

O

CO2CH3O OR1

N N

N N

O

Mg

O

CO2CH3O OR2

Co-chromatography

(i)

(i)

(ii)

(iii)

B7

8

3

3

BChl-a

BChl-b

3-Acetyl-Chl-a

3-Acetyl-Chl-a

Scheme 2. Synthesis of 3-acetyl-Chl-a derivatives esterifi ed with Δ2,14- and Δ2,10-phytadienyl groups at R1 and R2, respectively: (i) extraction and separation; (ii) oxidation with DDQ; (iii) isomerization under acid conditions

Fig. 2. Electronic-absorption spectra of BChl-b from Hlr. halochloris (dotted), BPhe-b prepared by demetalation of BChl-b from Hlr. halochloris (thick solid), 3-acetyl-Chl-a pre-pared by isomerization of BChl-b (dashed) and the membranes from the cells of Blc. viridis (thin solid). All the spectra were recorded in acetone except for the membranes in an aqueous buffer and were normalized at the most intense bands

Fig. 3. (a) RP-HPLC profi les for the extracts from purple bacte-ria containing BChl-b: Hlr. halochloris cultured at ~45 °C and (b) Blc. viridis at ~30 °C. The inset in (b) shows a 100-fold expansion at the retention time of 10–25 min



(907.4) and 909.4, corresponding to their protonated form ([MH]+). The peak at 909.4 for major BChl-b #4′ in Blc. viridis as shown in Fig. 3b is consistent with that of protonated BChl-b esterifi ed with Phy (calcd. for 909.53). The peaks for BChls-b #1(#1′), #2(#2′) and #3(#3′) were smaller than that for BChl-b #4′ by 6.1 (5.9), 4.1 (3.9) and 2.1 (2.0) Da, respectively. These results indicated that #1(#1′) to #3(#3′) were assigned to BChls-b esterifi ed with GG, phytatrienyl and phytadienyl, respec-tively. Furthermore, they have identical fragment ions at m/z = 631.4 (631.1), 631.0 (631.1) and 631.1 (631.1) due to loss of the corresponding esterifying groups, to be assigned to bacteriochlorophyllide-b (calcd. for 631.23). Thus, the 2.0 Da differences among the molecular ions of BChls-b are assigned to a structural difference in the esters. Although BChl-b #3′ from Blc. viridis was detected as only a trace amount, a slight difference in the retention times of BChls-b #3 and #3′ was observed, indicating that the ester group in the two components was different (see Fig. 3). This was confi rmed clearly in the following section (vide infra).

Acid treatment of BChl-b from Hlr. halochloris

Because BChl-b was reported to be quite unstable [15, 16], we demetalized BChl-b from Hlr. halochloris to get more stable, metal-free BPhe-b. When the intact extract was treated with an aqueous diluted HCl solution, BChls-b seen in Fig. 3a were completely eliminated, with an increase in the corresponding BPhe-b. Figure 4 presents the HPLC profi le after acid treatment of BChl-bat two different detection wavelengths, 680 (dotted) nm and 780 (solid) nm. Time-dependent isomerization of BPhe-b to a chlorin-type derivative (3-acetyl-pheophy-tin-a) under acid conditions has been demonstrated in the literature [22]. During the present acid treatment, a slight amount of undesirable components originating from chlorin-type derivatives was produced. This was confi rmed by the HPLC analysis monitored at 680 nm as shown by the dotted line of Fig. 4. The main peak eluted at ~6.5 min was assigned to be the desired BPhe-b

having a Δ2,10-phytadienyl group in the 17-propionate. This was confi rmed by on-line APCI-LCMS analysis: the molecular-ion was observed at m/z = 885.8 for [MH]+ as the protonated form (calcd. for 885.55). Furthermore, the fragment ion at m/z = 609.4 was assigned to loss of the esterifying group. The absorption spectrum of the result-ing BPhe-b is shown by the thick solid line of Fig. 2. The presence of a phytadienyl-type group in BPhe-b was confi rmed by mass spectrometry, although the positions of two C=C double bonds could not be determined. It is noteworthy that the transformation of BChl-b to BPhe-bby the action of an acid induced no structural change of other substituents in the molecule, all of which retained their stereochemistry.

Structural determination of the Δ2,10-phytadienyl substituent in BPhe-b from Hlr. halochloris

For the structural determination of the 17-propionate ester substituent in BPhe-b, we measured the NMR spectra. The 1H NMR spectrum of phytylated BChl-b

Fig. 4. RP-HPLC profi le monitored at 780 (solid) and 680 nm (dotted) for the BChl-b extract from Hlr. halochloris after acid treatment

Table 1. APCI-mass spectrometric data of BChls-b from the cells of Hlr. halochloris and Blc. viridis

Observed peaks

HPLC peak # Molecular ion Fragment Calcd. For MH+ Esterifying group (R)

1 903.3 631.4 903.48 GG

2 905.3 631.0 905.50 phytatrienyl

3 907.3 631.1 907.51 phytadienyl

1′ 903.5 631.1 903.48 GG

2′ 905.5 631.1 905.50 phytatrienyl

3′ 907.4 631.1 907.51 phytadienyl

4′ 909.4 631.1 909.53 Phy



is available in the literature [17, 23]. Figure 5 presents the partial 1H NMR spectrum of BPhe-b from Hlr. hal-ochloris, showing the region where proton signals of the esterifying alcohol occur. In the inset of the fi gure, two olefi nic proton signals can be clearly observed, although the signal from ring-B (7-H) indicated by an asterisk partially overlaps 2-H of the ester. This corresponds to the number of C=C double bonds in a phytadienyl-type group in BPhe-b. Another characteristic feature in the 1H NMR spectrum was observed in the signals from the methyl branches in the ester as shown in Fig. 5. Two singlet methyl peaks at 1.5 to 1.6 ppm and three doublet methyl peaks around 0.8 ppm were clearly observed. This showed that the two methyl groups were attached to qua-ternary (olefi nic, sp2) carbon atoms and three to tertiary (saturated, sp3) ones. Of the latter, higher-fi eld shifted, methyl groups, a pair of methyl signals overlapped each other, indicating that the two methyl groups were equiva-lent and at the 15-position. This type of characterization of the methyl branches in the ester group confi rmed the

presence of the C14–C15 single bond, although the posi-tions of two C=C double bonds could not be identifi ed: two of C2=C3, C6=C7 and C10=C11.

To determine the positions of two C=C double bonds in a phytadienyl group, we applied 2D-NMR spectros-copy. The DEPT spectra using 90° and 135° pulses were also used to differentiate the primary (p), secondary (s),tertiary (t) and quaternary (q) carbon atoms, and to con-fi rm the assignments of 13C signals (13C-1D and DEPT135 spectra with the assignment of the 13C-signals of the ester group are shown in Figs S1 and S2 in the Support-ing Information). Figure 6 summarizes the correlations which were found in COSY, ROESY and 1H-13C HMBC spectra, although the correlations in the central part of the ester, 7-H ↔ 8-H in COSY, could not be identifi ed due to their similar chemical shifts. The types of carbon atom differentiated by DEPT spectra are also indicated in parenthesis (see Figs S1 and S2). These correlations in 2D NMR and the differentiated carbon atom types were used to establish the structure of the phytadienyl group, starting from the unambiguously assigned C173(=O) and 71-H3 signals. The presence of the C2=C3 double bond was identifi ed by the following sequential correlations in HMBC: C173(q) ↔ 1-H2 ↔ C2(t), 2-H ↔ C31(p) and 1-H2 ↔ C3(q). The presence of the C10=C11 double bond in the ester was also identifi ed by the correlations 71-H3 ↔ 8-C(s) ↔ 9-H2 and 9C(s) ↔ 10-H ↔ C111(p)in HMBC. The presence of C6–C7 and C14–C15 single bonds in the ester were confi rmed in a similar way. Thus, the positions of two C=C double bonds in the phytadienyl ester group are established unambiguously as C2=C3 and C10=C11 shown in the middle of Fig. 1b. Figure 6 also indicates NOE correlations in the ROESY spectrum. They were used to establish the molecular confi gurations of the tri-substituted olefi ns in the ester group. E-confi guration of the C2=C3 double bond (trans in C1–C2=C3–C4, see Fig. 1b) in the ester was identifi ed using NOE correla-tions between 1-H2 ↔ 31-H3 and 2-H ↔ 4-H2. Similarly, E-confi guration of the C10=C11 double bond in the ester was identifi ed.

In addition to the above structural determination of the ester group, we characterized the stereochemistry at rings-B, -D and -E of BPhe-b. After the assignments of the 1H signals in BPhe-b were established, we analyzed

Fig. 5. Partial 1H NMR spectrum of BPhe-b having a Δ2,10-phytadienyl group in CDCl3 at room temperature. The inset shows the olefi nic protons in the ester. Characteristic signals in the ester are labeled. The numbering corresponds to that in Fig. 1(b)

Fig. 6. Observed 2D NMR correlations in the Δ2,10-phytadienyl group. Closed and open arrows indicate proton and carbon atoms, respectively. The numbering corresponds to that in Fig. 1(b). Primary, secondary, tertiary and quaternary carbon atoms differentiatedby DEPT spectra are indicated by p, s, t and q, respectively, in parenthesis



its ROESY spectrum. The apparent NOE correlations between 81-H and 10-H as well as 82-H3 and 7-H/71-H3

shown in Fig. 7a clearly indicate the E-confi guration for the tri-substituted double bond at the 8-ethylidene group on ring-B of BPhe-b. The result is compatible with the stereochemistry of the 8-ethylidene group in the struc-turally related BChl-b having a phytyl group [23] and BChl-g having a farnesyl group [24].

To confi rm the confi gurations at the chiral 132-, 17- and 18-positions on rings-D and -E, we used NOE cor-relations and CD spectra. The syn-orientations of 132-Hand 17-CH2CH2 as well as of 17-CH2CH2 and 18-H (17-H and 18-CH3) were confi rmed based on the NOE correla-tions shown in Fig. 7b: 132-H ↔ 171-H2, 17-H ↔181-H3,171-H2 ↔ 18-H, and 172-H2 ↔ 18-H. From these NOE correlations, the confi gurations around rings-D and -E in BPhe-b were established to be 132-(R)-, 17-(S)- and 18-(S)-confi gurations or their mirror image 132-(S)-, 17-(R)-and 18-(R)-confi gurations.

To distinguish the pair of enantiomers confi rmed above, we used the CD spectrum. Figure 8 shows the CD spectrum of BPhe-b in THF. BPhe-b exhibited a weak positive CD signal at the Qy, a negative signal at the Qx and a positive signal at the Soret-bands. Comparing the above intrinsic CD signals with previous data of structurally determined phytylated BChl-b and farnesylated BChl-g [25], the absolute confi guration at the 132-position of the present BPhe-b can be confi rmed as 132-(R)-confi guration. Thus,

the confi gurations at the 132-, 17- and 18-positions on rings-D and -E of BPhe-b were characterized to be 132-(R)-, 17-(S)- and 18-(S)-confi gurations generally seen in chlorophyllous pigments. The remaining undetermined stereochemistry was at the 7-position on ring-B. By analogy with the stereochemistry of BChl-a [26], the R-confi guration is expected for the 7-position of BPhe-b.

Synthesis and HPLC analysis of 3-acetyl-Chls-ahaving different types of phytadienyl substituents

To obtain two types of Chl derivatives having different phytadienyl substituents at the 17-propionate, we used structure-determined BChl-a having a Δ2,14-phytadi-enyl from Rhodopseudomonas sp. Rits [5] and BChl-b having Δ2,10-phytadienyl determined in this study from Hlr. halochloris (the structures of these ester groups are shown in Fig. 1b). Scheme 2 depicts the chemical modi-fi cation of BChl-a (the upper route) and BChl-b (the bot-tom route) to 3-acetyl-Chl-a. Oxidation of BChl-a withDDQ affords 3-acetyl-Chl-a having a Δ2,14-phytadienyl group as shown in the upper route of Scheme 2, according to the procedures in the literature [14]. On the other hand, 3-acetyl-Chl-a having a Δ2,10-phytadienyl group was prepared by isomerization of BChl-b from Hlr. halochlo-ris as shown in the bottom route of Scheme 2 [16]. The resulting 3-acetyl-Chls-a possessing phytadienyl groups obtained from different sources were characterized and confi rmed by on-line LC-MS analysis. The absorption spectrum of 3-acetyl-Chl-a prepared by isomerization of BChl-b is shown by the dashed line in Fig. 2.

The HPLC profi les of each 3-acetyl-Chl-a are shown in Figs 9a and 9b. Figure 9c shows the co-chromatographic analysis of the two 3-acetyl-Chls-a, thus obtained. Each derivative exhibited its intrinsic retention time: 3-acetyl-Chl-a having C2=C3 and C14=C15 double bonds in the 17-ester group was eluted more slowly than that having C2=C3 and C10=C11 bonds. Therefore, the position of C=C double bonds in the phytadienyl-type ester group affected RP-HPLC elution, indicating that the two

Fig. 7. (a) Partial structures of BPhe-b around ring-B and (b) rings-D and -E together with observed NOE correlations indicated by arrows

Fig. 8. A CD spectrum of BPhe-b recorded in THF



3-acetyl-Chls-a had different molecular hydrophobicity. The Δ2,14-phytadienyl ester was more hydrophobic than the Δ2,10-isomer. This difference would be ascribable to the fl exibility in the ester. The Δ2,14-phytadienyl ester possessing longer C–C single bonds might make a more fl exible ester chain than the Δ2,10-isomeric ester pos-sessing parts of two C–C chains.

CONCLUSION

We determined the structure of a phytadienyl-type group in the 17-propionate ester of BChl-b from Hlr. halochloris to be Δ2,10-E,E-phytadienyl by means of NMR spectroscopy. In addition, we confi rmed the stereo-chemistry at the 8-ethylidene group on ring-B to be E-confi guration as well as that at the chiral 132-, 17- and 18- positions on rings-D and -E to be 132-(R)-, 17-(S)- and 18-(S)-confi gurations, but neither the absolute confi gura-tion of the 7-position on the phytadienyl group, nor that of the 7-position on ring-B could be determined by the present NMR data: the former and latter was specul ated

to be (S) and (R), respectively. The presence of ger anyl-geranylated BChl-b (peak 1 of Fig. 3a) and the absence of phytylated BChl-b (peak 4′ of Fig. 3b) in the cells of Hlr. halochloris showed that the Δ2,10-phytadienyl group was biosynthetically prepared by stepwise hydrogenation: GG → Δ2,10,14-phytatrienyl or Δ2,6,10-phytatrienyl →Δ2,10-phytadienyl (see Fig. S3). Peak 2 has almost the same retention time as peak 2′ shown in Figs 3a and 3b, indicating that the Δ2,10,14-phytatrienyl group is the probable intermediate and the C6=C7 double bond of GG would be reduced fi rst. Moreover, we demonstrated the effect of the positions of two C=C double bonds in the ester group upon RP-HPLC elution using two types of Chl derivatives which have regioisomeric phytadienyl groups. The results supported the idea that the positions of C=C double bonds in the propionate ester group regulated molecular hydrophobicity of (B)Chl molecules; as a result, effective construction of photosynthetic appa-ratus might be achieved under various bacterial growth conditions.

Acknowledgements

We would like to thank Prof. Kazuhito Inoue at Kanagawa University for his kind provision of a purple photosynthetic bacterium, Blc. viridis DSM 133. We are also grateful to Ms. Fukimi Katayama at Ritsumeikan University for her experimental assistance. This work was partially supported by Grants-in-Aid for Scientifi c Research (B) (No. 19350088) (to HT) from the Japa-nese Society for the Promotion of Science (JSPS) and for Young Scientists (B) (No. 19750148) (to TM) from the Ministry of Education, Culture, Sports, Science and Technology. JH was supported by a JSPS Fellowship for Young Scientists (No. 183495).

Supporting information13C NMR and DEPT135 spectra of the isolated BPhe-b

having a Δ2,10-phytadienyl group in CDCl3 and a propo-sed biosynthetic route to Δ2,10-phytadienylated BChl-bare given in the supplementary material. This material is available at http://www.worldscinet.com/jpp/jpp.shtml.

REFERENCES

Scheer H. In 1. Chlorophylls, Scheer H. (Ed.). CRC Press: Boca Raton, 1991; 3–30.Tamiaki H, Shibata R and Mizoguchi T. 2. Photochem. Photobiol. 2007; 83: 152–162.Scheer H. In 3. Chlorophylls and Bacteriochloro-phylls: Biochemistry, Biophysics, Functions and Applications, Grimm B, Porra RJ, Rüdiger W and Scheer H. (Eds.). Springer: Dordrecht, 2006; 1–26.Cogdell RJ, Gall A and Köhler J. 4. Quart. Rev. Bio-phys. 2006; 39: 227–324.Mizoguchi T, Harada J and Tamiaki H. 5. FEBS Lett.2006; 580: 6644–6648.

Fig. 9. RP-HPLC profi les of isomeric 3-acetyl-Chls-a having phytadienyl groups in the 17-propionate prepared (a) by oxidation of BChl-a from Rhodopseudomonas sp. Rits and (b) by isomerization of BChl-b from Hlr. halochloris, and the co-chromatography of (a) and (b) is shown in (c)



Steiner R, Schäfer W, Blos I, Wieschhoff H and 6.Scheer H. Z. Naturforsch. 1981; 36c: 417–420.Kobayashi M, Oh-oka H, Akutsu S, Akiyama 7.M, Tominaga K, Kise H, Nishida F, Watanabe T, Amesz J, Kizumi M, Ishida N and Kano H. Photo-synth. Res. 2000; 63: 269–280.Harada J, Mizoguchi T, Yoshida S, Isaji M, Oh-oka H 8.and Tamiaki H. Photosynth. Res. 2008; 95: 213–221. Imhoff JF and Trüper HG. 9. Arch. Microbiol. 1977; 114: 115–121. Imhoff JF and Süling J. 10. Arch. Microbiol. 1996; 165:106–113.Drews G and Giesbrecht P. 11. Arch. Mikrobiol. 1966; 53: 255–262.Hiraishi A. 12. Int. J. Syst. Bacteriol. 1997; 47: 217–219.Maki H, Matsuura K, Shimada K and Nagashima 13.KV. J. Biol. Chem. 2003; 278: 3921–3928.Tamiaki H, Takeda K, Kondo S and Tanikaga R. 14.Tetrahedron: Asymmetry 1998; 9: 2101–2111.Steiner R, Cmiel E and Scheer H. 15. Z. Naturforsch.1983; 38c: 748–752.Kobayashi M, Yamamura M, Akutsu S, Miyake J, 16.Hara M, Akiyama M and Kise H. Anal. Chim. Acta.1998; 361: 285–290.

Scheer H, Svec WA, Cope BT, Studier MH, Scott 17.RG and Katz JJ. J. Am. Chem. Soc. 1974; 96: 3714–3716.Rüdiger W. In 18. The Porphyrin Handbook,Kadish KM, Smith KM and Guilard R. (Eds.), Vol. 13, Academic Press: San Diego, 2003; 71–108.Addlesee HA and Hunter CN. 19. J. Bacteriol. 1999; 181: 7248–7255. Shioi Y and Sasa T. 20. J. Bacteriol. 1984; 158: 340–343.Steiner R, Wieschhoff H and Scheer H. 21. J. Chroma-tography 1982; 242: 127–134.Kunieda M, Mizoguchi T and Tamiaki H. 22. Tetrahe-dron 2004; 60: 11349–11357.Risch N. 23. J. Chem. Res. (S). 1981; 116–117.Mizoguchi T, Oh-oka H and Tamiaki H. 24. Photochem. Photobiol. 2005; 81: 666–673.Kobayashi M, Akiyama M, Kano H and Kise H. 25.In Chlorophylls and Bacteriochlorophylls: Bio-chemistry, Biophysics, Functions and Applications,Grimm B, Porra RJ, Rüdiger W and Scheer H. (Eds.). Springer: Dordrecht, 2006; pp. 79–94. Barkigia KM, Gottfried DS, Boxer SG and Fajer J. 26.J. Am. Chem. Soc. 1989; 111: 6444–6446.




INTRODUCTION

Photodynamic therapy (PDT) combines a photosensiti-zer, visible or near-infrared light, and oxygen to produce necrosis and/or apoptosis in target tissues [1–4]. As a rapi-dly growing methodology, PDT has been used in the treat-ment of various cancers (e.g. melanoma, early and advanced stage cancer of the lung, digestive tract and genito-urinary tract) and the wet form of age-related macular degenera-tion [3–5]. Most of the currently available porphyrin-based pho to sen sitizers are highly hydro phobic compounds (e.g.he mato porphyrin deri vative, tetra (meta-hy dro xyphenyl)-chlorin, 2-(1-hexyloxyethyl)-2-devinyl-py ro pheo phor bide a, Sn(IV)-ethyl etio purpurin), often diffi cult to formu-late and occasionally with long retention times in tissues [6, 7]. Several strategies have been investigated to enhance the water solubility of PDT photo sensitizers in aqueous solutions, as well as their tumor specifi city, including conjugation to carrier peptides [8, 9], oligo nucleotides

[10, 11], monoclonal antibodies [12, 13], and carbohy drates [14–22]. Among these, the carbo hydrate-functio nalized photosensitizers have shown promise in PDT, due to their enhanced interaction with tumor cell receptors via car bo-hydrate-mediated cell recognition processes, increased cellular uptake, and solubility [23–27].

The main challenge in the development of porphyrin-carbohydrate conjugates for application in PDT is the lack of expeditious methodologies for their syntheses. Cur-rently known porphyrin-carbohydrate conjugates were synthesized from either the condensation of carbohydrate-containing benzaldehydes with pyr roles [14–17], or by introducing carbohydrates on a pre-formed macrocycle [18–22]. However, the syn thetic approaches to stable C-linked (vs O- and S-lin ked) porphyrin-carbo hydrate conjugates remain a chal lenge and often result in low yields of the target com pounds [17].

The Cu(I)-catalyzed azide-alkyne 1,3-dipolar cy clo-addition, i.e. “click” chemistry, is an attractive strategy for the expeditious preparation of porphyrin-car bo hydrate conjugates in high yields from readily available carbo-hydrate azides. Due to the high regio selectivity and the mild reaction conditions compa tible with a large number

Synthesis of porphyrin-carbohydrate conjugates using “click”

chemistry and their preliminary evaluation in human

HEp2 cells

Erhong Hao, Timothy J. Jensen and M. Graça H. Vicente*

Department of Chemistry, Louisiana State University, Baton Rouge, LA 70803, USA


ABSTRACT: Using a Cu(I)-catalyzed carbohydrate azide-alkynylphenylporphyrin cycloaddition (the so-called “click” chemistry), we have synthesized in high yields, a series of four new porphyrin-car bo-hydrate conjugates containing either one or four galactose or lactose moieties linked via tria zole units to a meso-phenyl group of a TPP or tetrabenzoporphyrin (TBP) macrocycle. The time-de pen dent uptake and subcellular distribution of this series of porphyrin-carbohydrate conjugates were eva luated in human carcinoma HEp2 cells. While the three TPP conjugates accumulated to a simi lar extent within cells and localized mainly in the ER and endosomes, the TBP-galactose conju gate was the one most effi ciently taken up by the HEp2 cells, accumulating approximately 5 times more than the TPP conjugates, and localized preferentially within the cell lysosomes.

KEYWORDS: porphyrin, carbohydrate, PDT, click chemistry.


*Correspondence to: M. Graça H. Vicente, email: [email protected], tel: 225-578-7405, fax: 225-578-3458


52 E. HAO ET AL.

of functional groups, the “click” methodology has found wide applications in the fi elds of chemistry, biology, and material science since it was fi rst reported [28, 29]. Por-phyrins linked via triazole bridges to cyclodextrin [30], crown ethers [31], glucans [32], thioporphyrins [33] and electro de surfaces [34] have been reported using “click” che mis try. Herein, by adopting “click” chemistry, we ha ve developed an effi cient facile methodology for the conjugation of carbohydrates with porphyrins, and have used it in the preparation of a new series of tri azole-linked por phyrin-carbo hydrate conjugates. Using this strategy we prepared galactose- and lactose-conjugated porphyrins, since it has been shown that galactose and lactose conjugates might display enhanced affi nity for binding to galectin proteins that are often over-expressed in tumor cells, and con sequently improve their PDT effi cacy [18–20]. Fur thermore, we extended this methodo-logy to the syn the sis of tetra benzo por phyrin-carbo hydrate conju ga tes. Te tra benzo por phyrins (TBPs) are poten tially promi sing PDT photo sensitizers because of their stron-ger absorptions in the red region of the optical spectrum, where light penetration is con siderably deeper, compa-red with porphyrins [3, 35, 36]. Furthermore, due to the extension of their p-conjugated systems, TBPs display high chemical stability, increased basici ty, and decreased oxidation potentials and solubility compared with the parent porphyrins [37, 38]. The water-solubility of TBPs has been increased via the introduction of sulfonic acid [39], carboxylic acid [40] and nido-carborane [41] groups. The methodolo gy developed herein allows the syntheses of high ly water-soluble carbohydrate-substituted TBPs.

EXPERIMENTAL

Syntheses

All reactions were monitored by TLC using 0.25 mm silica gel plates with or without UV indicator (60F-254). Silica gel Sorbent Technologies 32–63 m was used for fl ash column chromatography. 1H and 13C NMR spectra were obtained on either a DPX-250 or an ARX-300 Bru-ker spectrometer. Chemical shifts ( ) are given in ppm relative to residual protons in acetone-d6 (2.05 ppm, 1H; 206.0 ppm, 13C) or CDCl3 (7.26 ppm, 1H; 77 ppm, 13C). Exact mass was obtai ned using ESI-TOF and the isotope peaks were mat ched with calculated patterns (only the most abun dant peaks are listed below). All materials obtai-ned from commercial suppliers were used without fur ther p urifi cation. 5,10,15,20-tetra(4-hydroxyphenyl) porphy rin 1 was purchased from Porphyrin Products, Inc. and carbo-hydrates 2a–2c were purchased from Sigma-Aldrich.

Porphyrin-carbohydrate 3. Porphyrin 1 was mo no-alkylated using the methodology described in the literature [42]. In brief, to a solution of 5,10,15,20-te tra (4-hydroxy phenyl)porphyrin (170 mg, 0.250 mmol) in anhydrous DMSO (40 mL) was added K2CO3

(280 mg, 2.03 mmol). After heating to 60 °C for 15 min under argon, propargyl bromide (32 mg, 0.26 mmol) was added in one portion and the fi nal mixture was stirred at 60 °C overnight. The reaction mixture was cooled to rt, poured into brine (200 mL), extracted with ethyl acetate and dried over anhy drous sodium sulfate. The solvent was removed under va cuum, and the resulting purple residue was purifi ed by fl ash column chromatography on silica gel using a gradient of chloroform/ethyl acetate (1:1 to 100% ethyl acetate) for elution. The free-base porphyrin re sidue was dissolved in chloroform (20 mL) and zinc acetate (0.5 g) in methanol (10 mL) was added. The fi nal solution was refl uxed for 1 h. The solvent was re moved under vacuum, and the mono-alky lated Zn(II)-porphyrin was purifi ed by fl ash column chro ma to graphy on silica gel using chloro form/methanol 95:5 for elution to give 55 mg, 28% yield. MS (MALDI): m/z 781.5 [M + H]+. 1H NMR (CDCl3): δ, ppm 8.70–8.78 (m, 8H), 8.00–8.06 (m, 8H), 7.17–7.31 (m, 8H), 4.90 (s, 2H), 2.61 (s, 1H). The OH was not observed. This Zn(II)-porphyrin (18 mg, 0.023 mmol), carbohydrate 2a (6.3 mg, 0.029 mmol), CuSO4

.5H2O (1.5 mg), and sodium ascorbate (2 mg) were mixed in a 20 mL reaction vial. A solution of tert-BuOH/H2O 1:1 (10 mL) was added and the fi nal mixture was stirred at 70 °C for 24 h. The reaction mix-ture was washed with water and extracted with ethyl ace-tate. The orga nic layer was concentrated and the resulting residue was purifi ed on a pad of silica gel; a trace amount of unreacted starting porphyrin was removed by elution with ethyl acetate and porphyrin 3 was eluted u sing ace-tone. Porphyrin 3 was dried under vacuum to gi ve 21.3 mg (92% yield). 1H NMR (acetone-d6, 250 MHz): δ, ppm 8.92 (6H, br s), 8.80 (2H, br s), 8.37 (1H, s), 8.12 (2H, br s), 8.02–8.06 (6H, m), 7.46–7.49 (2H, m), 7.28-7.33 (6H, m), 7.06 (3H, br s), 5.65 (1H, d, J = 7.66 Hz), 5.40 (2H, s), 4.61 (1H, br s), 4.34 (2H, br s), 4.06 (2H, br s), 3.96 (2H, br s), 3.86–3.88 (1H, m), 3.75-3.76 (4H, m). HRMS (ESI-TOF): m/z 1006.2195 [M + Na]+. Calcd. for C53H41N7NaO9Zn 1006.2155.

Porphyrin-carbohydrate 4. This conjugate was pre-pared as described above, using mono-aspargyl Zn(II)-porphyrin (18 mg, 0.023 mmol), carbohydrate 2b (11 mg, 0.03 mmol), CuSO4

.5H2O (1.5 mg), and sodium ascorbate (2 mg). A solution of tert-BuOH/H2O 1:1 (10 mL) was added and the fi nal mixture was stirred at 70 °C for 24 h. After work-up as descri bed above, porphyrin 4was obtained (23.6 mg) in 88% yield. 1H NMR (acetone-d6, 250 MHz): 8.78 (8H, br s), 8.35 (1H, br s), 7.91–7.99 (8H, m), 7.09–7.26 (8H, m), 6.91 (3H, br s), 5.64 (1H, br s), 5.35 (2H, br s), 4.61 (1H, br s), 4.34 (2H, br s), 3.44–3.99 (17H, m). HRMS (ESI-TOF): m/z 1168.2697 [M + Na]+. Calcd. for C59H51N7NaO14Zn 1168.2683.

Carbohydrate-aldehyde 5. 4-ethynylbenzaldehy de(260 mg, 2 mmol), the carbohydrate azide (700 mg, 1.86 mmol) and CuI (38.2 mg, 0.2 mmol) were added to CH3CN (20 mL). DIPEA (0.9 mL) was then added and the fi nal reaction mixture was stirred at rt overnight. The


EVALUATION OF PORPHYRIN-CARBOHYDRATE CONJUGATES 53

solvent was removed under vacuum and the resulting residue partitioned between dichloro methane and water. The organic layer was concentrated under vacuum and the residue was purifi ed by silica gel column chromato-graphy using 50% ethyl acetate in dichloromethane for elution. The title aldehyde was obtained (883.0 mg) in 93.4% yield. 1H NMR (CDCl3, 250 MHz): δ, ppm 9.88 (1H, s), 8.12 (1H, s), 7.89 (2H, d, J = 8.26 Hz), 7.79 (2H, d, J = 8.38 Hz), 5.86 (1H, d, J = 9.26 Hz), 5.46–5.54 (2H, m), 5.21 (1H, dd, J = 10.24, 3.32 Hz), 4.22–4.25 (1H, m), 4.06–4.09 (2H, m), 2.11 (3H, s), 1.90 (3H, s), 1.89 (3H, s), 1.77 (3H, s). 13C NMR (CDCl3, 62.9 MHz): δ, ppm 193.0, 172.4, 171.6, 171.2, 171.1, 148.3, 137.3, 137.0, 131.6, 127.5, 120.7, 87.5, 75.4, 71.9, 69.3, 68.3, 62.6, 22.3, 21.9, 21.8, 21.5. MALDI-TOF: m/z 504.66 [M + H]+. Calcd. for C23H26N3O10 504.47.

Porphyrin-carbohydrate 6. Aldehyde 5 (252.3 mg, 0.5 mmol) and pyrrole (35 L) were dissolved in dry di-chloromethane (50 mL). The reaction mixture was stirred under argon for 15 min, before 20 L of a BF3

.OEt2

solution (2.5 M in dichloromethane) was added. The mixture was stirred at rt for 1 h. DDQ (227 mg, 1 mmol) was added and the fi nal mixture was stirred for another 1 h. The solvent was evaporated under vacuum and the resulting residue was puri fi ed by silica gel chromatogra-phy, using fi rst di chlo romethane and then ethyl acetate for elution. Porphy rin 6 was dried and obtained (58.0 mg) as a purple solid in 21% yield. 1H NMR (CDCl3, 250 MHz): δ, ppm 8.95 (8H, s), 8.36 (4H, s), 8.26–8.34 (16H, m), 6.04 (4H, d, J = 9.29 Hz), 5.77 (4H, t, J = 9.63 Hz), 5.65–5.66 (4H, m), 5.36–5.41 (4H, m), 4.35–4.40 (4H, m), 4.27–4.29 (8H, m), 2.30 (12H, s), 2.11 (12H, s), 2.08 (12H, s), 2.04 (12H, s). 13C NMR (CDCl3, 62.9 MHz): δ, ppm 170.3, 170.0, 169.8, 169.3, 148.3, 142.3, 135.0, 129.5, 124.3.5, 119.7, 118.3, 86.4, 74.1, 70.8, 67.9, 67.0, 61.3, 26.8, 20.7, 20.5, 20.3. MALDI-TOF: m/z 2203.81 [M]+. Calcd. for C108H106N16O36 2204.08.

Porphyrin-carbohydrate 7. Porphyrin 6 (44.1 mg, 0.02 mmol) was dissolved in methanol (20 mL). A 100 Lsolution of sodium methoxide 0.5 M in me thanol was added. The mixture was stirred at rt for 1 h. After solvent reduction under vacuum, the crude product was redissol-ved in 20 mL methanol and 36.5 mg Zn(OAc)2 was added to the solution. The mixture was stirred at rt for another 2 h and purifi ed by gel fi l tration on a Sephadex LH20 column eluted with me thanol. The pure porphyrin 7 was crystallized from me thanol/water to gi ve 29.1 mg (91% yield). 1H NMR (py ridine-d5, 250 MHz): δ, ppm 9.31 (8H, s), 9.13 (4H, s), 8.43–8.52 (16H, m), 6.50 (4H, d, J = 9.07 Hz), 4.50–5.39 (24H, m). MALDI-TOF: m/z1596.14 [M + H]+, calcd. for C76H73N16O20Zn 1595.90. Anal. calcd. for C76H73N16O20Zn.3H2O: C, 54.43; H, 4.45; N, 13.36; found: C, 54.39; H, 4.60; N, 12.98.

TBP-carbohydrate 8. Aldehyde 5 (504 mg, 1 mmol) and tetrahydroisoindole (125 mg, 1 mmol) were dissolved in dry dichloromethane (100 mL). The mixture was stir-red under argon for 15 min, and then 60 L of a BF3

.OEt2

solution (2.5 M in dichloro methane) was added. The reac-tion mixture was stir red at rt for 5 h in the dark after which DDQ (300.0 mg, 1.32 mmol) was added and the mixture was stirred overnight. The solvent was evaporated under vacu um and the resulting residue was purifi ed by silica gel column chromatography using fi rst dichloro me thaneand then ethyl acetate for elution. The resul ting porphyrinwas isolated (248.0 mg) as a green solid in 41% yield. 1H NMR (CDCl3, 250 MHz): δ, ppm 9.02 (4H, s), 8.29 (8H, d, J = 7.48 Hz), 8.20 (8H, d, J = 7.99 Hz), 6.00 (4H, d, J = 9.48 Hz), 5.78 (4H, t, J = 9.73 Hz), 5.59–5.63 (4H, m), 5.32–5.37 (4H, m), 4.26–4.33 (12H, m), 2.25–2.36 (16H, m), 2.18 (12H, s), 2.09 (12H, s), 2.03 (24H, s), 1.25 (16H, br). The inner NH was not observed due to protonation. MALDI-TOF: m/z 2419.87 [M + H]+, calcd. for C124H131N16O36 2419.89. This porphyrin (61 mg, 0.025mmol) and Pd(OAc)2 (44 mg, 0.18 mmol) were dissolved in CH3CN (30 mL). The reaction mixture was stirred at 60 °C for 30 min; the color of the solution changed to red after a few minutes. The sol vent was evaporated under vacuum and the resulting residue was partitioned between ethyl acetate and water. The organic layer was concen-trated under vacuum, and the residue was dissolved in THF (10 mL). DDQ (200 mg, 0.8 mmol) was added and the mixture was refl uxed for 1.5 h. The color of the solu-tion changed from red to green, and the reaction progress was monitored by UV-vis. After cooling to rt, ethyl ace-tate (100 mL) was added and the mixture was washed with water, 0.1 M HCl, and brine to remove excess of DDQ. The organic layer was dried and purifi ed by silica gel column chromatography using 50% of ethyl acetate in dichloromethane for elution. The fi rst green band was collected and dried to give porphyrin 8 as a green solid(25.5 mg) in 41% yield. 1H NMR (acetone-d6, 300 MHz): δ, ppm 9.03 (4H, s), 8.55 (8H, d, J = 7.89 Hz), 8.37 (8H, d, J = 7.73 Hz), 7.32 (8H, br s), 7.32 (8H, br s), 6.42 (4H, d, J = 9.30 Hz), 5.91 (4H, t, J = 9.69 Hz), 5.65 (4H, d, J= 2.44 Hz), 5.56 (4H, dd, J = 10.24, 3.34 Hz), 4.73–4.77 (4H, m), 4.21–4.36 (8H, m), 2.26 (12H, s), 2.09 (12H, s), 2.06 (12H, s), 2.01 (12H, s). MALDI-TOF: m/z 2507.67 [M + H]+. Calcd. for C124H123N16O36Pd 2507.65.

TBP-carbohydrate 9. Porphyrin 8 (18.5 mg, 0.007 mmol) was dissolved in methanol (10 mL). 20 L of a sodium methoxide solution 0.5 M in methanol was added. The fi nal mixture was stirred at rt for 1 h. Af ter solvent evaporation under vacuum, the crude pro duct was purifi ed by gel fi ltration on a Sephadex LH20 column eluted with methanol. After recrys tal li za tion from methanol/water, porphyrin 9 (11.3 mg) was ob tained in 89% yield. MAL-DI-TOF: m/z 1836.71 [M + H]+, calcd. for C92H80N16O20-

Pd 1837.14. Anal. calcd. for C92H79N16O20Pd.3H2O: C, 58.49; H, 4.54; N, 11.86; found C, 58.40; H, 4.60; N, 11.81.

Cell culture

All cell culture media and reagents were obtained from Invitrogen. Human carcinoma HEp2 cells were


54 E. HAO ET AL.

obtained from ATCC and maintained in a 50:50 mixtureof DMEM:Advanced MEM containing 5% FBS. The cells were sub-cultured biweekly to maintain sub-confl uent stocks.

Time-dependent cellular uptake

The HEp2 cells were plated at 7500 cells/well in 96 well plates and allowed to grow for 48 h. The por phyrin-carbohydrate conjugates were dissolved in Cre mophor/DMSO to make 100 × stock solutions be fore diluting in medium to a fi nal concentration of 10 M of porphyrin, 0.9% DMSO and 0.1% Cre mo phor. The cells were exposed to each conjugate for 0, 1, 2, 4, 8, and 24 h. The cellular uptake was stopped by removing the loading medium, washing with PBS and solubilizing the cells with 0.25% Tri ton X-100 in PBS. The internal porphyrin concentra tion was determined by measuring the fl uores-cence at 410/650nm using a FLORstar Plate reader. The cell numbers were quantifi ed using CyQuant and measu-ring at 480/520nm.

Microscopy

The HEp2 cells were plated on two Chamber co verslips and grown for 48 h. 10 M of each porphy rin-carbohydrate conjugate was added to the cells and incu-bated overnight. Individual organelles were identifi ed using Molecular Probe organelle stains from Invitro-gen (MitoTracker 100 nM for 30 min for mitochondria, LysoSensor 50 nM for 30 min for lysosomes, ERTracker 2 mM for 60 min for ER, and BODIPY Ceramide 250 nM for 30 min for Golgi Com plex). After staining, the cells were washed three times with medium and then twice with medium plus 50 mM HEPES pH 7.2 and examined using a Zeiss AxioVert 200M inverted microscope fi tted with standard FITC and Texas Red fi lter sets (Chroma). The images were acquired with a Zeiss Axiocam MRM CCD camera fi tted to the microscope.


Synthesis

Mono-substituted TPP-carbohydrate conjugates 3 and 4 were prepared as shown in Scheme 1, from com merciallyavailable meso-tetra (4-hy droxy phenyl) por phyrin 1. Mono-alkylation and zinc(II) metalation of porphyrin 1 gave a pro pargy loxy-terminated Zn(II)-porphyrin which reac-ted under typical “click” che mistry conditions with azi-de-containing galac tose 2a, in the presence of CuSO4

and sodium ascor bate in tBuOH/H2O 1:1. This reaction was performed at 70 °C for 24 h and gave TPP-galactose conjugate 3 in 92% yield. Interestingly, the correspon-ding free-base propargyloxy-terminated porphyrin failed to genera te the desired target conjugate under the same reaction conditions. The observed reactivity difference between the free-base and Zn(II)-complexed porphyrin

might result from copper insertion into the free-base macrocycle, rendering it unreactive under the mild reac-tion conditions. In any case, the free-base carbohydrate conjugates could be obtained by facile demetalation of the corresponding Zn(II) complexes. Furthermore, we discovered that a reaction temperature of at least 50 °C was essential to drive the “click” chemistry reaction to completion. Using similar reaction conditions but with azide-containing lactose 2b instead of galactose 2a, TPP-lactose conjugate 4 was obtained in 88% yield. It is worth noticing that the un protected carbohydrate 2b is able to survive these mild reaction conditions. The structures of conjugates 3 and 4 were confi rmed by 1H NMR spectros-copy and high-resolution MALDI-TOF mass spectro-metry, the latter showing the expected molecular ions at m/z 1006.2195 (calcd. 1006.2155) and 1168.2697 (calcd. 1168.2683) for porphyrins 3 and 4, respectively.

In order to extend the “click” chemistry to the pre-paration of tetra carbo hydrate-por phyrin conjuga tes, the Zn(II) complex of meso-tetra(4-propargyloxyphenyl)porphyrin was prepared by condensation of 4-pro pargyl-oxy benzal de hyde with pyrrole under Lindsey condi-tions [43] followed by zinc insertion. However, the tetra-conjugation of this por phyrin and galactose 2a under the “click” conditions des cri bed above gave a mixture of partially conju gated pro ducts, and the use of various reaction catalysts, rea gent ratios, reaction times and tem-peratures, also failed to drive the reaction to completion. We attemp ted to isolate the major reaction product from the “click” reaction above, and from a similar reaction using the ester-protected carbohydrate 2c in place of 2a,but the purifi cation was diffi cult.

An alternative route was therefore developed for the preparation of tetra carbo hydrate-por phy rin con ju gates, as shown in Scheme 2. The carbohydrate-con taining benzal-dehyde 5 was synthe sized in 93.4% yield from the “click” reaction between 2c and 4-ethy nyl benzal de hyde in CH3CN at rt. The opti mized conditions used CuI as the catalyst and DIPEA as the base, at rt overnight. Under typical “click” chemistry conditions (tBuOH/H2O 1:1, CuSO4

and sodium as corbate) at 60 °C for two days, the desired product 5 was obtained in lower (72.6%) yield. Conden-sation of benzal dehyde 5 with pyrrole under Lindsey conditions [43] gave the corresponding tetracarbohydra-te-porphyrin in 21% yield. Benzal dehyde 5 also reacted with tetra hydro iso indole [41] giving the corres pondingporphyrin in 41% yield. The structures of these porphy-rins were confi rmed by 1H and 13C NMR spectros copy, and by MALDI-TOF mass spectrome try. The insertion of palla dium(II) and subsequent oxi dation with DDQ in THF using the li terature procedures [41, 44], afforded the target TBP con jugate 8 in a combined 49% yield. On the other hand, the DDQ oxidation of the free-base porphyrin followed by metalation gave only traces of the targeted TBP 8, due to competing macrocycle oxidations and opening, as we have previously reported [45]. During the me ta lation reac tion in the presence of palladium(II)



NH

N

N

HN

N

N

N

NO

3. R-N3, Cu(I), tBuOH/THF

1. , K2CO3

N NN R

OH

OH

OH

HO

OH

HO

OH

Zn

Br

2. Zn(OAc)2, CH3OH

O

HO

H

H

HO

H

HOHH

OH

OHO

H

HHO

H

HOH

H

OHOH

O

H

HO

H

HOHH

HO

R =

1

2a

2b

3: R = 2a4: R = 2b

N

N

N

NPd

OAcO

H

H

AcO

H

HOAcH

OAc

N3

CHO

CuI, CH3CN

CHO

N

N

N

NZn

1. pyrrole, BF3·OEt22. DDQ

4. Zn(OAc)2

1. tetrahydroisoindole, BF3·OEt22. DDQ3. Pd(OAc)2 followed by DDQ

3. 0.5M CH3ONa/CH3OH4. 0.5M CH3ONa/CH3OH

5: R = 2c

O

AcO

H

H

AcO

H

HOAcH

OAc

R =

2c

O

HO

H

H

HO

H

HOHH

OH

2a

6: R = 2c7: R = 2a

8: R = 2c9: R = 2a

NNN

R

NN

N

NN

N

NN

N

NN

N

NNN

NN

N

NN N

NNN

R

R

R

R

R

R

R

R

Scheme 1.

Scheme 2.


56 E. HAO ET AL.

ace tate we ob served partial cleavage of the ester protec-ting group of the sugar moiety, with generation of polar green by-products. The complete and effi cient deprotec-tion of the ester groups was achieved using 0.5 M sodium methoxide in methanol, giving conjugates 7, after zinc metalation, and 9 in 91% and 89% yields, respectively.

Cellular studies

The time-dependent uptake of TPP-carbohydrate conjugates 3, 4 and 7, and of TBP conjugate 9, by hu man carcinoma HEp2 cells was investigated at a concentration of 10 M and the results obtained are shown in Fig. 1. All conjugates displayed similar up take kinetics showing a rapid uptake at low time points (< 2 h) after which a Fig. 1. Time-dependent uptake of porphyrin-carbohydrate

conjugates 3 (black), 4 (red), 7 (green) and 9 (blue) at 10 Mby human carcinoma HEp2 cells

Fig. 2. Subcellular localization of porphyrin-carbohydrate 3in human HEp2 cells at 10 M for 24 h. (a) Phase contrast, (b) overlay of porphyrin 3 fl uorescence and phase contrast, (c) BODIPY Ceramide fl uorescence, (e) LysoSensor Green fl uo-rescence, (g) MitoTracker Green fl uorescence, (i) ERTracker fl uorescence, (d), (f), (h), (j) overlays of organelle tracers with porphyrin 3 fl uorescence. Scale bar: 10 m




plateau was reached. The TBP conjugate 9 accumulatedwithin cells to a much higher extent than the TPP conju-gates, approximately 5 times more at all time points in-vestigated. Among the TPP conjugates, the TPP-galactose 3 accumula ted the most within the HEp2 cells followed by the lactose conjugate 4 and the tetra-galactose conju-gate 7. The cellular uptake of this series of compounds is therefore dependent on the structure of the conjugates andtheir hydrophobic character. TBP conjugate 9, the conju-gate most taken up by cells, due to the presence of the four

, ’-fused benzene rings, is more hydrophobic than the corresponding TPP conju ga te 7, which accumulated the least within cells. On the o ther hand, the most hydropho-bic TPP conjugate 3 ac cumulated the most within cells

of all the TPP con jugates. Although there was a slight difference in the uptake of the TPP-galactose conjugate 3 compared with the corresponding lactose conjugate 4,this difference was not signifi cant and probably refl ects the abi lity of these conjugates to bind to specifi c cell re-ceptors rather than their hydrophobic character. No ne ofthe conjugates was toxic to the HEp2 cells at concentra-tions up to 100 M (results not shown).

The intracellular localization of conjugates 3, 4, 7 and 9 was evaluated by fl uorescence microscopy, as can be seen in Figs. 2–5. The organelle-specifi c fl uo res cent pro-bes MitoTracker Green, LysoSensor Green, ERTracker Blue and BODIPY Ceramide were used to identify the cell mitochondria, lysosomes, ER, and Golgi Complex,


Fig. 5. Subcellular localization of porphyrin-carbohydrate 9in human HEp2 cells at 10 μM for 24 h. (a) Phase contrast, (b) overlay of porphyrin 9 fl uorescence and phase contrast, (c) BODIPY Ceramide fl uorescence, (e) LysoSensor Green fl uo-rescence, (g) MitoTracker Green fl uorescence, (i) ERTracker fl uorescence, (d), (f), (h), (j) overlays of organelle tracers with porphyrin 9 fl uorescence. Scale bar: 10 μm


58 E. HAO ET AL.

respectively, and to perform co-loca lization investigations by overlay with the con ju gate fl uo rescence. Whereas the TPP conjugates 3 and 4 were found mainly in the ER and endo somes, the tetra-galactose conjugates 7 and 9 loca-lized pre fe rentially within the lysosomes. In agreement with these results, the preferential site of intracellular localization of mTHPC [meso-tetra (3-hydro xy phe nyl)-chlorin] is the ER [46], and those of a water-soluble nido-carbo ranyl-TBP are the cell lyso somes [41].

Acknowledgements

This research was partially funded by the US Na tional Institutes of Health, grant number CA098902.

REFERENCES

1. Dougherty TJ, Gomer CJ, Henderson BW, Jori G, Kessel D, Korbelik M, Moan J and Peng Q. J. Natl. Cancer Inst. 1998; 90: 889–905.

2. Ali H and van Lier J. Chem Rev. 1999; 99: 2379–2450.

3. Pandey RK and Zhang G. In The Porphyrin Hand-book, Smith KM, Kadish KM and Guilard R. (Eds.). Academic Press: Boston, 2000; 6: 157–230.

4. Vicente MGH. Curr. Med. Chem. Anti-Cancer Agents 2001; 1: 175–194.

5. Brown S, Brown EA and Walker I. Lancet Oncol.2004; 5: 497–508.

6. Oku N, Saito N, Namba Y, Tsukada H, Solphin D and Okada S. Biol. Pharm. Bull. 1997; 20: 670–673.

7. Osterloh J and Vicente MGH. J. Porphyrins Phthalocyanines 2002; 6: 305–324.

8. Gijsens A, Missiaen L, Merlevede W and de Witte P. Cancer Res. 2000; 60: 2197–2202.

9. Sibrian-Vazquez M, Jensen TJ, Hammer RP and Vicente MGH. J. Med. Chem. 2006; 49: 1364–1372.

10. Mestre B, Pitie M, Loup C, Claparols C, Pratviel G and Meunier B. Nucl. Acids Res. 1997; 25: 1022–1027.

11. Li H, Fedorova OS, Trumble WR, Fletcher TR and Czuchajowski L. Bioconjugate Chem. 1997; 8:49–56.

12. Del Governatore M, Hamblin MR, Piccinini EE, Ugolini G and Hasan T. Br. J. Cancer 2000; 82:56–64.

13. Hudson R, Carcenac M, Smith K, Madden L, Clar-ke OJ, Pelegrin A, Greenman J and Boyle RW. Br. J. Cancer 2005; 92: 1442–1449.

14. Zhang X and Pandey RK. Anti-Cancer Agents Med. Chem. 2008; 8: 241–268.

15. (a) Hombrecher HK, Schell C and Thiem J. Bioorg. Med. Chem. Lett. 1996; 6: 1199–1202. (b) Cornia M, Valenti C, Capacchi S and Cozzini P. Tetrahe-dron 1998; 54: 8091–8106. (c) Li HP. Bioorg. Med. Chem. Lett. 2006; 16: 6298–6301.

16. (a) Oulmi D, Maillard P, Guerquin-Kern JL, Huel C and Momenteau M. J. Org. Chem. 1995; 60:1554–1564. (b) Sol V, Blais JC, Carre V, Granet R, Guilloton M, Spiro M and Krausz PJ. Org. Chem.1999; 64: 4431–4444.

17. (a) Pasetto P, Chen X, Drain CM and Frank RW. Chem. Commun. 2001; 81–82. (b) Stepek P, Dukh M, Saman D, Moravcova J, Kniezo L, Monti D, Venanzi M, Mancini G and Drasar P. Org. Biomol. Chem. 2007; 5: 960–970.

18. Fuhrhop JH, Demoulin C, Boettcher C, Koening J and Siggel U. J. Am. Chem. Soc. 1992; 114: 4159–4165.

19. (a) Chen X, Hui L, Foster DA and Drain CM. Bio-chemistry 2004; 43: 10918–10929. (b) Samaroo D, Vinodu M, Chen X and Drain CM. J. Comb. Chem.2007; 9: 998–1011.

20. (a) Laville I, Pigaglio S, Blais JC, Doz F, Loock B, Maillard P, Grierson DS and Blais J. J. Med. Chem.2006; 49: 2558–2567. (b) Asayama S, Mizushima K, Nagaoka S and Kawakami H. Bioconjugate Chem.2004; 15: 1360–1363.

21. (a) Zheng G, Graham A, Shibata M, Missert JR, Oseroff AR, Dougherty TJ and Pandey RK. J. Org. Chem. 2001; 66: 8709–8716. (b) Li G, Pandey SK, Graham A, Dobhal MP, Mehta R, Chen Y, Gryshuk A, Rittenhouse-Olson K, Oseroff A and Pandey RK. J. Org. Chem. 2004; 69: 158–172.

22. (a) Silva AMG, Tome AC, Neves MGPMS, Silva AMS, Cavaleiro JAS, Perrone D and Dondoni A. Tetrahedron Lett. 2002; 43: 603–605. (b) de C da SF, Ferreira VF, de Souza MCBV, Tome AC, Neves MGPMS, Silva AMS and Cavaleiro JAS. Synlett.2008; 8: 1205–1207.

23. Sharon N and Lis H. Science 1989; 246: 227–234.24. Hakomori SI. Adv. Cancer Res. 1989; 52: 257–332.25. Kim YS, Gum J, Brouckhausen I. Glycoconjugate

J. 1996; 13: 693–707.26. Chiatriotti L, Berlingieri MT and Battaglia C. Int. J.

Cancer 1995; 64: 171–175.27. Lotan R, Belloni PN and Tressler RJ. Glycoconju-

gate J. 1994; 11: 462–468.28. Kolb HC, Finn MG and Sharpless KB. Angew.

Chem. Int. Ed. 2001; 40: 2004–2021.29. Rostovtsev VV, Green LG, Fokin VV and Shar-

pless KB. Angew. Chem. Int. Ed. 2002; 41: 2596–2599.

30. Liu Y, Ke C-F, Zhang H-Y, Cui J and Ding F. J. Am. Chem. Soc. 2008; 130: 600–605.

31. Marois J-S, Cantin K, Desmarais A and Morin J-F. Org. Lett. 2008; 10: 33–36.

32. Hasegawa T, Umeda M, Numata M, Li C, Bae A-H, Fujisawa T, Haraguchi S, Sakurai K and Shinkai S. Carbohydrate Res. 2006; 341: 35–40.

33. Punidha S, Sinha J, Kumar A and Ravikanth M. J. Org. Chem. 2008; 73: 323–326.



34. Decréau RA, Collman JP, Yang Y, Yan Y and Deva-raj NK. J. Org. Chem. 2007; 72: 2794–2802.

35. Lavi A, Johnson FM and Ehrenberg B. Chem. Phys. Lett. 1994; 231: 144–150.

36. Wolford ST, Novicki DL and Kelly B. Fundam.Appl. Toxicol. 1995; 24: 52–56.

37. Rogers JE, Nguyen KA, Hufnagle DC, McLean DG, Su W, Gossett KM, Burke AR, Vinogradov SA, Pachter R and Fleitz PA. J. Phys. Chem. A 2003; 107: 11331–11339.

38. Vicente MGH and Smith KM. J. Porphyrins Phthalocyanines 2004; 8: 26–42.

39. Vinogradov SA and Wilson DF. J. Chem. Soc. Per-kin Trans. 2 1995; 1: 103–112.

40. Murashima T, Tsujimoto S, Yamada T, Miyazawa T, Uno H, Ono N and Sugimoto N. Tetrahe dron Lett.2005; 46: 113–116.

41. Gottumukkala V, Ongayi O, Baker DG, Lomax LG and Vicente MGH. Bioorg. Med. Chem. 2006; 14:1871–1879.

42. Sibrian-Vazquez M, Hao E, Jensen TJ and Vicente MGH. Bioconjugate Chem. 2006; 17: 928–934.

43. Lindsey JS, Schreiman IC, Hsu HC, Kearney, PC and Marguerettaz AM. J. Org. Chem. 1987; 52:827–836.

44. Finikova OS, Cheprakov AV, Beletskaya IP, Carroll PJ and Vinogradov SA. J. Org. Chem. 2004; 69:522–535.

45. Ongayi O, Fronczek FR, Vicente MGH. Chem.Commun. 2003: 2298–2299.

46. Kessel D. J. Porphyrins Phthalocyanines 2004; 8:1009–1014.




INTRODUCTION

In earlier work, Milgrom and coworkers demon-strated that attaching antioxidant butylated hydroxy toluene groups to the porphyrin macrocycle through its meso positions yields unusual redox activity of the resulting phenol-substituted compounds [1]. In particu-lar, 5,10,15,20-tetrakis(3,5-di-t-butyl-4-hydroxyphenyl) porphyrin, TDtBHPP, undergoes a 2-electron oxidation to a stable quinonoid form 5,10,15,20-tetrakis(3,5- di-tert-butyl-4-oxocyclohexadien-2,5-ylidene)-porphyrinogen, Ox[TDtBHPP], abbreviated to OxP [2]. There also exists a stable intermediate semiconducting porphyrin cation radical state, PH3

+• [3]. These three forms each possess unique properties, thus composing a neat tri-stable switch based on the porphyrin chromophore, and which can be most conveniently operated in the solution state. While the porphyrinic state TDtBHPP behaves as an arche-typal tetraphenylporphyrin in neutral solution, and in its

metal complexing abilities, the oxoporphyrinogen OxP possesses different reactivities, especially of the central cavity of the molecule which is much more accessible to an incoming reagent than is the porphyrin. Thus, OxP can be repeatedly and regioselectively N-alkylated, leading to a series of compounds containing the OxP chromophore [4]. The OxP chromophore is electron-defi cient, which suggests its use as a mediator in electron transfer pro-cesses, and also contains site(s) for binding of guests through hydrogen bonding, by analogy with the well-known anion binding porphyrinogens of Sessler, Gale and coworkers [5]. Our original interest in this system stemmed from the aforementioned tri-stable chemical switching accessible in the solution state, but our work on the N-alkylation reaction has resulted in the prepara-tion of many potential host compounds and oligochro-mophores and some of these are described below.

STRUCTURE

X-ray crystallography of the N-alkylated OxP com-pounds is indispensable for their characterization, especially given the possibility of structural isomerism in these compounds. Initial work on the parent porphyrin

Structures and properties of hemiquinone-substituted

oxoporphyrinogens

Jonathan P. Hill*a , Katsuhiko Arigaa and Francis D’Souza*b

a World Premier International (WPI) Research Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), 1-1 Namiki, Tsukuba 305-0044, Japanb Department of Chemistry, Wichita State University, 1845 Fairmount, Wichita, Kansas 67260-0051, USA


ABSTRACT: The structure and physical properties of a series of N-substituted, hemiquinone-substituted oxoporphyrinogens is presented and discussed. Structures of the compounds are dictated by the nature of the substituent, with the substitution pattern being in turn dictated by regioselectivity of N-alkylation. X-ray crystallography and other aggregation properties of the compounds are discussed. Redox reactions are also strongly infl uenced by N-substitution and substituent identity. Also presented are properties related to guest binding and photophysical properties of oligochromophoric host-guest complexes, involving oxoporphyrinogen, N-substituted with porphyrins and appropriately substituted fullerene guest electron acceptors.

KEYWORDS: porphyrin, oxoporphyrinogen, electrochemistry, oligochromophore.


*Correspondence to: Jonathan P. Hill, email: [email protected], fax: +81 29-860-4832 and Francis D’Souza, email: [email protected], fax: +1 316-978-3431


STRUCTURES AND PROPERTIES OF HEMIQUINONE-SUBSTITUTED OXOPORPHYRINOGENS 61

and oxoporphyrinogen OxP revealed the structures shown in Fig. 1a,b [2, 6]. The differing form of the compounds is obvious, with key features being the porphyrinogen-type conformation of the pyrrole rings in OxP and the

oxidation-state dependence of molecular shape. N-alky-lation of OxP occurs without altering the conformation of the OxP macrocycle and leads fi rst to cofacial N21,N23-dialkyl (Fig. 1c), then tetraalkyl (Fig. 1d) compounds,

Fig. 1. X-ray crystal structures of (a) 5,10,15,20-tetrakis(3,5-di-t-butyl-4-hydroxyphenyl)porphyrin, TDtBHPP, (b) 5,10,15,20-tetrakis(3,5-di-tert-butyl-4-oxocyclohexadien-2,5-ylidene)-porphyrinogen, Ox[TDtBHPP], (c) OxPBz2, and (d) OxPBz4. Chemical structures are shown below each structure

HN

NH

HNNH

O

OO

O

(b)

N

N

HNNH

OH

OHHO

HO

(a)

HN

NH

NN

O

OO

O

(c)

N

N

NN

O

OO

O

(d)


62 J.P. HILL ET AL.

Fig. 2. X-ray crystal structure of N,N′,N′′,N′′′-tetrakis (n-hexadecyl)-OxP. Chemical structure is similar to that of OxPBz4

shown in Fig. 1(d) but with n-hexadecyl groups replacing ben-zyl groups; (a) relative dispositions of the n-hexadecyl chains viewed parallel with the average macrocyclic plane, (b) pack-ing of the molecules illustrating the segregation of alkyl chains and OxP molecules (grey dumbell: dimer-like stacks of the OxP moieties; grey square: alkyl chain packing region)

(a)

(b)

as the main products [7]. Mono N-alkyl compounds are easily accessible by varying reaction times while tri-N-alkylates are available by varying the base used in the N-alkylation reaction. The predominance of the dialkyl and tetraalkyl compounds is thought to arise from the anion-binding properties of the OxP macrocycle and this can be seen as amine-group protection through anion complexation. The relatively poor anion-binding ability of a single NH group also explains the low yield of the tri-substituted compounds when inorganic bases are used in the N-alkylation reaction.

Intermolecular interactions in compounds with simple alkyl groups are generally weak, leading to unstable crystals and lower-quality X-ray diffraction data. For tetra-N-substituted compounds, their symmetry tends

to improve this situation. Thus, we were able to obtain structural data for the tetra-n-hexadecyl-OxP [8], tetrakis(4-bromobenzyl)-OxP [9] and tetrakis(2-methyl-naphthyl)-OxP molecules [10]. Tetra-n-hexadecyl-OxP has an attractive packing motif in which pairs of mol-ecules form stacks, which are further arranged in a hatch-like structure with n-hexadecyl chains packed in the intervening voids (Fig. 2). Tetrakis(2-methylnaphthyl)-OxP contains an interesting intermolecular – stackingmotif which causes the molecules to be arranged into one- dimensional arrays within the crystals (Fig. 3).

Increasing the complexity of the N-substituent leads to a corresponding increase in diffi culty of crystal growth. Thus, while the tetrakis(2-methylpyrenyl)-OxP [11] does crystal-lize from acetone solution, the crystals are of insuffi cient size for analysis. In compounds bearing larger N-substituents, crystals cannot be obtained and aggregates are more likely formed in the case of higher oligochromophores. An exam-ple of this can be seen in the fi ber formation by a bis-por-phyrinyl-substituted OxP [8]. Scanning electron microscopic imaging of material prepared by a solvent diffusion method revealed organogel-like aggregates of N21,N23-bis(5,10,15-tris (4-t-butylphenyl)-20-(4-methylphenyl)porphyrinyl)-OxP whose chemical structure and a proposed model of

Fig. 3. One-dimensional array of oxoporphyrinogens con-tained in the crystal structure of tetrakis(2-methylnaphthyl)-OxP. Chemical structure is similar to that of OxPBz4 shown in Fig. 1(d) but benzyl groups are replaced by 2-methylnaphthyl groups



the aggregate assembly mechanism is shown in Fig. 4. Essentially, the mechanism of aggregation is thought to be similar to that in the crystals of tetrakis(2-methyl naphthyl)-OxP where N- substituents are involved in intermolecular – stacking interactions except that, whereas tetrakis(2-

methylnaphthyl)-OxP has naphthyl groups on opposing sides of the molecule stacked, porphyrinyl groups attached to the same side of the molecule interact. A similar arrange-ment has been observed in aggregates of merocyanine dyes [12]. With respect to aggregate formation, the OxP moi ety is a rather unwieldy unit and possesses no points for signifi -cantly strong intermolecular interactions. This has hampered

our efforts in preparing structured aggregates, although we are currently undertaking work which has permitted us to remedy this problem.

ELECTROCHEMISTRY

We will discuss the electrochemical and spectroelectro-chemical properties of the 2-methylnaphthyl-substituted OxP compounds [10] since the whole family of those N-alkylates is available. The cyclic voltammograms of the porphyrinogens OxP and OxP(Np1–4) are shown in Fig. 5, together with that of the porphyrin parent, TDtBHPP. For tetra-oxocyclohexadienylidene porphyrinogen, OxP, bearing no N-substituents, the cyclic voltammogram contains two reversible oxidations located at E1/2 = 0.27 and 0.48 V vs Fc/Fc+ and an irreversible reduction at Epc = -1.33 V vs Fc/Fc+. A two-electron reduction process is suggested for the latter process since doubling the cur-rent of the fi rst oxidation is required. Varying scan rate, multicyclic, and low temperature (0 °C) voltammetric studies did not signifi cantly affect the irreversibility of the reduction wave. Two more poorly defi ned reductions at E1/2 = -1.81 and -2.18 V were found upon scanning the potential to -2.5 V. Compounds OxP(Np1–4) with an increasing number of 2-methylnaphthyl groups at the macrocyclic nitrogens also exhibit reversible oxidation processes, while at the same time undergoing a system-atic anodic shift of their oxidation potentials.

Reduction of the N-substituted porphyrinogens results in several further interesting observations. For OxP, reduction is an irreversible two-electron process, while the reductions of compounds OxP(Np1–4) become increas-ingly reversible, with the initial two-electron process being resolved into two reversible one-electron processes for tetra-N-substituted compound, OxP(Np4). Two more one-electron reductions were revealed by scanning the poten-tial further in the cathodic direction, and these reductions also become reversible, with increase of the multiplicity of N-substituents on the macrocycle. This wave involves two one-electron processes for compound OxP(Np4). The fi rst reduction process occurs at similar potentials for all the derivatives. Conversely, the potentials of oxidation processes are anodically shifted by 70–130 mV for each additional N-substituent, indicating an increasing stabili-zation of the HOMO level upon N-substitution [7]. Previ-ous ab initio calculations suggest a stabilized HOMO level for these N-substituted compounds. As a result, the experi-mental HOMO-LUMO gap, calculated from the potential difference between the fi rst oxidation and fi rst reduction, follow a trend in which the HOMO-LUMO gap increases with increasing N-substitution, as predicted by earlier computational studies [7].

Spectroelectrochemistry

UV-vis spectral changes observed during the redox processes of N-naphthyl derivatives of OxP in

NH

HN

N

NO

O

O

O

NN

NNZn

NN

NN Zn

(a)

(b)

(c)

(d)

Fig. 4. Aggregate formation for (TPP)2OxP. (a) Chemical struc-ture (TPP)2OxP, (b) calculated structure of (TPP)2OxP (MM2), (c) scanning electron micrograph of aggregates of (TPP)2OxPformed in chloroform/methanol mixtures and (d) plausible mechanism for self-assembly of (TPP)2OxP


64 J.P. HILL ET AL.

o-dichlorobenzene containing 0.1 M (TBA)ClO4 are presented in Fig. 6. Spectral changes during the fi rst oxi-dation of compounds OxP(Np1–4) revealed a reduction in intensity of the absorption maximum located at ~ 500 nm with the appearance of new bands in the visible or near-IR region, characteristic of the formation of a π-cationradical species. The presence of two species at equilibri-um in solution is indicated by observation of one or more isosbestic points.

These spectral changes were found to be reversible. The wavelength of the band corresponding to the for-mation of the π-cation radical is dependent upon the N-substituent multiplicity of the porhyrinogen macro-cycle. The newly developed, low-intensity absorption band undergoes a systematic red shift and, for OxP(Np3)and OxP(Np4), the new absorption band is split into two peaks at 717 nm and 910 nm, respectively.

Spectral changes observed during the fi rst reduction of compounds OxP and OxP(Np1) are irreversible, in accord with the cyclic voltammetric observations, although new absorption peaks emerge during this process. Conversely, for compounds OxP(Np2–4), the spectral changes are near-ly reversible. Since the fi rst reduction is either overlapped or located close to the second reduction process for the compounds studied, a potential of 130 mV past the fi rst/second reduction process was applied for spectroelec-trochemical measurements. Under these conditions, two sets of spectral changes were observed, presumably cor-responding to the one- and two-electron reduced prod-ucts. The spectral changes shown in Fig. 6 correspond to the fi rst spectral changes (i.e. fi rst reduction). Under these conditions, the peak position of the emerging band, which corresponds to the formation of the porphyrino-gen π–anion radical (i.e. porphyrin π–cation radical), was found to depend on the number of N-substituents at the porphyrinogen macrocycle, an observation similar to that observed for the π–cation radical species.

ANION BINDING AND DETECTION

N-alkyl OxP compounds, such as OxPBz2, can bind anionic species including the halide anions and other more complex anions [13]. Furthermore, there is some specifi city in the chromogenic response elicited from OxP and its derivatives in the presence of certain anions. Thus, fl uoride ions can be distinguished from the other halides by the more substantial change in color that occurs upon fl uoride ion-binding. This difference is even more substantial in spectrophotometric terms with fl uo-ride ion binding resulting in a strong absorption band at longer wavelength. In addition, derivatives of OxP exhibit solvatochromism, allowing some chromogenic distinc-tion between other ions, such as acetate or cyanide, and fl uoride. Both solvatochromic and anion-binding effects are infl uenced by the hydrogen-bonding ability of the pyrrolic NH groups so that increased H-bonding between

Fig. 5. Cyclic voltammograms of (a) OxP, (b) OxPNp1, (c) OxPNp2, (d) OxPNp3, (e) OxPNp4, (f) TDtBHPP. Np = (2- methylnaphthyl). Solvent/electrolyte: 1,2-dichlorobenzene/0.1

M (TBA)ClO4. Scan rate = 100 mV.s

-1

(a)

(b)

(c)

(d)

(e)

(f)

N

N

R2NNR1

O

OO

O

R3

R4

OxPNp1: R1 = 2-methylnaphthyl, R2 = R3 = R4 = HOxPNp2: R1 = R2 = 2-methylnaphthyl, R3 = R4 = HOxPNp3: R1 = R2 = R3 = 2-methylnaphthyl, R4 = HOxPNp4: R1 = R2 = R3 = R4 = 2-methylnaphthyl.



OxP, OxPBz2 and the solvent obscures the effect of anion binding. Hydrogen bonding by OxPBz2 to solvent ace-tone molecules was observed in the crystal structure of that solvate. Furthermore, OxP tends to undergo a tau-tomerism which can be weighted in favor of the porphy-rinogen form by complexation with appropriate anions. Solvatochromism and anion-binding behavior of OxPBz2

are illustrated in Fig. 7.The oxoporphyrinogen receptor investigated, OxPBz2,

acts as an excellent electrochemical anion recognition probe as revealed by the large anion-dependent electro-chemical responses for F-, NO3

-, CH3COO-, H2PO4-, and

ClO4- in o-dichlorobenzene [14]. The magnitude of the

potential shift was found to depend upon the nature of the anion and correlated well with the anion-binding con-stants. That is, anions with higher binding constants gave pronounced cathodic shifts of the oxidation potentials of the oxoporphyrinogen. The large potential shift upon anion binding is attributable to the nature of the recep-tor in which the anion binding site and the macrocylic

-system are integral parts.

BUILDING DONOR-ACCEPTOR CONJUGATES USING PORPHYRINS

The fi rst example of a porphyrin-quinonoid donor-acceptor triad featuring tetraphenylporphinatozinc(II) moieties covalently attached to an oxoporphyrinogen through its macrocyclic nitrogen atoms was reported [15]. This arrangement of chromophores resulted in an interesting interplay between the electron-donating zinc porphyrin(s) and the electron/energy accepting oxopor-phyrinogen. To describe these compounds, we initially used the term P-OxP, i.e. Porphyrin-OxoPorphyrinogen.Later, we used the term alligator porphyrins to describe

the clip-like disposition of two porphyrin units when cofacially linked to the oxoporphyrinogen unit. Since direct structural data was often diffi cult to obtain for these compounds, we used computational methods to probe the molecular and electronic structures of the OxP derivatives [15, 16]. In fact, we used these methods to predict electrochemical and other properties because

Fig. 7. (a) X-ray crystal structure of the acetone solvate of OxP-Bz2 illustrating H-bonding by the pyrrolic NH groups. (b) Spec-tral changes occurring upon titration of OxPBz2 with fl uoride anions. Inset: Benesi-Hildebrand plot for the binding constant

(a)

(b)

Fig. 6. Spectral changes accompanying the fi rst oxidations (1st row) and fi rst reductions (2nd row) of the N-alkylated derivatives of OxP. Solvent: 1,2-dichlorobenzene containing 0.1 M (TBA)ClO

4. A potential of 130 mV past the fi rst oxidation/reduction process

was applied for the spectroelectrochemical measurements


66 J.P. HILL ET AL.

initial studies on the N-alkylated OxP derivatives gave an excellent agreement between measured and calculated quantities (such as HOMO-LUMO gap, etc.) [7]. These methods also permit us to visualize the likely geometries and electronic structures of the oligochromophoric supra-molecular hosts. For example, the computed structure of one of the bis-porphyrin-linked OxP compounds is shown in Fig. 8. In addition, the locations of the HOMO and LUMO for the complex are indicated. This data helps us predict the likelihood of electron transfer and/or energy transfer events. The optical absorption of the triad reveals features corresponding to both the donor and acceptor entities. The electrochemical redox states of the triad were established from a comparative electrochemistry of

the triad and the reference compounds. Both steady-state and time-resolved emission studies revealed quench-ing of the singlet excited state of zinc porphyrin in the triad; free-energy calculations performed using Weller’s approach indicate the possibility of electron transfer from the singlet excited zinc porphyrin group to the oxopor-phyrinogen in polar solvents. Time-resolved fl uorescence studies reveal excited state energy transfer from zinc porphyrin to oxoporphyrinogen in non-polar solvents, while nanosecond transient absorption studies combined with time-resolved fl uorescence studies in polar solvents are indicative of the occurrence of photoinduced charge separation from the singlet excited zinc porphyrin to the oxoporphyrinogen.

HN

NH

NN

O

OO

O

N

N

N

NZn

O

O

O

O

N

N

N

N Zn

O

O

O

O

(a)

Fig. 8. (a) Structure of (TPPZn)2-OxP, and the frontier orbitals (b)(i) HOMO and (b)(ii) LUMO calculated at the B3LYP/3-21G* level

(b)



SUPRAMOLECULAR COMPLEXES OF PORPHYRIN-OXP CONJUGATES WITH FULLERENE GUESTS

Bis(zinc porphyrin)-oxoporphyrinogen [(ZnTPP)2-OxP] was prepared as an example of a donor-acceptor type triad [9]. Upon excitation of the zinc porphyrin(s), an effi cient energy or electron transfer occurs in this novel triad and this depends on the polarity of the solvent used in the studies. To further elaborate this system, we complexed this compound and similar derivatives with fullerene derivatives capable of binding at the porphyrin zinc metal ions [9, 15, 16]. This results in the complexes illustrated in Fig. 9.

These novel supramolecular complexes composed from zinc porphyrin(s), fullerene(s), and oxoporphyrinogen redox-/photo-active entities were characterized by spec-tral and electrochemical techniques. Photoinduced elec-tron transfer from singlet excited porphyrin to the fullerene entity was demonstrated from time- resolved emission and transient absorption studies. The experimentally measured charge-separation rates (kCS) were higher for the pentad, (C60Im:ZnP)2-OxP than that for the corre-sponding triad, C60Im:ZnP-OxP. The lifetimes of the radical ion-pair ( RIP) were found to be hundreds of nano-seconds, indicating some degree of charge stabilization in the supramolecular systems studied here. The results reveal that some modulation of the porphyrin-fullerene

Fig. 9. (a) Bis-porphyrinatozinc-substituted OxP complex with fullerene guest. (b) Complex of a tetra-porphyrinatozinc- substituted OxP derivative with two bis-pyridyl-substituted fullerenes. Chemical structures are given below the respective model. P and Mdenote para or meta connectivity at the N-benzyl group

O

O

NH NHN

N

OO

OO

N N

NNZn

N N

NNZn

R R

NH NN

R

R

R =

PP

(a)

NN

NN

OO

OO

N N

NN Zn N N

NNZnR R

NH NN

R

R

NN

N NZnNN

N NZn

RR

NH

N N

RR

M

MM

M

O

OR =

(b)


68 J.P. HILL ET AL.

energy/electron transfer processes can be achieved by careful design of the polychromophoric system. In the case of the bis-pyridyl-substituted fullerene and the host, we used the two-point binding mode resulting in compounds where charge transfer occurring in the non- complexed OxP-(ZnPx)n host molecules could be switched to effi cient photoinduced, electron transfer pro-cesses from singlet excited porphyrin to fullerene upon formation of the novel supramolecular donor-acceptor complexes and the subsequent transient absorption spec-tral studies indicate charge stabilization [9].

OXOCORROLOGEN — A NOVEL QUINONOID CORROLE

Most of the previous work described in this paper con-cerns the oxoporphyrinogen macrocycle derived from a symmetrical tetraphenylporphyrin. However, there exist a signifi cant number of oligopyrrolic macrocycles which

might also be of interest from scientifi c or materials points of view if they were (a) substituted with hemiqui-nonoid groups similar to those in OxP, and (b) available for N-substitution at macrocyclic nitrogen atoms. With this in mind, we prepared the corrole-type compound, oxocorrologen OxC, and investigated its activity in the N-alkylation reaction [17]. The X-ray crystal structure of N-(4-nitrobenzyl)-OxC is shown in Fig. 10. While the availablity of the N-alkylation reactions on OxC presents some interesting synthetic possibilities, we were more intrigued by the extremely complex proton NMR data resulting from this system. We deduced that this phe-nomenon is due to a tautomeric mixture when OxC is dis-solved in non-polar solvents. In fact, there are ten possible isomers of OxC and their calculated structures are shown in Fig. 11. Of these ten, at least three are available by vari-ation of solvent polarity, with isomers 1, 2a and 3a being available in yields up to 80%, with a further unidentifi ed

Fig. 11. Calculated structures of the ten possible tautomers of oxo-corrologen. Energies are given in eV relative to the tautomer 1

HN

NH N

N

OH

O

O

O2N

Fig. 10. X-ray crystal structure and chemical structure of N-(4-nitrobenzyl)-oxocorrologen. In the crystal structure, t-butyl groups have been omitted for clarity. Note the intramolecular

– interaction between a meso-substituent and the N-benzyl substituent, and the intramolecular hydrogen bond at the mac-rocyclic amine groups



tautomer also detected in solution at the lowest solvent polarity used. Tautomer identity was confi rmed using a combination of correlation NMR spectroscopy and nuclear Overhauser enhancement (nOe) spectroscopy.

CONCLUSION

In summary, the highly conjugated tetrapyrrole frame-work presented by oxoporphyrinogen OxP is suitable for modifi cation by N-alkylation at the pyrrolic nitrogen atoms. Prior to N-alkylation, the OxP group can also be considered as a molecular switching unit with access to three stable states: porphyrin, cation/anion radical and oxoporphyrinogen. Some control over structure at the molecular level can be exerted by varying the N-alkyl group identity, and non-crystalline aggregates can also be obtained. The novel N-alkyl-oxoporphyrinogens can be used in the preparation of electrochemical anion detec-tion agents and in the formation of novel supramolecu-lar donor-acceptor complexes suitable for consideration in light-harvesting applications. Further studies of the N-alkyl oxoporphyrinogens along these and other lines, including their self-assembling properties [18], are in progress in our laboratories.

Acknowledgements

This research was supported by Grant-in-Aid for Scientifi c Research in a Priority Area “Chemistry of Coordination Space” and a Grant-in-Aid for Sci-ence Research in a Priority Area “Super-Hierarchical Structures” from Ministry of Education, Science, Sports, and Culture, Japan, World Premier International Research Center Initiative (WPI Initiative) on Materials Nano-architronics, from Ministry of Education, Science, Sports, and Culture, Japan, and Grants-in-Aid for Scientifi c Research (B) from Japan Society for the Promotion of Science. This work was also supported by the National Science Foundation (Grant 0804015 to FD) and the donors of the Petroleum Research Fund administered by the American Chemical Society.

REFERENCES

(a) Albery WJ, Bartlett PN, Jones CC and Milgrom 1.LR. J. Chem. Res. Synopses 1985; 12: 364–365. (b) Milgrom LR, Mofi di N, Jones CC and Harriman A. J. Chem. Soc. Perkin Trans. 2 1989; 301–309. (c) Mil-grom LR, Mofi di N and Harriman A. J. Chem. Soc. Perkin Trans. 2 1989; 805–809. (d) Milgrom LR, Mofi di N and Harriman A. Tetrahedron 1989; 45:7341–7352. (e) Milgrom LR, Flitter WD and Short EL. J. Chem. Soc. Chem. Commun. 1991; 788–790. (f) Milgrom LR and Hill JP. J. Heterocyclic Chem.1993; 30: 1629–1633. (g) Milgrom LR, Hill JP and

Flitter WD. J. Chem. Soc. Perkin Trans. 2 1994; 521–524. Milgrom LR. 2. Tetrahedron 1983; 39: 3895–3898. Milgrom LR, Hill JP and Bone S. 3. Adv. Mater. Opt. Electron. 1993; 2: 143–147.(a) Milgrom LR, Hill JP and Yahioglu G. 4. J. Hetero-cyclic Chem. 1995; 32: 97–101. (b) Milgrom LR, Hill JP and Dempsey PF. Tetrahedron 1994; 50:13477–13484. (c) Dolušić E, Toppet S, Smeets S, Meervelt LV, Tinant B and Dehaen W. Tetrahedron2003; 59: 395–400.(a) Sessler JL, Camiolo S and Gale PA. 5. Coord. Chem. Rev. 2003; 240: 17–55. (b) Sessler JL and Gale PA. In The Porphyrin Handbook, Kadish KM, Smith KM and Guilard R. (Eds.), Academic Press: San Diego, CA, 2000; Vol. 6, pp. 257–258. (c) Ses sler JL, Gale PA and Cho W.-S. Anion Recep-tor Chemistry, Royal Society of Chemistry: Cam-bridge, UK, 2006.Hill JP, Wakayama Y, Schmitt W, Tsuruoka T, 6. Nakanishi T, Zandler ME, McCarty AL, D’Souza F, Milgrom LR and Ariga K. Chem. Commun. 2006; 2330–2322.Hill JP, Hewitt IJ, Anson CE, Powell AK, McCarty 7.AL, Karr PA, Zandler ME and D’Souza F. J. Org. Chem. 2004; 69: 5861–5869.Hill JP. Unpublished results.8.Xie Y, Hill JP, Schumacher AL, Sandanayaka ASD, 9.Araki Y, Karr PA, Labuta J, D’Souza F, Ito O, Anson CE, Powell AK and Ariga K. J. Phys. Chem. C 2008; 112: 10559–10572.Hill JP, Schmitt W, McCarty AL, Ariga K and 10.D’Souza F. Eur. J. Org. Chem. 2005; 2893–2902. Hill JP, Ariga K, Schumacher AL, Karr PA and 11.D’Souza F. J. Porphyrins Phthalocyanines 2007; 11: 390–396.W12. ürthner F, Yao S and Beginn U. Angew. Chem. Int. Ed. 2003; 42: 3247–3250.Hill JP, Schumacher AL, D’Souza F, Labuta J, Red-13.shaw C, Elsegood MJR, Aoyagi M, Nakanishi T and Ariga K. Inorg. Chem. 2006; 45: 8288–8296.Schumacher AL, Hill JP, Ariga K and D’Souza F. 14.Electrochem. Commun. 2007; 9: 2751–2754.Hill JP, Sandanayaka ASD, McCarty AL, Karr PA, 15.Zandler ME, Charvet R, Ariga K, Araki Y, Ito O and D’Souza F. Eur. J. Org. Chem. 2006; 595–603.Schumacher AL, Sandanayaka ASD, Hill JP, Ariga 16.K, Karr PA, Araki Y, Ito O and D’Souza F. Chem.Eur. J. 2007; 13: 4628–4635.Xie Y, Hill JP, Schumacher AL, Karr PA, D’Souza 17.F, Anson CE, Powell AK and Ariga K. Chem. Eur. J. 2007; 13: 9824–9833.Ariga K, Hill JP, Lee MV, Vinu A, Charvet R and 18.

Acharya S. Sci. Technol. Adv. Mater. 2008; 9:

014109.




INTRODUCTION

It has long been believed that the phase transition behavior and clearing points (c.p.s) of phthalocyanine (Pc)-based metallomesogens (metal-containing liquid crystals) scarcely change regardless of the kind of central metal [2, 3]. Therefore, their columnar mesophases may be formed by the balance between the intermolecular π-π

interaction of central phthalocyanine macrocycles and the thermal disturbance of the peripheral molten alkyl chains. Hence, it has been thought that the additional intermo-lecular interaction by the central metal may be negligibly small. Previously, we synthesized sandwich-type, Pc-based Lu complexes, bis[2,3,9,10,16,17,23,24-octakis(3,4-dialkoxyphenoxy)phthalocyaninato]lutetium(III) (abbreviated as {[(CnO)2PhO]8Pc}2Lu (n = 8–16) (1a–1i))shown in Fig. 1, and investigated their phase transition behavior and spectroscopic properties in detail [1, 4, 5]. In this work, we have synthesized a series of novel ho-mologs, bis[2,3,9,10,16,17,23,24-octakis(3,4-dialkoxy-phenoxy)phthalocyaninato]europium(III) (abbreviated as

Discotic liquid crystals of transition metal complexes 40†:unique double-clearing behavior of a series of novel discotic metallomesogens based on bis(phthalocyaninato)-europium(III) complexes

Hidetomo Mukai, Miho Yokokawa, Kazuaki Hatsusaka and Kazuchika Ohta*

Smart Material Science and Technology, Department of Bioscience and Textile Technology, Interdisciplinary Graduate School of Science and Technology, Shinshu University, Ueda, 386-8567, Japan


ABSTRACT: It has long been believed that the phase transition behavior and clearing points (c.p.s) of phthalocyanine-based metallomesogens scarcely change regardless of the kind of central metal. In this work, we have synthesized a series of novel sandwich-type, europium complexes, bis[2,3,9,10,16,17,23,24-octakis(3,4-dialkoxyphenoxy)phthalocyaninato]europium(III) (abbreviated as {[(CnO)2PhO]8Pc}2Eu(n = 8–16) (2a–2i)). Their mesomorphic properties were investigated by polarization microscope, differential scanning calorimeter and temperature dependent X-ray diffraction techniques to make a comparison with the phase transition behavior and c.p.s of the previously reported lutetium homologs ({[(CnO)2PhO]8Pc}2Lu (n = 8–16)) (1). Very interestingly, the c.p.s of {[(CnO)2PhO]8Pc}2Eu complexes (2) are higher by 35–61 °C than those of {[(CnO)2PhO]8Pc}2Lu complexes (1) having the same chain length (n). Furthermore, these Eu complexes (2) exhibited unique double-clearing behavior, whereas the Lu complexes (1) did not. For instance, the Eu complex (2i: n = 16) exhibited the following phase transition sequence: crystalline phase (K) 49 °C Colh 133 °C Cub(Pn3-m) 158 °C Cub(Pm3-n) 216 °C I.L.; the I.L. very slowly transformed into a Coltet mesophase; on further heating, Coltet cleared into I.L. again at 240 °C. On the other hand, the corresponding Lu complex (1i: n = 16) exhibited a phase transition sequence of K 50 °C Colh 124 °C Cub(Pm3-n) 148 °C Coltet 205 °C I.L., and did not exhibit the double-clearing behavior. Thus, it was revealed for the fi rst time, in the phthalocyanine-based metallomesogens, that the central metal had a large infl uence on their c.p.s and their phase transition behavior.

KEYWORDS: discotic liquid crystal, metallomesogen, sandwich-type, lanthanoid phthalocyanine, double-clearing.


*Correspondence to: Kazuchika Ohta, email: [email protected]

†Part 39, see reference 1


DISCOTIC LIQUID CRYSTALS OF TRANSITION METAL COMPLEXES 40 71

{[(CnO)2PhO]8Pc}2Eu (n = 8–16) (2a–2i) in Fig. 1). The molecular structure difference between complexes 1 and 2 is only a central metal. Lu3+ in the previous Lu com-plexes (1) is non-magnetic, whereas Eu3+ in the present Eu complexes (2) is largely magnetic due to the total spin quantum number of S = 3. Therefore, if these discotic Pc molecules are one-dimensionally stacked to form the columns, quasi-one-dimensional chains of metals can be formed in the center of the columns. In the case of Eu complexes (2), an additional intermolecular interaction due to the large magnetism can be expected. Hence, we had expected that the Eu complexes (2) would exhibit phase transition behavior that was different from the pre-vious Lu complexes (1) to show higher c.p.s. According

to our expectation, the c.p.s of the Eu complexes (2) were higher by 35–61 °C than those of the Lu complexes (1)having the same chain length (n). It is also interesting that the Eu complexes (2) showed unique double-clearing behavior [6–8], whereas most of the Lu complexes (1)showed the usual single clearing behavior. Thus, when comparing Lu complexes (1) with Eu complexes (2), it is apparent that the central metal infl uences their phase transition behavior and c.p.s.

In this paper, we report the synthesis and phase transi-tion behavior of a series of the novel Eu complexes (2)showing such very unique phase transition behavior.

EXPERIMENTAL

Syntheses

The Eu complexes (2) were prepared by the previ-ously reported method for Lu complexes (1) [4, 9]. For the purifi cation of the Eu complexes (2), a Japan Ana-lytical Industry LC-918 recycling preparative HPLC equipped with UV detector (UV-50) was used (column: JAIGEL 3H+4H, pressure; 14–15 kg.cm-2, fl ow rate: 3.9 mL.min-1). Yields, MALDI-TOF mass spectral data and elemental analysis data of the Eu complexes (2) are summarized in Table 1.

Measurements

The purifi ed Eu complexes (2) were identifi ed by ele-mental analysis (Perkin-Elmer elemental analyzer 2400), MALDI-TOF mass spectroscopy (Perseptive Biosystem Voyager, matrix: dithranol) and electronic absorption spectroscopy (Hitachi U-4100 spectrophotometer). The electronic absorption spectral data were summarized in Table 2. The phase transition behaviors were observed by using a polarization microscope (Olympus BH2), equipped with a heating plate (Mettler FP-82HT hot

Fig. 1. Formulae of {[(CnO)2PhO]8Pc}2Lu (1) and {[(CnO)2

PhO]8Pc}2Eu (2)

OCnH2n+1

M

NNN

N

NN

NNN

N

N NNN

NN

ORRO

OR

OR

OR

RORO

OROR

RORO OR

OR

RORO

OR

OCnH2n+1

n = 8n = 10n = 12n = 14n = 16

: a: c: e: g: i

R =

This work

2: M = Eu

1: M = Lun = 8n = 9n = 10n = 11n = 12n = 13n = 14n = 15n = 16

: a: b: c: d: e: f: g: h: i

Previous work

Table 1. Yields, MALDI-TOF mass spectral data and elemental analysis data of {[(CnO)2PhO]8Pc}2Eu (2a–2i)

Elemental analysis; found, % (calculated %)

Compound Yield, % Mol. formula (Mol. wt) Mass spectra C H N

n = 8: 2a 31C416H608N16O48Eu

(6753.45)6752.60 74.37 (73.99) 9.44 (9.07) 3.26 (3.32)

n = 10: 2c 12C480H736N16O48Eu

(7651.17)7651.08 75.20 (75.35) 10.06 (10.04) 2.82 (2.93)

n = 12: 2e 25C544H864N16O48Eu

(8548.89)8548.59 76.52 (76.43) 10.51 (10.19) 2.83 (2.62)

n = 14: 2g 40C608H992N16O48Eu

(9446.61)9445.86 77.76 (77.30) 11.04 (10.58) 2.73 (2.37)

n = 16: 2i 25C672H1120N16O48Eu

(10344.33)10341.90 77.86 (78.03) 11.21 (10.91) 2.39 (2.17)


72 H. MUKAI ET AL.

stage) controlled by a thermoregulator (Mettler FP-90 Central Processor), measured by a differential scanning calorimeter (Shimadzu DSC-50). Temperature-depen-dent X-ray diffraction measurements were performed with Cu-Kα radiation by using a Rigaku RAD X-ray diffractometer equipped with a handmade heating plate controlled by a thermoregulator [10, 11]. The X-ray data of the Eu complexes (2) were summarized in Table S1 in the supplementary information.


Double-clearing behavior

The phase transition behaviors and enthalpy changes of the Eu complexes (2) are summarized in Table 3. As can be seen from this table, the Eu complexes (2) exhib-ited plural mesophases (Colr, Colh, Cub and Coltet). These plural mesophase structures are illustrated in Fig. 2.

Table 2. UV-vis spectral data in CHCl3 of {[(CnO)2PhO]8Pc}2Eu (2a–2i)

CompoundConcentration†

(10-6 M)

λmax, nm (log ε)

Soret-band Q-band

Q0-1-band * Q0-0-band

n = 8: 2a 6.71 290.4 (5.16) 331.7 (5.14) 358.9 (5.07) 475.6 (4.56) 618.9 (4.69) 657.2 (4.67) 686.0 (5.23)

n = 10: 2c 6.73 290.3 (5.16) 332.7 (5.16) 361.9 (5.09) 477.1 (4.58) 618.5 (4.71) 657.7 (4.69) 684.9 (5.26)

n = 12: 2e 6.67 290.2 (5.16) 332.1 (5.15) 358.4 (5.09) 472.6 (4.58) 617.0 (4.70) 657.4 (4.69) 685.8 (5.25)

n = 14: 2g 6.72 288.1 (5.17) 333.4 (5.15) 357.9 (5.09) 475.6 (4.58) 617.0 (4.69) 658.0 (4.71) 685.4 (5.22)

n = 16: 2i 6.67 289.9 (5.19) 333.2 (5.19) 360.4 (5.11) 475.6 (4.61) 618.1 (4.72) 657.2 (4.72) 684.7 (5.26)

†In chloroform, *Aggregation band of Q0-0-band

Colr3(P21/a) ColhColr2(P21/a)Colr1(P21/a)

Cub (Pm3n) ColtetCub (Pn3m)Colh

Cub (Pm3n) ColtetCub (Pn3m)Colh

Cub (Pm3n) ColtetCub (Pn3m)ColhK

Cub (Pm3n) ColtetCub (Pn3m)ColhK

PhaseT/ oC [ Δ H/ kJmol-1]Compound

Phase

n = 10:2c

n = 8:2a

{[(CnO)2PhO]8Pc}2Eu relaxation

I.L.145[33.6] 210[1.43] 303

I.L.256[10.1]

I.L.144[29.3] 178[15.0] 297

I.L.247[14.1]

n = 12:2e I.L.146[33.9] 177[21.2] 281

I.L.247[25.1]

n = 14:2g I.L.136[27.7] 165[19.3] 271

I.L.228[3.19]

37[329]

I.L.133[28.3] 158[18.8] 240

I.L.216[6.77]

49[426]n = 16:2i

*

*

*

*

*

**

**

**

**

Table 3. Phase transition temperatures and enthalpy changes of {[(CnO)2PhO]8Pc}2Eu (2a–2e)

Phase nomenclature: K = crystal, Colr = rectangular columnar mesophase, Colh = hexagonal columnar mesophase, Cub = cubic mesophase, Coltet = tetragonal columnar mesophase and I.L. = isotropic liquid. *: very slow; **: gradual decomposition above ca 270 °C



where the columnar structures are seen in the Colr(P21/a)and Colh mesophases. The columns are branched in the Cub(Pn3-m) mesophase, but the branched columns are cut into short rods in the higher temperature Cub(Pm3-n)mesophase. In the highest temperature mesophase Coltet,the short rods are again stacked one-dimensionally. The de-tails of these mesophase structure changes will be reported elsewhere [12]. In this paper, we wish to report only our fi ndings of the difference of clearing behavior between the Lu and Eu complexes. It is very interesting that the Eu complexes (2) showed a unique double- clearing behavior. This behavior was confi rmed by our careful microscopic observations. For example, the double- clearing behavior of the derivative {[(C16O)2PhO]8Pc}2Eu (2i) is shown in Fig. 2A–F. Figure 2A shows a photomicrograph of the Cub(Pm3-n) mesophase of the derivative 2i at 170 °C. When it was heated up to 216 ºC, it started clearing into an isotropic liquid (I.L.) with fl uidity, as shown in Fig. 2B; after 5 s, it cleared perfectly into I.L. (Fig. 2C). During our careful observations when maintaining this temperature for 5 min, we found that slight texture appeared from the I.L. (Fig. 2D). After 10 min, it grew to a dendritic texture having C4 symmetry characteristic of a Coltet mesophase (Fig. 2E). On further heating up to 240 °C, it started clearing into I.L. again (Fig. 2F). Thus, the derivative 2i (n = 16) showed unique double-clearing behavior. The other derivatives for n = 14, 12 and 10 also showed similar double-clearing behavior. As can be seen from Table 3 and Fig. 2G, the derivative 2a (n = 8) only reformed from I.L. into a Colh mesophase, showing a dendritic texture having C6 symmetry characteristic to a Colh mesophase. It took a very long time for the I.L. to result from the fi rst clearing relaxed into the highest tem-perature mesophase (Colh (n = 8) or Coltet (n = 10–16)). It took more than one day to completely relax into the high-est temperature mesophase. Hence, the c.p.s and enthalpy changes of these highest temperature mesophases could not be determined by differential scanning calorimeter (DSC) measurements even for the slowest heating rate of 2.5 °C.min-1. The c.p.s of these highest temperature mesophases could be determined only by laborious mi-croscopic observations. As mentioned above, each of the Eu complexes (2) exhibited double-clearing behavior.

In contrast, most of the corresponding Lu homologs (1)showed single-clearing behavior [1]. For example, when the Lu complex (1i) was heated, it showed a phase transi-tion sequence of K 49 °C Colh 133 °C Cub(Pn3-m) 158°C Coltet 216 °C I.L. Thus, the Lu complexes (1) showed only the usual single-clearing behavior.

G-T diagram

The usual single clearing behavior of the Lu com-plexes (1) and the unique double-clearing behavior of the Eu complexes (2) can be rationally explained by using two schematic Gibbs free energy (G) vs temperature (T) diagrams. As illustrated in Fig. 3A, when the Colh

mesophase of the {[(C16O)2PhO]8Pc}2Lu complex (1i)is heated, it proceeds following the direction of the small arrows, and runs along the bold line in this fi gure. The phase transition occurs at each intersection. As a result, it shows a phase transition sequence of Colh →Cub(Pn3-m) → Coltet → I.L. In this case, only one clear-ing occurs at the white circle of the intersection of the Coltet phase line and the I.L. line at 205 °C. On the other hand, when the Col h phase of {(C16O)2PhO}8Pc}2Eu complex (2i) is heated, it runs along the bold line as shown in Fig. 3B, and shows a phase transition seq-uence of Colh → Cub(Pn3-m) → Cub(Pm3-n) → I.L. The Cub(Pm3-n) phase line crosses with the I.L. line at 216 °C, where the fi rst clearing occurs. When it is held at this temperature, the I.L. phase is relaxed into the more thermally stable Coltet phase. When the resulting Coltet

phase is further heated, this phase clears into I.L. again at the intersection of the Coltet phase line and the I.L. line at 240 °C.

Thus, both the double- and single-clearing behaviors can be rationally explained by using these schematic G-Tdiagrams. In comparison with G-T diagrams of Fig. 3A (the Lu complex (1i)) and B (the Eu complex (2i)), the Coltet phase line of the Eu complex (2) goes down to the lower side, and the intersection (c.p.) of the Coltet line and the I.L. line moves to higher temperature. This implies that the Coltet phase of the Eu complex (2i) is more stable than that of the Lu complex (1i). Hence, it is suggested that the highest temperature Coltet mesophase may be stabilized by

Colh ColtetCub(Pm3n)Cub(Pn3m)_ _

Colr(P21/a)

Fig. 2. Schematic models of the mesophase structures for the {[(CnO)2PhO]8Pc}2Eu (2) complexes


74 H. MUKAI ET AL.

some additional intermolecular force while changing from Lu to Eu.

Infl uence of central metal on the clearing points

In Fig. 4, the phase transition temperatures of the Lu complexes (1) and the Eu complexes (2) were plotted against the number of carbon atoms in the alkoxy chain (n). As can be seen from this fi gure, the highest temper-ature phase is the Coltet phase in both the Lu complexes (A: n ≥ 10) and the Eu complexes (B: n ≥ 10). In com-parison with the c.p.s of Coltet phase of these complex-es, their c.p.s are at 205–248 °C for the Lu complexes (1) and at 240–303 °C for the Eu complexes (2). The c.p.s of Coltet phases of the Eu complexes (2) were

higher by 35–61 °C than the c.p.s of the Coltet phases of the corresponding Lu complexes (1), that have same chain length (n). It has long been believed that the phase transition behavior and c.p.s of Pc-based metal-lomesogens scarcely change regardless of the kind of central metals [2, 3]. Moreover, what has never been discussed is the infl uence of central rare-earth metal on their c.p.s in all the sandwich-type rare-earth metal Pc liquid crystals [13, 14]. In comparison with the c.p.s of Coltet phases of the present Eu complexes (2) and the previous Lu complexes (1), it is apparent that the central rare-earth metal ions have a large infl uence on their c.p.s, as described above. The molecular structure difference between the Lu complexes (1) and the Eu complexes (2) is only a central rare-earth metal ion.

n = 16: 2i

n = 8: 2a

Fig. 3. Photomicrographs of {[(C16O)2PhO]8Pc}2Eu (2i) recorded with uncrossed polarizers; (A) at 170 °C, (B) at 216 °C, (C) after 5 s at 216 °C, (D) after 5 minutes at 216 °C, (E) after 10 min at 216 °C, (F) at 240 °C and a photomicrograph of {[(C8O)2PhO]8Pc}2Eu(2a) recorded with crossed polarizers and (G) at 270 °C



Lu3+ in the Lu complexes (1) is non-magnetic due to the total spin quantum number of S = 0, whereas Eu3+ in the Eu complexes (2) is largely magnetic as S = 3. When these discotic Pc molecules are one-dimensionally stacked to form the columns, quasi-one-dimensional chains of metals can be formed in the columnar centers.

Therefore, the Eu complexes (2) may have an additional intermolecular interaction due to the large magnetism of the Eu3+ ions. On the other hand, the Lu complexes (1) may not have this interaction. Hence, the c.p.s of Coltet phases of the Eu complexes (2) may be higher than those of the Lu complexes (1).

I.L.

I.L.

Coltet

Coltet

Cub (Pm3n)

Cub (Pm3n)

Colh

Colh

Cub (Pn3m)

Cub (Pn3m)

133 158 216 240

Cub (Pm3n) ColtetCub (Pn3m)ColhK I.L.133 158 240

I.L.216

49n = 16: 2i

Temperature, oC

G

__

_

_

Fig. 4. G-T diagrams of the present clearing behavior, (A) the single-clearing behavior of {[(C16O)2PhO]8Pc}2Lu (1i) and (B) the double-clearing behavior of {[(C16O)2PhO]8Pc}2Eu (2i). �: clearing point, �: mesophase-mesophase transition

Cub (Pm3n)

Colh

Coltet I.L.

Colh

Cub (Pm3n)

Coltet

I.L.

124 148 205

Temperature, oC

G

ColtetCub (Pn3m)ColhK I.L.124 148 20550

n = 16: 1i

_

_

(A)

(B)


76 H. MUKAI ET AL.

CONCLUSION

A novel series of sandwich-type, Pc-based Eu complex-es, bis[2,3,9,10,16,17,23,24-octakis(3,4-dial k oxy phenoxy)phthalocyaninato]europium(III) ({[(CnO)2PhO]8Pc}2Eu (n = 8–16) (2)), were synthesized and these phase-transition behaviors were investigated to compare with those of the corresponding Lu homologs ({[(CnO)2PhO]8Pc}2Lu (n = 8–16) (1)). The Eu complexes (2) showed unique, double-clearing behavior which was not observed for the Lu com-plexes (1). Furthermore, compared with the c.p.s of the Coltet

phase of these two complexes, the c.p.s of the Eu complexes were higher by 35–61 °C than those of the corresponding Lu complexes. Since these two complexes are composed of the same ligands, it was concluded that this phenome-non results from the difference of the central metal. It was revealed for the fi rst time in the Pc-based metallomesogens that the additional intermolecular interactions by the mag-netism of the central metal ions may largely infl uence their clearing points and phase transition behaviors.

Acknowledgements

This work was supported by a Grant-in-Aid for Global COE Program and a Grant-in-Aid for Science Research (19022011) in a Priority Area “Super-Hierarchical Structures” from Ministry of Education, Culture, Sports, Science and Technology, Japan.

Supporting Information

X-ray data of {[(CnO)2PhO]8Pc}2Eu are given in the supplementrary material. This material is available at http://www.worldscinet.com/jpp/jpp.shtml.

REFERENCES

Mukai H, Hatsusaka K and Ohta K. 1. J. Porphyrins Phthalocyanines 2007; 11: 846–856.Sleven J, Gorller-Walrand C and Binnemans K. 2.Mater. Sci. Eng. C, 2001; 18: 229–238.Binnemans K, Sleven J, de Feyter S, de Schryver 3.FC, Donnio B and Guillon D. Chem. Mater. 2003; 15: 3930–3938.Hatsusaka K, Kimura M and Ohta K. 4. Bull. Chem. Soc. Jpn. 2003; 76, 781–787.Hatsusaka K, Kimura M and Ohta K. 5. J. Porphyrins Phthalocyanines 2003; 7(2): 105–111.Ohta K, Ema H, Muroki H, Yamamoto I and Matsuzaki 6.K. Mol. Cryst. Liq. Cryst. 1987; 147: 61–78.Komatsu T, Ohta K, Fujimoto T and Yamamoto I. 7. J. Mater. Chem. 1994; 4: 533–536.Ohta K, Azumane S, Watanabe T, Tsukada S and 8.Yamamoto I. Appl. Organometallic Chem. 1996; 10: 623–635.Hatsusaka K, Ohta K, Yamamoto I and Shirai H. 9. J. Mater. Chem. 2001; 11: 423–433.Ema H. Master Thesis, Shinshu University, Ueda, 10.1988; Chap. 7.Hasebe H. Master Thesis, Shinshu University, Ueda, 11.1991; Chap. 5.Mukai H, Yokokawa M, Ichihara M and Ohta K. 12.Liq. Cryst. In preparation.Ban K, Nishizawa K, Ohta K, van de Craats AM, 13.Warman JM, Yamamoto I and Shirai H. J. Mater.Chem. 2001; 11: 321–331.Gürek AG, Basova T, Luneau D, Lebrun C, Kol’tsov 14.E, Hassan AK and Ahsen V. Inorg. Chem. 2006; 45:1667–1676.

Colr3(P21/a)

300

200

100

8 10 12 14 16

K

Colh

Coltet

I.L.

Colr1(P21/a)

Colr2(P21/a)

Colh

Number of the carbon atoms in the alkoxy chain (n)

Tem

pera

ture

,o C

Tem

pera

ture

,o C

0

300

200

100

0

I.L.

ColtetCub (Pm3n)

Cub (Pn3m)

Colh

K

X

8 10 12 14 169 11 13 15

K

Cub (Pn3m)

Cub (Pm3n)_

_

_

_

Fig. 5. Phase transition temperature vs n for (A) {[(CnO)2PhO]8Pc}2Lu (1a–1i) and (B) {[(CnO)2PhO]8Pc}2Eu (2a–2i). �: melting point, �: clearing point, �: mesophase–mesophase transition




INTRODUCTION

Porphyrins and related tetrapyrroles [1] are well-known as functional materials for gravimetric [2] and optical [3, 4] chemical sensing. Optical responses to target stimuli include selective changes in the fluo-rescence and absorption spectra, which can only be fully characterized by fluorescence spectroscopy and Excitation Emission Matrix (EEM) spectroscopy [5–7]. Unfortunately, the instrumentation required for these measurements is unaffordable for distributed chemical sensing uses.

On the other hand, consumer electronic devices such as CDs and DVD drives [8], flatbed scanners [3, 9] and computer screens with web cameras (computer screen photo-assisted technique, CSPT [10–14]) constitute in-creasingly sophisticated platforms capable of extended

functions such as chemical sensing. The ubiquity of these devices, which makes them a readily available and highly distributed infrastructure, is among the motiva-tions for exploring such uses. Concurrently, the associ-ated computing power serves to compensate for instru-mental weaknesses [14, 15] by making the setups ver-satile enough to support diverse assays formats [10, 11, 13, 15], and detection principles [11, 12, 14, 15]. Com-puter platforms associated to any of these cases are also geographically traceable network terminals, compatible with distributed detection and centralized monitoring of detected events.

In particular, CSPT measurements of fluorescent indicators are able to retain spectral fingerprints of these substances, providing a robust platform for distribut-ed chemical sensing. In this work, we review the most recent progress in chemical sensing based on this tech-nique, which includes the demonstration of handy assays operating as electronic noses [16], as well as the theoreti-cal description of CSPT signatures as fingerprints of the EEM of the tested indicators.

*a, b, c, a,c, b , c

a

a Division of Applied Physics, IFM-Linköping University, 58183 Linköping, Swedenb Department of Chemical Sciences and Technologies, University of Rome “Tor Vergata”, Via della Ricerca Scientifica, 00133 Rome, Italyc Department of Electronic Engineering, University of Rome “Tor Vergata”, Via del Politecnico 1, 00133 Rome, Italy


ABSTRACT: Recent progress in spectral fingerprinting of fluorescent indicators using distributed instrumentation based on consumer electronic devices is reviewed. In particular, the evaluation of disposable assays using a computer screen photo-assisted technique (CSPT) is discussed. Sample identification and optimization strategies are analyzed as well as the underlying theoretical background for polychromatic spectral fingerprinting.

KEYWORDS: optical sensing, chemical sensing, porphyrins, optical fingerprinting, computer screen illumination.


*Correspondence to: Daniel Filippini, email: [email protected], fax: +46 13-288969


78 D. FILIPPINI ET AL.

EXPERIMENTAL

(5,10,15,20-tetraphenylporphyrinato)zinc, (ZnTPP), (5,10,15,20-tetraphenylporphyrinato)iron(III) chloride, (FeTPPCl), and 2,3,17,18-tetraethyl-7,8,12,13-tetrame-thyl-a,c-biladiene dihydrobromide (BD), dispersed in a polyvinylchloride (PVC) matrix, were used as sensing materials. These indicators are deposited as 1 mm dia-meter spots on a glass slide; six spots (each indicator duplicated) constitute the sensing array. Metalloporphy-rins and BD were prepared according to literature meth-ods [1, 7]. The array was prepared by deposition of drops of PVC membrane solutions (1 wt % of porphyrin, PVC/bis(2-ethylhexyl)sebacate (1:2) polymeric matrix) in tet-rahydrofuran (THF). Evaporation of the solvent led to the formation of a PVC/porphyrin membrane.

The sensing array was introduced into a ~20 mm × 60 mm × 5 mm gas cell with glass walls (Fig. 1a), where it was exposed to controlled concentrations of different gases with the aid of an automatic gas-mixing system. NH3, NOx, CO, and triethylamine (TEA) at concentra-tions of 10, 20, 500 and 1500 ppm in nitrogen were used for the results presented here.

CSPT measurements utilized a regular LCD screen (Philips 170S4) as a light source and a Logitech Quick-cam Pro4000 web camera operating at a resolution of 320 × 240 pixels as the image detector. A sample holder provided a controlled geometry for the measurements and shielded the experiments from ambient illumination.

Software written in Matlab commanded the measure-ment (illuminating sequence and video acquisition) and extracted the information (CSPT fingerprints) from man-ually selected regions of interest (ROI).

Experimental results used to corroborate the modeling consisted of the spectral characterization of fluorescent molecules, with a concentration of 2 μM for absorption measurements and 10 μM for the emission measure-ments. Spectroscopic-grade THF (Fluka) and distilled water passed through a Milli-Q purification system were used as solvents. Fluorescein was diluted in water, whereas ZnTPP, BD and (5,10,15-triphenylcorrolato) germanium(IV) chloride (GeTPCCl) [18] were diluted in THF. Absorption spectra were measured with a Shimad-zu UV-1601PC, spectrophotometer. EEM measurements were carried out on a spectrofluorimeter (Hitachi F4500), operating in the 390–700 nm detection range, with excitations at 20 nm intervals in the same range. Samples were contained in 1 cm light-path quartz cuvettes.

Polychromatic excitation measurements were per-formed using a LCD screen (Philips 170/s2, 1280 × 1024 pixels resolution at a 60 Hz refresh frequency) operating at normal conditions of intensity and contrast.

The screen illuminated the cuvettes, which were con-tained in a holder with two transversal optical fibers [19], one pointing toward the screen and the second perpen-dicular to it. In the direction transversal to the illumi-nation, a mirror was used to increase the collection of fluorescence in that channel. Optical fibers were connected

Fig. 1. (a) Scheme of a CSPT measuring setup. A sensing array composed of three sensing spots deposited on a glass slide is contained in a tight gas cell and exposed to controlled atmospheres. The cell has glass sides and the array is imaged by a web camera under screen illumination. (b) Picture of the sensing area under cyan illumination. White frames indicate regions of interest (ROI) used to characterize the illumination and indicators. (c) Intensity profiles of 6 ROIs corresponding to three indicators and three background areas beside them. The signals are the concatenation of the intensity profiles of each camera channel for each particular ROI


SPECTRAL FINGERPRINTING OF PORPHYRINS 79

to an Ocean Optics USB2000 spectrometer. During the measurements, the screen displayed a sequence of 50 colors. The transmission and emission spectrum of both the reference (solvent) and the sample were recorded for each color.


The result of a CSPT measurement of fluorescent substances is a sequence of true-color pictures of such substances taken under different illuminating condi-tions provided by the computer screen [10]. CSPT is an imaging technique and, as such, the number of substanc-es imaged or their layout in an array can be completely arbitrary [10, 11, 13]. This aspect confers versatility on the technique, enabling it to deal with different assays always using the same instrument, which is an advanta-geous situation for distributed sensing.

The information acquired in CSPT measurements is retrieved in a post-processing step where regions of inter-est (ROI) corresponding to indicators and background information are identified and used to compose finger-prints of the chemical stimuli. The selection of these ROIs can be manual, as in the present case, or automatic, such as in more advanced detection schemes [20].

Figure 1b highlights ROIs with circular white frames for the indicators and squared frames for the background. The average intensity of these ROIs produces three-index color values for each illuminating color. This multidimen-sional information can be treated in different ways [21] that confer slightly different classification performances, which can be adapted to actual problems [22]. For the case of fluorescent substances [23, 24], the information is composed of concatenated intensity profiles from each camera channel. Figure 1c shows an example of such in-tensity profiles for Zn(tpp), Fe(tppCl) and BD substances on the left column of the array (Fig. 1b) and their closest background ROIs.

The idea of looking at a tri-chromatic intensity pro-file for a sequence of illuminations with different spectral radiances is common to spectral reconstruction meth-ods [25], a closely related approach to hyper-spectral imaging [26, 27], adapted to problems such as artwork digitalization. Due to the smooth spectral absorbances of color substances, spectral reconstruction enables the spectral characteristics of a sample to be fully recreat-ed using a minimum of spectral bands [28], a condition proven achievable with the wide-band partially overlap-ping excitations and detection windows present in CSPT arrangements [29].

Since the purpose of the CSPT experiment is not the spectral reconstruction itself but the spectral character-ization of the indicators within ROIs and especially their spectral changes upon interaction with target stimuli, CSPT evaluations are truncated reconstructions, where spectral fingerprints are directly classified and the main challenge consists of finding the best illuminating condi-

tions to optimize the spectral identification of the chemi-cal responses [30, 31].

CSPT measurements of fluorescent indicators intro-duce the additional challenge to describe signatures of such substances, a subject not treated by the conven-tional reconstruction theory, and an important aspect for understanding the CSPT behavior which has been recent-ly elucidated [19].

Intensity profiles such as those in Fig. 1c are the com-position of transmitted light and fluorescence in different proportions, depending on the material, illuminating radi-ance, imaging band and geometry of the experiment [19]. As can be seen from Fig. 1c, for identical experimental conditions, the intensity profiles from the indicators cap-ture distinctive information that can be used to identify them. The proportion of fluorescent and transmitted sig-nals incorporated in the signatures can be selected with the geometry of the setup or integrated in a microstruc-ture, such as recently demonstrated using SU-8 micro-pillar forests as substrates [13].

Actually, the purpose of the CSPT experiments is not the identification of the indicators themselves, but their spectral changes upon exposure to target environments. Thus, a more efficient way to highlight these responses is to subtract the intensity profile of the indicator from the sensing profile before exposure [23]. In this way, positive peaks correspond to increases of transmittance or fluores-cence, whereas negative peaks correspond to decreases of transmitted light or fluorescence quenching. For the tetrapyrroles exploited in this work, fluorescence occurs in the spectral region isolated by the red camera chan-nel, which accordingly reports most of the fluorescent response. Figure 2 shows several of these array signatures tracing the presence of NH3, CO, triethylamine (TEA) and NOx at concentrations of 10, 20, 500 and 1500 ppm in nitrogen, respectively. In this case, duplicated spots are also concatenated in the array signature. As can be seen, different gases produce different peak distributions; the molecules can be identified in this way. These finger-prints are too complex for direct inspection, and multi-variate analysis methods [32] are used to summarize the information they contain and to eventually implement automatic classification schemes [15].

A powerful way to summarize and compare informa-tion in complex signals is to perform Principal Compo-nent Analysis (PCA) on the fingerprints [32, 33]. Figure 2 shows the first three principal components of the whole set of fingerprints (white circular symbols), which explain more than 90% of the original information. This plot shows the classification of the considered molecules as well-separated points in the 3D principal components’ space. Automatic classification strategies can operate intro-ducing the scores of an unknown sample into this PC’s space and measuring distances to these known categories, thus determining, by proximity, the character of the unknown sample. This is a similar process to the visual inspection used for the evaluation of colorimetric quick



tests, but in the case of CSPT, it can be automatically implemented by software [15]. One important difference, however, is that colorimetric evaluations, in contrast to spectral fingerprints, are not univocal representations of the spectral responses, and metameric matches [34] can affect the measurements.

Figure 2 also illustrates how scores can be associated with the experimental conditions leading to the identifica-tion, an important possibility to optimize the measurement of target molecules or to screen for particular indicators.The bi-plot representation [32] in Fig. 2 addresses the correlation of the classification with the source of the CSPT signals. In these graphics, each symbol can be inter-preted as the tip of a vector from the origin; scores that are 100% correlated with a measuring condition lie in the same direction, while perpendicular directions are not correlated. The particular measuring conditions and indi-cators are identified by colors and symbols. Squares, dia-monds and circles identify the ZnTPP, FeTPPCl and BD responses, respectively, while the center of these symbols indicate the illuminating color and the frame of the cor-responding camera channel.

In this way, it is possible to identify the CSPT mea-suring conditions that determine the ability of the CSPT fingerprints to distinguish the different target stimuli. For example, the detection of TEA is highly correlated with the response of BD, which is at 180° to the score

and is mostly observed in the red camera channel for orange-yellow illuminations, physically corresponding to the quenching of the BD fluorescence. From this kind of analysis, it is possible to identify substance-specific subsets of optimum illuminating conditions that could produce the same identification with a shorter illuminat-ing sequence. In this example, just orange-yellow illu-mination could be enough to detect TEA using BD as indicator.

The process of selecting measuring conditions that maximize the discrimination of a given set of indicators is central to CSPT performance and has been studied from practical [35] and theoretical [30, 31] perspectives, from which the automatic selection of measuring con-ditions using correlation analysis, as illustrated in the bi-plot example, is the most recent development [35]. In this case, best measuring conditions are selected by setting a threshold for the correlation of scores with the loads. Accordingly, a possibly measure of correlations is expressed in Equation 1:

i j i iS L, , (1)

where Si is the n-dimensional vector corresponding to the score i, and Lj is the n-dimensional vector of the load j.Thus, maximum values for a particular score i are de-tected and used to identify the experimental condition of the load j that makes i,j maximum. This approach has the advantage of strictly operating on experimental re-sults valid for each CSPT platform and has been recently demonstrated for the evaluation of creatinine, potassium and glucose tests [35].

As mentioned at the beginning of this section, CSPT introduces the problem of describing the fingerprints of fluorescent substances, a subject not treated in the con-ventional spectral reconstruction formalism, and which implies the characterization of fluorescent substances by polychromatic stimulation of their excitation emission matrices.

The signal reaching the camera (Fi( )) in a CSPT experiment is a weighted sum of fluorescence (Ei( )) and transmitted (Ti( )) light [19] as follows (Equation 2):

Fi( ) = n1 · Ti( ) + n2 · Ei( ), (2)

where n1 depends on the concentration and the molar absorptivity of the considered substance, and on the path length of the cuvette, whereas n2 depends on the geo-metry of the setup and quantum yield of the substance, while the sub-index i identifies the illuminating spectral radiance displayed by the screen. In order to corroborate the model discussed below, spectroscopic measurements of Fi( ) under screen illumination were performed.

Thus, for each illuminating color i displayed on a com-puter screen, Ei( ) and Ti( ) are measured with a spectro-photometer as detailed in the experimental part.

Fig. 2. Bi-plot of CSPT fingerprints of CO, TEA, NH3 and NOx. 3D scores are indicated by white circles and loads in col-ors. In the bi-plots, the contribution from the different colors of the illumination (inner color of the symbols), the camera chan-nels (outer color of the symbols), and the different indicators (squares, diamonds, and circles for ZnTPP, FeTPPCl, and BD, respectively) are indicated. The insets show the different finger-prints used to construct the bi-plot



EI

Iii

i

( )( )

( ),

0 90

(3)

and

TI

Iii

i

( )( )

( ).

0 0

(4)

The emission and transmission spectra of tested sub-stances, for each particular illuminating color i, are given by Equations 3 and 4, in which I and I0 are intensities measured through the dissolved target substance and the pure solvent used to dissolve it, respectively, for the fiber optics at 90° from the illumination and facing the illumi-nation (0°).

In contrast to regular spectrophotometers that use wide-band light sources, computer screens emit polychro-matic spectral radiances as a result of the combination of three partially overlapping bands, corresponding to the radiances of the screen primaries R( ), G( ), B( ). Thus, for any given color i defined by the triplet of weights (ri, gi, bi) the exciting spectral radiance is given by Equa-tion 5 in which is the correction for the non-linearity of the intensity with respect to the displayed color value [19, 36]:

ci( ) = (ri · R( ) + gi · G( ) + bi · B( )) . (5)

The absorbance of a substance (A( )), as convention-ally measured with a spectrometer, is independent of the illumination, since it is obtained for a unique broadband illumination. In CSPT, depending on the illuminating col-ors, some bands are highly modulated, and the resulting absorbance ( i( ) and the associated transmittance i( ))does depend on the illumination, according to Equation 6, in which c~i( ) is the normalized ci( ).

i( ) = c~i( ) · A( ), (6)

i( ) = 10- i . (7)

Each polychromatic excitation ci( ) also produces a collection of emissions i( em), (for each single excita-tion line stimulated by ci( )) that can be calculated from the EEM of the tested substance (where i( , em) is the normalized EEM landscape) obtained with a regular fluorescence spectrophotometer:

i ic d( ) ( ) ( , )em em . (8)

In this way, the calculated total transmittance can be written as follows in Equation 9:

i( ) = 1 · i( ) + 2 · i( ), (9)

where 1 + 2 = 1 are the free variables used to fit the model with the measured Fi( ). According to Equations 6–9,

illuminations with varied spectral compositions (e.g.a sequence of illuminating colors) highlight different features of the EEM, which become embedded in the total transmittance.

In Fig. 3, we illustrate the ability of this model to explain and predict the experimental results. Figure 3a shows the contour plots of the excitation emission spectra of three substances measured with regular monochromatic excita-tion (fluorescence spectrometer). The substances are two fluorescent indicators in the sensing array and fluorescein is a common fluorescent dye. Figure 3b shows the mea-sured spectra (Fi( )) of these substances for 50 polychro-matic illuminations. The red line in the fluorescein panel (Fig. 3b) indicates F( ) = 1; values larger than 1 corre-spond to predominant fluorescence, where the detectorreceives more light than through the pure solvent used as reference. For clarity, only ten Fi( ) are highlighted in Fig. 3b with their respective illuminating colors.

Fig. 3. (a) Excitation-emission matrices of Zn(Tpp), and BD and fluorescein. (b) Measured total transmittance of the sub-stances in (a) for an illuminating sequence of 50 polychromatic colors. Ten of these illuminating colors are highlighted in the figure. Black solid lines correspond to the modeled spectra



As can be seen, absorption features in the polychro-matic spectra are well-aligned with the absorption peaks of the excitation-emission spectra. Different illuminating colors specifically highlight characteristic features of the EEM. For instance, in the case of fluorescein, for blue illumination, the whole absorption band can be obtained, while toward cyan illumination the absorption peak nar-rows, becoming a more efficient excitation just limited to the excitation band, showing a larger proportion of fluorescent signal. Illuminating colors in the red region are not able to excite fluorescence and this is reflected in the flat spectra for red light. The numerous features of these spectra and their transitions with the illumination constitute a detailed fingerprint of the EEM that becomes captured in Fi( ).

The complexity of the fingerprints in Fig. 3b under-scores the challenge to provide a unified model able to reproduce all nuances and to generalize to different sub-stances by using a minimum number of free parameters. Modeled results are displayed as black solid lines in Fig. 3b for the ten experimental results highlighted in color. With only two free parameters ( 1 and 2), the model properly reproduces the location of absorption and emission features as well as their specific shape and behavior along the illuminating sequence. Standard mea-sures of fitting goodness [37, 38], applied to a set of five substances including those in Fig. 3b plus rhodamine B and GeTPCCl, corroborated the fingerprinting of the EEM described by the model. Pearson’s r coefficient [37, 38] evaluates the fitting to trends in the data and the root mean square deviation (rmsd) assesses the deviations from exact values. A value of r = 1 indicates a perfect fit and, in our case, an average r = 0.8541 value indicates the ability of the model to generalize to different substances, while the average rmsd = 0.0247 is well within the close fit category (< 0.06) [38].

Finally, in order to obtain an overall picture of the performance of the model and CSPT to retain spectral information from the EEM, we analyzed the classifica-tion patterns of the EEMs. Figure 4 shows the PCA of these data sources and it shows all five substances, well-separated from each other for the EEMs, which is the reference technique (blue asterisks in Fig. 4). Similari-ties between the ZnTPP and GeTPCCl are reflected by their proximity in the PCs space. The classification of the polychromatic fingerprints (black squares in Fig. 4) shows a similar pattern with the comparable amount of information on the first two PCs (about 74%). Red dia-monds in Fig. 4 correspond to the classification of the modeled polychromatic signatures, which is equivalent to the measured fingerprints. The similarity and equiva-lent amount of variance explained in the first two PCs reflects the substantial amount of spectral information retained in the CSPT fingerprints. This is a counterintui-tive result, given the reduced number of spectral bands in CSPT [29], but in good agreement with spectral recon-struction principles [28].

To summarize, in this contribution we have reviewed recent progress in chemical sensing using disposable optical arrays of porphyrins evaluated with CSPT plat-forms. The discussed results show the possibility of performing complex evaluations comparable to e-Nose experiments, but using ubiquitous and highly distributed platforms complemented by handy assays. In particu-lar, the spectral fingerprinting of fluorescent indicators involve complex signatures, and the underlying theo-retical framework explaining them as polychromatically excited signatures of the EEM has been presented. The combination of a friendly assay structure, able to be evaluated on an existing infrastructure, and supported by a robust description of their performance, makes the CSPT fingerprinting of porphyrin arrays an attractive tar-get to explore for distributed sensing applications such as environmental monitoring or diagnostics. The concepts reviewed here are the foundations of that pursuit.

CONCLUSION

Recent progress in spectral fingerprinting of fluores-cent indicators for chemical sensing, using computer screen photo assisted techniques, has been reviewed. Key results collected in this work demonstrate the ability of disposable and handy arrays of porphyrins to support complex classification tasks comparable to e-Nose evalu-ations, but based on highly distributed and readily avail-able infrastructure already owned by potential users, such as computer sets with web cameras.

The theoretical description of polychromatically stim-ulated fluorescent indicators provides the background to understanding these measurements and establishes that CSPT signatures are fingerprints of the excitation emis-sion matrix of fluorescent indicators, able to capture a

Fig. 4. Score plot of the first two principal components of ZnTPP, GeTPC, BD, fluorescein and rhodamine B excitation emission matrices (blue asterisks), polychromatic fingerprints (black squares), and modeled polychromatic fingerprints (red diamonds). Gray areas highlight the different classes



substantial amount of spectral information, a remarkable capacity considering the affordability of the method.

Acknowledgements

This work was supported by grants from the Swedish Research Council, the Swedish Foundation for Strategic Research (SSF) through the program NanoSense [A3 05:204] and a Senior Individual Grant to IL, and by a Marie Curie Early Stage Research Training Fellowship of the European Community’s Sixth Framework Program under contract number MEST-CT-2004-504272 to EG.

REFERENCES

a) Kadish KM, Smith KM and Guilard R. (Eds.). 1.The Porphyrin Handbook, Vol. 1–10, Academic Press: San Diego; 2000. b) Kadish KM, Smith KM and Guilard R. (Eds.). The Porphyrin Handbook,Vol. 11–20, Academic Press: San Diego; 2003.Paolesse R, Mandoj F, Marini A and Di Natale C. In 2.Porphyrin Based Chemical Sensors, Encyclopedia of Nanoscience and Nanotechnology, Nalwa H. (Ed.), Vol. 9, American Science Publishers: 2004; 21.Rakow NA and Suslick KS. 3. Nature. 2000; 406:710–713.Di Natale C, Mignani AG, Paolesse R, Macag-4.nano A and Mencaglia AA. IEEE Sensors J. 2005; 1165–1174.Hart SJ and JiJi RD. 5. Analyst 2002; 127: 1693–1699.Kubista M, Ghasemi J, Sjörgreen B and Forootan 6.A. Spectrosc. Eur. 2004; 16: 8–13.Hall GH and Kenny JE. 7. Anal. Chim. Acta. 2007; 581: 118–124.Potyrailo RA, Morris WG, Leach AM, Sivavec 8.TM, Wisnudel MB and Boyette S. Anal. Chem.2006; 78: 5893–5899, and references therein.Taton TA, Mirkin CA and Letsinger RL. 9. Science2000; 289: 1757–1760.Filippini D, Svensson SPS and Lundström I. 10. Chem.Commun. 2003; 240–241.Filippini D, Alimelli A, Di Natale C, Paolesse R, 11.D’Amico A and Lundström I. Angew. Chem. Int.Ed. 2006; 45: 3800–3803.Bakker JWP, Arwin H, Lundström I and Filippini 12.D. Appl. Opt. 2006; 45: 7795–7799.Macken S, Lundström I and Filippini D. 13. Appl.Phys. Lett. 2006; 89: 25410.Filippini D, Winquist F and Lundström I. 14. Anal.Chim. Acta 2008; 625: 207–214.Filippini D and Lundström I. 15. Analyst 2006; 131:111–117.Röck F, Barsan N and Weimar U. 16. Chem. Rev. 2008; 108: 705–726.

a) Buchler JW. In 17. The Porphyrins, Dolphin D. (Ed.), Vol. I, Academic Press: New York, 1979; 389. b) Grigg R. In The Porphyrins, Dolphin D. (Ed.), Vol. II, Academic Press: New York, 1979; 327.Nardis S, Mandoj F, Paolesse R, Fronczek FR, 18.Smith KM, Prodi L, Montalti M and Battistini G. Eur. J. Inorg. Chem. 2007; 2345–2352.Gatto E, Malik MA, Di Natale C, Paolesse R, 19.D’Amico A, Lundström I and Filippini D. Chem.Eur. J. 2008; 14: 6057–6060.Di Natale C, Martinelli E, Paolesse R, D’Amico A, 20.Filippini D and Lundström I. PLoS One. 2008; 3:e3139–e3150.Filippini D and Lundström I. 21. Analyst 2006; 131:118–125.Filippini D, Comina G and Lundström I. 22. Sensors Actuators B 2005; 107: 580–586.Filippini D, Bakker J and Lundström I. 23. Sensors Actuators B 2005; 106: 302–310.Alimelli A, Pennazza G, Santonico M, Paolesse R, 24.Filippini D, D’Amico A, Lundström I and Di Na-tale C. Anal. Chim. Acta 2007; 582: 320–328.Westland S and Ripamonti C. In 25. Computational Colour Science, Wiley: Chichester, 2004; Chapter 10.Gat N. 26. Proc. SPIE 2000; 4056: 46–50.Bell III JF. 27. et al. Science 2004; 306: 1703–1709.Westland S, Shaw J and Owens H. 28. Sens. Rev. 2000; 20: 50–55.Filippini D and Lundström I. 29. J. Appl. Phys. 2006; 99: 114518.Filippini D and Lundström I. 30. Appl. Phys. Lett.2005; 86: 084101.Filippini D and Lundström I. 31. Anal. Chim. Acta 2006; 557: 393–398.Johnson R and Wichern D. In 32. Applied Multi-variate Statistical Analysis, Pearson Education Ltd., 2002.Jolliffe T. In 33. Principal Component Analysis,Springer: Heidelberg, 1986.Foster DH, Amano K, Nascimento SMC and 34.Foster MJ. J. Opt. Soc. Am. A. 2006; 23: 2359–2372.Bjorklund R, Filippini D and Lundström I. 35. Sens.Actuators B 2008; 134: 199–205.Wyszecki G and Stiles W. In 36. Color Science: Con-cepts and Methods, Quantitative Data and Formu-lae, Wiley: New York, 1982.Read TRC and Cressie NAC. In 37. Goodness-of-Fit Statistics for Discrete Multivariate Data, Springer: Heidelberg, 1988.Hu L and Bentler PM. 38. Struct. Equation Model1999; 6: 1–55.




PHTHALOCYANINE: A SENSING ELEMENT FOR GAS SENSORS

What is a gas sensor?

A gas sensor, and generally, a chemical sensor, is a device that allows a response in real time after receiv-ing a stimulus of a chemical nature. In a general way, it allows the direct decodifi cation of the qualitative and quantitative information of an analyte (gas) in a physical signal. Basically, a gas sensor device consists of a series of elements (Fig. 1), as follows: a) a sensitive membrane

or receptor of the stimulus, b) a transducer that converts it into an electrical or optical signal that can be straight-forwardly measured, and c) a data acquisition system to make possible a signal read-out.

To catalogue a gas sensor, one has to consider the transduction method under operation; therefore, if the species to be detected leads to changes in the number of charge carriers present in the sensing receptor material, one refers to an electrical sensor. On the other hand, if the change is more related to variation in the absorption or emission patterns of electromagnetic radiation from the sensing material, one refers to an optical sensor.

The role of phthalocyanines

Phthalocyanine derivatives (MPc) constitute a notice-able fi gure-of-merit in the fi eld of Organic Electronics. Their broad and unusual physicochemical properties con-fi rm them as possibly the most important, small molecular weight organic material ever used. Three main properties are considered responsible for such a past, present and future in the aforesaid topic of research: i) a high thermal

Electrical transduction in phthalocyanine-based gas sensors:

from classical chemiresistors to new functional structures

Marcel Bouvet*a , Vicente Parrab†, Clémentine Locatellib† and Hui Xiongb

a Institut de Chimie Moléculaire de l’Université de Bourgogne, UMR CNRS 5260, Université de Bourgogne, 9 avenue Alain Savary, BP 47870, 21078 Dijon cedex, Franceb Laboratoire de Chimie Inorganique et Matériaux Moléculaires, UMR CNRS 7071, UPMC Paris 6, 4 place Jussieu, 75252 Paris cedex 05, France

Received 14 October 2008Accepted 25 November 2008

ABSTRACT: Phthalocyanines are organic-based materials which have attracted a lot of research in recent times. In the fi eld of sensors, they present interesting and valuable potentialities as sensing elements for real gas sensor applications. In the present article, and taking some of our experiments as representative examples, we review the different ways of transduction applied to such applications. Some of the new tendencies and transducers for gas sensing based on phthalocyanine derivatives are also reported. Among them, electrical transduction (resistors, fi eld-effect transistors, diodes, etc.) has been, historically, the most commonly exploited way for the detection and/or quantifi cation of gas pollutants, vapors and aromas, according to the conducting behavior of phthalocyanines. We will focus precisely on these systems.

KEYWORDS: transducer, resistor, diode, transistor, heterojunction, sensor, molecular material, molecular semiconductor, phthalocyanine, polluting gas.


*Correspondence to: Marcel Bouvet, email: [email protected], fax: +33 380396098, tel: + 33 380396086

†Current address: Vicente Parra: Pevafersa, Renewable Energies; R&D Department, Parque Tecnológico de Boecillo, Avenida Juan de Herrera 12-14, 47151 Boecillo, Valladolid, Spain; Clémentine Locatelli: Laboratoire Matière Molle et Chimie, ESPCI, 10 rue Vauquelin, 75252 Paris cedex 05, France


ELECTRICAL TRANSDUCTION IN PHTHALOCYANINE-BASED GAS SENSORS 85

and chemical stability compared with other similar organic materials, ii) a rich substitution chemistry leading to an enormous amount of compounds with modulated features, and iii) a very practical processability to build devices, leading to the achievement of a good variety of solid phases by several deposition and growth methods.

As stated before, organic thin fi lms with semicon-ducting properties have been intensively studied due to their very promising applications in Organic Electronics [1, 2]. Contrary to what happens in the case of inorganic semiconductors like silicon, the electrical properties of molecular materials are better described by localized states [3, 4]. Thus, the creation of charge carriers results from an electron transfer between two molecular units, implying an activation energy near 2 eV or higher. The intrinsic density of charge carriers is very low and the so-called molecular semiconductors are actually doped insulators. Among molecular materials, due to their rich substitution chemistry, metallophthalocyanines (MPc) are commonly used in the design of molecular electronic devices, such as organic fi eld-effect transistors (OFETs) [5, 6], solar cells [7, 8], organic light-emitting diodes (OLEDs) [9] and chemical sensors [10]. The well-known reversible electrical property changes of organic materi-als, after exposure to gases, have already been exploited to create highly sensitive gas sensors and transducers [11]. It has been known for some time that oxidizing gases increase the electrical conductivity of metallophthalo-cyanines, whereas fl uorinated MPcs show an increase of conductivity in the presence of donating and/or reduc-ing species. Among them, it is worth mentioning ozone (O3), which is a fairly good oxidizing agent responsible for pollution of urban atmosphere, and ammonia (NH3),which is a donating species used as a freezing liquid in place of freons.

A particular case is that of lanthanide bisphthalocya-nines (LnPc2) [12]. Due to their radical nature, they can easily lose and gain an electron, yielding very low acti-vation energy (0.5 eV) for the creation of charge carri-ers, ca 1016 cm-3. The lutetium derivative (LuPc2) was the fi rst intrinsic molecular semiconductor discovered, lead-ing to thin-fi lm conductivities of ca 5.10-5 (Ω.cm)-1 [13, 14]. For the same reason, electrochemical sensors were studied with lanthanide bisphthalocyanines and electron-ic tongues were developed for wine discrimination [15]. Additionally, the various redox states differ by their opti-cal properties, making optical transduction feasible [16].

The energy levels, which are responsible for electrical properties of the main phthalocyanines studied hereafter, were determined recently by ultraviolet photoelectron spectroscopy (UPS) and inverse emission photoelec-tron spectroscopy (IEPS) (Fig. 2) [17]. The energy gap of non-radical phthalocyanines is ca. 2 eV, which noticeably contrasts with that of LuPc2 [18, 19], but their energy levels (HOMO and LUMO) are stabilized from non-substituted to perfl uorinated molecules.

TRANSDUCERS

Apart from resistors, organic fi eld-effect transistors (OFET) have also been introduced as transducers for chemosensing of ozone [20–22], volatile organic com-pounds [23], and glucose [24]. All transducers discussed are depicted in Fig. 3. Very recently [25], the authors associated p- and n-type molecular doped insulators in diode-like p-n junctions and studied them as potential transducers for gas sensing. Thus, a bilayer made from hexadecafl uoro-nickel phthalocyanine, Ni(F16Pc), and the unsubstituted analogue (NiPc), exhibited strong variations in the built-in potential of the junction and in

Fig. 1. Functional scheme of a gas sensor

Est ímulo qu ímico(mol éculas de gas,

líquido , etc.)

Receptorsensible

Transductor

SEÑAL


líquido , etc.) sensible

Transductor

SEÑAL


líquido , etc.)

Receptorsensible

Transductor

SEÑAL

líquido , etc.)líquido , etc.)

Receptorsensible

Transductor

SEÑAL

ReceptorTransducer

SIGNAL

Generation ofchemical information

De-codification toan physical signal

Signal acquisition,treatment

and read-out

Stimulusby gas molecules


86 M. BOUVET ET AL.

its rectifi cation ratio, after exposure to ammonia. Even though FETs were prepared with LuPc2, such a structure is not adapted to highly conductive molecular materi-als in which the bulk conductivity can only be slightly modulated by gate voltages [26]. In the present paper, we will develop the use of these phthalocyanines, fi rst in resistors made from pure compound, as well as from polymer-phthalocyanine hybrid material and, second, in a new type of transducer called molecular semiconduc-tor-doped insulator heterojunction (MSDI) [27].

Resistors

The resistor remains the most studied semiconductor-based transducer. Gas detection, by measuring the con-ductivity variation of semiconducting materials, has been performed mainly with metal oxides such as SnO2-x [28]. The main drawbacks of these metal oxides are, fi rst, their lack of selectivity, even if metallic catalyzers like Pd, Pt and Ir are added and, second, the high operating tempera-tures required (200–400 °C).

In molecular materials, redox processes can cause an important variation of the charge carrier density. Thus, the sensitivity of phthalocyanine-based resistors to nitro-gen dioxide (NO2) is well established [29, 30]. Contrary to NO2, only a few studies devoted to ozone (O3) sens-ing have been reported [20, 21, 31]. The main reason for such a limitation being the short lifetime of ozone, which makes low concentration of its production in the laboratory diffi cult to achieve. Both gases are responsible for the main pollution peaks in urban areas. Particular methodologies were applied to improve the O3 vs NO2

selectivity, which is based on different interaction kinet-ics with molecular materials [32].

The NO2 effect on the conductivity of LuPc2 thin fi lms appears very different from that observed on non-radical phthalocyanines. LuPc2 fi lms exposed to NO2 concen-trations in the range 2–300 ppm, then de-doped under a nitrogen fl ow, show that after a sharp increase, conductivity goes through a maximum, then decreases, whereas the doping gas is still present (Fig. 4) [33]. In the fi rst step,

Fig. 2. Schematic view of the energy levels (LUMO and HOMO or SOMO) involved in the electrical properties of phthalocya-nines studied here, as deduced from cyclic voltammetry measurements (LuPc2) [18] and by photoelectron spectroscopies (UPS and IEPS) [17]

a) b)

c)

d)

Fig. 3. Schematic view of the different transducers reviewed in this paper: a) resistors, b) OFETs, c) diode-like structures, and d) the brand-new MSDIs



Fig. 4. Conductivity variation of radical phthalocyanines’ thin fi lms as a function of time during exposure to NO2, a) for a polycrystalline fi lm of LuPc2 [33], and b) for an amorphous fi lm of a naphthalocyanine analogue of LuPc2 [34]. In the latter case, the mass variation is indicated

0

0.25

0.5

σ.10

3 , Ω−1

cm−1

0.75

1

1.25

0 0.5 1 1.5 2 2.5 3t, hr

m

t

σ

mNO2 (2 ppm)

N2

σ.10

3 , Ω−1

cm−1

0

1

2

3

4

5

6

7

8

0

0.5

1

1.5

2

0 100 200 300 400 500t, min

mΔ,μ

g

NO2 (10 ppm) N

2

a)

b)

0 400 800 12000 400 800 1200

I x 1

0-5,

A

t, s

NH3 / dry Ardry Ardry Ar dry Ardry Ar

93 ppm

280 ppm

510 ppm

778 ppm

NH3 / dry ArNH3 / dry Ardry Ardry Ardry Ardry Ar dry Ardry Ardry Ardry Ar

93 ppm93 ppm

280 ppm280 ppm

510 ppm510 ppm

778 ppm778 ppm

6.80

6.84

6.88

6.92

6.80

6.84

6.88

6.92

Fig. 5. Variation of the current of a LuPc2 resistor submitted to various NH3 concentrations

NO2 acts as a doping agent and conductivity increases; therefore, when LuPc2

+ concentration increases notice-ably, intermolecular charge transfer with a neutral neigh-bouring molecule becomes less probable, leading the conductivity to decrease. The latter step can actually be explained in terms of a drop in charge carrier mobility.

and σ = [LuPc2+] e μ.

The time tm required to attain the highest variation depends on the crystalline properties of the fi lm. Thus, 3000 Å-thick fi lms prepared on a substrate maintained at 200 °C (polycrystalline, α-phase) exhibit a tm longer than the 300 Å-thick fi lms (amorphous) deposited at 80 °C; additionally, the lower the gas concentration, the longer the tm. The mass variation Δm, due to the gas adsorption (which is measured with a quartz microbalance through-out the doping process) is not directly related to the conductivity variation (Fig. 4). It was shown that 4 NO2

molecules are adsorbed per phthalocyanine molecule at tm, for LuPc2 fi lms. Moreover, the stronger the oxidiz-ing agent, the lower the number of oxidizing molecules

needed to attain the maximum conductivity value, e.g.only 0.3 in the case of bromine. This indicates that, with NO2, the electron transfer rate from phthalocyanine is lower than unity. The adsorption process can continue longer than tm, up to 8 NO2 molecules per phthalocya-nine and 6 bromine atoms. These results show the strong effects of adsorption and diffusion phenomena during the sensing experiments and therefore, the necessity to control and master the structure and morphology of the sensing layers becomes a critical issue. As a result, it was demonstrated that amorphous fi lms prepared from LuPc2

analogues led to shorter reaction times than polycrystal-line fi lms [12, 34].

Nevertheless, with weaker redox active species, that behavior is markedly different [35, 36]. Figure 5 illustrates the effect of different NH3 concentrations(from different mass fl ow rates), achieved by diluting with dry argon, on the electrical properties of a LuPc2

vacuum-evaporated fi lm. As expected for an electron donor gas interacting with a p-type semiconductor, the rapid decrease of the device’s current is clearly visible just after introducing the NH3 fl ow. This trend is more marked when increasing the concentration, but it is worth commenting that the larger the amount of NH3,the lesser the tendency to attain a plateau-like response during the exposure, especially at higher concentrations. This phenomenon is associated, after gas adsorption, with bulk effects caused by the diffusion of gas mole-cules. Then, the original surface of the fi lm is renewed. Once the NH3 fl ow is discontinued, the current begins to increase, due to the molecular gas desorption (dynam-ic rest in presence of dry argon). Indeed, the kinetics of this process are directly related to the NH3 concentration, i.e. they depend on the above-mentioned bulk effects. This retains the diffused gas molecules within the fi lm (in grains, structural defects, etc.). In any case, even after a 15 min exposure to high NH3 concentrations, conduc-tivity does not go through a maximum value, as with NO2.

Recently, experiments to control the attainment of that undesirable maximum conductivity in LuPc2 when exposed to a strong oxidizing agent were leading us to LuPc2 + NO2 LuPc2

+ + NO2-


88 M. BOUVET ET AL.

consider the formation of hybrid fi lms by incorporating an insulating polymer matrix in the fi lm, namely poly(methyl methacrylate) (PMMA). Best results were obtained with 80/20 (w/w) LuPc2/PMMA mixtures, deposited by sol-vent cast from a CH2Cl2 solution. Under O3 (400 ppb), the current increases sharply for 300 s, then slightly (Fig. 6a), but does not go through a maximum value, even after 103 s. When the exposure time is reduced to 120 s, only a continuous, but not linear, sharp increase is observed. The differentiation of the curve I = f(t) gives access to a response of the sensor independent of the exact exposure time, and allows the typical current drift to be eliminated. Thus, at 100 ppb, the response given by (dI/dt)max, which corresponds to the maximum of the slope in the I = f(t) curve, is about 0.045 nA.s-1, showing a good reproduc-ibility (Fig. 6b). This experiment has been carried out in static rest conditions, i.e. without any fl ow between exposure periods to O3, in repeated exposure-static rest cycles (2/8 min). This study highlights the potential interest of molecular semiconductor-insulating polymer hybrid fi lms as novel sensing materials.

Organic fi eld-effect transistors (OFETs)

In the past, we presented the fi rst OFETs based on MPcs for gas sensing applications [20, 21]. The device

was prepared from a silicon wafer using a ten-stage method by classical microelectronic techniques. The MPc fi lm (250 Å thick), in this case NiPc, was deposited by vacuum sublimation.

The sensitivity of the device was determined by exposure-static rest cycles (2/8 min) at O3 concentrations in the range 10–300 ppb. A conditioning period is neces-sary before obtaining a steady state.

The response ΔI (Iexposure - Irecovery) of the device as a function of the O3 concentration has a value of 65 pA.ppb-1

(or 0.40%.ppb-1) during the 2 min-long exposure periods, i.e. 32 pA per ppb and min, in the range 10–300 ppb (Fig. 7). Nevertheless, the IDS current does not vary lin-early as a function of time during the exposure periods. This is the reason why a better view of the response of the sensor was given by the derivative dIDS/dt. The maximum value of the derivative was directly correlated to the O3

concentration (Fig. 8).The current at saturation, IDSsat, can be taken as the

response of the sensor, but also the ratio Ion/Ioff, i.e. the ratio between the current at a given gate voltage and the current at VG = 0. Actually, it is not trivial to vary the applied voltages to obtain a stable response of the device

Fig. 6. a) Variation of the current of a LuPc2/PMMA (80/20) hybrid fi lm exposed to ozone (about 400 ppb, then pure air, then no gas), b) derivative dI/dt as a function of time for a resistor made from a LuPc2/PMMA (80/20) hybrid fi lm submitted to 110 ppb ozone exposure-static rest cycles

a)

b) Fig. 7. Variation of the IDS current of a NiPc fi eld-effect transis-tor as a function of the O3 concentration for 2 min/8 min expo-sure-static rest cycles, and the corresponding relative variation ΔI/I (VG = VD = -3 V, VS = 0 V)

Fig. 8. Variation of the temporal derivative dIDS/dt of a NiPc fi eld-effect transistor exposed to different O3 concentrations (VG = VD = -3 V, VS = 0 V)



in the time scale of the exposure-static rest cycles. Thus, such a multimodal detection has never been applied.

Low energy consumption and a miniaturization of the signal processing allowed the preparation of an autono-mous system as well as a portable apparatus. Figure 9 shows the response of a sensor exposed to the ambient air enriched in O3 by a UV lamp, compared to the con-centration given by a commercial O3 analyzer.

The autonomous system has been used for one month in suburb of Paris (France). At a day scale, the (dIDS/dt)max

value corresponds to the O3 concentration except at low concentrations (Fig. 10a). Indeed, when the NO2 con-centration is maximal, the sensor gives an overestimated value of the O3 concentration. This is not surprising since the O3 concentration is minimal when the NO2 value is maximal (Fig. 10b). The dynamic methodology used improves the sensitivity to O3 because the kinetics of reaction of O3 is higher than that of NO2, as shown by Pauly et al. [37]. Clearly, the response of the sensor is more related to O3 than to the sum of the two oxidizing gases, which indicates the total oxidizing pollution. That sum varies much less than the individual concentrations (Fig. 10b).

Whatever the methodology applied, the use of a selec-tive fi lter highly improves the response of sensors, as has been recently shown [32, 38].

Diode-like structures

These devices are certainly interesting, thanks to their unique electrical features compared to, for instance, classical resistors: i) the rectifi cation phenomenon; ii) the potential variety of parameters that can be exploited, considering a multimodal detection. The literature reports few papers on sensors based on Schottky diodes [39, 40]. Even more scarce is the investigation on p-ndiode heterojunctions. In fact, in all these cases, the sensing elements are made from inorganic materials or conducting polymers [41].

The use of organic p-n heterojunctions for organic electronics applications has currently become a very active research fi eld, with many high-quality applications in photovoltaic cells [42], OFETs [43] and organic electroluminescent devices (OLEDs) [44]. The next stage will be the development of sensors that can take advan-tage of the electrical parameters according to the physics of the device.

Recent investigations have demonstrated that sand-wich structures Au|Ni(F16Pc)|NiPc|Al show interesting features (Figs 11a and 11b) when being exposed to Ar and NH3 atmospheres, owing to the electrical behavior exhibited by both the metal/organic and organic/organic junctions (pseudo-Schottky device’s structure); namely, i) the monolayered devices show no rectifi cation behav-ior, ii) NH3 gives rise to a sensible increase of forward and direct currents in the sandwich device, iii) the rec-tifi cation ratio is clearly modifi ed (up to 95%), but the I-V pattern, in both regimes of polarization, never attain a symmetrical aspect.

These devices make possible a multimodal prin-ciple of detection, controlled by a combined effect between the organic heterojunction and the metal/ organic contacts.

10

20

30

40

50

0

20

40

60

80

100

0 500 1000 1500

Sens

or R

espo

nse,

mV

[O3]

mic

rog,

m3

t, min

Fig. 9. Response of a sensor exposed to the ambient air enriched in O3 by a UV lamp compared to the concentration given by a commercial O3 analyzer

0

20

40

60

80

100

120

0

20

40

60

80

100

120

0 500 1000 1500

O3 (microg/m3)

NO2 (microg/m3)

O3 + NO2 (microg/m3)

mic

rog,

m

t, min

3

Fig. 10. Response of a sensor measured for one day in an air quality control site of Airparif (Paris, France), after signal pro-cessing, compared to the concentration given by a commercial O3 analyzer a); and a typical daily variation of the O3, and NO2

concentrations as well as their sum

a)

b)


90 M. BOUVET ET AL.

Molecular semiconductor-doped insulator heterojunctions (MSDIs)

New organic devices, namely Molecular Semiconduc-tor-Doped Insulator (MSDI) heterojunctions have been proposed as a new transducer for gas chemosensing [27]. They are built around a heterojunction between a thin fi lm of a molecular semiconductor (MS), the lutetium bisphthalocyanine (LuPc2) complex and a thin fi lm of a doped-insulator (ΔI) one, non-substituted or fl uorinated copper phthalocyanine (Cu(FnPc), where n = 0, 8, 16). Whereas a LuPc2 resistor exhibits linear behavior of its I(V) characteristics, the MSDIs show non-linear but sym-metrical I(V) curves (not shown here). The energy barrier is higher for F16 due to the nature of the majority charge carriers, namely negative in F16 and positive in LuPc2.F8 leads to an intermediate electrical signature.

The current of the MSDI LuPc2/F0 prepared with CuPc as sub-layer increases under O3 (90 ppb) but does

not decrease after a long exposure, even after 103 s-long exposure period. Clearly, the heterojunction between the MS and the ΔI plays a key role in the MSDI behavior. Under NH3 (50 ppm), the same device shows a current decrease. The electrical behavior is related to the den-sity of majority charge carriers, which increases under O3

and decreases under NH3. On the contrary, MSDI LuPc2/F16, prepared with Cu(F16Pc) as sub-layer, yield a nega-tive response to O3 and a positive response to NH3, as a result of an increase of the energy barrier at the MSDI heterojunction. Due to the particular redox properties of LuPc2, the MSDIs are very sensitive, but without the sat-uration effect observed in LuPc2 resistors. In the MSDI devices, the semiconductor layer is the only organic fi lm exposed to the atmosphere, whereas the doped-insulator one plays the role of modulator of the effective charge carrier’s nature, by tuning the electronic characteristics of the organic heterojunction.

CONCLUSION

Phthalocyanines are very attractive organic-based materials for sensing elements in gas sensors, thanks to the molecule’s tunability, and the varied thin fi lm states (i.e. crystalline structures and morphologies). In gas sensors, the responses depend on multiple parameters, namely: the physical properties of the gases, the sensing element features and the presence of interfering species. To optimize the effi ciency of the device, different trans-ducers (electrical, optical, etc.) can be designed according to the large series of physicochemical properties of those functional molecular materials, especially the radical-stable lanthanide bisphthalocyanine derivatives, allowing profi table multimodal detection.

Among the transduction methods reported up to now, one can emphasize resistors and fi eld-effect transistors. Other devices based on diodes and new structures such as the molecular semiconductor-doped insulator fi lms (MSDI) give rise to a new generation of gas sensors based on organic semiconductors, directed towards the accomplishment of pioneering, sensing performances from advanced but accessible functional materials. This can improve on the best-known models thus far, the classical and rather operation-limited, single-layer chemiresistors, without the complexity of transistor technology.

Acknowledgements

The authors acknowledge the Agence Nationale de la Recherche (A. N. R., France) for funding under POLL-CAP project Blan-06-154000. Dr. M. L. Rodríguez-Mé-ndez and Prof. A. Pauly are acknowledged for supplying the electrode substrates. V. P. thanks the City of Paris for a post-doctoral grant. H. X. thanks the Erasmus Mundus program MONABIPHOT (E.N.S. Cachan) and M. B. the E.S.P.C.I. for fi nancial support.

-10 -5 0 5 10

-4x10-6

-3x10-6

-2x10-6

-1x10-6

0

1x10-6

2x10-6

3x10-6

4x10-6

5x10-6

6x10-6

I,A

Bias, V

AlAu

NiPc

glass

AlAu

NiPc

glass

AlAu

Ni(F16Pc)

glass

AlAu

Ni(F16Pc)

glass

Al

AuNi(F16Pc)

NiPc

glass

Al

AuNi(F16Pc)

NiPc

glass

-10 -5 0 5 10

-2.0x10-5

-1.0x10-5

0.0

1.0x10-5

2.0x10-5

Bias, V

Ar

i/A

NH3

exposure

NH3

NH3

Fig. 11. a) Current-voltage characteristics of Au|NiPc|Al (�),Au|Ni(F16Pc)|Al (�) and Au|NiPc|Ni(F16Pc)|Al (�) devices. Inset: schematic views of the samples’ structures; b) Effect of NH3 (from a 10% aqueous solution) on the I–V behavior of Au|NiPc|Ni(F16Pc)|Al, having both organic fi lms exposed to the atmosphere

a)

b)



REFERENCES

Forrest SR. 1. Chem. Rev. 1997; 97: 1793–1896.Forrest SR and Thompson ME. 2. Chem. Rev. 2007; 107: 923–925.Silinsh EA. 3. Organic Molecular Crystals: Their Electronic States, Springer Verlag: Berlin, 1980.Silinsh EA, Bouvet M and Simon J. 4. Mol. Mater.1995; 5: 1–24.Madru R, Guillaud G, Al Sadoun M, Maitrot M, 5.Clarisse C, Le Contellec M, André J-J, and Simon J. Chem. Phys. Lett. 1987; 142: 103–105.Madru R, Guillaud G, Al Sadoun M, Maitrot M, 6.André J-J, Simon J and Even R. Chem. Phys. Lett.1988; 145: 343–346.Tang CW. 7. Appl. Phys. Lett. 1986; 48: 183–185.Tsuzuki T, Shirota Y, Rostalski J and Meissner D. 8.Sol. Energy Mater. 2000; 61: 1–8.Armstrong NR. 9. J. Porphyrins and Phthalocyanines2000; 414–417.Bott B and Thorpe SC. Metal phthalocyanine gas 10.sensors. In Techniques and Mechanisms in Gas Sensing, Moseley PT, Norris J and Williams DE (Eds.). Adam Hilger: Bristol, 1991.Bouvet M and Pauly A. Molecular semiconductor-11. based gas sensors. In The Encyclopedia of Sensors,Grimes CA, Dickey EC and Pishko MV (Eds.). American Scientifi c Publishers: 2006; Vol. 6, pp. 227–270.Bouvet M. Radical phthalocyanines and intrin-12.sic semiconduction. In The Porphyrin Handbook,Kadish KM, Guilard R and Smith K. (Eds.). Aca-demic Press: 2003; Vol. 19, pp. 37–103.André J-J, Holczer K, Petit P, Riou M-T, Clarisse 13.C, Even R, Fourmigué M and Simon J. Chem. Phys. Lett. 1985; 115: 463–466.Turek P, Petit P, André J-J, Simon J, Even R, Boud-14.gema B, Guillaud G and Maitrot M. J. Am. Chem. Soc. 1987; 109: 5119–5122.Parra V, Arrieta AA, Fernandez-Escudero JA, Gar-15.cia H, Apetrei C, Rodriguez-Mendez ML and De Saja JA. Anal. Chim. Acta 2006; 563: 229–237.Valli L. 16. Adv. Colloid Interface Sci. 2005; 116:13–44.Murdey R, Sato N and Bouvet M. 17. Mol. Cryst. Liq. Cryst. 2006; 455: 211–218.L’Her M, Cozien Y and Courtot-Coupez J. 18. J. Elec-troanal. Chem. 1983; 157: 183–187.Bufl er J, Abraham M, Bouvet M, Simon J and Göpel 19.W. J. Chem. Phys. 1991; 95: 8459–8466.Bouvet M, Guillaud G, Leroy A, Maillard A, Spirk-20.ovitch S and Tournilhac F-G. Sensors Actuators B2001; 73: 63–70.Bouvet M, Leroy A, Simon J, Tournilhac F, Guillaud 21.G, Lessnick P, Maillard A, Spirkovitch S, Debliquy M, de Haan A and Decroly A. Sensors Actuators B2001; 72: 86–93.

Bouvet M. 22. Anal. Bioanal. Chem. 2006; 384: 366–373.Crone B, Dodabalapur A, Gelperin A, Torsi L, Katz 23.HE, Lovinger AJ and Bao Z. Appl. Phys. Lett. 2001; 78: 2229–2231.Bartic C, Campitelli A and Borghs S. 24. Appl. Phys. Lett. 2003; 82: 475–477.Muzikante I, Parra V, Dobulans R, Fonavs E, Latv-25.els J and Bouvet M. Sensors 2007; 7: 2984–2996.Guillaud G, Al Sadoun M, Maitrot M, Simon J and 26.Bouvet M. Chem. Phys. Lett. 1990; 167: 503–506.Parra V and Bouvet M. Submitted.27.Watson J, Ihokura K and Coles GSV. 28. Meas. Sci. Technol. 1993; 4: 711–719.Jones TA and Bott B. 29. Sensors Actuators B 1986; 9:27–37.Dogo S, Germain J-P, Maleysson C and Pauly A. 30.Thin Solid Films 1992; 219: 251–256.Schütze A, Pieper N and Zacheja J. 31. Sensors Actua-tors B 1995; 23: 215–217.Brunet J, Pauly A, Mazet L, Germain J-P, Bouvet 32.M and Malézieux B. Thin Solid Films 2005; 490:28–35.Pauly A. Ph.D., Université Blaise Pascal-Clermont 33.II, Clermont-Ferrand, 1993.Passard M. Ph.D., Université Blaise Pascal-Cler-34.mont II, Clermont-Ferrand, 1995.Parra V, Bouvet M, Brunet J, Rodriguez-Mendez 35.ML and De Saja JA. Thin Solid Films 2008; 516:9012–9019.Rodriguez-Mendez ML, Souto J, De Saja R, Mar-36.tinez J and De Saja JA. Sensors Actuators B 1999; 58: 544–551.Brunet J, Germain J-P, Pauly A, Lauron B and 37. Varenne C. Methodology of measurement for phthalocyanines gas sensors dedicated to the detec-tion of ozone and nitrogen dioxide. Results in urban air pollution context. In Sensors for Environmental Control, Siciliano P. (Ed.). World Scientifi c Pub-lishing: Singapore, 2003; pp. 188–192.Viricelle JP, Pauly A, Mazet L, Brunet J, Bouvet M, 38.Varenne C and Pijolat C. Mater. Sci. Eng. C 2006; 26: 186–195.Talazac L, Barbarin F, Varenne C and Cuminal Y. 39.Sensors Actuators B 2001; 77: 447–454.Talazac L, Barbarin F, Varenne C, Mazet L, Pellier 40.S and Soulier S. Sensors Actuators, B 2002; 83:149–159.Janata J and Josowicz M. 41. Nat. Mater. 2003; 2:19–24.Sicot L, Geoffroy B, Lorin A, Raimond P, Sentein 42.C and Nunzi J-M. J. Appl. Phys. 2001; 90: 1047–1054.Wang H, Wang J, Huang H, Yan X and Yan D. 43. Org. Electr. 2006; 7: 369–374.Lai SL, Chan MY, Fung MK, Lee CS and Lee ST. 44. J. Appl. Phys. 2007; 101: 014509.




INTRODUCTION

Previously, we have shown for the fi rst time that incorporation of the ruthenium(II) tetra-15-crown-5-pht-halocyaninate with axially coordinated triethylenediamine molecules (R4Pc)Ru(TED)2 (R4Pc2- = [4,5,4′,5′,4′′,5′′,4′′′,5′′′-tetrakis-(1,4,7,10,13-pentaoxapentadecamethylene)-phthalocyaninate-anion], TED – triethylenediamine) into electroconducting polymers with high glass transi-tion temperature Tg leads to solid-state photorefractive composites [1–3]. It has been proposed that photoelectric

and photorefractive sensitivities of polymer composites to IR radiation is caused by the formation of supramo-lecular ensembles of complexes. In particular, it should be mentioned that it has been experimentally proven that phthalocyanines are not only photorefractive sensitizers in the hard-hitting IR region, but can also be, simulta-neously, non-linear optical (NLO) chromophores pro-viding a photorefractive effect. It is signifi cant that the photorefractive polymer composites with spectral sensi-tivity in the near-IR region offer wide opportunities for numerous photonic applications. It is especially inter-esting for biomedical diagnostics and optical communi-cation. The chaotic orientation of the chromophores is frozen in the composites with high Tg. In this case, the possibility of orientation polarization is eliminated; the

Solvent-induced supramolecular assemblies of

crown-substituted ruthenium phthalocyaninate:

morphology of assemblies and non-linear optical properties

Antonina D. Grishinaa, Yulia G. Gorbunova∗a,b , Victor I. Zolotarevskya,

Larisa Ya. Pereshivkoa, Yulia Yu. Enakievaa , Tatiana V. Krivenkoa,

Vladimir Savelyeva, Anatoly V. Vannikova and Aslan Yu. Tsivadzea,b

a Russian Academy of Sciences, Frumkin Institute of Physical Chemistry & Electrochemistry, Leninsky pr., 31, Moscow, GSP-1, 119991, Russiab Russian Academy of Sciences, Kurnakov Institute of General & Inorganic Chemistry, Leninsky pr., 31, Moscow, GSP-1, 119991, Russia

Received 26 October 2008Accepted 21 November 2008

ABSTRACT: The information concerning the architecture of supramolecular assemblies, which are composed of ruthenium(II) tetra-15-crown-5-phthalocyaninate with axially coordinated triethy-lenediamine molecules, has been obtained with the application of atomic force microscopy. The ensembles are formed in Ru(II) complex solutions in chloroform and tetrachloroethane. The number of molecules and their relative orientation in supramolecular assemblies were estimated in dependence of the solvent (chloroform, tetrachloroethane) and the heating temperature. Samples fabricated after heating of the Ru(II) complexes solution in tetrachloroethane formed stable supramolecular wires of 600 nm and more in length. The third-order non-linear optical characteristics of complexes in tetrachloroethane solution were studied by the z-scanning method. Molecular polarizability of the complex is about 4.5 × 10-32 esu. The polarizability attributed to one molecule increases by a factor of 3.6 when the individual molecule assembles into a supramolecular aggregate.

KEYWORDS: phthalocyanine, ruthenium, crown-ether, supramolecular ensembles, atomic force microscopy, z-scanning, photorefractive properties, non-linear optical properties.


*Correspondence to: Yulia G. Gorbunova, email: [email protected], fax: +7 495-952-25-66, tel: +7 495-955-48-74


SOLVENT-INDUCED SUPRAMOLECULAR ASSEMBLIES OF CROWN-SUBSTITUTED RUTHENIUM PHTHALOCYANINATE 93

second-order susceptibility goes to zero and only the third-order susceptibility has a non-zero value as shown in Equation 1:

χ(3) = Nγ f 4�cos4ξ�, (1)

where N is the concentration of the chromophores, γ is the third-order molecular polarizability, f = (n0 + 2)/3 is the Lorentz fi eld factor. The mean value of �cos4ξ � equals 1/5 at a random distribution of orientation angles ξ of NLO chromophores.

The high glass transition temperature polymers improve the stability of the characteristics of photore-fractive composites but require the application of nano-dimensional NLO chromophores. Indeed, third-order susceptibility is very low for media containing short-dipole chromophores but increases as χ(3) ~ l 2.4 with the increase of the exciton delocalization length l [4, 5] and achieves the highest values for nano-dimensional forma-tions. Thus, elaboration of the photorefractive materials based on the high Tg polymers requires the use of non-linear optical chromophores with an extended system of the collective electron excitation.

Molecular phthalocyanines (Pc) exhibit large third-order non-linearity [6–9]. The measurements of the third harmonic generation (THG) in related compounds have shown that molecular aggregation, in some cases, enhances the THG response [8]. Using the degenerate

four-wave mixing experiments for silicon phthalocya-nine oligomers SiPcO, it was found that the formation of the slipped-stack arrangement (SiPcO)n enhances γas n3 [9].

Previously, the structure of the (R4Pc)Ru(TED)2⋅7CHCl3

single crystals has been studied using the X-ray dif-fraction method [10]. As shown in Fig. 1, molecules of Ru(II)-complex form a “brickwork” structure in the crys-tal cell. It leads to a considerable broadening of optical absorption at the Q-band [11]. It has been shown that only in the case when the aromatic polyimide (API, Tg = 240 °C) [1, 2] or polyvinylcarbazole (Tg = 200 °C) [3] was added to the repeatedly heated tetrachloroethane solution of (R4Pc)Ru(TED)2, the corresponding composites pos-sessed optical absorption, photoelectric and photorefrac-tive sensitivities at 1064 nm [1–3]. For example, the main measured photorefractive characteristics of composites based on API, namely, the two-beam gain coeffi cient Γ, the net gain coeffi cient – α

0 (α0 = 37.7 cm-1 is the

absorption coeffi cient of the composite at 1064 nm), and the response time τ, are given in Table 1. It is evident that Γ grows and τ reduces with an increase of the applied fi eld E0. As seen in Table 1, the two-beam gain coeffi -cient Γ and the net gain coeffi cient ( – α

0) achieve high

values equal to 140 and 102.3 cm-1, correspondingly, at 110 V.μm-1. The response time equals 0.1 s at 110 V.μm-1

[1, 2]. It was suggested that these characteristics are due

Fig. 1. (a) Structure of (R4Pc)Ru(TED)2 and (b) X-ray diffraction data of its crystal packing


94 A. D. GRISHINA ET AL.

to the formation of supramolecular ensembles as the optical absorption of the (R4Pc)Ru(TED)2 tetrachloro-ethane solution expands to the IR-region, including the wavelength at 1064 nm, after heating of the solution [1–3]. The goal of this work is to study solvent-induced supramolecular assembling of ruthenium(II) tetra-15-crown-5-phthalocyaninate with axially coordinated tri-ethylenediamine molecules and their non-linear optical characteristics by direct methods.

EXPERIMENTAL

Crown-substituted ruthenium phthalocyaninate (R4Pc)Ru(TED)2 was synthesized by a procedure described in the literature [10]. The individual complex was isolated in 90% yield by column liquid chromatography on neutral alumina using a CHCl3/MeOH (99:1) mixture as eluent. The samples for AFM measurements were produced by the casting of the (R4Pc)Ru(TED)2 solutions (1.2 × 10-4

M) in chloroform or tetrachloroethane on mica discs of 15 mm in diameter. Fresh tetrachloroethane solutions of (R4Pc)Ru(TED)2 were sonicated for 5 min, after which they were transferred to the mica discs: (1) immediately after sonication (fresh solution), (2) after keeping the

solution for 18 h at rt and (3) after 4 heating cycles (3 min heating of the solution at 70 °C and cooling to room temperature) (heated solution). In the case of chloroform, the fresh (R4Pc)Ru(TED)2 solution, as well as the solu-tion heated for 7 cycles to 50 °C, were used.

The AFM investigation of complex assembling was carried out with AFM model MultiMode and control-ler Nanoscope IV, Veeco Corporation. Image acquisi-tion was performed in tapping mode. Silicon Cantilevers NSG 3 (resonance frequency 80 kHz, force constant 1.1 Nm-1, curvature radius 10 nm) and NSG 01 (resonance frequency 150 kHz, force constant 5.5 Nm-1, curvature radius 10 nm) from NT-MDT Corporation (Russia, Mos-cow) were used. Data were processed by software WSxM 4.11 [12] Nanotec Electronica (Spain, Madrid).

The electronic absorption spectra were recorded on a Shimadzu UV-3101PC spectrophotometer. The third-order optical non-linearity was estimated by the z-scanmethod using a tetrachloroethane solution of the (R4Pc)Ru(TED)2 complex.


AFM measurements

Figure 2a demonstrates the image of (R4Pc)Ru(TED)2

ensembles on the mica surface. The ensembles were formed in fresh chloroform solution and cast onto mica disks. These ensembles are cylindrical formations of approximately 2 nm in height (Fig. 2b) and an average of 50 nm in diameter. In all cases, the term height was used for the distance between the mica surface and the top of aggregates.

As mentioned above, we have previously studied the structure of the single crystal containing seven chloro-form molecules per one (R4Pc)Ru(TED)2 molecule in a crystal cell using the X-ray diffraction method [10]. Mole-cules of Ru-complex form a structure of “brickwork” in the crystal packing (Fig. 1b). Chloroform molecules

Table 1. The electric fi eld dependences of the two-beam gain coeffi cient Γ, the net gain coeffi cient Γ– α0 and response time τat 1064 nm in the composites consisting of aromatic polyimide, (R4Pc)Ru(TED)2 and ferrocene. The cw Nd : YAG laser was used

E0, V.μm-1 , cm-1 – 0, cm-1 , s

37 48 10.3 2.2

57 63.5 25.3 1.6

67 79 41.3 0.6

90 100 62.3 0.25

110 140 102.3 0.1

Fig. 2. (a) AFM image of (R4Pc)Ru(TED)2 ensembles cast on mica surface from the fresh solution in chloroform, (b) the heights of Z nm along the profi le indicated on the image by a white line, (c) the relative distribution of the heights of ensembles



in the given crystal package are located between the crown-ether fragments of neighboring phthalocyanine molecules and bind them into extended supramolecular assemblies of a “brickwork” structure. It is signifi cant that the obtained single crystals are not stable at room temperature and decompose while chloroform molecules evaporate. We have suggested that the formation of simi-lar supramolecular ensembles, as in the case of single crystals, should be observed in thin fi lms of Ru-complex fabricated from the solution in chloroform. Using data from X-ray analysis for (R4Pc)Ru(TED)2⋅7CHCl3 and the results of AFM investigations, we were able to esti-mate a mean quantity of molecules and molecular layers in the formed ensembles. The volume of the molecular cell in the single crystal equals 3258.6 Å3 and the N...Ndistance between outer nitrogens of the triethylene-diamine molecules in a Ru(II) complex is defi ned as 9.130 Å. The following calculations (250 Å × 250 Å ×3.14 × 20 Å/3258.6 Å3 = 1204 molecules) lead to the conclusion that approximately 1200 complex molecules are located in the ensemble formed from chloroform solution. As the height of particles is 2 nm, they are prob-ably packed in three molecular layers (see Fig. 1). It was shown that the ensemble height grows to 4 nm as a result of the repeated heating of the chloroform solution at 50 °C.

To study the infl uence of solvent nature on the mor-phology of supramolecular assemblies, solutions of the Ru complex in tetrachloroethane were prepared. Tetra-choloethane has a higher boiling point (147 °C) than chlo-roform (61 °C). The AFM image of the fresh sample was analogous to the image of the sample obtained from chlo-roform. It also contained cylindrical particles from 50 to 100 nm in diameter and the mean height of the cylinders was about 2–3 nm. The image of the sample fabricated after the storage of a fresh solution for 18 h indicates that such storage gives rise to spontaneous self-assembling of the cylindrical formations to more large-sized wires of various lengths right up to 150 nm (Fig. 3). As mentioned

above, photorefractive polymer layers were cast from the tetrachloroethane solution of the components [1–3]. Better results were achieved after 4 heating cycles (3 min heating of the (R4Pc)Ru(TED)2 solution at 70 °C and cooling to room temperature). As shown in Table 1, at E0 = 110 V.μm-1 the two-beam coupling gain coeffi cient equals = 140 cm-1 and the response time is about τ = 0.1 s at 1064 nm. Optical absorption of the (R4Pc)Ru(TED)2

solution extends to 1500 nm as a result of the heating. For this reason, representations of AFM images from the heated solution are of special interest; therefore, an exp-eriment including heating of the solution to 70 °C with further cooling was carried out. Figure 4a shows the image of supramolecular wires obtained as a result of a 4-cycle heating of the solution to 70 °C for 3 min and cooling to rt. Figure 4b presents supramolecular wires (marked with a white ring on Fig. 4a). Figure 4c shows the wire (right-hand in Fig. 4b) in the 3D modes. Figures 4d and 4e show the cross-sections along the lines 1, 2 and 3, respectively. As seen, the sample consists of wires 7–8 nm in height, 100–150 nm in width (Fig. 4d), and 600 nm (Fig. 4e) or more in length. Presented in Fig. 4b (right-hand) and 4c supramolecular wire consists of 160 000 molecules packed in 10 molecular layers (calculated from X-ray diffraction data). Thus, the nature of the solvent plays a crucial role in the self-assembling processes of the inves-tigated complexes.

Z-scan measurements of the third-order non-linearity

Z-scan method allows the estimation of the suscepti-bility χ(3) and the non-linear optical absorption in both the solid media and the solutions [13, 14]. The optical transmittance through the sample is measured in con-ditions where the cell moves along axis z of the laser beam focused by a lens and intersects the Rayleigh range z0 = πw0

2/λ (here 2w0 is the beam waist, i.e. diameter of the beam at focus). At high intensity of laser irradiation

Fig. 3. (a) AFM image of (R4Pc)Ru(TED)2 ensembles cast on mica surface from the tetrachloroethane solution after their storage for 18 h, (b) 3D image of the wire indicated in Fig. 3(a) by a white oval



in the Rayleigh range, the volume polarization of the sample P(E) obtains an appreciable contribution from the non-linear components. In the solution, at random centro-symmetrical orientation of the non-linear chromophores, the second-order susceptibility vanishes to zero and P(E)= χ(1)E + χ(3)E 3 (E is the electric fi eld of an electromag-netic wave, χ(1) is the linear susceptibility). As n2 = 1 + 4πP/E, hence in Equation 2:

n = n0 + n2I 0, (2)

where n0 = (1 + 4πχ(1))0.5 and n2 ≈ 2πχ(3)/n0. According to Equation 2, the refractive index increases or decreases depending on the n2 sign. The intense laser irradiation induces a two-photon excitation of NLO chromophores, therefore in the Rayleigh range the optical absorption coeffi cient includes the linear (α 0) and non-linear (β)terms as expressed in Equation 3:

α(I) = α 0 + βI0. (3)

According to a known formula [13, 14], the real part of the complex third-order susceptibility is related to n2

as Reχ(3) (esu) = n2 × [10–7 × cn02/12π2] or as follows in

Equation 4:

Reχ(3) (esu) = n2 × (n02/0.0394), (4)

where n2 is given in cm2.W-1, n0 = 1.5 for tetrachloroethane. Non-linear optical absorption determines the imaginary part of the susceptibility as shown in Equation 5:

Imχ(3) (esu) = (βλ/4π) × (n02/0.0394) (5)

(βλ/4π is given in cm2.W-1).Experimental points of the transmittance were measured

using the pulse Nd:YAG laser (1064 nm, pulse energy 0.005 J and width 8 ns). The laser beam was focused with a 6.5 cm focal length lens. The 1 mm quartz cell fi lled with 6.4 × 10-5 mol.L-1 solution of (R4Pc)Ru(TED)2 in tetrachlo-roethane was scanned along focused laser beam. The single laser pulses were switched on after each displacement of the cell with 15 s intervals. The beam waist and the intensity in the focal point were w0 = 23 μm and I0 = 1.8 × 1010 W.cm-2, respectively. The aperture of 1.5 mm diameter was placed in front of the detector. Normalized closed- and open-aperture transmittances were measured. A fresh solution of (R4Pc)Ru(TED)2 in tetrachloroethane and a solution after its 3 min heating at 70 °C were used. Closed-aperture measur-ing demonstrated that the (R4Pc)Ru(TED)2 solution acts as the focusing lens, both before and after heating. So, n2 in both cases has a positive value, and hence the effect is not caused by radiation heating.

Fig. 4. (a) AFM image of (R4Pc)Ru(TED)2 ensembles cast on mica surface from the tetrachloroethane solution after 4 cycles heating/cooling (3 min heating of the solution to 70 °C and cooling to room temperature), (b) ensembles indicated by white ring in Fig. 4a, (c) 3D mode of the wire shown at the right in Fig. 4b, (d) dimensions of ensembles along lines 1, 2 and (e) 3



Positive value n2 provides the phase distortion ΔΦ 0 at z = 0 according to Equation 6:

ΔΦ 0 = 2πn2I0 Leff/λ, (6)

where Leff = {[1 – exp(-α 0L)]/α 0}, L = 0.1 cm, λ = 1064 nm. Optical absorption at 1064 nm of the (R4Pc)Ru(TED)2

solution (6.4 × 10-5 mol.L-1) in tetrachloroethane was A ≅0.002 in the fresh solution and 0.017 after its 3 min heat-ing at 70 °C. The corresponding absorption coeffi cients, α 0 = 2.303A/L, are equal to 0.04606 cm-1 and 0.392 cm-1.Thus, Leff ≈ 0.1 cm in the fresh solution and in the solu-tion after heating.

Figure 5 shows the experimental points measured with a closed aperture in the fresh solution (curve 1), after 3 min heating at 70 °C (curve 2) and in pure tetrachloro-ethane (curve 3). Solid curves in Fig. 5 correspond to the equation T(CA) = 1 – 4ΔΦ0x/(x2 + 9)(x2 + 1) plotted by fi tting the non-linearly-induced phase distortion ΔΦ0

to experi mental data. Here, x = z/z0 and z0 = πw02/λ =

2 mm. Solid curve 1 corresponds to the value ΔΦ0 = 1.2. It follows from Equation 2 and Equation 6, that Δn = n2I0

= 0.0002, n2 = 1.14 × 10-14 cm2.W-1 and the real part of the third-order susceptibility equals χ(3) = 6.5 × 10-13 esu(Equation 4). Let us assume that the susceptibility of the fresh solution is determined only by the molecular form of (R4Pc)Ru(TED)2, the number of molecules of which is equal to N = NAv × 6.4 × 10-5, NAv being the Avogadro number. In this case, the molecular polarizability of (R4Pc)Ru(TED)2 estimated from Equation 1 is near to γ = 4.5 ×10-32 esu. After heating the cell, the optical absorption at 627 nm decreases by half from its original value (from 0.6 to 0.3). This implies a decrease of the susceptibility from 6.5 × 10-13 to 3.25 × 10-13 esu. However, in contrast to this, the phase distortion increases to ΔΦ0 = 2.7 (Fig. 5, curve 2) after heating. Equation 2 and Equation 6 give values Δn = 0.00046 and n2 = 2.6 × 10-14 cm2.W-1; therefore the

susceptibility (Equation 4) of the solution increases to χ(3) = 1.5 × 10-12 esu. It follows that after assembling half of the (R4Pc)Ru(TED)2 molecules into supramolecular ensem-bles, the common susceptibility includes contributions of both molecular forms chromophores, χ(3) = 3.25 × 10-13

esu, and the ensembles, χ(3) = (1.5 × 10-12–3.25 × 10-13) = 1.17 × 10-12 esu. As a result, the polarizability referred to one molecule increases 1.17 × 10-12 /3.25 × 10-13 = 3.6 times when an individual molecule forms a supramolecular ensemble. This effect may be explained by the enlarged collective electron excitation system in the supramolecu-lar ensembles, compared with molecular forms.

The open-aperture normalized transmittances were measured after heating of the solution. The experimental points, shown in Fig. 5 (curve 4), are well fi tted by the following equation:

Tz = ln(1+q0/(1+x2)]/[q0/(1+x2)], (7)

where q0 = βI0Leff = 0.8. Hence β = 4.5 × 10-10 cm.W-1,βλ/4π = 3.8 × 10-15 cm2.W-1 and the imaginary suscep-tibility Im(χ(3)) = 2.2 × 10-13 esu (Equation 5). Thus, the summary susceptibility of the solution containing supramolecular ensembles equals χ(3) = {[Re(χ(3))]2 + [Im(χ(3))]2}0.5 = 1.52 × 10-12 esu. It is possible that the measured values of the third-order susceptibility are somewhat understated due to the negative contribution of the thermal non-linearity.

CONCLUSION

Atomic force microscopic measurements allow us to obtain images of supramolecular ensembles of the com-plexes of Ru(II) with tetra-15-crown-5-phthalocyanine and axially coordinated triethylenediamine molecules formed from different solvents. On the basis of compara-tive analysis of obtained experimental data and X-ray dif-fraction results on the Ru(II) complexes, the number of molecules and their relative orientation in supramolecu-lar ensembles were estimated depending on the solvent nature and thermal treatment of solutions. It was found that the samples prepared by heating the Ru(II) complex solutions in tetrachloroethane demonstrate stable supra-molecular wires of 7–8 nm in height, 100–150 nm in width and 600 nm or more in length.

The magnitudes and positive sign of the third-order susceptibility were measured in the solutions of molecu-lar complexes (R4Pc)Ru(TED)2 and their supramolecu-lar ensembles. The molecular polarizability of (R4Pc)Ru(TED)2 is about = 4.5 × 10-32 esu. The polarizability attributed to one molecule increases by a factor of 3.6 when the individual molecule forms a supramolecular ensemble. Supramolecular ensembles architecture de-fi nes the photoelectric and photorefractive sensitivities of polymer layers to IR radiation (1064 nm). These layers have the high two-beam gain coeffi cient Γ = 140 cm-1

and the response time 0.1 s at 110 V.μm-1 [1, 2].

Fig. 5. The normalized closed-aperture transmittances of the cell containing the fresh solution of (R4Pc)Ru(TED)2 in (1) tetrachloroethane, (2) the same solution after 5 min heating at 70 °C, (3) the tetrachloroethane solvent, (4) open-aperture data for the solution after heating. The solid curves are theo-retical fi ts (see text). (R4Pc)Ru(TED)2 concentration is 6.4 × 10–5 M



Acknowledgements

This work was supported by the International Science and Technology Center (Grant No. 3718) and the Russian Foundation for Basic Research (Grants No. 07-03-13547, 08-03-00125 and 08-03-00835).

REFERENCES

Vannikov AV, Grishina AD, Gorbunova YG, Enaki-1.eva YY, Pereshivko LY, Krivenko TV, Savelyev VV and Tsivadze AY. Dokl. Phys. Chem. 2005; 403:137–141.Vannikov AV, Grishina AD, Gorbunova YG, Enaki-2.eva YY, Krivenko TV, Savelyev VV and Tsivadze AY. Russ. J. Phys. Chem. 2006; 80: 453–460.Grishina AD, Konnov FY, Gorbunova YG, Enaki-3.eva YY, Pereshivko LY, Krivenko TV, Savelyev VV, Vannikov AV and Tsivadze AY. Russ. J. Phys. Chem. 2007; 81: 982–989. Knoester J. 4. Chem. Phys. Lett. 1993; 203: 371–377.Kusyk MG. 5. Phys. Rev. Lett. 2000; 85: 1218–1221.De la Torre G, Vazquez P, Agullo-Lopez F and 6. Torres T. Chem. Rev. 2004; 104: 3723–3750.

Claessens CG, De la Torre G and Torres T. In 7.Non-Linear Optical Properties of Matter. From Molecules to Condensed Phases. Papadopoulos MG, Sadlej AJ and Leszczynski J. (Eds.). Springer: New York, 2006; pp. 509–535.Coe BJ. In 8. Non-Linear Optical Properties of Matter. From Molecules to Condensed Phases.Papadopoulos MG, Sadlej AJ and Leszczynski J. (Eds.). Springer: New York, 2006; pp. 571–608.Manas ES, Spano FC and Chen LX. 9. J. Chem. Phys.1997; 107: 707–719. Enakieva YY, Gorbunova YG, Nefedov SE and 10.Tsivadze AY. Mendeleev Commun. 2004; 14: 193–194.Arslanov VV, Gorbunova YG, Selektor SL, Shei-11.nina LS, Tselykh OG, Enakieva YY and Tsivadze AY. Russ. Chem. Bull. 2004; 53: 2532–2541.Horcas I, Fernandez R, Gomes-Rodriguez JM, 12.Colchero J, Gomes-Herrero J and Baro AM. Rev. Sci. Instrum. 2007; 78: 013705(1–8). Sheik-Bahae M, Said AA, Wei T-H, Hagan DJ and 13.Van Stryland EW. IEEE J. Quant. Electron. 1990; 26: 760–769. Sutherland RL. 14. Handbook of Nonlinear OpticsMarcel Dekker: New York, 1996.




INTRODUCTION

2,7,12,17-tetraphenylporphycene (TPPo, Chart 1) is a second-generation, photodynamic therapy (PDT) photo-sensitizer that meets all the photophysical requirements: intense absorption above 650 nm, high fluorescence for diagnostic purposes ( F = 0.15), and high triplet ( T = 0.33) and singlet oxygen ( = 0.23) quantum yields 1–3 . Although its photodynamic activity has been

confirmed in several studies of cell photoinactivation 4, 5], its mechanism of action remains unknown. While

singlet oxygen, 1O2, is regarded as the most important cytotoxic species (type-II mechanism), photo-oxidative damage caused by the direct interaction between the photoexcited sensitizer and nearby biomolecules (type-I mechanism) may also play a role. High sensitizer photo-reactivity, binding to susceptive biomolecules, or low ox-ygen concentrations are factors that may favor the type-I mechanism over the type-II [6–8]. In previous work, we reported the electron transfer photoreactivity of TPPo and characterized its radical ions [9]. Now, we report on the ground- and excited-state interaction of TPPo with essential biomolecules that may favor the type-I mecha-nism in the overall photodynamic process. Specifically, we have examined the ground- and excited-state interac-tions of TPPo with model biomolecules such as guanos-ine (Guo), 7,8-dihydro-8-oxoguanosine (8-Guo), human

Ground- and excited-state interactions of 2,7,12,17-tetraphenylporphycene with model target biomolecules for type-I photodynamic therapy

Noemí Rubio†a, Víctor Martínez-Junza§a, Jordi Estrugaa, José I. Borrella,Margarita Morab, M. Lluïsa Sagristáb and Santi Nonell*a

a Grup d’Enginyeria Molecular, Institut Químic de Sarrià, Universitat Ramon Llull, Via Augusta 390, 08017 Barcelona, Spainb Departament de Bioquímica i Biología Molecular, Facultat de Biologia, Universitat de Barcelona, Avinguda Diagonal 645, 08028 Barcelona, Spain


ABSTRACT: Biosubstrate-sensitizer binding is one of the factors that enhances the type-I mechanism over the type-II in the whole photodynamic process. 2,7,12,17-Tetraphenylporphycene (TPPo), a second-generation photosensitizer, is a hydrophobic compound with good photophysical properties for photodynamic therapy applications that has proved its ability for the photoinactivation of different cell lines. Nevertheless, little is known about its mechanism of action. This paper focuses on the study of the interaction/binding of TPPo with different model biomolecules that may favor the type-I mechanism in the overall photodynamic process, including nucleosides, proteins, and phospholipids. Compared with more hydrophilic photosensitizers, it is concluded that TPPo is more likely to undergo type-II (singlet oxygen) than type-I (electron transfer) photodynamic processes in biological environments.

KEYWORDS: porphycenes, photodynamic therapy, type-I mechanism, human serum albumin, gramicidin, nucleosides, guanosine, phospholipids, lipid A, singlet oxygen.

*Correspondence to: Santi Nonell, email: [email protected],fax: +34 93-205-62-66, tel: +34 93-267-20-00

†Current address: Wellman Center for Photomedicine, Harvard Medical School, Massachusetts General Hospital, Thier-224, Boston, MA 02114, USA

§Current address: Center for Nanotechnology, Westfälische Wilhelms-Universität Münster, Heisenbergstrasse 11, D-48149 Münster, Germany


100 N. RUBIO ET AL.

serum albumin (HSA), gramicidin A (gA), phosphati-dylethanolamine (PE) and lipid A.

Guanine (G) is the easiest oxidizable base among those making up DNA and RNA, and G stacks are more readilyoxidizable sites than G itself [10–13]. 7,8-dihydro-8-oxoguanosine (8-Guo) is an important product of oxida-tive attack on nucleic acids [14–16], generated by both type-I and type-II mechanisms [17–21]. 8-Guo is more easily oxidized than the parent Guo by ca 0.43 V [15], and has therefore been used as a model for 5 G in the 5 -GG-3 sequence in a -DNA conformation [22]. The same approach is adopted in this work.

On the other hand, serum proteins such as albumin, high-density lipoproteins (HDL) and low-density lipo-proteins (LDL) [23] are natural carriers for the sensitizers.They participate in transport to the tumor sites and in the uptake into the cells [24–26]. Serum albumin is the major protein of the extracellular fluids with a concentra-tion in blood plasma of ca 0.6 mM; therefore, we found the study of the binding of TPPo to HSA to be of inter-est. A simpler protein is gramicidin A (gA), a pentade-gramicidin A (gA), a pentade-capeptide that contains four Trp residues in an alternating sequence of d- and l-amino acids [27]. This sequence allows the molecule to fold as a helix, forming membrane channels that are specific for transport of monovalent cations [28, 29]. In organic solvents, gA forms a vari-ety of double-stranded (DS) intertwined dimers [30–33]. In lipid bilayers, the DS dimers unfold and refold into the Trp-anchored single-stranded (SS) structure, where the Trp residues are placed near the membrane/solution interfaces [33]. Gramicidin A was chosen as the third target for our studies.

Finally, phosphatidylethanolamine (PE) is a phosphol-ipid found both in the inner and outer membrane of bac-teria. Likewise, lipid A is a part of the lypopolysaccharide complex found in the outer membrane of gram(-) bacte-ria, which contains a larger number of fatty-acid chains per molecule as compared to PE. In order to evaluate the potential of TPPo for photoinactivation of bacteria, the interaction between PE and lipid A with TPPo was studied similarly.

MATERIALS AND METHODS

Chemicals

TPPo was synthesized as described in the literature [1, 34, 35]. Palladium(II)-5,10,15,20-tetraphenylpor-phine (PdTPP) was purchased from Aldrich. L- -dimiristoyl-phospathidylcholine (DMPC) was purchased from Avanti Polar Lipids (Birmingham, AL, USA). Polycarbonate membranes and imidazole were from Poretics Products (Livermore, CA, USA) and Sigma-Aldrich Chemical Co. (St. Louis, MO, USA), respec-tively. Dimyristoyl- and dioleoyl-phosphatidyl etha-nolamine (PE), PE from E. coli, human serum albumin (HSA) and gramicidin A (gA) were from Sigma. 2 ,3 ,5 -tris((tert-butyldimethylsilyl)oxy)guanosine (Guo-S) and 2 ,3 ,5 -tris((tert-butyldimethylsilyl)oxy)-7,8-dihydro-8-oxoguanosine (8-Guo-S) were synthesized as described by Sheu and Foote [15]. Benzonitrile (PhCN, 99.9%, HPLC grade) and triethylamine (99%) from Aldrich, ben-zene, toluene and methanol (spectroscopic grade) from SDS, and DMSO from Panreac were used as received. Argon 5.0 was from Abelló-Linde.

Preparation of liposomes

Large unilamellar liposomes (LUVs) for TPPo and gramicidin A (gA) incorporation were prepared by micro-emulsification following standard procedures as described in the literature for some photosensitizers [36, 37]. Briefly, lipid films containing DMPC and TPPo or DMPC and a mixture of TPPo and gA were prepared. The films were hydrated with PBS buffer and the suspension was frozen and thawed 5 times, bath sonicated for 60 min at 45 °C and, finally, micro-emulsified (EmulsiFlex B3 device, Avestin, Ottawa, Canada). Micro-emulsification was car-ried out by pumping the fluid 15 times through the inter-action chamber (45 ºC, 200 kPa) for 13 mM DMPC. TPPo and gA were incorporated at 6.25 and 10 mol.%. Non-in-corporated TPPo and gA were removed by centrifugation at 4000 rpm for 20 min.

Photophysical measurements

Absorption spectra were recorded on a Varian Cary 4 spectrophotometer. Fluorescence measurements were performed using a SPEX Fluoromax-2 spectroflu-orometer. Triplet state lifetimes were determined by nanosecond laser flash photolysis using an OPO laser (SL OPO, Continuum; 5 ns pulse-width, 1–10 mJ per pulse) pumped by a Q-switched Nd-YAG laser (Sure-lite I-10, Continuum) in argon-saturated solutions. The second harmonic of the Nd-YAG laser was also used for excitation. Photoinduced absorbance changes were monitored at 90º by an analyzing beam produced by a Xe lamp (PTI, 75W) in combination with a dual-grating monochromator (mod.101, PTI) coupled to

Chart 1. Structure of 2,7,12,17-tetraphenylporphycene


GROUND- AND EXCITED-STATE INTERACTIONS OF 2,7,12,17-TETRAPHENYLPORPHYCENE 101

a photomultiplier (Hamamatsu R928, 300–800 nm). Production and quenching of singlet oxygen were determined by time-resolved, near-IR emission spec-troscopy (TNIR) using a germanium detector (North Coast EO-717P). Kinetic analysis of the individual transients was achieved with the FitLW software developed in our laboratory. For photobleaching stud-ies, the samples were irradiated in 1 cm quartz cells using a Xe lamp appropriately filtered to remove the UV and IR components.


Interaction with nucleosides

Neither ground- nor singlet-excited state interaction of TPPo with the nucleosides could be detected by eitherabsorption or fluorescence techniques. The quenching of 3TPPo by Guo-S and 8-Guo-S was then studied by laser flash photolysis irradiating at exc = 532 nm, where only the porphycene absorbs. The quenching rate con-stants were derived from the TPPo triplet lifetime, T,measured in the absence and in the presence of Guo-S and 8-Guo-S at the highest concentration allowed by solubility in PhCN, 27 mM and 2.9 mM, respectively. The values are collected in Table 1 along with those for palladium(II)-5,10,15,20-tetraphenylporphine (PdTPP), a related tetrapyrrole that shows almost the same redox potential as 3TPPo, and those reported for two different fullerenes [38].

The rate constants correlate well with the free energy of the electron-transfer process, calculated using Equation 1:

G F E E

EN e

dA

r

0 0 0

00

2

4

( / ) ( )

,

D D A/A -

0

(1)

where E0 (D /D) and E0 (A/A -) are the standard redox potentials of the electron donor and acceptor, respectively.

E00 is the excitation energy, and N e

dA

r

2

04 is the coulom

bic term (NA, Avogadro’s constant; e, electronic charge; 0,vacuum permittivity; r, relative permittivity of the solvent; d, encounter distance of the radical ions). The coulombic term is assumed to be +12.2 kJ.mol-1 in PhCN [38]. Owing to the unfavorable thermodynamics, elec-tron transfer is several orders of magnitude slower than the production of singlet oxygen, with a rate constant close to diffusional, 2.9 109 M-1.s-1 [2]. The significance of these results for PDT applications is that TPPo is not likely to cause oxidative damage to DNA through a type-I mechanism.

Interaction with proteins

Human serum albumin. The association between TPPo and HSA was revealed by changes in the absorption spectra of TPPo (2.1 M) after addition of different concentrations of HSA in PBS containing 1.3% volume of DMSO. The absorption spectrum of TPPo in this solvent shows broad bands and low Soret/Q-bands, consistent with TPPo being extensively aggregated. The addition of HSA shifts the absorption spectrum of TPPo hypsochromically and increases the Soret/Q-band ratio (Fig. 1), consistent with a disaggregation of the macrocycle upon protein binding.

The association constant of the dye-protein complex, Ka, was calculated from the changes in the absorption spec-tra using Equation 2, which assumes that TPPo is bound at equivalent sites on separate protein molecules. [HSA]/[TPPo] ratios in the range 15–90 were used for the study.

TPPo HSA TPPo HSAKa

(2)Ka

TPPo HSA

TPPo HSA.

Table 1. Observed rate constants and calculated free energy for electron transfer between Guo-S and 8-Guo-S and several triplet photosensitizers. All standard potentials are vs the saturated calomel electrode in benzonitrile

Guo-S 8-Guo-S

(E0 (D /D) = 1.40 V)a (E0 (D /D) = 0.97 V)a

E0 (3A/A -), V kq , M-1.s-1 G0ET , kJ.mol-1 kq , M-1.s-1 G0

ET , kJ.mol-1

TPPo 0.63b 104 62.2 3.7 × 104 20.7

PdTPP 0.68c 3.0 × 104 57.3 2.7 × 106 15.8

DHF 0.91d < 5.0 × 104 35.1 1.1 × 107 -6.4

C60 1.14a 3.3 × 106 15.9 1.1 × 108 -28.8

aRef. [38]; bRef. [2]; cRef. [39]; dRef. [40]


102 N. RUBIO ET AL.

A modified form of the Benesi-Hildebrand expression [41] (Equation 3) was used to determine the value of Ka:

TPPo HSA HSA

app app

0

0 0

1

A Ka ( ) ( ), (3)

where A is the change in absorbance measured upon protein addition, and 0 and app are the molar absorp-tion coefficients of the dye and the dye-protein complex, respectively. The Benesi-Hildebrand plot at = 392 nm is shown in Fig. 1. The association constant is Ka = (1.5 0.7) 104 M-1, ca 1 order of magnitude lower than those reported for the association of anionic sensitizers such as AlPcS (Ka = (1.3 0.2) 105 M-1) [41] to HSA, or ZnPcSmix (Ka = 1.0 105 M-1) to BSA [42] and the neutral sensitizer BPDMA (Ka = 5.2 105 M-1) to HSA [43]. Notwithstanding, it is comparable to the association constants of neutral sensitizers such as SnET2 to bovine serum albumin (Ka = 2.8 104 M-1) [44].

Fluorescence measurements on the effect of [HSA] on TPPo were also carried out. Solutions of TPPo (1.05 M)in PBS containing 1.3% DMSO did not show any fluores-cence emission excited at either 390 or 402 nm, whereas the sharp emission at 677 nm corresponding to TPPo appeared in neat DMSO. This confirms the extensive aggregation of TPPo in PBS solutions. When HSA was added to the PBS containing-TPPo solution up to a con-centration of 0.45 mM ([HSA]/[TPPo] ratio up to 428), no fluorescence could still be recorded. At this [HSA]/[TPPo] ratio, 87% of TPPo is bound to HSA in the form of monomers, which are distinctly fluorescent. It must be concluded that the fluorescence is quenched by the pro-tein. The principal binding regions on HSA are located

in subdomains IIA and IIIA, where residues Trp214, Lys 199 and Tyr 411, located strategically in the hydrop-hobic pockets, have been involved in the binding process [45]. Free-energy values for electron transfer between 1TPPo and Tyr or Trp, calculated according to Equation 1, are -63.3 and -53.7 kJ.mol-1, respectively, assuming that they are in close vicinity [2, 46]. From the perspec-tive of PDT applications, electron transfer from Tyr or Trp to 1TPPo is indeed feasible and likely to occur when-ever TPPo is located close to such amino acids.

Gramicidin A. The interaction between TPPo and gramicidin A (gA) was studied both in solution (benzo-nitrile or DMSO) and in dymiristoylphosphatidylcho-line (DMPC) liposomes as model membranes. Neither ground-state absorption nor fluorescence spectra showed any evidence for the interaction between TPPo and gA, suggesting that TPPo is not localized close to its Trp resi-dues, the only amino acids that could undergo electron transfer. However, laser flash photolysis experiments in deaerated benzonitrile showed that the rate constant for 3TPPo decay decreased to a plateau value as gramicidin concentration increases (Fig. 2). The decays followed single exponential kinetics at all gA concentrations, sug-gesting that the protein shields TPPo from the external solvent quenchers as it is often observed [47–49].

Monoexponential decay kinetics upon binding to HSA have been previously observed for other photosensitiz-ers [41, 50–52]. A dynamic equilibrium model was pro-posed by Foley et al. to account for this observation [52] (Scheme 1).

The model predicts monoexponential decay kinetics, if there is a dynamic exchange between free and protein-bound dye molecules during the lifetime of the excited

Fig. 1. Benesi-Hildebrand plot corresponding to the change of TPPo absorbance at 392 nm vs. HSA concentration. Inset:Effect of HSA on the absorption spectrum of TPPo (spectra have been shifted vertically for greater clarity)

Fig. 2. 3TPPo experimental deactivation rates as a function of protein concentration. The solid line corresponds to the fit of Equation 4 to the experimental data. Inset: Logarithmic plot of the transient decays for 3TPPo in the absence and presence (1.3 mM) of gramicidin



triplet state. The relevant rate constants can be obtained by fitting Equation 4 to the data [52] as follows:

k k kk k

k kca

obs

protein

protein

12

1 2

1

( )** *

* (4)

where kobs is the observed decay rate constant; ka and kc

are the deactivation rate constants for the triplet state of the free and bound TPPo, respectively; and k1

* and k2*

refer to the rate constants for association and dissocia-tion in the excited state, their ratio giving the equilibrium constant K*

eq (Equation 5):

Kk

keq*

*

*.1

2

(5)

Fitting Equation 4 to the data in Fig. 2, yields k1* = (1.0

0.6) 109 M-1.s-1, k2* = (2.8 0.1) 104 s-1 and kc = (1.6

0.8) 104 s-1, from which the equilibrium constant for the binding of 3TPPo to gA is determined as K*

eq = (3.7 1.5) 104 M-1. This value is within the same order as those reported for the interaction of sulphonated phthalo-cyanines 3AlPcSn with HSA: K*

eq = (7.1 1.5) 104 M-1

for n = 2, (6.2 2.0) 104 M-1 for n = 3, and (7.0 3.0) 104 M-1 for n = 4 [41].

Measurements of the system TPPo-gA were also performed in dymiristoylphosphatidylcholine (DMPC) liposomes. In lipid bilayers, gramicidin A shows a sin-gle-stranded structure where the Trp residues are placed near the membrane/solution interface [33]. Two kinds of liposomes were used in our studies: in the first, TPPo and gA were encapsulated together, while in the second TPPo was first encapsulated and gA was added later by incuba-tion. This procedure ensures that the protein adopts the same conformation as in cell membranes. Irrespective of the preparation procedure, no changes in the 3TPPo life-time could be observed, indicating that TPPo does not interact with gramicidin in the liposomes, in contrast to the behavior observed for several more hydrophilic sul-phonated aluminum phthalocyanines [53]. This is con-sistent with a localization of the hydrophobic TPPo deep in the bilayer rather than association with gA, while the sulphonated phthalocyanines are likely at the interface, close to the Trp residues.

Additional information on the interaction between 3TPPo and gA in benzonitrile was obtained examining the formation and decay of singlet oxygen (1O2) through its characteristic 1270 nm emission. The amplitude of the decay part of the signal, S(0), normalized by the energy of the laser pulses, was measured as a function of the protein concentration. This amplitude is directly propor-tional to the 1O2 production quantum yield. Consistent with the increase in the 3TPPo lifetime, it was observed that S(0) decreased in the presence of gA (Fig. 3A). Thus, gA also shields TPPo from oxygen.

Likewise, the rate constant of 1O2 decay, k , was found to increase linearly, which indicates that 1O2 is quenched by gA, in accordance with the presence of Trp residues in this protein. Analysis of the concentration dependence according to Equation 6 yielded the quenching rate con-stant, kgA = (2.9 0.1) 107 M-1·s-1 (Fig. 3B)

k k kgA0 [ ].Gramicidin A (6)

This value is comparable to that reported for Trp, e.g.3.3 107 M-1.s-1 in D2O:ethanol (1:1) [54]. Proteins such as HSA react with 1O2 with a rate constant kHSA = 5 108

M-1.s-1 [41]. In conclusion, our results indicate that gA is more likely to be photoinactivated by TPPo through singlet oxygen than by a type-I mechanism. The same

Fig. 3. Gramicidin A effects on the formation and decay of 1O2. (A) Laser-energy normalized intensity of the 1270 nm 1O2

emission (B) Rate constant for 1O2 decay. Inset:Time profile of the 1270 nm-emission decays at two different protein concentrations

Scheme 1. Model for the binding of the excited state of TPPo to gramicidin A. TPPo* and (TPPo-gramicidin)* represent free and bound excited state molecules, respectively. ka and kc are the rates for triplet state deactivation for the free and bound forms, respectively, while k1

* and k2* correspond to the association and

dissociation rates


104 N. RUBIO ET AL.

conclusion was reached by Rokitskaya et al. using the more hydrophilic trisulphonated and tetrasulpho-nated aluminum phthalocyanines (AlPcS3 and AlPcS4)[55, 56].

Interaction with phospholipids

The interaction of TPPo with different phosphatidyle-thanolamines (PEs) containing different fatty acids, as well as with lipid A was examined, as in the previous sections. Once again, we were unable to observe any changes in the absorption and fluorescence behavior of TPPo, but the triplet state was slightly affected by these biomolecules. Specifically, dimyristoyl-PE was found to decrease the 3TPPo lifetime, while lipid A increased it. Table 2 shows the values of the rate constant for quench-ing of 3TPPo by different kinds of PEs. All fall in the range 104–105 M-1.s-1, below the value that was reported for AlPcS2 for its triplet-state reaction with PE from E. coli, 2.0 106 M-1.s-1 [57]. The interaction is favored in the less polar solvent toluene and is also more favorable for the shorter-chain dimyristoyl-PE, which is consis-tent with a shorter distance between TPPo and the active amino group, the likely deactivating agent.

Indeed, irradiation of TPPo in the presence of trieth-ylamine caused photobleaching of the sensitizer (Fig. 4). Fluorescence and laser flash photolysis studies demon-strated the participation of both the singlet and the triplet state of TPPo in this process and the rate of photobleach-ing was decreased in the presence of oxygen.

In the case of lipid A, the only noticeable change was a 15% increase of the 3TPPo lifetime at 1.4 mM lipid A, which, as in the case of gramicidin A, indicates that TPPo is shielded from the surrounding medium by this biomole-cule and is not quenched by it. Once again, this finding contrasts with the observation of effective quenching of a sulphonated phthalocyanine by lipid A [57]. Thus, the higher lipophilic character of TPPo confers this sensitizer with lower direct photochemical reactivity towards phos-pholipids, which in the end suggests that its photody-namic activity is most likely through a type-II, i.e. singlet oxygen, mechanism.

CONCLUSION

The interactions between the hydrophobic photosensi-tizer TPPo and a number of representative biomolecules have been studied through a variety of spectroscopic tech-niques. The triplet state of TPPo is quenched by highly oxi-dizable molecules such as Guo-S, 8-Guo-S, dimyristoyl-PE and triethylamine, the latter also being able to quench the singlet state. TPPo binds to several biomolecules such as HSA, gA, and lipid A, although quenching is observed only in the case of HSA. We have found that binding to gA and lipid A results in a longer triplet lifetime instead. Taken together and compared with the results described for more hydrophilic photosensitizers, the present results indicate that the hydrophobic TPPo is more likely to under-go type-II (singlet oxygen) than type-I (electron transfer) photodynamic processes in biological environments.

Acknowledgments

This work was supported by grants from the Spanish Ministerio de Ciencia e Innovación (SAF2002-04034-C02-02 and CTQ2007-67763-C03/BQU). N. Rubio thanks Fundació Institut Químic de Sarrià for a predoc-toral fellowship.

Table 2. Rate constants for quenching of 3TPPo by PE

Compound kq/M-1.s-1 Conditions

Dimyristoyl-PE5.8 × 105

1.1 × 105Toluene (rt)b

Tol:MetOH (1:1) (47 ºC)

Dioleoyl-PE 1.0 × 105

3.9 × 104 1.0 105

Tol:MetOH (1:1) (rt)b

PE from E. colia 1.1 × 105 Toluene (rt)b

aPE made up of a mixture of different fatty acids; brt: room temperature

Fig. 4. Photobleaching of TPPo in the presence of 1 M trieth-ylamine in argon- and air-saturated benzene solutions. The irra-diation source was a Xe lamp with the UV and IR components removed. The spectra were recorded at 0, 2, 3, 5, 11, and 19 min



REFERENCES

Nonell S, Bou N, Borrell JI, Teixidó J, Villanueva A, 1.Juarranz A and Cañete M. Tetrahedron Lett. 1995;36: 3405–3408.Rubio N, Prat F, Bou N, Borrell JI, Teixidó J, 2.Villanueva A, Juarranz A, Cañete M, Stockert JC and Nonell S. New J. Chem. 2005; 2: 378–384.Stockert JC, Cañete M, Juarranz A, Villanueva A, 3.Horobin RW, Borrell JI, Teixidó J and Nonell S. Curr. Med. Chem. 2007; 14: 9–997.Villanueva A, Cañete M, Nonell S, Borrell JI, 4.Teixidó J and Juarranz A. Anti Canc. Drug Des.1996; 11: 89–99.Cañete M, Lapeña M, Juarranz A, Vendrell V, Borrell5.JI, Teixidó J, Nonell S and Villanueva A. Anti Canc. Drug Des. 1997; 12: 543–554.Foote CS. 6. Type I and Type II Mechanism of Photo-dynamic Action in Light-activated Pesticides. Amer-ican Chemical Society: California, 1987; 22–38.Spikes JD. 7. Photosensitization in the Science of Photobiology (2nd edn.). Plenum Press: New York, 1989; 79–110.Ochsner M. 8. J. Photochem. Photobiol. B: Biol. 1997;39: 1–18.Rubio N, Borrell JI, Teixidó J, Cañete M, Juarranz 9.A, Villanueva A, Stockert JC and Nonell S. Photo-chem. Photobiol. Sci. 2006; 5: 376–380.Saito I, Takayama M, Sugiyama H, Nakatani K, 10.Tsuchida A and Yamamoto M. J. Am. Chem. Soc.1995; 117: 6406–6407.Sugiyama H and Saito I. 11. J. Am. Chem. Soc. 1996;118: 7063–7068.Stemp EAD, Arkin MR and Barton JK. 12. J. Am. Chem. Soc. 1997; 119: 2921–2925.Ito K and Kawanishi S. 13. Biochemistry 1997; 36:1774–1781.Kasai H and Yamaizumi Z. 14. J. Am. Chem. Soc. 1992;114: 9692–9694.Sheu C and Foote CS. 15. J. Am. Chem. Soc. 1995; 117:6439–6442.Cullis PM, Malone ME and Merson-Davies LA. 16.J. Am. Chem. Soc. 1996; 118: 2775–2781.Ravanat JL, Berger M, Benard F, Langlois R and 17.Ouellet R. Photochem. Photobiol. 1992; 55: 809–814.Kvam E, Berg K and Steen HB. 18. Biochim. Biophys. Acta 1994; 1217: 9–15.Cadet J, Berger M, Buchko GW, Joshi PC, Raoul 19.S and Ravanat JL. J. Am. Chem. Soc. 1994; 116:7403–7404.Adam W, Saha-Möller CR, Schönberger A, Berger 20.M and Cadet J. Photochem. Photobiol. 1995; 62:231–238.Adam W, Andler S, Saha-Möller CR and Schön-21.berger A. J. Photochem. Photobiol. B: Biol. 1996;34: 51–58.

Prat F, Houk KN and Foote CS. 22. J. Am. Chem. Soc.1998; 120: 845–846.Kessel D. 23. Canc. Lett. 1986; 33: 183–188.Kessel D. 24. J. Photochem. Photobiol. B:Biol. 1992;14: 261–262.Hamblin MR and Newman EL. 25. J. Photochem. Photobiol. B: Biol. 1994; 26: 45–56.Hamblin MR and Newman EL. 26. J. Photochem. Photobiol. B: Biol. 1994; 26: 147–157.Dubos RJ. 27. J. Exp. Med. 1939; 70: 1–10.Urry DW, Goodall MC, Glickson JD and Meyers 28.DF. Proc. Natl. Acad. Sci. USA 1971; 68 : 1907–1911.Pressman BC. 29. Proc. Natl. Acad. Sci. USA 1965; 53:1076–1083.Veatch WR and Blout ER. 30. Biochemistry 1974; 13:5257–5264.Burkhart BM, Li N, Langs PA, Pangborn WA and 31.Duax WL. Proc. Natl. Acad. Sci. USA 1998; 95:12950–12955.Duaux WL, Pletnev V and Burkhart BM. 32. J. Mol. Struct. 2003; 647: 97–111.Andersen OS, Koeppe II RE and Roux B. 33. IEEETrans. Nanotechnol. 2005; 4: 10–20.Cañete M, Ortiz A, Juarranz A, Villanueva A, No-34.nell S, Borrell JI, Teixidó J and Stockert JC. AntiCanc. Drug Des. 2000; 15: 143–150.Gavaldà A, Borrell JI, Teixidó J, Nonell S, Arad O, 35.Grau R, Cañete M, Juarranz A, Villanueva A and Stockert JC. J. Porphyrins Phthalocyanines 2001;5: 846–852.Postigo F, Mora M, De Madariaga MA, Nonell S 36.and Sagristá ML. Int. J. Pharm. 2004; 278: 239–254.Postigo F, Sagristá ML, De Madariaga MA, Nonell 37.S and Mora M. Biochim. Biophys. Acta 2006; 1758:583–596.Bernstein R, Prat F and Foote CS. 38. J. Am. Chem. Soc. 1999; 121: 464–465.Takeda J and Sato M. 39. Chem. Lett. 1995; 24: 939–940.Guarr TF, Meier MS, Vance VK and Clayton M. 40.J. Am. Chem. Soc. 1993; 115: 9862–9863.Phillips D. 41. Prog. React. Kinet. 1997; 22: 175–300.Ogunsipe A and Nyokong T. 42. Photochem. Photobiol. Sci. 2005; 4: 510–516.Aveline BM, Hasan T and Redmond RW. 43. J. Photo-chem. Photobiol. B: Biol. 1995; 30: 161–169.Pogue BW, Redmond RW, Trivedi N and Hasan T. 44.Photochem. Photobiol. 1998; 68: 809–815.He X-M and Carter DC. 45. Nature 1992; 358: 209–215.Milligan JR, Aguilera JA, Ly A, Tran NQ, Hoang O 46.and Ward JF. Nucl. Acids Res. 2003; 31: 6258–6263.Geacintov NE and Brenner C. 47. Photochem. Photo-biol. 1989; 50: 841–858.


106 N. RUBIO ET AL.

Roberts JE, Atherton SJ, Gaillard ER and Dillon J. 48.Photochem. Photobiol. 1995; 62: 339–341.Lang K, Mosinger J and Wagnerová DM. 49. Coord. Chem. Rev. 2004; 248: 321–350.Reddi E, Lambert CR, Jori G and Rodgers MAJ. 50.Photochem. Photobiol. 1987; 45: 345–351. Davila J and Harriman A. 51. J. Am. Chem. Soc. 1990;112: 2686–2690.Foley MSC, Beeby A, Parker AW, Bishop SM and 52.Phillips D. J. Photochem. Photobiol. B: Biol. 1997;38: 10–17.

Rokitskaya TI, Block M, Antonenko YN, Kotova 53.EA and Pohl P. Biophys. J. 2000; 78: 2572–2580.Michaeli A and Feitelson J. 54. Photochem. Photobiol.1994; 59: 284–289.Rokitskaya TI, Antonenko YN and Kotova EA. 55.FEBS Lett. 1993; 329: 332–335.Rokitskaya TI, Antonenko YN and Kotova EA. 56.Biochim. Biophys. Acta 1996; 1275: 221–226.Lacey JA and Phillips D. 57. Phys. Chem. Chem. Phys.2002; 4: 232–238.




INTRODUCTION

The organic syntheses of corroles have seen a tre-mendous increase in research activity during the past ten years [1]. Many different synthetic approaches to this macrocycle, which can be regarded as a one-carbon short relative of the porphyrins [2], have been investigated and published [3], and some of them could be developed into powerful synthetic methods [4]. With these methods in hand, the focus in corrole preparations has now shifted towards functionalization chemistry [5] and approaches for preparation of superstructured systems [6]. While these are well-established for porphyrins [7] and have been used in many different areas in the past, superstruc-tured corroles are still somewhat under-represented. A good example is the picket fence approach. This utilizes one or more (usually four) reactive o-aminophenyl-, m-hydroxyphenyl- or other reactive meso-situated sub-stituents in order to functionalize and distinguish the

distal and proximal sides of a porphyrin plane, usually for confi nement design and towards selective binding issues. A signifi cant number of different picket fence porphyrins and related o-amidophenyl-substituted compounds are known in the literature [8] — the fi eld is still growing [9] — while only four approaches exist for the preparation of respective corroles [6a,b,e,f], One can easily imagine that this narrow synthetic entry also renders the prepa-ration of chiral corrole ligands quite diffi cult. Indeed, only a few conceptually different approaches have been reported in the literature [6a,f, 10], and only two of these employ picket fence methodologies.

Very recently, we reported the introduction of steri-cally hindered achiral and chiral amido pickets into the o-phenyl positions of corroles [6f], using a catalytic sys-tem developed by Buchwald and Yin [11] and transferred to the high-yield multiple modifi cation of porphyrins by Zhang et al. [12] before. We have now illuminated the use of this elegant synthetic transformation for corrole modifi cation in some more detail, in particular, concern-ing the preparation of the target molecules 1–3 (Chart 1) on different routes, and our results are discussed here.

Investigating different strategies towards the preparation of chiral and achiral bis-picket fence corroles

Martin Bröring* , Markus Funk and Carsten Milsmann

Fachbereich Chemie, Philipps-Universität, Hans-Meerweinstraße, 35043 Marburg, Germany

Received 28 July 2008Accepted 15 October 2008

ABSTRACT: Three different strategies have been designed and tested for the preparation of sterically hindered chiral and achiral bis-picket fence corroles. All attempts to prepare an A3 corrole, with six pivaloylamido pickets in the ortho-positions of the meso-aryl substituents, failed. Mass spectrometrical investigations of the reaction mixtures prove the very slow formation of tetrapyrrolic condensation products from strongly hindered benzaldehydes. For sterical reasons, however, these potential precursors seem to be forced into a stretched conformation incapable of oxidative macrocyclization. For A2B-typebis-picket fence corroles, the formation of trace amounts of four-fold pivaloylamido substituted species with a sterically innocent aryl group at the 10-position can be shown. A reverse order of steps, with the palladium-catalyzed amidations of o-bromophenyl groups at the 5- and 15-positions of the completed corrole macrocycle as the last step, has fi nally been found as the key to a successful and versatile synthesis of such chiral and achiral sterically encumbered corrole ligands.

KEYWORDS: corrole, amidation, chiral ligands, picket fence corrole, palladium.

SPP member in good standing

*Correspondence to: Martin Bröring, email: [email protected], fax: +49 6421-2825653


108 M. BRÖRING ET AL.

EXPERIMENTAL

Materials and methods

2,6-dibromobenzaldehyde [13], dipyrrylmethane 7and corroles 2, 3 and 8 were prepared as previously reported [6f]. Reagents were obtained from commercial sources and used as received, except for pyrrole which was distilled under argon and stored at -23 °C in the dark. Solvents were dried and distilled under argon, prior to use. All reactions were performed in an inert atmo-sphere of argon by using standard Schlenk techniques and vacuum-line methods. Column chromatography was carried out under nitrogen, using silica (particle size 0.063–0.200 nm). Free base corroles were purifi ed using dry column vacuum chromatography (DCVC [14]) on silica PF 60 (Merck) as the stationary phase. NMR spec-tra (1H, 13C) were obtained at ambient temperature on a Bruker ARX 300 instrument. Chemical shifts are given in ppm relative to the residual proton resonance of the solvent. High resolution mass spectra were recorded using the ESI method on a Finnigan TSQ 700, or using the EI method at 70 eV on a Varian CH7. MALDI-TOF spectra were measured on a Bruker BiFlex IV instrument, using trans-3,5-dimethoxy-4-hydroxycinnamic acid as the matrix. The UV-vis spectra were measured on a Shimadzu UV-1601 PC spectrometer in concentrations of about 10-5 mol.L-1. The CD spectrum of 3 was record-ed in dichloromethane at c ~ 10 mmol.L-1 on a JASCO J-810 spectropolarimeter.

Preparation of 2,6-di(pivaloylamido)benzaldehyde 4. 2,6-dibromobenzaldehyde (2.34 g, 8.9 mmol) was transformed to the respective ethylene glycol acetal as described in the literature [15]. Under a blanket of argon, this acetal, pivaloylamide (2.1 g, 20.9 mmol), Cs2CO3

(8.5 g, 26.1 mmol), Pd(OAc)2 (78 mg, 0.35 mmol, 4 mol.%) and Xantphos (302 mg, 0.52 mmol, 6 mol.%) were dissolved in dioxane (35 mL) and heated to refl ux for 18 h. After cooling to room temperature, the brown mixture was diluted with dichloromethane (100 mL), repeatedly washed with water and dried with sodium sulfate. The solvent was removed in vacuo and the residue purifi ed by

column chromatography on silica with dichloromethane/diethylether (4:1). The compound was obtained as an off-white, waxy solid (2.78 g, 91% yield). 1H NMR (CDCl3):δ, ppm 9.05 (br.s, 2 H, NH), 7.95 (d, J = 8.3 Hz, 2 H, meta-H), 7.33 (t, J = 8.3 Hz, 1 H, para-H), 5.86 (s, 1 H, CH), 4.25–4.08 (m, 4 H, OCH2CH2O), 1.29 (s, 18 H, tBu-H).13C NMR (CDCl3): δ, ppm 176.6, 137.3, 129.9, 119.0, 114.6, 102.7, 64.6, 39.8, 27.5. HRMS (EI+): m/z calcd. for [C19H28N2O4]

+ 348.2049; found 348.2052; Δ = -0.3 mmu [M]+. The pivaloylamidobenzaldehyde acetal was dissolved in acetonitrile (200 mL) and treated dropwise, with stirring, with hydrochloric acid (60 mL, 2 mol.L-1).The mixture was stirred for 15 min and dichloromethane (300 mL) was added. The organic layer was separated, washed repeatedly with a saturated solution of sodium bicarbonate and then water, and dried with sodium sul-fate. The product was obtained, after removal of the sol-vent, in light yellow needles (2.36 g, 97% yield). 1H NMR (CDCl3): δ, ppm 9.89 (s, 1 H, CHO), 9.65 (br.s, 2 H, NH), 7.70 (d, J = 8.2 Hz, 2 H, meta-H), 7.48 (t, J = 8.3 Hz, 1 H, para-H), 1.33 (s, 18 H, tBu-H). 13C NMR (CDCl3): δ,ppm 192.3, 178.5, 141.2, 135.8, 118.9, 116.5, 39.9, 27.4. HRMS (EI+): m/z calcd. for [C17H24N2O3]

+ 304.1787; found 304.1791; Δ = -0.4 mmu [M]+.

Preparation of 5,10,15-tris(2,6-dibromophenyl)corrole 5. All steps had to be performed in red light. The synthetic protocol follows that in the literature [16]. A solution of 2,6-dibromobenzaldehyde (264 mg, 1 mmol) in pyrrole (350 μL, 5 mmol) is treated with trifl uoroacetic acid (16 μL, 0.21 mmol) and stirred at ambient tempera-ture for 16 h. Dichloromethane (10 mL) and triethylam-ine (29 μL, 0.21 mmol) were then added, and the mixture was dropped into thoroughly stirred dichloromethane (40 mL) simultaneously, with a solution of DDQ (227 mg, 1 mmol) in THF (10 mL). After 30 min, all volatiles were removed in vacuo. The dark residue was purifi ed by column chromatography on silica with n-hexane/di-chloromethane (3:2), and the title compound obtained as a violet solid (18 mg, 5% yield). 1H NMR: δ, ppm 8.99 (d, J = 4.2 Hz, 2 H, β-H), 8.52 (d, J = 4.7 Hz, 2 H, β-H), 8.39–8.35 (m, 4 H, β-H), 8.03 (d, J = 8.1 Hz, 6 H, meta-H), 7.56–7.49 (m, 3 H, para-H); NH protons not

Chart 1. General overview of achiral and chiral target molecules 1–3

R1 =

R2 =N N

NN

R2

H HH

NH

R1

O

NH

R1

O

HN

R1

O

HN

R1

O

OMe

OMe

OMe

R1 =

R2 =

R1 =

R2 =

H OMe

OMe

OMe

HNO

HNO

1

2

3


INVESTIGATING DIFFERENT STRATEGIES OF BIS-PICKET FENCE CORROLES 109

detected. UV-vis (CH2Cl2): λmax, nm (εrel) 412 (1.00), 424 (0.87), 571 (0.16), 608 (0.10), 636 (0.03); a quan-titative measurement has not been attempted due to the sensitivity of 5 to light. HRMS (ESI+): m/z calcd. for [C37H21N4Br6]

+ 1000.6800; found 1000.6822; Δ = -2.2 mmu [M]+.

Preparation of pivaloylamide substituted dipyr-rylmethane 6. The synthetic protocol follows that in the literature [17]. A solution of benzaldehyde 4 (1.5 g, 5 mmol) in pyrrole (35 mL, 0.5 mol) was purged with argon for 30 min. Magnesium bromide diethylether-ate (0.46 g, 2.5 mmol) was then added and the mixture stirred at room temperature for 90 min. Finely ground NaOH (1 g, 25 mmol) was then added to quench the reac-tion, followed by another 45 min of stirring. The mixture was allowed to sit for at least 2 h, fi ltered through a plug of celite, and the solids washed with additional pyrrole. All volatiles were distilled in vacuo at a maximum of 65 °C, and the residue was thoroughly dried in a high vacuum prior to chromatographic purifi cation on silica with dichloromethane/diethylether (9:1, then 4:1). After removal of the solvent, the title compound was obtained as a light yellow solid (560 mg, 27% yield). 1H NMR (CDCl3): δ, ppm 8.89 (br.s, 2 H, NH), 7.46–7.30 (m, 5 H, NH and Ph-H), 6.74–6.70 (m, 2 H, α-H), 6.18–6.08 (m, 4 H, β-H), 5.80 (s, 1 H, meso-H), 1.14 (s, 18 H, tBu-H). 13CNMR (CDCl3): δ, ppm 178.8, 136.8, 130.6, 129.1, 128.4, 125.5, 118.2, 108.0, 106.8, 39.3, 36.3, 27.3. HRMS (EI+): m/z calcd. for [C25H32N4O2]

+ 420.2525; found 420.2515; Δ = 1.0 mmu [M]+.


Three different preparative strategies have been designed and tested in order to synthesize the target mole-cules 1–3 in Chart 1. Strategy I (Scheme 1) follows known work from Rose et al. [6a, 18] and Collman and Decreau [6b] and makes use of 2,6-dinitrobenzaldehyde as the sin-gle precursor for the picket nitrogen atoms. This strategy was not successful. Neither a method described by Gryko and Koszarna [16] nor an alternative approach reported by Paolesse et al. [19] yielded the desired tris(dinitrophenyl)corrole. Moreover, the respective porphyrin is known to be formed in minor amounts [20]. The more soluble

4-tert-butyl-2,6-dinitrobenzaldehyde [18a] was found, by Rose et al., to produce much higher yields of the por-phyrin, and even the respective corrole has been observed as a by-product of this synthesis [18]. Furthermore, in a different paper, the same group describes the successful introduction of this aldehyde in the preparation of A2Bcorroles [6a]; however, due to the time-consuming syn-thesis of this aldehyde, we abandoned this strategy.

Strategy II makes use of the palladium-catalyzed ami-dation of bromoaryls reported by Buchwald and Yin [11] and successfully applied to 2,6-dibromophenyl substitut-ed porphyrins by Zhang et al. [12]. 2,6-dibromobenzal-dehyde [13] was amidated in three steps, using a simple acetal protection protocol for the redox sensitive formyl group [15] to 4 in an overall yield of 88%. Nonetheless, all attempts to use 4 in the one-pot macrocyclization to the superstructured corrole 1, failed, and the mass spectro-scopic analyses of the reaction mixtures obtained using different approaches [16, 19] proved the formation of mainly dipyrrolic products.

If the order of steps is inverted, it is observed that the tris(2,6-dibromophenyl)corrole 5 forms in a one-pot reaction under Gryko’s pseudo-high dilution conditions [16], but that unexpectedly this material is extremely light-sensitive. Mass spectrometrical analyses indicated the formation of oxygenated open-chain products upon exposure to light and air. This fi nding is in marked con-trast to the reported stability of respective chloro- and fl u-oro-derivatives [21], but is similar to other cases reported for corroles without electron-withdrawing groups [22]. We assume that the increase in steric hindrance, in addition to the increase in electron density, is responsible for the instability of 5. The new corrole can be isolated after chromatography and recrystallization in the dark as violet powder in 5% yield. All spectroscopic data of 5 point to a normal A3 corrole with the steric hindrance indicated by the typical split Soret band at 412/424 nm. When 5 was treated with the catalytic mixture for ami-dation, a dark solid of presumably elemental palladium formed with consumption of the starting material; neither an amidation product nor other corrole was obtained after workup. Several changes of the reaction conditions did not result in an improvement or in trace amounts of ami-dation products; again the strategy was abandoned.

N N

NNH H

H

NO2

NO2

O2N

O2N

CHONO2O2N

R

NH

R = H, tBu NO2O2N

R

R

R1+

[H+]

Strategy I

Scheme 1. Attempted preparation of superstructured A3 corrole 1 via the multiple reduction of nitro groups (Strategy I)



Since the formation of only dipyrrolic condensation products was observed during the attempted one-pot macrocyclization of 4 and all attempts to prepare the A3 type bis-picket fence corrole 1 failed, we switched to the building strategy for A2B corroles [23] to prepare 1, as well as 2 and 3, as new target molecules with less steric constraints. As depicted in Scheme 3, two path-ways were again investigated containing the amidation step at different stages of the synthesis. Starting from 4, the dipyrromethane 6 can, in fact, be easily prepared under appropriate conditions in 27% yield. 6, however, was found to be extremely unreactive with respect to the further condensation with the bis(pivaloylamido)benzal-dehyde 4, and even after two weeks reaction time, only the dipyrromethene 9 was identifi ed as the major product after DDQ oxidation (Chart 2).

The condensation step was monitored by MALDI-TOF mass spectrometry. After two days, the fi rst traces of condensation products appeared as the respective [M + 1]+ ion signals and slowly grew in intensity over the course of one week. Besides the dominating signal for 11, which forms under the conditions of the measure-ment from 6, only two signals at m/z = 707 and 1126 were found and interpreted as the 1:1 and the (partial-ly oxidized) 2:1 condensation products from 6 and 4, respectively. After DDQ oxidation, the spectrum chang-es markedly and shows signals at m/z = 721, 1138 and 839, besides the intense peak for 11. Chart 2 summarizes these results and presents the proposed structures of the detected molecular ions. Remarkably, no corrole forms, even in traces, despite the presence of the tetrapyrrolic precursor 11. We believe that the steric congestion of 11

Scheme 2. Attempted preparation of superstructured A3 corrole 1 using a palladium-catalyzed amidation step (Strategy II)

N N

NNH H

H

Br

Br

Br

Br

CHOBrBr

BrBr

1

Strategy II

+NH

[H+]

+O

H2N

CHONH

NH

O O1. glycol, H+

2. [Pd]3. hydrolysis

[Pd]

4

5

[H+]

Scheme 3. Successful preparation of superstructured A2B corroles 2 and 3 using a late palladium-catalyzed amidation step (Strategy III)

H H

N N

NNH H

H

Br

Br

Br

Br

CHOBrBr

1 or 2

Strategy III

+NH

4

8

[H+]

O

H2N R

[H+]NH

NH

O O

NN

6

CHO

MeOOMe

OMeor 4

H H

2 and 3[H+]BrBr

NN

7CHO

MeOOMe

OMe

MeO OMeOMe

[Pd]

[Pd]

H

R =

[H+] +NH



in combination with specifi c hydrogen bonds between some of the six amide functional groups introduces a stretched conformation of the open-chain intermediate, which renders macrocyclization impossible. When 6 is condensed with trimethoxybenzaldehyde or the very reactive pentafl uorobenzaldehyde, the transformation remains extremely slow. The respective A2B corroles can now be detected by mass spectrometry in both cases, but are present only in trace amounts and cannot be isolated from the reaction mixtures; therefore, this pathway was also abandoned.

Finally, we switched the order of the condensation and amidation steps again, in such a way that the four-fold amidation of the preassembled corrole 8 becomes the last step in the sequence. This approach was fi nally successful. The preparation of dipyrromethane 7 is straightforward and can be achieved in yields of 45% [6f]. The subsequent macrocyclization using trimethoxybenz-aldehyde affords corrole 8 in 22% yield [6f] and with-out any remarkable light sensitivity. Palladium-catalyzed amidation of 8 with either pivaloylamide or the chiral 2,2′-dimethylcyclopropylamide proceeds under similar conditions as described by Zhang for porphyrin systems [12] and yields the desired bis-picket fence corroles 2and 3 in 30 and 41% yield, respectively, in addition to di- and tri-substituted side products [6f]. No racemization or ring-opening was observed for 3. The chiral pickets are situated directly atop and below the corrole π system and

strongly infl uence the electronic transitions of the macro-cycle π-system as apparent from the strong Cotton effects in the CD spectrum of the compound (Fig. 1).

CONCLUSION

In summary, we have presented the exploration of several pathways towards superstructured bis-picket fence corroles and found an entry for chiral and achiral examples of sterically demanding A2B corroles. During the study, it became apparent that two points have to be considered with respect to the infl uence of steric congestion on the

Chart 2. Detected ions (MALDI-TOF) and proposed structures of the byproducts from the reaction of 6 and 4 with acid before (top) and after oxidation with DDQ (bottom)

N N

NNH H

H

piv

piv

piv

piv

before oxidation

11

H

pivpiv

NN

9pivpiv

H

pivpiv

NN

10

m/z([M+1]+) = 419

piv

piv

m/z([M+1]+) = 707 m/z([M+1]+) = 1126

N

N N

NHHH

pivpiv

pivpiv

after oxidation

14

m/z([M+1]+) = 839

N N

NNH H

piv

piv

piv

piv

13

pivpivH

pivpiv

NN

12

piv

piv

m/z([M+1]+) = 721 m/z([M+1]+) = 1138

H

O

O

Fig. 1. CD spectrum (CH2Cl2) of 3



a) Koszarna B and Gryko DT. 4. J. Org. Chem. 2006; 71: 3707–3717. b) Gryko DT. Eur. J. Org. Chem.2002; 1735–1743.a) Saltsman I, Mahammed A, Goldberg I, Tkachen-5.ko E, Botoshansky M and Gross Z. J. Am. Chem. Soc. 2002; 124: 7411–7420. b) Hiroto S, Hisaka I, Shinokubo H and Osuka A. Angew. Chem. Int. Ed.2005; 44: 6763–6766. c) Stefanelli M, Mastroianni M, Nardis S, Liccocia S, Fronczek FR, Smith KM, Zhu W, Ou Z, Kadish KM and Paolesse R. Inorg. Chem. 2007; 46: 10791–10799. a) Rose E and Andrioletti B. 6. J. Chem. Soc. Perkin Trans. 1 2002; 715–716. b) Collman JP and Decreau RA. Org. Lett. 2005; 7: 975–978. c) Barbe J-M, Canard G, Brandès S and Guilard R. Eur. J. Org. Chem. 2005; 4601–4611. d) Gryko TD, Tasior M, Peterle T and Bröring M. J. Porphyrins Phthalocyanines 2006; 10: 1360–1370. e) Maes W, Ngo TH, Vanderhaeghen J and Dehaen W. Org. Lett. 2007; 9: 3165–3168. f) Bröring M, Milsmann C, Ruck S and Köhler S. J. Organomet. Chem. 2008; DOI:10.1016/j.jorganchem.2008.04.039.Jaquinod L. In 7. The Porphyrin Handbook, Kadish KM, Smith KM and Guilard R. (Eds.), Academic: San Diego, CA, 2000; Vol. 1, 201–237.For an overview see: a) Momenteau M and Reed 8.CA. Chem. Rev. 1994; 94: 659–698. b) Collman JP, Boulatov R, Sunderland CJ and Fu L. Chem. Rev. 2004; 104: 561–588. c) Kim E, Chufán EE, Kamaraj K and Karlin KD. Chem. Rev. 2004; 104:1077–1134.a) Halime Z, Lachkar M, Roisnel T, Furet E, Halet 9.J-F and Boitrel B. Angew. Chem. Int. Ed. 2007; 46:5120–5124. b) Dürr K, Macpherson BP, Warratz R, Hampel F, Tuczek F, Helmreich M, Jux N and Ivanović-Burmazović I. J. Am. Chem. Soc. 2007; 129:4217–4228. c) Collman JP, Devaraj NK, Décreau RA, Yang Y, Yan Y, Ebina W, Eberspacher TA and Chidsey CED. Science 2007; 315: 1565–1568.a) Gross Z, Golubkov G and Simkhovich L. 10. Angew. Chem. Int. Ed. 2000; 39: 4045–4047. b) Maha-mmed A and Gross Z. J. Am. Chem. Soc. 2005; 127: 2883–2887.a) Buchwald SL and Yin J. 11. Org. Lett. 2000; 2:1101–1104. b) Buchwald SL and Yin J. J. Am. Chem. Soc. 2002; 124: 6043–6048.Zhang XP, Chen Y and Fields KB. 12. J. Am. Chem. Soc. 2004; 126: 14718–14719.Serwatowski J and Lulinski S. 13. J. Org. Chem. 2003; 68: 5384–5387.Pedersen DS and Rosenbohm C. 14. Synthesis 2001; 2431–2434.Schweikart K-H, Hanack M, Lüer L and Oelkrug 15.D. Eur. J. Org. Chem. 2001; 293–302.Gryko DT and Koszarna B. 16. Org. Biomol. Chem.2003; 1: 350–357.

preparation of corroles. On one hand, steric demand low-ers the condensation rate to impractical values. On the other, steric strain leads to unfavorable, although as yet unassigned, conformations of the open-chain tetrapyr-rolic precursor of corrole and thus inhibits macrocycle formation, in part, or completely. Due to these issues, and despite the ease of oxidizing and destroying signifi cant amounts of the corrole precursor by the reactive palladium precatalyst, the palladium- catalyzed amidation step has to be performed as the last step in the synthesis of bis-picket fence corroles for a successful preparation.

Acknowledgement

Funding of this work from the Deutsche Forschungs-gemeinschaft (DFG) is gratefully acknowledged.

REFERENCES

a) Nardis S, Monti D and Paolesse R. 1. Mini-Rev. Org. Chem. 2005; 2: 355–372. b) Gryko DT, Fox JP and Goldberg DP. J. Porphyrins Phthalocya-nines 2004; 8: 1091–1105. c) Aviv I and Gross Z. Chem. Commun. 2007; 1987–1999.a) Paolesse R. In 2. The Porphyrin Handbook, Vol. 2, Kadish KM, Smith KM and Guilard R. (Eds.), Academic: San Diego, CA, 2000; pp. 201–232. b) Kadish KM, Will S and Erben C. In The Porphyrin Handbook, Vol. 2, Kadish KM, Smith KM and Gui-lard R. (Eds.), Academic: San Diego, CA, 2000; pp. 233–300. c) Guilard R, Barbe J-M, Stern C and Kadish KM. In The Porphyrin Handbook, Vol. 18, Kadish KM, Smith KM and Guilard R. (Eds.), Elsevier Science: Boston, 2003; pp. 303–349.a) Gross Z, Galili N, Simkhovich L, Saltsman I, 3.Botoshansky M, Blaser D, Boese R and Goldberg I. Org. Lett. 1999; 1: 599–602. b) Paolesse R, Jaqui-nod L, Nurco DJ, Mini S, Sagone F, Boschi T and Smith KM. Chem. Commun. 1999; 1307–1308. c) Gryko DT. Chem. Commun. 2000; 2243–2244. d) Paolesse R, Nardis S, Sagone F and Khoury RG. J. Org. Chem. 2000; 66: 550–556. e) Gryko DT and Jadach K. J. Org. Chem. 2001; 66: 4267–4275. f) Bröring M and Hell C. Chem. Commun. 2001;2336–2337. g) Bröring M, Hell C, Brandt CD and Cónsul Tejero E. J. Porphyrins Phthalocyanines 2003; 7: 214–219. h) Steene E, Dey A and Ghosh A. J. Am. Chem. Soc. 2003; 125: 16300–16309. i) Jeandon C, Ruppert R and Callot HJ. Chem. Com-mun. 2004; 1090–1091. j) Geier III GK and Grin-drod SC. J. Org. Chem. 2004; 69: 6404–6412. k) Luguya RJ, Fronczek FR, Smith KM and Vicente MGH. Tetrahedron Lett. 2005; 46: 5365–5368. l) Bröring M, Brégier F, Cónsul Tejero E, Hell C and Holthausen MC. Angew. Chem. Int. Ed. 2007; 46:445–448. m) Bröring M, Hell C and Brandt CD. Chem. Commun. 2007; 1861–1862.



Lindsey JS, Laha JK, Dhanalekshmi S, Taniguchi 17.M and Ambroise A. Org. Process Rev. Dev. 2003; 7: 799–812.a) Rose E, Kossanyi A, Quelquejeu M, Soleil-18.havoup M, Duwavran F, Bernard N and Lecas A. J. Am. Chem. Soc. 1996; 118: 1567–1568. b) Rose E, Andrioletti B and Pandian RP. J. Porphyrins Phthalocyanines 2002; 6: 602–606.Paolesse R, Marini A, Nardis S, Froiio A, Mandoj 19.F, Nurco DJ, Prodi L, Montalti M and Smith KM. J. Porphyrins Phthalocyanines 2003; 7: 25–36.a) Corden BB and Drain CM. 20. Inorg. Chem. 1989; 28: 4374–4376. b) Shelnutt JA, Quintana CA

and Assink RA. Inorg. Chem. 1989; 28: 3421–3425.Gross Z, Galili N and Saltsman I. 21. Angew. Chem. Int. Ed. 1999; 38: 1427–1429.a) Geier III GR, Chick JFB, Callinan JB, Reid 22.CG and Auguscinski WP. J. Org. Chem. 2004; 69:4159–4169. b) Ziegler CJ, Ding T, Aleman EA and Modarelli DA. J. Phys. Chem. A 2005; 109:7411–7417. c) Tardieux C, Gros CP and Guilard R. Heterocycl. Chem. 1998; 35: 965–970.a) Dehaen W, Asokan CV and Smeets S. 23. Tetrahe-dron Lett. 2001; 42: 4483–4485. b) Gryko DT and Koszarna B. Synthesis 2004; 13: 2205–2209.




INTRODUCTION

Tetrathiafulvalene (TTF) and its derivatives have been extensively studied as electron donors in organic conductors, organic solar cells, and so on [1–3] becau-se these can be obtained by simple synthetic proceduresto allow fine-tuning of their redox properties. The one-electron oxidation of TTF gives its cation radical with an oxidation potential of ca. 0.34 V vs Ag/AgCl, whereas the one-electron oxidation and reduction potentials of tetraphenylporphyrins are around 0.8 to 1.0 V and -1.11to -1.33 V, respectively [4, 5]. This suggests that the porphyrin will become an acceptor unit when linked with TTF to afford interesting D-A molecules as potential applications for nonlinear optics as well as organic conductors.

Recently, much attention has been focused on materialsthat exhibit high two-photon absorption (2PA) efficiency because of their potential applications for two-photon

photodynamic therapy (2PA-PDT) [6–9], 3-D optical me- [6–9], 3-D optical me-, 3-D optical me-mory [10–15], optical power limiting [16], 3-D microfa- [10–15], optical power limiting [16], 3-D microfa-, optical power limiting [16], 3-D microfa- [16], 3-D microfa-, 3-D microfa-brication [17, 18], and fluorescence microscopy [19–21].Porphyrins are potential candidates for 2PA materials because they have a highly conjugated -electron system.However, monomeric porphyrins without donor/acceptor groups have only small (2) values of less than a few tens of GM (1 GM = 10-50 cm4.s.molecule-1.photon-1) [22, 23].We have reported that the connection of -conjugation between chromophores using an acetylene bridge and in-troduction of a donor (D) and acceptor (A) are significantfor the enhancement of 2PA [24, 25].

Although, several D-A molecules using TTF and por-phyrin or phthalocyanine have been synthesized [26–33], no study of a TTF-porphyrin system connected with acetylene has been reported. Thus, we have designed a new porphyrin-TTF conjugate 1, where two porphyrins are bridged by TTF using acetylene bonds, giving a class of A-D-A configuration (Scheme 1). Herein, we report the synthesis, electrochemical and photophysical properties of 1, including steady-state absorption and emission spectra, and nonlinear absorption properties measured using a nanosecond Z-scan method.

Bisporphyrin connected by tetrathiafulvalene

Kazuya Ogawa* and Yasunori Nagatsuka

Graduate School of Materials Science, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma, Nara 630-0101, Japan


ABSTRACT: A new porphyrin-tetrathiafulvalene composite, where two porphyrins are bridged by tetrathiafulvalene (TTF) using acetylene bonds was synthesized. The Q-band of the monomeric porphyrin appears at 590 nm whereas that of the composite is red-shifted to 620 nm and intensified. The Soret band is also red-shifted from 427 nm to 435 nm and much broadened, indicating the expansion of -conjugation over the porphyrin and tetrathiafulvalene units. The HOMOs and LUMOs were calculated using time-dependent density functional theory. Voltammetric experiments revealed that the fi rst oxidation poten-. Voltammetric experiments revealed that the fi rst oxidation poten-Voltammetric experiments revealed that the fi rst oxidation poten-revealed that the fi rst oxidation poten-the first oxidation poten-tial of the TTF moiety in the composite was shifted by +155 mV compared with TTF in the absence of composite. The effective two-photon absorption (2PA) cross section values were measured by using a nanosecond open aperture Z-scan method. The maximum effective 2PA cross section values were obtained at 760 nm, as 7300 GM in benzonitrile and 5900 GM in toluene. The values obtained in the polar solvent were 1.2 to 1.5 times larger than those in the nonpolar solvent.

KEYWORDS: porphyrin, TTF, two-photon absorption, acetylene.


*Correspondence to: Kazuya Ogawa, email: [email protected], fax: +81 743-72-6219


BISPORPHYRIN CONNECTED BY TTF 115


Scheme 2 shows the synthetic route for triad 1. Diiode-TTF 9 was synthesized according to the literature [34]. Porphyrin 4 was synthesized from aldehyde 2 and dipyr-romethane 3 with trifluoroacetic acid (TFA) followed by chloranil oxidation in a 58% yield [35]. One of the meso-positions was brominated by N-bromosuccinimide (NBS). After the zinc insertion, a cross-coupling reaction with the TMS-acetylene group using a Pd2(dba)3/AsPh3 catalytic system was performed. Finally, deprotected compound 8 was reacted with 9 using the Pd2(dba)3/AsPh3 catalytic system to afford the target 1. After purification by silica gel column chromatography, the MALDI-TOF mass spectrum and TLC showed a single species with an appropriate mass peak at m/z = 1584.18. However, the 1H NMR spectrum of 1 showed two isomers with no tem-perature dependency on the spectrum shape. The differ-ences in chemical shift between the two species for the

porphyrin ring were within 0.01 ppm (Fig. 1). However, the difference of the characteristic signals arising from the TTF unit, at 7.00 and 6.97 ppm, was relatively large allowing the integral ratio of 0.70:1.30. Since the largest difference in the chemical shift was observed for TTF protons, the two species are probably attributed to trans/cis isomers at the TTF moiety. The two peaks at 7.00 and 6.97 ppm may be assigned to trans and cis isomers, respectively, because the trans isomer, which has a shorter distance between the TTF proton and the porphyrin at the far position by 0.4 Å compared to the cis isomer, would more largely be affected by ring current effects from the porphyrin rings. At first, we expected that only a transisomer would be obtained since the NMR of 9 showed a single peak and only a trans isomer was described in the literature [34]. The two isomers could not be separated by column chromatography. In order to gain the electronic structures of both isomers, time-dependent density functional theory (TDDFT) calculations with B3LYP function and 6-31G(d) basis set of Gaussian-03 Package were performed. The frontier HOMO and LUMO orbitals of each isomer of 1 are shown in Fig. 2. Aryl groups were omitted for simplification. In both isomers, the HOMO and LUMO were found to be mainly localized on the TTF and porphyrin moieties, respectively, indicating a favorable property for photoinduced electron transfer.

The UV-vis absorption spectrum of 1 in toluene is shown in Fig. 3 (solid line) along with that of 8 (dashed line). The Q-band of 1 observed at 620 nm was

p+

Scheme 1. Structure TTF-bridged bisporphyrin 1

Scheme 2. Synthetic route to 1


116 K. OGAWA AND Y. NAGATSUKA

red-shifted by 30 nm compared with that of the startingcompound 8. The Soret band of 1 was also red-shiftedfrom 427 nm to 435 nm and much broadened. These observations suggest an expansion of the -conjugation over the porphyrin and TTF units in 1. Figure 4 shows the normalized fluorescence spectra of 1 (solid line) and 8 (dashed line) in toluene. The fluorescence quantum yield of 1 in toluene was determined to be 0.022, which

was smaller than that of 8 (0.030), and further decreased to 0.017 in the polar benzonitrile solvent, suggesting that electron transfer occurs between the porphyrin and TTF moieties upon photoirradiation. The time-resolved fluorescence of 1 was measured by using a pulsed-N2-dye laser with a streak camera and determined to be 1.6 ns in toluene and 1.0 ns in benzonitrile, respectively, with single exponential decay profiles in both cases. These

Fig. 1. 1H NMR spectrum of 1 in CDCl3

Fig. 3. UV-visible absorption spectra of 1 (solid line) and 8(dotted line) in toluene

Fig. 4. Fluorescence spectra of 1 (solid line) and 8 (dotted line)in toluene

Fig. 2. Electron density maps of HOMO and LUMO levels of trans and cis isomers of 1. Aryl groups were omitted for simplification



measurements showed that there is an ignorable diffe-rence in spectroscopic properties between the trans and cis isomers.

The electrochemical properties of 1 were examined bydifferential pulse voltammogram (DPV) measurementsin benzonitrile (0.1 M tetrabutylammonium perchlorate).Figure 5 shows deconvoluted DPV charts (vs Fc+/Fc) of 8 (top), TTF (middle), and 1 (bottom). The first and sec-ond oxidation potentials of monomer 8 and TTF wereobserved at 440, 630, and -80, 325 mV, respectively. The DPV of 1 showed two broad peaks at around 75 mV and 300 mV. The former is assignable to the first oxidation potential of the TTF moiety in 1 which was shifted by +155 mV compared with TTF. This indicates that the connection of -conjugation between the porphyrin and TTF in 1 shifts the potential to a more difficult oxidation.The larger peak which appeared at 300 mV may be the

first oxidation potential of the porphyrin part, probably overlapping with the second oxidation of the TTF moiety. Unfortunately, a clear reduction peak potential could not be observed.

The nonlinear absorption measurements were per-formed using a nanosecond Z-scan method in benzo-nitrile and toluene, and were analyzed as the effective 2PA including the excited-state absorption [25]. Figure6 represents typical Z-scan traces of 1 (2.0 × 10-3 M in benzonitrile) with theoretically fitted curves at 760, 770, 780, 800, 810, and 820 nm. In these fi gures, the normali-. In these fi gures, the normali-these fi gures, the normali- figures, the normali-s, the normali-, the normali-zed transmittance was plotted as a function of the sample position, z. The dip was observed around the focal posi-he dip was observed around the focal posi-tion. The curve fit was performed according to the theore-as performed according to the theore- performed according to the theore-tical equations [24, 25, 36, 37] assuming 2PA:

(1)

qq

( ) 021

(2)

q0 = (2) (1 - R)I0 Leff (3)

Leff = [1 - exp(- (1)L)]/ (1) (4)

eff (2) = h- (2)/N (5)

where is the normalized z-position ( = (z - z0)/zR), and z0 and zR are the focal position and the Rayleigh range, respectively. (1) is a one-photon absorption coefficient,R denotes the Fresnel reflectance, and L is the path length (2 mm). (2) is the 2PA coefficient, Leff denotes the effec-tive path length, and I0 is the peak intensity at the focal position. N is the number density of the molecule and h-

is the photon energy of the incident light. Finally, the ef-fective 2PA cross-section value, eff (2), was obtained from Equation 5. The Z-scan curves at all wavelengths could be well fitted with the reasonable Rayleigh range of 5 mm, which agreed with that obtained for a standard sam-ple of Rhodamine B [35, 38].

It should be noted here that the cross-section value measured using nanosecond laser pulses is 2–3 orders of magnitude larger than that obtained with femtosecond pulses [25, 39–41] because of the contribution of ESA to the data. This effect is difficult to separate from the pure 2PA; therefore, we cannot compare directly between them. However, in some applications using 2PA mate-However, in some applications using 2PA mate-rials such as 2PA-PDT, a laser with a nanosecond pulse,which is already available in many medical institutions,may be more appropriate because of the ease of operation. For this reason, we have investigated the nanosecond non- have investigated the nanosecond non-investigated the nanosecond non-investigated the nanosecond non- the nanosecond non-linear absorption properties of porphyrin compounds [9,25, 35, 38] as well as femtosecond studies.

2PA spectra of 1 are shown in Fig. 7. The maxi- shown in Fig. 7. The maxi-7. The maxi-. The maxi-The maxi-he maxi-mum eff (2) values obtained were found to be 7300 GMin benzonitrile and 5900 GM in toluene at 760 nm in

Fig. 5. Differential pulse voltammograms of 8 (top), TTF (mid-dle), and 1 (bottom) in benzonitrile containing 0.1 M TBAP as a supporting electrolyte with sweep rate of 10 mV.s-1



a wavelength range from 760 nm to 820 nm. The values obtained in polar solvent of benzonitrile were 1.2 to 1.5 times larger than those in the nonpolar solvent,toluene. This suggests that the eff (2) value correlates with fluorescence quenching, i.e. electron transfer efficiency. The eff (2) values increased with shortening of the wa-s increased with shortening of the wa- increased with shortening of the wa-velength of incident light. Such behavior was observed in porphyrin monomers such as H2TPP and ZnOEP andwas explained as a resonance enhancement nearby one-photon absorption of the Q-band [22, 23, 42]. Unfortuna-22, 23, 42]. Unfortuna-]. Unfortuna-Unfortuna-Unfortuna-nfortuna-tely, the eff (2) value for 8 was too small to be determined and probably less than 50 GM from the experimental conditions (the cross-section value was also less than 20 GM in the femtosecond measurement [15]). Therefore, the eff (2) value for 1 is two orders of magnitude larger

Fig. 7. Two-photon absorption spectra of 1 in benzonitrile ( )and toluene (X)

Fig. 6. Typical Z-scan traces of 1 (2.0 × 10-3 M in benzonitrile) with theoretically fitted curves at 760, 770, 780, 800, 810, and 820 nm



than that of 8. This enhancement is obviously attributed to the connection of the TTF moiety to the porphyrins. The eff (2) value of 7300 GM is comparable to that obtai-ned for directly connected bisporphyrin by butadiyne bond (7700 GM at 860 nm), where stronger interactionexists between two porphyrins [35]. The femtosecond 2PA measurement for 1 will be measured and reported elsewhere. The conductivity of 1 in the film is also now under active investigation and will be reported in the near future.

EXPERIMENTAL

General procedures1H NMR spectra were recorded on a JEOL ECP 600

with Me4Si as the internal standard ( = 0 ppm). UV-vis spectra were measured using a Shimadzu UV-1650PC UV-visible spectrophotometer. Fluorescence measurements were performed on a Hitachi F-4500 fluorescence spectro-photometer. MALDI-TOF mass spectra were obtained on a PerSeptive Biosystems Voyager DE-STR or a Shimadzu/KRATOS Axima-CFR MALDI with dithranol (Aldrich) as the matrix. Voltammetric experiments were performed using a BAS CV-50W potentiostat. Reactions were monito-red on silica gel 60 F254 TLC plates (Merck). The silica gel utilized for column chromatography was purchased from Kanto Chemical Co. Inc.: Silica Gel 60N (Spherical, Neu-tral) 60–210 μm and 40–50 μm (Flash). Dry tetrahydro-furan (THF) was obtained by refluxing with sodium and benzophenone followed by distillation. Diiode-TTF 2 wassynthesized according to the literature [34].

Synthesis

Preparation of 5,15-bis(ethoxycarbonylphenyl)porphyrin (4). Under nitrogen atmosphere, trifl uoroace-nitrogen atmosphere, trifl uoroace- atmosphere, trifl uoroace-trifl uoroace-trifluoroace-tic acid (0.25 mL, 3.1 mmol) was added to a solution of 2(570 mg, 3.1 mmol) and 3 (460 mg, 3.1 mmol) in CHCl3

(1 L). After stirring for 20 h in the dark at rt, triethylami-1 L). After stirring for 20 h in the dark at rt, triethylami-L). After stirring for 20 h in the dark at rt, triethylami-h in the dark at rt, triethylami-h in the dark at rt, triethylami-ne (0.45 ml, 3.1 mmol) and chloranil (2.20 g, 9.00 mmol) were added. The reaction mixture was further stirred at rtfor 12 h. After removing the solvent, the crude material was purified by silica gel column chromatography with CHCl3 as the eluent which yielded 4 as a purple solid (1.09 g, 1.79 mmol, 57.7%). 1H NMR (600 MHz; CDCl3;Me4Si): H, ppm 10.36 (s, 2H, Por meso), 9.41 (d, J = 4.6Hz, 4H, Por ), 9.04 (d, J = 4.6 Hz, 4H, Por ), 8.50 (d, J = 8.1 Hz, 4H, Ph), 8.36 (d, J = 8.1 Hz, 4H, Ph), 4.61(q, J = 7.0 Hz, 4H, –CH2CH3), 1.58 (t, J = 7.0 Hz, 6H,–CH2CH3), -3.12 (s, 2H, NH). MS (MALDI-TOF; dithra-2 (s, 2H, NH). MS (MALDI-TOF; dithra- (s, 2H, NH). MS (MALDI-TOF; dithra-; dithra- dithra-nol): m/z 606.79. Calcd. for [M]+ 606.23.

Preparation of 5,15-bis(ethoxycarbonylphenyl)-10-bromoporphyrin (5). N-bromosuccinimide (17 mg, 93μmol,) was added to a solution of 4 (113 mg, 0.19 mmol) in CHCl3 (10 mL) at -40 °C. After stirring for 20 min in the dark, the solvent was removed. The crude product was extracted with CHCl3, washed with water and dried over

anhydrous Na2SO4. The desired fraction was separated via silica gel column chromatography (eluent: CHCl3)to yield 5 (36.2 mg, 56 μmol, 30%) and 64% of 4 were recovered. 1H NMR (600 MHz; CDCl3; Me4Si): H, ppm10.20 (s, 1H, Por meso), 9.72 (d, J = 4.6 Hz, 2H, Por ),9.25 (d, J = 4.6 Hz, 2H, Por ), 8.88-8.87 (d × 2, J = 4.6Hz, 4H, Por ), 8.47 (d, J = 8.1 Hz, 4H, Ph), 8.22 (d, J = 8.1 Hz, 4H, Ph), 4.60 (q, J = 7.0 Hz, 4H, –CH2CH3),1.57 (t, J = 7.0 Hz, 6H, –CH2CH3), -3.02 (s, 2H, NH). MS (MALDI-TOF; dithranol): m/z 685.62. Calcd. for[M + H]+ 685.15.

Preparation of 5,15-bis(ethoxycarbonylphenyl)-10-trimethylsilylethynylporphyrinatozinc (7). Zinc ace-tate dihydrate (593 mg, 2.7 mmol) in MeOH (2 mL) was added to 5 (116.7 mg, 0.18 mmol) in CHCl3 (10 mL). After stirring for 1 h, the mixture was washed well with water, extracted with CHCl3 and dried over anhydrous Na2SO4. The crude product was purified via silica gel column chromatography (eluent: CHCl3) to give a pink solid of 6 (143.9 mg, 0.19 mmol, 99.9%). A vacuum-dried Schlenk flask was charged with 6 (143.9 mg, 0.19 mmol) and filled with argon atmosphere. THF (10 mL), TEA(56.0 mg, 0.57 mmol), Pd(PPh3)2Cl2 (16.2 mg, 23 μmol), CuI (2.2 mg, 12 μmol), and TMS-acetylene (56.0 mg,0.57 mmol) were added under argon atmosphere. After stirring in the dark at rt for 3 h, the reaction mixture was washed with saturated aqueous NaHCO3 and water, extracted with CHCl3 and dried over anhydrous Na2SO4.The crude product was purified by silica gel column chromatography (eluent: 8:1 hexane/EtOAc) to give 7 (110.9 mg, 0.15 mmol, 76%). 1H NMR (600 MHz;CDCl3; Me4Si): H, ppm 9.77 (s, 1H, Por meso), 9.67 (d, J = 4.8 Hz, 2H, Por ), 9.02 (d, J = 4.8 Hz, 2H, Por ),8.81–8.71 (d × 2, J = 4.8 Hz, 4H, Por ), 8.23 (d, J = 7.8Hz, 4H, Ph), 8.11 (d, J = 7.8 Hz, 4H, Ph), 4.34 (q, J = 7.2Hz, 4H, –CH2CH3), 1.43 (t, J = 7.2 Hz, 6H, –CH2CH3),0.55 (s, 9H, TMS). MS (MALDI-TOF; dithranol): m/z764.52. Calcd. for [M]+ 764.18.

Preparation of 5,15-bis(ethoxycarbonylphenyl)-10-ethynylporphyrin (8). 7 (27 mg, 35 μmol) was dissolved in CHCl3 (5 mL) and 1 M tetrabutylammonium fluoride (TBAF) (53 μL) in THF was added. After stirring for 30 min at rt in the dark, the reaction solution was washed with saturated aqueous NaHCO3 and water, extracted with CHCl3 and dried over anhydrous Na2SO4. The target com-pound was further purified by silica gel column chroma-tography (eluent: CHCl3) to give the title compound (23.6mg, 34 μmol, 96%). 1H NMR (600 MHz; CDCl3; Me4Si):

H, ppm 10.14 (s, 1H, Por meso), 9.77 (d, J = 4.8 Hz, 2H, Por ), 9.30 (d, J = 4.8 Hz, 2H, Por ), 8.95–8.92 (d × 2, J = 4.8 Hz, 4H, Por ), 8.41 (d, J = 7.8 Hz, 4H, Ph), 8.25 (d, J = 7.8 Hz, 4H, Ph), 4.54 (q, J = 7.2 Hz, 4H, –CH2CH3), 4.15 (s, 1H, –C CH), 1.55 (t, J = 7.2 Hz, 6H,–CH2CH3). MS (MALDI-TOF; dithranol): m/z 692.20.Calcd. for [M]+ 692.14.

Preparation of compound 1. Under an argon atmo-sphere, 9 (9.2 mg, 20.2 μmol) and 8 (28.7 mg, 40.0 μmol)



were dissolved in THF (5 mL) after which triethylamine (1 mL), Pd2(dba)3 (2.5 mg, 2.4 μmol), and AsPh3 (6.2 mg, 20.2 μmol) were added. After stirring for 9 h, the reac- μmol) were added. After stirring for 9 h, the reac-9 h, the reac- h, the reac-tion solution was evaporated. The residue was dissolved in chloroform and washed with water. The crude pro-duct was purified by silica gel column chromatography (eluent: 35:1 chloroform/methanol) to give 1 (5.9 mg, 18.4%). 1H NMR (600 MHz; CDCl3; Me4Si): H, ppm 10.19 and 10.18 (two singlets, 2H, meso), 9.69 and 9.68(two doublets, 4H, ), 9.32 and 9.31 (two doublets, 4H,

), 8.96 and 8.95 (two doublets, 4H, ), 8.92 and 8.91(two doublets, 4H, ), 8.47 and 8.45 (two doublets, 8H, Ph), 8.32 and 8.31 (two doublets, 8H, Ph), 7.00 (s, 0.75H, TTF), 6.97 (s, 1.25H, TTF), 4.65 (two quartets, 8H, Et), 1.59 (two triplets, 12H, Et). UV-vis (CHCl3): max, nm 434, 560, 624. Fluorescence ( ex = 435 nm; CHCl3): max,nm 625. MS (MALDI-TOF; dithranol): m/z 1584.18. Calcd. for [M]+ 1584.17.

Measurement of nonlinear absorption

Nanosecond open-aperture Z-scan measurements were performed by using an optical parametric oscillator (Conti-nuum Surelight OPO) pumped with a THG beam (355 nm) from a Q-switched Nd:YAG laser (Continuum Sureli-ght I-10) ranged from 760 nm to 820 nm [25]. The pulse width and the repetition rate were 5 ns and 10 Hz, respec-tively. The laser beam was focused using a plano-convex lens (f = 100 mm) with a beam waist of around 0.035 mm at the focal point. The peak intensity of the incident pulses at the focal point I0 were 1.0 to 1.3 × 1014 W.m-2.The sample solution was placed in a 2 mm quartz cuvette and stirred using a magnetic bar during the measurement.

Acknowledgements

This work was supported by Grant-in-Aids for Young Scientists (B) (no. 18750118) from the Ministry of Edu-cation, Culture, Sports, Science and Technology, Japan (Monbu Kagakusho), and by Iketani Science Technology Foundation.

REFERENCES

Yamada J-I, Akutsu H, Nishikawa H and Kikuchi K. 1.Chem. Rev. 2004; 104: 5057.Iyoda M, Hasegawa M and Miyake Y. 2. Chem. Rev.2004; 104: 5085.Lorcy D and Bellec N. 3. Chem. Rev. 2004; 104: 5185.Hodge JA, Hill MG and Gray HB. 4. Inorg. Chem.1995; 34: 809.Ravikanth M, Reddy D, Misra A and Chandrashe-5.kar TK. J. Chem. Soc. Dalton Trans 1993; 1137.Bhawalkar JD, Kumar ND, Zhao CF and Prasad 6.PN. J. Clin. Laser Med. Surg. 1997; 15: 201.Wachter EA, Partridge WP, Fisher WG, Dees HC 7.and Petersen MG. Proc. SPIE-Into. Soc. Opt. Eng.1998; 3269: 68.

Ogawa K, Hasegawa H, Inaba Y, Kobuke Y, Inouye 8.H, Kanemitsu Y, Kohno E, Hirano T, Ogura S and Okura I. J. Med. Chem. 2006; 49: 2276.Dy JT, Ogawa K, Satake A, Ishizumi A and Kobuke 9.Y. Chem. Eur. J. 2007; 13: 3491.Parthenopoulos DA and Rentzepis PM. 10. Science1989; 245: 843. Strickler JH and Webb WW. 11. Opt. Lett. 1991; 16:1780.Dvornikov AS and Rentzepis PM. 12. Opt. Commun.1995; 119: 341. Shen Y, Friend CS, Jiang Y, Jakubczyk D, Swiat-13.kiewicz J and Prasad PN. J. Phys. Chem. B 2000; 104: 7577. Dvornikov AS, Liang Y, Cruse CS and Rentzepis 14.PM. J. Phys. Chem. B 2004; 108: 8652.Dy JT, Maeda R, Nagatsuka Y, Ogawa K, Kamada 15.K, Ohta K and Kobuke Y. Chem. Commun. 2007; 5170.Nalwa HS and Miyata S. (Eds.). 16. Nonlinear Optics of Organic Molecules and Polymers, CRC Press: Boca Raton, FL, 1997.Kawata S, Sun H-B, Tanaka T and Takada K. 17.Nature 2001; 412: 697.Zhou W, Kuebler SM, Braun KL, Yu T, Cammack 18.JK, Ober CK, Perry JW and Marder SR. Science2002; 296: 1106.Kohler RH, Cao J, Zipfel WR and Webb WW. 19.Science 1997; 276: 2039.Caylor CL, Dobrianov I, Kimmer C, Thorne RE, 20.Zipfel W and Webb WW. Phys. Rev. 1999; E59:R3831.Miller MJ, Wei SH, Parker I and Cahalan MD. 21.Science 2002; 296: 1869.Drobizhev M, Karotki A, Kruk M and A. Rebane. 22.Chem. Phys. Lett. 2002; 355: 175.Drobizhev M, Karotki A, Kruk M, Mamardashvili 23.NZ and Rebane A. Chem. Phys. Lett. 2002; 361:504.Ogawa K, Ohashi A, Kobuke Y, Kamada K and Ohta 24.K. J. Am. Chem. Soc. 2003; 125: 13356.Ogawa K, Ohashi A, Kobuke Y, Kamada K and 25.Ohta K. J. Phys. Chem. B 2005; 109: 22003.Blower MA, Bryce MR and Devonport W. 26. Adv. Mater. 1996; 8: 63.Wang C, Bryce MR, Batsanov AS, Stanley CF, 27.Beeby A and Howard JAK. J. Chem. Soc. Perkin Trans. 1997; 2: 1671.Becher J, Brimert T, Jeppesen JO, Pedersen JZ, Zu-28.barev R, Bjørnholm T, Reitzel N, Jensen TR, Kjaer K and Levillain E. Angew. Chem. Int. Ed. 2001; 40:2497.Liddell PA, Kodis G, de la Garza L, Bahr JL, Moore 29.AL, Moore TA and Gust D. Helv. Chim. Acta 2001; 84: 2765.Farren C, Christensen CA, FitzGerald S, Bryce MR 30.and Beeby A. J. Org. Chem. 2002; 67: 9130.



Li H, Jeppesen JO, Levillain E and Becher J. 31. Chem.Commun. 2003: 846.Sadaike S-I, Takimiya K, Aso Y and Otsubo T. 32.Tetrahedron Lett. 2003; 44: 161.Loosli C, Jia C, Liu S-X, Haas M, Dias M, Levillain 33.E, Neels A, Labat G, Hauser A and Decurtins S. J. Org. Chem. 2005; 70: 4988.Ranganathan A, El-Ghayoury A, Mézière C, Harté E, 34.Clérac R and Batail P. Chem. Commun. 2006; 2878.Ogawa K, Dy J and Kobuke Y. 35. J. Porphyrins Phthalocyanines 2005; 9: 735Sheik-Bahae M, Said AA, Wei T-H, Hagan DG and 36.van Stryland EW. IEEE J. Quant. Electr. 1990; 26:760.

Audebert P, Kamada K, Matsunaga K and Ohta K. 37.Chem. Phys. Lett. 2003; 367: 62.Tanihara J, Ogawa K and Kobuke Y. 38. J. Photochem. Photobiol. A. 2006; 178: 140.Swiatkiewicz J, Prasad PN and Reinhardt BA. 39. Opt.Commun. 1998; 157: 135. Kim O-K, Lee K-S, Woo HY, Kim K-S, He GS, 40.Swiatkiewicz J and Prasad PN. Chem. Mater. 2000; 12: 284.Lei H, Wang HZ, Wei ZC, Tang XJ, Wu LZ, Tung 41.CH and Zhou GY. Chem. Phys. Lett. 2001; 333:387.Kamada K, Ohta K, Iwase Y and Kondo K. 42. Chem.Phys. Lett. 2003; 372: 386.




INTRODUCTION

Recently, organic dye has attracted great attention as optical, optoelectronic, and electronic materials for the application of multi-photon absorbing materials, organic light-emitting diodes (OLED) and organic fi eld-effect transistors (OFET). Porphyrins, phthalocyanines, and their related compounds are recognized as promising candidates for such materials [1] because they are fairly stable under aerobic conditions in spite of their large π-systems. Many synthetic methods have enabled us to access a variety of their derivatives, properties of which are modifi ed from their parent π-systems by substitution at meso and/or β positions. Although preparation of the parent compounds such as porphine, phthalocyanine, and tetrabenzoporphine (TBP) is well known, purifi ca-tion of these completely fl at and rigid molecules is rather

diffi cult. Parent monobenzo-, dibenzo-, and tribenzo-por-phyrins with no peripheral substituent had not been pre-pared, probably due to the lack of a reliable preparation method and diffi cult purifi cation, although benzoporphy-rins bearing substituents at meso and/or β positions have been prepared by several groups [2]. We have developed the preparation method of highly π-conjugated molecules [3] in high purity. Our strategy is based on pericyclic cy-cloreversion represented by the thermal retro-Diels-Alder reaction of bicyclo[2.2.2]octadiene to benzene and has many advantages for the preparation of such highly conju-gated macrocyclic compounds. The fi nal pericyclic cyclo-reversion of precursors to the fl at macrocycles needs no chemical substance such as solvents and reagents. Only energy for activation is required, and the by-products are gaseous neutral molecules which instantaneously separate from the product. Reactive intermediates such as radicals and ions are not formed during the reaction. In this paper, the preparation of these benzoporphyrins (2a–c) and their BCOD precursors (1a–c), as well as their structures and electronic spectra, is discussed (Fig. 1).

Synthesis, structures, and properties of BCOD-fused

porphyrins and benzoporphyrins

Hiroki Uoyamaa, Takahiro Takiuea, Kazuyuki Tominagaa, Noboru Onob

and Hidemitsu Uno*a

a Division of Synthesis and Analysis, Department of Molecular Science, Integrated Center for Sciences (INCS), Ehime University, Matsuyama 790-8577, Japanb Department of Chemistry and Biology, Graduate School of Science and Engineering, Ehime University, Matsuyama 790-8577, Japan

Received 15 August 2008Accepted 13 October 2008

ABSTRACT: Bicyclo[2.2.2]octadiene (BCOD)-fused porphyrins with no other substituents were prepared by [2 + 2] and [3 + 1] porphyrin syntheses from ethanodihydroisoindole derivatives in fairly good yields. Thermal retro-Diels-Alder reactions of BCOD-fused porphyrins gave the corres ponding benzoporphyrins with no substituent in quantitative yields. Their UV-vis spectra and crystal structures were carefully examined in terms of π-system expansion of the porphyrin ring current. In the cases of monobenzo- and adj-dibenzo-porphyrins, a single Soret band in their UV-vis spectra and no bond alteration in the benzene rings of their crystal structures were observed, while the split Soret bands and the obvious bond alteration in their benzene rings were recorded in the case of opp-dibenzoporphyrin.

KEYWORDS: benzoporphyrin, bicyclo[2.2.2]octadiene, retro-Diels-Alder reaction, electronic spectra, crystal structure.


*Correspondence to: Hidemitsu Uno, email: [email protected], fax: +81 89-927-9670, tel: +81 89-927-9660


SYNTHESIS, STRUCTURES, AND PROPERTIES OF BCOD-FUSED PORPHYRINS AND BENZOPORPHYRINS 123


Preparation of BCOD-fused pyrrole derivatives

We decided to prepare the BCOD-fused porphyrins by [3 + 1] and [2 + 2] porphyrin syntheses, which were proved to be reliable synthetic methods for the prepara-tion of asymmetric porphyrins [4]. For utilization of these methods for the preparation of BCOD-fused porphyrins, the BCOD-fused pyrroles with an acetoxymethyl group 6 and with two hydroxymethyl groups 9 were required. Starting from BCOD-fused pyrroles 3 [3a–d], the required derivatives were prepared as shown in Scheme 1. The Vilsmeier reaction of ester 3 gave formyl derivative 4 in a 92% yield [3c, 3d]. Reduction of the formyl group with NaBH4 afforded (hydroxymethyl)pyrrole 5a (98%),

which was then acetylated with Ac2O and pyridine to give (acetoxymethyl)pyrrole 6 in a 98% yield [3c, 3d]. An attempted reduction of 4b to bis(hydroxymethyl)pyrrole 9 with LiAlH4 in THF at room temperature gave ethyl (hydroxymethyl)pyrrolecarboxylate 5b in 82% yield similar to the reduction of t-butyl ester 4a with NaBH4.Under the THF-refl ux condition, excessive reduction of ester and formyl groups occurred to give the correspond-ing dimethylpyrrole in a 47% yield. Thus, the substrate was changed to diformylpyrrole 8, which was directly prepared by diformylation of α-free BCOD-fused pyr-role 7 with trimethyl orthoformate and TFA in a 78% yield [3b]. Reduction of diformylpyrrole 8 with NaBH4

provided bis(hydroxymethyl)pyrrole 9, quantitatively. This bis(hydroxymethyl)pyrrole 9 was, however, very sensitive towards an acid. This compound deteriorated

N

N N

N

ZnN N

NN

ZnNN

N N

Zn

NN

N N

ZnN N

NN

Zn

NN

N N

Zn

NN

N N

ZnN N

NN

Zn

N N

NN

Zn

N

N N

N

Zn

1a

2a

opp-1b

opp-2b

adj-1b

adj-2b

1c

2c

1d

2d

5β

γδ

5 5 5 5

15 15 15 15 15

21

22

23

24 7172

73

74

Fig. 1. BCOD-fused porphyrin 1a–d and benzoporphyrin 2a–d

NH

CO2R

CHO

NH

CO2R1

OR2

NH

CO2R

NH NH

CHO

NH

CHO OH

OH

i) ii) or iii)

iv)

vi)

81%

quant.

3a: R = t-Bu3b: R = Et

9

5a: R1 = t-Bu, R2 = H5b: R1 = Et, R2 = H6: R1 = t-Bu, R2 = Ac

7 8

92%98%

78%

v)

98%

4a: R = t-Bu4b: R = Et 92%

82%

iii)

Scheme 1. Reagents and conditions: i) DMF, POCl3, CH2Cl2, rt; NaOAc, H2O. ii) NaBH4, MeOH/THF, rt. iii) LiAlH4, THF, rt. iv) Ac2O, pyridine, rt. v) 3b, KOH, ethylene glycol, 175 °C. vi) CH(OMe)3, TFA, rt


124 H. UOYAMA ET AL.

rapidly even during the NMR measurement in CDCl3,which contained a trace of acid; therefore, this material was used without purifi cation.

Preparation of BCOD-fused porphyrins

The [3 + 1] method was applied for the preparation of singly BCOD-fused porphyrin 1a (Scheme 2). Con-densation of tripyrrane 10 [5] and BOCD-fused pyr-role 9 with BF3·OEt2 followed by neutralization with Et3N and oxidization with chloranil gave free-base BCOD-fused porphyrin, which could not be puri-fi ed or completely metalated with zinc, probably due to its low solubility. The zinc derivative 1a was obtained when oxidation with chloranil was carried out in the presence of zinc acetate. The porphyrin 1a was purifi ed by GPC, followed by recrystallization from

CHCl3 and MeOH. The yield of 1a based on tripyrrane 10 was 12%. Singly BCOD-fused porphyrin 1a was slightly soluble in common solvents such as CHCl3, CH2-

Cl2, THF, and pyridine.Preparation of adjacently di-BCOD-fused porphyrin

adj-1b was achieved by MacDonald [2 + 2] condensation (Scheme 3). The required BCOD-fused dipyrromethane 12 was prepared according to the literature: condensa-tion of BCOD-fused pyrrole 3b and methylal with the aid of a catalytic amount of H2SO4 (92%), followed by removal of the ester moieties (92%) [6]. The reaction of 12 with DMF and POCl3 gave diformyl derivative 14 in a 67% yield [7]. This dicarbaldehyde could be obtained via dipyrromethanedicarboxylic acid 13, which was in turn obtained by saponifi cation of diester 11 with lithium hydroxide (70%). Decarboxylative formylation of dicar-boxylic acid 13 with CH(OMe)3 in TFA provided 14 in a 63% yield. The reaction of diformyldipyrromethane 15 [8] and BCOD-fused dipyrromethane 12 with TFA, followed by oxidation with DDQ and treatment with Zn(OAc)2·2H2O, gave the target porphyrin adj-1b in only a 3% yield. Another combination of the starting materials gave a better result. The reaction of BCOD-fused diform-yldipyrromethane 14 and diformyldipyrromethane 16gave porphyrin adj-1b in a 10% yield. This result can be understood by the steric bulkiness and electron-donating nature of the BCOD moieties.

NN

N N

ZnHN

NH HNHN

101a

+

9

12%

HO OH

N N

NN

ZnNH HN

HN

t-BuO2C

t-BuO2C

i)

17

NH

CO2t-Bu

OAc

ii)

16%

opp-1b

6

Scheme 2. Reagents and conditions: BF3·OEt2; chloranil; Zn(OAc)2·2H2O

N N

NN

Zn

HNNH

HNNH

R R

11: R = CO2Et12: R = H13: R = CO2H14: R = CHO

RR

15: R = H16: R = CHO

adj-1b: 3% from 12 and 16 10% from 14 and 15

i) 92%

ii) 67%

iii) 70%

iv) 63%

ii) 46%

v)

Scheme 3. Reagents and conditions: i) KOH, ethylene glycol, 180 °C. ii) DMF, POCl3, CH2Cl2; NaOAc, H2O. iii) LiOH·H2O, THF/MeOH, H2O, refl ux. iv) CH(OMe)3, TFA, rt. v) TFA, CH2Cl2; DDQ; Zn(OAc)2·2H2O

Scheme 4. Reagents and conditions: i) pyrrole, montmorillonite K-10, CH2Cl2. ii) pyrrole-2,5-dicarbaldehyde, TFA; DDQ; Zn(OAc)2·2H2O



First, we attempted to prepare oppositely di-BCOD-fused porphyrin opp-1b by the [2 + 2] method with α-free pyrrole and bis(hydroxymethyl)pyrrole. Thus, the reactions of 2,5-bis(hydroxymethyl)pyrrole [5] and α-free BCOD-fused pyrrole 7, and pyrrole and 2,5-bis(hydroxymethyl) BCOD-fused pyrrole 9 were con-ducted. However, both reactions gave intractable mixtures. Therefore, the [3 + 1] condensation approach was exam-ined (Scheme 4). Condensation of pyrrole and acetoxym-ethylpyrrole 6 with montmorillonite K-10 as a mild acid catalyst gave tripyrrane 17, which was used in the next step without purifi cation. The ester groups of tripyrrane 17 were fi rst removed by treatment with TFA and then reacted with pyrrole-2,5-dicarbaldehyde, which was prepared by the oxidation of 2,5-bis(hydroxymethyl)pyr-role with MnO2. Oppositely di-BCOD-fused porphyrin opp-1b was obtained in a 16% yield.

Triply BCOD-fused porphyrin 1c was prepared by the [3 + 1] method (Scheme 5). Tripyrrane 18 was prepared by the reaction of BCOD-fused pyrroles 6 and 7 accord-ing to the procedure reported in the literature [3c, 3d], and was condensed with pyrrole-1,5-dicarbaldehyde to give 1c in a 14% yield from 6.

Thermal conversion of BCOD-fused porphyrins to benzoporphyrins

The BCOD-fused porphyrins readily included some solvent molecules in their crystals [3e, 3f]. In fact, X-ray analysis of these compounds revealed the existence of solvent molecules in the single crystal of adj-1b (videinfra). In order to know the molecular composition of the obtained BCOD-fused porphyrins, elemental analyses were carried out using powdery samples of BCOD-fused porphyrins, which were obtained by recrystalliza-tion from CHCl3/MeOH followed by evacuation (ca 1mmHg) overnight. In all cases, water (1a) or chloroform (adj-1b, opp-1b, and 1c) was involved in the samples. Thermogravimetric (TG) analyses were performed using these samples and the results are shown in Table 1. The weight loss started from 80 (1c), 127 (opp-1b), 131 (adj-1b), and 135 (1a) °C, and stopped at similar tempera-tures (182–197 °C). There is no obvious rate change of the weight loss in all cases. Since the starting temperature was almost the same in the BCOD-fused porphyrins [3e], the difference should be due to tightness of the crystal packing of the samples; in fact, the larger the difference

Scheme 5. Reagents and conditions: i) 7, montmorillonite K-10, CH2Cl2; ii) pyrrole-2,5-dicarbaldehyde, TFA; DDQ; Zn(OAc)2·2H2O

NH HN

HN

CO2t-Bu

t-BuO2C

N

NN

N

Zni)

NH

CO2t-Bu

OAc14%

ii)

1c

6

18

Table 1. Elemental analyses and TG experiments of BCOD-fused porphyrin

Elemental analysisa, % TG, oC Weight loss, %

BCOD-fusedporphyrin

C H NComposition calcd.

from EAStart Stop Obsd. Calcd.a

1a 67.79 4.15 12.03 C26H18N4Zn +1/4H2O 135 182 6.8 7.1(6.2)

(67.76) (4.16) (12.16)

adj-1b 70.60 4.67 10.27 C32H24N4Zn +1/8CHCl3 131 188 13.4 13.0(10.6)

(70.81) (4.46) (10.28)

opp-1b 68.92 4.67 9.85 C32H24N4Zn +1/4CHCl3 127 197 13.8 15.4(10.6)

(69.19) (4.37) (10.01)

1c 71.78 5.10 8.61 C38H32N4Zn +1/4CHCl3 80 186 14.3 18.9(13.8)

(72.02) (4.78) (8.78)

a The elemental percentage values are based on the molecular composition. The value in parentheses is the calculated value includ-ing solventsb The calculated weight loss is based on the calculated molecular composition. The value in parentheses is based on the molecularformula of the porphyrin



between the observed and calculated weight losses, the lower the starting temperatures. This means that the sol-vent molecules involved in the samples were partially lost during storage after the elemental analysis. The bulk conversion of BCOD-fused porphyrins 1a, adj-1b, opp-1b, and 1c to the corresponding benzoporphyrins 2a–cwas undertaken using a glass tube oven at 200 °C for 1 h under reduced pressure. The obtained materials of 2a–cgave satisfactory spectroscopic data as well as elemental analyses.

UV-vis spectra of BCOD- and benzene-fused porphyrins

Zinc BCOD-fused porphyrins 1a–d were readily solu-ble in common solvents such as chloroform (Fig. 2), while zinc benzoporphyrins 2a–d were only soluble in solvents with a coordinating ability such as pyridine (Fig. 3). The

absorption bands of zinc BCOD-fused porphyrins 1a–dwere observed at almost the same positions and closely resembled those reported for the β-octaalkyl-substituted porphyrins [9]. The strength of the bands was notewor-thy. As the number of BCOD-ring fusion increased, Qx-bands with the lowest energy became stronger.

From Fig. 3, the Q-band I absorptions with the lowest energies were bathochromically shifted from 586 (log10

ε = 4.15, 2a) to 592 (4.62, adj-2b), 612 (4.58, opp-2b), 617 (4.76, 2c), and 627 (5.00, 2d) nm; the intensi-ties also increased in the similar order. The Soret-band absorptions with lower energies were red-shifted from 416 (log10 ε = 5.63, 2a) to 422 (5.70, adj-2b), 428 (5.60, opp-2b), 433 (5.52, 2c), and 433 (5.49, 2d) nm. In the cases of 2a, adj-2b, and 2d, only one strong Soret band was observed, while opp-2b and 2c showed obvious splitting in the strong Soret bands. In the case of opp-dibenzoporphyrin, the similar splitting was predicted for

the opp-dibenzoporphyrin(Ni) [2e]. No splitting in the Soret band region was well in accord with the theoretical calculation in the case of TBPZn 2d [10]. UV-vis spectra of substituted monobenzo- [2b] and adj-dibenzo-porphyrins [2e] showed no splitting in the Soret region, similar to the cases of 2a and adj-2b.

X-ray analyses of BCOD- and benzene-fused porphyrins

As electronic spectra of porphyrinoids were greatly affected by the structures, we carried out X-ray analysis of the porphyrins. BCOD-fused porphyrins 1 and benzoporphyrins 2 were dis-solved in chloroform or pyridine and the solutions were placed in methanol vapor. Good single crys-tals of 1a, adj-1b·1/2CHCl3, opp-1b, 2a·C5H5N,adj-2b·C5H5N, and opp-2b·C5H5N·1/4MeOH were obtained. The crystallographic results are summa-rized in Table 2.

The samples of doubly BCOD-fused porphy-rins adj- and opp-1b used for the X-ray experi-ment were diastereomeric mixtures. Therefore, the ethylene and ethynylene bridges of bicyclo[2.2.2]octadiene moieties were treated as disordered structures, and the populations were calculated. In the case of 1a, the ethylene and ethynylene bridges showed no disorder in the crystal and the molecules stacked between special positions (-1 symmetry). In the cases of adj-2b·C5H5N and opp-2b·C5H5N·1/4MeOH, two and four crystallo-graphically independent molecules were found in asymmetric unit cells, respectively. In the case of opp-1b, the porphyrin molecules occupied special positions with symmetry of -1. Therefore, the out-of-plane distortion of the porphyrin molecule in this crystal must be waved. The solvent chloroform molecules in adj-1b·1/2CHCl3 occupied special

Fig. 2. UV-vis spectra of BCOD-fused porphyrins 1a (bold line), adj-1b(solid line), opp-1b (bold dotted line), 1c (dotted line), and 1d (dotted broken line) in chloroform

Fig. 3. UV-vis spectra of benzoporphyrins 2a (bold line), adj-2b (solid line), opp-2b (bold dotted line), 2c (dotted line), and 2d (dotted broken line) in pyridine



positions (-1) in a disordered fashion. As expected from the previous results [3e], the porphyrin rings were almost fl at in the cases of the BCOD-fused porphyrins. Porphy-rins 1a and opp-1b showed a slightly waved, out-of-plane distortion and adj-1b showed a slightly ruffl ed, out-of-plane distortion (Fig. 4).

On the other hand, benzoporphyrin molecules tended to adopt out-of-plane distortions [3]. In all crystals, por-phyrin molecules occupied normal positions. Therefore, there was no crystallographical requirement for the por-phyrin structures. In the porphyrin structures of adj-2b,domed, out-of-plane distortion was observed, while all porphyrin molecules of opp-2b showed very slightly sad-dled, out-of-plane distortion and mono-benzoporphyrin 2a showed waved out-of-plane distortion (Fig. 5). In all cases, the center zinc atom deviated to the coordinated pyridine. In the porphyrin structures of adj-2b, the direc-tions of the porphyrin doming and zinc-atom deviation were the same, while the directions of the benzene tilting and zinc-atom deviation were the same in opp-2b. These data, as well as the bond lengths of the benzene rings, are shown in Tables 3 and 4.

A macrocyclic aromatic π-system containing local aromatic π-systems is very interesting in terms of bond alteration in their crystal structures. We previously showed that the bond alteration was more distinctive in the center benzene moieties than in the edge benzene moieties of the tribenzoporphyrin derivatives [3e]. From Table 2, no bond alteration in the benzene ring of 2a was observed, while obvious bond alteration was observed in opp-2b. The Cγ–Cδ bonds [1.373(5)–1.388(5) Å] were clearly shorter than the Cδ–Cδ bonds [1.399(5)–1.406(5) Å]. The benzene ring of adj-2b showed slight bond alter-ation: Cγ–Cδ, 1.376(10)–1.394(9) Å; Cδ–Cδ, 1.368(9)–1.404(9) Å. This data showed that the porphyrinic mac-rocyclic ring current affected the aromaticity of benzene rings differently in opp-2b, adj-2b, and 2a.

The crystal packing of benzoporphyrins is worth men-tioning. The crystal system of 2a was tetragonal and the space group was chiral P41212 (#92). Two porphyrin mol-ecules stacked on the face opposite to the coordinated pyridine molecules by the diagonal special axes (x, -x,1/4), (x, -x, 3/4), (x, x, 0), and (x, x, 1/2) with a symme-try of 2 (Fig. 6). In this stacking porphyrin pair, the zinc

Table 2. Crystallographic summary

1aadj-1b ·

1/2CHCl3

opp-1b 2a·C5H

5N adj-2b·C

5H

5N

opp-2b

·C5H

5N·1/4MeOH

Crystal formula C26

H18

N4Zn C

65H

49Cl

3N

8Zn

2C

32H

24N

4Zn C

29H

19N

5Zn C

33H

21N

5Zn C

133H

88N

20OZn

4

Space group P-1 P2/n P-1 P412

12 P2

12

12

1P-1

a[Å] 7.1474(14) 17.121(6) 6.1069(5) 17.152(8) 8.408(3) 16.392(3)

b[Å] 10.232(2) 6.781(2) 8.8903(10) 17.152(8) 20.278(7) 17.824(3)

c[Å] 13.046(2) 21.695(8) 11.2271(13) 14.813(7) 27.949(10) 19.901(3)

[°] 95.703(3) 90 110.898(5) 90 90 64.300(6)

[°] 93.551(4) 101.695(4) 91.174(4) 90 90 72.483(7)

[°] 101.836(4) 90 91.527(4) 90 90 84.800(8)

V[Å3] 925.8(3) 2466.6(15) 568.95(10) 4358(4) 4765(3) 4991.1(15)

Z 2 2 1 8 8 2

[mm-1] 1.35 1.191 1.111 1.157 1.066 1.02

unique refl ns 4148 5617 2593 5001 10606 39855

Requiv

0.0459 0.0765 0.0245 0.0463 0.0713 0.0292

Obsd. -refl ns 2869 3717 2443 4622 8515 22567

Parameters 281 371 170 317 705 1449

R1[I > 2 (I)] 0.0535 0.0744 0.0329 0.0595 0.078 0.0554

wR2(All) 0.157 0.0806 0.0806 0.1539 0.1561 0.1518

GOF 1.087 1.063 1.053 1.157 1.113 1.058

T[K] 150 100 100 150 150 150

CCDC No. 698126 698124 698128 698127 698125 698129



Table 3. Bond lengths of the benzene moieties in 2a and adj-2b

2a adj-2b

Distortion Waved Domed Domed

Zn deviationa 0.334(1) 0.4336(14) 0.4645(15)

[0.294(2)] [0.292(3)] [0.311(3)]

Dihedral angleb 6.1(1)° 10.2(2)° 12.0(2)° 13.2(2)° 10.6(2)°

Cβ–Cβ 1.388(6) 1.417(8) 1.410(9) 1.380(9) 1.381(8)

Cβ–Cγ1.382(6) 1.375(8) 1.404(8) 1.401(9) 1.392(8)

1.390(6) 1.392(8) 1.392(9) 1.425(9) 1.417(8)

Cβ–Cδ1.394(6) 1.380(8) 1.379(9) 1.376(10) 1.390(9)

1.380(7) 1.380(9) 1.394(9) 1.360(10) 1.375(9)

Cδ–Cδ 1.390(8) 1.404(9) 1.368(9) 1.402(10) 1.403(9)a Zinc atom deviations from the mean plane of porphyrin 24 atoms. Numeral in brackets is the deviation from the mean plane of four nitrogen atomsb Dihedral angle between the benzene moiety and the porphyrin mean plane calculated by 24 atoms. The negative value means that the plane tilts toward the zinc atom

Fig. 4. Ortep drawing of BCOD-fused porphyrins 1a, adj-1b, and opp-1b. Upper: top views vertically to the porphyrin planes. Lower: side views along from the meso-C5 to meso-C15. Hydrogen atoms and the chloroform solvent molecule in adj-1b·1/2CHCl3

were omitted for clarity

Fig. 5. Ortep drawing of independent molecules of a) 2a·C5H5N, b) adj-2b·C5H5N and c) opp-2b·C5H5N·1/4MeOH. Hydrogen atoms are omitted for clarity



atom was located above the benzene center of the partner molecule. The coordinated pyridine molecules occupied positions around the axes (0, 0.5, z) and (0.5, 0, z) of 41

symmetry.The space group of adj-2b·C5H5N was P212121. Two

independent porphyrin molecules were stacked on the opposite faces of the coordinated pyridines (Fig. 7). In

this stacking porphyrin pair, the zinc atom was located above one of the benzene peri-positional carbons of the partner molecule. The coordinated pyridine molecules occupied positions around the axes (x, 0, 1/4) and (x, 1/2, 3/4) of 21 symmetry.

The space group of opp-2b·C5H5N·MeOH was P-1. In the crystal structure of opp-2b, four independent

Table 4. Bond lengths of the benzene moieties in opp-2b

opp-2b

Distortion Domed Domed Domed Domed

Zn deviationa 0.2878(7) 0.2423(8) 0.2655(8) 0.2969(8)[0.2710(14)] [0.2425(14)] [0.2537(14)] [0.2702(15)]

Dihedral angleb -1.75(8)° -7.45(9)° -3.51(9)° -6.73(2)° -1.96(9)° -5.02(9)° -3.85(9)° -6.69(9)°

Cβ–Cβ 1.408(4) 1.405(4) 1.406(4) 1.406(4) 1.400(4) 1.406(4) 1.403(4) 1.403(4)

Cβ–Cδ1.401(4) 1.402(4) 1.400(4) 1.401(4) 1.390(4) 1.406(5) 1.398(4) 1.394(4)

1.388(4) 1.400(4) 1.382(4) 1.397(4) 1.397(5) 1.403(5) 1.396(4) 1.398(5)

Cγ–Cγ1.378(4) 1.381(5) 1.373(5) 1.373(5) 1.384(5) 1.375(5) 1.384(5) 1.376(5)

1.388(4) 1.378(5) 1.387(5) 1.376(5) 1.387(5) 1.380(5) 1.380(5) 1.388(5)

Cγ–Cγ 1.402(4) 1.399(5) 1.406(5) 1.406(5) 1.401(5) 1.399(5) 1.401(5) 1.399(5)a Zinc atom deviations from the mean plane of porphyrin 24 atoms. Numeral in brackets is the deviation from the mean plane of four nitrogen atomsb Dihedral angle between the benzene moiety and the porphyrin mean plane calculated by 24 atoms. The negative value means that the plane tilts toward the zinc atom

Fig. 6. Crystal packing views of 2a·C5H5N: a) along c, b) along a-b and c) along a + b axes

Fig. 7. a) Stacking view of the independent molecular pair of adj-2b·C5H5N perpendicularly to the porphyrin planes and b) the crystal packing view of adj-2b·C5H5N along a-axis



molecules were found in an asymmetric cell and each molecule formed a dimeric stacking pair by the symmet-ric operation (-1). The coordinated pyridine molecules were aligned as a windmill (Fig. 8a). This unit contained one heavily disordered molecule of methanol. In the unit cell, this tetramer motif faced each other (Fig. 8b).

EXPERIMENTAL

General

Melting points were measured on a Yanagimoto micromelting point apparatus and are uncorrected. NMR spectra were obtained with a JEOL AL-400 or EX-400 spectrometer at ambient temperature. IR spectra were measured with a Horiba FT-720 infrared spectrophotometer. EI and FAB spectra were measured with a JEOL JMS-700. The MALDI-TOF MS spec-trum was measured with a Voyager DE Pro instrument. Elemental analyses were performed with a Yanaco MT-5 elemental analyzer. UV-vis spectra were measured with a HITACHI U-2810 spectrophotometer. All solvents and chemicals were reagent-grade quality, obtained com-mercially and used without further purifi cation, except as noted. Dry dichloromethane and THF were purchased from Kanto Chemical Co. Toluene, triethylamine, and pyridine were distilled from calcium hydride and then stored on the appropriate molecular sieves, 4A. Solvents for chromatography were purifi ed by distillation. For spectral measurements, spectral grades of pyridine and chloroform were purchased from Nacalai Tesque Co. Thin-layer (TLC) and column chromatography were per-formed on Art. 5554 (Merck KGaA) and Silica Gel 60N (Kanto Chemical Co.), respectively. t-butyl 4,7-dihydro-4,7-ethano-2H-isoindole-1-carboxylate (3a) [3d], ethyl 4,7-dihydro-4,7-ethano-2H-isoindole-1-carboxylate (3b)[3c, 3d], t-butyl 3-formyl-4,7-dihydro-4,7-ethano-2H-

isoindole-1-carboxylate (4a) [3d], t-butyl 4,7-dihydro-3-hydroxymethyl-4,7-ethano-2H-isoindole-1-carboxylate (5a) [3d], t-butyl 3-acetoxymethyl-4,7-dihydro-4,7-eth-ano-2H-isoindole-1-carboxylate (6) [3d], 4,7-dihydro-4,7-ethano-2H-isoindole (7) [3b], 4,7-dihydro-4,7- ethano-2H-isoindole-1,3-dicarbaldehyde (8) [3b], 5,10,15,17-tetrahydrotripyrrin (10) [5], diethyl 21,24,5,71,74,10-hexahydro-21,24;71,74-diethanodibenzo[b,g]dipyrrin-1,9-dicarboxylate (11) [6], 21,24,5,71,74,10-hexahydro-21,24;71,74-diethanodibenzo[b,g]dipyrrin (12) [6], and 5,10-dihydrodipyrrin (15) [11] were prepared according to the literature procedures. Other commercially available materials were used without further purifi cation.

Ethyl 4,7-dihydro-3-formyl-4,7-ethano-2H-isoin-dole-1-carboxylate (4b). To a solution of DMF (0.23 mL, 3.0 mmol) in dry CH2Cl2 (5.0 mL) was slowly added POCl3 (0.27 mL, 3.0 mmol) at 0 °C under N2. The mix-ture was stirred for 15 min at the same temperature and added to a solution of ethyl 4,7-ethano-4,7-dihydro-2H-isoindole-1-carboxylate (3b; 543 mg, 2.50 mmol) in CH2

Cl2 (5.0 mL). The mixture was warmed to room tempera-ture and stirred for 2 h. After being cooled to 0 °C, the reaction was quenched with 10% aqueous NaOAc. The mixture was extracted with CH2Cl2. The organic extract was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by col-umn chromatography on silica gel (EtOAc/hexane: 30%) to give 566 mg (2.31 mmol, 92%) of the title compound as colorless crystals, mp 111–113 °C; Rf = 0.35 (30% EtOAc/hexane). 1H NMR (400 MHz, CDCl3; Me4Si): δH,ppm 1.40–1.62 (7H, m), 4.28–4.40 (4H, m), 6.52 (2H, m), 9.12 (1H, br s), 9.74 (1H, s). 13C NMR (100 MHz, CDCl3;Me4Si): δC, ppm 14.34, 26.00, 26.38, 32.51, 33.26, 60.93, 119.24, 125.07, 134.54, 135.83, 136.69, 140.86, 160.92, 178.41. IR (KBr): ν, cm-1 3284, 1697, 1678, 1223, 1142, 1101. MS (EI): m/z (%) 245 ([M]+, 7), 217 (100), 171 (84), 115 (29). Anal. calcd. for C14H15NO3: C, 68.56; H, 6.16; N, 5.71. Found: C, 68.34; H, 6.17; N, 5.72.

Fig. 8. Crystal packing views of opp-2b·C5H5N·1/4MeOH: a) along c-axis and b) perpendicularly to b plane. Disordered methanol molecules are omitted for clarity



Ethyl 4,7-dihydro-3-hydroxymethyl-4,7-ethano-2-H-isoindole-1-carboxylate (5b). To a stirred solution of formylpyrrole 4b (123 mg, 0.500 mmol) in THF (5.0 mL) was added a 1.0 M solution of LiAlH4 in THF (1.0 mL, 1.0 mmol) at 0 °C under N2. The mixture was warmed to room temperature and stirred for 2 h. The reaction was quenched with water and the suspension was fi ltered through a Celite pad. The fi ltrate was extracted with EtOAc. The organic extract was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by column chromatography on silica gel (EtOAc/hexane: 50%) to give 101 mg (0.410 mmol, 82%) of the title compound as pale brown crystals, mp 94–96 °C; Rf = 0.3 (EtOAc/hexane: 50%). 1H NMR (400 MHz, CDCl3; Me4Si): δH, ppm 1.21–1.52 (4H, m), 1.33 (3H, t, J = 7.3 Hz), 1.72 (1H, br s), 3.77 (1H, m), 3.83 (1H, m), 4.25 (2H, q, J = 7.3 Hz), 4.58 (2H, m), 6.44 (2H, m), 9.63 (1H, m). 13C NMR (100 MHz, CDCl3; Me4Si):δC, ppm 14.25, 26.16, 26.75, 32.17, 33.69, 55.67, 60.01, 112.91, 126.91, 128.65, 135.02, 135.91, 137.03, 162.60. IR (KBr): ν, cm-1 3384, 3291, 1695, 1221, 1142, 1088. MS (EI): m/z (%). 247 ([M]+, 14), 219 (100), 202 (13), 173 (16), 144 (35), 117 (29). Anal. calcd. for C14H17NO3·H2O:C, 63.38; H, 7.22; N, 5.28. Found: C, 63.65; H, 6.84; N, 5.38.

Attempted reduction of 4b to bis(hydroxymethyl)pyrrole 9. To a stirred solution of 4b (123 mg, 0.500 mmol) in THF (5.0 mL) was added a 1.0 M solution of LiAlH4 in THF (1.0 mL, 1.0 mmol) at 0 °C under N2.The mixture was refl uxed for 3 h and then cooled to 0 °C. The reaction was quenched with water and the resulting suspension was fi ltered through a Celite pad. The fi ltrate was extracted with EtOAc. The organic extract was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by col-umn chromatography on silica gel (10% EtOAc/hexane) to give 41 mg (0.23 mmol, 47%) of 1,3-dimethyl-4,7-dihydro-4,7-ethano-2H-isoindole as a colorless powder, mp 152–154 °C; Rf = 0.4 (10% EtOAc/hexane). 1H NMR (400 MHz, CDCl3; Me4Si): δH, ppm 1.52 (4H, m), 2.15 (6H, s), 3.69 (2H, m), 6.46 (2H, m), 6.88 (1H, br s). 13CNMR (100 MHz, CDCl3; Me4Si): δC, ppm 10.97, 27.36, 32.35, 114.72, 124.94, 136.06. IR (KBr): ν, cm-1 3348, 1601, 1435, 1415, 1379, 1292, 1149. MS (EI): m/z (%). 173 ([M]+, 33), 145 (100), 103 (8), 77 (7). Anal. calcd. for C12H15N·1/5H2O: C, 81.49; H, 8.78; N, 7.92. Found: C, 81.54; H, 8.45; N, 7.93.

1,3-bis(hydroxymethyl)-4,7-dihydro-4,7-ethano-2-H-isoindole (9). To a stirred solution of pyrroledicarb-aldehyde 8 (161 mg, 0.800 mmol) in THF (8.0 mL) and MeOH (0.8 mL) was added NaBH4 (38 mg, 1.0 mmol) at 0 °C under N2. The mixture was warmed to room temper-ature and stirred for 3 h. The mixture was quenched with water and extracted with EtOAc. The organic extract was washed with water and brine, dried over Na2SO4, and con-centrated in vacuo to give 163 mg (0.794 mmol, 99%) of the crude title compound, which was very labile towards

even a trace amount of an acid. Therefore, this material was used without purifi cation. 9: a brown foam. 1H NMR (acetone-d6): δ, ppm 1.36–1.48 (4H, m), 3.60 (2H, br s), 3.77 (2H, m), 4.45 (4H, s), 6.37 (2H, m), 8.84 (1H, br s). 13C NMR (100 MHz, acetone-d6; Me4Si): δC, ppm 28.10, 33.00, 55.92, 122.00, 126.81, 136.62. IR (KBr): ν, cm-1

3398, 1698, 1624, 1340, 1340, 1068.2 1 , 2 4 , 5 , 7 1 , 7 4 , 1 0 - hexahydro- 2 1 , 2 4 ; 7 1 , 7 4 -

diethanodibenzo[b,g]dipyrrin-1,9-dicarbaldehyde (14) [7]. A solution of diethyl 21,24,5,71,74,10-hexa-hydro-21,24;71,74-diethanodibenzo[b,g]dipyrrin-1,9-dicarboxylate [6] (11; 1:1 diastereomeric mixture; 223 mg, 0.500 mmol) and LiOH·H2O (611 mg, 14.5 mmol) in THF (10 mL), EtOH (10 mL), and H2O (10 mL) was refl uxed for 16 h. After the mixture was cooled to room temperature, aqueous 1.0 M HCl (14.5 mL, 14.5 mmol) was added. The mixture was extracted with EtOAc. The organic layer was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by column chromatography on silica gel (EtOAc/hexane) to give 137 mg (70%) of 21,24,5,71,74,10-hexahydro-21,24;71,74-diethanodibenzo[b,g]dipyrrin-1,9-dicarboxylic acid (13) as a diastereomeric mixture: col-orless crystals, mp. 152–154 °C; Rf = 0.40 (70% EtOAc/hexane); 1H NMR (400 MHz, acetone-d6; Me4Si; diaste-reomer mixture): δH, ppm 131–1.63 (both isomers; 8H, m), 3.78–4.04 (both isomers; 4H, m), 4.37 (both isomers; 2H, m), 6.48 (both isomers; 4H, m), 10.25 (both isomers; 2H, br s), 10.67 (both isomers; 2H, br s). 13C NMR (100 MHz, acetone-d6; Me4Si; diastereomer mixture; typi-cal signals): δC, ppm 23.31, 23.42, 27.02, 27.05, 27.56, 27.59, 32.98, 33.08, 34.67, 112.82, 112.60, 124.03, 124.17, 128.27, 128.35, 135.29, 135.91, 137.45, 161.98. IR (KBr): ν, cm-1 3321, 1670, 1511, 1461, 1253, 1223, 1146, 1087. MS (FAB): m/z 391 [M+ + 1]. Anal. calcd. for C23H22N2O4 + H2O: C, 67.63; H, 5.92; N, 6.68. Found: C, 67.67; H, 5.63; N, 6.82.

Dicarboxylic acid, 13 (117 mg 0.300 mmol) was treat-ed with TFA (0.60 mL) for 5 min at 0 °C under N2 in the dark. To the mixture was added trimethyl orthoformate (0.60 mL) at 0 °C, after which it was stirred for 3 h at room temperature. The reaction was quenched with 1.0 M aqueous NaOH at 0 °C. The mixture was extracted with EtOAc. The organic extract was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by column chromatography on silica gel (30% EtOAc/hexane) to give 68 mg (63%) of the title compound as colorless crystals: Rf = 0.30 (50% EtOAc/hexane:). 1H NMR (400 MHz, CDCl3; Me4Si; diastereomer mixture): δH, ppm 1.22–1.79 (both isomers; 8H, m), 3.85–4.00 (both isomers; 4H, m), 4.21 (both isomers; 2H, s), 6.49 (both isomers; 4H, m), 9.49 (one isomer; 2H, s), 9.49 (another isomer; 2H, s), 10.49 (one isomer; 2H, br s), 10.55 (another isomer; 2H, br s). 13CNMR (100 MHz, CDCl3; Me4Si; diastereomer mixture; typical signals): δC, ppm 23.08, 23.13, 26.45, 26.92, 26.98, 27.12, 32.44, 32.52, 32.66, 32.77, 32.96, 123.73,



123.76, 123.82, 128.99, 129.00, 129.12, 129.40, 129.43, 129.53, 133.87, 134.51, 134.59, 134.63, 136.05, 136.14, 136.23, 136.28, 136.37, 144.07. IR (KBr): ν, cm-1 3400, 1654, 1655, 1576, 1506, 1448, 1373, 1085. MS (EI): m/z(%). 358 ([M]+, 42), 330 (84), 302 (100), 273 (13), 243 (20), 158 (25), 130 (11), 117 (9). HR-MS (EI): m/z calcd. for C23H22N2O4: 358.1682. Found: 358.1682.

5,10-dihydrodipyrrin-1,9-dicarbaldehyde (16) [8]:To a stirred solution of DMF (2.3 mL, 30 mmol) in CH2Cl2 (10 mL), was added dropwise POCl3 (2.8 mL, 30 mmol) at 0 °C under N2. The mixture was stirred at room temperature for 15 min. The reaction mixture was added dropwise to a solution of 5,10-dihydrodipyrrin (15) [15] (1.46 g, 10.0 mmol) in 1,2-dichloroethane (20 mL) at 0 °C and the mixture was then refl uxed for 1.5 h. A 10% aqueous NaOAc solution was added at 0 °C, after which the mixture was refl uxed for 15 min and extracted with CH2Cl2. The organic extract was washed with water and brine, dried over Na2SO4, and concent-rated in vacuo. The residue was purifi ed by column chro-matography on silica gel (40% EtOAc/hexane) to give 0.930 g (46%) of the title compound as pale brown crys-tals: mp 219–220 °C; Rf = 0.3 (40% EtOAc/hexane). 1HNMR (400 MHz, CDCl3; Me4Si): δH, ppm 4.17 (2H, s), 6.14 (2H, d, J = 3.4 Hz), 6.90 (2H, d, J = 3.4 Hz), 9.42 (2H, s), 11.08 (2H, br s). 13C NMR (100 MHz, acetone-d6;Me4Si): δC, ppm 26.77, 110.51, 121.84, 133.91, 138.95, 178.84. IR (KBr): ν, cm-1 3234, 1633, 1491, 1419, 1354, 1307, 1186, 1036. MS (EI): m/z (%); 202 ([M]+, 100), 173 (26), 145 (28), 108 (11), 95 (4).

Pyrrole-2,5-dicarbaldehyde [12]. A suspension of bis(hydroxymethyl)pyrrole (0.634 g, 4.99 mmol) and MnO2 (13.2 g) in CH2Cl2 was stirred for 48 h at room temperature. The mixture was fi ltered and the fi ltrate was concentrated in vacuo. The residue was purifi ed by column chromatography on silica gel to give 0.283 g (46%) of the title compound as a colorless powder, mp 120–122 °C; Rf = 0.40 (30%, EtOAc/hexane). 1H NMR (400 MHz, CDCl3; Me4Si): δ, ppm 7.01 (2H, d, J = 2.4 Hz), 9.77 (s, 2H), 9.90 (br s, 1H). 13C NMR (100 MHz, CDCl3; Me4Si): δ, ppm 119.5, 135.8, 181.6. IR (KBr): ν,cm-1 3166, 1683, 1635, 1546, 1431, 1178, 802. MS (EI): m/z123 ([M]+,100), 94 (9), 66 (24). Anal. calcd. for C6H5NO2:C, 58.54; H, 4.09; N, 11.38. Found: C, 58.61; H, 4.04; N, 11.42.

21, 24-dihydro-21,24-ethanobenzo[b]porphyrinato zinc (1a). To a stirred solution of tripyrrane 10 [5] (180 mg, 0.800 mmol) and bis(hydroxymethyl)pyrrole 9 (164 mg, 0.800 mmol) in dry CH2Cl2 (160 mL) was slowly added BF3·OEt2 (20 μL, 0.016 mmol) at room tempera-ture under N2 in the dark. After the reaction mixture had been stirred for 24 h, the mixture was stirred with chlo-ranil (295 mg, 1.20 mmol) and Zn(OAc)2·2H2O (878 mg, 4.00 mmol) for 12 h. The mixture was washed with water and brine, dried over Na2SO4, and concentrated in vacuo.The residue was purifi ed successively by column chroma-tography on silica gel (CHCl3), by GPC chromatography,

and by recrystallization from CHCl3/MeOH to give 43.0 mg (12%) of the title porphyrin as a red powder: mp 146 °C (decomp.); Rf = 0.6 (CHCl3). UV-vis (CHCl3):λmax, nm (log ε) 398 (5.47), 528 (4.17), 560 (3.82). 1HNMR (400 MHz, THF-d8; Me4Si): δH, ppm 1.83 (2H, m), 2.15 (2H, m), 5.71 (2H, m), 7.08 (2H, m), 10.27 (2H, s), 10.33 (2H, s). 13C NMR (100 MHz, THF-d8; Me4Si):δC, ppm 27.93, 36.58, 100.64, 104.25, 131.06, 131.23, 131.53, 136.68, 143.25, 148.89, 149.04, 149.53. MS (MALDI-TOF): m/z 450 (calcd. for [M + H]+ 450). Anal. calcd. for C26H18N4Zn + 1/4H2O: C, 67.76; H, 4.16; N, 12.16. Found: C, 67.79; H, 4.15; N, 12.03.

2 1,2 4,7 1,7 4- te trahydro-2 1,2 4;7 1,7 4-d ie than-odibenzo[b,g]porphyrinato zinc (adj-1b). To a stirred solution of BCOD-fused dipyrromethane 12 (302 mg, 1.00 mmol) and diformyl dipyrromethane 16 (202 mg, 1.00 mmol) in CH2Cl2 (200 mL) was added TFA (5.0 mL) at room temperature under N2 in the dark. After the mix-ture had been stirred for 14 h at room temperature, the mixture was neutralized with triethylamine followed by treatment with DDQ (341 mg, 1.50 mmol) for 6 h and Zn(OAc)2·2H2O (1.10g, 5.00 mmol) for 8 h. The mixture was washed with water and brine, dried over Na2SO4,and concentrated in vacuo. The residue was purifi ed by column chromatography on silica gel (CHCl3) followed by recrystallization from CHCl3/MeOH to give 14.2 mg (3%) of the title porphyrin as a red powder, mp 150 °C (decomp.); Rf = 0.30 (CHCl3). UV-vis (CHCl3): λmax,nm (log ε) 399 (5.42), 526 (4.18), 560 (3.95). 1H NMR (400 MHz, CDCl3; Me4Si; diasteomer mixture): δH, ppm 1.98 (both isomers; 4H, m), 2.25 (both isomers; 4H, m), 5.78 (both isomers; 4H, m), 7.16 (both isomers; 4H, m), 10.35 (both isomers; 1H, s), 10.41 (both isomers; 2H s) 10.44 (one isomer; 1H, s), 10.45 (another isomer; 1H, s). 13C NMR (100 MHz, THF-d8; Me4Si; diasteomer mix-ture): δC, ppm 27.95, 28.00, 36.59, 36.61, 97.08, 97.11, 100.69, 100.70, 104.39, 130.68, 131.17, 136.66, 136.71, 136.79, 136.81, 142.68, 143.28, 143.31, 148.56, 149.13, 149.76, 150.16, 150.18. MS (FAB): m/z 529 [M+ + H]. Anal. calcd. for C32H24N4Zn + 1/8CHCl3: C, 70.81; H, 4.46; N, 10.28. Found: C, 70.60; H, 4.67; N, 10.27. The reaction of BCOD-fused diformyl dipyrromethane 14 (358 mg, 1.00 mmol) and pyrrolylmethane 15 (146 mg, 1.00 mmol) was performed under similar conditions to give 52.9 mg (10%) of the title compound.

21,24,121,122-tetrahydro-21,24;121,122-diethan-odibenzo[b,l]porphyrinato zinc (opp-1b). To a stirred solution of pyrrole (14.5 mg, 0.100 mmol) and ace-toxymethylpyrrole 6 (63.5 mg, 0.200 mmol) in CH2Cl2

(3 mL) was added montmorillonite K-10 (100 mg) at room temperature in the dark. The mixture was stirred for 14 h and fi ltered. The fi ltrate was successively washed with water, a saturated aqueous NaHCO3, and brine, dried over Na2SO4, and concentrated in vacuo to give crude 21,24,5,10,121,124,15,17-octahydro-21,24;121,124-diethanodibenzo[b,l]tripyrrin (17). This crude material was treated with trifl uoroacetic acid (0.5 mL) for 10 min



at room temperature under N2 in the dark. The mixture was diluted with CH2Cl2 (18 mL) and a solution of pyr-role-2,5-dicarbaldehyde (12.3 mg, 0.100 mmol) in CH2-

Cl2 (22 mL) was added. After 14 h, the reaction mixture was neutralized with triethylamine and stirred with DDQ (34.0 mg, 0.150 mmol) for 4 h. To the mixture was added Zn(OAc)2·2H2O (0.110 mg, 5.00 mmol), after which it was stirred overnight. The mixture was then washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The residue was purifi ed by column chromatog-raphy on silica gel (CHCl3), followed by recrystallization from CHCl3/MeOH to give 8.3 mg (16%) of the title porphyrin as a diastereomeric mixture: a red powder, mp 165 °C (decomp.); R

f = 0.30 (CHCl

3). UV-vis (CHCl

3):

max,nm (log ) 400 (5.46), 528 (4.22), 562 (4.00).

1H

NMR (400 MHz, CDCl3; Me

4Si; diasteomer mixture):

H, ppm 1.98 (both isomers; 4H, m), 2.24 (both isomers;

4H, m), 5.76 (both isomers; 4H, m), 7.16 (both isomers;

4H, m), 9.60 (one isomer; 2H, s), 10.44 (another isomer;

2H, s); 13

C NMR (100 MHz, CDCl3; Me

4Si; diasteomer

mixture; typical signals): C, ppm 27.93, 36.45, 101.53,

101.55, 131.57, 131.59, 136.76, 142.67, 148.85, 148.86,

150.27. MS (FAB): m/z 529 [M+ + H]; Anal. calcd. for

C32

H24

N4Zn + 1/4CHCl

3: C, 69.19; H, 4.37; N, 10.01.

Found: C, 68.92; H, 4.67; N, 9.85.

2 1 , 2 4 , 7 1 , 7 4 , 1 2 1 , 1 2 2 - hexahydro- 2 1 , 2 4 ; 7 1 ,74;121,122-triethanotribenzo[b,g,l]porphyrinato zinc (1c). To a stirred solution BCOD-fused pyrrole 7 (145 mg, 1.00 mmol) and pyrrole 6 (635 mg, 2.00 mmol) in CH2Cl2

(30 mL) was added montmorillonite K-10 (1.0 g) at room temperature in the dark. The mixture was stirred for 14 h and montmorillonite K-10 was removed by fi ltration. The fi ltrate was washed with water, saturated aqueous NaH-CO3, and brine, dried over Na2SO4, and concentrated in vacuo to give crude 21,24,5,71,74,10,121,124,15,17-decahydro-21,24;71,74;121,124-triethanotribenzo[b,g,l]tripyrrin (16) [3d]. This material was used without puri-fi cation. The residual material was treated with trifl uo-roacetic acid (5.0 mL) for 10 min at room temperature in the dark. The mixture was diluted with CH2Cl2 (180 mL), and a solution of pyrrole-2,5-dicarbaldehyde (123 mg, 1.00 mmol) in CH2Cl2 (20 mL) was added. After 23 h, the reaction mixture was neutralized with triethylamine and stirred with DDQ (340 g, 1.50 mmol) for 4 h. To the reaction, solution was added Zn(OAc)2·2H2O (1.10 g, 5.00 mmol), after which the mixture was stirred over-night. The mixture was washed with water and brine, dried over Na2SO4, and concentrated in vacuo. The resi-due was purifi ed by column chromatography on silica gel (CHCl3) followed by recrystallization from CHCl3/MeOH to give 87.6 mg (14%) of the title porphyrin as a mixture of diastereomers: mp 143 °C (decomp.); Rf = 0.30 (CHCl3). UV-vis (CHCl3): λmax, nm (log ε) 399 (5.48), 527 (4.24), 561 (4.11). 1H NMR (400 MHz, CDCl3;Me4Si; diasteomer mixture): δH, ppm 2.05 (all isomers; 6H, m), 2.30 (all isomers; 6H, m), 5.79 (one isomer; 6H, m), 5.87 (other isomers; 6H, m), 7.22 (all isomers; 6H, m),

9.55 (one isomer; 2H, m), 9.57 (other isomers; 2H, m), 10.38 (one isomer; 2H, s), 10.41 (other isomer; 2H, m), 10.46 (one isomer; 2H, s), 10.50 (other isomers; 2H, m). 13CNMR (100 MHz, CDCl3; Me4Si; diasteomer mixture, typi-cal signals): δC, ppm, 27.96, 27.99, 28.03, 36.50, 36.57, 97.91, 97.95, 97.98, 98.00, 101.55, 101.60, 101.63, 131.16, 131.22, 131.23, 136.89, 136.92, 136.95, 136.99, 137.00, 142.16, 142.71, 142.76, 142.78, 142.85, 148.39, 148.48, 149.80, 149.85, 149.88, 150.14, 150.19, 150.22, 150.29, 150.31, 150.34. MS (FAB): m/z 607 [M+ + H]. Anal. calcd. for C38H30N4Zn + 1/4CHCl3: C, 72.02; H, 4.78; N, 8.78. Found: C, 71.78; H, 5.10; N, 8.61.

General procedure for preparative retro-Diels-Alder reaction

The precursor BCOD-fused porphyrin in a sample tube was placed in a 25 mL round-bottomed fl ask. The fl ask was evacuated by a rotary vacuum pump and then heated at 200 °C for 1 h in a glass tube oven. After cool-ing, the targeted benzoporphyrin was obtained in a quan-titative yield.

Benzo[b]porphyrinato zinc (2a). Red powder, mp > 300 °C. UV-vis (pyridine): λmax, nm (log ε) 395 (4.63), 416 (5.63), 529 (sh; 3.87), 547 (4.24), 586 (4.15). 1HNMR (400 MHz, pyridine-d5; Me4Si): δH, ppm 8.16 (2H, m), 9.60–9.69 (8H, m), 10.45 (2H, s), 10.86 (2H, s). 13CNMR (100 MHz, pyridine-d5; Me4Si): δC, ppm 98.59, 106.47, 121.47, 127.52, 131.10, 131.47, 132.80, 140.34, 146.52, 148.62, 148.81, 150.82. MS (FAB): m/z 423 [M+

+ H]. Anal. calcd. for C26H18N4Zn + 1/2H2O: C, 66.69; H, 3.49; N, 12.95. Found: C, 66.67; H, 3.56; N, 13.02.

Dibenzo[b,g]porphyrinato zinc (adj-2b). Red pow-der, mp > 300 °C. UV-vis (pyridine): λmax, nm (log ε) 400 (4.72), 422 (5.70), 546 (sh; 4.11), 554 (4.31), 583 (sh; 4.41), 592 (4.62). 1H NMR (400 MHz, THF-d8; Me4Si):δH, ppm 8.13 (4H, m), 9.37 (2H, m), 9.41 (2H, m), 9.52 (2H, m), 9.64 (2H, m), 10.21 (1H, s), 10.54 (2H, s), 10.90 (1H, s). 13C NMR (100 MHz, THF-d8; Me4Si): δC, ppm 91.53, 99.01, 107.45, 120.87, 121.22, 126.78, 127.17, 129.22, 131.19, 139.61, 140.34, 144.82, 147.13, 147.44, 149.55. MS (FAB): m/z 473 [M+ + H]. Anal. calcd. for C26H18N4Zn + 1/4H2O: C, 70.30; H, 3.48; N, 11.71. Found: C, 70.12; H, 3.61; N, 11.39.

Dibenzo[b,l]porphyrinato zinc (opp-2b). Red pow-der, mp > 300 °C. UV-vis (pyridine): λmax, nm (log ε) 390 (4.39), 413 (5.29), 428 (5.60), 515 (sh; 3.54), 526 (sh; 3.85), 536 (3.96), 561 (4.24), 568 (4.38), 581 (sh; 3.77), 612 (4.58). 1H NMR (400 MHz, pyridine-d5; Me4Si): δH,ppm 6.68 (4H, m), 8.17 (4H, m), 8.24 (4H, s), 9.45 (4H, s). 13C NMR (100 MHz, pyrdine-d5; Me4Si): δC, ppm 101.26, 122.53, 128.38, 132.61, 141.16, 146.34, 150.53. MS (FAB): m/z 473 [M+ + H]. Anal. calcd. for C28H16N4Zn+ 1/2H2O: C, 69.65; H, 3.55; N, 11.60. Found: C, 69.85; H, 3.67; N, 11.58.

Tribenzo[b,g,l]porphyrinato zinc (2c). Red powder, mp > 300 °C. UV-vis (pyridine): λmax, nm (log ε) 401



(sh; 4.54), 423 (5.42), 433 (5.52), 512 (sh; 3.34), 545 (sh; 3.78), 553 (sh; 3.91), 564, (sh; 4.09), 578 (4.21), 597 (4.56), 617 (4.76). 1H NMR (400 MHz, THF-d8; Me4Si):δH, ppm 9.94 (4H, m), 9.96 (4H, m), 11.25 (2H, m), 11.35 (2H, m), 11.49 (4H, m), 12.43 (2H, s), 12.78 (2H, s). 13CNMR (100 MHz, THF-d8; Me4Si): δC, ppm 92.64, 100.69, 120.72, 121.09, 121.12, 126.34, 126.81, 126.91, 129.40, 139.13, 139.71, 140.03, 143.13, 145.94, 146.11, 147.72. MS (FAB): m/z 473 [M+ + H]. Anal. calcd. for C32H18N4Zn:C, 73.36; H, 3.46; N, 10.69. Found: C, 73.26; H, 3.77; N, 10.54.

X-ray analysis

BCOD-fused porphyrin 1 (ca 2 mg) was dissolved in chloroform (ca 5 mL) and the solution was divided equally between four sample tubes. One, two, three, and four drops of methanol, respectively, were added to the four sample tubes, which were then placed in a jar containing methanol and left for a few days. Similarly, crystal preparation of benzoporphyrin 2 was performed using pyridine and methanol. The suitable crystals were placed in Lindeman capillary tubes with a very small amount of the mother liquor, after which the capillary tubes were sealed using a candle fl ame. Determination of cell parameters and collection of refl ection intensi-ties were performed on Rigaku Mercury-8 (3-kW sealed tube) instrument equipped with graphite monochromated Mo Ka radiation or Rigaku Mercury-7 (18 kW rotating anode). The data were corrected for Lorentz, polariza-tion, and absorption effects. The structures were solved by direct methods (SIR-97 [13] or Shelxs-97 [14]) and expanded using the Fourier technique [15]. Hydrogen atoms were placed in calculated positions and refi ned by using riding models. All calculations were performed by using the CrystalStructure crystallographic software package [16] or WinGX [17]. Shelxl-97 [14] was used for structure refi nement. The structure data obtained were validated by the PLATON program [18]. These data can be obtained free of charge at www.ccdc.com.ac.uk.

CONCLUSION

Zinc BCOD-fused porphyrins with no peripheral substituent were prepared by 2 + 2 and 3 + 1 porphyrin syntheses and the thermal retro-Diels-Alder reactions gave zinc complexes of the parent benzoporphyrin, adj-dibenzoporphyrin, opp-dibenzoporphyrin, and triben-zoporphyrin with no substituent in quantitative yields. Splitting in the Soret band was observed in zinc opp- benzo- and tribenzo-porphyrins. Monobenzo- and adj-dibenzo-porphyrins showed a single Soret band and no bond alteration in the benzene rings of their crystal struc-tures was observed. Four independent molecules were found in the crystal structure of opp-dibenzoporphyrin.Obvious bond alteration in the benzene rings of all four molecules was observed. These fi ndings may suggest

the porphyrinic macrocyclic ring current, with its major contribution, expands effectively into the benzene rings of opp-dibenzoporphyrin, slightly into those of adj-dibenzoporphyrin, and scarcely into that of mono- benzoporphyrin.

Acknowledgements

This work was partially supported by Grants-in-Aid

for the Scientifi c Research C (20550047) from the Japa-

nese Ministry of Education, Culture, Sports, Science and

Technology. We appreciate the Institute for Molecular

Science for the permission of X-ray measurements using

Rigaku Mercury-7.

REFERENCES

a) Boa Z, Lovinger AJ and Dodabalapur A. 1. Adv. Mater. 1997; 9: 42–44. b) Zeis R, Seigrist T and Kloc CH. Appl. Phys. Lett. 2005; 86: 0221031–0221033. c) Aramaki S, Sakai Y and Ono N. Appl.Phys. Lett. 2004; 84: 2085–2087. d) Shea PB, Cao Y, Ono N and Kanicki J. Appl. Phys. Lett. 2005; 87: 1735061–1735063. e) Shea PB, Pattison LR, Kawano M, Kanicki J, Petroff P, Yamada H and Ono N. J. Appl. Phys. 2006; 100: 0345021–0345027. f) Shea PB, Pattison LR, Yamada H, Chen C, Petroff P, Ono N and Kanicki J. Appl. Phys. Lett. 2007; 90: 2331071–2331073. g) Shea PB, Pattison LR, Kawano M, Chen C, Chen J, Petroff P, Martin DC, Yamada H, Ono N and Kanicki J. Synth. Metal.2007; 157: 190–197.

a) Lash TD, Chandrasekar P, Osuma AT, Chaney ST and Spence JD. J. Org. Chem. 1998; 63: 8455–8469. b) May DAJ and Lash TD. J. Org. Chem. 1992; 57: 4820–4828. c) Silva AMG, Oliveira KT, Faustino MAF, Neves MGPMS, Tomé AC, Silva AMS, Cavaleiro JAS, Brndão P and Felix V. Eur. J. Org. 2008; 704–712. d) Inokuma Y, Ono N, Uno H, Kim DY, Noh SB, Kim D and Osuka A. Chem.Commun. 2005; 3782–3784. e) Gerhardt SA, Lewis JW, Zhang JZ, Bonnett R and McManus KA. Pho-tochem. Photobiol. Sci. 2003; 2: 934–938. f) Bon-nett R and McManus KA. J. Chem. Soc. Perkin. Trans. 1 1996; 2461–2466. g) Jiao L, Hao E, Fronc-zek FR, Vicente MGH and Smith KM. Chem. Com-mun. 2006; 3900–3902. h) Zeng Z, Liu C, Jin L-M, Guo C-C and Chen Q-Y. Eur. J. Org. Chem. 2005; 306. i) Liu C, Shen D-M, Zeng Z, Guo C-C and Chen Q-Y. J. Org. Chem. 2006; 71: 9772–9783. j) Uno H, Masuda G, Tukiji M, Nishioka Y and Iida T. Tetrahedron Lett. 2007; 48: 7512–7515.a) Ito S, Murashima T, Uno H and Ono N. 2. Chem.Comunn. 1998; 1661–1662. b) Ito S, Ochi N, Murashima T, Uno H and Ono N. Heterocycles 2000; 52: 399–411. c) Okujima T, Komobuchi N, Shimi-zu Y, Uno H and Ono N. Tetrahedron Lett. 2004;



45: 5461–5464. d) Okujima T, Komobuchi N, Uno H and Ono N. Heterocycles 2006; 67: 255–267. e) Uno H, Shimizu Y, Uoyama H, Tanaka Y, Okujima T and Ono N. Eur. J. Org. Chem. 2008; 87–98. f) Mack J, Bunya M, Shimizu Y, Uoyama H, Komobu-chi N, Okujima T, Uno H, Ito S, Stillman MJ, Ono N and Kobayashi N. Chem. Eur. J. 2008; 14: 5001–5020. g) Uno H, Nakamoto K, Kuroki K, Fujimoto A and Ono N. Chem. Eur. J. 2007; 13: 5773–5784. h) Ito S, Ochi N, Murashima T, Uno H and Ono N. Chem. Commun. 2002; 893–894. i) Uno H, Masu-moto A and Ono N. J. Am. Chem. Soc. 2003; 125:12082–12083.

a) Boundif A and Momenteau M. J. Chem. Soc. Chem. Commun. 1994; 2069–2070. b) Boundif A and Momenteau M. J. Chem. Soc. Perkin Trans 1.1996; 1235–1242. c) Sessler JL, Genge JW, Urbachi A and Sanson P. Synletters 1996; 187–188. d) Lash TD. Chem. Eur. J. 1996; 2: 1197–1200. e) Smith KM. In The Porphyrin Handbook, Kadish KM, Smith KM and Guilard R. (Eds.). Vol. 1, Academic Press: San Diego, 1999, pp. 2–44.Taniguchi S, Hasegawa H, Yanagiya S, Tabeta Y, 3.Nakano Y and Takahashi M. Tetrahedron 2001; 57:2103–2108.a) Yamada H, Kushibe K, Okujima T, Uno H, and 4.Ono N. Chem. Commun. 2006; 383–385. b) Yamada H, Kushibe K, Mitsuogi S, Okujima T, Uno H and Ono N. Tetrahedron Lett. 2008; 49: 4731–4733.Giraud-Roux M, Proni G, Nakanishi K and Berova 5.N. Heterocycles 2003; 61: 417–432.Taniguchi M, Balakumer A, Fan D, McDowell BE 6.and Lindsey JS. J. Porphyrins Phthalocyanines2005; 9: 554–574Broadhurst MJ, Grigg R and Johnson AW. 7. J. Chem. Soc. C 1971; 3681–3690.

McHugh AJ, Gouterman M and Weiss C Jr. 8. Theoret. Chim. Acta 1972; 24: 346–370.a) Fang Z and Breslow R. 9. Org. Lett. 2006; 8: 251–254. b) Singh K, Behal S and Hundal MS. Tetrahe-dron 2005; 61: 6614–6622.a) Bray BL, Hess P, Muchowski JM and Scheller 10.ME. Helv. Chim. Acta 1988; 71: 2053–2057. b) Mu-chowski JM and Hess P. Tetrahedron Lett. 1988; 29:777–780.SIR-97: A package for crystal structure solution 11.and refi nement, Istituto di Cristallografi a, Italy; Altomare A, Burla MC, Camalli M, Cascarano G, Giacovazzo C, Guagliardi A, Moliterni AGG, Polidori G and Spagna R. J. Appl. Cryst. 1999; 32:115–119.Shelxs-97 and Shelxl-97, Program for the refi ne-12.ment of crystal structures from diffraction data, University of Göttingen, Göttingen, Germany; Sheldrick G and Schneider T. Meth. Enzymol. 1997; 277: 319–343.DIRDIF-99 program system: Beurskens PT, Ad-13.miraal G, Beurskens G, Bosman WP, de Gelder R, Israel R and Smits JMM. Technical Report of the Crystallography Laboratory, University of Nijme-gen, The Netherlands, 1999.CrystalStructure Ver. 3.7.0, Crystal structure anal-14.ysis package, Rigaku (3–9-12 Akishima, Tokyo, Japan) and Rigaku/MSC (9009 New Trails Dr., The Woodlands, TX 77381 USA), 2000–2004.WinGX 1.4.0 suite for small-molecule single-crys-15.tal crystallography, University of Glasgow, Farrugia LJ. J. Appl. Cryst. 1999; 32: 837–838.PLATON, Spek AL (2003), A Multipurpose Crys-16.tallographic Tool, Utrecht University, Utrecht, The Netherlands; Spek AL. J. Appl. Cryst. 2003; 36:7–13.



Copyright © 2009 World Scientific Publishing Company

INTRODUCTION

The thermal or photochemical reaction of ethyl dia-zoacetate with arenes to give a mixture of isomeric cyclo-heptatrienes is a prototypical example of the Büchnerreaction, a process that has been known for over 100 years[1]. Copper-catalyzed processes were found to reduce thecomplexity of the Büchner product mixtures, and in the

early 1980s Rh2(OAc)4 and its analogues became the cat-alysts of choice for this reaction [2]. Rh2(OAc)4-catalyzedcyclopropanations of arenes, especially intramolecularversions, have remained widely used due to their highregio- and stereo-selectivity [3]. The related sequentialcyclopropanation–Cope rearrangement method has beenapplied fruitfully to a variety of synthetic challenges andincludes examples of Büchner-type cyclopropanations ofbenzenes, pyrroles, and furans [4].

The valence isomer, bicyclo[4.1.0]hepta-2,4-diene(norcaradiene, NCD), and its unusually rich rearrange-ment chemistry has been extensively studied for half acentury. In 1957, Woods observed the thermal valence

Addition of carbenes derived from aryldiazoacetatesto arenes using chloro(tetraphenylporphyrinato)ironas catalyst

Harun M. Mbuvi and L. Keith Woo*�

Department of Chemistry, Iowa State University, Ames, IA 50011, USA


ABSTRACT: Chloro(tetraphenylporphyrinato)iron, Fe(TPP)Cl, is an active catalyst for the Büchneraddition of para-substituted methyl 2-phenyldiazoacetates, 1a–d, to substituted benzenes. Yieldsgreater than 70% have been achieved at temperatures ranging from 60–100 °C. Reactions of substi-tuted methyl 2-phenyldiazoacetates with benzene gave rapidly equilibrating mixtures of norcaradiene-cycloheptatriene valence isomers, 2a–d/2′a–d, in yields over 70%. Treatment of chlorobenzene withmethyl 2-phenyldiazoacetate produced a regio-isomeric mixture of 7-carbomethoxy-2-chloro-7-phenylnorcaradiene/7-carbomethoxy-2-chloro-7-phenylcycloheptatriene, 3a/3′a, and 7-carbomethoxy-3-chloro-7-phenylnorcaradiene/7-carbomethoxy-3-chloro-7-phenylcycloheptatriene, 4a/4′a. Whenp-methylanisole was treated with methyl 2-phenyldiazoacetate at 80 °C, a product that largely favored afused cyclopropane structure, 7-carbomethoxy-2-methoxy-5-methyl-7-phenylnorcaradiene, 12a, wasobtained along with the benzylic C–H insertion product methyl 3-( p-methoxyphenyl)-2-phenylpropi-onate, 13a. Heating the norcaradiene product 12a at 110 °C yielded the ring-opened diarylacetate, 14a.The diene forms of the fluxional norcaradiene-cycloheptatriene systems were trapped with benzyne togive one stereoisomer of 3,3-disubstituted benzhomobarralenes, 18a–d. The norcaradiene-cyclohepta-triene valence isomers were quantitatively converted into ring-opened diaryl acetate products upon acid-ification in acetonitrile. Rates for the addition of methyl ( p-chlorophenyl)diazoacetate to benzene werefirst order with respect to the diazo reagent. A concerted mechanism involing an iron carbene complexis proposed for these iron porphyrin-catalyzed Büchner reactions.

KEYWORDS: catalysis, iron porphyrin, carbene, diazo reagent, Büchner reaction, valence isomer.

�SPP full member in good standing

*Correspondence to: L. Keith Woo, email: [email protected], tel: +1515-294-5854, fax: +1515-294-9623

isomerization of bicyclo[2.2.1]heptadiene (norbornadi-ene) to 1,3,5-cycloheptatriene (CHT) [5]. Considerablemechanistic and synthetic interests continue to evolvewith these processes.

The scope of iron-porphyrin-mediated diazo chem-istry includes dimerization, cyclopropanation [6], olefi-nation [7], N–H insertion [8], and C–H activation [9].While continuing to examine metal-mediated diazoprocesses, we examined the Fe(TPP)Cl-catalyzed reac-tion of methyl 2-phenyldiazoacetate (MPDA) with ben-zene. Unlike the aromatic ring C–H insertion observed inour previous work [9], the reaction afforded a mixturewhose NMR spectra indicated that it consisted of rapidlyinterconverting norcaradiene–cycloheptatriene valenceisomers. Furthermore, upon acidification in acetonitrilethese fluxional NCD/CHT products quantitatively con-verted to ring-opened diaryl acetate products, structuralmotifs that are present in a number of important pharma-ceuticals [10] such as tolterodine [11], CDP-840 [12],and nomifensine [13]. The details of Fe(TPP)Cl-catalyzed reactions with a variety of arenes and p-substituted MPDAs are reported herein.

EXPERIMENTAL

General

Fe(TPP)Cl was obtained from Aldrich. Toluene wasdried by passing through a column of catalytic copperand alumina as described by Grubbs et al. [14].Substituted methyl phenyldiazoacetates were preparedas outlined in the literature [15]. Proton NMR and 13CNMR spectra were recorded on a Varian VXR 300 or aBruker DRX400 spectrometer. 1H NMR peak positionswere referenced against residual proton resonances ofdeuterated CDCl3 (δ, 7.27 ppm). Gas chromatographyanalysis was performed on a HP 5890 series II or aFinnigan GC-MS. Dodecane was used as an internalstandard. All reactions were performed under an atmos-phere of nitrogen.

General procedure for Büchner and C–H insertionreactions. About 30.0 mg of the diazo reagent wereaccurately weighed and placed in a 50-mL round bottomflask containing a stir bar. A condenser fitted with a rub-ber septum was then attached to the round bottom flaskand the contents thoroughly flushed with nitrogen.Fe(TPP)Cl (2 mol.%) was placed in a separate flask, dis-solved in 5 mL of substrate, and the contents were bub-bled with dry nitrogen for 15 min. This solution wastransferred to the diazo reagent by a cannula. The mix-ture was then heated to an appropriate temperature whilecontinuously stirring until the diazo reagent wasconsumed. The products were separated or purified byeluting on a silica gel column (4 cm diameter × 30 cm,hexane/ethyl acetate; 30:1).

General procedure for trapping fluxionalnorcaradiene-cycloheptatriene systems with benzyne.About 10.0 mg of the norcaradiene–cycloheptatrieneproduct was dissolved in 10.0 mL of acetonitrile andplaced in a 4-dram vial that contained a stir bar. To thissolution, 1.2 equiv. of 2-(trimethylsilyl)phenyl triflatewere added and the mixture was stirred briefly beforeadding 2 equiv. of cesium fluoride. The vial was thensealed and the mixture stirred for 8 h. The product waspurified by silica gel chromatography (4 cm diameter ×30 cm long, hexane/ethyl acetate; 20:1).

Synthesis

Reaction of methyl 2-phenyldiazoacetate with ben-zene. The general procedure was used with methyl 2-phenyldiazoacetate (30.1 mg, 0.171 mmol), Fe(TPP)Cl(2.40 mg, 1.99 mol.%), and 10.0 mL of benzene. Themixture was stirred at 80 °C for 32 h. The product wasisolated by eluting through a silica gel column using a30:1 hexane/ethyl acetate mixture. The products, 7-car-bomethoxy-7-phenylnorcaradiene/7-carbomethoxy-7-phenylcycloheptatriene, 2a/2′a (29.3 mg, 0.130 mmol,76% yield based on methyl 2-phenyldiazoacetate) wereobtained as a white solid. The proton NMR and 13C NMRmatched literature values [16]. 2a/2′a. 1H NMR (300MHz): δ, ppm 7.19 (m, 5H, aryl C–H), 6.28 (m, 2H,vinyl C–H), 6.03 (m, 2H, vinyl C–H), 4.31 (m, 2H, H1,6), 3.65 (s, 3H, OCH3).

13C NMR (75.4 MHz): δ, ppm176.0, 135.6, 131.3, 127.4, 127.1, 126.9, 125.1, 72.3,53.0. MS (EI): m/z 226. Calcd. for [M]+ 226. Anal. foundfor C15H14O2 (calcd.) %C: 79.85 (79.62), %H: 6.26(6.24).

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith benzene. The general procedure was used withmethyl 2-( p-chlorophenyl)diazoacetate (30.7 mg, 0.146mmol), Fe(TPP)Cl (2.10 mg, 2.04 mol.%), and 10.0 mLof benzene. The mixture was stirred at 80 °C for 32 h.The product was isolated by eluting through a silica gelcolumn using a 30:1 hexane/ethyl acetate mixture. Theproducts, 7-carbomethoxy-7-( p-chlorophenyl)norcaradi-ene/7-carbomethoxy-7-( p-chlorophenyl)cyclohepta-triene, 2b/2′b (29.6 mg, 0.114 mmol, 78% yield based onmethyl 2-( p-chlorophenyl)diazoacetate) were obtainedas a white solid. 1H NMR (300 MHz): δ, ppm 7.14 (d,2H, JH = 8.4, aryl C–H), 7.09 (d, 2H, JH = 8.4 aryl C–H),6.24 (m, 2H, vinyl C–H), 5.98 (m, 2H, vinyl C–H), 4.13(m, 2H, H1,6), 3.66 (s, 3H, OCH3).

13C NMR (75.4 MHz):δ, ppm 175.9, 133.7, 133.2, 132.8, 129.3, 127.6, 127.5,125.2, 67.6, 53.2. MS(EI): m/z 260. Calcd. for [M]+

260.7. Anal. found for C15H13O2Cl (calcd.) %C: 69.32(69.10), %H: 5.21 (5.03).

Reaction of methyl 2-( p-tolyl)diazoacetate withbenzene. The general procedure was used with methyl 2-( p-tolyl)diazoacetate (30.4 mg, 0.160 mmol), Fe(TPP)Cl(2.30 mg, 2.04 mol.%), and 10.0 mL of benzene. Themixture was stirred at 80 °C for 32 h. The product was


ADDITION OF CARBENES DERIVED FROM ARYLDIAZOACETATES TO ARENES 137

isolated by eluting through a silica gel column using a30:1 hexane/ethyl acetate mixture. The products, 7-carbo-methoxy-7-( p-tolyl)norcaradiene/7-carbomethoxy-7-( p-tolyl)cycloheptatriene, 2c/2′c (30.3 mg, 0.126 mmol,79% yield based on methyl 2-( p-tolyl)diazoacetate),were obtained as a white solid. 1H NMR (400 MHz): δ,ppm 7.08 (d, 2H, JH = 8.0, aryl C–H), 7.00 (d, 2H, JH =8.0, aryl C–H), 6.28 (m, 2H, vinyl C–H), 6.04 (m, 2H,vinyl C–H), 4.31 (m, 2H, H1,6), 3.65 (s, 3H, OCH3), 2.29(s, 3H, ArCH3).

13C NMR (100.5 MHz): δ, ppm 176.4,136.6, 132.7, 131.2, 128.2, 127.6, 125.3, 73.1, 53.2,41.0, 21.5. MS (EI): m/z 241. Calcd. for [M + 1]+ 241.Anal. found for C16H16O2 (calcd.) %C: 79.52 (79.97),%H: 6.83 (6.71).

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with benzene. The general procedure was usedwith methyl 2-( p-methoxyphenyl)diazoacetate (30.5 mg,0.148 mmol), Fe(TPP)Cl (2.10 mg, 2.02 mol.%), and10.0 mL of benzene. The mixture was stirred at 80 °C for32 h. The products were isolated by eluting through asilica gel column using a 30:1 hexane/ethyl acetate mix-ture. The products, 7-carbomethoxy-7-( p-methoxyphenyl)norcaradiene/7-carbomethoxy-7-( p-methoxyphenyl)cycloheptatriene, 2d/2′d (31.1 mg, 0.121 mmol, 82%yield based on methyl 2-( p-methoxyphenyl)diazoac-etate) were obtained as a white solid. The 1H NMR datawas found to match literature values [16]. 2d/2′d.1H NMR (400 MHz, CDCl3): δ, ppm 7.09 (d, 2H, JH =8.8, aryl C–H), 6.72 (d, 2H, JH = 8.8 aryl C–H), 6.25 (m,2H, vinyl C–H), 6.02 (m, 2H, vinyl C–H), 4.21 (m, 2H,H1,6), 3.76 (s, 3H, ArOCH3), 3.65 (s, 3H, OCH3).

13CNMR (100.5 MHz): δ, ppm 176.4, 158.2, 132.4, 127.3,125.0, 112.6, 70.4, 60.4, 55.1, 52.9. MS(EI): m/z 256.Calcd. for [M]+ 256. Anal. found for C16H16O3 (calcd.)%C: 75.12 (74.98), %H: 6.33 (6.29).

Reaction of methyl 2-phenyldiazoacetate withchlorobenzene. The general procedure was used withmethyl 2-phenyldiazoacetate (30.1 mg, 0.171 mmol),Fe(TPP)Cl (2.40 mg, 1.99 mol.%), and 5.0 mL ofchlorobenzene. The mixture was stirred at 80 °C for 32 h. The products were partially separated by elutingthrough a silica gel column using a 30:1 hexane/ethylacetate mixture. It was not possible to fully separate thetwo products. A combined yield of 34.6 mg, 0.133 mmol,of both products, 3a/3′a and 4a/4′a (78% yield based onmethyl 2-phenyldiazoacetate were obtained). However,8.20 mg (18% yield) of pure 7-carbomethoxy-2-chloro-7-phenylnorcaradiene/7-carbomethoxy-2-chloro-7-phenylcycloheptatriene, 3a/3′a, was obtained from acenter-cut of a band. 3a/3′a. 1H NMR (400 MHz): δ, ppm7.20–7.25 (m, 3H, aryl C–H), 7.11–7.15 (m, 2H, arylC–H), 6.13 (dd, 1H, JH = 9.2, JH = 5.2 vinyl C–H), 5.83(d, JH = 6.8, 1H vinyl C–H), 5.71 (dd, JH = 9.2, JH = 6.8,1H, vinyl C–H), 3.67 (s, 3H, OCH3), 3.36 (d, JH = 8.8,1H, H1,6), 3.30 (dd, JH = 8.8, JH = 5.2, 1H, H1,6).

13C NMR(100.5 MHz): δ, ppm 176.1, 132.3, 132.0, 131.4, 127.5,127.4, 125.6, 123.1, 123.0, 53.3, 44.6, 43.0. MS (EI): m/z

260. Calcd. for [M]+ 260. Anal. found for C15H12ClO2

(calcd.) %C: 68.97 (69.10), %H: 5.11 (5.03). 7-carbomethoxy-3-chloro-7-phenylnorcaradiene/7-car-bomethoxy-3-chloro-7-phenylcycloheptatriene, 4a/4′a(11.4 mg, 25%) containing ~10% of 3a/3′a were obtained.4a/4′a. 1H NMR (400 MHz): δ, ppm 7.19–7.24 (m, 3H,aryl C–H), 7.12–7.18 (m, 2H, aryl C–H), 6.28 (m, 2H,vinyl C–H), 6.03 (d, JH = 8.4, 1H, vinyl C–H), 4.38 (m,2H, H1,6), 3.66 (s, 3H, OCH3).

13C NMR (100.5 MHz): δ,ppm 175.2, 134.9, 132.2, 130.9, 128.3, 127.4, 127.3,126.3, 123.5, 74.7, 73.4, 53.2, 38.7. MS (EI): m/z 260.Calcd. for [M]+ 260.

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith chlorobenzene. The general procedure was usedwith methyl 2-(chlorophenyl)diazoacetate (30.6 mg,0.145 mmol), Fe(TPP)Cl (2.20 mg, 2.16 mol.%) and 5.0mL of chlorobenzene. The mixture was stirred at 80 °Cfor 32 h. The products were partially separated by elut-ing through a silica gel column using a 30:1 hexane/ethylacetate mixture. It was not possible to fully separate thetwo products. A combined yield of 30.9 mg, 0.105 mmol,72% yield based on methyl 2-(chlorophenyl)diazoac-etate) was obtained. However, collecting a center frac-tion of the 7-carbomethoxy-2-chloro-7-( p-chlorophenyl)norcaradiene/7-carbomethoxy-2-chloro-7-( p-chloro-phenyl)cycloheptatriene, 3b/3′b, band produced a puresample (10.4 mg, 24% yield). 3b/3′b. 1H NMR (400MHz): δ, ppm 7.20 (d, 2H, JH = 8.6, aryl C–H), 7.06 (d,2H, JH = 8.6, aryl C–H), 6.11 (dd, 1H, JH = 9.2, JH = 5.6,vinyl C–H), 5.84 (d, 1H, JH = 6.8, vinyl C–H), 5.73 (dd, 1H,JH = 9.2, JH = 6.8, vinyl C–H), 3.67 (s, 3H, OCH3), 3.32(d, 1H, JH = 9.0, H1,6), 3.25 (dd, 1H, JH = 9.0, JH = 5.6,H1,6).

13C NMR (100.5 MHz): δ, ppm 175.6, 133.7, 133.3,131.3, 130.4, 127.9, 125.8, 123.1, 122.8, 53.3, 43.7, 42.0.MS (EI): m/z 295. Calcd. for [M]+ 295. Anal. found forC15H12Cl2O2 (calcd.) %C: 59.96 (61.04), %H: 4.21 (4.10).7-carbomethoxy-3-chloro-7-( p-chlorophenyl)norcaradi-ene/7-carbomethoxy-3-chloro-7-( p-chlorophenyl)cyclo-heptatriene, 4b/4′b (9.40 mg, 22% yield) containing~ 2% of 3b/3′b was obtained. 4b/4′b. 1H NMR (300MHz): δ, ppm 7.18 (d, 2H, JH = 8.7, aryl C–H), 7.08 (d,2H, JH = 8.7, aryl C–H), 6.27 (m, 2H, vinyl C–H), 6.02(dd, 1H, JH = 8.7, JH = 1.5, vinyl C–H), 4.25 (m,2H, H1,6), 3.68 (s, 3H, OCH3).

13C NMR (75.4 MHz):δ, ppm 175.0, 133.8, 133.3, 132.7, 132.4, 131.7, 128.7,127.9, 126.6, 123.4, 53.4. MS (EI): m/z 295. Calcd. for[M]+ 295.

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with chlorobenzene. The general procedure wasused with methyl 2-(methoxyphenyl) diazoacetate (30.3mg, 0.147 mmol), Fe(TPP)Cl (2.20 mg, 2.13 mol.%) and5.0 mL of chlorobenzene. The mixture was stirred at80 °C for 32 h. The products were partially separatedby eluting through a silica gel column using a 30:1hexane/ethyl acetate mixture. The products, 7-carbometh-oxy-2-chloro-7-( p-methoxyphenyl)norcaradiene/7-carbo-methoxy-2-chloro-7-( p-methoxyphenyl)cycloheptatriene,


138 H.M. MBUVI AND L.K. WOO

3d/3′d, and 7-carbomethoxy-3-chloro-7-( p-methoxy-phenyl)norcaradiene/7-carbomethoxy-3-chloro-7-( p-methoxyphenyl)cycloheptatriene, 4d/4′d, and unreactedp-MeO-MPDA could not be fully separated from eachother. The proton NMR of the product was assigned bycomparison with the spectra of 3b/3′b and 4b/4′b. AnNMR yield of 16% of 3d/3′d and 17% of 4d/4′d wasobtained. 3d/3′d. 1H NMR (400 MHz): δ, ppm 7.05 (d,2H, JH = 8.8, aryl C–H), 6.76 (d, 2H, JH = 8.8, aryl C–H),6.12 (dd, 1H, JH = 9.0, JH = 5.6, vinyl C–H), 5.84 (d, 1H,JH = 6.8, vinyl C–H), 5.72 (dd, 1H, JH = 9.0, JH = 6.8,vinyl C–H), 3.77 (s, 3H, ArOCH3), 3.65 (br, 3H, OCH3),3.31 (d, 1H, JH = 9.0, H1,6), 3.25 (dd, 1H, JH = 9.0, JH =5.6, H1,6). MS (EI): m/z 290. Calcd. for [M]+ 290. 4d/4′d.1H NMR (300 MHz): δ, ppm 7.07 (d, 2H, JH = 8.3, arylC–H), 6.74 (d, 2H, JH = 8.3, aryl C–H), 6.25 (m, 2H,vinyl C–H), 6.03 (m, 1H, vinyl C–H), 4.30 (m, 2H, H1,6),3.77 (s, 3H, ArOCH3), 3.67 (b, 3H, OCH3). MS (EI): m/z290. Calcd. for [M]+ 290.

Reaction of methyl 2-phenyldiazoacetate withtoluene. The general procedure was used with methyl 2-phenyldiazoacetate (30.1 mg, 0.171 mmol), Fe(TPP)Cl(2.40 mg, 1.99 mol.%), and 5.0 mL of toluene. The mix-ture was stirred at 80 °C for 32 h. The benzylic C–Hinsertion product methyl 2,3-diphenylpropionate, 5aand two Büchner products 7-carbomethoxy-2-methyl-7-phenylnorcaradiene/7-carbomethoxy-2-methyl-7-phenylcycloheptatriene, 6a/6′a and 7-carbomethoxy-3-methyl-7-phenylnorcaradiene/7-carbomethoxy-3-methyl-7-phenylcycloheptatriene, 7a/7′a, were partially sepa-rated using silica gel chromatography and hexane/ethylacetate 30:1 as eluent. It was not possible to fully separatethe three products. A combined yield of 31.6 mg, 0.132mmol 77% yield was obtained. A mixture (6.40 mg) of 5acontaining ~ 30% of 7a/7′a was used for spectroscopicanalysis. The proton NMR data for the benzylic C–Hproduct, methyl 2,3-diphenylpropionate, 5a, was found tomatch literature values [17]. Pure 6a/6′a (8.60 mg, 21%yield) was obtained from a center-cut of a band. 1H NMR(400 MHz): δ, ppm 7.17–7.22 (m, 3H, aryl C–H), 7.01-7.07 (m, 2H, aryl C–H), 6.04 (dd, 1H, JH = 8.4, JH = 5.6,vinyl C–H), 5.73 (dd, 1H, JH = 8.4, JH = 6.4 vinyl C–H),5.56 (d, 1H, JH = 6.4, vinyl C–H), 3.65 (s, 3H, OCH3),3.24 (dd, 1H, JH = 8.6, JH = 5.6, H1,6), 3.10 (d, 1H, JH =8.6, H1,6), 2.10 (s, 3H, vinyl-CH3).

13C NMR (100.5MHz): δ, ppm 177.3, 133.7, 133.1, 132.5, 127.2, 127.0,126.1, 122.1, 121.8, 53.0, 44.6, 42.1, 24.1. MS(EI): m/z240. Calcd. for [M]+ 240. An impure sample of 7a/7′acontaining ~ 30% of 6a/6′a was used for NMR analysis.1H NMR (400 MHz): δ, ppm 7.11–7.22 (m, 5H, arylC–H), 6.16 (m, 1H, vinyl C–H), 5.96 (m, 1H, vinyl C–H),5.74 (d, 1H, JH = 9.2, vinyl C–H), 3.88 (m, 2H, H1,6), 3.65(s, 3H, OCH3), 1.61 (s, 3H, vinyl-CH3).

13C NMR (100.5MHz): δ, ppm 176.4, 135.3, 135.1, 132.0, 129.0, 128.7,128.6, 125.1, 122.3, 52.8, 21.9.

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith toluene. The general procedure was used with

methyl 2-( p-chlorophenyl)diazoacetate (30.4 mg, 0.144mmol), Fe(TPP)Cl (2.00 mg, 1.97 mol.%), and 5.0 mL oftoluene. The mixture was stirred at 80 °C for 32 h. Thebenzylic C–H insertion product, 5b and two Büchnerproducts 7-carbomethoxy-2-methyl-7-( p-chlorophenyl)norcaradiene/7-carbomethoxy-2-methyl-7-( p-chloro-phenyl)cycloheptatriene, 6b/6′b and 7-carbomethoxy-3-methyl-7-( p-chlorophenyl)norcaradiene/7-carbomethoxy-3-methyl-7-( p-chlorophenyl)cycloheptatriene, 7b/7′b,were partially separated using silica gel chromatographyand hexane/ethyl acetate 30:1 as eluent. It was not possi-ble to fully separate the three products. A combined yieldof 27.2 mg (0.0993 mmol, 69% yield) was obtained. Itwas impossible to separate 5b from (7b/7′b) and about15.2 mg, 38% yield, containing a 1:1 ratio of these prod-ucts were obtained. The 1H NMR of the benzylic C–Hproduct, 5b, was found to match literature values [17]. Amixture (9.20 mg, 23% yield) of 6b/6′b containing~15% of 7b/7′b was used for spectroscopic analysis.6b/6′b. 1H NMR (300 MHz): δ, ppm 7.15 (d, 2H, JH = 8.4,aryl C–H), 6.95 (d, 2H, JH = 8.4, aryl C–H), 6.02 (dd, 1H,JH = 9.3, JH = 5.6, vinyl C–H), 5.74 (dd, 1H, JH = 9.3, JH

= 6.3, vinyl C–H), 5.58 (d, 1H, JH = 5.6, vinyl C–H), 3.65(s, 3H, OCH3), 3.20 (dd, 1H, JH = 8.6, JH = 5.6, H1,6), 3.06(d, 1H, JH = 8.6, H1,6), 2.08 (s, 3H, vinyl-CH3). 7b/7′b. 1HNMR (400 MHz): δ, ppm 7.05 (d, 2H, JH = 8.4, arylC–H), 6.95 (d, 2H, JH = 8.4, aryl C–H), 6.14 (m, 1H,vinyl C–H), 5.93 (m, 1H, vinyl C–H), 5.73 (d, 1H, JH =9.2, vinyl C–H), 3.73 (m, 2H, H1,6), 3.64 (s, 3H, OCH3),1.61 (s, 3H, vinyl-CH3). MS (EI): m/z 275. Calcd. for [M + 1]+ 275.

Reaction of methyl 2-( p-tolyl)diazoacetate withtoluene. The general procedure was used with methyl 2-( p-chlorophenyl)diazoacetate (30.3 mg, 0.159 mmol),Fe(TPP)Cl (2.20 mg, 1.97 mol.%), and 5.0 mL of toluene.The mixture was stirred at 80 °C for 32 h. The benzylicC–H insertion product, 5c and two Büchner products 7-carbomethoxy-2-methyl-7-( p-tolyl)norcaradiene/7-car-bomethoxy-2-methyl-7-( p-tolyl)cycloheptatriene, 6c/6′cand 7-carbomethoxy-3-methyl-7-( p-tolyl)norcaradiene/7-carbomethoxy-3-methyl-7-( p-tolyl)cycloheptatriene,7c/7′c, were partially separated using silica gel chro-matography and hexane/ethyl acetate 30:1 as eluent. Itwas not possible to separate the three products. A com-bined yield of 30.1 mg, 0.119 mmol, 75% yield wasobtained. A mixture (7.20 mg, 18%) of benzylic C–Hproduct 5c containing 20% 7c/7′c was obtained. The 1HNMR of 5c was found to match literature values [17].Pure 6c/6′c (7.40 mg, 18% yield) was obtained froma center-cut of a band. 1H NMR (300 MHz): δ, ppm 7.00(d, 2H, JH = 8.4, aryl C–H), 6.91 (d, 2H, JH = 8.4, arylC–H), 6.03 (dd, 1H, JH = 9.3, JH = 5.6, vinyl C–H),5.74 (dd, 1H, JH = 9.3, JH = 6.2, vinyl C–H), 5.57 (d, 1H,JH = 6.2, vinyl C–H), 3.65 (s, 3H, OCH3), 3.23 (dd, 1H,JH = 8.4, JH = 6.5, H1,6), 3.08 (d, 1H, JH = 8.4, H1,6), 2.29(s, 3H, ArCH3), 2.09 (s, 3H, vinyl-CH3).

13C NMR (75.4MHz): δ, ppm 177.7, 136.6, 133.8, 132.4, 130.2, 129.3,



128.2, 126.3, 122.3, 122.0, 53.2, 24.2, 21.5. MS(EI): m/z254. Calcd. for [M]+ 254. Anal. found for C17H18O2

(calcd.) %C: 80.02 (80.28), %H: 6.97 (7.13). A mixture(2.20 mg, 5% yield) of 7c/7′c containing ~15% 5c wasused for spectroscopic analysis. 7c/7′c. 1H NMR (300MHz): δ, ppm 7.04 (d, 2H, JH = 8.1, aryl C–H), 6.99 (d,2H, JH = 8.1, aryl C–H), 6.15 (m, 1H, vinyl C–H), 5.96(m, 1H, vinyl C–H), 5.77 (d, 1H, JH = 8.4, vinyl C–H),3.93 (m, 2H, H1,6), 3.63 (s, 3H, OCH3), 2.28 (s, 3H,ArCH3), 1.64 (s, 3H, R vinyl-CH3). MS (EI): m/z 254.Calcd. for [M]+ 254.

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with toluene. The general procedure was used withmethyl 2-( p-methoxyphenyl)diazoacetate (30.5 mg,0.148 mmol), Fe(TPP)Cl (2.10 mg, 2.02 mol.%), and 5.0mL of toluene. The mixture was stirred at 80 °C for 32 h.Pure benzylic C–H insertion product 5d (11.6 mg, 29%yield), pure Büchner product 7-carbomethoxy-2-methyl-7-( p-methoxyphenyl)norcaradiene/7-carbomethoxy-2-methyl-7-( p-methoxyphenyl)cycloheptatriene, 6d/6′d(6.50 mg, 16% yield), and partially pure 7-carbo-methoxy-3-methyl-7-( p-methoxyphenyl)norcaradiene/7-carbomethoxy-3-methyl-7-( p-methoxyphenyl)cyclo-heptatriene, 7d/7′d (13.5 mg, 34%) were obtained fromsilica gel chromatography with hexane/ethyl acetate 30:1as eluent. The 1H NMR spectrum of 5d was found tomatch literature data. 6d/6′d. 1H NMR (400 MHz): δ,ppm 6.94 (d, 2H, JH = 8.8, aryl C–H), 6.72 (d, 2H, JH =8.8, aryl C–H), 6.01 (dd, 1H, JH = 9.2, JH = 5.6, vinylC–H), 5.74 (dd, 1H, JH = 9.2, JH = 6.5, vinyl C–H), 5.57(d, 1H, JH = 6.5, vinyl C–H), 3.76 (s, 3H, ArOCH3), 3.65(s, 3H, OCH3), 3.20 (dd, 1H, JH = 8.4, JH = 5.6, H1,6), 3.05(d, 1H, JH = 8.4, H1,6), 2.08 (s, 3H, R vinyl-CH3).

13CNMR (100.5 MHz): δ, ppm 177.6, 158.2, 133.6, 133.4,126.2, 125.0, 122.1, 121.7, 112.7, 55.0, 53.0, 44.3, 41.8,24.1. MS (EI): m/z 270. Calcd. for [M]+ 270. A mixture(4.20 mg, 9%) of 7d/7′d containing ~ 25% of 5d wasused for spectroscopic analysis. 1H NMR (300 MHz): δ,ppm 7.04 (d, 2H, JH = 8.7, aryl C–H), 6.72 (d, 2H, JH =8.7, aryl C–H), 6.14 (m, 1H, vinyl C–H), 5.94 (m, 1H,vinyl C–H), 5.75 (d, 1H, JH = 9.0, vinyl C–H), 3.81 (m,2H, H1,6), 3.77 (s, 3H, ArOCH3), 3.64 (s, 3H, OCH3),1.63 (s, 3H, R vinyl-CH3). MS (EI): m/z 270. Calcd. for[M]+ 270.

Reaction of methyl 2-phenyldiazoacetate withanisole. The general procedure was used with methyl 2-phenyldiazoacetate (30.1 mg, 0.171 mmol) Fe(TPP)Cl(2.40 mg, 1.99 mol.%), and 5.0 mL of anisole. The mixturewas stirred at 80 °C for 32 h. The products were partiallyseparated by eluting through a silica gel column using a30:1 hexane/ethyl acetate mixture. Two Büchner products7-carbomethoxy-2-methoxy-7-phenylnorcaradiene/7-carbomethoxy-2-methoxy-7-phenylcycloheptatriene,8a/8′a and 7-carbomethoxy-3-methoxy-7-phenylnor-caradiene/7-carbomethoxy-3-methoxy-7-phenylcyclo-heptatriene, 9a/9′a, were partially separated using silicagel chromatography and hexane/ethyl acetate 30:1 as

eluent. It was not possible to fully separate the two prod-ucts. A combined yield of 16.6 mg, 0.0648 mmol, 38%yield based on methyl 2-phenyldiazoacetate was obtained.A sample of 8a/8′a containing ~ 5% of 9a/9′a was usedfor spectroscopic analysis. 1H NMR (400 MHz): δ, ppm7.19 (m, 5H, aryl C–H), 6.24 (dd, 1H, JH = 11.0, JH = 9.6,vinyl C–H), 5.58 (dd, 1H, JH = 11.0, JH = 2.4, vinylC–H), 5.51 (dd, 1H, JH = 6.0, JH = 2.4, vinyl C–H), 4.25(dd, 1H, JH = 9.6, JH = 9.6, H1,6), 4.09 (dd, 1H, JH = 9.6,JH = 9.6, H1,6), 3.64 (s, 3H, ArOCH3), 3.43 (s, 3H,ROCH3). MS(EI): m/z 256. Calcd. for [M]+ 256. 9a/9′a.1H NMR (400 MHz): δ, ppm 7.18 (m, 3H, aryl C–H),7.09 (m, 2H, aryl C–H), 5.72 (m, 2H, vinyl C–H), 4.79(d, 1H, JH = 8.0, vinyl C–H), 3.62 (s, 3H, ROCH3), 3.61(s, 3H, ROCH3), 3.10 (m, 2H, H1,6). MS (EI): m/z 256.Calcd. for [M]+ 256.

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with anisole. The general procedure was used withmethyl 2-( p-methoxyphenyl)diazoacetate (30.4 mg,0.148 mmol), Fe(TPP)Cl (2.10 mg, 2.02 mol.%), and5.0 mL of anisole. The mixture was stirred at 80 °C for32 h. The products were partially separated by elutingthrough a silica gel column using a 30:1 hexane/ethylacetate mixture. Two Büchner products 7-carbomethoxy-2-methoxy-7-( p-methoxyphenyl)norcaradiene/7-car-bomethoxy-2-methoxy-7-( p-methoxyphenyl)cycloheptatriene, 8d/8′d and 7-carbomethoxy-3-methoxy-7-( p-methoxyphenyl)norcaradiene/7-carbomethoxy-3-methoxy-7-( p-methoxyphenyl)cycloheptatriene, 9d/9′d,were partially separated using silica gel column andhexane/ethyl acetate 30:1 as eluent. It was not possible tofully separate the two products. A combined yield of 17.8mg, 0.0622 mmol, 42% yield based on methyl 2-( p-methoxyphenyl)diazoacetate was obtained. A sample of8d/8′d containing some minor impurities was used forspectroscopic analysis. 1H NMR (400 MHz): δ, ppm 7.09(d, 2H, JH = 8.8, aryl C–H), 6.72 (d, 2H, JH = 8.8, arylC–H), 6.22 (m, 1H, vinyl C–H), 5.59 (m, 1H, JH = 2.6,vinyl C–H), 5.48 (m, 1H, vinyl C–H), 4.13 (m, 1H, H1,6),3.99 (m, 1H, H1,6), 3.76 (s, 3H, ArOCH3), 3.64 (s, 3H,ROCH3), 3.45 (s, 3H, ROCH3). MS(EI): m/z 286. Calcd.for [M]+ 286. A sample of 9d/9′d containing minor impu-rities was used for spectroscopic analysis. 1H NMR (400MHz): δ, ppm 7.01 (d, 2H, JH = 8.4, aryl C–H), 6.72 (d,2H, JH = 8.4, aryl C–H), 5.73 (m, 2H, vinyl C–H), 4.82 (m,1H, vinyl C–H), 3.77 (s, 3H, ArOCH3), 3.63 (s, 3H,ROCH3), 3.62 (s, 3H, ROCH3), 3.08 (m, 2H, H1,6).MS(EI): m/z 286. Calcd. for [M]+ 286.

Reaction of methyl 2-phenyldiazoacetate withp-xylene. The general procedure was used with methyl2-phenyldiazoacetate (30.3 mg, 0.172 mmol), Fe(TPP)Cl(2.40 mg, 1.98 mol.%), and 5.0 mL of p-xylene. Themixture was stirred at 80 °C for 8 h. The cyclopropana-tion product 7-carbomethoxy-2,5-dimethyl-7-phenylnor-caradiene, 10a and the benzylic C–H insertion product11a were partially separated using silica gel column andhexane/ethyl acetate 30:1 as eluent. A combined yield of



35.9 mg, 0.141 mmol, 82% yield based on methyl 2-phenyldiazoacetate was obtained. However, 12.6 mg, 29%yield of pure 10a were obtained: 1H NMR (300 MHz): δ,ppm 7.19 (m, 3H, aryl C–H), 6.97 (m, 2H, aryl C–H), 5.47(s, 2H, vinyl C–H), 3.65 (s, 3H, ROCH3), 2.94 (s, 2H,H1,6), 2.09 (s, 6H, RCH3).

13C NMR (75.4 MHz): δ, ppm177.8, 133.1, 132.0, 130.7, 127.5, 127.2, 122.2, 53.1, 41.4,23.9. MS(EI): m/z 254. Calcd. for [M]+ 254. Anal. foundfor C17H18O2 (calcd.) %C: 80.06 (80.28) %H: 7.28 (7.13).Pure 11a (8.60 mg, 20% yield) was obtained from a cen-ter-cut of a band. 1H NMR and 13C NMR data for 11a werefound to match literature values. 11a. 1H NMR (300 MHz):δ, ppm 7.28 (m, 5H, aryl C–H), 7.06 (d, 2H, JH = 8.4, arylC–H), 7.02 (d, 2H, JH = 8.4, aryl C–H), 3.84 (dd, 1H, JH =9.0, JH = 6.6, methine C–H), 3.61 (s, 3H, OCH3), 3.39 (dd,1H, JH = 13.8, JH = 9.0, ArCH2R), 2.99 (dd, 1H, JH = 13.8,JH = 6.6, ArCH2R), 2.30 (s, 3H, ArCH3).

13C NMR (75.4MHz): δ, ppm 174.1, 139.8, 136.2, 136.1, 129.3, 129.0,128.9, 128.2, 127.6, 53.9, 52.2, 39.6, 21.3. MS (EI): m/z254. Calcd. for [M]+ 254.

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith p-xylene. Methyl 2-( p-chlorophenyl)diazoacetate(30.2 mg, 0.143 mmol) was placed in a round bottomflask. In a different round bottom flask, the catalystFe(TPP)Cl (2.00 mg, 1.99 mol.%) was dissolved in 5.0mL of p-xylene under an atmosphere of dry nitrogen.Both flasks were thoroughly flushed with nitrogenbefore the dissolved catalyst was transferred to the flaskcontaining the diazo by a cannula. The mixture wasstirred at 80 °C for 32 h. The products were partially sep-arated by eluting through a silica gel column using a 30:1hexane/ethyl acetate mixture. The cyclopropanationproduct 7-carbomethoxy-2,5-dimethyl-7-( p-chlorophenyl)norcaradiene, 10b and the benzylic C–H insertion prod-uct 11b were partially separated using silica gel columnand hexane/ethyl acetate 30:1 as eluent: a combinedyield of 34.4 mg, 0.119 mmol, 83% yield based onmethyl 2-( p-chlorophenyl)diazoacetate) were obtained.Pure 10b (16.8 mg, 41% yield) was obtained from acenter-cut of a band. 1H NMR (400 MHz): δ, ppm 7.15(d, 2H, JH = 8.4, aryl C–H), 6.88 (d, 2H, JH = 8.4, arylC–H), 5.49 (s, 2H, vinyl C–H), 3.66 (s, 3H, ROCH3),2.94 (s, 2H, H1,6), 2.07 (s, 6H, RCH3).

13C NMR (100.5MHz): δ, ppm 177.1, 133.2, 132.8, 131.4, 130.3, 127.7,122.3, 53.0, 41.2, 23.7. MS (EI): m/z 289. Calcd. for[M+1]+ 289. Anal. found for C17H17ClO2 (calcd.) %C:70.46 (70.71) %H: 5.71 (5.93). A mixture of 11b (5.40mg, 13% yield) containing ~ 25% 10b was used for NMRanalysis. 1H NMR (400 MHz): δ, ppm 7.28 (d, 2H, JH =8.4, aryl C–H), 7.23 (d, 2H, JH = 8.4, aryl C–H), 7.07 (d,2H, JH = 8.0, aryl C–H), 6.98 (d, 2H, JH = 8.0, aryl C–H),3.81 (dd, 1H, JH = 8.0, JH = 7.6, methine C–H), 3.62 (s,3H, ROCH3), 3.36 (dd, 1H, JH = 14.6, JH = 8.0, ArCH2R),2.97 (dd, 1H, JH = 14.6, JH = 7.6, ArCH2R), 2.30 (s, 3H,RCH3).

13C NMR (100.5 MHz): δ, ppm 173.8, 137.7,137.3, 135.3, 132.2, 130.4, 129.5, 128.5, 127.8, 53.1, 52.1,39.1, 21.1. MS (EI): m/z 289. Calcd. for [M + 1]+ 289.

Reaction of methyl 2-( p-tolyl)diazoacetate with p-xylene. The general procedure was used with methyl 2-( p-tolyl)diazoacetate (30.6 mg, 0.161 mmol), Fe(TPP)Cl(2.20 mg, 1.94 mol.%), and 5.0 mL of p-xylene. The mix-ture was stirred at 80 °C for 32 h. The products were par-tially separated by eluting through a silica gel columnusing a 30:1 hexane/ethyl acetate mixture. The cyclo-propanation product 7-carbomethoxy-2,5-dimethyl-7-( p-tolyl)norcaradiene, 10c and the benzylic C–Hinsertion product 11c were partially separated using sil-ica gel column and hexane/ethyl acetate 30:1 as eluent:A combined yield of 37.1 mg, 0.138 mmol, 86% yieldbased on methyl 2-( p-tolyl)diazoacetate were obtained.A pure sample of 10c (12.6 mg, 29% yield) was obtainedfrom a center-cut of a band. 1H NMR (300 MHz): δ, ppm7.01 (d, 2H, JH = 8.1, aryl C–H), 6.86 (d, 2H, JH = 8.1,aryl C–H), 5.49 (s, 2H, vinyl C–H), 3.66 (s, 3H,ROCH3), 2.93 (s, 2H, H1,6), 2.23 (s, 3H, ArCH3), 2.08 (s,6H, RCH3).

13C NMR (75.4 MHz): δ, ppm 178.0, 136.7,131.8, 130.7, 130.0, 128.4, 122.2, 53.1, 41.5, 25.8, 23.9,21.5. MS(EI): m/z 268. Calcd. for [M]+ 268. Anal. foundfor C18H20O2 (calcd.) %C: 79.86 (80.56), %H: 7.32(7.51). Pure benzylic C–H product, 11c, (10.6 mg, 25%yield) was obtained from a center-cut of a band. 1H NMR(400 MHz): δ, ppm 7.23 (d, 2H, JH = 8.0, aryl C–H), 7.14(d, 2H, JH = 8.0, aryl C–H), 7.07 (d, 2H, JH = 8.0, arylC–H), 7.03 (d, 2H, JH = 8.0, aryl C–H), 3.83 (dd, 1H, JH

= 9.0, JH = 6.8, methine C–H), 3.61 (s, 3H, ROCH3), 3.39(dd, 1H, JH = 13.4, JH = 9.0, ArCH2R), 2.99 (dd, 1H, JH

= 13.4, JH = 6.8, ArCH2R), 2.35 (s, 3H, ArCH3), 2.31 (s,3H, RCH3).

13C NMR (100.5 MHz): δ, ppm 174.3, 137.2,136.3, 136.0, 129.6, 129.3, 129.0, 128.0, 53.5, 52.2,39.6, 21.3, 21.2. MS (EI): m/z 268. Calcd. for [M]+ 268.Anal. found for C18H20O2 (calcd.) %C: 80.32 (80.56),%H: 7.21 (7.51).

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with p-xylene. The general procedure was usedwith methyl 2-( p-methoxyphenyl)diazoacetate (30.4 mg,0.148 mmol), Fe(TPP)Cl (2.30 mg, 2.21 mol.%), and 5.0mL of p-xylene. The mixture was stirred at 80 °C for 32h. The products were partially separated by elutingthrough a silica gel column using a 30:1 hexane/ethylacetate mixture. The cyclopropanation product 7-car-bomethoxy-2,5-dimethyl-7-( p-methoxyphenyl)norcara-diene, 10d and the benzylic C–H insertion product 11dwere partially separated using silica gel column andhexane/ethyl acetate 30:1 as eluent: a combined yieldof 39.9 mg, 0.140 mmol, 88% yield based on methyl2-( p-methoxyphenyl)diazoacetate was obtained. Pure10d (9.70 mg, 21% yield) was obtained from a center-cutof a band. 1H NMR (400 MHz): δ, ppm 6.88 (d, 2H, JH =9.6, aryl C–H), 6.73 (d, 2H, JH = 9.6, aryl C–H), 5.46 (s,2H, vinyl C–H), 3.73 (s, 3H, ArOCH3), 3.63 (s, 3H,ROCH3), 2.88 (s, 2H, H1,6), 2.04 (s, 6H, RCH3).

13CNMR (100.5 MHz): δ, ppm 178.3, 158.3, 138.6, 132.8,130.5, 122.1, 112.8, 55.0, 53.0, 41.3, 23.7. MS (EI) : m/z284. Calcd. for [M]+ 284. Pure benzylic C–H product,



11d (12.3 mg, 27%) was obtained from a center-cut of aband and used for spectroscopic analysis. 1H NMR (400MHz): δ, ppm 7.21 (d, 2H, JH = 8.0, aryl C–H), 7.02 (d,2H, JH = 8.0, aryl C–H), 7.00 (d, 2H, JH = 8.0, aryl C–H),6.83 (d, 2H, JH = 8.0, aryl C–H), 3.77 (s, 3H, ArOCH3),3.76 (m, 1H, methine C–H, overlaps with ArOCH3), 3.58(s, 3H, ROCH3), 3.32 (dd, 1H, JH = 13.8, JH = 8.6,ArCH2R), 2.94 (dd, 1H, JH = 13.8, JH = 6.8, ArCH2R),2.27 (s, 3H, ArCH3).

13C NMR (100.5 MHz): δ, ppm174.4, 159.0, 136.3, 136.0, 133.0, 131.1, 129.2, 129.0,114.2, 55.5, 53.0, 52.2, 39.7, 21.3. MS (EI): m/z 284.Calcd. for [M]+ 284. Anal. found for C18H20O3 (calcd.)%C: 75.58 (76.03), %H: 6.97 (7.09).

Reaction of methyl 2-phenyldiazoacetate with p-methylanisole. The general procedure was used withmethyl 2-phenyldiazoacetate (30.3 mg, 0.172 mmol),Fe(TPP)Cl (2.40 mg, 1.98 mol.%), and 5.0 mL ofp-methylanisole. The mixture was stirred at 80 °C for 32 h.The products were partially separated by eluting througha silica gel column using a 30:1 hexane/ethyl acetatemixture. The cyclopropanation product 7-carbomethoxy-2-methoxy-5-methyl-7-phenylnorcaradiene, 12a and thebenzylic C–H insertion product 13a were partially sepa-rated using silica gel column and hexane/ethyl acetate30:1 as eluent: a combined yield of 34.4 mg, 0.127mmol, 74% yield based on methyl 2-phenyldiazoacetatewere obtained. An impure sample of 12a (11.3 mg, 24%yield) was used for NMR analysis. 1H NMR (400 MHz):δ, ppm 7.20 (m, 3H, aryl C–H), 7.06 (m, 2H, aryl C–H),5.45 (d, 1H, JH = 6.8, vinyl C–H), 4.74 (d, 1H, JH = 6.8,vinyl C–H), 3.64 (s, 3H, ArOCH3), 3.60 (s, 3H, ROCH3),3.05 (d, 1H, JH = 9.6, H1,6), 2.96 (d, 1H, JH = 9.6, H1,6),2.05 (s, 3H, RCH3).

13C NMR (100.5 MHz): δ, ppm177.1, 154.5, 133.0, 131.9, 127.6, 127.3, 124.6, 121.6,95.9, 55.5, 53.1, 46.8, 36.8, 25.1, 23.3. MS (EI): m/z 270.Calcd. for [M]+ 270. Anal. found for C17H18O3 (calcd.)%C: 75.32 (75.53), %H: 6.53 (6.71). Pure 13a (7.70 mg,17% yield) was obtained from a center-cut of a band. The1H and 13C NMR data of 13a was found to match litera-ture values [18]. 13a 1H NMR (400 MHz): δ, ppm 7.21-7.31 (m, 5H, aryl C–H), 7.04 (d, 2H, JH = 8.4, aryl C–H)6.78 (d, 2H, JH = 8.4, aryl C–H) 3.81 (dd, 1H, JH = 8.8,JH = 6.8, methine C–H) 3.78 (s, 3H, ArOCH3), 3.61 (s,3H, ROCH3), 3.36 (dd, 1H, JH = 13.8, JH = 8.8, ArCH2R),2.98 (dd, 1H, JH = 13.8, JH = 6.8, ArCH2R). 13C NMR(100.5 MHz): δ, ppm 174.0, 158.1, 138.7, 131.1, 130.0,128.7, 128.0, 127.4, 113.8, 55.2, 53.9, 52.0, 39.0. MS (EI): m/z 270. Calcd. for [M]+ 270.

Thermal rearrangement of 12a. Compound 12a(10.2 mg, 0.0378 mmol) was dissolved in 5 mL ofp-methylanisole and the solution stirred at 110 °C for3 h. All the solvent was removed under reduced pressureand the ring-opened product 14a was purified usingsilica gel chromatography and hexane/ethyl acetate 30:1to afford 9.90 mg (97% yield). 1H NMR (400 MHz): δ,ppm 7.28–7.38 (m, 5H, aryl C–H), 7.05 (dd, 1H, JH =8.4, JH = 2.0, aryl C–H), 6.86 (d, 1H, JH = 2.0, aryl C–H),

6.78 (d, 1H, JH = 8.4, ring C–H), 5.31 (s, 1H, methineC–H), 3.83 (s, 3H, ArOCH3), 3.74 (s, 3H, ROCH3), 2.23(s, 3H, ArCH3).

13C NMR (100.5 MHz): δ, ppm 173.6,154.8, 137.9, 129.8, 129.1, 128.8, 128.6, 127.4, 127.2,110.5, 55.7, 52.3, 50.8, 20.7. MS (EI): 270 [M]+. Anal.found for C17H18O3 (calcd.) %C: 75.41 (75.53), %H: 6.64(6.71).

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith p-methylanisole. The general procedure was usedwith methyl 2-( p-chlorophenyl)diazoacetate (30.3 mg,0.144 mmol), Fe(TPP)Cl (2.00 mg, 1.97 mol.%), and 5.0mL of p-methylanisole. The mixture was stirred at 60 °Cfor 32 h. The benzylic C–H insertion product 13b and 7-carbomethoxy-2-methoxy-5-methyl-7-( p-chlorophenyl)norcaradiene, 12b, were partially separated using silicagel chromatography and hexane/ethyl acetate 30:1 aseluent. A combined yield of 32.8 mg, 0.107 mmol, 74%yield based on methyl 2-( p-chlorophenyl)diazoacetatewas obtained. Pure 12b (13.4 mg, 31% yield) wasobtained from a center-cut of a band. 1H NMR (400MHz): δ, ppm 7.17 (d, 2H, JH = 8.0, aryl C–H), 6.96 (m,2H, JH = 8.0, aryl C–H), 5.46 (d, 1H, JH = 6.6, vinylC–H), 4.75 (d, 1H, JH = 6.6, vinyl C–H), 3.64 (s, 3H,ArOCH3), 3.59 (s, 3H, ROCH3), 3.04 (d, 1H, JH = 9.2,H1,6), 2.95 (d, 1H, JH = 9.2, H1,6), 2.03 (s, 3H, RCH3).

13CNMR (100.5 MHz): δ, ppm 176.4, 154.1, 133.1, 133.0,131.3, 127.8, 124.2, 121.7, 95.9, 55.3, 53.0, 46.8, 36.6,24.2, 23.1. Anal. found for C17H17ClO3 (calcd.) %C:66.46 (67.00), %H: 5.53 (5.62). Pure 13b (8.30 mg, 19%yield) was obtained from a center-cut of a band and usedfor spectroscopic analysis. 1H NMR (400 MHz): δ, ppm7.28 (d, 2H, JH = 8.4, aryl C–H), 7.23 (d, 2H, JH = 8.4,aryl C–H), 7.01 (d, 2H, JH = 8.6, aryl C–H), 6.78 (d, 2H,JH = 8.6, aryl C–H), 3.78 (dd, 1H, overlapped with CH3),3.78 (s, 3H, ArOCH3), 3.62 (s, 3H, ROCH3), 3.33 (dd,1H, JH = 13.6, JH = 8.4, ArCH2R), 2.94 (dd, 1H, JH =13.6, JH = 7.2, ArCH2R). 13C NMR (100.5 MHz): δ, ppm173.6, 158.2, 137.1, 133.3, 130.6, 130.0, 129.9, 129.4,128.8, 113.8, 55.2, 53.3, 52.2, 39.0. MS (EI): m/z 305.Calcd. for [M + 1]+ 305.

Thermal rearrangement of 12b. Compound 12b(11.2 mg, 0.0368 mmol) was dissolved in 5 mL ofp-methylanisole and the solution stirred at 100 °C for3 h. All the solvent was removed under reduced pressureand the ring-opened product 14b was purified usingsilica gel chromatography and hexane/ethyl acetate30:1 as eluent to afford 10.8 mg (96% yield). 1H NMR(400 MHz): δ, ppm 7.31 (d, 2H, JH = 8.8, aryl C–H), 7.26(d, 2H, JH = 8.8, aryl C–H), 7.07 (dd, 1H, JH = 8.2, JH =2.0, aryl C–H), 6.87 (d, 1H, JH = 2.0, aryl C–H), 6.79 (d,1H, JH = 8.2, aryl C–H), 5.27 (s, 1H, methine C–H), 3.80(s, 3H, ArOCH3), 3.74 (s, 3H, ROCH3), 2.25 (s, 3H,ArCH3).

13C NMR (100.5 MHz): δ, ppm 173.3, 154.7,136.6, 133.1, 130.5, 130.0, 129.6, 129.4, 129.4, 129.0,128.7, 127.0, 55.7, 52.4, 50.2, 20.7. MS (EI): m/z 305.Calcd. for [M + 1]+ 305. Anal. found for C17H17ClO3

(calcd.) %C: 66.53 (67.00), %H: 5.56 (5.62).



Reaction of methyl 2-( p-tolyl)diazoacetate with p-methylanisole. The general procedure was used withmethyl 2-( p-tolyl)diazoacetate (30.6 mg, 0.161 mmol),Fe(TPP)Cl (2.30 mg, 2.03 mol.%), and 5.0 mL of p-methylanisole. The mixture was stirred at 80 °C for 32 h.The cyclopropanation product 7-carbomethoxy-2-methoxy-5-methyl-7-( p-tolyl)norcaradiene, 12c and thebenzylic C–H insertion product 13c were partially sepa-rated using silica gel chromatography and hexane/ethylacetate 30:1 as eluent. A combined yield of 35.2 mg,0.124 mmol, 77% yield based on methyl 2-( p-tolyl)dia-zoacetate were obtained. Pure 12c (12.4 mg, 27% yield)was obtained from a center-cut of a band. 1H NMR (400MHz): δ, ppm 7.01 (d, 2H, JH = 7.6, aryl C–H), 6.93 (d,2H, JH = 7.6, aryl C–H), 5.46 (d, 1H, JH = 7.0, vinylC–H), 4.76 (d, 1H, JH = 7.0, vinyl C–H), 3.64 (s, 3H,ROCH3), 3.60 (s, 3H, ROCH3), 3.03 (d, 1H, JH = 9.4,H1,6), 2.94 (d, 1H, JH = 9.4, H1,6), 2.30 (s, 3H, ArCH3),2.04 (b, 3H, RCH3).

13C NMR (100.5 MHz): δ, ppm177.1, 154.3, 136.6, 131.5, 129.7, 128.3, 124.5, 121.4,95.7, 55.3, 52.9, 40.3, 36.6, 24.5, 23.1, 21.4. MS (EI): m/z284. Calcd. for [M + 1]+ 284. Pure benzylic C–H product,13c (8.90 mg, 19% yield) was obtained from a center-cutof a band. 1H NMR (400 MHz): δ, ppm 7.20 (d, 2H, JH =8.0, aryl C–H), 7.13 (d, 2H, JH = 8.0, aryl C–H), 7.05 (d,2H, JH = 8.0, aryl C–H), 6.79 (d, 2H, JH = 8.0, aryl C–H),3.79 (m, 1H, overlapped with CH3), 3.78 (s, 3H,ArOCH3), 3.60 (s, 3H, ROCH3), 3.35 (dd, 1H, JH = 13.8,JH = 8.8, ArCH2R), 2.96 (dd, 1H, JH = 13.8, JH = 6.6,ArCH2R), 2.33 (s, 3H, ArCH3).

13C NMR (100.5 MHz):δ, ppm 174.1, 158.1, 137.0, 131.3, 129.9, 129.4, 127.8,113.7, 55.2, 53.5, 52.0, 39.0, 21.1. MS (EI): m/z 284.Calcd. for [M + 1]+ 284.

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with p-methylanisole. The general procedure wasused with methyl 2-( p-methoxyphenyl)diazoacetate(30.5 mg, 0.148 mmol), Fe(TPP)Cl (2.10 mg, 2.02mol.%), and 5.0 mL of p-methylanisole. The mixturewas stirred at 80 °C for 32 h. The cyclopropanationproduct 7-carbomethoxy-2-methoxy-5-methyl-7-( p-methoxyphenyl)norcaradiene, 12d and the benzylic C–Hinsertion product 13d were partially separated using sil-ica gel chromatography and hexane/ethyl acetate 30:1 aseluent. A combined yield of 34.6 mg, 0.115 mmol, 78%yield based on methyl 2-( p-tolyl)diazoacetate wasobtained. A sample of 12d containing about 10% of 13dwas used for spectroscopic analysis. 1H NMR (400MHz): δ, ppm 6.95 (d, 2H, JH = 8.2, aryl C–H), 6.73 (d,2H, JH = 8.2, aryl C–H), 5.47 (d, 1H, JH = 6.8, vinylC–H), 4.76 (d, 1H, JH = 6.8, vinyl C–H), 3.77 (s, 3H,ArOCH3), 3.64 (s, 3H, ROCH3), 3.59 (s, 3H, ROCH3),3.03 (d, 1H, JH = 9.6, H1,6), 2.93 (d, 1H, JH = 9.6, H1,6),2.03 (s, 3H, R-CH3). MS (EI): m/z 300. Calcd. for [M + 1]+

300. Pure 13d (15.6 mg, 35%) was obtained from a cen-ter-cut of a band. The 1H and 13C NMR data of 13dwas found to match literature values [18]. 13d 1H NMR(400 MHz): δ, ppm 7.23 (d, 2H, JH = 8.4, aryl C–H),

7.03 (d, 2H, JH = 8.8, aryl C–H), 6.85 (d, 2H, JH = 8.8,aryl C–H), 6.78 (d, 2H, JH = 8.4, aryl C–H), 3.80 (s, 3H,ArOCH3), 3.77 (s, 3H, ArOCH3), 3.76 (dd, 1H, JH = 8.8,JH = 6.8, methine C–H), 3.60 (s, 3H, ROCH3), 3.33 (dd,1H, JH = 14.0, JH = 8.8, ArCH2R), 2.94 (dd, 1H, JH =14.0, JH = 6.8, ArCH2R). 13C NMR (100.5 MHz): δ, ppm174.3, 158.9, 158.1, 131.2, 130.8, 130.2, 130.0, 129.0,114.0, 113.7, 55.3, 55.2, 53.0, 39.1. MS (EI): m/z 300.Calcd. for [M + 1]+ 300.

Reaction of methyl 2-phenyldiazoacetate with p-chlorotoluene. The general procedure was used withmethyl 2-phenyldiazoacetate (30.3 mg, 0.172 mmol),Fe(TPP)Cl (2.40 mg, 1.98 mol.%) was dissolved in 5.0mL of p-chlorotoluene. The mixture was stirred at 80 °Cfor 32 h. The products were partially separated by elut-ing through a silica gel column using a 30:1 hexane/ethylacetate mixture. The cyclopropanation product cyclo-propanation product 7-carbomethoxy-2-chloro-5-ethyl-7-phenylnorcaradiene, 16a and the benzylic C–H insertionproduct 17a were partially separated using silica gel col-umn and hexane/ethyl acetate 30:1 as eluent: A com-bined yield of 29.3 mg, 0.107 mmol, 62% yield based onmethyl 2-phenyldiazoacetate were obtained. Pure 16a(4.20 mg, 9% yield) was obtained from a center-cut of aband. 1H NMR (400 MHz): δ, ppm 7.23 (br, 3H, arylC–H), 7.09 (br, 2H, aryl C–H), 5.74 (d, 1H, JH = 6.6,vinyl C–H), 5.46 (d, 1H, JH = 6.6, vinyl C–H), 3.67 (s,3H, ROCH3), 3.20 (d, 1H, JH = 9.2, H1,6), 3.00 (s, 1H, JH

= 9.2, H1,6), 2.09 (s, 3H, R–CH3).13C NMR (100.5 MHz):

δ, ppm 176.4, 132.4, 132.0, 131.7, 128.2, 127.7, 127.5,123.1, 121.4, 53.2, 42.1, 41.2, 23.6. MS (EI): m/z 275.Calcd. for [M + 1]+ 275. Benzylic C–H product 17a (15.7mg, 33% yield) of containing trace impurities wasobtained from a center-cut of a band. 1H NMR (400MHz): δ, ppm 7.23–7.35 (m, 5H, aryl C–H), 7.21 (d, 2H,JH = 7.8, aryl C–H), 7.04 (d, 2H, JH = 7.8, aryl C–H),3.81 (dd, 1H, JH = 8.6, JH = 6.6, methine C–H), 3.62 (s,3H, ArOCH3), 3.38 (dd, 1H, JH = 13.6, JH = 8.6,ArCH2R), 3.00 (dd, 1H, JH = 13.6, JH = 6.6, ArCH2R).13C NMR (100.5 MHz): δ, ppm 173.7, 138.3, 137.5,132.2, 130.4, 128.8, 128.5, 128.0, 127.6, 53.5, 52.1,39.1. MS (EI): m/z 275. Calcd. for [M + 1]+ 275.

Reaction of methyl 2-( p-chlorophenyl)diazoacetatewith p-chlorotoluene. The general procedure was usedwith methyl 2-( p-chlorophenyl)diazoacetate (30.4 mg,0.144 mmol), Fe(TPP)Cl (2.00 mg, 1.97 mol.%), and5.0 mL of p-chlorotoluene. The mixture was stirred at80 °C for 32 h. The products were partially separated byeluting through a silica gel column using a 30:1hexane/ethyl acetate mixture. The 7-carbomethoxy-2-chloro-5-methyl-7-( p-chlorophenyl)norcaradiene, 16band the benzylic C–H insertion product 17b were par-tially separated using silica gel column and hexane/ethylacetate 30:1 as eluent. A combined yield of 25.9 mg,0.0841 mmol, 58% yield based on methyl 2-( p-chlorophenyl)diazoacetate) was obtained. Pure 16b (3.20mg, 7% yield) was obtained from a center-cut of a band.



1H NMR (400 MHz): δ, ppm 7.21 (d, 2H, JH = 8.6, arylC–H), 7.01 (d, 2H, JH = 8.6, aryl C–H), 5.76 (d, 1H, JH =6.8, vinyl C–H), 5.48 (d, 1H, JH = 6.8, vinyl C–H), 3.68(s, 3H, ROCH3), 3.18 (d, 1H, JH = 9.2, H1,6), 2.99 (d, 2H,JH = 9.2, H1,6), 2.08 (b, 3H, RCH3).

13C NMR (100.5MHz): δ, ppm 175.9, 133.3, 133.0, 132.2, 130.4, 128.1,128.0, 123.3, 121.7, 53.3, 42.1, 41.1, 23.6. MS (EI): m/z309. Calcd. for [M + 1]+ 309. A mixture 17b (14.2 mg,32%) containing trace impurities was obtained. 1H NMR(300 MHz): δ, ppm 7.29 (d, 2H, JH = 8.4, aryl C–H), 7.21(d, 4H, JH = 8.4, aryl C–H), 7.02 (d, 2H, JH = 8.4, arylC–H), 3.78 (dd, 1H, JH = 8.1, JH = 7.2, methine C–H),3.62 (s, 3H, ROCH3), 3.35 (dd, 1H, JH = 13.8, JH = 8.1,ArCH2R), 2.97 (dd, 1H, JH = 13.8, JH = 7.2, ArCH2R). 13CNMR (75.4 MHz): δ, ppm 173.5, 137.2, 133.7, 131.1,130.5, 129.7, 129.5, 129.1, 128.8, 53.0, 52.4, 39.2. MS (EI): m/z 309. Calcd. for [M + 1]+ 309.

Reaction of methyl 2-( p-tolyl)diazoacetate with p-chlorotoluene. The general procedure was used withmethyl 2-( p-tolyl)diazoacetate (30.6 mg, 0.161 mmol),Fe(TPP)Cl (2.30 mg, 2.03 mol.%), and 5.0 mL of p-chlorotoluene. The mixture was stirred at 80 °C for 32 h.The cyclopropanation product 7-carbomethoxy-2-chloro-5-methyl-7-( p-methylphenyl)norcaradiene, 16c and thebenzylic C–H insertion product 17c were partially sepa-rated using silica gel chromatography and hexane/ethylacetate 30:1 as eluent. A combined yield of 29.3 mg,0.103 mmol, 64% yield based on methyl 2-( p-tolyl)dia-zoacetate was obtained. It was not possible to get pure16c (1H and 13C NMR data were assigned by subtractingthe spectrum of pure 17c from that of the mixture). 1HNMR (300 MHz): δ, ppm 7.13 (d, 2H, JH = 8.1, arylC–H), 6.95 (d, 2H, JH = 8.1, aryl C–H), 5.75 (d, 1H, JH =6.6, vinyl C–H), 5.47 (d, 1H, JH = 6.6, vinyl C–H), 3.60 (s,3H, ROCH3), 3.18 (d, 1H, JH = 9.0, H1,6), 3.00 (d, 2H, JH =9.0, H1,6), 2.31 (s, 3H, ArCH3), 2.09 (b, 3H, RCH3).

13CNMR (75.4 MHz): δ, ppm 176.8, 137.3, 132.6, 131.6,129.0, 128.4, 123.3, 121.6, 53.4, 42.4, 41.5, 23.7, 21.5. MS (EI): m/z 289. Calcd. for [M + 1]+ 289. Pure benzylicC–H product 17c (5.90 mg, 13%) was obtained from a cen-ter-cut of a band. 1H NMR (400 MHz): δ, ppm 7.22 (d, 2H,JH = 8.4, aryl C–H), 7.19 (d, 2H, JH = 8.4, aryl C–H), 7.13(d, 2H, JH = 8.0, aryl C–H), 7.06 (d, 2H, JH = 8.0, arylC–H), 3.78 (dd, 1H, JH = 8.8, JH = 6.8, methine C–H), 3.62(s, 3H, ROCH3), 3.36 (dd, 1H, JH = 14.0, JH = 8.8,ArCH2R), 3.00 (dd, 1H, JH = 14.0, JH = 6.8, ArCH2R),2.34 (s, 3H, ArCH3).

13C NMR (100.5 MHz): δ, ppm173.8, 137.6, 137.3, 135.3, 132.2, 130.4, 129.5, 128.5,127.8, 53.1, 52.1, 39.1, 21.1. MS (EI): m/z 289. Calcd. for[M + 1]+ 289.

Reaction of methyl 2-( p-methoxyphenyl)diazoac-etate with p-chlorotoluene. The general procedure wasused with methyl 2-( p-methoxyphenyl)diazoacetate(30.4 mg, 0.148 mmol), Fe(TPP)Cl (2.10 mg, 2.02mol.%), and 5.0 mL of p-chlorotoluene. The mixture wasstirred at 80 °C for 32 h. A cyclopropanation product7-carbomethoxy-2-chloro-5-methyl-7-( p-methoxyphenyl)

norcaradiene, 16d and a benzylic C–H insertion product17d were partially separated using silica gel chromatog-raphy and hexane/ethyl acetate 30:1 as eluent. A com-bined yield of 24.7 mg, 0.0813 mmol, 55% yield basedon methyl 2-( p-methoxyphenyl) diazoacetate wasobtained. It was not possible to isolate pure 16d (1H and13C NMR data were assigned by subtracting the spectrumof pure 17d from that of the mixture). 1H NMR (400MHz): δ, ppm 7.00 (d, 2H, JH = 8.4, aryl C–H), 6.78 (d,2H, JH = 8.4, aryl C–H), 5.76 (d, 1H, JH = 6.4, vinyl-H),5.47 (d, 1H, JH = 6.4, vinyl-H), 3.77 (s, 3H, ArOCH3), 3.68(s, 3H, ROCH3), 3.17 (d, 1H, JH = 9.6, H1,6), 2.97 (d, 1H,JH = 9.6, H1,6), 2.08 (s, 3H, RCH3).

13C NMR (100.5MHz): δ, ppm 176.7, 158.7, 132.7, 128.2, 123.9, 123.2,121.5, 113.2, 55.1, 53.2, 42.2, 41.3, 24.8, 23.6. Pure ben-zylic C–H product 17d (6.30 mg, 14%) was obtained froma center-cut of a band. 1H NMR (300 MHz): δ, ppm 7.20(m, 4H, aryl C–H), 7.03 (d, 2H, JH = 8.7, aryl C–H), 6.85(d, 2H, JH = 8.7, aryl C–H), 3.80 (s, 3H, ArOCH3), 3.75(dd, 1H, JH = 8.4, JH = 6.4, methine C–H), 3.61 (s, 3H,ROCH3), 3.35 (dd, 1H, JH = 14.0, JH = 8.4, ArCH2R), 2.97(dd, 1H, JH = 14.0, JH = 6.4, ArCH2R). 13C NMR (75.4MHz): δ, ppm 173.9, 159.0, 137.6, 132.2, 130.4, 130.3,129.0, 128.5, 114.1, 55.3, 52.6, 52.1, 39.2. MS (EI): m/z305. Calcd. for [M + 1]+ 305. Anal. found for C17H17ClO3

(calcd.) %C: 66.56 (67.00), %H: 5.83 (5.62).Reaction of 7-carbomethoxy-7-( p-chlorophenyl)

norcaradiene/7-carbomethoxy-7-( p-chlorophenyl)cycloheptatriene, 2b/2′b, with benzyne. The generalprocedure was used with 2b/2′b (20.6 mg, 0.0792mmol), 2-(trimethylsilyl)phenyl triflate (28.1 mg, 0.0951mmol, 1.2 equiv.), and CsF (24.1 mg, 0.158 mmol, 2equiv.). The product was isolated by eluting through asilica gel column using a 20:1 hexane/ethyl acetate mix-ture. Pure benzhomobarrelene 18b (25.8 mg, 0.0768mmol, 97% yield) was obtained as a white solid. 1HNMR (400 MHz): δ, ppm 7.26 (m, 4H, aryl C–H), 7.05(m, 4H, aryl C–H), 5.54 (m, 2H, vinyl C–H), 4.26 (br,2H, methine-H), 3.53 (s, 3H, ROCH3), 2.33 (t, 2H, JH =2.0, H1,6).

13C NMR (100.5 MHz): δ, ppm 173.0, 147.2,136.5, 134.6, 132.3, 128.4 125.0, 123.4, 57.0, 52.9, 47.4,41.5, 36.1. MS (EI): m/z 336. Calcd. for [M]+ 336. Anal.found for C21H17ClO2 (calcd.) %C: 74.62 (74.89), %H:5.15 (5.09).

Reaction of 7-carbomethoxy7-( p-tolyl)norcara-diene/7-carbomethoxy7-( p-tolyl)cycloheptatriene,2c/2′c, with benzyne. The general procedure was usedwith 2c/2′c (10.3 mg, 0.0429 mmol), 2-(trimethylsilyl)phenyl triflate (15.2 mg, 0.0515 mmol, 1.2 equiv.), andCsF (13.1 mg, 0.0862 mmol, 2 equiv.). The productwas isolated by eluting through a silica gel columnusing a 20:1 hexane/ethyl acetate mixture to producepure benzhomobarrelene 18c (12.7 mg, 0.0402 mmol,94% yield) as a white solid. 1H NMR (400 MHz): δ,ppm 7.22 (dd, 2H, JH = 5.2, JH = 3.2, aryl C–H), 7.09(d, 2H, JH = 7.8, aryl C–H), 7.02 (dd, 2H, JH = 5.2,JH = 3.2, aryl C–H), 6.99 (d, 2H, JH = 7.8, aryl C–H),



5.51 (dd, 2H, JH = 4.4, JH = 3.2, vinyl C–H), 4.24 (m,2H, methine-H), 3.52 (s, 3H, ROCH3), 2.39 (s, 3H,ArCH3), 2.33 (t, 2H, JH = 2.2, H1,6).

13C NMR (100.5MHz): δ, ppm 173.5, 147.4, 136.0, 133.9, 130.7, 128.8124.6, 123.1, 52.6, 41.3, 35.7, 21.3. MS (EI): m/z 316.Calcd. for [M]+ 316. Anal. found for C22H20O2 (calcd.)%C: 83.36 (83.51), %H: 6.39 (6.37).

Reaction of 7-carbomethoxy-2-chloro-7-( p-chloro-phenyl)norcaradiene/7-carbomethoxy-2-chloro-7-( p-chlorophenyl)cycloheptatriene, 3b/3′b, withbenzyne. The general procedure was used with 3b/3′b(15.4 mg, 0.0524 mmol), 2-(trimethylsilyl)phenyl triflate(18.6 mg, 0.0629 mmol, 1.2 equiv.), and CsF (15.9 mg,0.105 mmol, 2 equiv.). The product was isolated by elut-ing through a silica gel column using a 20:1 hexane/ethylacetate mixture. Pure product 19b (18.6 mg, 0.0503mmol, 96% yield) was obtained as a colorless oil. 1HNMR (300 MHz): δ, ppm 7.22–7.38 (m, 4H, aryl C–H),7.00–7.21 (m, 4H, aryl C–H), 5.57 (m, 2H, vinyl C–H),4.23 (m, 1H, methine-H), 3.56 (s, 3H, ROCH3), 2.67(dd, 1H, JH = 7.2, JH = 0.6, H1,6), 2.52 (dd, 1H, JH =7.2, JH = 2.7, H1,6).

13C NMR (75.4 MHz): δ, ppm 172.0,146.1, 145.1, 138.7, 135.1, 132.7, 131.8, 128.7, 128.5,125.7, 122.8, 121.1, 69.6, 52.9, 47.8, 42.8, 40.6,37.4. MS (EI): m/z 371. Calcd. for [M + 1]+ 371. Anal.found for C21H16Cl2O2 (calcd.) %C: 67.66 (67.94), %H:4.56 (4.34).

Reaction of 7-carbomethoxy-3-chloro-7-( p-chloro-phenyl)norcaradiene/7-carbomethoxy-3-chloro-7-( p-chlorophenyl)cycloheptatriene, 4b/4′b, with ben-zyne. The general procedure was used with 4b/4′b (17.9mg, 0.0609 mmol), 2-(trimethylsilyl)phenyl triflate (21.6mg, 0.0731 mmol, 1.2 equiv.), and CsF (18.4 mg,0.121 mmol, 2 equiv.). The product was isolated by elut-ing through a silica gel column using a 20:1 hexane/ethylacetate mixture. Pure product 20b (21.8 mg, 0.0589mmol, 97% yield) was obtained as a colorless oil. 1HNMR (400 MHz): δ, ppm 7.20–7.34 (m, 4H, aryl C–H),7.02–7.11 (m, 4H, aryl C–H), 5.42 (dd, 1H, JH = 7.2, JH

= 2.6,vinyl C–H), 4.30 (dt, 1H, JH = 6.8, JH = 2.0,methine-H), 4.15 (q, 1H, JH = 2.4, methine-H), 3.53 (s,3H, ROCH3), 2.38 (t, 2H, JH = 2.4, H1,6).

13C NMR(100.5 MHz): δ, ppm 172.5, 145.4, 135.9, 134.3, 132.3,132.1, 131.6, 131.3 128.7, 127.9, 125.1, 123.5, 123.2,52.8, 49.2 47.6, 42.1, 35.7, 35.1. MS (EI): m/z 370.Calcd. for [M]+ 370.

Reaction of 7-carbomethoxy-2-methyl-7-phenyl-norcaradiene/7-carbomethoxy-2-methyl-7-phenylcy-cloheptatriene, 6a/6′a, with benzyne. The generalprocedure was used with 6a/6′a (7.40 mg, 0.0308 mmol),2-(trimethylsilyl)phenyl triflate (11.0 mg, 0.0370 mmol,1.2 equiv.), and CsF (9.40 mg, 0.0618 mmol, 2 equiv.).The product was isolated by eluting through a silica gelcolumn using a 20:1 hexane/ethyl acetate mixture. Pureproduct 21a (9.30 mg, 0.0294 mmol, 96% yield based ondiene) was obtained as a colorless oil. 1H NMR (400MHz): δ, ppm 7.22–7.34 (m, 5H, aryl C–H), 7.00–7.19

(m, 4H, aryl C–H), 5.51 (t, 1H, JH = 6.8, vinyl C–H),5.26 (m, 1H, JH = 7.2, vinyl C–H), 4.22 (m, 1H, methine-H), 3.53 (s, 3H, ROCH3), 2.43 (dd, 1H, JH = 9.2, JH = 4.4,H1,6), 2.11 (d, 2H, JH = 9.2, H1,6), 1.92 (s, 3H, CH3).

13CNMR (100.5 MHz): δ, ppm 173.4, 138.5, 137.7, 134.0,131.1, 130.9, 128.1, 128.0, 126.3, 124.3, 122.9, 120.1,52.6, 48.1, 44.1, 41.5, 37.5, 33.9, 20.0. MS (EI): m/z 316.Calcd. for [M]+ 316.

Acidification of 7-carbomethoxy-7-( p-tolyl)norcaradiene/7-carbomethoxy-7-( p-tolyl)cyclohepta-triene, 2c/2′c. Valence isomers 2c/2′c (14.2 mg, 0.0592mmol) were dissolved in 3 mL of acetonitrile. Threedrops of concentrated sulfuric acid were added to thesolution and the mixture was warmed to 60 °C while stir-ring. After allowing the mixture to cool, 5 mL of waterwere added and the product was extracted with methyl-ene chloride (3 × 5 mL). Removal of CH2Cl2 under reducedpressure afforded methyl 2-phenyl-2-( p-tolyl)benze-neacetate, 22c. The 1H and 13C NMR data was found tomatch literature values [19].

Acidification of 7-carbomethoxy-2,5-dimethyl-7-( p-chlorophenyl)norcaradiene, 10b. 7-carbomethoxy-2,5-dimethyl-7-( p-chlorophenyl)norcaradiene 10b (15.8mg, 0.0590 mmol) was dissolved in 3 mL of acetonitrile.Three drops of concentrated sulfuric acid were added tothe solution and the mixture was warmed to 60 °C whilestirring. After cooling to ambient temperature, 5 mL ofwater were added and the product was extracted withmethylene chloride (3 × 5 mL). Removal of CH2Cl2

under reduced pressure afforded methyl 2-(2,5-dimethylphenyl)-2-( p-chlorophenyl)benzeneacetate23b (15.2 mg, 0.0567 mmol, 96% yield). 1H NMR (400MHz): δ, ppm 7.29 (d, 2H, JH = 8.6, aryl C–H), 7.17 (d,2H, JH = 8.6, aryl C–H), 7.08 (d, 1H, JH = 8.4, arylC–H), 7.03 (br, 1H, aryl C–H), 7.02 (d, 1H, JH = 8.4,aryl C–H), 5.16 (s, 1H, methine C–H), 3.76 (s, 3H,ROCH3), 2.31 (s, 3H, Ar–CH3), 2.22 (s, 3H, Ar–CH3).13C NMR (100.5 MHz): δ, ppm 173.1, 136.5, 136.2,136.0, 133.2 133.1, 130.8, 130.4, 128.7, 128.5, 128.4,53.0, 52.5, 21.2, 19.3. MS (EI): m/z 289. Calcd. for [M + 1]+ 289. Anal. found for C17H17ClO2 (calcd.) %C:70.34 (70.71), %H: 5.76 (5.93).


The iron(III) tetraphenylporphyrinato complexFe(TPP)Cl catalyzes the decomposition of para-substi-tuted methyl 2-phenyldiazoacetates, 1a–d, in benzeneto produce rapidly equilibrating mixtures of substitutednorcaradiene-cycloheptatriene adducts, 2a–d and 2′a–d(Equation 1, Table 1). This reaction required tempera-tures of 80 °C for 12 to 16 h with 2 mol.% catalyst. Nocarbene dimerization to form butenedioates wasobserved. The rapidly equilibrating valence isomerswere characterized by 1H NMR and 13C NMR spec-troscopy, mass spectrometry and elemental analysis.




For example, the formation of 7-carbomethoxy-7-phenylnorcaradiene, 2a, and its ring expansion valenceisomer 7-carbomethoxy-7-phenylcycloheptatriene, 2′a,was established by 1H NMR spectroscopy at 20 °C withthe appearance of diagnostic resonances for the olefinichydrogens at 6.28 (m, 2H) and 6.03 (m, 2H) ppm. Therapid equilibration of the two species was indicated bythe appearance of the 1,6-proton (see Equation 1 fornumbering) resonance at 4.31 ppm, a position interme-diate between those expected for the cyclopropylhydrogens of static 7-methoxycarbonyl-7-phenyldiben-zonorcaradiene (~3.67 ppm) [20] and the 1,6-hydro-gens of the static 7-carbomethoxycycloheptatriene(~5.43 ppm) [21]. Furthermore, the 13C NMR spectrumexhibited a broad signal at 72.3 for the 1,6-carbons.This also represented a time-averaged value due to theexchanging vinyl and cyclopropyl carbon atoms at the1,6-positions. Equilibration of the valence isomersremained rapid even at -60 °C. The composition of theproduct was verified further by its molecular mass of226 m/z.

When chlorobenzene was treated with methyl2-phenyldiazoacetate, 1a, and heated at 80 °C with2 mol.% Fe(TPP)Cl for 16 h, two regioisomers ofthe fluxional NCD–CHT products were produced(Equation 2). In all cases, positions 1 and 6 of thevalence isomers maintained a C–H group. Thus, theBüchner addition did not involve the chlorinated


Table 1. Summary of catalytic reactions of substituted methyl 2-phenyldiazoacetate compoundsa

Entry Substrate Catalyst Diazo Products Yields Ratio

1 benzene Fe(TPP)Cl 1a 2a/2′a 76b n/a2 benzene Fe(TPP)Cl 1b 2b/2′b 78b n/a3 benzene Fe(TPP)Cl 1c 2c/2′c 79b n/a4 benzene Fe(TPP)Cl 1d 2d/2′d 82b n/a5 C6H5Cl Fe(TPP)Cl 1a 3a/3′a:4a/4′a 78c 1:16 C6H5Cl Fe(TPP)Cl 1b 3b/3′b:4b/4′b 72c 1:17 C6H5Cl Fe(TPP)Cl 1c 3c/3′c:4c/4′c < 10d 1:18 C6H5Cl Fe(TPP)Cl 1d 3d/3′d:4d/4′d 33c 1:19 toluene Fe(TPP)Cl 1a 5a:6a/6′a:7a/7′a 77c 1:1.5:1.5

10 toluene Fe(TPP)Cl 1b 5b:6b/6′b:7b/7′b 68c 1:2:211 toluene Fe(TPP)Cl 1c 5c:6c/6′c:7c/7′c 75c 1:1:112 toluene Fe(TPP)Cl 1d 5d:6d/6′d:7d/7′d 79c 2:1:113 anisole Fe(TPP)Cl 1a 8a:9a 38c 1:0.714 anisole Fe(TPP)Cl 1b 8b:9b < 10d 1:1.315 anisole Fe(TPP)Cl 1c 8c:9c < 10d 1:116 anisole Fe(TPP)Cl 1d 8d:9d 42c 1:117 p-xylene Fe(TPP)Cl 1a 10a:11a 82c 2:118 p-xylene Fe(TPP)Cl 1b 10b:11b 83c 3:119 p-xylene Fe(TPP)Cl 1c 10c:11c 86c 1.1:120 p-xylene Fe(TPP)Cl 1d 10d:11d 88c 1:121 p-Me-anisole Fe(TPP)Cl 1a 12a:13a 74c 1.5:122 p-Me-anisole Fe(TPP)Cl 1b 12b:13b 74c 2:123 p-Me-anisole Fe(TPP)Cl 1c 12c:13c 77c 1.2:124 p-Me-anisole Fe(TPP)Cl 1d 12d:13d 78c 1:325 p-Cl-toluene Fe(TPP)Cl 1a 16a:17a 62c 1:326 p-Cl-toluene Fe(TPP)Cl 1b 16b:17b 58c 1:427 p-Cl-toluene Fe(TPP)Cl 1c 16c:17c 64c 1:228 p-Cl-toluene Fe(TPP)Cl 1d 16d:17d 55c 1:1

a Substrates used as solvent (5 mL), substituted methyl phenyldiazoacetates (0.19 mmol), (TPP)FeCl (2.5 mg, 2.0 mol.%), mixture thoroughly purged with dry nitrogen, heated at 100 °C for 12–16 h with stirring. b Isolatedyields. c Combined isolated yields of all products. d Detected but not isolated.

1 mol.%

80 ˚C.12-16 h

Fe(TPP)Cl+

ON2 O

O

O

O

O

XH = 1aCl = 1bMe = 1cOMe = 1d Yields > 75%

X

X

X2'a-d

+ N2

2a-d

12

3

45

6

7

12

3

4

5 6

7

(1)



double bond of the substrate. The ratio of the 2-chloro-products, 3a/3′a to the 3-chloro-isomers, 4a/4′a was1:1. Separation of the product mixture by silica gelchromatography using hexane/ethyl acetate (30:1) asthe eluent produced a pure sample of the 2-chloro iso-mers, 3a/3′a. The 3-chloro fraction, 4a/4′a, containedsome residual 2-chloro products and could not beobtained cleanly. The 1H NMR spectrum of 3a/3′aexhibited signals for the 1,6-protons at 3.30 (dd) and3.36 (d) ppm, respectively. Furthermore, the 13C NMRsignal of C7 appeared at 26.2 ppm while C1 and C6appeared at 43.0 and 44.6 ppm, respectively. In con-trast, the 1H NMR signals for H1 and H6 of 4a/4′aappeared together at 4.38 (m) ppm while 13C NMRsignals for C7 appeared at 38.7 and those of C1 and C6appeared as broad peaks at 74.7 and 73.5, respectively.This NMR data suggested that the equilibratingvalence isomers of 3a/3′a favor the norcaradiene formmore than is the case with 4a/4′a. Similar norcaradi-ene–cycloheptatriene adducts 3b/3′b and 4b/4′b wereobtained when chlorobenzene was treated under thesame conditions with p-Cl-MPDA, 1b. The reactionmixture was separated on a silica gel column to give apure sample of the 2-chloro isomers, 3b/3′b. The 3-chloro isomers 4b/4′b contained traces of 3b/3′b andcould not be purified further. Proton NMR signals forH1 and H6 appeared at 3.32 (d) and 3.25 (dd) ppm,respectively for the 2-chloro isomer, 3b/3′b while therelated protons appeared together at 4.25 (m) ppm forthe 3-chloro isomer, 4b/4′b. Although the C1 and C6

13C NMR signals for the 2-chloro isomer,3b/3′b, were detected as broad peaks at 43.7and 42.0, the related carbons were notdetectable for the 3-chloro isomer case 4b/4′b,presumably due to coalescence of the signalsinto the baseline. The reaction of chlorobenzenewith p-MeO-MPDA, 1d, was found to producelow yields of both regioisomers 3d/3′d and4d/4′d. Refluxing for 3 days using 2% catalystresulted in less than 50% conversion and about32% combined yield of the NCD–CHT isomers3d/3′d and 4d/4′d.

Toluene underwent concomitant C–H activa-tion and Büchner addition reactions with substitutedmethyl 2-phenyldiazoacetates to give benzylic insertionproducts, 5a–d, and the two regioisomers of the flux-ional NCD–CHT systems 6a–d/6′a–d and 7a–d/7′a–d(Equation 3). Although it was not possible to obtain allof the products in high purity, some were separatedcleanly. This provided sufficient samples to allowassignment of all the proton NMR signals for this fam-ily of products by comparative analysis. MPDA, 1a,reacted with toluene to give the benzylic insertionproduct methyl 2,3-diphenylpropionate, 5a and the flux-ional NCD–CHT products 7-carbomethoxy-2-methyl-7-phenylnorcaradiene/7-carbomethoxy-2-methyl-7-phenylcycloheptatriene, 6a/6′a and 7-carbomethoxy-3-methyl-7-phenylnorcaradiene/7-carbomethoxy-3-methyl-7-phenylcycloheptatriene, 7a/7′a. It was notpossible to cleanly isolate methyl 2,3-diphenylpropi-onate, 5a. The benzyl product contained ~20% of7a/7′a after eluting the reaction mixture through a sil-ica gel column with a hexane/ethyl acetate (30:1) sol-vent mixture. Nonetheless, the proton NMR resonancesof 5a were found to match literature values. The 2-methyl isomers 6a/6′a were obtained in pure formwhile the 3-methyl fraction 7a/7′a could not be fullyseparated from 6a/6′a. As in the case of the chloroben-zene products, the 1,6 protons of 6a/6′a appeared sepa-rately at 3.10 ppm and 3.24 ppm, respectively whilethose of 7a/7′a appear together as multiplets at 3.88ppm. These signals are all upfield relative to those inthe chloro analogues 4a/4′a. The electronic nature of

2 mol.%

Fe(TPP)Cl

O

O

O

O

X

X

Cl

Cl

Cl

2 (1a-d)

3'a-d

-2 N2

O

O

O

O

X

X4'a-d

Cl

Cl

+80 ˚C.

12-16 h

2

3a-d 4a-d

(2)

O

O

X

+

O

O

X

X

Fe(TPP)Cl

O

O

3 (1a-d)

5a-d

6'a-d

O

O

X

X

O

O

7'a-d

+3

2 mol.%

-3 N2

80 ˚C.16-24 h

6a-d 7a-d

(3)



the para-substituent of the substituted methyl 2-phenyl-diazoacetates influenced the benzyl insertion: Büchnerproduct ratios slightly. The less electron-donating chloridein methyl 2-( p-chlorophenyl)diazoacetate, 1b, resultedin Büchner addition as the major products,5b:6b/6′b:7b/7′b = 1:2:2. As the electron donor abilityof the aryl substituent on the diazo reagent increased,the ratio of the Büchner product decreased. For exam-ple, the p-methoxy substituted MPDA, 1d, producedmore benzylic C–H insertion (5d:6d/6′d:7d/7′d = 2:1:1,Table 1, entries 9–12).

Contrary to expectations, the more electron-richanisole substrate (Equation 4) was found to give lowyields of the Büchner-type products that subsequentlydecomposed to unknown compounds on standing. In thisreaction, GC analysis showed that most of the diazoreagent was not consumed after 12 h underrefluxing conditions. Column chromatographydid not result in clean separation of the productmixture. Unlike the chlorobenzene and toluenecases, products with methoxy substitution at posi-tion 2 gave 1,6-proton NMR signals at lower fieldas compared to those with substitution at position 3.For example, the 1,6-protons of 7-carbomethoxy-2-methoxy-7-phenylnorcaradiene/7-carbo-methoxy-2-methoxy-7-phenylcycloheptatriene,8a/8′a, gave NMR signals at 4.25 and 4.09 ppm,whereas the 1,6-hydrogens of the 3-methoxy iso-mers 9a/9′a appeared at 3.10 ppm.

Substituted methyl 2-phenyldiazoacetatesreacted with p-xylene to give both cyclopropana-tion and benzylic C–H insertion products(Equation 5). The equivalence of the vinyl methylgroups of the norcaradiene products indicated thatthese compounds retained mirror symmetry and

that the cyclopropanation occurred only at the 2,3-xylyl double bond. The 1,6-protons of 7-car-bomethoxy-2,5-dimethyl-7-phenylnorcaradiene,10a, gave 1H NMR signals at 2.94 ppm while 13CNMR signals for the 1,6-carbons appear at 41.4ppm. These data suggest that 10a favors the nor-caradiene form. Electron withdrawing groups onthe aryl of the diazo reagents favored cyclopropa-nation over benzylic insertion with the chloro sub-stituent giving a 10/11 product ratio of 3:1. As theelectron donating nature of the diazo substituent isincreased, this ratio decreased (Table 1 entries17–20). Thus, the p-methoxy-substituted MPDAproduced a 1:1 product ratio.

When p-methylanisole was treated with methyl2-phenyldiazoacetate at 80 °C for 16 h in thepresence of 2 mol.% Fe(TPP)Cl, both the cyclo-propanation product, 7-carbomethoxy-2-methoxy-5-methyl-7-phenylnorcaradiene, 12a, and thebenzylic C–H insertion product, methyl 3-( p-methoxyphenyl)-2-phenylpropionoate, 13a, wereobtained (Equation 6). In these reactions thevalence isomerization of 12a–d to the cyclohepta-

triene form appears to be minor as indicated by the 1HNMR resonances for the cyclopropyl protons. For exam-ple, in 12a, these signals were found at 2.96 (d, 1H) and3.05 (d, 1H). The higher consumption of the diazo reagentswith p-methylanisole relative to anisole (Table 1) is duein part to the competing benzylic insertion reaction. At110 °C, the reaction of p-methylanisole and MPDAafforded a ring C–H insertion product, methyl 2-(2-methoxy-5-methylphenyl)-2-phenylacetate, 14a, insteadof the cyclopropanation product (Equation 7). Similarly,p-Cl-MPDA, 1b, was found to yield the cyclopropanationproduct only when the reactions were done at 60 °C,otherwise a ring C–H insertion product 14b, formed athigher temperature. To probe this behavior further, pure 7-carbomethoxy-2-methoxy-5-methyl-7-phenylnorcaradiene,

OO

O

X

2 mol.%

8'a-d

OMe

MeO

O

O

X

9'a-d

MeO

+Fe(TPP)Cl

2 (1a-d)

-2 N2

80 ˚C.12-16 h

O

O

O

O

X X

MeO

2 8a-d 9a-d

(4)

O

O

X

10a-d

+

O

X

O

11a-d

2 mol.%Fe(TPP)Cl

2 (1a-d)

-2 N2

80 ˚C.12-16 h

2(5)

O

O

X12a-d

OMeOMe

+

OMe

O

X

O

13a-d

2 mol.%Fe(TPP)Cl

2 (1a-d)

-2 N2

80 ˚C.12-16 h

2 (6)

13a,b

MeO

OMe

+

OMe

O

X

O

14a,b

OO

X

2 mol.%Fe(TPP)Cl

2 (1a or 1b)

-2 N2

110 ˚C12-16 h

2(7)



12a, was dissolved in p-methylanisole and the mixture washeated at 110 °C for 6 h. Quantitative conversion to thediarylacetate, 14a, (Equation 8) was observed. Similarly,heating 7-carbomethoxy-7-( p-chlorophenyl)-2-methoxy-5-methylnorcaradiene, 12b, at 110 °C resulted in therearrangement to the product 14b.

Although the cyclopropane moiety in product 12a cantheoretically rearrange to two ring-opened products, 14aand 15a, only one isomer was observed by 1H NMRspectroscopy based on the appearance of a singlemethine signal at 5.31 ppm. The product stereochemistrywas confirmed by a 2D NOESY NMR experiment. Thearyl proton H3, appearing as a doublet at 6.87 ppm (seeEquation 8 for proton numbering) and definitivelyassigned by its small four-bond coupling (J = 2.0 Hz) toH5, interacted through space with the aryl methyl group(at C4) at 2.23 ppm. This NOE result is only possible inproduct 14a. In isomer 15a, the aryl proton H2 is too farto exhibit a NOE interaction with the aryl methyl groupat C4.

When p-chlorotoluene was stirred with methyl2-phenyldiazoacetate, 1a, at 80 °C for 16 h in thepresence of a 2 mol.% catalyst, both the cyclopropa-nation product, 7-carbomethoxy-2-chloro-5-methyl-7-phenylnorcaradiene, 16a, and the benzylic C–Hinsertion product 17a were obtained (Equation 9).Chloro and methyl substituted methyl 2-phenyl-diazoacetates gave similar products, but the proportionof the cyclopropanation product was found to behigher for the chloro substituent (Table 1 entries25–26). However, methyl 2-( p-methoxyphenyl)diazoacetate was found to give the cyclopropanationproduct 16d and benzylic insertion product 17d inalmost equal amounts.

Trapping fluxional dienes

To explore further these fluxional NCD–CHTsystems, trapping experiments with benzyne were under-taken. Benzyne was generated in situ using 2-(trimethylsilyl)phenyl triflate and CsF dissolved inacetonitrile [22] in the presence of the NCD–CHT sys-tems. In each case, a single isomer was produced inyields greater than 94%. 1H NMR spectroscopy allowedunambiguous characterization and identification of theproducts as 3-carbomethoxy-3-aryl-8,9-benzhomobar-ralenes, 18a–d, resulting from [2+4]-cycloaddition reac-tions (Equations 10–13). For example, after the reactionof benzyne with 7-carbomethoxy-7-( p-methylphenyl)-orcaradiene/cycloheptatriene valence isomers, 2c/2′c,the 1,6-hydrogen signals moved upfield from 4.32 ppmto 2.33 (m, 2H) ppm. The retention of chemical equiva-lence of these protons and their peak position indicatedthat the new product retained mirror symmetry and con-tained a cyclopropyl fragment. Furthermore, a newchemical shift at 4.24 (m, 2H) ppm is consistentwith bridgehead protons as reported previously for

MeO

14a,b

OOOMe O

O 110 ˚C

p-methylanisole

12a,b

OMe

O

OX

15a,b

H3

H2

H6

H5

O

O

X16a-dClCl

+

Cl

O

X

O

17a-d

2 mol.%Fe(TPP)Cl

2 (1a-c)

-2 N2

80 ˚C.12-16 h

2 (9)

(8)

O

O

O

O

X

X2'b,c

18b,c

2b,c OO

X

b: X = Clc: X = Me

O

O

O

O

Cl

Cl3'b

19b

3b

OO

Cl

Cl

Cl

Cl

(11)

(10)

O

O

O

O

Cl

Cl4'b

20b

4b OO

ClCl

Cl

Cl

(12)



benzhomobarrelene compounds [23]. The existence of amirror plane in the product ruled out [2+4] addition ofbenzyne to the cycloheptatriene form of the valence iso-mers. Moreover, all [2+2] adducts can be ruled out by thepresence of bridgehead protons in the product. This wassomewhat unexpected since the treatment of cyclohepta-triene with benzyne produced a [2+2] product along with7-phenylcycloheptatriene as a minor product [24].However, in the cases reported here, it is apparent thatbenzyne reacts preferentially with the norcaradiene formof the valence isomers. Precedence for the [2+4] prod-ucts is provided by the work of Abbasoglu et al. [25] inwhich the reaction of 7-carbomethoxycycloheptatrienewith benzyne resulted in the formation of only one ben-zhomobarrelene isomer. To our knowledge, benzhomo-barralenes 18a–d are the first examples withdisubstitution at the 3-position.

Two norcaradiene stereoisomers (I and III, Scheme 1)can interconvert with each other through the cyclo-heptatriene structure (II). Since these norcaradiene–cycloheptatriene systems are dynamic and the protonNMR signals of the products produce a simple time aver-aged spectrum, it is not possible to determine a prioriwhich norcaradiene isomer reacts with benzyne. An endoattack by benzyne on either norcaradiene is likely to bedisfavored due to steric reasons. The most likely benzho-mobarrelene products are isomers IV and V (Scheme 1).

A NOESY NMR experiment performed on benzhomo-barrelene 18c revealed that the cyclopropyl hydrogens at2.33 ppm interact through space with the methyl hydro-gens of the ester group. In addition, a through-space inter-action between the hydrogens of the p-tolyl group (6.99ppm) with the vinyl protons (5.51 ppm) was observed.These NOE relationships indicate that the benzhomobar-relenes adopt structure V.

Acidification of fluxional dienes

Acid treatment of the NCD–CHT valence isomersresulted in ring opening of the cyclopropane and forma-tion of diaryl acetates. For example, 7-carbomethoxy-7-( p-tolyl)-norcaradiene/7-carbomethoxy-7-( p-tolyl)cycloheptatriene, 2c/2′c, quantitatively converted to methyl2-phenyl-2-p-tolylacetate, 22c, upon acidification withsulfuric acid and warming to 60 °C in acetonitrilesolution (Equation 14). Similarly, 2,5-dimethyl-7-( p-chlorophenyl)-7-carbomethoxy norcaradiene/2,5-dimethyl-7-( p-chlorophenyl)-7-carbomethoxycyclohep-tatriene, 10b/10′b, readily converted to methyl2-(2,5-dimethylphenyl)-2-p-chlorophenylacetate (23b)upon acidification (Equation 15).

Mechanistic considerations

In the present study, MPDAs substituted with electronwithdrawing groups such as Cl tended to favor Büchneraddition to arenes whereas the electron rich methoxy sub-stituted MPDA preferred benzylic insertion products. Thus,an increase in the electrophilicity of the intermediate car-bene favors the cyclopropanation of arenes. This concurswith earlier research done by Anciaux et al. [2b] whichshowed that catalysis by tetrakis(carboxylato)dirhodium(II)complexes with carboxylate ligands of very strong organicacids such as trifluoroacetic and perfluorobenzoic acids ledto cyclopropanation of benzene and toluene with EDA inquantitative yields while the parent Rh2(OAc)4 catalystgave yields of less than 30%. Whether the addition of car-benoid to the aromatic molecule is concerted or takes placevia a stepwise ionic mechanism remains to be answered.

O

O

O

O

6'a

21a

6a OO

(13)

OO

O

OO

O

III III

O

O

OO

Exo attack

IVV

X

X

X

X

X = Cl, Me

Exo attack

X

Scheme 1.

OO H2SO4

CH3CN

OO

2c 22c

(14)

OO

Cl

H2SO4

CH3CN

OO

Cl10b 23b

(15)



However, if the addition were a stepwise ionic mechanism,one would expect some 1,2-hydrogen shift by-product thatwould produce ring C–H insertion products (Scheme 2)[26]. The lack of these hydrogen transfer products supportsa concerted mechanism.

Initial rates of catalysis with p-Cl-MPDA and benzenewere determined under pseudo-first-order conditionswith different diazo reagent concentrations. These rateswere found to be first order with respect to the concen-tration of the diazo compound ([diazo] = 0.095 mM, kobs

= 2.40 μM/h; 0.19 mM, 4.75 μM/h; 0.29 mM, 7.0 μM/h)and indicated that the formation of a metal carbene com-plex was the rate-determining step. This is consistentwith the metal complex catalyzed cyclopropanation viadiazocarbonyl reagents. These reactions are generallyassumed to involve an intermediate metal carbene com-plex generated by metal-mediated extrusion of nitrogenfrom the diazo compound, followed by a concertedcyclopropanation process [27, 28].

The NCD–CHT valence isomers readily convert todiarylacetates when acidified. Scheme 3 shows how a

1,2-hydrogen transfer ring opening mechanism is facili-tated by the addition of acid.

CONCLUSION

Fe(TPP)Cl is an effective catalyst for the cyclo-propanation of arenes using p-substituted methylphenyldiazoacetates as carbene sources. This reactionproduced valence isomeric norcaradiene–cyclo-heptatriene products. The compounds obtained fromthe insertion of carbenes derived from MPDAs gave1H NMR diagnostic resonances for olefinic hydrogensat approximately 6.3 ppm (m, 2H) and 6.0 (m, 2H)ppm and rapidly equilibrating 1,6-proton resonanceat 4.3 ppm when spectra are run at 20 °C. Whenthe same NMR sample was run at -60 °C, the olefinicprotons broadened, but remained at similar chemicalshifts. The 1,6-protons also broadened and movedto slightly higher field. This indicates that even at-60 °C these isomers are in rapid equilibration onthe NMR time scale, thus giving a time-averagedsignal due to the exchanging vinyl and cyclopropylcarbon atoms at the 1,6-positions. Insertion of MPDAinto chlorobenzene produces 2-chloro 3a/3′a and3-chloro 4a/4′a isomers. These regioisomers areeasily distinguished by their 1H NMR signals with1,6-proton resonances that are surprisingly verydifferent. The 1,6-proton signals of the 2-chloro 3a/3′aproduct appeared at 3.30 (dd ) and 3.36 (d ) ppm,respectively, while those of 4a/4′a appeared together at4.38 (m) ppm.

The fluxional nature of the NCD–CHT system wasconfirmed further by trapping with benzyne to givehigh yields (>90%) of one isomer of substituted ben-zhomobarralenes resulting from [2+4] cycloadditionreactions. Trapping 2-chloro 3b/3′b and 3-chloro4b/4′b with benzyne helped confirm the formation oftwo regioisomers. The benzhomobarralene productfrom trapping the 2-chloro 3b/3′b isomer exhibited abridgehead signal at 4.23 ppm (m) that integratedto one proton and a vinyl signal at 5.57 that integratedto two protons. However, trapping the 3-chloro4b/4′b isomer produced 20b which had two differentbridgehead signals at 4.15 ppm (m) and 4.30 ppm(m) integrating to one proton each and a vinyl signalat 5.42 ppm that integrated as one proton. Thesefluxional systems were also aromatized by acidifica-tion to afford biaryl products that are the same as onewould expect from carbene insertion into an aromaticC–H bond.

Studies on the initial rate of reactions for p-ClMPDA addition to benzene indicate that theformation of an iron carbene complex is the rate-limiting step. Absence of 1,2-hydrogen shift by-products suggested that a concerted mechanism isinvolved.

Fe

R1 R2

+ Fe

R1

R2

YX

Y

X

X

Y

R1

R2

X

Y

H

H

X

Y

+

cyclopropanation

1,2-hydrogen shift

R1

R2

R2

R1

Scheme 2.

O O

X

OO

X

H

H

H

O O

X

H H

Scheme 3.



Acknowledgment

The authors are grateful to the funding provided bythe Petroleum Research Fund administered by theAmerican Chemical Society and an award from theNational Science Foundation.

REFERENCES

1. a) Doyle MP, McKervey MA and Ye T. ModernCatalytic Methods for Organic Synthesis withDiazo Compounds, Wiley: New York, 1998;Chapter 6. b) Maas G. Top. Curr. Chem. 1987; 137:75–253. c) Davies HML. In ComprehensiveOrganic Synthesis, Trost BM (Ed.), Vol. 4,Pergamon: Oxford, 1991; Chapter 4.8.

2. a) Anciaux AJ, Demonceau A, Hubert AJ, NoelsAF, Petiniot N and Teyssié P. J. Chem. Soc. Chem.Commun. 1980; 765–766. b) Anciaux AJ,Demonceau AF, Noels AF, Hubert AJ, Warin R andTeyssié P. J. Org. Chem. 1981; 46: 873–876.

3. a) Kennedy M, McKervey MA, Maguire AR,Tuladhar SM and Twohig MF. J. Chem. Soc. PerkinTrans. 1990; 1: 1047–1054. b) Duddeck H, FergusoG, Kaitner B, Kennedy M, McKervey MA andMaguire AR. J. Chem. Soc. Perkin Trans. 1: 1990;1055–1063. c) Cordi AA, Lacoste J-M and HennigP. J. Chem. Soc. Perkin Trans. 1993; 1: 3–4. d)Manitto P, Monti D and Speranza G. J. Org. Chem.1995; 60: 484–485. e) Maguire AR, Buckley NR,O’Leary P and Ferguson G. J. Chem. Soc. Chem.Commun. 1996; 2595–2596. f) Manitto P, Monti D,Zanzola S and Speranza G. J. Org. Chem. 1997; 62:6658–6665. g) Moody CJ, Miah S, Slawin AMZ,Mansfield DJ and Richard IC. J. Chem. Soc. PerkinTrans. 1998; 1: 4067–4075. h) Maguire AR,Buckley NR, O’Leary P and Ferguson G. J. Chem.Soc. Perkin Trans. 1998; 1: 4077–4091. i) DoyleMP, Ene DG, Forbes DC and Pillow TH. J. Chem.Soc. Chem. Commun. 1999; 1691–1692.

4. Davies HML, Smith HD. Hu B, Klenzak SM andHegner FJ. J. Org. Chem. 1992; 57: 6900–6903.

5. Woods WG. J. Org. Chem. 1960; 23: 110.6. a) Hamaker CG, Mirafzal GA and Woo LK.

Organometallics 2001; 20: 5171–5176. b) Du G,Andrioletti B, Rose E and Woo LK. Organometallics2002; 21: 4490–4495.

7. a) Mirafzal GA, Cheng G and Woo LK. J. Am.Chem. Soc. 2002; 124: 176–177. b) Cheng G,Mirafzal GA and Woo LK. Organometallics 2003;22: 1468–1474.

8. a) Baumann LK, Mbuvi HM, Du G and Woo LK.Organometallics, 2007; 26: 3995–4002. b) Aviv Iand Gross Z. Synletters 2006; 951–953.

9. Mbuvi HM and Woo LK. Organometallics 2008;27: 637–645.

10. Nilvebrant L, Andersson K-E, Gillberg P-G,Stahl M and Sparf B. Eur. J. Pharmacol. 1997;327: 195.

11. Nilvebrant L, Hallen B and Larsson G. Life Sci.1997; 60: 1129.

12. Hughes B, Howat D, Lisle H, Holbrook M, JamesT, Gozzard N, Blease K, Hughes P, Kingaby R,Warrellow G, Alexander R, Head J, Boyd E, EatonM, Perry M, Wales M, Smith B, Owens R, CatterallC, Lumb S, Russell A, Allen R, Merriman M,Bloxham D and Higgs G. Br. J. Pharmacol. 1996;118: 1183.

13. a) Hoffmann I, Ehrhart G and Schmitt K. Arzneim.-Forsch. 1971; 21: 1045. b) Lowe JA III, HagemanDL, Drozda SE, McLean S, Bryce DK, CrawfordRT, Zorn S, Morrone J and Bordner J. J. Med.Chem. 1994; 37: 3789.

14. Pangborn AB, Giardello MA, Grubbs RH, RosenRK and Timmers FJ. Organometallics 1996; 15:1518.

15. Davies HML, Bruzinski P, Hutcheson DK,Kong N and Fall MJ. J. Am. Chem. Soc. 1996;118: 6897.

16. Hall GE and Roberts JD. J. Am. Chem. Soc. 1971;93: 2203

17. Xiao X, Morrell, A, Fanwick PE and Cushman M.Tetrahedron 2006; 62: 9705.

18. Davies HML and Jin Q. Tetrahedron: Asym. 2003;14: 941–949.

19. Davies HML, Stafford DG and Hansen T. Org. Lett.1999; 1: 233–236.

20. Bogdanova A and Popik VV. Org. Lett. 2001; 3:1885–1888.

21. Morilla ME, Diaz-Requejo MM, Belderrain TR,Nicasio MC, Trofimenko S and Perez PJ.Organometallics 2004; 23: 293–295.

22. Himeshima Y, Sonoda T and Kobayashi H. Chem.Lett. 1983; 1211–1214.

23. Volz WE and Paquette LA. J. Org. Chem. 1976; 41:57–62.

24. Crews P and Beard J. J. Org. Chem. 1973; 38:522–528.

25. Abbasoglu R, Menzek A and Balci M. Turkish J.Chem. 1998; 22: 201–207.

26. Padwa AT, Krumpe KE, Gareau YS andChiacchio U. J. Org. Chem. 1991; 56: 2523–2530.

27. Nowlan DT, Gregg TM, Davies HML andSingleton DA. J. Am. Chem. Soc. 2003; 125:15902–15911.

28. Rasmussen T, Jensen JF, Ostergaard N, Tanner D,Ziegler T and Norrby P-O. Chem. Eur. J. 2002; 8:177–184.



Synthesis and photophysical behavior of axiallysubstituted phthalocyanine, tetrabenzotriazaporphyrin,and triazatetrabenzcorrole phosphorous complexes

Edith M. Antunes and Tebello Nyokong*♦

Department of Chemistry, Rhodes University, Grahamstown 6140, South Africa


ABSTRACT: The synthesis of phosphorous phthalocyanines, triazatetrabenzcorroles, and tetraben-zotriazaporphyrins with a variety of axial ligands is reported. The new complexes are: phospho-rous dihydroxy tetrabenzotriazaporphyrin (5, PV(OH)2TBTAP), diphenyl phosphorous phthalocyanine(6, [PV(Ph)2Pc](OH)), diphenyl phosphorous triazatetrabenzcorrole (7, PV(Ph)2TBC), and dioctylphosphorous triazatetrabenzcorrole (8, PV(C8H17)2TBC). The complexes are not aggregated indimethylsulfoxide (DMSO) and pyridine. Upon axial coordination of a phenyl or octyl group,the complexes are soluble (and not aggregated) in dichloromethane (DCM) and tetrahydrofu-ran (THF). The triplet lifetimes range from 395 to 546 μs (for complexes 5 to 8), withthe P(Ph)2TBC (7) complex showing the longest triplet lifetime (546 μs), while the small-est triplet quantum yield (�T = 0.27) was obtained for the [P(Ph)2Pc](OH) (6) complex.[P(OH)2Pc](OH) (3) showed the shortest triplet lifetime (113 μs) and the largest triplet quantum yield(�T = 0.52).

KEYWORDS: phosphorous phthalocyanines, phosphorous triazatetrabenzcorroles, phosphorous tetra-benzotriazaporphyrins, triplet yields, triplet lifetime, fluorescence quantum yield.

INTRODUCTION

There has been enormous interest in phthalocyanines(Pcs) due to their promising potential in diverse applica-tions including: chemical sensing [1], liquid crystals [2],electrocatalysis [3], nonlinear optics [4, 5], and medicine(primarily photodynamic therapy [6–9]). Considerableresearch has thus been devoted to these complexes [5, 10].Nearly all metals or metalloids in the periodic tablehave been incorporated into the Pc core. Of particularinterest are the metallophthalocyanine (MPc) complexes

♦SPP full member in good standing∗Correspondence to: Tebello Nyokong, email: [email protected], tel: +27 46 6038260; fax: +27 46 6225109

containing Groups IVA and VA elements due to theirability to retain two different valence/oxidation states onthe central atom.

During the development of phthalocyanine chem-istry, the unsymmetrical tetrabenzotriazaporphyrins(TBTAPs) were formed by the reaction of a phthalonitrilewith a Grignard reagent (viz. methylmagnesium iodide)[11]. TBTAPs differ from phthalocyanines by the incor-poration of a methine group, instead of nitrogen, at ameso position. Using a Grignard reaction, a series TBTAPderivatives substituted at the benzo and/or meso posi-tions have been prepared [12]. Recently, we reportedon the synthesis of non-peripherally-octaalkyl sub-stituted dichlorotin(IV) tetrabenzotriazaporphyrin com-pounds [13].

Copyright © 2009 World Scientific Publishing Company

154 E.M. ANTUNES AND T. NYOKONG

A third analogue based on the phthalocyanine core,the triazatetrabenzcorroles (TBCs), was synthesized in theearly 1980s. The TBCs are characterized by the loss of oneof the bridging nitrogen atoms [14, 15] and a sharp Soretband (at approximately 440 nm). The syntheses of Ge, Si,Ga, and Al TBC complexes have been reported [16, 17],with our group recently reporting on a microwave synthe-sis of a sulfonated SnTBC [18]. Phosphorous TBCs havebeen reported [15]. In addition, a number of octa- andtetra-substituted P(O)TBC derivatives have also been syn-thesized [19–25] with most complexes containing alkylchain ring substituents. These complexes have displayedsignificant fluorescence [23] and singlet oxygen quan-tum yields [20], while the water-soluble tetrasulfonatedP(O)TBCs have been shown to possess good photody-namic therapy toward HeLa cells [24] and significantphotocleavage of supercoiled pUC19 DNA [26]. In addi-tion, TBCs have found potential application in nonlinearoptics [27].

Apart from the studies described above, the photophys-ical behavior of phosphorous macrocycles has receivedless attention. The effect of axial ligand exchange on thephotophysical behavior of phosphorous complexes has notyet been explored, and hence the photophysical behavior

N

N

N N

N

NNN

N P

O

DBU / Li metal, 1-Pentanol,

N2, 130−140 °C,12 h

POBr3, Pyridine,

N2, 90−110 °C, 2 h

HNN N

N

NNH

N

N

PBr3, Pyridine,

N2, 90−110 °C, 2 h

NN N

N

NNN

N P

OH

C6H5MgBr, THF,N2, rt, 30 min C6H5MgBr, THF,

N2, rt, 30 min

NN N

N

NNN

N P

R1

6R1 = 7

R1 =

NN

N

NNN

N P

R1

C8H17MgI, Et2O,N2, rt, 2 h

8R1 =

R1

1

34

CH3MgI, Et2O,N2, reflux, 4 h

HNN CH

N

NNH

N

N

NN CH

N

NNN

N P

2

5

POBr3 / PBr3, Pyridine,

N2, 90−110 °C, 2 h

OH

HO

OH

R1

Scheme 1. Synthetic pathway for the various phosphorous derivatives

of PPc, PTBC, and PTBTAP derivatives containing dif-ferent axial ligands is herein reported.

The new complexes synthesized include phospho-rous oxide tetrabenzotriazaporphyrin (5, P(OH)2TBTAP),diphenyl phosphorous phthalocyanine (6, [P(Ph)2Pc](OH)), diphenyl phosphorous triazatetrabenzcorrole (7,P(Ph)2TBC), and dioctyl phosphorous triazatetrabenzcor-role (8, P(C8H17)2TBC) (Scheme 1).

EXPERIMENTAL

Materials

Methanol (MeOH), phosphorous tribromide (PBr3),phosphorous oxybromide (POBr3), deuterated dimethyl-sulfoxide (DMSO-d6), deuterated pyridine (Pyr-d5),dicyanobenzene, 1,8-diazabicyclo[5.4.0]undec-7-ene(DBU), lithium metal (Li), magnesium turnings, andiodomethane (CH3I) were purchased from Aldrichand used as received. Pyridine, dimethylsulfoxide(DMSO), diethyl ether (Et2O), tetrahydrofuran (THF),dichloromethane (DCM), and 1-pentanol were alsoobtained from Aldrich and dried prior to use. Column


BEHAVIOR OF AXIALLY SUBSTITUTED PHTHALOCYANINE, TBTAP AND TRIAZATETRABENZCORROLES 155

chromatography was performed on silica gel 60 (0.04–0.063 mm) and neutral alumina.

Equipment

Ground-state electronic absorption spectra wererecorded on a Varian Cary 500 UV-vis-NIR spectropho-tometer. 1H NMR spectra were obtained using a BrukerAvance II+ 600 MHz and a Bruker AMX 400 MHz NMRspectrometer, with 31P NMR spectra recorded on the latterinstrument. FT-IR spectra (KBr pellets) were recorded ona Perkin-Elmer spectrum 2000 FT-IR spectrometer. Flu-orescence excitation and emission spectra (corrected forinstrument response) were recorded on a Varian Eclipsespectrofluorimeter using a 1 cm pathlength cuvette at roomtemperature. Laser flash photolysis experiments were per-formed with light pulses produced by a Quanta-Ray Nd:YAG laser providing 400 mJ, 90 ns pulses of laser light at10 Hz, pumping a Lambda-Physik FL3002 dye (Pyridin 1dye in methanol). Single pulse energy was 7 mJ. The ana-lyzing beam source was from a Thermo Oriel xenon arclamp, and a photomultiplier tube was used as a detector.Signals were recorded with a digital real-time oscilloscope(Tektronix TDS 360), with the kinetic curves averagedover 256 laser pulses. The triplet lifetimes were deter-mined by exponential fitting of the kinetic curves usingthe program OriginPro 7.5. Solutions with an absorp-tion of 1.5 at the excitation wavelength were prepared forthe triplet yield and lifetime determinations and purgedof oxygen by bubbling with argon for 15 min beforebeing irradiated with laser. Mass spectral information wasobtained on MALDI-TOF or ESI-mass spectrometers byDr M. Stander at Stellenbosch University, South Africa.Elemental analyses were carried out on a Vario EL IIIMicroCube CHNS Analyzer.

Photophysical methods

Fluorescence quantum yields (�F) were determined bycomparative method [28] (Equation 1):

�F = �StdF × FSample × AStd × n2

FStd × ASample × (nStd)2, (1)

where FSample and FStd are the areas under the fluorescencecurves of the MPc derivatives and the reference, respec-tively. ASample and AStd are the absorbances of the sampleand reference at the excitation wavelength, and n and nStd

are the refractive indices of solvents used for the sampleand standard, respectively. ZnPc in DMSO was used asa standard, �F = 0.20 [29]. At least three independentexperiments were performed for the quantum yield deter-minations. Both the sample and the standard were excitedat the same relevant wavelength.

Triplet quantum yields were determined using a com-parative method based on triplet decay, using Equation 2:

�SampleT = �Std

T × �ASampleT × εStd

T

�AStdT × ε

SampleT

, (2)

where �ASampleT and �AStd

T are the changes in the tripletstate absorbance of the sample derivative and the stan-dard, respectively. ε

SampleT and εStd

T are the triplet statemolar absorption coefficients for the sample and standard,respectively. �Std

T is the triplet state quantum yield for thestandard, ZnPc (�Std

T = 0.65 in DMSO [30]). εSampleT and

εStdT were determined from the molar absorption coeffi-

cients of their respective ground singlet state (εSampleS and

εStdS ) and the changes in absorbances of the ground sin-

glet states (�ASampleS and �AStd

S ) according to Equations 3aand 3b:

εSampleT = ε

SampleS

�ASampleT

�ASampleS

, (3a)

εStdT = εStd

S

�AStdT

�AStdS

. (3b)

Natural or radiative lifetimes (τN ) were estimatedusing PhotochemCAD program which uses the Strickler–Berg equation [31]. The fluorescence lifetimes (τF ) wereevaluated using Equation 4:

�F = τF

τN

. (4)

Syntheses

Unmetallated phthalocyanine (1), unmetallated tetra-benzotriazaporphyrin (2),dihydroxy phosphorous phthalo-cyanine (3, [P(OH)2Pc](OH)), and phosphorous oxotriazatetrabenzcorrole (4, P(O)TBC) were synthesizedand characterized according to the reported procedures[11, 14, 15, 17, 32, 33].

Phosphorous oxo tetrabenzotriazaporphyrin (5,[PV (OH)2 TBTAP]). Firstly, complex 2 was synthe-sized according to the method reported in the literature[11] with slight modification as follows: A solution ofmethylmagnesium iodide was prepared from Mg (2.4 g,0.099 mol) and iodomethane (6.5 mL, 0.104 mol) in100 mL of dry ether (Et2O). This Grignard reagent wassubsequently added to a mixture of 12.8 g of dicyanoben-zene (0.100 mol) and 50 mL of anhydrous Et2O, which,after 2 min, began to boil and this gentle reflux was con-tinued for 4 h. The remaining Et2O was then removedin vacuo and H2O (5 mL) carefully added to the reactionmixture until iodine vapor was produced. The residue wasthen heated at 200 ◦C, after which the dry residue wascooled, crushed, and repeatedly extracted (using a Soxh-let apparatus) with an ethanol (EtOH):acetone mixture.The extract was then dried in vacuo. This H2TBTAP com-plex 2 was then used to synthesize complex 5 by heating(under reflux) a mixture of 2 (100 mg, 0.20 mmol) andPBr3 (0.68 mL, 7.20 mmol) or POBr3 (1.5 g, 6.97 mmol)in pyridine at 90 to 110 ◦C for 2 h. The reaction was thenallowed to cool to room temperature and poured care-fully into water and allowed to stand overnight. The greenproduct obtained was then centrifuged and washed with



copious amounts of water. Upon drying, the precipitatewas chromatographed on a silica gel column using MeOHand pyridine. A dark green band containing 5 was elutedwith pyridine. Yield: 23%. UV-vis (DMSO): λmax, nm(log ε) 673 (4.61), 650 (4.42), 612 (3.87), 603 (3.83),395 (4.23). Calcd. for C33H19N7O2P: C 68.73, H 3.32,N 17.01; found C 67.32, H 3.70, N 15.30. ESI-MS: m/zcalcd. 576.2; found [M-OH]+ 558. 1H NMR (400 MHz,DMSO-d6): δ, ppm 11.02 (m, 1H), 9.65 (m, 3H), 9.46 (m,5H), 8.26–8.17 (m, 8H). 31P NMR (162 MHz, DMSO-d6):δ, ppm -195.93.

A modified procedure [34] was used to prepare thevarious phenyl and octyl axially ligated complexes asdescribed in the following paragraphs.

Preparation of diphenyl phosphorous phthalocya-nine (6, [PV(Ph)2Pc]+OH-). A solution of 50 mg(0.089 mmol) of [P(OH)2Pc] (OH) (3) in 10 mL DCMand 2 mL MeOH was stirred for 30 min to generate areactive dimethoxy derivative. The solvent was removedin vacuo, after which 10 mL dry THF and phenylmag-nesium bromide (4 equivalents, 1 M) in anhydrous Et2Owere added. After 30 min of stirring at room tempera-ture, the mixture was hydrolyzed and the organic layerwashed with water and evaporated to dryness. Purifica-tion was achieved on a short silica gel column with DCMfollowed by elution with methanol to give a light greencomplex, after purification with a BioBead column (usingDCM). Yield: 12%. UV-vis (DMSO): λmax, nm (log ε) 671(4.78), 604 (3.91), 352 (4.42). Calcd. for C38H21N8P (bro-mobenzene): C 68.87, H 3.70, N 12.86; found C 69.85, H5.45, N 9.53%. 1H NMR (400 MHz, DMSO-d6): δ, ppm9.69–9.65 (m, 4H), 8.54–8.50 (m, 4H), 7.72–7.66 (m,8H), 4.15–4.10 (m, 2H), 3.47–3.40 (m, 2H), 1.65–1.59(m, 1H).

Preparation of diphenyl phosphorous triazatetra-benzocorrole (7, PV(Ph)2TBC). A similar procedureto the one used for 6 was used to prepare complex 7,using 4 instead of 3, where 150 mg (0.28 mmol) ofP(O)TBC (4) and phenylmagnesium bromide (4 equiv-alents, 0.73 mmol, 1 M). The dry residue, obtainedafter hydrolysis, was passed through a BioBead (BIO-RAD) column using DCM as an eluent, and this fractionwas further purified using a short silica gel columnusing DCM and then methanol. A light green complexwas obtained upon elution with 100% MeOH. Yield:10%. UV-vis (DMSO): λmax, nm (log ε) 655 (4.70), 628(4.44), 442 (5.00). Calcd. for C44H26N7P (bromoben-zene): C 71.41, H 3.72, N 11.66; found C 71.19, H 5.48,N 7.50% MS (MALDI-TOF): m/z Calcd. 683.2; found[M-C6H5]+ 606.2. 1H NMR (400 MHz, DMSO-d6): δ,ppm 10.07–9.61 (m, 4H), 9.67–9.37 (m, 2H), 8.42–8.21(m, 10H), -1.89 (d, 2H), -2.05 (d, 2H), -2.56 (d, 2H), -3.70(t, 2H), -3.79 (t, 2H). 31P NMR (162 MHz, Pyr-d5): δ, ppm-212.02.

Preparation of dioctyl phosphorous triazatetraben-zocorrole (8, PV(C8H17)2TBC). A solution of 150 mg(0.275 mmol) of P(O)TBC (4) in 10 mL DCM and 2 mL

methanol was stirred for 30 min to generate a reactivedimethoxy derivative. The solvent was removed in vacuo,after which 10 mL dry Et2O and 4 equivalents of octyl-magnesium iodide (1.1 mmol, 1.48 M) in Et2O wereadded. After 2 h of stirring at room temperature, themixture was hydrolyzed and the organic layer washedwith water and evaporated to dryness. The residue waspassed through BioBeads (BIO-RAD) using DCM asan eluent and then further purified using a short silicagel column. The light green complex was eluted withDCM. Yield: 14.4%. UV-vis (DMSO), λmax, nm (log ε):661 (4.33), 638 (4.06), 605 (3.65), 454 (4.72). Calcd.for C48H50N7P (iodooctane): C 60.97, H 6.72, N 7.78;found C 59.09, H 6.07, N 7.91. MS (MALDI-TOF):m/z calcd. 755.4; found [M-C8H17] 642.4. 1H NMR(600 MHz, DMSO): δ, ppm 9.93–9.98 (d, 2H), 9.76–9.68(m, 6H), 8.52 (t, 2H), 8.46 (t, 2h), 8.44–8.37 (m, 4H), 0.49(m, 4H), 0.36 (t, 6H), 0.19-0.11 (m, 4H), -0.29to -0.36 (m, 4H), -0.83 to -0.90 (m, 4H), -1.54to -1.58 (m, 4H), -4.15 to -4.23 (m, 4H), -5.55to -5.60 (m, 4H). 31P NMR (162 MHz, Pyr-d5):δ, ppm -202.48; (DMSO-d6): δ, ppm -208.21.


Synthesis and characterization

Complexes 1 to 4 are known [11, 14, 15, 24, 33]. Eventhough the synthesis of H2TBTAP (2) is known [11], thephosphorous derivative has however not been reported.The reaction of 2 with either PBr3 or POBr3 results in theformation of the new P(OH)2TBTAP (5). Complexes 6 to8 are also new. The axial ligand exchange of OH in 3 or theoxo group in 4 for a phenyl or octyl group was obtainedusing a method similar to that reported by Kadish et al.,employing the Grignard reaction with the appropriate Mgreagent to form 6–8 [34]. The synthetic pathway of thesephosphorous derivatives is given in Scheme 1.

The phosphorous macrocycles were characterized byUV-vis, MS, and NMR spectroscopies (including 31PNMR) and elemental analyses. The analyses were con-sistent with the predicted structures as shown in theexperimental section. Examination of 1H NMR spectrarevealed the typical signals expected for the unsubstitutedring core of the complexes, occurring in the region of∼ δ = 8–11 ppm, confirming a strong diamagnetic currenteffect. The TBTAP compound 5 displayed the characteris-tic methine proton at approximately 11 ppm in the protonspectrum. As expected, the 1H NMR spectra of the cor-roles (viz. 7 and 8) were more complex than that of itsphthalocyanine derivative (6), due to the loss of symmetry.The protons of the axially bound ligands gave rise to sig-nals characteristic of strong shielding due to the magneticanisotropy of the aromatic macrocycle and coupling to thecentral phosphorous atom. The proton signals of the TBC(7, 8) complexes were found to be shielded to a far greaterextent (Fig. 1) than that of the Pc complex (6).For complex



ppm (t1)-5.0-4.0-3.0-2.0-1.00.0

Fig. 1. 1H NMR (600 MHz, DMSO-d6) spectrum illustrating the strongly shielded protons of the axial ligand protons of complex 8

7, for example, the proton at the ortho position appears asa doublet at δ = -1.89 with a large 1H–31P coupling con-stant of ∼ 24 Hz. In addition, it would appear that the twophenyl axial ligands centered on the phosphorous atomare in slightly different chemical environments becauseof the small difference in chemical shifts observed forthese axial protons. Phosphorous (31P) NMR proved to bea useful technique in determining the coordination numberfor all complexes synthesized. The 31P resonances for allcomplexes were located in the high-field region viz. -190to -240 ppm with shifts at ∼ δ = −200 ppm being dis-tinctive of six coordinate complexes [19, 22]. Elementalanalyses and mass spectra confirmed the presence of twoaxial ligands.

Difficulties in interpreting mass spectral data wereencountered because of the number of signals observedfor a single compound. Breusova et al. [25] found that,in addition to the peaks of the singly charged ions cor-responding to the expected molecular masses, peaksconsistent with the successive elimination of the axial lig-ands induced by laser ionization were also found. Theauthors observed the most intense signals to correspondto the unsubstituted phosphorous complex, while peaksattributed to the free ligand were also observed [25]. Massspectral analyses gave expected masses as shown in theexperimental section. The elemental analyses (%C) wereconsistent with the expected values. It has been previ-ously observed [19, 35] that MPc molecules, due to thelarge size, do not give satisfactory elemental analyses asobserved for %N for the complexes reported in this work.These discrepancies might possibly be due to the pres-ence of solvent molecules. The %C results obtained inthis study, however, were found to be suitable.

Generally, phthalocyanine complexes are insoluble inmost organic solvents; however, introduction of sub-stituents to the ring or axial ligands increases solubility

substantially. The alkyl/phenyl axial substituted com-plexes (6–8) exhibited excellent solubility in organicsolvents such as pyridine, DMSO, DCM, and THF.

The Pc complexes (3 and 6) show typical Pc UV-visspectra with a sharp Q-band at ∼ 670 nm and a broadB-band at ∼ 350 nm (Fig. 2 and Table 1). For these com-plexes, the metal-free Pc derivative (1) was formed first,followed by the reaction with an excess of POBr3 to givethe desired [P(OH)2Pc](OH) derivative (Scheme 1). Uponcoordination of the axial phenyl groups to 3 to form 6, lit-tle or no shift in Q-band was observed (Table 1); both areat 671 nm. However, the solubility of complex 6 improvedsignificantly in a variety of solvents.

TBC complexes display characteristic UV-vis spectrawith a sharp peak at 440 nm [21–24] (Fig. 2), which canbe used to confirm the formation of the compound. TBCcomplexes lose the fourth azomethine nitrogen of the MPcmoiety, with the molecules no longer retaining the sym-metrical Pc structure (Scheme 1). The presence of excessmetal halide is thought to cause the bridging nitrogen lossfrom the Pc core to form the corrole [17]. In this study, themetal-free Pc derivative (1) was again formed first. Upon

Fig. 2. Comparison of the UV spectra of complexes 3, 4, and 5



Table 1. Photophysical parameters of the phosphorous macrocycles in DMSO

Compound λQ-band, nm (log ε) �F τF , ns �T τT , μs �I C

[PV(OH)2Pc](OH) (3) 671 (4.89) 0.13 2.02 0.52 113 ± 1 0.35PV(OH)2TBTAP (5) 672 (4.61) 0.17 2.53 0.30 449 ± 9 0.53[PV(Ph)2Pc](OH) (6) 671 (4.78) 0.08 0.96 0.27 456 ± 5 0.65PV(Ph)2TBC (7) 655 (4.70) 0.18 2.90 0.45 546 ± 4 0.37(PV(C8H17)2TBC) (8) 661 (4.33) 0.15 4.40 0.30 395 ± 7 0.55

reaction with PBr3, the resultant PV(O)TBC derivative for-mation can be monitored spectroscopically by observingthe collapse of the sharp Q-bands in the visible regionof unmetallated Pc to form three distinct bands (in theQ-band region), together with the formation of the sharpSoret band at 440 nm (Fig. 2). These distinctive bandsare still observed upon axial ligand exchange of the oxogroup for either the phenyl or the octyl groups; this isaccompanied by a marked improvement in solubility in avariety of solvents. A prominent shift in the Q-band to theblue compared to the corresponding Pc complexes wasobserved (Fig. 2).

Aggregation in phthalocyanines and related complexesis usually depicted as a coplanar association of ringsprogressing from monomer to dimer and higher ordercomplexes. It is dependent upon the concentration ofthe complex, the nature of the solvent, the nature ofthe substituents, and the complexed metal ions. In thisstudy, the aggregation behavior of the PPc, PTBC, andPTBTAP derivatives was investigated in DMSO (Fig. 3).The complexes did not show aggregation at concentra-tions less than 3 × 10-5 M. The linear ranges were:P(OH)2TBTAP (5) 0.2–1.7×10-5 M, [P(Ph)2Pc](OH) (6)0.5–2.1 × 10-5 M, P(Ph)2TBC (7) 0.5–2.8 × 10-5 M, andP(C8H17)2TBC (8) 0.1–1.3 × 10-5 M.

Photophysical studies

Figure 4 shows the absorption, excitation, and emissionspectra of P(OH)2TBTAP (5). The excitation spectrum ofcomplex 5 is similar to the absorption spectrum in termsof overlap of the low energy component of the Q-band.The former is however broad and does not show clearsplitting, suggesting some changes in symmetry follow-ing excitation. Low-symmetry Pc complexes, such as thatof unmetallated Pcs [36], fluoresce with only one mainpeak, assigned as the 0–0 transition of the fluorescence[36], hence the observation of one main emission peak inFig. 4. The fluorescence quantum yields (�F ) obtainedfor the phosphorous derivatives are given in Table 1.These values are within the range reported for MPc com-plexes, except for PV(Ph)2Pc-2 (6) which has a much lower�F value. Axial ligand exchange of the oxo group of4 (�F = 0.12 [33] for a phenyl (7) or octyl (8) axialsubstituent increased the quantum yields marginally to�F = 0.18 (for 7) and �F = 0.15 (for 8) (Table 1).Comparison of identical axial ligands but different core

ring systems, reveals a much lower �F value (= 0.08)for the Pc complex 6 compared to the TBC complex 7(�F = 0.18) (Table 1).

Since good correlations using the Strickler-Berg equa-tion have been found between experimentally and theoreti-cally determined lifetimes for the unaggregated molecules[37], the fluorescence lifetimes (τF , Table 1) of the com-plexes synthesized were calculated. Complex 6 also showsa much lower fluorescence lifetime (Table 1). The τF and�F values of 3 showed a decrease upon the exchange ofthe axial OH ligands for phenyl axial ligands.

Studies of the triplet behavior of the complexes uponexcitation were also conducted. Transient absorptionspectra were recorded in argon-degassed solutions byexciting the photosensitizer (in DMSO) in the Q-bandregion and recording the transient absorption spectra pointby point from 400 to 780 nm (Fig. 5). The Q- and Soret

(a)

(b)

Fig. 3. (a) UV spectra of PV(Ph)2TBC (7) over the concentrationrange studied: 2.8 × 10-5–5.2 × 10-6 M; (b) Beer-Lambert lawplot for spectral peaks in (a)



Fig. 4. (a) Comparison of the absorbance, (b) excitation, and(c) emission spectra of complex 5 (Excitation wavelength=599 nm)

Fig. 5. Transient differential spectrum of complex P(OH)2TBTAP (5) in DMSO; Excitation wavelength = 672 nm; theinset shows the weak transient absorption of the triplet state

0.000 0.001 0.002 0.003 0.004 0.005-0.01

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

ΔA

Time, s

Fig. 6. Triplet decay curve of P(Ph)2TBC (6) in DMSO at 490 nm(Excitation wavelength = 660 nm)

bands showed negative absorption (bleaching) bands,withthe transient spectra showing a broad positive absorp-tion around 500 nm (Fig. 5). Figure 6 shows the tripletdecay curve for the complexes that obeyed second-orderkinetics. This is typical of MPc complexes at high con-centrations (> 1 × 10-5 M) [38] due to the triplet–tripletrecombination. The concentrations employed in this workwere in this range; hence, triplet–triplet recombination isexpected.

In terms of triplet quantum yields (�T ) and lifetimes,the Pc complex (3) gave the largest value of the tripletquantum yield when compared to the rest of the complexesin Table 1. It has been reported before [13] that the pres-ence of the a-CH group instead of one of the aza-nitrogenatoms of the phthalocyanine core to form a TBTAP inthiol-substituted phthalocyanines, results in an increasein �T values. However, in this work, the Pc complex (3)gave the largest �T value, suggesting that the axial andperipheral ligands, as well as the central metal have a sig-nificant effect on �T values of these macrocycles. TheTBC derivative (7) gave a larger �T value than that forTBTAP (Table 1). TBC derivatives have been reported[18] to give lower �T value compared to the correspond-ing Pc complexes. Table 1 shows that the TBC complexes(7 and 8) have lower �T values compared to complex 3;however, the value for 8 is almost similar to those of thephenyl-substituted complex 6 and the reported complex 4(�T = 0.27) [33].

Overall, in terms of triplet lifetimes, the valuesreported here for the phosphorous derivatives, except for[P(OH)2Pc](OH) (3), are high when compared to MPccomplexes in general [36]. This could be due to the smallsize of the P central element which would result in low�T values, but long triplet lifetimes. The TBC derivative(7) gave the longest lifetime when compared to the corre-sponding Pcs containing similar axial ligands Pc (6) andcompared to 8 or TBTAP derivative (5) (Table 1). A longlifetime has also been previously reported for complex 4[33]. The lowest triplet lifetime was observed for 3, whichcorresponded to the highest triplet quantum yield. Thequantum yields of internal conversion (�I C) were foundto be high (Table 1). This can be attributed to the pres-ence of OH groups, especially for complexes 3, 5, and 6,which are known to play a significant role in the internalconversion from S1 to S0 [39].

CONCLUSION

In conclusion, a series of new phosphorous macrocy-cles containing phenyl and octyl ligands were synthesized(complexes 6–8), and their photophysical data were com-pared with those of their unsubstituted complexes.We alsoreport for the first time on the synthesis of P(OH)2TBTAP(5). The derivatives all showed long triplet lifetimes rang-ing from 395 to 546 μs. Correspondingly low �T values(again except for [P(OH)2Pc](OH)) ranging from �T =0.27 to 0.45 were therefore observed for these complexes.The triplet lifetime and �T values obtained for complexesdiscussed in this work make them plausible candidates forPDT.

Acknowledgments

This work was supported by the Department of Scienceand Technology (DST) and National Research Foundation(NRF), South Africa through DST/NRF South African



Research Chairs Initiative for Professor of MedicinalChemistry and Nanotechnology as well as Rhodes Uni-versity and Medical Research Council of South Africa.EMA wishes to acknowledge MRC for a post-doctoralscholarship.

REFERENCES

1. McKeown NB. Phthalocyanine Materials: Structure,Synthesis and Function, Cambridge University Press:Cambridge, UK, 1998; p. 200.

2. Guillaud G, Simon J and Germain JP. Coord. Chem.Rev. 1998; 180: 1433–1484.

3. Nyokong T. In N4-Macrocyclic Metal Com-plexes: Electrocatalysis, Electrophotochemistry, andBiomimetic Electrocatalysis, Zagal JH, Bedioui F,Dodelet J-P. (Eds.). Springer: New York, 2006;Ch. 7.

4. de la Torre G, Torres T and Agulló-López F. Adv.Mater. 1997; 9: 265–269.

5. de la Torre G, Claessens CG and Torres T. Chem.Commun. 2007; 2000–2015.

6. Bown SG, Tralau CJ, Coleridge-Smith PD, AkdemirD and Wieman TJ. Br. J. Cancer 1986; 54: 43–52.

7. Bonnett R. Chemical Aspects of Photodynamic Ther-apy, Gordon and Breach Science Publishers: Ams-terdam, 2000.

8. Macdonald IJ and Dougherty TJ. J. PorphyrinsPhthalocyanines 2001; 5: 105–129.

9. Spikes JD. J. Photochem. Photobiol. B: Biol. 1990;6: 259–274.

10. Kadish KM, Smith KM and Guilard R (Eds.). Por-phyrin Handbook, Academic Press: New York, 2003;pp. 15–20.

11. Barret PA, Linstead RP, Tuey GAP and RobertsonJM. J. Chem. Soc. 1939; 106: 1809–1820.

12. Tse Y-H, Goel A, Hu M, Lever ABP and Leznoff CC.Can. J. Chem. 1993; 71: 742–753.

13. Khene S, Cammidge AN, Cook MJ andNyokong T. J. Porphyrins Phthalocyanines 2007;11: 761–770.

14. Gouterman M, Sayer P, Shankland E and Smith JP.Inorg. Chem. 1981; 20: 87–92.

15. Sayer P, Gouterman M and Connel CR. Acc. Chem.Res. 1982; 15: 73–79.

16. Fujiki M, Tabei H and Isa K. J. Am. Chem. Soc. 1986;108: 1532–1536.

17. Li J, Subramaniam LR and Hanack M. J. Chem. Soc.Chem. Commun. 1997; 679–680.

18. Khene S, Ogunsipe A, Antunes E and Nyokong T. J.Porphyrins Phthalocyanines 2007; 11: 109–117.

19. Li J, Subramanian LR and Hanack M. Eur. J. Org.Chem. 1998; 2759–2767.

20. Liu J, Zhao Y, Zhao F, Zhang F, Song X and ChauF-T. J. Photochem. Photobiol. A: Chem. 1996; 99:115–119.

21. Kasuga K, Liu L, Handa M, Sugimore T, Isa K, Mat-suura K and Takanami Y. Inorg. Chem. 1999; 38:4174–4176.

22. Fox JP and Goldberg DP. Inorg. Chem. 2003; 42:8181–8191.

23. Liu J, Zhang F, Zhao F, Tang Y, Song X and Yao G. J.Photochem. Photobiol. A: Chem. 1995; 91: 99–104.

24. Song Z, Zhang F, Li X, Zhek-Kiu C, Zhao F and TangY. J. Porphyrins Phthalocyanines 2002; 6: 484–488.

25. Breusova MO, Pushkarev VE and Tomilova LG.Russ. Chem. Bull. Int. Ed. 2007; 56: 1456–1460.

26. Huang L, Zhong C, Zhang F and Tung C-H. Bioorg.Med. Chem. Lett. 2008; 18: 2152–2155.

27. Huang L, Li Z, Zhang F, Tung C-H and Kasatani K.Optics Commun. 2008; 281: 1275–1279.

28. Fery-Forgues S and Lavabre DJ. Chem. Ed. 1999; 76:1260–1264.

29. Ogunsipe A, Chen J-Y and Nyokong T. New J. Chem.2004; 28: 822–827.

30. Tran-Thi TH, Desforge C and Thies C. J. Phys.Chem.1989; 93: 1226–1233.

31. Strickler SJ and Berg RA. J. Chem. Phys. 1962; 37:814–822.

32. Leznoff CC. In Phthalocyanines: Properties andApplications, Leznoff CC and Lever ABP (Eds.),Vol. 1, VCH Publishers: New York, 1989; Ch. 1.

33. Antunes E and Nyokong T. Metal-based Drugs 2008;2008: 498916.

34. Kadish KM, Ou Z, Adamian VA, Guilard R, Gros CP,Erben C, Will S and Vogel E. Inorg. Chem. 2000; 39:5675–5682.

35. Marcuccio S, Svirskaya PI, Greenberg S and LeverABP. Can. J. Chem. 1985; 63: 3057–3069.

36. Nyokong T. Coord. Chem. Rev. 2007; 251: 1707–1722.

37. Maree D, Nyokong T, Suhling K and Phillips D.J. Porphyrins Phthalocyanines 2002; 6: 373–376.

38. Debacker MG, Deleplanque O, Van Vlierberge B andSauvage FX. Laser Chem. 1988; 8: 1–11.

39. Beeby A, Parker AW, Simpson MSC and Phillips, D.J. Photochem. Photobiol. B: Biol. 1998; 16: 73–81.





INTRODUCTION

The chemistry of phthalocyanines (Pcs) was fi rst directed toward the synthesis of asymmetrically substi-tuted Pcs, with a recent and dramatic increase as the quest for new materials is constantly enhanced. Three types of asymmetric patterns have been reported so far (Fig. 1). Obtaining AB3-type and A2B2-type Pcs is ensured by well-known selective syntheses [1]. These asymmetric substitution patterns give rise to many properties, with interesting effects for self-assembly [2, 3], electronic push-pull effects [4, 5] or labelling of potential photosen-sitizers [6], among others. Syntheses are either selective, based on mixtures of chemically different precursors reacting selectively with one another [7, 8], or statisti-cal, by mixing two precursors of the same type [9, 10].In the latter case, it is possible to favor the formation of the desired product by playing on the relative ratio of the two precursors. The method for the synthesis of cross-wise Pcs (commonly named ABAB) has been known since 1990 (Scheme 1). This method was described by Young et al. [11] following a patent methodology [12].It relies on the specifi c reactivity of a trichloroisoindo-lenine derivative with compounds of general formula R-NH-R [13]. Trichloroisoindolenine derivatives undergo

selective cross-condensation with diiminoisoindolines, leading to the desired crosswise Pc, in the presence of a base and a reductive agent. Despite a yield said to be quite high in this fi rst paper [11], later descriptions of ABAB Pcs never yielded more than 35% [14–18].

As far as we know, and despite the obvious interest they could present, no Pc having three different substitu-ents has ever been selectively synthesized. Only closely related structures, having a triazole moiety have been reported [19]. Expecting very interesting properties of these new materials regarding the common Pcs appli-cations [20], we challenged the synthesis of the highly asymmetric Pc 1, bearing three different substituents. This Pc may be named ABAC in a quite descriptive way (Fig. 1). Obtaining such a Pc is, of course, not conceiv-able by mixing three Pc precursors of the same type, as the number of potential products cannot lead to an acceptable yield, and would in any case present outstand-ing diffi culties regarding the separation process. We thus developed a novel strategy based on the use of a directed selective synthesis: the crosswise Pc selective synthesis, applied to a statistical mixture of precursors, i.e. two different diiminoisoindolines. For the fi rst ABAC (1),substituents were chosen to allow easy separation of the resulting compounds. B is a hydrophobic didodecyloxy unit facing C, which is a hydrophilic moiety consisting of a solketal unit (Chart 1). This solketal unit also offers the advantage of allowing further transformations after deprotection of the acetal group, liberating two hydroxyl functions with versatile substitution potential.

A fi rst ABAC phthalocyanine

Fabienne Dumoulin* , Yunus Zorlu, M. Menaf Ayhan, Catherine Hirel, Ümit Isci and Vefa Ahsen*

Gebze Institute of Technology, Department of Chemistry, P. O. Box 141, Gebze, 41400 Kocaeli, Turkey

Received 7 August 2008Accepted 6 October 2008

ABSTRACT: The derouting of the selective synthesis of crosswise phthalocyanines (Pcs) applied to a statistical precursors mixture leads to the fi rst member of a new family of Pcs bearing three different substituents: the ABAC phthalocyanines. By playing on the precursors’ relative ratio, the yields and selectivity of the method have been optimized.

KEYWORDS: selectivity, asymmetric phthalocyanine, trichloroisoindolenine, diiminoisoindoline.


*Correspondence to: Fabienne Dumoulin and Vefa Ahsen, email: [email protected], [email protected], fax: +90 262-605-31-01, tel: +90 262-605-31-23


162 F. DUMOULIN ET AL.

EXPERIMENTAL

General

Solvents were dried as described in Perrin and Armar-ego [21]. 3 was prepared following the described method [22]. Chromatographic purifi cation was performed on silica gel (Merck, 0.04–0.063 mm) with the indicated eluting systems. Absorption spectra in the UV-vis region were recorded with a Shimadzu 2001 UV Pc spectropho-tometer using a 1 cm path-length cuvette at room temper-ature. LC-ESI mass spectra were recorded with a Bruker microTOF spectrometer. Elemental analyses were per-formed on a ThermoFinnigan Flash 1112 instrument. 1Hand 13C NMR spectra were recorded in CDCl3 solutions with a Varian 500 MHz spectrometer. The HPLC system is an Agilent 1100 series HPLC system (ChemStation software) equipped with a G1311A pump and G1315B diode array detector monitoring the range 254–900 nm. A normal phase column Lichrosorb-SI-60 (250 × 4.6 mm) from Alltech. Associates, Inc. was also used.

Synthesis

Preparation of 5-(2,2-dimethyl-1,3-dioxolan-4-yl)-methoxy-1,3-dihydro-1,3-diiminoisoindole (4). A sus-pension of 4-solketalphthalonitrile [2] (5 g, 19 mmol) in dry methanol (250 mL) in the presence of MeONa (400 mg) was placed under argon with bubbling of NH3 at rt for 1 h, then refl uxed for 6 h under an ammonia fl ow. Upon heating, the 4-solketalphthalonitrile readily dissolved in methanol. After cooling, the solvent was evaporated and the solid washed with hexane and fi ltered by suction. The resulting powder was stirred in chloroform and fi ltered. Evaporation of the fi ltrate yielded 4 as a greenish pow-der, which was then used without further purifi cation. Yield 4.9 g (94%). IR (KBr): ν, cm-1 3241, 2987, 2937, 1620, 1536, 1485, 1447, 1373, 1312, 1227, 1138, 1061, 841. ESI-MS: m/z 276.1 [M + H]+. 1H NMR (CDCl3): δ,ppm 7.91 (bs, NH), 7.48, 6.85 (2d, 2H, H-6, H-7), 7.14 (H-4), 4.41–3.78 (m, 5H, CH2CHCH2), 1.34, 1.40 (2t, 6H, 2 CH3).

13C NMR (CDCl3): δ, ppm 167.35, 161.42, 137.13, 127.13, 122.33, 117.83, 109.82, 106.62, 73.69, 69.19, 66.41, 26.68, 25.26, mp 161–163 °C. Anal. calcd for C14H17N3O3: C, 61.08; H, 6.22; N, 15.26. Found: C, 61.24; H, 6.17; N, 15.55.

Preparation of 1. A mixture of diiminoisoindolines 3 (775 mg, 1.5 mmol) and 4 (830 mg, 3.0 mmol) was suspended in dry THF (150 mL) in the presence of trieth-ylamine (1.2 mL). This suspension was cooled in an ice bath and stirred at 0 °C for 1 h. 1,3,3-trichloroisoindole-nine 2 [12] (2 g, 9.0 mmol) dissolved in THF (100 mL) was added dropwise. The reaction mixture was stirred for 1 h at 0 ºC, then for 5 h at rt. After fi ltration under argon to remove the triethylammonium chloride, hyd-roquinone (500 mg) and MeONa (250 mg) were added to the fi ltrate, stirring of which continued under refl ux

ABAC

A3B ABAB

AABB

Fig. 1. Schematic representation of the different asymmetric Pcs types. is a schematic representation of a Pc core

N

Cl

Cl Cl

NH

NH

NH

N

Cl

Cl Cl

NH

NH

NH

+ + NH

NH

NH

+

Classical reaction

ABAC reaction

Scheme 1. Crosswise Pcs’ synthesis and ABAC reaction

N

HN N

N

N

NH

NN

OC12H25

OC12H25

O

O

O

1

Chart 1. ABAC phthalocyanine


A FIRST ABAC PHTHALOCYANINE 163

overnight. The solvent was then removed, the crude product thoroughly washed with ethanol and purifi ed on silica gel (dichloromethane:methanol, 20:1), yielding 137 mg (9%) of dark green powder. 1H NMR (CDCl3):δ, ppm 9.01–10.7 (2m, 13H, aromatics), 4.26–3.37 (3m, 9H, CH2CHCH2, OCH2), 1.58 (br t, 4H, 2 OCH2CH2),1.43 (s, 3H, CH3), 1.37 (s, 3H, CH3), 1.26 (s, 18H, 9 CH2), 0.88 (t, 6H, 2 CH2CH3).

13C NMR (CDCl3): δ, ppm 160.51, 151.79, 134.84, 128.77, 127.92, 124.12, 121.82, 121.39 (aromatics), 110.27 (C(Me)2), 74.60 (CH), 69.51, 69.45, 67.39 (CH2O), 25.85–25.63 (C(CH3)2), 32.28–23.00 (CH2 alkyl), 14.36 (CH3). The purity of 1 has been checked by HPLC. The mobile phase was a 98/2 (v/v) mixture of hexane–IPA (2-propanol). The fl ow-rate was set at 1 mL.min-1 and the sample was dissolved in dichloromethane at a concentration of 1 mg.mL-1 for the normal phase. Under these conditions, the retention time for ABAC phthalocyanine was found to be 43.08 min. UV-vis (CHCl3): max, nm (log ) 340 (3.94), 661 (4.15), 697 (4.27).


Trihaloisoindolenines and, more generally, trichlor-oisoindolenines [13] are known to react specifi cally with R-NH2 compounds in the presence of an acid acceptor and a reductive agent. The patented described method [13] was optimized after being split into two steps: fi rst, a mixture of trichloroisoindolenine and diiminoisoindo-line (1:1) were stirred in tetrahydrofuran in the presence of triethylamine, after which the triethylammonium salts were removed, the hydroquinone was added, and the solution was refl uxed while the characteristic green color of Pc appeared [11]. Derouting this crosswise ABAB Pcs synthetic method, the key point of our strategy is the use of a statistical mixture of two different diiminoisoindo-lines (3 and 4, bearing the B and C substituents, respec-tively) to be condensed with a trichloroisoindolenine (2)bearing the A substituent (Scheme 1 and 2). The pos-sible products are, in this case, quite numerous: the three possible crosswise Pcs (the desired ABAC and the two

“symmetric” ones: ABAB and ACAC), the Pcs resulting from the condensation of three diiminoisoindoline units on one trichloroisoindolenine (a rarely reported case [18]giving six potential combinations: ABBB, ACCC, ABCC, ABBC, ABCB and ACBC), and the six possible Pcs result-ing from the condensation of four diiminoisoindoline units (BBBB, BCCC, BBCC, BCBC, BBBC and CCCC).

To limit the number of these products and increase the yield of the ABAC, we experimented by varying the precursors’ ratio. We used two equivalents of trichlor-oisoindolenine 2, relative to the total molar amount of diiminoisoindolines (3 + 4). This aimed to enhance the formation of the crosswise Pcs, limiting the formation of the other adducts. We also experimented by varying the ratio of the two diiminoisoindolines, employing two equivalents of 4 (bearing the C moity), relative to 3 (bear-ing the B moiety) which aimed to avoid the formation of the crosswise ABAB.

After completion of the reaction and cooling of the reaction mixture, the tetrahydrofuran was evaporated and the Pcs were cleaned by washing several times in hot ethanol and ethylacetate. Analyses of the formed Pcs were carried out on this mixture, then 1 was easily puri-fi ed on a silica-gel column chromatography thanks to the different polarity of the ABAC (1) and ACAC. The fi rst ABAC 1 was obtained in 9% yield, which was an excel-lent one, given that it is the fi rst time that an attempt has been made to mix three different phthalocyanine precur-sors. Yields of all the possible products are summarized in Table 1. In addition to the high yield for ABAC, the formation of side-products has been successfully limited (Table 1). Only traces of four diiminoisoindolines adducts were observed, and two major Pcs (the desired ABAC (1)and the symmetric crosswise ACAC) were isolated, in the same molar amounts. The overall yield of crosswise Pcs fi t the best reported yields, with the expected selectivity, as almost only crosswise Pcs was obtained.

The ABAC 1 has been unambiguously character-ized. 1H and 13C NMR spectrum recorded in deuterated

N

Cl

Cl Cl

NH

NH

NH

+ + NH

NH

NH

C12H25O

C12H25O

O

O

O

2 3 4

1i, ii

Scheme 2. Formation of 1. Reagents and conditions: (i) trieth-ylamine, THF, ice bath, 1 h, then rt, 5 h, then after fi ltration (ii) hydroquinone, MeONa, THF, refl ux overnight

Table 1. Yields obtained for each of the obtainable products during the formation of 1.

Entry Possible Pcs Yields, %

1 ABAC 1 9% (based on 3)

2 ABAB not detected

3 ACACsame molar amount

as 1

4 one trichloroisoindolenine unit 2 – three diiminoisoin-doline units (3 + 4) adducts

not detected

5a 4 diiminoisoindoline units adducts

traces

adetection by ESI-MS


164 F. DUMOULIN ET AL.

chloroform show the expected peaks despite a slight aggregation broadening the signals. On the 1H NMR spectrum, the study of the integrals revealed that only one solketal moiety was present on the molecule bearing two dodecyl chains. The aromatic area showed several peaks corresponding to 13 aromatic protons. Due to the

aggregation, internal NH were not visible. The LC-ESI-MS spectrum fi ts the theoretical one perfectly (Fig. 2). UV-vis spectra of this fi rst ABAC Pc, recorded in various solvents, are presented in Fig. 3. It is characteristic of a free-based Pc. The purity of 1 was analyzed by HPLC experiments, using a normal phase column with a 98/2 (v/v) mixture of hexane–IPA (2-propanol) as mobile phase. This method allows proof of absence of the other

Fig. 2. (a) ESI-MS mass spectrum of 1 (b) Detail of isotopic pattern obtained by ESI-MS for molecular ion of 1 (c) Theoretical isotopic pattern: m/e: 1012.59 (100.0%), 1013.60 (70.3%), 1014.60 (25.4%), 1015.60 (6.7%), 1013.59 (3.0%), 1014.59 (2.1%)

300 400 500 600 700 800

0.00

0.05

0.10

0.15

0.20

Absorb

ance

Wavelenght, nm

Fig. 3. UV-vis spectra of 1 in chloroform (dashed black), tetrahydrofuran (dotted red) and toluene (straight blue). max,nm (log ), chloroform: 340 (3.94), 661 (4.15), 697 (4.27), tetrahydrofuran: 694 (4.20), 659 (4.08), 339 (3.83), toluene: 694 (4.14), 660 (4.00), 341 (3.61)

Fig. 4. HPLC profi le of 1


A FIRST ABAC PHTHALOCYANINE 165

possible Pcs, of which traces may not be detectable with elemental analyses, due to close elemental composition. The only peak on the spectrogram has a retention time of 43 min with no traces of other products (Fig. 4).

CONCLUSION

In summary, and to our knowledge, this is the fi rst description of selective synthesis leading to a Pc having three different substituents on expected and chosen positions: the ABAC phthalocyanine (1). Selective synthesis is based on the specifi c reactivity of two different types of precursor. The number of possible products was limited by varying the ratio of the precursors, the trichloroisoindolenine:diimininoisoindolines one on one hand, and the relative ratio of the two diiminoisoindolines bearing, respectively, the substit-uents B and C, on the other. The selectivity of the method is proved as only two major products are formed, among which is the expected ABAC Pc, 1, obtained in excellent yield. From now onwards, the next challenges reside in the production of many of these novel molecules to obtain new materials, expected to exhibit interesting properties, especially concerning electronic effects and from a self-assembly point of view. This should provide innovative materials for applications currently focusing on the inter-ests of the phthalocyanine community, i.e. self-assemblies and non-linear optics. Synthesis of other ABAC phthalo-cyanines is currently being achieved by our group.

Acknowledgements

The authors wish to thank Ilker Un and Bünyamin Cosut for NMR, MALDİ and LC-ESI-MS analyses. The fi nancial support of the Scientifi c and Techno-logical Research Council of Turkey TUBITAK (project 106T376) is gratefully acknowledged.

REFERENCES

de la Torre G, Claessens CG and Torres T. 1. Eur. J. Org. Chem. 2000; 2821–2830.Kimura M, Kuroda T, Ohta K, Hanabusa K, Shirai H 2.and Kobayashi N. Langmuir 2003; 19: 4825–4830.

Eichhorn H. 3. J. Porphyrins Phthalocyanines 2000; 4: 88–102.Garcia-Frutos EM, Diaz DD, Vazquez P and Tor-4.res T. Synletters 2006; 19: 3231–3236. Sastre A, del Rey B and Torres T. 5. J. Org. Chem.1996; 61: 8591–8597.Hudson R and Boyle RW. 6. J. Porphyrins Phthalo-cyanines 2004; 8: 954–975.de la Torre G and Torres T. 7. J. Porphyrins Phthalo-cyanines 2002; 6: 274–284. Nolan KJM, Hu M and Leznoff CC. 8. Synletters1997; 5: 593–594.Calvete M and Hanack M. 9. Eur. J. Org. Chem. 2003; 2080–2083.Jung R and Hanack M. 10. Synthesis 2001; 9: 1386–1394.Young JG and Onyebuagu W. 11. J. Org. Chem. 1990; 55: 2155–2159.Farbenfabriken Bayer, US Patent 2,701,255, 12. February 1, 1955.Idelson EM. Polaroid Corp., US Patent 4,061,654 13.December 6, 1977; Chemical Abstract., 1978; 88, P171797m.Dabak S and Bekaro14. ğlu Ö. New. J. Chem. 1997; 21:267–271.Stihler P, Hauschel B and Hanack M. 15. Chem. Ber. Recueil 1997; 130: 801–806. Lin MJ and Wang JD. 16. J. Mol. Struct. 2007; 837:284–289.Youngblood JW. 17. J. Org. Chem. 2006; 71: 3345–3356.Hanack M and Stihler P. 18. Eur. J. Org. Chem. 2000; 303–311.Cabezon B, Quesada E, Esperanza S and Torres T. 19.Eur. J. Org. Chem. 2000; 2767–2775.de la Torre G, Claessens CG and Torres T. 20. Chem.Commun. 2007; 2000–2015.Perri21. n DD and Armarego WLF. In Purifi cation of Laboratory Chemicals (2nd edn.), Pergamon Press: Oxford, 1989.Masurel D, Sirlin C and Simon J. 22. New J. Chem.1987; 11: 455–456.




INTRODUCTION

Chlorophylls, [1] among them chlorophyll a 1, the green pigment of photosynthesis in plants and cyanobac-teria, form the main structure of the chlorin (dihydropor-phyrin) class of porphyrinoids [2].

The common structural feature of this class of pig-ments is the chlorin macrotetracycle 2 with one partially reduced pyrrole ring. This minor structural modifi cation of the chlorin chromophore over the completely unsaturated porphyrins disturbs the symmetry of the chromophore, with striking effects on the color and photophysical prop-erties, predestinating chlorin to pigments of photosynthe-sis [3]. To mimic the light-induced electron transfer of photosynthesis, artifi cial photosynthetic systems consist-ing of porphyrin donors and electron-acceptor moieties (e.g. quinones and fullerenes) were designed [4]. The majority of these dyads were built from the completely unsaturated porphyrin macrocycles as donor units. The naturally occurring chlorin chromophore was used only

occasionally as a donor subunit for artifi cial photosyn-thesis devices [5]. In our ongoing investigations on the synthesis of photosynthetic model systems based on pep-tide-linked chlorin subunits [6], we aimed for preparation of the chlorin-anthraquinone dyad 6.

EXPERIMENTAL

Starting materials were either prepared according to literature procedures or were purchased from Fluka, Merck, or Aldrich and used without further purifi ca-tion. All solvents were purifi ed and dried using standard methods. All reactions were carried out under argon. 1HNMR spectra: Bruker DPX-200 Avance spectrometer; AVANCE NB-360 MHz; all chemical shifts were ref-erenced to TMS lock signal. MS and HR-MS: Finnigan MAT 8200 spectrometer [EI (70 eV), DCI (NH3, 8 mA.s-1)and ESI (solvent)]. IR: Perkin-Elmer Paragon 500 FT-IR spectrometer. UV-vis: Varian Cary 50 spectrophotometer. Fluorescence: Perkin-Elmer LS50B fl uorescence spectro-photometer. Column chromatographic separations were performed on silica gel (32–63 μm, 60 Å, ICN). Melting points are uncorrected and were determined on a Reich-ert Thermovar hot-stage apparatus or on a Gallenkamp apparatus. CD spectra: JASCO J-600 spectropolarimeter with a water-jacket cell of 1 dm. HPLC: Knauer HPLC Pump 64, Knauer RP-18 column LiChrosorb.

Synthesis of an amide-linked chlorin-anthraquinone dyad and its structural and spectroscopic properties

Thorsten Könekampa,b, Tobias Borrmanna and Franz-Peter Montforts*†a

a Institut für Organische Chemie, FB 2, Universität Bremen, Loebener Str. NW2/C, D-28359 Bremen, Germanyb Department of Chemistry, Dartmouth College, Hanover, NH 03755-3564, USA


ABSTRACT: An enantiomerically pure chlorin-anthraquinone dyad 6 was synthesized in respect to a model compound for light-induced electron transfer in the photosynthetic center. Therefore, the enantiomerically pure chlorin 4 was linked with the anthraquinone amine 5 to form a peptide-like dyad. The conformation of the dyad follows NMR investigations and a PM3 calculation. Observed quenching of the fl uorescence indicates an intramolecular energy and/or electron transfer from the chlorin donor moiety to the anthraquinone subunit of the dyad.

KEYWORDS: chlorin, anthraquinone, light-induced electron transfer, donor-acceptor dyad.


*Correspondence to: Franz-Peter Montforts, email: [email protected]

†Current address: Institute of Organic Chemistry, FB2, University of Bremen, PO Box 330440, D-28334 Bremen, Germany


SYNTHESIS OF AN AMIDE-LINKED CHLORIN-ANTHRAQUINONE DYAD 167

Preparation of {(2',5'-dioxo-pyrrolidin-1'-yl)[1S,2S ,3S ,4R)-3-cyano-1,2 ,3 ,4 ,8 ,9-hexahydro-8,8,13,14,18,19-hexamethyl-1,4-methano-23H,25H-benzo[b]porphin-2-carboxylato]} zinc-(II) 4. To a so-lution of chlorin 3 (2.3 mg, 3.8 μmol) in dry THF (1 mL) under argon atmosphere, a 5N KOH solution (0.5 mL) in MeOH/H2O (9+1) was added and heated for 45 min at 60 °C. The reaction mixture was then diluted with CH2Cl2

(10 mL) and washed with sat. NaHCO3 solution (10 mL). The aquous layer was exhaustively back-extracted with CH2Cl2. The combined organic layers were fi ltered over cotton wool and concentrated in vacuo. The blue residue was dried in a high vacuum to yield the crude chlorin carboxylic acid (2.2 mg, 3.8 μmol, 99%) after which the free carboxylic acid was dissolved in dry CH2Cl2 (1 mL). To this solution, dicyclohexyl carbodiimide (15.5 mg, 75.5 μmol) and n-hydroxy succinimide (8.7 mg, 75.5 μmol) were added and stirred for 2 h at rt under argon atmosphere. The reaction mixture was then concentrated and the blue residue was chromatographed over silica (10 g) with CH2Cl2/EtOAc (15+1) as eluent. Isother-mic recrystallization from CHCl3/n-hexane gave 1.7 mg (2.5 μmol, 65%) of the n-hydroxysuccinimide ester 4 as blue green solid, mp 200 °C (decomposed). IR (KBr): ν,cm-1 3326 (m, OH of unreacted carboxylic acid), 2928 (s, CH), 2852 (m, CH), 2238 (w, CN), 1817 (w), 1788 (w), 1742 (s, C=O ester), 1724 (m, C=O lactam), 1694 (w), 1659 (w), 1627 (s, arylC=C), 1581 (m, arylC=C), 1507 (w, arylC=C), 1443 (w, CH), 1383 (w), 1306 (w), 1254 (w), 1204 (m), 1143 (m), 1086 (m), 1062 (m), 948 (w), 714 (w), 699 (w). 1H NMR (200 MHz, CDCl3):δH, ppm 2.03, 2.07 (6H, 2s, 2 × 8-CH3), 2.47 (4H, bs, 2 × 3′-,4′-CH2), 3.00 (2H, m, 2 × 27-CH2), 3.20 (1H, m, 3-CHCN), 3.22 (3H, s, 19-CH3), 3.33, 3.34 (6H, 2s, 14-,18-CH3), 3.42 (3H, s, 13-CH3), 4.32 (1H, dd (t), 3J = 2 × 4.1 Hz, 2-CHCO2R), 4.54 (2H, bs, 9-CH2), 5.01 (1H, bs, 4-CH), 5.30 (1H, d, 3J = 4.2 Hz, 1-CH), 8.52 (1H, s, 6-CH), 8.61 (1H, s, 11-CH), 9.40 (1H, s, 16-CH), 9.66 (1H, s, 21-CH). UV-vis (CHCl3, c = 1.18 × 10-5 mol.L-1):λmax, nm (log ε) 620 (37802), 579 (4880), 502 (3816), 401 (103971), 294 (10892). HPLC (RP-18, 1.2 mL.min-1,MeOH): tR = 2 min. MS (ESI, positive, CH2Cl2/MeOH(1:1000)): m/z 690 [M]+, 713 [M + Na]+, 792 [M + K]+,

(ESI, negative (1:1000)): m/z 592 [M - C4H4NO2]-, 689

[M - H]-, 725 [M + Cl]-.

Chlorin-anthraquinone dyad 6

A solution of chlorin 4 (2.3 mg, 3.3 μmol) in dry THF (1 mL) was prepared under argon atmosphere in the dark. Aminomethyl-9,10-anthraquinone-hydrobromide (4.2 mg, 13.2 μmol, 4 equiv.) and a few drops of dry Et3Nwere added to this solution. After stirring for 18 h at rt, the reaction mixture was diluted with sat. NaHCO3 solu-tion and exhaustively extracted with CH2Cl2. The organic layer was dried by fi ltration over cotton wool, and the solvent was removed in vacuo. Column chromatography on silica gel with CH2Cl2/EtOAc (9+1) as eluent yielded 1.8 mg (2.2 μmol, 68%) of the pure chlorin-anthaquino-ne-dyad 6 as a green solid. 1H NMR (360 MHz, CD2Cl2

+ 0.05% pyridin-d5): δH, ppm 1.92, 1.94 (6H, 2s, 2-CH3

A/B), 2.90 (2H, AB system, 3J = 2 × 5.4 Hz, 2J = 19 Hz, 36-CH2), 3.10 (3H, s, 13-CH3), 3.22 (3H, s, 12-CH3), 3.29 (3H, s, 7-CH3), 3.35 (3H, s, 8-CH3), 3.60 (1H, dd, 3J = 1.5 Hz, 4J = 4.1 Hz, 38-CH), 3.86 (1H, dd, 3J = 6.1 Hz, 2J = 15.5 Hz, 42-CH2 part A), 4.14 (1H, dd, 3J = 5.6, 2J = 15.4 Hz, 42-CH2 part B), 4.08 (1H, dd, 3J = 4.1 Hz, 3J = 8.2 Hz, 39-CH), 4.5 (2H, s, 3-CH2), 4.93 (1H, s, 37-CH), 5.08 (1H, d, 3J = 4.2 Hz, 35-CH), 6.88 (1H, dd, 4J = 1.7 Hz, 3J= 8.0 Hz, 34-CH), 7.25 (1H, d, 3J = 8.0 Hz, 33-CH), 7.36 (1H, d, 4J = 1.4 Hz, 22-CH), 7.71, 7.58 (1H, ddd, 4J = 2.4 Hz, 3J = 6.3 Hz, 3J = 7.8 Hz; 1H, m, 27/28-CH), 7.58, 8.14 (dd, 4J = 2.4 Hz, 3J = 7.5 Hz, 1H; m, 1H, 26/29-CH),8.57 (1H, s, 20-CH), 8.66 (1H, s, 5-CH), 8.80 (1H, t, J= 2 × 5.0 Hz, 41-NH), 9.32 (1H, s, 10-CH), 9.34 (1H, s, 15-CH). 13C NMR (90 MHZ, CDCl3): δc, ppm 11.5 (12-CH3), 11.6 (13-CH3), 11.7 (7-CH3), 11.8 (8-CH3), 30.9, 31.2 (2-CH3 A/B), 33.8 (38-CH), 43.7 (42-CH2), 45.5 (2-C), 47.0 (35-CH), 47.6 (37-CH), 51.6 (3-CH2), 55.2 (39-CH), 55.6 (36-CH2), 92.8 (20-CH), 92.9 (5-CH), 99.4 (10-CH), 103.7 (15-CH), 124.0 (38-CN), 125.8 (22-CH), 127.0 (33-CH), 127.0 (26/29-CH), 132.1 (32-C), 132.9 (34-CH), 133.0 (23-C), 133.5 (7-C), 133.5 (25/30-C), 133.9, 134.0 (27/28-CH), 133.95 (12-C), 134.7 (13-C), 134.9, 136.6 (16-NC), 137.9 (8-C), 142.6 (19-NC), 144.5 (14-NC), 145.5 (21-C), 146.1 (18-C), 147.0 (9-NC), 147.0 (11-NC), 150.0 (17-C), 153.2 (6-NC), 159.5 (4-NC), 168.7 (1-NC), 170.7 (40-CO), 182.2 (24/31-CO). UV-vis (CH2Cl2, c = 8.77 × 10-6 mol.L-1): λmax, nm (log ε)643 (11592), 621 (31850), 578 (4878), 497 (4990), 399 (114688), 277 (18905), 257 (32202). CD (CH2Cl2, c = 8.77 × 10-6 mol.L-1): λmax, nm ([Θ]) = 275 (7819), 305 (-7798), 325 (4998), 397 (38263), 418 (-9404), 533 (28548), 553 (23747), 568 (19669), 595 (14231), 608 (-6783), 619 (14356), 651 (-25517), 689 (-29687 deg.cm2.dm-1). HPLC (RP-18, 1.2 mL.min-1, MeOH): tR = 3.4 min. MS (EI, 70 eV): m/z (%) 812 (10) [M]+, 496 (60) [M - C19H12O3N2 (RDA)]+, 466 (26) [M - C19H12O3N2

(RDA) - 2⋅CH3]+, 451 (12) [M - C19H12O3N2 (RDA)

- 3⋅CH3]+, 316 (29) [C19H12O3N2 (RDA)]+, 236 (100)

Fig. 1. Chlorophyll a 1 and chlorin 2


168 T. KÖNEKAMP ET AL.

[M - C15H10O2N]+, 209 (11). HR-MS (EI): m/z 812.24549. Calcd. for C48H40O3N6Zn 812.24533.


The synthesis of the chlorin-anthraquinone dyad 6 requires an enantiomerically pure chlorin 3 [6] with respect to the possibility of using the dyad as a subunit for larger photosynthetic devices. To obtain stereochemi-cally homogenous oligomeric systems, each of its sub-units has to be homochiral. Chlorin 3 obtained by total synthesis [6] from four monocyclic building blocks was hydrolyzed under standard reaction conditions with potassium hydroxide.

Activation of the obtained carboxylic acid was best achieved by the N-hydroxy succinimide group to yield the N-hydroxy succinimide ester 4. The carboxylic func-tion is in a neo-pentyl like environment and is therefore severely sterically hindered. Mixed anhydride-based activation groups like iso-butyl chloroformate did not work for the desired coupling reaction because the amino component reacted preferentially at the carbonyl func-tion of the activation group. The coupling of the activated chlorin 4 with the anthraquinone amine 5 was achieved in the presence of triethyl amine in tetrahydrofurane under very gentle conditions (rt, 18 h). Column chromatogra-phy yielded chlorin-anthraquinone dyad 6 which was completely structurally characterized by application of different spectroscopic methods.

Scheme 1. a) 1. KOH (MeOH/H2O), THF, 60 °C, 45 min, 99%; 2. NHS, DCC, CH2Cl2, rt, 2 h, 65%. b) 4, THF, Et3N, 68%


SYNTHESIS OF AN AMIDE-LINKED CHLORIN-ANTHRAQUINONE DYAD 169

The luminescence spectrum of the dyad 6 compared to the spectra of the simple chlorin 3 and its chlorin–chlorin dyad indicated the expected light-induced electron and/or energy transfer of the donor-acceptor system. A com-parison (Fig. 2) of the bare chlorin 3, its chlorin–chlorin dyad and the chlorin-anthraquinone dyad 6 after excita-tion at the wavelength of the Soret band (400, 394 and 399 nm, respectively) showed very strong emissions at 627 nm for the chlorin 3 and its dimer, but complete quenching of the luminescence at 627 nm for the chlorin-anthraquinone dyad 6.

Quite unique emissions at the wavelength of the Soret band are observed for the chlorin–chlorin dyad (394 nm) and the chlorin-anthraquinone dyad 6 (401.5 nm). These emissions correspond to S2 - S0 transitions, which, according to Kasha’s rule [7], should not be allowed. Further photophysical studies are required to elucidate why Kasha’s rule is violated by the dyads and not by the simple chlorin 3.

The structure of the dyad 6 was established by 1H NMR and 13C NMR spectroscopy, which both showed the com-plete set of proton and carbon signals. In addition, mass spectrometry showed the molecular mass of 812 Da and a major fragment ion of 496 Da (M+ - C19H12O3N2) cor-responding to a retro-Diels-Alder-type fragmentation at the norbornene moiety, with the anthraquinone unit split-ting off from the chlorin together with a fumaric nitrile amide part.

NOE measurements indicated mainly the proximity between the protons of the terminal ring of the anthraqui-none and 7,8-methyl groups of ring B of the chlorin. A second set of NOEs were found for the substituted aromatic ring of the anthraquinone unit and the ring C/15-methine region of the chlorin moiety. A semiem-pirical PM3 calculation [8] taking these NOEs into

account shows a cofacial arrangement of the chlorin and the anthraquinone units with a minimum distance of 5.2 (Fig. 3) at the linkage region and 6.6 between the centers of both subunits.

The minimum distance of 5.2 between the two sub-units should not allow any ground-state electronic interac-tions (charge transfer) between donor and acceptor. This is refl ected by the UV-vis spectrum of 6 which shows a superposition of the single spectrum of anthraquinone 5 and the corresponding chlorin 3 and no spectral perturbation.

CONCLUSION

With synthesis of the chlorin-anthraquinone dyad 6 we have a donor-acceptor unit ready which shows light-induced electron or/and energy transfer. The cyano

Fig. 2. Luminescence spectra of a) single chlorin 3 (solid bold line), b) chlorin–chlorin dimer (dotted line), and c) chlorin- anthraquinone 6 (dashed thin line) in CHCl3 (c = ~ 10-6 M) after excitation at the individual Soret bands (400, 394, and 399 nm, respectively)

Fig. 3. Conformational analysis of dyad 6 based on NOE correlations and PM3 calculations


170 T. KÖNEKAMP ET AL.

function at the norbornene moiety of the dyad shouldalso allow peptide-like linkages to other units, e.g. donor–donor structures giving more complicated photosynthetic devices coming structurally very close to the the natural photosynthetic reaction centers.

Acknowledgements

This work was supported by the Deutsche Forsc-hungsgemeinschaft (DFG). We thank Mr. J. Stelten for NMR spectroscopic investigations, and Dr. T. Dülcks and Mrs. D. Kemken for recording the MS spectra.

REFERENCES

a) Kadish KM, Smith KM and Guilard R. 1. The Por-phyrin Handbook, Vol. 1, Academic Press: San Diego, 2000. b) Scheer H. Chlorophylls, CRC Press: Boca Raton, 1991. c) Montforts FP and Glasenapp-Breiling M. In Progress in Heterocyclic Chemistry, Gribble GW and Gilchrist TL. (Eds.), Vol. 10, Pergamon: Oxford, 1998; 1–24.a) Montforts FP, Gerlach B and Höper F. 2. Chem. Rev.1994; 94: 327. b) Montforts FP, Glasenapp-Breiling M, Kusch D. In Houben-Weyl Methods of Organic Chem-istry, Vol. E9d, Schaumann E. (Ed.), Tnieme: Stuttgart, New York, 1998; 577–716. c) Montforts FP, Glasenapp-Breiling M. In Progress in the Chemistry of Organic Natural Products, Vol. 84, Heiz W, Falk H, Kirby GW. (Ed.), Springer: Wien, New York, 2002, 1–51.

d) Galezowski M and Gryko DT. Current Org. Chem.2007; 11: 1310–1338.a) Deisenhofer J. 3. The Photosynthetic Reaction CenterNorris JR. (Eds.), Academic Press: San Diego, 1993. b) Huber R. Angew. Chem. 1989; 101: 849; Angew. Chem. Int. Ed. Engl. 1989; 28: 848. c) Deisenhofer J and Michel H. Angew. Chem. 1989; 101: 872; Angew. Chem. Int. Ed. Engl. 1989; 28: 829.a) Gust D and Moore TA. In 4. Topic in Current Chemis-try, Photoinduced Electron Transfer III Mattey J. (Ed.), Vol. 159, Springer: Berlin 1991, 103. b) Wasielewski MR. Chem. Rev. 1992; 92: 435. c) Gust D, Moore TA and Moore AL. Acc. Chem. Res. 1993; 26: 198. c) Kurreck H and Huber M. Angew. Chem. 1995; 107:929; Angew. Chem. Int. Ed. Engl. 1995; 34: 849.a) Tauber AY, Kostiainen RK and Hynninen PH. 5.Tetrahedron 1994; 50: 4723. b) Borovkov VV, Gribkov AA, Kozyrev AN, Brandis AS, Ishida A and Sakata Y. Bull. Chem. Soc. Jpn. 1992; 65: 1533. c) Maruyama K, Yamada H and Osuka A. Chem. Lett. 1989; 5: 833. d) Tauber AY, Helaja J, Kilpeläinen I and Hynninen PH. Acta Chem. Scand. 1997; 51: 88. e) Abel Y and Montforts FP. Tetrahedron Lett. 1997; 38: 1745.Könekamp T, Ruiz A, Duwenhorst J, Schmidt W, 6. Borrmann T, Stohrer WD and Montforts FP. Chem.Eur. J. 2007; 13: 6595–6604.Kasha M. 7. Discuss. Faraday Soc. 1950; 9: 14. HyperChem program package Professional 8.0; 8. Hypercube, Inc: Gainesville (FL, USA).


CONTENTS


Synthesis of all-cis and all-trans tetrakis (phenyl vi ny lene)phthalocyanines 1Alexander Efi mov*, Essi Sariola and Helge Lemmetyinen

Photoinduced electron transfer in supramolecu lar as semblies involving saddle-distorted 14porphy rins and phthalocyanines

Takahiko Kojima*, Tatsuaki Nakanishi, Tatsuhiko Honda and Shunichi Fukuzumi*

New aspects of porphyrins and related com pounds: self-assembled structures in two-di men sional 22molecu lar arrays

Katsuhiko Ariga*, Jonathan P. Hill*, Yutaka Wakayama, Misaho Akada, Esther Barrena and Dimas G. de Oteyza

Interaction of nitric oxide with Ru(II) complexes of deu teroporphyrindimethylester derivatives 35Rudy Martin, Roberto Cao*, Ana M. Esteva and Franz-Peter Montforts*

Structural determination of the 2,10-phytadienyl subs tituent in the 17-propionate of bacteriochlorophyll-b 41from Halorhodospira halochloris

Tadashi Mizoguchi*, Megumi Isaji, Jiro Harada, Kazuyuki Watabe and Hitoshi Tamiaki*

Synthesis of porphyrin-carbohydrate conjugates using “click” chemistry and their preliminary evaluation 51in hu man HEp2 cells

Erhong Hao, Timothy J. Jensen and M. Graça H. Vicente*

Structures and properties of hemiquinone-subs tituted oxoporphyrinogens 60Jonathan P. Hill*, Katsuhiko Ariga and Francis D’Souza

Discotic liquid crystals of transition metal com ple xes 40: unique double-clearing behavior of a se ries of novel discotic metallomesogen based on bis (phtha lo cyaninato)europium(III) complexes 70

Hidetomo Mukai, Miho Yokokawa, Kazuaki Hatsusaka and Kazuchika Ohta*

Spectral fi ngerprinting of porphyrins for dis tri buted chemical sensing 77Daniel Filippini*, Emanuela Gatto, Adriano Alimelli, Muhamad Ali Malik, Corrado Di Natale,

Roberto Paolesse, Arnaldo D’Amico and Ingemar Lundström

Electrical transduction in phthalocyanine-based gas sensors: from classical chemiresistors to new func tional structures 84

Marcel Bouvet*, Vicente Parra, Clémentine Locatelli and Hui Xiong

Solvent-induced supramolecular assemblies of crown-substituted ruthenium phthalocyaninate: morphology of assemblies and non-linear optical properties 92

Antonina D. Grishina, Yulia G. Gorbunova*, Victor I. Zolotarevsky, Larisa Ya. Pereshivko, Yulia Yu. Enakieva, Tatiana V. Krivenko, Vladimir Savelyev, Anatoly V. Vannikov and Aslan Yu. Tsivadze

Ground- and excited-state interactions of 2,7,12,17-te tra phenylporphycene with model target 99biomolecules for type-I photodynamic therapy

Noemí Rubio, Víctor Martínez-Junza, Jordi Estruga, José I. Borrell, Margarita Mora, M. Lluïsa Sagristá and Santi Nonell*

Investigating different strategies towards the prepara tion of chiral and achiral bis-picket fence corroles 107Martin Bröring*, Markus Funk and Carsten Milsmann

Bisporphyrin connected by tetrathiafulvalene 114Kazuya Ogawa* and Yasunori Nagatsuka

Synthesis, structures, and properties of BCOD-fused por phyrins and benzoporphyrins 122Hiroki Uoyama, Takahiro Takiue, Kazuyuki Tominaga, Noboru Ono and Hidemitsu Uno*

Addition of carbenes derived from aryldiazoacetates to arenes using chloro(tetraphenylporphyrinato)iron 136as catalyst

Harun M. Mbuvi and L. Keith Woo*

Synthesis and photophysical behavior of axially subs tituted phthalocyanine, tetrabenzotriazaporphyrin 153and triazatetrabenzcorrole phosphorous complexes

Edith M. Antunes and Tebello Nyokong*

A fi rst ABAC phthalocyanine 161Fabienne Dumoulin*, Yunus Zorlu, M. Menaf Ayhan, Catherine Hirel, Ümit Isci and Vefa Ahsen*

Synthesis of an amide-linked chlorin-anthraquinone dyad and its structural and spectroscopic properties 166Thorsten Könekamp, Tobias Borrmann and Franz-Peter Montforts*

Transfer of Copyright Agreement and Publishing Agreement

The transfer of copyright from the Author(s) should be explicitly stated to enable the Publisher to disseminate the Work to thefullest extent. The following transfer must be signed and returned to the Publisher’s offi ce. The work will not be published until reception of the signed agreement.

World Scientifi c Publishing Co Pte Ltd5 Toh Tuck Link,

Singapore 596224.Fax: +65 6467 7667 - Email: [email protected]

The work entitled: ______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

The Author(s) guarantees that the above article is the Author(s)’ original Work, and has never been published elsewhere before and that the text, the illustrations, and any other materials included in the article do not infringe upon any existing copyright.

The Author(s) transfers to JPP, during the full term of copyright, free of charge, the exclusive rights comprised in the copyrightof the Work, including but not limited to the right to publish the Work and the material contained therein throughout the world, in all languages, and in all media of expression now known or later developed, and to license or permit others to do so. However, the following rights are reserved by Author(s):

– the right to use, free of charge, all or part of this article in future Works of their own, such as books or lectures; – the right to reproduce the article for their own purposes provided the copies are not offered for sale; – the right to make copies of all or part of the Work for the Author(s)’ use in classroom teaching.

The Author(s) indemnifi es the Publisher and the Editors of the Journal of Porphyrins and Phthalocyanines for, and holds them harmless from, any loss, expense or damage occasioned by a claim or suit by a third party for copyright infringement, or any suit arising out of any breach of the foregoing warranties as a result of publication of his article.

Note: corrections, additions or deletions introduced at the proof stage which constitute a departure from the originally submitted and accepted work will be incorporated at the Authors’ expense.

Signature* ___________________________ Date: ________________________________

Printed name: ___________________________________________________________________

*If signed by only one author on behalf of all coauthors, the additional statement must be signed.

« I represent and warrant that I am authorized to execute this copyright on behalf of all the authors of the article referred above »

Signature ____________________________ Date: ________________________________

The source for:❖ Synthetic Porphyrins❖ Custom Porphyrins❖ R&D Porphyrins

❖ Porphyrins for Pharmaceutical and Diagnostic Applications❖ Natural Porphyrins and Derivatives

❖Metalloporphyrins❖Chlorins❖ Pyrroles

Contact details:U.S. Office

Frontier Scientific, Inc.P.O. Box 31

Logan, Utah 84323-0031USA

Tel.: +1 (435) 753-1901Fax: +1 (435) 753-6731

Email: [email protected]: [email protected]

www.frontiersci.com

European Office

Frontier Scientific Europe, Ltd.Carnforth, Lancashire, LA6 1DE

United KingdomTel.: +44 (0) 1524 730551Fax: +44 (0) 1524 730551

Email: [email protected]

Japan Office

Funakoshi Co., Ltd.9-7 Hongo 2-ChomeBunkyo-ku, Tokyo 113-0033JapanTel.: +81 33 292 6641Fax: +81 33 292 6645Email: [email protected]

PORPHYRINS

Documents

Jan2009-Journal of Porphyrin and Pthcyns