292 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ß substances such as 03, N2, and the rare gases of the zero period of the per. iodic table, show characteristic and rather complex absorption spectra. This region is therefore very useful for qualitative analysis of many compounds. In the region 2 to 7 •, the absorp- tion characteristic is of individual groups in the molecule. For example, all compounds with free OH will ab- sorb at about 3 /•. All compounds with C•---O will absorb in the 5 to 6 u region. There are also character- istic frequencies that can be assigned to other simple groups such as CHa, C:C, C:N, N--H, etc. The po- sition of these peaks is affected some- what by the rest of the molecules, but is not shifted to other regions. In the region 8 to 25 u, the absorption peaks are characteristic of the molecules as a whole. This region is therefore particularly ad- vantageous for distinguishing be- tween closely related compounds such as isomers, or members of homologous series. Most of the work done in the infrared has been in the 2 to 15 u range, using NaC1 prisms. A con- siderable amount has also been done with KBr prisms that cover the region to 28 u. Infrared spectrophotometry is especially suited to analysis of mixtures because of the relatively large number of absorption peaks shown by most compounds it is especially useful for analysis of mixtures of aliphatic hydrocarbons that have little or no absorption in the visible or ultra-violet regions. Many publications are available t½ show application of infrared technique to analysis of fractions of petroleum. In our laboratories, we have used infrared to determine the presence and relative amounts of six components in a mixture of isomeric awlamine compounds. One of the chief difficulties in the application of infrared techniques is the limited number of solvents that do not themselves show strong absorption over wide ranges of the spectrum. Carbon tetrachloride and carbon disulfide are the most widely used, although other solvents can be used over limited regions. It is necessary that the solvents used do not dissolve the salt plates used as cells. When only qualitative determinations are desired, the sol- vent difficulties are not so impor- tant, since solid compounds can be used directly either as films de- posited on salt plates, or as "mulls" with mineral oil or other fairly transparent .liquids. In our labora- tories we have had some promising results from emulsions of materials in carbon disulfide and carbon tetrachloride. TECHNIQUES OF ANALYTICAL SPECTROPHOTOMETRY Most of the qualitative and quan- titative determinations made by spectrophotometric means depend upon comparison of the absorption of the unknown with that of a standard obtained in exactly the same manner as was the unknown. In the visible and ultraviolet regions it is usually advisable to

SPECTROPHOTOMETRY IN ANALYSIS OF COSMETIC PRODUCTS 293 obtain the absorption spectrum of a compound under at least two con- ditions. Compounds soluble in water or in dilute alcohol may best be measured in both alkali and acid. Other compounds may be measured in two or more solvents. The spectra so obtained are usually suflSciently distinctive to serve as a means of identification of the com- pounds. The optimum concentra- tion of solutions usually lies between 10 and 100 rag. per liter of solvent when 1-cm. cells are used. Glass cells are satisfactory for use in the visible region. Fused quartz can be used down to about 210 m/• in the ultraviolet region. In the infrared region, sodium chloride cells are most commonly used. Potassium bromide is sometimes used, as are the plates of silver chlo- ride. The cells used in infrared work must be much shorter liquids are usually measured in cells be- tween 0.15 ram. and 0.001 mm. in length. Gases are measured in cells from 1 cm. to 20 cm. in length. Operations in the determination of infrared spectra are similar to those carried out in obtaining spec- tra at shorter wavelengths. So- lutions examined are usually much more concentrated than those needed for visible or ultraviolet work 100 rag. per ml. of solvent is probably a fairly good average. The use of dilute solutions is rarely possible because the absorp- tion due to the solvent may mask that due to the solution. A recent development in spectro- photometric procedures is the vari- able reference technique (1). In this procedure, the unknown and stand- ard solutions are directly com- pared. The spectrophotometer is actually used as a null indicator, and the composition of the unknown is determined entirely by deter- mining the amount of standard or standards required to balance the absorption due to the unknown. This technique is not readily adapted for use on other than double-beam instruments. A comprehensive treatise on ab- sorption spectrophotometry is avail- able in "Analytical Absorption Spectrophotometry" edited by Mellon (3). Study of Williams' (4) article on "Infrared Instrumentation and Techniques" is also recommended. INSTRUMENTS AVAILABLE The so-called "double-beam" re- cording spectrophotometers have so much to recommend them that the laboratories of the Division of Cosmetics have adopted them exclusively. Their principal advan- tage is in the speed with which data, already accurately plotted, can be obtained. They have in the detector stage a means for measuring differences in intensity of light incident upon the photo- sensitive element. Thus the ab- solute intensity of the illumination can vary over wide limit• without affecting the precision of measure- ment of differences. The two beams of light are switched, or alternatively, the detector output is sampled, at suflSciently short