ESR OF
Bas ~
4 - . A L ~ ~ ~ ~SULFIDES ~ N Y LAND )
ETHERS
411
n Resonance Spectra of Some
Bis(4-alko:xy henyl) Sulfides and Ethers
Department of Chemistry, Ohw University, Athens, Ohw @701
Henry J. Shine Deprwrtwent of Chemistry, Texas Teeh University, Lubbock, Texas 70409 (Received August 91,1970)
P u b i i ~ ~ tcosts i ~ n assisted by the Petroleum Research Foundation and the Ohw University Research Institute
The em spectra of the cation radicals of bis(Phydroxypheny1) (Z), bis(4-methoxyphenyl) (3) sulfides and bis(4hydroxyphenyl) (e),bis(4-methoxyphenyl) (5) ethers have been investigated. The spectra of 2 and 3 have been obtained with much better resolution than previously obtained while the spectra of 4 and 5 have not been previously reported. The spectra were easily analyzed in terms of splittings from all the available protons. Molecular orbital calculations of the spin density distributions enable the g factors of these radicals to be correlated with the spin densities on the sulfur and oxygen atoms.
Introduction The formation of cation radicals from aromatic sulfides and sulfoxides has been firmly established in recent years, The same cation radical is formed by the one-electron oxidation of the sulfide and by the homolytic deoxygenation of the sulfoxide. For the most part these reactions have been carried out in concentrated sulfuric acid. This acid will remove an electron from the sulfide (and thus itself be reduced eventually to sulfur dioxide in the process) and will initiate by protonation the deoxygenation of the sulfoxide. As fm as is now known, cation radicals can be obtained from diary1 sulfides only if the aryl rings bear substituents which can stabilize the cation radical by both charge and unpaired-electron delocalization. Because of this the use of concentrated sulfuric acid as the oxidant CtLUEieS sevex e problems. Ring sulfonation is enhanced by the same substituent properties which stabilize cation-radical formation. Thus, it has been shown that in concentrated sulfuric acid diphenyl sulfide (I-) is disulfonatedi and does not form a cation radical. ~ 4 - ~ y d r o x y ~ t(2) ~ ~and e n ~bis(4-methoxyphenyl) ~~ sulfides forin cation radicals, but in low concentration and of shori liie. These molecules are too susceptible to further rcaction with concentrated sulfuric acid for sucoesisful cation-radical work. The use of concentrated sulfuric acid has another disadvantage. It cannot be used a,t low temperatures since it usually leads t o esr spectra with faiily broad lines. For example, the ew spectra of 2 rand 3 in concentrated sulfuric acid at room temperature have 5 and 9 broad lines, respectively, with which the correct assignment of on-coupling; iriteractions has been impossible. formation of eation radicals from diary1 ethers has never been reported. l z 8
I n recent years solutions of aluminum chloride in nitromethane have been used very successfully for obtaining highly resolved esr spectra at low temperatures."e The technique has been adapted not only to organosulfur c~mpounds,~J but also to hydroquinones and their ethem4t5 We now report the use of this method for the formation of the cation radicals of compounds 2 and 3 and the corresponding ethers 4 and 5. The method has proved adaptable only to the ethers and thio ethers carrying hydroxy and methoxy groups. Di-p-tolyl sulfide (6), which gave a cation radical in concentrated sulfuric acid, could not be oxidized with aluminum chloride in nitromethane. The esr spectra obtained with the compounds 2-5, however, are well resolved and permit a complete analysis of the proton hyperfine splittings. Consideration of the signs and magnitudes of these splittings constants as compared to theoretical predictions together with the measured g value variations enable one to compare the relative eflects of the oxygen and sulfur bridging atoms on the spin density distributions.
Experimental Section Bis(4-hydroxyphenyl) (2) and bis(4-xnethoxyphenyl) (3) sulfides were obtained and purified as described earlier.3 Bis(4-hydroxyphenyl) (4) and bis(4-methoxy(1) H. J. Shine, Organosulfur Chem., 93 (1967). (2) H. Schmidt, Angew. Chew. Int. Ed. Engl., 3,602 (1964). ( 3 ) H. J. Shine, M. Rahman, H. Seeger, and G.-8. Wu, J . Org. Chem., 3 2 , 1901 (1967). (4) W. F. Forbes and P. D. Sullivan, J . Amer. Chem. Soc., 88, 2862 (1966). (5) W. F. Forbes, P. D. Sullivan, and H. M. Wang, $bid., 89, 2705 (1967). (6) P. D. Sullivan, ibid., 90, 3618 (1908). (7) H. J. Shine and P. D. Sullivan, J . Phys. Chm., 7 2 , 1390 (1968).
The Journal of Physical Chemistry, Vol. 7'6,No. 3,1971
PAULD. SULLIVAN AND HENRY J. SHINE
412 phenyl) (E;) ethers were obtained commercially (Eastman Kodak) and were used without further purification. The cation radicds were formed in the A1C13-CHsN02 system as previously d e ~ c r i b e d , ~and - - ~were found to be stable for several hours at -20". For this series (2-5) of compounds the radical concentration was found to be a critical factor in obtaining good esr spectra. A certain amount of trial and error was therefore involved in obtaining optimum conditions of concentration8 with respect to the esr spectra. The g values were obtained in a dual sample cavity using the perylene radicaJ anion as a secondtq standard. The errors quoted are the errors involved in reproducing the results. The absolute g v a l ~ e sare, however, not expected to vary significantly from the values quoted since additional checks with other compounds of known g values resulted in good agreement with literature values. The computer programs for spectral simulation and molecular orbital calculations were also those previously de~cribed.~
Results ~~s(4-hydroxyphenyl) Sulfide (2). This compound gave an esr spectrum a t -50" with line widths of 70 mG which was easily interpreted in terms of two groups of four equivalent protons and a group of two equivalent protons Rith splitting constants of 1.61, 0.115, and 1.02 6 , respectively. The 1.02-G splitting may be assigned to the hydroxyl protons but the assignment of the other groups to the ortho or meta protons will be deferred until later. The g factor for this radical was found to be 2.00687 It: 0.000103. Attempts to measure the sulfur33 hyperfine interaction in this and the following compound were not successful due to the difficulty in preparing solutions with large radical concentrations. ~ s ( 4 ~ r n ~ t h ~ ~ ~ pSulfide h e n y l )(3). This compound gave a well resolved esr spectrum with line widths of @a.70 mG at --50". The analysis was again straight in terms of a group of six equivalent proton 1 nssigned to the methoxyl protons and two groups of four equivalent protons (1.55 and 0.16 G) assigned to thc ring protons. The analysis was verified with a computer-simulated spectrum. The y factor of this compound was measured at 2.00686 rt 0.00002. is(4-hydroxyphmyl) Ether (4) A well resolved esr spectrum was again obtained and could be readily analysed in terms of a group of two equivalent protons (1.64 G) asgigned to the hydroxyl protons and two groups of four-ring protons (1.18 and 0.94 G). ~s(4-rneth~xyp~~eny~) Ether (5) The well resolved (line width CLE.60 mci) esr spectrum of this compound a t --50" was analyzed in terms of a 1.76 G splitting from the !six equivalent methox3l protons and splittings of 1.11 and 1.01 G from the two groups of four equivalent protons The analysis was corroborated with a computer siinulation. The g factor for this radical was found to be 2.00364 f 0.00002. The Journal of ,?hy.&al
chemistry, Vol. 76,No. 3,1971
Discussion There are several points oi interest about these results which warrant further consideration. It is obvious from our experiments that the unpaired electron is completely delocalized over both phenyl rings in all the compounds. I n other words there is a rapid intramolecular electron transfer between the two rings. This may be contrasted with the intermediate and slow rates of intramolecular electron transfer found in the anion radicals of bis(4-nitrophenyl) sulfide and ethers and bis(2-nitrophenyl) ether.Io There are also no effects observed in the spectra from the restricted rotation of the methoxyl and hydroxyl groups as observed previously in substituted benzenesL1and biphenyls. l2 This suggests that either the potential barriers to rotation for these groups are much smaller in the compounds now being reported or that additional rotations about the central C-X-C bonds effectively average out splitting constants. An interesting correlation is observed when the splitting constants of 4 and 5 are compared with the splitting constants of tbe hydroquinone and p-dimethoxybenzene cation radicals. It is noted that the hydroxyl and the average ring proton splitting constants of hydroquinone (3.30 and 2.25 6, respectively) are almost twice the corresponding values of 4 (1.64 and 1.04 G, respectively). I n the same way, pdimethoxybenzene and 5 are related, ie., the methoxyl splittings are 3.40 and 1.76 G and the average ring proton splittings are 2.27 and 1.06 G, respectively. This type of behavior is similar to that observed for dimeric cation radicals13 and one might therefore think of 4 and 5 as similar to the dimeric cation radicals of hydroquinone and p-dimethoxybenxene. Similar correlations may a,lso exist between 2 and 3 and the cation radicals of p-(methy1thio)phenol and p (methy1thio)anisole. To date, the esr spectra of these latt,er radicals have not been completely analyzed;" however, using the data of 3 one might predict that the cation radical of p-(methylthio)ankole should have a methoxy splitting of -2.2 G and two pairs of proton splittings of 3.10 and 0.32 G. Similar predictions can also be made for the cation radical of p-(methylthio) phenol. Molecular Orbital Calculations and Assignment 0.f Splitting Constants. I n order to compare the spin (8) This is thought to be due to the formation of a two-electron -e
-e
oxidation product,
