296 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 3. Can the concept of employing ultraviolet absorption to interpret di- sulfide strain be generalized to include compounds other than cyclic disul- fides ? Among the compounds which we studied were Lcystine, oxidized gluta- thione, propyl disulfide, (t) butyl disulfide, penicillamine disulfide, dithio- diglycolic acid, dithiodisalicylic acid, dihydroxy dinaphthyl disulfide, di- thiodimorpholine, dithiodimalic acid, g, •' diaminodiethyl disulfide and 0,0' dibromodibenzyl disulfide. The ultraviolet spectra of these compounds were determined as a function of pH, i.e., examinations were made under acid, neutral, and basic conditions. The conclusions that were arrived at as a result of this investigation may be summarized as follows: 1. All disulfide-containing compounds, regardless of whether or not they showed an absorption or transparency under conditions of examination employed by previous workers, revealed absorption maxima either under alkaline or acid conditions. This generalization does not apply to tertiary disulfides. 2. This behavior of disulfide in alkali or acid media permits a general classification of disulfides to be made. First, there are those compounds such as the tertiary disulfides that give a non-specific absorption which is independent of pH. The remaining disulfide compounds will show absorp- tion maxima either in acid or in alkali media. 3. The extinction coefficients of the observed maxima are continuous functions of pH. That is, if the maximum occurs in acidic medium--in- creased acidity will enhance the absorption. Likewise if the maximum oc- curs in alkali medium above a critical pH, the absorption is enhanced with increasing alkali. 4. The absorption spectra of those disulfides which show a response to changes in pH, exhibit three distinct phases: (a) non-specific absorption, (b) shoulders or inflections, and (c) absorption maximum. Phase "b" spec- trum is an artifact in the sense that it can be shown to arise as a composite of the spectra of phase "a" and "c". Therefore there are essentially only two types of SS absorption--a non-specific and a specific absorption. An interpretation of these data can be made if one assumes that it is R I doubly bonded sulfur :S-- such as --C=S-- which is behaving as the ultraviolet chromophore. Compounds such as tetra methyl thiourea which possess the --C•S group absorb at 250 mu--coincident with the region in which simple unstrained disulfides such as n-propyl disulfide are observed to R I absorb. Implicit in the postulation of the --C:S-- chromophore in di- sulfides is the assumption that some positive group such as a proton has

PROGRESS IN THE CHEMISTRY OF DISULFIDES 297 migrated from the carbon atom adjacent to the sulfur. The consequence of this step is to leave the carbon with an electron pair, i.e., a carbanion, with which it can form a double bond with the sulfur atom. This assumption seems justified in view of the fact that tertiary disulfides which have no hydrogens on the carbon joined to the sulfur only show non-specific ultra- violet absorption. Furthermore, compounds which show absorption max- ima in alkali show only non-specific absorption identical with that of terti- ary disulfides, when placed in acid medium. It is interesting to note that regardless of whether alkali or acid is re- sponsible for the generation of the ultraviolet absorption maximum of the disulfide compound, the absorp cion maximum always occurs in the region of 330 mu. This is precisely the wavelength that Calvin has found to corre- spond to the absorption maxima of the very highly strained five-membered disulfide rings found in 6,8 thioctic acid and trimethylene disulfide. It can be shown from an analysis of all the possible structures in which these com- pounds can exist that the chromophore responsible for absorption at 330 mu is the planar conjugated --C=S--S:C-- grouping. Aromatic disul- fide compounds readily assume this configuration (some in acid, others in alkali) as a consequence of an enhancement of resonance stabilization which results from an assumption of such a structure. Thus the energy required to deform the dibedral angle is more than compensated for by the gain in resonance stabilization of the final structure. Aliphatic disulfides having hydrogen atoms attached to the carbon which is joined to sulfur (these shall henceforth be designated as B hydrogens in conformity with our pos- tulated B elimination mechanism) also show this absorption. This occurs only in very strong alkali and increases with increasing time. These hy- drogens being less acidic are more reluctant to leave the carbon atom to which they are joined. In Calvin's strained cyclic 5-membered disulfide rings the • hydrogens are so acidic that they ionize readily in the presence of solvent to give the planar conjugated --C:S--S:C-- with a dihedral angle approaching that of zero. The planar ring is apparently less strained than the 5-membered ring containing the 90 ø dihedral angle. It is of in- terest to recall that the essential feature of our postulation of the group as being the fundamental ultraviolet chromophore is that we were required to assume that the loss of a proton occurred and that the resulting carbanion with its free electron pair was capable of donating the electron pair to sulfur to form the doubly bonded sulfur structure --C:S--. According to independent views of Kimball (40) and Rothstein (41) theoretically, an electron pair may be donated to one of the untilled 3d orbitals of the sulfur atom. It is therefore possible that molecules in which negatively charged sulfur atoms occur make definite contributions to the excited state. What we are essentially measuring in our ultraviolet spectra studies of disulfide compounds is an acid-base phenomenon