344 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS tion is a diffusion-controlled process however, the evidence supporting this contention is inconclusive at this time. In fact, determination of the rate-controlling step in this process is a large order task, since it is difficult to follow quantitatively the loss of pigment in hair, and since two impor- tant side reactions consume oxidizing agent--i.e., the previously described oxidation of amino acid residues (7), and dibasic amino acid residues as- sociate strongly with many oxidizing agents including hydrogen peroxide and persulfate (34, 35). The current knowledge of the structure of melanic pigments, the most preponderant pigments of human hair, has been described recently by Nicolaus (8) and Mason (9). Nicolaus has proposed that melanin is a complex random polymer formed from several species of the Raper scheme for the biological formation of 5,6-dihydroxyindole (36, 37) (Fig. 4), while Mason proposes that melanin is a homopolymer of 5,6-dihy- droxyindole. Regardless of the differences, both theories suggest that the indole quinone grouping (or its reduced form) is a major repeating structural unit of melanins (Fig. 5). TYROSINE DOPA H 2 H2 N/CH -- COzH • H 0 ,,,,.•...,•,• C H HO • • 2 HO HO•N•C H - C02H • I H H .5, 6 - DIHYDROXYINDOLE Figure 4. Rapcr's scheme for formation of 5,6-(lihydroxyindolc -•-co•, H H H }NDOLE QUINONE GROUPING m•E-2,•- DI•RBOXYLIC ACID PYRmE-2 •.D- TRI•RBOXYLIC AOD Figure 5. Repealing unit and fragmenls of melanins Binns and Swan (38) oxidized synthetic melanins with alkaline hy- drogen peroxide and identified only very small quantities of pyrole 2,3- dicarboxylic acid and pyrole 2,3,5-tricarboxylic acid as end products (Fig. 5). These same authors suggested that xnelanic acids are also products of this reaction however, no evidence was provided to support this sug- gestion.

BLEACHING HUMAN HAIR 345 Small quantities of these same two pyrolic acids (and other pyrolic acids) have been isolated from the oxidation of sepiomelanin with per- manganate or with hydrogen peroxide in acetic acid (39, 40). This latter reaction most likely involves attack by peracetic acid on melanin rather than attack by the hydroperoxide anion or the hydrogen peroxide mole- cule, as probably occurs in hair bleaching (41). The isolation of pyrolic acids from the oxidation of melanin is con- sistent with the previous suggestion that the indole quinone grouping is a major repeating structural unit. The low yields of isolated pyrolic acids (generally less than 1%) may be explained on the basis that these fragments are themselves sensitive to oxidation however, a second ex- planation suggests a multiplicity of sites in the pigment macromolecules that are susceptible to attack by oxidizing agents producing many unre- covered fragments. In more pertinent experiments, Wolfram et al. (10) have studied the reaction of alkaline hydrogen peroxide with melanin granules isolated from human hair. These authors suggest an initial solubilization followed by decolorization. Several products derived from melanoprotein were isolated from these reactions, including protein- aceous species rich in the dibasic amino acids, and several other products ranging up to 15,000 in molecular weight. Further identification of these fragments should provide interesting information concerning the struc- ture of melanin and the mechanism of its oxidative degradation during bleaching. The pigments in red hair are structurally different from the brown- black melanins (42). The chromophoric unit (Fig. 6) of the red pigments is of low molecular weight (560) as compared to the polymeric melanins. However, the aromatic rings of this structure are of high electron density and, as a consequence, should be sensitive to attack by oxidizing agents. One may conclude from the current knowledge of the structures of the melanins and the trichosiderins that they are similar enough to be sensi- tive to oxidation, yet dissimilar enough to react through different mecha- nistic schemes during oxidative deg•'adation. A more complete under- standing awaits further study. OH H OH __ I C02H • HO- C--CH •' NH 2 CH 2 -- •H-NH 2 C02H Figure 6. Structurc of the pigment trichosiderin suggested by Prota