SOLID STATE

ELSEVIER

Nuclear Magnetic Resonance

Solid State Nuclear Magnetic Resonance 4 (1995) 47-51

Phosphorus chemical shift tensors in dithiadiphosphetane disulfides determined by solid-state 31P nuclear

magnetic resonance

Gang Wu, Roderick E. Wasylishen * Department of Chemistry, Dalhousie Unicersity, Halifax, Nova Scotia B3H 4J3, Canada

Received 9 May 1994; accepted 17 August 1994

Abstract

The phosphorus chemical shift tensors of two dithiadiphosphetane disulfides, 2,4-bis(4-methoxyphenyl)-1,3-dithia- 2,4-diphosphetane-2,4-disulfide (1) and 2,4-bis(methylthio)-1,3-dithia-2,4-diphosphetane-2,4-disulfide (21, have been characterized by solid-state 31P nuclear magnetic resonance (NMR) measurements. The weak homonuclear dipolar interaction between the two 31P nuclei in each of these compounds has a significant influence on the 3*P NMR line shapes of static powder samples. From the weak dipolar splittings it is possible to deduce the orientation of the phosphorus chemical shift tensor. In both compounds, the intermediate components of the phosphorus chemical shift tensors, a,,, were found to be perpendicular to the plane defined by the two phosphorus atoms and the two terminal sulphur atoms in the molecule, S=P . P=S. The smallest shift components, S,,, were found to deviate approximately 14” from the P=S bond direction.

Keywords: Solid-state nuclear magnetic resonance; Dipolar-chemical shift nuclear magnetic resonance; Homonuclear dipolar coupling; Phosphorus chemical shift tensors; Dithiaphosphetane disulfides

1. Introduction

In general, the chemical shift of a nucleus in a molecule is anisotropic. That is, the precise mag- netic field experienced by a particular nucleus in a molecule depends upon the orientation of the molecule in the external magnetic field. The chemical shift of a nucleus is generally described by a symmetric second-rank tensor which can be characterized by three principal components and

* Corresponding author.

three angles that define the orientation of the principal-axis system (PAS) of the chemical shift tensor. Although the three principal components of phosphorus chemical shift tensors have been measured in numerous phosphorus-containing compounds [l], the orientations of such tensors are often unknown. The ideal technique for ob- taining information concerning the orientation of a chemical shift tensor is the single-crystal NMR method (see, e.g., Ref. [2]). However, large single crystals suitable for NMR studies are often not readily available and the analysis of the data is also often tedious. The dipolar-chemical shift NMR method [3-61 is a useful alternative, which

0926.2040/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved

SSDI 0926-2040(94)00033-6

G. Wu, R. E. Wasylishen /Solid State Nuclear Magnetic Resonance 4 (1995) 47-51

OCH;

s\ /p\

s\p/ SW,

/\ S

\ 2 S \ 2

pk S ,P\

CH3S S

CH,O 1 2

Fig. 1. Structures of compounds 1 and 2.

requires only powder samples. This method has been used to study many chemica1 shift tensors, and several studies involving the characterization of phosphorus chemical shift tensors have been reported [7-111. However, all previous studies have dealt with spin systems consisting of two directly bonded nuclei.

Here we report the determination of the phos- phorus chemical shift tensors in two dithiadiphos- phetane disulfides: 2,4-bis(4-methoxyphenyl)-1,3- dithia-2,4-diphosphetane-2,4-disulfide (1) and 2,4-bis(methylthio)-1,3-dithia-2,4-diphosphetane- 2,4-disulfide (2) (Fig. 1). It will be shown that the weak dipolar interaction between the two non- bonded 3’P nuclei has a significant influence on 31P NMR spectra of static polycrystalline sam- ples.

2. Experimental

Compounds 1 and 2 were obtained from Aldrich and used without further purification. All 31P NMR spectra were recorded on Bruker MSL- 200 (B, = 4.70 T) and AMX-400 NMR (B, = 9.40 T) spectrometers operating at “P NMR frequen- cies of 81.03 and 161.98 MHz, respectively. Cross-polarization (CP) under the Hartmann- Hahn match and high-power proton decoupling were used in acquiring all 31P NMR spectra. The 90” pulse width was 4.0 ps and the contact time was 5 ms. A recycle time of 10 s was used. The 31P NMR chemical shifts were referenced to 85% H,PO, (aqueous) by using a solid NH,H,PO, sample, which has a 31P peak at t-O.8 ppm rela- tive to 85% H,PO, (aqueous).

The static NMR line shape simulations were based upon the exact analytical expression for a homonuclear spin pair [4] and carried out with a FORTRAN program which incorporates the POWDER routine of Alderman et al. [12] for powder averaging. The calculated line shape was convoluted with a Gaussian line-broadening func- tion.

3. Results and discussion

Compound 1, also known as Lawesson’s reagent, is often used for thiation of ketones, carboxamides, esters, lactones, lactams, imides, enamines and S-substituted thioesters [13]. The crystal structure of 1 indicates that the two phos- phorus atoms are related by an inversion centre [14], hence, the two 31P nuclei in 1 are magneti- cally equivalent. Under the condition of high- power proton decoupling, the two “P nuclei in 1 constitute an isolated A, spin pair. The 3’ P magic-angle spinning (MAS) NMR spectrum of 1 is shown in Fig. 2. In the MAS spectrum, the isotropic peak is found at 41.5 ppm, flanked by many spinning sidebands, which indicates a signif- icant chemical shift anisotropy at the 3’P nucleus. The 3’P MAS NMR experiments were carried out over a large range of sample spinning fre- quencies. The line widths for peaks in the MAS NMR spectra were found to be relatively inde- pendent of the sample spinning frequency. This indicates that the two phosphorus chemical shift tensors are indeed coincident; otherwise, spin- ning frequency-dependent MAS NMR spectra

1 I I i

300 200 100 0 -100 -200 -300

ppm

Fig. 2. “P MAS NMR spectrum of 1 obtained at 4.70 T. The

sample spinning frequency was 4022 Hz.

G. Wu, R.E. Wasylishen /Solid State Nuclear Magnetic Resonance 4 (1995) 47-51 49

0

PPm Fig. 3. Observed (bottom) and calculated (top) static “P

NMR spectra of 1 at 4.70 T. A total of 512 transients were

recorded. Parameters used to calculate the static spectrum

are given in Table 1.

may be expected [15-201. For spin systems con- sisting of isolated dilute nuclei, Maricq and Waugh [211 and Herzfeld and Berger [22l have developed useful methods for extracting the prin- cipal components of the chemical shift tensor from spinning sideband intensities. However, their procedures are not applicable for spin systems consisting of two dipolar-coupled nuclei such as is the case in 1.

Therefore, one is faced with a problem that may have an infinite number of solutions. Usually, either an independent measurement of R or the symmetry that the molecule possesses may help limit the number of solutions. In the case of 1, the molecule contains a local mirror plane de- fined by two phosphorus atoms and two terminal sulphur atoms, S=P . . . P=S. Hence, one of the three principal components must lie perpendicu- lar to the mirror plane. Under this condition, it can be shown from analytical expressions that it is impossible for the component with the smallest splitting to be perpendicular to the mirror plane. Therefore, the number of possible solutions is reduced to two.

The static “P NMR spectrum of 1 obtained at From the static 31 P NMR spectrum of 1 ob- 4.70 T is shown in Fig. 3. The static spectrum tained at 4.70 T (Fig. 3), the three dipolar split- spans about 400 ppm and consists of two subspec- tings were measured as: A, = 1853 Hz, A, = 1075 tra, which is typical for an A, spin system [4]. Hz and A, = 732 Hz at the highest (S,,), interme- This is in contrast to a previous study [231 where diate (S,,) and smallest (6,J shift components, severe line broadenings in the static spectrum respectively. The error in these splittings is esti- prevented the authors from resolving the split- mated to be +50 Hz. These splittings were con- tings. It is straightforward to show that each of firmed in the static 31P NMR spectrum of 1

the two subspectra in the static NMR spectrum obtained at 9.40 T, indicating that Eq. 1 is valid of an A, spin system has the same characteristic in the present case. Otherwise, field-dependent line shape as that arising from an isolated spin splittings may be observed. The two possible solu- with anisotropic chemical shift [41. The two dipo- tions for the phosphorus chemical shift tensor lar subspectra observed in the static NMR spec- orientations are: (i) the 6,, component is perpen- trum of 1 can be simulated using the exact analyt- dicular to the molecular mirror plane or (ii) the ical expressions for the two transitions in an A, S,, component is perpendicular to the mirror spin system [4]. However, when v0 I(c?~~ - Sjj> I B plane. For the first solution, one has R = 1235 1.5 R, where R is the dipolar coupling constant Hz, wDhich corresponds to a P-P separation of between the two coupled homonuclear spins, a 2.52 A, and the orientation of the phosphorus more straightforward method can be applied. Un- chemical shift tensor is given by +i = 90”, C& = 43” der this condition, the three splittings observed in and C& = 47”. For the second solut@, R = 717 the static spectrum are directly related to the Hz, which corresponds to rpp = 3.02 A and c$, = orientation of the internuclear vector in the PAS 19”, 42 = 90” and (b3 = 71”.

of the chemical shift tensor by Eq. 1 [24].

Aj = lSR(1 - 3 cos2&), i = 1, 2, 3 (I)

where c$~ is the angle between the internuclear vector and S,,. Although the three splittings can be measured directly from the experimental NMR powder spectrum, they are not independent, i.e.,

CA;=0 (2) i= 1

50 G. Wu, R. E. Wasylishen /Solid State Nuclear Magnetic Resonance 4 (I 995) 47-51

Fig. 4. Orientation of the phosphorus chemical shift tensor in

1.

In the discussion which follows, we shall pro- vide arguments to differentiate between these two solutions. First, based on the cpstal structure of 1, the P-P separation is 3.08 A [14]. Hence, the second solution clearly yields a reasonable P-P distance, whereas the P-P separation de- rived from the first solution is much too short. Second, a close examination of the orientations of the phosphorus chemical shift tensors given by the two possible solutions also indicates that the second solution is more reasonable. According to the second solution, the smallest shift compo- nent, Ss3, is close to the P=S bond, which is in agreement with results found in phosphine sul- fides [25-271. Furthermore, preliminary GIAO calculations on the phosphorus chemical shift tensor in a simple model compound, 2,4- bis(methyl)-1,3-dithia-2,4-diphosphetane-2,4-di- sulfide, [CH, . PS212, support this assignment [28]. Therefore, the first solution may be discarded. The orientation of the phosphorus chemical shift tensor in 1 is depicted in Fig. 4. This orientation was also confirmed by carrying out a line shape simulation which is based on analytical expres- sions for a homonuclear spin pair [4] (see Fig. 3).

Compound 2, also known as Davy’s methyl reagent, was studied using the same procedure as

outlined above. The results are listed in Table 1. Using the direct dipolar coupling constant ob- tained from the analysis of the static powder 31P NMRDspectrum, 841 Hz, the P-P separttion is 2.86 A, shorter than that in 1 by 0.16 A. This value agrees well with the P-P separ$tion found in [CH, . PS,],, in which rpp = 2.906 A [30]. The orientations of the phosphorus chemical shift ten- sors in the two closely related compounds, 1 and 2, were found to be identical. As well, the princi- pal components of the phosphorus chemical shift tensors in these two compounds are similar.

4. Conclusions

We have shown that useful information about the orientation of the phosphorus chemical shift tensors can be obtained by studying NMR line shapes of static powder samples. Even when two 31P nuclei are not directly bonded, the dipolar interaction between them may still introduce in- formative features in static NMR line shapes. As more phosphorus chemical shift tensors are char- acterized, a better understanding of the aniso- tropic nature of chemical shifts will be possible. It is also important to determine both the magni- tude and orientation of the principal components in order to test the quality of the results from ab initio MO calculations.

Acknowledgements

We wish to thank Michael D. Lumsden for the GIAO calculation and for several helpful discus- sions. Also we wish to thank Dr. Klaus Eichele for helpful suggestions. This research was sup-

Table 1

Phosphorus chemical shift tensors a and dipolar coupling constants ’ in 1 and 2

Compound Si,” 6 11 s 22 6 33 nc Kd R 41 42 43

1 41.5 184 138 - 197 381 0.25 717 19” 90 71”

2 39.1 192 131 - 205 397 0.23 841 19 90 71”

a All chemical shifts are in ppm and are accurate to +2 ppm.

b Dipolar coupling constants are in Hz and are accurate to +50 Hz. ’ Span R = S,, - 6,s as defined in Ref. [29].

d Skew K = 3(S,, - 6,,)/0 as defined in Ref. [29].

G. Wu, R.E. Wasylishen /Solid State Nuclear Magnetic Resonance 4 (1995) 47-51 51

ported by the Natural Sciences and Engineering Research Council (NSERC) of Canada. All 31P NMR spectra were recorded at the Atlantic Re- gion Magnetic Resonance Centre (ARMRC), which is also supported by NSERC of Canada.

References

[I] T.M. Duncan, A Compilation of Chemical Shift Anisotropies, The Farragut Press, Chicago, IL, 1990.

[2] M.A. Kennedy and P.D. Ellis, Concepts Magn. Resort., 1

(1989) 35 and 109.

[3] D.L. VanderHart and H.S. Gutowsky, J. Chern. Phys., 49 (1968) 261.

[4] K.W. Zilm and D.M. Grant, J. Am. Chem. Sot., 103

(1981) 2913.

[5] R.K. Harris, K.J. Packer and A.M. Thayer, J. Magn. Reson., 62 (19851 284.

[lo] R.D. Curtis, M.J. Schriver, and R.E. Wasylishen, J. Am. Chem. Sot., 113 (1991) 1493.

[6] R.E. Wasylishen, R.D. Curtis, K. Eichele, M.D. Lums-

den, G.H. Penner, W.P. Power and G. Wu, in J.A.

Tossell (Ed.), Nuclear Magnetic Shieldings and Molecular Structure (NATO ASI Series C, Vol. 3861, Khtwer Aca-

demic Publishers, Dordrecht, 1993, p. 297.

[7] K.W. Zilm, G.G. Webb, A.H. Cowley, M. Pakulski and

A. Orendt, J. Am. Chem. Sot., 110 (1988) 2032.

[8] G. Wu, W. P. Power, R. E. Wasylishen and G. Baccolini,

Can. I. Gem., 70 (1992) 1229.

[9] W.P. Power, R.E. Wasylishen and R.D. Curtis, Can. I. Chem., 67 (1989) 454.

[ll] R.D. Curtis, B.W. Royan, R.E. Wasylishen, M.D. Lums- den and N. Burford, Inorg. Chem., 31 (1992) 3386.

[121

(131

1141

WI

[I61

[171

[I81

]191

[201

1211

D.W. Alderman, M.S. Solum and D.M. Grant, J. Chem. Phys., 84 (1986) 3717. B.S. Pedersen and S.-O. Lawesson, Tetrahedron, 35 (1979)

2433.

R. Kempe, J. Sieler, H. Beckman and G. Ohms, 2.

Kristallogr., 202 (1992) 159.

S. Hayashi and K. Hayamizu, Chem. Phys. Lett., 161 (1989) 158. A. Kubo and CA. McDowell, J. Chem. Phys., 92 (1990) 7156. R. Challoner, T. Nakai and C.A. McDowell, J. Chem.

Phys, 94 (1991) 7038.

K. Eichele, G. Wu and R. E. Wasylishen, J. Magn. Resort. Series A, 101 (1993) 156. G. Wu and R.E. Wasylishen, L Chem. Phys., 98 (1993)

6138.

G. Wu and R.E. Wasylishen, Inorg. Chem.. 33 (1994) 2774. M.M. Maricq and J.S. Waugh, J. Chem. Phys., 70 (1979) 3300.

[22] J. Herzfeld and A.E. Berger, J. Chem. Phys., 73 (1980)

6021.

[23] G. Klose, A. Mops, G. Grossmann and L. Trahms, Chem. Phys. Lett., 175 (1990) 472.

1241 K. Eichele and R.E. Wasylishen, .I. Magn. Reson. Series A, 106 (1994) 46.

[25] J.B. Robert and L. Wiesenfeld, Mol. Phys., 44 (1981) 319. [26] P.N. Tutunjian and J.S. Waugh, J. Chem. Phys., 76 (1982)

1223.

[28] M.D. Lumsden and R.E. Wasylishen, unpublished re- sults.

1271 P.N. Tutunjian and J.S. Waugh, J. Magn. Resort., 49 (1983) 155.

[29] J. Mason, Solid State Nucl. Magn. Reson.. 2 (1993) 285.

[30] J.J. Daly, J. Chem. Sot., (1965) 4065.