264 JOURNAl. OF THE SOCIETY OF COSMETIC CHEMISTS TABLE 1--STABILITY CONSTANTS TABLE 2--EFFECT OF pH ON Fe •3 CHELA- TION Order of pH Increasing Efficiency 2-3 Log Log Log 4.0-8.5 Ligand KFe +3A KcuA KcaA 8.5-10.0 HEDTA DTPA EDTA HEDTA EDTA DTPA EDTA HEDTA DTPA DHG P 8.15 EFFECT OF PH ON CALCIUM CHELATION HEDTA 19.6 17.4 •10 EDTA 25.1 18.8 10.6 pH Range of DTPA 28.6 21.03 11.0 90-100% Ligand Chelation HEDTA 8.0 EDTA 7.0 DTPA 6.5 generally present, it is always advisable to use a greater proportion than the stoichiometric or mole to mole ratio. This empirical approach is nothing new but it is really the only course to follow in problems with trace metal ion contamination. One can then observe the formulation under study to arrive at the optimum concentration required to immunize the product against the trace metal ion effects. One would be expected to use the strongest chelator available and a look at the log K values would mean the selection of DTPA. Thus, for cos- metic preparations where shelf life is an important factor, the most efficient chelating agent should be used giving consideration to the final pH of the finished product. To summarize, if one has a problem with ferric ion, Fe +a, at a pH of 7.0 to 9.5, DTPA would be the choice. If the pH is 4.0 to 7.0, the choice should either be EDTA or DTPA. If only divalent ions were present then either EDTA or DTPA could be employed. Thus, it appears that DTPA has the most suitable properties for practically all aqueous-based cosmetic systems, though the advantage of DTPA over EDTA may be marginal in many uses, but where most efficient chelation is required, DTPA should be investigated first. At this point, we would like to analyze the last few standard statements vn chelation of ferric ion that are based on facts obtained with ideal condi- tions. Suppose we have a very practical system containing a molar con- centration of ferric ion as ferric chloride and the pH of this system is adjusted to a pH of 4.0 and then 8.5, ferric hydroxide would form in each case and the addition of any amount of the chelating agents mentioned even at molar ratios 10:1 or 100:1 will show little precipitable dissolving action on the precipitated iron hydroxide. However, in our very practical system should we add about one and one-half moles of EDTA or DTPA before the ferric ion is introduced and then buffer to a pH of 4.0 or 8.5, we would find no iron hydroxide to form. Should the ferric ion be added in

ASPECTS OF CHELATION IN COSMETIC PRODUCTS 265 excess of the chelating power of the EDTA or DTPA so that ferric hydrox- ide would form, the solution would still contain chelated ferric ion prac- tically equal to the former system. If one would wait long enough in the case of EDTA at a pH of 8.5, the iron chelate in both of the latter cases would hydrolyze and would give up the iron to form iron hydroxide. At the pH condition of 4.0 and up to 6.5, the iron chelate of EDTA is exceed- ingly stable and the ferric ion is reduced to such a low value that iron hydroxide cannot form. In the case of DTPA, at both 4.0 and through 8.5, the iron is always lower than the solubility product of iron hydroxide so no precipitation will occur. Thus, the order of addition is exceedingly important in the performance of the chelating agent and in any aqueous system the chelating agent should always be the first ingredient. In prac- tical cosmetic systems where fatty acids, bacteriostats and hydroxide ion will compete for the ferric ion or other trace metal ions, it is obviously necessary to first have a sufficient amount of the chelating agent and in the most advantageous position, by introduction before the other com- peting anions. To illustrate this fact, we were approached by a manufacturer of sur- factants to help solve an iron problem in neutralizing their alkyl aryl sulfonate fraction with caustic. Phase separation and iron hydroxide flock were the problems after reaction with caustic. Addition of the Chelating agent to the neutralized or sodium alkyl aryl sulfonate solution did not dissolve the iron hydroxide nor did phase separation occur of the unreacted hydrocarbon. Once the chelating agent, DTPA, which proved to be the most effective, was added to the sulfonic acid fraction and then neutralized to a pH of 7.5 or 8.0 with caustic, clear-cut phase separation of unreacted hydrocarbon was observed and no iron hydroxide flock developed in the surfactant fraction. EDTA was investigated in this problem and it performed as well as DTPA. Shelf-life of the surfactant solution was a factor and with the final pH of 7.5-8.0, if one would wait long enough, the ferric chelate of EDTA would gradually hydrolyze and iron hydroxide would form. This did happen, but DTPA was the solution to this*:type• of product problem. The most suitable means of incorporating EDTA or DTPA into cosmetic products is the use of the acid form. This allows the use of a suitable base such as sodium, potassium, ethanolamines and ammonia for any system where compatibility is a necessity. However, there are a number of prod- uct formulations where the sodium derivative of either EDTA or DTPA can be used. The best known use of the sodium derivative of EDTA in cosmetic products is for shelf life clarity of liquid soaps. This use has developed that it is unnecessary to chill or filter the liquid soap preparation before packaging. However, filtering is still employed as a precautionary measure.