Copper and calcium uptake in colored hair

Transcription

1 J. Cosmet. Sci., 60, (May/June 2009) Copper and calcium uptake in colored hair K. E. SMART, M. KILBURN, M. SCHROEDER, B. G. H. MARTIN, C. HAWES, J. M. MARSH, and C. R. M. GROVENOR, Department of Materials, University of Oxford, Parks Road, Oxford, OX1 3PH, UK (K.E.S, M.S, C.R.M.G.), School of Life Sciences, Oxford Brookes University, Gipsy Lane, Oxford, OX3 0BP, UK (B.G.H.M., C.H.), Procter & Gamble, Miami Valley Innovation Center, E. Miami River Road, Cincinnati, OH (J.M.M.), and Centre for Microscopy, Characterisation & Microanalysis, University of Western Australia, 35 Stirling Highway, Crawley, 6009 WA, Australia (M.K.). Accepted for publication December 29, Synopsis During hair coloring a number of disulfi de bonds in cystine are oxidized (1) to create cysteic acid, forming binding sites for metal ions such as Ca 2+ and Cu 2+ from tap water (2). The increased uptake of these metals can have a detrimental impact on fi ber properties for example, reducing shine and causing a poor wet and dry feel (3). In addition, the increased uptake of copper can also contribute to further fi ber damage during subsequent coloring due to its ability to take part in metal-induced radical chemistry (4). It is important to know where in the fi bers these metals are located in order to either effectively remove these metals or control their chemistry. Nanoscale secondary ion mass spectrometry (NanoSIMS) has been used to locate the calcium and copper within hair that has been treated with a colorant and washed multiple times in tap water containing these ions. Untreated hair is used as a baseline standard material. Images with up to 50-nm spatial resolution of the preferential locations of calcium uptake were obtained, showing a high concentration of calcium in the cuticle region of colored hair, specifi cally in the sulfur-rich regions (A-layer and exocuticle). INTRODUCTION Typical permanent hair colorants contain hydrogen peroxide buffered to a ph of 10 with ammonium hydroxide. The role of the oxidant is to bleach the melanin, lightening the underlying substrate color, and to oxidize the dye precursors to form chromophores. The final color the consumer achieves is a combination of lightening of the natural color and deposition of synthetic color inside the hair. However, the oxidant can also react with the hair proteins and lipids, leading to changes in the fiber properties that can be experienced by consumers, especially over multiple cycles. These properties include a reduction in shine, reduced manageability, poor wet and dry feel, and increased likelihood of split ends (3). One of the proposed mechanisms of fiber damage is reaction of the hair proteins and lipids with hydroxyl radical species that are formed from the catalytic reaction between hydrogen Address all correspondence to J. M. Marsh. 337

2 338 JOURNAL OF COSMETIC SCIENCE peroxide and redox metals such as copper (4). The source of the copper is the tap water used during the shampoo and rinsing processes (5). Water hardness ions such as calcium and magnesium are also taken up by hair from the tap water. Although the calcium does not directly take part in the color chemistry, it can affect the shine and combing properties of the hair by forming insoluble calcium salts and soaps on the fiber surface (6). For both metals it has been demonstrated that their uptake increases as the hair undergoes multiple treatments from an oxidative colorant. It has been hypothesized that this is due to the enhanced number of metal binding sites that are formed in the hair during the oxidation process. The presence of metals in hair has been widely reported in the literature, with earlier work concentrating on measuring the range and levels of metals present in the fi ber (7). This work also demonstrated that there are two main sources of these metals, endogenous and exogenous. The endogenous metals are incorporated in the fi ber at the hair follicle and are thought to act as a pathway for the body to excrete unwanted metals (8). The exogenous metals come from the fi ber s exposure to the environment and in particular to the water used in its washing (9). Recent research has focused on identifying the specifi c location of metals in the fi ber structure and on identifying which metals are exogenous and which metals are endogenous. Of particular interest has been the presence of metals whose origin may have been due to exposure to pollution. Kempson and Skinner (10) investigated the hair from a worker exposed to lead, and Audinot et al. (11) investigated hair from an individual exposed to arsenic. In both cases the nature of the metal as either exogenous or endogenous was identifi ed. ToF SIMS and x-ray microfl uorescence imaging have been used to map calcium in hair and to identify the source as either exogenous or endogenous (12,13). The technique used here to investigate the location of metals is nanoscale secondary ion mass spectrometry (NanoSIMS). SIMS is a surface analysis technique in which a focused primary ion beam is scanned across the target surface. The bombardment of these primary ions results in the ejection of charged atomic and molecular species from the sample surface. The secondary ions are then separated on the basis of their mass-to-charge ratio and correlated to their spatial origin to form a chemical image (14,15). The NanoSIMS is the latest generation of dynamic SIMS facilities and is capable of achieving, at the same time, extremely high spatial resolution (< 50 nm) and mass resolving power (m/ m = 5,000) due to its coaxial lens confi guration and advanced ion optics. Limited previous work has been done using this kind of analysis on hair samples. The work of Hallegot and Corcuff (16) was the fi rst example of a SIMS study of a hair fi ber, while Collin et al. (17) studied the presence of isotopically labeled taurine in hair. In this study we have concentrated on studying the location of both calcium and copper in hair samples, both untreated and after treatment with an oxidative colorant. EXPERIMENTAL SAMPLE PREPARATION Three samples were used for analysis. The fi rst was untreated Caucasian hair, the second was colored Caucasian hair, and the third was colored Caucasian hair containing high levels of copper. The Caucasian hair containing high levels of copper was used as a control to check the sensitivity of the instrument to detect copper in the hair samples.

3 Cu AND Ca UPTAKE IN COLORED HAIR 339 Caucasian untreated mixed hair (natural white), obtained from a commercial source (IHIP, New York), was formed into swatches (16 cm, 1.5 g). The treated hair was obtained by treating the untreated hair with a commercial extra-light blonde shade that contained 4.5% hydrogen peroxide and ammonium hydroxide at a ph of 10. Four grams of the product was applied to each gram of hair. The product-covered hair was placed in an oven for 30 minutes at 30 C and then rinsed for 1 min in tap water containing about 300 ppm calcium and magnesium, and approximately 0.06 ppm copper. The hair was then washed twelve times using 0.1 g of a commercial shampoo. This whole process, involving one color and twelve washes, was repeated five times to create the treated Caucasian white hair sample. The colortreated hair was placed under running tap water dosed with copper chloride (1 ppm) for 30 minutes to create the copper-rich sample. The hair was then dried in front of a fan. To create a fl at surface for NanoSIMS analysis, the hair was encapsulated within a resin matrix. The hair was cut into 2-cm lengths, placed in a shallow aluminium dish containing medium-grade L.R. White resin, and fully coated in resin. Small bundles of resincoated hair were positioned vertically in 1.5-ml Eppendorf tubes and polymerized in the oven at 70 C for 9 hr. The hair-containing resin blocks were then cut into sections of 0.5-µm thickness using a RMC powertomexl ultramicrotome. The thin sections were placed on a droplet of water on a silicon substrate and warmed gently to give good adherence. All samples were gold coated to maintain conductivity in the NanoSIMS. HIGH-RESOLUTION SECONDARY ION MASS SPECTROMETRY The CAMECA NanoSIMS 50 can be operated with a Cs + primary beam to detect negative secondary ions, or with an O primary beam to detect positive secondary ions. Highresolution ion images were obtained using the Cs + primary ion beam, focused to approximately 50 nm. The ion species 12 C, 32 S, 12 C 14 N, and 40 Ca 16 O were simultaneously mapped over an area of 5 5 µm, with pixels in each image. The O primary beam, with a resolution of about 300 nm, was used to map 40 Ca + and 63 Cu + over an area of µm, with pixels in each image. Each area was cleaned and implanted with the primary beam prior to image acquisition. To improve the counting statistics, fi ve sequential images were added together. INDUCTIVELY COUPLED MASS SPECTROMETRY Samples (0.5 g) were taken from each of the three types of hair, washed, and then analyzed using inductively coupled plasma-mass spectroscopy (ICP-MS). RESULTS AND DISCUSSION CALCIUM MAPPING: O BEAM RESULTS In the fi rst set of experiments we investigated the use of the O beam for chemical mapping of both the uncolored and colored hair samples. The 40 Ca + and 63 Cu + images are shown respectively in Figure 1(a,b) for untreated hair and in Figure 1(d,e) for treated hair. The untreated hair shows an enhanced concentration of calcium in the cuticle region (~1000 counts per second) and also at the center of the fi ber, in the medulla region (2000

4 340 JOURNAL OF COSMETIC SCIENCE Figure 1. (a) NanoSIMS 40 Ca + map obtained using the O primary ion beam from untreated Caucasian hair. (b) NanoSIMS 63 Cu + map obtained using the O primary ion beam from untreated Caucasian hair. Bar = 10 µm. (c) NanoSIMS line profi les of the 40 Ca + and 63 Cu + signals obtained using the O primary ion beam on untreated Caucasian hair. The line scan direction is indicated with an arrow in the 40 Ca + image in (a). The 40 Ca + ion signal has been offset by 500 counts per second for clarity. (d) NanoSIMS 40 Ca + map obtained using the O primary ion beam from treated Caucasian hair. (e) NanoSIMS 63 Cu + map obtained using the O primary ion beam from treated Caucasian hair. Bar = 10 µm. The 40 Ca + ion signal has been offset by 500 counts per second for clarity. (f) NanoSIMS line profi les of the 40 Ca + signal obtained using the O primary ion beam on treated Caucasian hair. The line scan direction is indicated with an arrow in the 40 Ca + image in (d). to 3000 counts per second). This image is similar to that observed by Merigoux et al. (13) using x-ray fl uorescence. In comparison, the colorant-treated hair showed overall a much higher level of calcium all across the hair surface and a signifi cantly higher concentration of the calcium in the cuticle area (~7000 counts per second). The relative level of the calcium observed in these two images correlates well with the bulk calcium analysis data using ICP-MS (see Table I). The very clear increase in the concentration of calcium in the cuticle area indicates preferential binding sites located in this region of the cuticle vs a diffusion profi le, but unfortunately the relatively poor resolution achieved in this image does not allow us to come to any more detailed conclusions as to the precise binding sites within the cuticle. This poor resolution is a consequence of the low analyte signal and the resulting compromise between detectable signal resolution. CALCIUM MAPPING: Cs + BEAM RESULTS Due to the relatively large diameter of the O beam, its spatial resolution is intrinsically limited to ~200 nm. In order to improve the resolution for the calcium distribution, the Cs + ion beam was used. This experimental setup is capable of resolutions down to 50 nm for negative secondary ions, and 12 C, 32 S, 12 C 14 N, and 40 Ca 16 O ions were mapped. The images are shown in Figure 2. In all these images there are two distinct regions in each cuticle cell, thought to be due to the high sulfur regions (exocuticle and A-layer) and the

5 Cu AND Ca UPTAKE IN COLORED HAIR 341 Table I Bulk Analysis Data from ICP-MS for Natural White, Colored White, and Copper-Rich White Hair Untreated (mg/kg) Treated (mg/kg) Copper-rich (mg/kg) Mg 160 ± ± ± 23 Ca 1300 ± ± ± 720 Fe 19 ± 3 No data 7 ± 1 Cu 12 ± ± ± 1650 low-sulfur regions (endocuticle). The 32 S map (Figure 2b) supports this, as it shows an enhanced concentration of sulfur in the outer region of the cuticle, i.e., the A-layer and exocuticle. The 32 S signal in the endocuticle is about 50% of that in the A-layer and exocuticle. The distribution of sulfur in the cuticle regions of the 32 S map obtained here is in good agreement with the EELS study of sulfur in the cuticle (16). The high protein content of the exocuticle is also reflected in the 12 C 14 N map [a species known to be generated at high yield from proteins (18)]. The 12 C map shows the opposite variation, with a higher concentration in the endocuticle than the exocuticle. The composition of the endocuticle is very different from that of the exocuticle and consists of relatively large amounts of dibasic and diacidic amino acids (19). These regions have a high 12 C signal due to the presence of molecules with multiple C C bonds (see the 12 C map in Figure 1a). Although the resolution is too poor to assign the location of metal binding in the cortex, cortical cells may be distinguished underneath the lamella of the cuticle of the hair (see Figure 2a). The cortical cells are high in 12 C 14 N and 32 S (see Figure 2a,b) compared to the surrounding intercellular material, which is high in 12 C (see Figure 2a). The 40 Ca 16 O map also shows significantly different yields in the different regions of the cuticle cells, with a distinctly higher number of 40 Ca 16 O counts from the outer region (A-layer and exocuticle. The co-localization of 40 Ca 16 O and 32 S is illustrated in the color overlay in Figure 2f. In this image the signal from the 32 S ions is color-coded in green and the 40 Ca 16 O signal is color-coded in red. Where the 40 Ca 16 O and 32 S signals are overlapping, the signal is yellow. The overlay confirms that 40 Ca 16 O is strongly associated with 32 S in the outer cuticle region, as no significant areas of red are observed. If the same procedure is carried out using the 12 C signal as the green color and the 40 Ca 16 O signal as the red color, then the image in Figure 2e is obtained. It is clearly seen that the calcium does not coincide with the location of the 12 C signal, as clear areas of red are observed. A line scan across the cuticle (see Figure 2c for location of the line scan) was obtained for each of the four ion signals. Figure 3 shows the four line scans, and further emphasizes the co-localization of the 40 Ca 16 O and 32 S signals and the inverse behavior of the 12 C signal. The higher yield of calcium in the cuticle sulfur-rich areas for colored hair may imply binding of the metals to the sulfonate functional group of cysteic acid formed during oxidation. However, it is more likely that as a result of oxidation a charge rearrangement has taken place between the sulfonate and carboxylate groups and that it is the carboxylic acid groups that are involved in the binding (20,21). COPPER MAPPING: O BEAM RESULTS The chemical maps in Figure 1(b,e) show the maps obtained for the 63 Cu + ion. The 63 Cu + signal intensity was extremely low, and what signal there is appears to be concentrated in

6 342 JOURNAL OF COSMETIC SCIENCE Figure 2. High-resolution NanoSIMS maps from the treated Caucasian hair obtained using the Cs+ primary ion beam for 12C2, 32S, 12C14N, and 40Ca16O ions and color overlays to illustrate more clearly which signals are co-localized. The color yellow indicates the overlap of red and green, where the 32S and 40Ca16O signals are co-localized. Bar = 1 µm.

7 Cu AND Ca UPTAKE IN COLORED HAIR 343 Figure 3. NanoSIMS line profi les obtained using the Cs + primary beam on treated hair for 12 C 2, 32 S, 12 C 14 N, and 40 Ca 16 O signals. The line scan direction is indicated with an arrow in the 12 C 14 N image in Figure 2a. Abbr: Ex, exocuticle; En, endocuticle. the medulla and in isolated spots in the cortex in both untreated and treated samples. We had some concerns about whether these observations were representative of the extremely low copper concentration in these samples or artefacts of sample preparation. In an attempt to confi rm our ability to reliably map the location of copper ions in these samples, a hair swatch that contained a signifi cantly higher level of copper than would be expected in a consumer situation was prepared. The colorant-treated hair was washed in water containing 1 ppm copper ions for 30 min until the hair was visibly green. ICP-MS analysis confi rmed that the copper level had increased from 120 ppm to 11,000 ppm and that the calcium level had dropped from 8,700 ppm to 4,800 ppm (see Table I). The O ion beam was used for analysis of positive ions, and the detector was tuned to 40 Ca + and 63 Cu +. The images obtained are shown in Figure 4. The 40 Ca + image is similar to what we have seen previously, with the signal concentrated both in the medulla and the cuticle. However, the 63 Cu + signal intensity is still signifi cantly lower than that of 40 Ca +. One possible explanation for this result is that the yield of 63 Cu + ions from the hair is very much lower than that of 40 Ca + ions. The relative sensitivity factor, or RSF, for 40 Ca + in a silicon Figure 4. (a) NanoSIMS 40 Ca + map obtained using the O primary ion beam from copper-treated Caucasian hair. (b) NanoSIMS 63 Cu + maps obtained using the O primary ion beam from copper-treated Caucasian hair. Bar = 10 µm. (c) NanoSIMS line profiles of the 40 Ca + and 63 Cu + signal obtained using the O primary ion beam on copper-treated Caucasian hair. The line scan direction is indicated with an arrow in the 40 Ca + image in (a).

8 344 JOURNAL OF COSMETIC SCIENCE matrix is to atoms cm 3, whereas the RSF for 63 Cu + in a silicon matrix is an order of magnitude lower, to atoms cm 3 (22). In this high copper image, the 63 Cu + signal is observed in the medulla and in isolated spots in the cortex. It is also detected in the cuticle and in a distinct diffusion gradient into the cortex. The diffusion profi le observed in the cortex is likely due to the technique used to dose the fi ber. Thus at this stage we can only offer a tentative identifi cation of copper concentrating in the cuticle, in the same way as clearly shown for calcium. CONCLUSIONS The NanoSIMS has proved to be an excellent tool for mapping, with a resolution of up to 50 nm, the location of calcium in both untreated and colored hair. We have demonstrated that the additional uptake of the calcium in colored hair is to a signifi cant extent concentrated in the cuticle. The technique has also clearly demonstrated the co-location of the calcium with sulfur-rich regions of the cuticle, specifi cally the A-layer and exocuticle. REFERENCES (1) C. Robbins, Infra red analysis of oxidised keratins, Textile Res. J., 37, 811 (1967). (2) R. E. Noble, Uptake of calcium and magnesium by human scalp hair from waters of different geographical environments, Sci. Total Environ., 239, 1-3, (1999). (3) J. Jachowicz, Hair damage and attempts to its repair, J. Soc. Cosmet. Chem., 38, (1987). (4) J. M. Marsh, J. Flood, D. Domaschko, and N. Ramji, Hair coloring systems delivering color with reduced fi ber damage, J. Soc. Cosmet. Chem., 58, (2007). (5) G. R. Bhat, E. Lukenbach, and R. R. Kennedy, The green hair problem: A preliminary investigation, J. Soc. Cosmet. Sci., 30, 1 8 (1979). (6) K. V. Curry and S. Golding, Hair lipids. I. Extraction of fatty materials from hair clippings, J. Soc. Cosmet. Chem., 22, (1971). (7) G. Chittleborough, A chemist s view of the analysis of human hair for trace metals, Sci. Total Environ., 14(1), (1980). (8) M. Villian, V. Cirimele, and P. Kintz, Hair analysis in toxicology, Clin. Chem. Lab. Med., 42(11), (2004). (9) J. Bacso, L. Sarkadi, and E. Koltay, On endogenous and exogenous calcium content of hair samples used in XRF and PIXE measurements, Int. J. Appl. Rad. Isotopes., 33, 5 11 (1982). (10) I. M. Kempson and W. M. Skinner, Advanced analysis of metal distributions in human hair, Environ. Sci. Technol., 40, (2006). (11) J. N. Audinot, S. Schneider, M. Yegles, P. Hallegot, R. Weneg, and H. N. Migeon, Imaging of arsenic traces in human hair by nano-sims 50, Appl. Surf. Sci., , (2004). (12) I. M. Kempson, W. M. Skinner, and P. K. Kirkbride, Calcium distributions in human hair by ToF- SIMS, Biochim. Biophys. Acta, 1, 1624 (2003). (13) C. Merigoux, F. Briki, F. Sarrot-Reynauld, M. Salome, B. Fayard, J. Susini, and J. Doucet, Evidence for various calcium sites in human hair shaft revealed by sub-micrometer X-ray fl uorescence, Biochim. Biophys. Acta, 1619, (2003). (14) A. Benninghoven, F. G. Rüdenauer, and H. W. Werner, Secondary Ion Mass Spectroscopy: Basic Concepts, Instrumental Aspects, and Trends (John Wiley & Sons, New York, 1987). (15) J. C. Vickerman, A. Brown, and N. M. Reed, Secondary Ion Mass Spectrometry: Principles and Applications (Clarendon Press, Oxford, 1989). (16) P. Hallegot and P. Corcuff, High spatial resolution maps of sulfur from human hair sections; an EELS study, J. Microsc., 172, 2, (1993). (17) C. Collin, B. Gautier, O. Gaillard, P. Hallegot, S. Chabane, P. Bastien, M. Peyron, M. Bouleau, S. Thibaut, F. Pruche, A. Duranton, and B. A. Bernard, Protective effects of taurine on human hair follicle grown in vitro, Int. J. Cosmet. Sci., 28, (2006).

9 Cu AND Ca UPTAKE IN COLORED HAIR 345 (18) P. J. Heard, K. A. Feeney, G. C. Allen, and P. R. Shewry, Determination of the elemental composition of mature wheat grain using a modifi ed secondary ion mass spectrometer (SIMS), Plant J., 30, (2002). (19) J. A. Swift and B. Bews, Chemistry of human hair cuticle. II. Isolation and amino acid analysis of the cell membrance complex and A-layer, J. Soc. Cosmet. Chem., 5, (1974). (20) E. O. P. Thompson and I. J. T. O Donnell, Studies on oxidized wool, Aust. J. Biol. Sci., 12, (1959). (21) L. J. Wolfram, K. Hall, and I. Hui, The mechanism of hair bleaching, J. Soc. Cosmet. Chem., 21, (1970). (22) R. G. Wilson, SIMS quantifi cation in Si, GaAs, and diamond An update, Int. J. Mass Spectrom., 143, (1995).