Cryogenic stabilization of high vapor pressure samples for surface analysis underultrahigh vacuum conditionsJ. J. Bruckner, K. Wozniak, S. Hardcastle, A. Sklyarov, S. Seal, and T. L. Barr Citation: Journal of Vacuum Science & Technology A 17, 2668 (1999); doi: 10.1116/1.581928 View online: http://dx.doi.org/10.1116/1.581928 View Table of Contents: http://scitation.aip.org/content/avs/journal/jvsta/17/5?ver=pdfcov Published by the AVS: Science & Technology of Materials, Interfaces, and Processing Articles you may be interested in A ultra-high-vacuum wafer-fusion-bonding system Rev. Sci. Instrum. 83, 055108 (2012); 10.1063/1.4718357 Real-time observation of the dry oxidation of the Si(100) surface with ambient pressure x-ray photoelectronspectroscopy Appl. Phys. Lett. 92, 012110 (2008); 10.1063/1.2830332 Cryogenic variable temperature ultrahigh vacuum scanning tunneling microscope for single molecule studies onsilicon surfaces Rev. Sci. Instrum. 75, 5280 (2004); 10.1063/1.1818871 Construction and performance of an ultrahigh vacuum-compatible high temperature vapor dosing system for lowvapor pressure compounds J. Vac. Sci. Technol. A 21, 491 (2003); 10.1116/1.1550011 Ultrahigh vacuum surface analysis coupled with a high pressure system for the study of near critical andsupercritical fluid processing Rev. Sci. Instrum. 68, 184 (1997); 10.1063/1.1147806

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

Cryogenic stabilization of high vapor pressure samples for surfaceanalysis under ultrahigh vacuum conditions

J. J. Brucknera)

Wilson, Sonsini, Goodrich, and Rosati, Professional Corporation, Palo Alto, California 94304

K. WozniakChemistry Department, Warsaw University, ul. Pasteura 102 093 Warsaw, Poland

S. Hardcastle and A. SklyarovAdvanced Analytical Facility, University of Wisconsin–Milwaukee, Milwaukee, Wisconsin 53201

S. SealAdvanced Materials Processing and Analysis Center, Department of Mechanical, Materials and AerospaceEngineering, University of Central Florida, Orlando, Florida 32816

T. L. BarrDepartment of Materials and Laboratory for Surface Studies, University of Wisconsin–Milwaukee,Milwaukee, Wisconsin 53201

~Received 12 March 1998; accepted 11 June 1999!

A novel form of ultrahigh vacuum~UHV! cryogenic stabilization has been used to obtainhigh-resolution x-ray photoelectron spectroscopy~XPS! data from a complex amine,1,8-bis~dimethylamino!naphthalene, whose solid phase exhibited at room temperature anunacceptably high rate of sublimation. Protonated versions of the amine exhibit hydrogen bonding.Electron spectroscopy for chemical analysis chemical shifts can be used to describe the strength andasymmetry of hydrogen bonding formed in proton sponge complexes. Analyzing the binding energyshifts of N ~1s! induced by the presence of this hydrogen bonding required obtaining correspondingXPS spectra from the nonprotonated~reference! sample, but the reference sample sublimes undereven moderate vacuum conditions. The combined results suggest that other high vapor pressurematerials, particularly those that were previously considered to be too corrosive for routine analysis,can be cryogenically stabilized for surface analysis under similar UHV conditions. ©1999American Vacuum Society.@S0734-2101~99!09605-0#

I. INTRODUCTION

X-ray photoelectron spectroscopy~XPS!, also known aselectron spectroscopy for chemical analysis~ESCA!, pro-vides useful electron density data that are indicative of thechemical state of the atomic elements in the surface region ofa solid sample. XPS is therefore a surface analysis techniquethat analyzes photoelectrons emitted from the outer 10–80 Åof the surface of a sample.1 Generally in this region, therewill be some adsorption of, and chemical involvement with,air induced species including adsorbed carbonaceous species~particularly hydrocarbons!, which can be designated as ad-ventitious carbon. The latter are always found on air exposedsurfaces. The resulting adsorbed oxides and carbonaceousspecies are relatively inert to more deleterious environmentaleffects, thus providing a modest form of ‘‘natural’’ passiva-tion that can be useful to protect some samples.2

The nature of these environmental effects may usually berecognized and separated from the detected properties of theexperimental surface of interest. Moreover, the persistent ad-ventitious hydrocarbons often provide a basis for establish-ing the photoelectron binding energy scale needed for XPSchemical analysis, especially when the sample of material ofinterest exhibits charging~i.e., a lack of coupling to the

Fermi edge of the instrument! due to low electronicconductivity.3,4

Many spectroscopies cannot directly detect the lowestatomic number elements. One of the most often employednegative statements concerning electron spectroscopies,@i.e.,both XPS and Auger electron spectroscopy~AES!# is thatthey cannot be employed to detect hydrogen. In the minds ofmany this contention has often been overextended to thestatement that XPS is blind to the effects of hydrogen. Thisis incorrect. Such overgeneralization ignores the well knownfact that many of the chemical determinations in XPS arebased on the binding energy shifts induced into the spectra ofan atom A, for example, carbon, by another atom B, forexample, hydrogen, in an AxBy system. Thus, while it is truethat hydrogen cannot be directly detected by XPS, most sur-face scientists are aware that the effects of hydrogen can bedetected by XPS by observing the effects of hydrogen incompounds on its adjacent atoms~e.g., on the carbonspectra!.5–8 In the present case, the effect of hydrogen bond-ing was detected by XPS by observing the progressive ef-fects of that bonding on the nitrogen~1s! spectra of the vari-ous complex amines involved.

All commercial XPS systems require that the sample beheld in ultrahigh vacuum~UHV! of approximately 1(10)28

Torr or better. However, many materials that are solid ata!Electronic mail: john@bruckner.net

2668 2668J. Vac. Sci. Technol. A 17 „5…, Sep/Oct 1999 0734-2101/99/17 „5…/2668/8/$15.00 ©1999 American Vacuum Society

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

standard temperature and pressure~STP! vaporize by subli-mation at pressures well above 1(10)28 Torr, thereby pro-hibiting the use of conventional, room temperature XPStechniques for chemical state analysis of such volatile mate-rials.

It has been demonstrated that with the use of differentialpumping XPS analysis can be achieved at pressures as highas 1(10)23 Torr. However, most XPS systems are notequipped with this intricate and often costly capability. Fur-ther, at the typical XPS operating temperature of approxi-mately 300 K, many high vapor pressure materials of interestmay vaporize too much at pressures above 1(10)23 Torr inorder to successfully employ conventional pumping to main-tain these pressures.

A novel route around this dilemma is to cool such volatilesample materials to a sufficiently low temperature~e.g., liq-uid nitrogen temperatures! before vacuum conditions are im-posed. Cooling the sample sufficiently allows the investiga-tor to take advantage of the resulting lower free energy oftenreducing the vaporization rate below a level compatible withthe pumping system, thus stabilizing an otherwise volatilematerial as a solid, despite high vacuum and, in many cases,even UHV conditions.

However, in most XPS systems, the cooling subsystem isdesigned to cool a sample only after it has already beensituated in a preparation and/or analysis chamber under UHVconditions. If, due to the requirement for stabilization, thesample must be frozen before entry into the UHV analysischamber, the low temperatures of a liquid nitrogen regimecan wreak havoc with certain aspects of the sample handlingand transfer components~e.g., entry seals! due to dissimilari-ties in thermal expansion coefficients. More particularly, ifthe sample must be cooled in an introduction chamber toachieve stabilization prior to being moved into an UHVpreparation chamber, and then moved again into the UHVanalysis chamber, the process of transferring the frozensample can be problematic. Specifically, the UHV seals canbe adversely affected by the low temperature regime~e.g.,VG O rings!. An example of such an adverse effect includesdifferential shrinkage of gasket materials. Further, in the caseof polymeric gaskets, their substantially increased viscosityis also an adverse effect. The latter effect can be particularlyonerous where the polymeric gaskets themselves are movingparts during the sample handling dynamics~e.g., HP gas-kets!.

In the present case, these problems are addressed bymodifying the usual pumping protocol to one in which thepreparation chamber is first backfilled to atmospheric pres-sure with argon and the sample is then moved from the in-troduction chamber to the preparation chamber prior to im-posing UHV conditions.

II. EXPERIMENT

The high vapor pressure sample that was stabilized withthe extended pumping protocol was nonprotonated 1,8-bis~dimethylamino!naphthalene~DMAN !. Also known asN,N,N,N 8 tetramethyl-1,8-naphthalenediamine, this material

has been assigned Chemical Abstracts Service No. 20734-58-1. DMAN has a melting point from 49 to 51 °C at apressure of 760 Torr. DMAN is readily available commer-cially from Lancaster Synthesis Ltd., Windham, NH. In Fig.1~a!, the nonprotonated species of DMAN is illustrated.DMAN is a strong base that due to its propensity to incor-porate H1 units is often labeled as a proton sponge.9,10

The protonated species of DMAN include the sharedbonding of one hydrogen atom.11 In Fig. 1~b!, the hydrogenof interest is depicted between the two aniline/amino typenitrogen atoms. This hydrogen is known to bind between thetwo aniline/amino type nitrogen atoms into structures thathave been confirmed by nuclear magnetic resonance~NMR!and x-ray diffraction~XRD! studies.6,8,10,12–16

For those involved in studies of proton sponges it is wellknown ~both from theoretical and experimental studies! thatthere is only one hydrogen in the@N–H . . . N#1 hydrogenbridge.6–8,10,12,14This has been confirmed byab initio calcu-lation as well as by numerous solution and solid state NMRresults and x-ray and neutron diffraction studies. There is notenough space for two hydrogen atoms in the@N–H . . . N#1

hydrogen bridge although some proton sponge complexescan have disordered H bonding with the proton jumping overthe two sites close to the nitrogen atoms. This is a dynamicsituation with the proton tunneling between the two allowedpositions. On average this would give two hydrogens/protonswith the occupancy factors equal to 50% located in such a Hbridge but it does not mean that these two hydrogen posi-tions are occupied at the same time. On the contrary, onlyone of them is occupied. They both are occupied in a statis-tical manner with equal probability~50%!. However such adisordered H bond is somewhat of an exception and in gen-eral far more often the ordered asymmetric@N–H..N#1 hy-drogen bridge is found in proton sponge complexes.

Direct proof of the presence of H in the hydrogen bondcomes from solid state NMR which clearly shows the pres-

FIG. 1. ~a! Nonprotonated DMAN.~b! Protonated DMAN.

2669 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2669

JVST A - Vacuum, Surfaces, and Films

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

ence of only one H atom with extremely large chemicalshifts ~; 18 ppm compared with typical values for hydrogenatoms in the range from;1 ppm for the aliphatic H attachedto carbons, up to 10–12 ppm for protons involved in a typi-cal hydrogen bonding!. This extra deshielding of the protonin the proton sponge complexes comes from cooperative in-teractions and close proximity of the lone electron pairs ofthe nitrogens atoms.

As we will amplify below, different acids generally ex-hibit different degrees and positions of the resulting H1 in-corporation. Variations in the linearity and centricity of theresulting N–H bonds are crucial. They depend on thestrength of interactions with the closest counterion present inthe crystal lattices of the ionic complexes of DMAN. Usu-ally, the closest counterion to the DMANH1 cation is lo-cated in front of the bulky methyl groups and it has a stronginfluence on the proton involved in the@N–H . . . N#1 hydro-gen bridge. A nice manifestation of such secondary interac-tions are the values of the NHN hydrogen bond angle whichis smaller by;5° in the crystal complexes than the valueobtained from ab initio calculations for the isolatedDMANH1 cation not interacting with any anion.9

The present x-ray photoelectron spectroscopy analysiswas achieved using a modified Hewlett Packard ESCA spec-trometer~HP 5950A!. The base pressure of this XPS systemis generally on the order of 1(10)29 Torr.

This XPS system employs a three Si crystal x-ray mono-chromator to improve energy resolution and sensitivity. Anysample charging was removed with a specially designed lowenergy electron flood gun. A nonmonochromatic systemwould have a more difficult time with such sensitivesamples.

This XPS system uses a long cylindrical sample probe totransfer a sample located near the distal tip of the sampleprobe by sliding the sample probe in toward the center analy-sis chamber. The proximal end of the sample probe is alwaysexposed to ambient. There is a first vacuum seal between afirst pair of Teflon™ compression rings that are part of thepreparation chamber and the outer surface of the sampleprobe. In addition, there is an UHV seal between a secondand third pair of Teflon™ compression rings that are part ofthe preparation chamber and the outer surface of the sampleprobe. The second and third pair of rings are situated be-tween the analysis chamber flange and the first compressionring. In addition there is an UHV isolation valve between thepreparation and analysis chambers. This permits the pressurein the preparation chamber to be as high as ambient forsample loading while the analysis chamber remains at UHV.

One form of sample probes available is equipped with aliquid nitrogen cooling system. The liquid nitrogen coolingsystem controller can maintain a stable temperature as low as120 K. Liquid nitrogen supplied to an inlet at the proximalend of this probe is routed down its length through a conduitto a region adjacent to the sample holding slots, which arelocated near the distal tip of the probe. This N2 ~liquid! isthen routed back up the length of the sample probe throughanother conduit to an outlet at the proximal end. A solenoid

valve controlled by a predetermined setpoint acts to controlthe flow of liquid nitrogen through the probe. While the com-pression rings of the UHV seals were designed to maintainUHV when a sample probe is being repositioned for sampleloading, they were not designed to maintain UHV duringinsertion when the probe is at cryogenic temperatures.

Thus, extreme cooling of the sample causes both thesample probe and~if subjected to this cooling! the compres-sion rings to undergo differential shrinkage due to dissimilarthermal coefficients. Meanwhile, the viscosity of theTeflon™ rings increases substantially due to the loweredtemperature. In addition, frost that condenses on the outersurface of the sample probe, which is exposed to ambientmoisture, will contribute substantially to difficulties in main-taining UHV as the sample probe is inserted into the analysischamber. Thus, some other means of sample insertion wasnecessary.

From previous experience, it was determined that nonpro-tonated DMAN sublimes under UHV conditions at roomtemperature. Unfortunately, there is no published phase datafor nonprotonated DMAN. In an effort to estimate the ambi-ent temperature sublimation point for the nonprotonatedDMAN, a simple bell jar was employed together with a lowvacuum mechanical pump. Complete sublimation was foundto occur at 2(10)23 Torr and approximately 295 K.

A nonprotonated DMAN sample designed for cryogenicstabilization was prepared by grinding the material with amortar and pestle. The powder was packed into a sampleholder that was then inserted into the sample probe near thedistal tip of the sample probe in a manner designed to presenta smooth sample surface on the Rowland circle provided bythe analysis chamber under optimum conditions.2

In order to begin the present experiments, the loadedsample probe was inserted into the preparation chamber atambient pressure and the cooling system was turned on. Inview of the fact that no phase information was available forDMAN, it was not known to what temperature the samplewould have to be cooled in order to retain the solid statestatus of the DMAN. The sample was initially cooled to atemperature of 150 K since this was hoped to be adequate,given the 120 K minimum for this system under UHV. Whenthe monitored temperature of the sample probe at ambientbecame stable at 150 K, the mechanical roughing pump wasturned on. As the roughing process proceeded, a modestpresence of visible frost began to build up on the externalsurfaces of the sample probe exposed to ambient atmosphere.After approximately 1 h, the pressure in the preparationchamber became low enough so that the diffusion pump wasturned on. Based on the observed vacuum level, the sampleproceeded to outgas for an extended period of time as thecondensation of primarily H2O had to be evacuated as wellas the N2, O2, etc.

Meanwhile, liquid nitrogen was fed into the sample probeon an intermittent basis to maintain the low temperaturesince heat was continuously being transferred from the am-bient surroundings to the sample probe. Further, the amountof apparent frost on the sample probe continued to increase.

The temperature setpoint on the temperature controller

2670 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2670

J. Vac. Sci. Technol. A, Vol. 17, No. 5, Sep/Oct 1999

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

was slowly increased to determine the temperature sensitiv-ity of the high vapor pressure sample. At 180 K, an unknowncondensed phase formed on the sample probe. At 210 K, thecondensed phase appeared to evaporate. The temperature set-point was raised in 10° increments to 250 K and no discern-able pressure increase was observed. Only at 270 K wassome pressure increase perceived.

As the preparation chamber was pumped down to UHV, aheat gun was used to melt the frost that continued to accu-mulate on the sample probe. After it appeared that a suitableand stable UHV pressure had been achieved, the sample wasvery slowly moved into the analysis chamber. While movingthe sample, the pressure in the analysis chamber began torise. This is believed to have been due to imperfect sealingbetween the Teflon™ gasket seals and the sample probe.Therefore, the sample was moved in modest increments ofapproximately 1 mm at a time. With each pressure rise, thepressure in the analysis chamber was allowed to stabilize andthen the slow incremental advance was repeated.

During data collection, the temperature of the sample washeld at approximately 200 K. The pressure in the analysischamber was approximately 1(10)28 Torr during data col-lection.

After data collection was complete, the sample was takenfrom the analysis chamber back to the preparation chamber.While moving the sample, the pressure in the analysis cham-ber began to rise once again. Therefore, the sample wasagain moved in modest increments of approximately 2 mm ata time. After each incremental movement, the pressure in theanalysis chamber was allowed to stabilize and then the slowincremental retraction was repeated. Finally, the gate valvewas closed and the preparation chamber ion pump wasturned off.

The preparation chamber was then isolated and backfilledto ambient pressure with argon. The liquid nitrogen supply tothe sample probe was then shut off and the sample probeslowly withdrawn from the preparation chamber. By allow-ing the sample probe to rise to ambient temperature, no un-due stress was placed on the ring seals.

The nonprotonated DMAN sample was then removedfrom the sample probe. Despite all the elaborate pumpingand sample movement precautions, it was apparent based onthe diminished thickness of the removed sample that a sub-stantial part of the sample material had sublimed.

III. XPS RESULTS AND DISCUSSION

The XPS results for DMAN reveal a singular binding en-ergy study for nonprotonated DMAN that suggests a reason-ably pure version of that material. This has permitted a de-tailed study of proton induced hydrogen bonding throughcomparison of the latter with the measured binding energies@particularly those of the N 1~s! of protonated species ofDMAN #.

Further, in related fashion we have initiated a detailedstudy of the influence of hydrogen on the XPS results ofselect oxide hydroxide, hydrocarbon and hydrogen bondingcases.5

Figure 2 illustrates a survey scan of the cryogenically sta-bilized nonprotonated DMAN sample. Carbon, nitrogen andoxygen peaks are evident. There is no apparent involvementof species such as F, Cl and Br.

Figure 3 illustrates the carbon 1s region for the cryogeni-cally stabilized nonprotonated DMAN sample. The shoulderpeak on the right in Fig. 3 is assigned to adventitious carbonand/or the hydrocarbon part of the DMAN systems. As aresult, all binding energy values were adjusted approxi-mately 1.2 eV up field to be in accord with the standardvalue of 284.6 eV for these hydrocarbon species.3

In Fig. 4 the nitrogen 1s region for the cryogenically sta-bilized nonprotonated DMAN shows a single, relatively nar-row peak, suggesting a singular species with one type ofnitrogen. Our best estimate of the N~1s! binding energy fornonprotonated DMAN is 399.5 eV based on adventitious

FIG. 2. Survey spectra for nonprotonated DMAN.FIG. 3. Carbon 1s spectra for nonprotonated DMAN.

FIG. 4. Nitrogen 1s spectra for nonprotonated DMAN.

2671 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2671

JVST A - Vacuum, Surfaces, and Films

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

carbon being assigned a binding energy value of 284.6 eV.The absence of any significant up-field peaks~i.e., oxidizedH or the hydrogen bonding effect! is evidence of the nonpro-tonated state of this cryogenically stabilized~reference!sample. The N~1s! binding energy value of 399.5 eV isconsistent with the anticipated peak position for the nitrogenin DMAN.7,8

In fact, because the DMAN molecule crystallizes in ageneral position of the orthorhombic P212121 space group,there are two independent nitrogen atoms present in thisstructure. Although the isolated molecule of DMAN is sym-metric, the DMAN molecule in the crystal lattice seems to beslightly asymmetric due to some asymmetric interactionswith the closest crystal environment~the crystal environmentof the left half of the DMAN molecule is not the same as theenvironment of the right half of the moiety!. The N ~1s!ESCA spectrum of DMAN demonstrates then that such tinyasymmetric interactions are difficult to detect with XPS. Ofcourse, protonation is a far stronger perturbation than theabove interactions and this is why its consequences can bemore readily detected with ESCA.

In Fig. 5, for comparison, the corresponding nitrogen 1sregion for the DMAN species associated with hydrogen bro-mide is shown. A plurality of up-field peak~s! is clearly evi-dent and this has been shown to be proof of the creation ofthe protonated state in DMAN. The same signal amplitudeand photoelectron binding energy scales as in Fig. 4 are usedin Fig. 5, without any relative charge shifting or relativeamplitude modifications.

Correspondingly, in Fig. 6, the nitrogen 1s regions fromboth Figs. 4 and 5 are shown superimposed. It can be seen

that the up-field signals from the protonated species areclearly absent in the reference spectra. The ESCA N~1s!signal coming from the nitrogen atoms involved in the@N-H . . . N#1 hydrogen bonding is perturbed by weak inter-actions with a counterion present in such structures. Thecounterion form, together with the DMANH1 cation, formsthe crystal structures of the DMAN salts. Such counterionsare usually located~determined from x-ray and neutron stud-ies! close to the DMANH1 cation ~in front of the methylgroups!. In fact, the@N–H . . . N#1 hydrogen bonding also hasa minor component, which comes from interactions with thisclosest counterion, and it might be better to use@NH . . . N#1

. . . X2 nomenclature to stress the role of the counterion de-noted as X2. In summary, the stronger the acid used to pro-duce the complex, the more electronegative the X2 anion,the stronger the interactions of the major component of hy-drogen bonding@NH . . . N#1 with the X2 counterion, thelarger positive charge induced at the hydrogen involved inthe hydrogen bonding and the more electronegative the Natoms.

There are a number of well established three dimensionalstructures of DMAN and its complexes. They may be foundin the Cambridge Structural Database containing structuraldata for organic crystals. This Database as well as originalpapers were used to visualize and examine the structures.According to numerous structures of DMAN complexes, theDMANH1 cation is located in such a manner in the crystallattice that it is accompanied by a counterion that is usuallyplaced in front of the four methyl groups. The anion is lo-cated in an asymmetric way—closer to one of the dimeth-ylamino groups. The location of the counterion is due toweak interactions that constitute a minor component of themulticenter@N-H . . . N#1 H bond.

For example, in a complex of 1,8-bis~dimethylamino!naphthalene ~DMAN ! with 1,2-dichlormaliec acid, due to strong electrostatic forces substan-tially every DMANH1 cation is surrounded by sixnegatively charged ClMH- moieties and vice versa~see Ref.6!. Any sites in such complexes without adjacent counterionsare statistically insignificant.

All these facts are well known. As another example, seeRef. 16.

In a related manner, a series of samples of protonatedDMAN were achieved by reaction with a variety of acids.These were analyzed by XPS, as well as byNMR/XRD.6,8,10,12,14 The spectroscopic and diffractionaldata clearly confirm the presence of only one proton in the@NH . . . N#1 hydrogen bonding that is exhibited by protonsponges. Cryogenic stabilization was not required for theseprotonated species.

In Fig. 7, the nitrogen 1s XPS spectra for this variety ofprotonated species of DMAN is shown. The variable natureof the resulting hydrogen bonding is evidenced by the diver-sity of the up-field range of the peaks~labeled I! at the left ofeach spectrum. The peak on the right of each these spectra~labeled II! lines up with, and is attributed to, the parentDMAN.7,8 Spectral deconvolution in Fig. 7 was accom-

FIG. 5. Nitrogen 1s spectra for HBr protonated DMAN.

FIG. 6. Overlay of Figs. 4 and 5.

2672 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2672

J. Vac. Sci. Technol. A, Vol. 17, No. 5, Sep/Oct 1999

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

plished after removal of the Shirley background by minimiz-ing chi squared using a standard peak-fitting software pack-age. All of the core level photo emission peaks weresuccessfully fitted with a Gaussian-Lorentzian distribution.As noted above, the binding energy reference has been madewith respect to the carbon~1s! peak at 284.6 eV. Reasons fordifferent shifts depending on pH are discussed below.

It can be appreciated that the magnitude~relative height!of the hydrogen bonding peak is a recognized function of therelative amount of H1 accumulation and, therefore, may be afunction of the acidity of the proton donating reactant. In thisregard, the height of the hydrogen bonding peak~I! is muchgreater in the case of highly acidic hydrogen fluoride than inthe case of mildly acidic hydrogen bromide. This is not sur-

prising because more acidic donating reactants are morelikely to contribute protons to the DMAN matrix. Thus, thehigher acidity of the stronger proton donating reactants isevidenced by hydrogen bonding peaks of greater magnitude.

In these cases, hydrogen bonding results in the insertionof H1 units into the diamine. The more effective this inser-tion, the more the system replicates the 3.01 eV shift expe-rienced by NH3 in going to NH4

1 . It is anticipated that stron-ger acids are better at realizing this than weaker acids.Stronger acids tend to incorporate the H1 more effectively.

Moreover, it can be appreciated that the amplitude~lateraldisplacement! of the hydrogen bonding shift also appears tobe a function of the acidity of the proton donating reactant.In more detail, the binding energy shift for the~I! peak of thehighly acidic HF species appears to be approximately 3.5 eV.In contrast, the binding energy shift for the~I! peak of theslightly less acidic HC1 species appears to be approximately3.0 eV. Further, the binding energy shifts for the~I! peaks ofthe remaining, less acidic species appear to be from approxi-mately 2.5 to approximately 3.0 eV. Thus, the higher acidityof the stronger proton donating reactants apparently directlyinfluences the size of the binding energy shift.

Our best estimate for the effect of hydrogen bonding onnitrogen for these proton sponges is the inducement of abinding energy shift from that of DMAN at approximately399.5 eV~II ! to a range of from approximately 402 to ap-proximately 403 eV~I!. The model used to estimate the en-ergy shifts has already been reported elsewhere.6–8,10,12,14

The results for six protonated species of DMAN and theresult for nonprotonated DMAN are summarized in Table I.

It can be seen from Table I that the most acidic speciesexperience a slightly higher shift in binding energy. It shouldalso be noted that any change in the acidity of the speciesbeing reacted with DMAN seems to influence the complexityof the H bonded peak structure~i.e., peak I is generally ac-companied by a number of additional peaks to create a mani-fold!. This progressive manifold feature appears to relate tothe creation of numerous chemistries and structures. The dif-ference between the two peaks that can be deconvolutedfrom the N~1s! peak I manifold may describe the asymmetryof hydrogen bonding formed in proton sponges.

In Fig. 8 the oxygen 1s region for the cryogenically sta-bilized nonprotonated DMAN shows a single relatively nar-

FIG. 7. Nitrogen 1s spectra for six protonated DMAN species.

TABLE I. Nitrogen 1s binding energy~eV! for seven DMAN species.~NA5not available.!

Material Peak Ia Peak IIa

DMAN-HF 403.0 399.5DMAN-HCL 402.5 399.5DMAN-HI 402.0 399.5DMAN-HCIO4 402.0 399.5DMAN-HBF4 402.0 399.5DMAN-HBr 402.0 399.5DMAN nonprotonated NA 399.5

aBased on the adventitious carbon being assigned a binding energy value of284.6 eV.

2673 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2673

JVST A - Vacuum, Surfaces, and Films

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

row peak. Figure 8 is relevant because the absence of a peakat 534 eV ~i.e., OH! indicates that there was no apparentinvolvement of water~e.g., ice! with regard to the cryogeni-cally stabilized reference sample. Figure 8 proves that theresults have not been affected by water.

Turning now to the condensed phase that formed on thesample probe during the pumping, Fig. 9 is the nitrogen 1sregion for the condensed phase material. The condensedphase material was nitrogen rich. The absence of any up-fieldpeaks is evidence of the lack of any hydrogen bonding withthe nitrogen. The low nitrogen 1s binding energy for thiscondensate~compared to the stabilized reference sample!may be due to the operation of the flood gun being moreeffective with regard to the conformingly deposited conden-sate.

In addition to the spectral results, several pumping proto-cols can be gleaned from this study. As noted above, thesample must kept cool enough to prevent it from sublimingas the preparation chamber is pumped down. However, thesample probe must also be kept mechanically engaged withthe seals to maintain the vacuum and be kept free enoughfrom any condensate in order for the sample to be movedinto the analysis chamber when a sufficiently high vacuum isreached. Further, it must be noted that the cold surface of thesample probe can function as a cryotrap, condensing un-known phases that may well begin to out gas as the vacuumlevel is increased. Perhaps of greatest significance oneshould note that cryogenic transport may be substituted for

total UHV transport in this case of a vacuum sensitive ma-terial without damage of the requisite chemical analysis.

IV. CONCLUSIONS

DMAN is an excellent model system with which to studyconsequences of protonation. It is well known that theDMAN molecule sequesters a proton from a countermoietythus forming a DMANH1 cation with an asymmetric@N–H . . . N#1 H bonding. If the proton in this bonding werein the center, one would observe just one peak for N~1s!electrons. But this is not the case because the proton is closerto one of the nitrogens. This is why one can usually observea broad complex peak for N~1s! electrons, which can bedeconvoluted into two peaks for the donor and acceptor ni-trogen atoms. As a result the difference in the binding ener-gies for the donor and acceptor can measure the degree ofasymmetry of the@N–H . . . N#1 H bond. The asymmetry ofthis bond depends on the interactions with the counterion.Because N~1s! binding energies of both the donor and ac-ceptor nitrogen atoms are affected, one can compare thesevalues with the binding energies of the unsubstituted DMANbase. Such comparison allows one to estimate the overalleffect ~the strength! of hydrogen bonding in ionic complexesof proton sponges.

In the case of DMAN, we have shown that the rate ofsublimation is temperature sensitive. As a result ESCA spec-tra have been achieved for DMAN itself, thus permitting ananalysis of the total H-bonding effect induced by differentacids into this proton sponge. On the basis of these ESCAspectra some new parameters describing the strength andasymmetry of hydrogen bonding can be proposed. Specifi-cally, by comparison of the protonated and nonprotonatedforms, one can have a new measure of the strength of hydro-gen bonding. Also, the difference between the chemicalshifts for N atoms present in the molecule~which can beobtained when the deconvolution of nitrogen peaks is done!is a good measure of asymmetry of the hydrogen bonding.This makes ESCA spectroscopy useful in studies of weakinteractions.

We have shown that the degree of cooling required tostabilize a volatile sample under UHV may be modest andgood electron density data from samples that sublime at rela-tively high pressures~such as those under study here! may beobtained by moderately cooling the sample with inexpensiveliquid nitrogen. As a result, the disadvantageous cooling ef-fects on functioning vacuum equipment can be minimized bycooling the sample as little as possible and manipulating anytemperature affected parts of the instrument as slowly as pos-sible. This cryogenic stabilization technique may be usefullyextended to sample materials that were previously thought tobe too difficult to work with, such as phosphorus, or evenvarious phosphoric acids.

In conclusion, it should be noted that the ESCA peak sizeand shift arguments employed here to define degrees of hy-drogen bonding have previously been utilized to describe avariety of additional diamine systems. Interested readers arereferred to the appropriate papers.

FIG. 8. Oxygen 1s spectra for nonprotonated DMAN.

FIG. 9. Nitrogen 1s spectra for condensate phase material.

2674 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2674

J. Vac. Sci. Technol. A, Vol. 17, No. 5, Sep/Oct 1999

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30

ACKNOWLEDGMENTS

The authors are indebted to Susan J. Kerber, Ph.D., ofMaterial Interface, Sussex, Wisconsin, for her invaluable in-sights and encouragement.

1D. Briggs and M. P. Seah,Practical Surface Analysis, 2nd ed.~Wiley,Chichester, 1990!, Vol. 1.

2T. L. Barr, Modern ESCA~Chemical Rubber, Boca Raton, FL, 1994!.3T. L. Barr and S. Seal, J. Vac. Sci. Technol. A13, 1239~1995!.4G. Barth, R. Linder, and C. Bryson, Surf. Interface Anal.11, 307 ~1988!.5S. J. Kerber, J. J. Bruckner, K. Wozniak, S. Seal, S. Hardcastle, and T. L.Barr, J. Vac. Sci. Technol. A14, 1314~1996!.

6K. Wozniak, H. He, J. Klinowski, W. Jones, and T. L. Barr, J. Phys.Chem.99, 14667~1995!.

7K. Wozniak, H. He, J. Klinowski, W. Jones, T. L. Barr, and S. Hardcastle,J. Phys. Chem.100, 11408~1996!.

8K. Wozniak, H. He, J. Klinowski, T. L. Barr, and P. Milart, J. Phys.Chem.100, 11420~1996!.

9J. A. Platts, S. T. Howard, and K. Wozniak, J. Org. Chem.59, 4647~1994!.

10K. Wozniak, J. Mol. Struct.374, 317 ~1996!.11K. Wozniak, H. He, J. Klinowski, W. Jones, and E. Grech, J. Phys. Chem.

98, 13755~1994!.12K. Wozniak, H. He, J. Klinowski, and E. Grech, J. Phys. Chem.99, 1403

~1995!.13K. Wozniak, H. He, J. Klinowski, B. Nogaj, D. Lemanski, D. Hibbs, and

M. Hursthouse, J. Chem. Soc., Faraday Trans.91, 3925~1995!.14B. Nogaj, K. Wozniak, D. Lemanski, M. Ostafin, and E. Grech, Solid

State Nucl. Magn. Reson.4, 187 ~1995!.15K. Wozniak, C. C. Wilson, K. S. Knight, and E. Grech, Acta Crystallogr.,

Sect. B: Struct. Sci.52, 691 ~1996!.16K. Wozniak, T. M. Krygowski, D. Pawlak, W. Kolodziejski, and E.

Grech, Phys. Organ. Chem.10, 814 ~1997!.

2675 Bruckner et al. : Cryogenic stabilization of high vapor pressure samples 2675

JVST A - Vacuum, Surfaces, and Films

Redistribution subject to AVS license or copyright; see http://scitation.aip.org/termsconditions. Download to IP: 128.248.155.225 On: Tue, 25 Nov 2014 00:21:30