NMR STUDIES OF SULFUR COMPOUNDS

Stephen R. Heller

Division of Nuclear Medicine,

Walter Reed Army Institute of Research

Walter Reed Army Medical Center,

Washington, D. C. 20012

Nuclear Magnetic Resonance (NMR) has become one of the standard laboratory tools in the last few years (1). Thi s summary highlights some of the recent publications which use NMR as a primary tool for the study of sulfur compounds (2).

Until recently all NMR Studies of sulfur compounds were indirect, i.e. the NMR spectra were of neighboring nuclei (¹H, ¹⁹F, ³¹P, ¹³C, ¹¹B, etc.). While ³²S (natural abundance 99.24%,), the predominant sulfur isotope, cannot be detected by NMR due to its lack of nuclear spin, ³³S (natural abundance of 0. 76'%,) does have a nuclear spin (I = 3/2). Although its sensitivity is only 0. 226°%, that of ¹H, good ³³S wide line signals were reported by Karr and Schultz3 for sphalerite (ZnS) and pyrrhotite (Fe _x-1S_x), In the latter compound two signals were found, suggesting there are two different sulfur environments in the crystal lattice.

Sulfur Ylides

Trost (4) has prepared the stable sulfur-ylide, dimethylsulfonium phenacylide (I) which was found to exist as an interconverting mixture of two enolate ions (Figure 1). The equilibrium was studied by ¹H NMR. At 39^oC, a broad singlet ( v1/2 = 9.0 Hz, methine hydrogen) appeared at DELTA = 4.20 at -26°C, sharpened ( v 1/2 = 1.5 Hz) and appeared at DELTA = 4. 42. At 118°C the singlet (v 1/2 = 2. 0 Hz) appeared at DELTA = 4. 24. No further changes were observed at higher or lower temperatures. This was interpreted as indicating a single geometric isomer present at low temperature, which becomes a mixture at higher temperature when there is rapid rotation relative to the NMR time scale about the carbon-carbon bond.

Fig. I

Nozaki and coworkers (5) have reported a new method (Figure 2) for the preparation of certain sulfur ylides.

Fig. 2

The ¹H NMR spectra of these ylides were found to be very temperature-dependent, and the implications to conformational equilibria are thoroughly discussed. In one case nonequivalence of the methylene protons in the ylide below was interpreted as indicating a pyramidal structure for the sulfur atom (Figure 3).

Fig. 3

Novel Reactions and Compounds

Shelton and Davis (6) have presented ¹H NMR and IR spectral evidence for the existence of t-butylsulfenic acid ( II), prepared as follows:

Fig. 4

The ¹H NMR spectrum (at 60.0 MHz) of the acid was observed in seven solvents (C₆H₆, CCl₄, DMSO, CH₃CN, n-heptane, cyclooctatetracene and dioxane).

In the non-aromatic solvents, the t-butyl protons in the acid appeared about 2 Hz upfield from the t-butyl protons in the sulfoxide, and 7-20 Hz downfield from the sulfoxide protons in aromatic solvents. The diamagnetic shift in the latter case was attributed to PIE complexation between the acid and solvent. The absorption peak due to the acid proton was not located, (the region searched was not given).

Sharts and Fong (7) reported that the reaction of hydrogen disulfide with phthaloyl chloride did not lead to the expected disulfide product, but rather gave thionphthalic anhydride ( III), (which readily rearranged to thiophathalic anhydride ( IV )).

Fig. 5

Thionphthalic anhydride was differentiated from the symmetrical thiophthalic anhydride by its IH NMR spectrum. In the case of the former, the aromatic complex splitting pattern was ABCD vs. A₂B₂ for the symmetrical anhydride.

Conformation Studies.

Lambert and coworkers (8) have used ¹H NMR spectroscopy to investigate the conformational properties of the group VI heterocycles and their conjugate acids,

Fig. 6

The deformations of these heterocycles with respect to the conformation of cyclohexane was analyzed using J_H-H between the alpha and beta protons. The protonation of these heterocycles led to the hydrogen being in the axial position.

Tomaeda and coworkers (9) have observed a unique deshielding effect by a polar thio-function in their ¹H NMR studies on 4-ethylthio-4-en-3-oxo steroids and their analogs.

Fig. 7

A doublet centered at DELTA = 3.78 for five of the above steroids was found. This was interpreted as being the downfield half (H_B) of an AB splitting pattern (J = 14.5 Hx) of the C-6 methylene protons. (There was essentially no hyperfine coupling from protons at C-7). The unusually large downfield shift of this doublet assigned to the C-6 equatorial proton was presumed to be due to deshielding by the ethylthio group at C-4. This unique deshielding effect by the alkylthio functional group was shown to be general by the study of 2-ethylthio-3, 5, 5-trimethylcyclohex-2-enone and its 2-substituted analogs.

Other Studies.

Bullock and coworkers (10) used ¹H NMR to study the kinetics of D-H exchange of

ALPHA ALPHA -2 H₂) phenylsulphoxyacetic acid in pure water (Figure 8).

Fig. 8

The existence of the intermediate V was shown in the reported spectra.

Fackler and Covcouvanis (11), in their paper entitled "Further on the Myth of Nickel ( IV ) - Sulfur Chelates" give ¹H NMR evidence of analogous compounds ( VI, VII ) to show that species formed by oxidation of nickel ( II ) dithiobenzoate and nickel ( II ) N-phenyldithiocarbamate do not contain nickel ( IV ) but the ligand-oxidized nickel ( II ) species.

Fig. 9

The ¹H NMR spectrum of the chelate show two CH₃ resonances separated by approximately 2, 7 Hz, which is consistent with the non-symmetrical Ni ( II ) chelate ( VI ) and not the symmetric Ni ( IV ) chelate ( VII ) shown above.

Cantwell and Harrisonl2 have reported the preparation and properties of the 1, 3 dimethyl-2 -thiopentalenyl anion ( V III ).

Fig. 1 0

The ¹H NMR spectrum of VIII shows the magnetic equivalence of the protons in the anion indicating the proposed symmetrical structure is correct, The diamagnetic shift of the protons on C -4 and C-6 (4. 75 ppm) indicate a high negative charge density at these positions.

Crampton (13) has used ¹H NMR (along with UV) spectroscopy to study the interactions of a series of trinitrobenzene derivatives with sodium sulphite in water and water/DMSO mixtures, In water it was found that each nitro-compound gave a 1:1 adduct ( IX ) at low sulfite concentrations, while at higher concentrations di-adducts were formed (X).

Fig. 11

REFERENCES

1. For a thorough and up to date text see J. W. EMSLEY, J. FEENEY, and L. H. SUTCLIFFE, "High Resolution NMR Spectroscopy", Vols. I and 2, Pergamon Press, New York, (1965, 1966).

2. All chemical shifts, DELTA, reported here are given in ppm relative to TMS. All coupling constants are given in units of Hertz (Hz),

3. Abstract of papers, 19th Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, (Cleveland, Ohio), March, 1968.

4. B. M. TROST, J. Amer. Chem. Soc., 9, 138 (1967). Dimethylsulfonium Phenacylide.

5. H. NOZAKI, D. TUNEMOTO, Z. MORITA, K. NAKAMURA, K. WATANABE, M. TAKAKU, and K. KONDO, Tetrahedron, 23, 4279 (1967). Preparative and Structural Studies on Certain Sulphur-Ylides.

6.J. R. SHELTON and K. E. DAVIS, J. Amer. Chem, Soc., 89, 718 (1967). t-Butylsulfenic

Acid,

7. C. M. SMARTS and D. W. FONG, J. Org. Chem,, 32, 3709 (1967). Reaction of Hydrogen Disulfide with Phthaloyl Chloride.

8. J. B. LAMBERT, R. G. KESKE, and D. K. WEARY, J. Amer, Chem. Soc., 89, 5921(1967).

Conformational Choraterization of Simple Group VI Heterocycles.

9. M . TOMEDA, M. INUZUKA, T. FURUTA, and M. SHINOZUKA,Tetrahedron, 24, 959 (1968) Studies on Conformation and Reactivity - V. A Nuclear Magnetic Resonance Study of 4-Ethylthio-4-EN-3-OXO Steroids and their Analogs: The Unique Dishielding Effect of Polar Thio-function.

10. E. BULLOCK, J. M. W. SCOTT, and P. D. GOLDING, Chem. Commun., 168 (1967),An

Instance of a Stereoselective Isotopic Exchange Reaction

11. .J. P. FACKLER, JR. and D. COVCOWANIS, J. Amer. Chem. Soc., 89, 1745 (1967). Further_on the Myth of Nickel ( IV ) - Sulfur C helates.

12. T. S. CANTWELL and B. L. HARRISON, Tetrahedron Lett., 4477 (1967). The 2-thiapentalenyl Anion.

13. M. R. CRAMPTON, J. Chem. Soc., (a), 1341 (1967). A Spectroscopic Study of Complex Formation Between Aromatic Nitro Compounds and Sodium Sulfit.e.