]. Soc. Cosmet. Chem., 22, 635-654 (September 17, 1971) Correlation of Skin Feel Emollients to Their Chemical Structure of ROBERT L. GOLDEMBERG, B.S., and CONSUELO P. DE LA ROSA, B.S.* Presented January 12, 1971, before the Midwest Chapter, Chicago, Ill. Synopsis--It has been found possible to quantitate the SKIN FEEL of cosmetics by use of a Skin Feel Index (SFI) proposed by the authors. This index is the ratio of Initial Slip of a product (scale 1-5) to its Total End Feel (scale 4-20) after complete dry-out on the skin. SFI values were determined for 85 cosmetic EMOLLIENTS comprising 14 chemical groups, each incorporated in a standard base at 37% of the total "residue" left on the skin by this base after evaporation of its water content. Effects of molecular weight, "oiliness," polarity, unsaturation, and chain branching were studied. Due to the unavailability of some "key" compounds, a detailed statistical or graphical analysis was not possible, but definite trends could nevertheless be observed as various factors (such as chain length or polarity) were systematically varied in this initial study. INTRODUCTION Attempting to correlate the chemical structure of raw materials with various properties of cosmetic interest (such as skin feel, sheen on the hair, or effect on viscosity ot5 lotions) has always fascinated the cosmetic chemist. However, measuring the influence of various changes in chem- ical structure implies the ability to measure numerical changes in other properties of these raw materials. Properties such as skin feel or odor, however, have generally been considered to be subjective in nature, and not amenable to numerical description. This has been a major stum- bling block. * Van Dyk & Company, Inc., Belleville, N.J. 07109. 635

636 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Where it is possible to measure properties instrumentally (e.g., viscos- ity changes or hair sheen), numerical values are obtained which can be arranged into orderly tables or depicted graphically. Grouping of data allows the investigator to seek meaningful relationships, or to compare them with some external factor which may also be changing--i.e., mo- lecular structure. In the case of skin feel, no one heretofore has attempted to assign nu- merical values the situation was analagous to trying to describe odor changes verbally. Some words, such as "oily" or "dry-velvety" or "waxy," are over-used others, such as "scroopy" (the textile chemist's description of the somewhat rough, soft-draggy feel of raw silk), may be unfamiliar to most persons. In an attempt to become more precise, word modifiers are introduced to describe varying degrees of oiliness or velvetiness, etc. This often results in very complicated circumlocutions and language which mean different things to different investigators. Worst of all, such verbal descriptions cannot be ranked or summed up as can a series of nu- merical values. Various groups of compounds can be ranked in order of their in- creasing dryness, oiliness, dragginess, or tackiness on the skin. One can- not give absolute values to such properties, but it certainly can be de- cided whether one compound is oilier or more tacky than another. By systematic evaluation of many small homologous series of compounds (ranking for changes in skin fed within each such small series), it should be possible to discern trends of changes in skin fed which relate to chem- ical structure. For example, when a series of isopropyl esters are used as emollients in a particular lotion, what is the effect on skin feel of the molecular weight of the fatty alcohol portion of the ester molecule, of branching the fatty alcohol portion of the molecule, or of successively in- creasing the degree of unsaturated bonding in a particular series of Cls esters? This investigation was an attempt to answer questions of this type. EXPERIMENTAL At first, the ranking of various series of lotions, each containing the test emollient at a fixed percentage, was tried. Later, a numerical (1 to 5) rating system for each of various qualities was adopted. Each test series was limited to 4-6 samples, the number that could comfortably be applied to the inside of the tester's arm at one time. Emollients were not tested "pure," but incorporated at 7% in a basic lo- tion whose oil phase totalled 14%, which deposited an 18.9% total residue