j. Cosmet. Sci., 51, 183-192 (May/June 2000) Near-infrared spectroscopy: Applications in hair research CHANDRA M. PANDE and BRIAN YANG, Bristol-Myers Squibb World Wide Beauty Care, 2 Blachley Road, Stamford, CT 06922. Accepted for publication March 15, 2000. Synopsis We have evaluated the potential of NIR (near-infrared) spectroscopy as a tool in hair research. We find that this technique is ideally suited for measuring the relative moisture content of hair in situ under practical hair-grooming conditions. It also allows measurement of melanin absorption in the NIR region, where there is no interference from synthetic hair dyes. Thus, rapid evaluation of "lift," the bleaching produced by oxidation dye products, can be performed on hair swatches as -well as on live heads. INTRODUCTION Human hair fibers are roughly 50-80 l•m in diameter and are primarily composed of keratin proteins. Nearly 95 % of the dry hair mass is proteinaceous, and 10% of this derives from the disulfide-containing amino acid, cystine. The remaining is made up of lipids, pigments, and some bound ions (1). Under ambient conditions of 25øC and 50% RH (relative humidity), hair fibers bind as much as 10% water by weight (2). Hair fibers have a complex morphology, with three distinct regions: the outermost cuticle layers, the inner cortical cells, and occasionally an innermost and porous medulla. The cortical cells contain or-helical protein, assembled in a fibrillar arrangement, embedded in an amorphous protein matrix (3). Water acts as a plasticizer for hair and plays a critical role in determining its tactile, mechanical, and other cosmetic properties. The amount of water bound to hair depends on the ambient humidity, with more water bound at higher humidities. The affinity for water mainly arises from the polar amino acid side chains of keratin, with negligible contribution from the peptide bonds (4,5). As the humidity increases, one would expect the binding to become less specific, with water binding to low affinity sites as well as existing as free water. The moisture content of hair at any given RH, as measured by the regain from the dry state, is different and less than that obtained by way of dehydrating hair from 100% RH. Such hysteresis is also seen in other synthetic polymers and biopolymers and has been explained to arise from differences in the ratio of the "bound" to the "free" water (6). Watt (2) provides an excellent review of the water-binding properties of keratin, with ample references. Cosmetic hair treatments, either with heating appliances such as a blow dryer and curling iron or with grooming products such as hair fixatives, modify the water-binding ability of hair and, therefore, its physical 183

184 JOURNAL OF COSMETIC SCIENCE properties such as hold and feel. It is, therefore, very important to be able to measure the effects of hair care products on their ability to effect changes in the kinetic and equi- librium water-binding properties of hair. Human hair fibers derive their natural color from melanin pigment. Unlike synthetic hair colors that are mixtures of small molecules dispersed throughout the fiber, includ- ing the surface, the natural pigment is polymeric and exists as discreet granules only in the hair cortex. This structural difference between these two types of coloring matter has a profound influence on their optical properties. While the synthetic hair dyes do not absorb above 750 nm (red light), the natural pigment shows semiconductor-like optical properties and absorbs (scatters) light from the UV region all the way into the near- infrared, at least up to 1400 nm. NIR energy is relatively weak in causing electronic transitions to excited states in all but the highly delocalized electronic systems. The hair pigment melanin is unusual in this regard, and it has been suggested that its absorption does not fit the classical definition of chromophoric absorption and is more semiconductor-like (7). Typically, NIR absorp- tion by molecules results from overtones and combinations of the characteristic bond vibrations. For biological tissue, this region is dominated by bands due to O-H, mainly from water, and N-H from the protein backbone. NIR spectroscopy has been used extensively in skin research (8-11). Use of NIR second-derivative methodology to monitor the water content of hair has also been reported (12). Here we report the results of our study aimed at evaluating the potential of NIR spectroscopy as a tool in hair research. We find that this technique allows measurement of the water content of hair under hair-grooming conditions. Although these measure- ments provide relative water concentrations, they can be converted to absolute values by appropriate calibration. We have also exploited the difference in the light absorption characteristics of the natural hair pigment and the synthetic dyes to measure "lift," or the bleaching of the natural pigment due to the oxidative hair coloring process in the presence of the deposited dye. This latter aspect is particularly important when testing competitive products for which information on the ingredient concentrations is often lacking. EXPERIMENTAL Spectra were collected with a Magna 760 FT-IR system with NIR capability (Nicolet Instrument Co.), using the SAB-IR accessory. The latter is a bifurcated fiber-optic probe consisting of a ca. 3-mm fiber-optic bundle in which the fibers that bring in the incident light and those that take the reflected light to the detector are randomly arranged. The assembly tip is encased in a screw-in cover with a Sapphire angled window to cut down the specularly reflected light. The system uses a PbS detector and runs under the OMNIC © operating system. Background correction was performed using Spectralon ©. Typically 32 scans (38 sec) were collected for each sample at a resolution of 8 cm-•. The data were collected in the reflectance mode. Absorption curves were generated as log (l/R). Baseline correction was performed on the spectra using the packaged routine. In reflectance measurements, both absorption and diffuse scattering of the incident light contribute to the observed signal. The depth of penetration by light depends on the wavelength and the nature of the sample. In biological materials, for example, the longer wavelength NIR radiation will penetrate more deeply than the ultraviolet light. Mea-