Synopsis

OBJECTIVE:

Formaldehyde is an effective and popular semipermanent hair straightener, but the severe consequences for human health due to its toxicity have prompted the search for safer alternatives. Different carbonyl compounds, including glyoxylic acid, have recently been proposed as promising candidates. Despite the interest in this topic, there is a lack of information about the interactions between hair keratin and straightener agents. This study addresses this issue to gain new insights useful in the development of new products for safe, semipermanent hair deformation.

METHODS:

The possible reactions occurring between carbonyl groups and nucleophilic sites on amino acid residues belonging to the keratin were investigated using as model compounds some aldehydes and amino acid derivatives. Raman and IR analyses on yak hair subjected to the straightening treatment with glyoxylic acid in different conditions were carried out. Scanning electron microscope (SEM) analyses were carried out on yak and curly human hair after each step of the straightening procedure.

RESULTS

The reactions between aldehydes and N-a-acetyl-Llysine revealed the importance of the carbonyl electrophilicity and temperature to form imines. Raman and IR analyses on yak hair subjected to the straightening treatment evidenced rearrangements in the secondary structure distribution, conformational changes to the disulphide bridges, a decrease of the serine residues and formation of imines. It was also indicated that straightening produced major conformational rearrangements within the hair fibre rather than on the cuticle.

CONCLUSION:

This investigation revealed the role played by the electrophilicity of the carbonyl on the straightener agent and of the temperature, closely related to the dehydration process. Raman and IR studies indicated the involvement of imine bonds and the occurrence of a sequence of conformational modifications during the straightening procedure. SEM analyses showed the effectiveness of the treatment at the cuticular level.

Introduction:

Since the first decade of the last century, the permanent deformation of human hair has been achieved with the use of chemical agents. In fact, due to the elastic properties of hair [1], when both waving and straightening are obtained by thermal treatment, only a temporary deformation that lasts only until the next wash is obtained [2], as the set is given by hydrogen bonding which is subsequently broken, and the hair reverts to it unwaved/unstraightened state. The principal technique, mostly used for permanent hair waving, is based on the breaking of cystine disulphide bonds by treatment with an aqueous basic solution containing a reducing agent followed by the rinsing of hair, its rolling around a curling and finally its treatment with an oxidizing agent able to restore the disulphide bonds [3,4]. The second technique is a modification of the very old process using hot alkaline solutions. Now, it is carried out in milder conditions, usually 10–15 min of treatment at room temperature with water/oil emulsions of a hydroxide [4]. One mechanistic explanation of the permanent modification obtained with alkali invokes the socalled lanthionization reaction, which causes the cleavage of the cystine disulphide bond (–CH2–S–S–CH2–) and its replacement with a lanthionine bond (–CH2–S–CH2–), a stable thioether bond [1–5]. Until a few years ago, the permanent curling has been one of the most requested treatments by consumers in hair salons but, today, the hair straightening is increasingly in demand. In this latter case, the formulations containing reducing agents or alkali (also called chemical relaxers) often cause hair weakening and breaking, a damage that increases after the straightening with hot combs or irons [6]. These effects can be reduced on going from a permanent to a semipermanent deformation, a milder treatment that is effective for a few months. In this context, one of the most popular and efficient methods, used for a long time, to obtain a prolonged smoothing of frizzy hair was the so-called Brazilian hair straightening that used formulations with high contents of formaldehyde as active ingredient [7] (even more than 9% w/w has been found in home-made creams in Brazil) [8]. Formaldehyde is often used as chemical disinfectant and preservative, but its use and content have been subjected to severe restrictions because of the hazard [9,10] of this compound for the human health (causing cancer, leukaemia, eye irritations, respiratory problems, contact allergies). Therefore, its presence in hair straightener formulations is very dangerous for both salon workers and customers, due to the direct contact with it during the application and, even worse, to its vaporization and release in the environment occurring in the subsequent step, when hair is straightened through hot flattening irons (this treatment will be hereafter called ironing). Consequently, the interest in alternative safer and environmentally friendly hair straightening products has recently grown, and some patents and patent applications on formaldehyde-free formulations have been registered [4,11–17]. Some of them claim a-hydroxyacids or a-ketoacids as active hair straightening components. In particular, glyoxylic acid [13–17] has received great attention, and it is already in use in several commercial formulations for the semipermanent hair straightening, a process which is intended satisfactory when hair retains the smooth shape after not less than six consecutive washing cycles with water and shampoo [15–17]. Despite the interest in this topic, a survey in the literature has revealed lack of information about the chemistry involved in the interaction between hair fibres and carbonyl-based straightening compounds. This fact, together with the belief that a deeper knowledge of the mechanism and interactions involved in this hair treatment might be useful to find other compounds suitable for a safe semipermanent hair deformation, has suggested us to start a study on this topic using a multidisciplinary approach, and herein, we report the obtained results.

Materials and methods

Formaldehyde (36.5% in water), glyoxylic acid and the other chemicals used were purchased from Sigma-Aldrich (Milan, Italy). White yak hair was purchased from Socap (Napoli, Italy). Hair ironing was carried out using a BHS AF6000 titanium plate. The 1 H and 13C NMR spectra were recorded at 300, 400 or 600 MHz and 75.36, 100.57 or 150.82 MHz, respectively. 1 H Chemical shifts were measured in d (ppm) and referenced to the solvent (4.6 ppm for 1 H NMR in D2O and 2.50 and 39.50 ppm for 1 H NMR and 13C NMR, respectively, in DMSO-d6). J values are given in Hz. Mass spectra (ESI-MS) were recorded on a WATERS ZQ 4000 instrument (Milford, MA USA). pH Values are determined using an AMEL Instruments Mod. 2335 pH-METER (Milano, Italy) with combined electrode Hamilton 3 M KCl. The reactions carried out in water at temperature >100°C were performed in a sealed tube.

FT-Raman spectroscopy

Raman spectra were recorded on a Bruker MultiRam FT-Raman spectrometer equipped with a cooled Ge-diode detector. The excitation source was a Nd3+-YAG laser (1064 nm) in the backscattering (180°) configuration. The focused laser beam diameter was about 100 lm, the spectral resolution 4 cm
1 , the laser power at the sample about 60 mW. Due to their intrinsic orientation, the spectra were recorded by positioning the fibres along one specific direction. Spectra were recorded in duplicate at least. The Amide I spectral range was analysed by a curve-fitting procedure to evaluate the content of secondary structures. A linear correction in the 1740–1534 cm
1 spectral range brought the baseline of the Raman spectra to approximately zero intensity. The fourth-derivative spectra were calculated by applying the fourth-derivative function with 9-point smoothing according to the Savitsky–Golay method. The frequencies of the band centres found in the fourthderivative spectra were used as starting parameters for the curvefitting procedure. The curve-fitting analysis was performed using the OPUS version 6.5 program (Bruker Optik GmbH, Ettlingen, Germany), which uses the Levenberg–Marquardt algorithm. The Raman component profiles were described as a linear combination of Lorentzian and Gaussian functions. The content of a-helix, bsheet, b-turns and unordered conformations was calculated from the area of the individually assigned bands (at about 1655, 1670 and 1700, 1685, 1640 cm
1 , respectively [18–21]) and expressed as fraction of the total area. As reported by Kuzuhara [22,23], white yak fibres are preferable lacking melanines, which are highly fluorescent and are responsible for a worsening of the Raman signal.

FT-IR spectroscopy

IR spectra were recorded on a Nicolet 5700 FTIR spectrometer, equipped with a Smart Orbit diamond ATR accessory and a DTGS detector; the spectral resolution was 4 cm
1 . Due to their intrinsic orientation, the spectra were recorded by positioning the fibres along one specific direction. Spectra were recorded in triplicate.

Scanning electron microscope (SEM) evaluation

The scanning electron microscopic evaluation of the surface morphology of samples of yak and human samples was performed longitudinally with a Zeiss Evo 50-EP (Carl-Zeiss, Oberkochen, Germany). To minimize artefacts, sputtering was avoided and the samples were observed in variable pressure (VP) mode. All measures were carried out at an accelerating voltage of 20 kV and 100 Pa in the chamber pressure. The signal revealed secondary electrons.

Reactions with model compounds

Synthesis of (R)-2-acetamido-3-((hydroxymethyl)thio)propanoic acid (4): to a solution of N-acetyl-L-cysteine (1, 0.045 g, 0.276 mmol) in 1.0 mL of water, an equimolar amount of a 36.5% aqueous solution of formaldehyde was added. After about 30 min. at room temperature, the solvent was removed under vacuum and the NMR spectral data of the residue agreed with those reported in the literature for 4 [24,25]. Synthesis of (R)-2-acetamido-3-{[(R)-carboxy(hydroxy)methyl] sulfanyl}propanoic acid and (R)-2-acetamido-3-{[(S)-carboxy (hydroxy)methyl]sulfanyl}propanoic acid (5a and 5b): N-acetyl-Lcysteine (1, 0.045 g, 0.276 mmol) and glyoxylic acid (3, 0.020 g, 0.276 mmol) were dissolved in 1.0 mL of water (pH approximately 2), and after 30 min, the solvent was removed under vacuum at a temperature of about 35°C. NMR and mass spectral data of the residue revealed the presence of a mixture of the two diastereomers 5a and b, whose spectral data are reported in Appendix S1. Reactions between N-a-acetyl-L-lysine and aromatic aldehydes: Na-acetyl-L-lysine (6, 0.010 g, 5.31 9 10
5 mol), an equimolar amount of aldehyde 7, and DMSO-d6 (0.75 mL) were introduced in a NMR spectroscopy tube. Compound 6 was very slightly soluble in the reaction mixture. The progress of reaction was monitored through 1 H NMR spectroscopy after 22 and 46 h for reactions carried out at 25°C and after 4 and 22 h for those carried out at 50°C. The NMR spectral data of compounds 8a–c are reported in Appendix S1. Reactions between N-a-acetyl-L-lysine and glyoxylic acid: Na-acetyl-L-lysine (6) (0.010 g, 5.31 9 10
5 mol) and 3 (3.94 mg, 5,31 9 10
5 mol) were dissolved in water (1 mL). After 2 h at 65°C, the reaction mixture was heated in an oil bath kept at 110°C until complete removal of the solvent (about 30 min). The residue was dissolved in DMSO-d6, and the analysis of the NMR spectra (see Appendix S1) showed the presence of the two diastereomers 9a and 9b. The same results have been obtained carrying out the reaction between 3 and 6 in aqueous basic solution (pH approximately 8.7).

Preparation of the samples for Raman, ATR/FT-IR and SEM analyses

Each sample of yak hair used for Raman, ATR/FT-IR and SEM analyses was a lock of hair of about 20 cm length and 0.5 cm diameter. The basic pre-treatment was carried out bathing the lock for 5 min in an aqueous KOH solution at pH approximately 8.7, by rinsing with distilled water and drying with a hairdryer. The treatment with the straightening solution was carried out bathing the lock previously treated with the basic solution (unless otherwise specified) in a 6% aqueous solution of glyoxylic acid for 30 min then drying with a hairdryer without rinsing. The ironing process of the samples was performed by passing for 6 times the locks of yak hair (5 times in the case of human curly hair), throughout their length, under a straightener plate preheated at 230°C.

Results and discussion

This study began with an investigation on the reactions that can occur between aldehydes and some functional groups belonging to the amino acids present in hair. Then, the modifications induced in hair after each step of the usual straightening process have been studied through ATR/FT-IR and Raman spectrophotometry on solid yak fibres, chosen as a suitable model for human hair [26]. Finally, SEM analyses, carried out both on yak fibres and human hair before and after straightening, provided information on the changes caused by the treatment on the cuticular surface of the hair. For the sake of simplicity, the results will be presented and discussed in separate subheadings.

Study on model compounds: reactions between aldehydes and amino acid derivatives

To gain mechanistic information on the possible reactions that can occur during the semipermanent hair straightening process, we have started from the consideration that the most patented formulations for this treatment contain carbonyl compounds as active ingredients [4,11–17], in line with the well-known ability of formaldehyde to act as efficient straightener agent. Based on this, we focused our attention on the possible reversible reactions that can occur between hair and aldehydes. One of the most probable is the nucleophilic attack on the carbonyl carbon atom by nucleophiles situated on the end-chain of some amino acids of keratin. Among them, those bearing sulphur, oxygen and nitrogen nucleophilic centres are good candidates able to produce hemithioacetals, hemiacetals and imines, respectively, upon reaction with aldehydes. Below, we report the results of our investigation about the interactions between aldehydes and N-acetyl-L-cysteine and N-a-acetyl-L-lysine chosen as model for sulphur nucleophiles and nitrogen nucleophiles.

Interactions between cysteine derivatives and aldehydes

Among the above-mentioned amino acids, cysteine possesses the most reactive nucleophilic centre, but its presence in virgin hair is a rare event. The breaking of the cystine disulphide bond usually occurs after treatment with reducing agents or strongly basic solutions. In this latter case, after approximately 10–15 min [4,11] at room temperature, the lanthionization can take place with formation of a covalent thioether bond [27]; such an event seems improbable upon treatment with carbonyl compounds due to the semipermanent character of the straightening conferred by this technique. In addition, the breaking of the cystine disulphide bond causes a weakening of the hair fibre, and also, this contrasts with its healthy appearance and shiny aspect after the straightening; this might be another indication of the non-occurrence of the S–S bond rupture upon treatment with aldehydes. However, it has to be considered that sometimes the usual procedure for the hair straightening with formulations containing carbonyl compounds involves a quick hair pre-treatment (for about 5 min) with basic solutions or shampoos [13] and always a final step over a hot iron. In addition, it has been reported that thermal stresses above 200°C on wool and silk fibres can produce the breaking of the disulphide bonds [28,29]. These considerations have prompted us to check whether the conditions used for straightening might cause the breaking of the disulphide bond and the subsequent reaction of the thiol functional group with the carbonyl. Thus, in order to prepare reference compounds, we carried out the reactions between N-acetyl-L-cysteine (1, Scheme 1) and formaldehyde (2, Scheme 1) or glyoxylic acid (3, Scheme 1), currently a popular substitute of the former. We used N-protected cysteine in order to avoid intramolecular aminoalkylation with formation of thiazolidine, a reaction already reported to occur between cysteine and formaldehyde [30]. The reaction between 1 and an equimolar amount of 2 gave (R)- 2-acetamido-3-(hydroxymethyl)thio)propanoic acid (4, Scheme 1), whereas with 3 a mixture of the two diastereomers 5a and 5b (Scheme 1) is produced. The 1 H NMR spectra of these products revealed that the region in which the signals belonging to the SCH2O protons of 4 (d = 4.59 ppm) and the SCH(OH)COOH proton of 5 (d approximately 5.2 ppm) fall may be diagnostic for the formation of such adducts. Then, we treated commercially available N,N-dibenzoyl-L-cystine with formaldehyde or glyoxylic acid in water at different pH values (ranging from 2 to 10) and temperatures (30, 100 and 210°C), but no evidence of adducts was obtained. This suggests that the cleavage of the cystine disulphide bond during the semipermanent hair straightening process is unlikely.

Interactions between lysine derivatives and aldehydes

It is well known that the e-amino group of lysine is involved in a plethora of biological processes acting as nitrogen nucleophile [31]. For example, in transamination reactions, the terminal amino group of a specific lysine belonging to the enzyme plays a key role in the reversible formation of aldimines with pyridoxal-50 -phosphate. Recently, particular attention has been directed towards the so-called glycation, a non-enzymatic step occurring in the Maillard reaction between non-reducing sugars and amino groups [32]. In general, the reaction between a primary amine and a carbonyl compound gives an imine that, depending on the reagents and reaction conditions, can undergo different fates such as exchange and metathesis reactions [33] and, in many cases, return to the starting reagents if water is not removed. Due to the reversible character of the reaction, in some cases (especially in complex biological matrices), the formation of the imine is demonstrated by the detection (i.e. by mass spectrometry) of the amine derivative [34] obtained by reduction ‘in situ’ of the imine. To gain information on the possible factors that may be involved in the hair straightening, we carried out some reactions on simple compounds. We decided to use N-a-acetyl-L-lysine (6, Scheme 2) as model for basic amino acids. The presence of the protection on the aamino group permits to better mimic the situation of the amino acid in the polypeptide fibre and also to avoid the reaction of the aldehyde with this group [35]. As imines derived from one aromatic reactant at least are more stable than those in which one of the two reactants is aliphatic, we decided to carry out the reaction between 6 and a series of aromatic aldehydes, that is, 7a–c (Scheme 2). The reactions were carried out directly in the NMR spectroscopy tube in DMSO-d6 at 25 and 50°C between equimolar amounts of the reagents. In these conditions, despite the extremely poor solubility of 6, the reaction proceeded due to the solubility of the reaction product. 1 H and 13C NMR spectral data of compounds 8a–c (Scheme 2), never reported prior in the literature, agree with those of an iminic structure. An interesting feature of the 1 H NMR spectrum was the observation, usually rare, of coupling between the proton bound to the nitrogen atom of the amide group and the proton bound to the adjacent carbon atom. The progress of the reaction was monitored through 1 H NMR spectroscopy, and the results obtained are summarized in Table I. The data shown in Table I indicate that two factors affect the spontaneous progress of the reaction, temperature and the electrophilicity of the carbonyl reagent. In fact, at a reaction time of 22 h, the increase of temperature from 25 to 50°C resulted in an increase of the conversion by 72% for 7a (see entries 1 and 4) and 20% for both 7b (see entries 5 and 8) and 7c (entries 9 and 12). Keeping constant the temperature, a drastic decrease of the conversion was observed on going from the nitro-substituted aldehyde 7a to the less reactive 7b and 7c, which reflects the effect of the substituents, in agreement with the Hammett parameters (rp = 0.78,
0.17,
0.27 for NO2, CH3, and OCH3, respectively) [36] and with the already reported data for reactions between aniline and p-substituted benzaldehydes [37]. From these results, it can be expected that the high electrophilicity of the carbonyl carbon atom of formaldehyde and glyoxylic acid plays an important role in favouring the attack by amino end-groups of basic amino acids. Then, we prepared the product derived from 6 and glyoxylic acid (3) which, to the best of our knowledge, has never been reported so far, being it subjected ‘in situ’ to reductive amination [38]. In this case, as the formulations containing the hair straightener are in aqueous medium, we carried out the reaction in water, dissolving equimolar amounts of the two reagents and then removing the solvent. The 1 H and 13C NMR spectra in DMSO-d6 of the residue agreed with the presence of the two imines 9a and 9b (Scheme 3). The absence in the 1 H NMR spectrum of the signal at d = 9.18 ppm (belonging to the aldehydic proton of 3) and the presence of two signals at d = 8.02 and d = 7.91 ppm in 54/46 relative percentage ratio (ascribable to the CH=N proton of the two isomeric imines) were particularly diagnostic. A further experiment was carried out by dissolving equimolar amounts of 3 and 6 in D2O, directly in the NMR spectroscopy tube and monitoring the reaction course. At room temperature, the spectrum showed only the signals of the reagents (the CH signal of glyoxylic acid in water was at d = 5.18 ppm, being it present in the hydrated form) [39]. After 2 h at 60°C, only traces of 9 were present in the spectrum, whereas after 6 h at 60°C, the conversion (evaluated on the basis of the integrals of the signal belonging to the methylene group bound to the iminic nitrogen atom) was about 4%. An increase of the temperature until 90°C produced, after 3 h, a 12% of conversion and the formation of two products. The diagnostic signals were those at d = 7.96 (s, 1H), 7.86 (s, 1H), 3.26 (t, J = 7.0 Hz, 2H), 3.20 (t, J = 7.0 Hz, 2H), the first and the second ones agree with those of a CH=N proton, and the others are ascribable to the methylene group bound to the iminic nitrogen atom of 9a and 9b. As expected, only after complete removal of water were they obtained quantitatively. These findings agree with the literature, which reports a negligible reactivity at room temperature in aqueous solution for glycation reactions between 3 and N-a-acetyl-L-lysine and also in reactions between 3 and the amino group of lysine residues belonging to human serum albumin [40,41]; the obtained results confirm the importance of the water removal to complete the condensation step forming the imine. Briefly summarizing, the study carried out on model molecules showed that the high electrophilicity of the carbonyl carbon atom and the temperature are important factors that favour imine formation. In the case of molecules such as formaldehyde and glyoxylic acid, the electrophilicity of the carbonyl, the low molecular size (which can favour the penetration within the hair scales) and the air-drying (which favours the dehydration) followed by the hightemperature ironing (which very likely allows a dehydration in the inner layers of the hair thus promoting condensation reactions) might be among the key factors responsible for the efficient hair straightening obtained with these reagents. Experiments were thus conducted on hair fibres in attempt to determine whether this was the case. These are presented and discussed as follows.

Raman and ATR/IR investigation on model compounds

To investigate the hair modifications during the straightening process with glyoxylic acid-based formulations, we have considered some reported procedures that describe the steps used in the treatment. Hair is treated with a solution of glyoxylic acid in water (at different concentrations, but usually 6%) [42] for a certain different time (on average 60 30 min) [13] then is dried with a hairdryer. Once dried, hair is straightened with an iron heated at about 200 50°C. After that, many stylists prefer to wait 48–72 h before washing hair [42,43]. Optionally, hair can be pre-treated by washing with an alkaline shampoo (pH 7.5–9.5) [13]. Based on this information, we conducted a series of Raman and ATR/IR experiments on solid yak fibres focusing particular attention on the possible formation of imines. First, we analysed Raman and ATR/IR spectra of the imines 9 and of the related starting materials 3 and 6. The samples derived from the condensation reaction between 3 and 6 showed a broad and strong mC=N Raman stretching band at about 1655 cm
1 [44] (Fig. S1), compatible with the presence of 9, according to NMR spectroscopy data. The broad band at about 1725 cm
1 in the spectrum of 9 is assignable to the mC=O stretching mode of the COOH group. This band appeared shifted and broadened in comparison with pure glyoxylic acid (1720 cm
1 ), suggesting the presence of the product of the reaction, but also a change in the hydrogen-bonding pattern. The IR spectra, shown in Fig. S2, confirm the Raman findings; also in this case, the products obtained from the reaction between 3 and 6 show a broad and strong band at a similar wavenumber position, compatible with the presence of the imines; as above reported for Raman spectra, the profile of the mC=O band due to the COOH group is not perfectly coincident with that of pure glyoxylic acid. The IR marker bands of the aldehyde group of glyoxylic acid at 2790 and 2710 cm
1 were not detected. This result confirms the occurrence of the reaction and the quantitative obtainment of the products upon water removal. After that, we collected Raman and IR spectra of yak fibres before and after each step used in the usual hair straightening treatment with glyoxylic acid solutions.

Raman investigation on yak fibres

Figure 1 reports the Raman spectra of untreated yak starting fibres (control sample, spectrum a), yak fibres after immersion in an aqueous KOH solution at pH approximately 8.7 for 5 min, rinsing with water, air-drying, bathing in an aqueous 6% glyoxylic acid solution for 30 min and air-drying (spectrum b) and plus ironing at 230°C (spectrum c). Figure 2 reports the percentages of secondary structure conformations as obtained by the curve fitting of the Raman Amide I range of the analysed samples. With regard to the control sample (Fig. 1, trace a), the position of Amide I and III (at 1654 and 1280 cm
1 , respectively) was typical of a-helix, that is, the main conformation present in the more internal cortical cells of wool (the weaker component at 1251 cm
1 is assignable to unordered conformations). Actually, it has been reported that [45] the Raman spectrum of wool fibres was in excellent agreement with that obtained from the cortical cells. Therefore, the Raman spectrum is sensitive to the bulk of the fibre. On the other hand, the Raman spectrum of the control yak fibres was in agreement with that recorded on the cortical region of black hair [46], as well as with the findings [47] that cortical microfibrillae are constituted by a-helix structures. The curve fitting of the Amide I region confirmed that the prevailing conformation was a-helix; as shown in Fig. 2, its content attained 44%. Also, the band at 935 cm
1 is assignable to this conformation (mCC skeletal stretching) [18,48]. Some bands are attributable to specific vibrational modes of amino acids residues, as indicated in Fig. 1. The band at 505 cm
1 is due to the mSS stretching mode of cystine disulphide bridges in gauche-gauche-gauche conformation [18]. The position of this band revealed that in yak fibres the Ca-Cb-S-S-Cb-Ca linkage takes the lowest potential energy conformation (i.e. gauche-gauchegauche) [18,49]. Other weaker components at about 493, 519 and 540 cm
1 have been assigned to strained, gauche-gauche-trans and trans-gauche-trans Ca-Cb-S-S-Cb-Ca conformations, respectively [18]. The Raman tyrosine doublet at about 850–830 cm
1 has been widely used to describe the average hydrogen-bonding state of the tyrosine phenoxyl groups in globular and fibrous proteins [18,50–52]. In fact, the I850/I830 ratio achieves its minimum value of about 0.3 when tyrosine residues are buried and the phenolic OH group acts as a strong hydrogen bond donor to an electronegative acceptor, such as carboxyl oxygen [50]. When tyrosines are exposed at the surface of the protein, the phenolic OH acts as both a donor and an acceptor of moderate hydrogen bonds and the I850/I830 is approximately 1.25 [50]. If the phenoxyl oxygen is the acceptor of a strong hydrogen bond from an electropositive group, such as a lysyl NH3 + group, and does not participate in significant hydrogen bond donation, the I850/I830 approaches a presumed maximum value of 2.5 [50]. More recent studies on filamentous virus capsids [51] and silk fibroin in Silk I form [52] allowed to refine the correlation and to verify that the I850/I830 intensity ratio can even exceed the latter value. In the spectrum of untreated yak fibres (Fig. 1, trace a), the I850/I830 intensity ratio was 1.27, that is, analogous to that of wool fibres (I850/I830 = 1.25) [53]. This result suggests that phenolic OH groups act as both donor and acceptor of moderate hydrogen bonds, according to literature data [50] and may be explained in relation to the amino acid composition of yak fibres [54]; actually, in the presence of considerable amounts of serine, threonine, aspartic and glutamic acids, tyrosine residues are expected to act as both donor and acceptor of H-bonds. To evaluate the effect of only the basic pre-treatment, yak fibres were immersed in the aqueous basic solution at pH = 8.7 for 5 min, rinsed with water and then blow-dried. The Raman spectrum recorded on the yak fibres subjected to this treatment (not shown) was not noticeably different from that recorded on the untreated sample. Actually, as shown in Figs 2 and 3(A,B), no significant changes in the conformational distribution, A935/ A1450 and AS-S/A1450 were observed. On the basis of the obtained results, it can be affirmed that the simple treatment with an aqueous KOH solution for 5 min followed by washing with water and blow-drying did not substantially alter the fibre structure and the occurrence of the lanthionization reaction seems poorly probable, in agreement with the reported data indicating that the lanthionization takes place after a basic treatment for 10–15 min [4,11]. Figure 1, trace b shows the Raman spectrum of yak fibres first immersed in aqueous KOH solution (for 5 min, then rinsed and air-dried), then bathed in aqueous 6% glyoxylic acid for 30 min and air-dried. The spectrum was significantly different to that recorded for the control and KOH pre-treated fibres. The data obtained by curve fitting (Fig. 2, data set f) would suggest that, if compared with the case of treatment with KOH solution alone, the treatment with KOH and glyoxylic acid should have caused a significant increase in the a-helix content (i.e. in the relative area of the band at about 1655 cm
1 ), which attained 50% (compare Fig. 2 data sets b and c). On the contrary, the A935/A1450 ratio did not undergo an analogous increase in this sample (Fig. 3A, cases b and c). This apparent contradiction may be explained by ascribing the increase of the area of the 1655 cm
1 component to the formation of iminic compounds; actually, as above reported, the mC=N stretching mode has been observed in this spectral range. The treatment with KOH and glyoxylic acid also caused changes in tyrosine environment and distribution of the S-S conformations. The I850/I830 intensity ratio increased (I850/I830 = 1.46), suggesting a change towards a more exposed environment for tyrosine residues. On the other hand, a certain broadening around 1730 cm
1 was detected, which suggests the incorporation of the COOH group of glyoxylic acid in the yak fibres. However, it may be observed that this mCOOH band was shifted with respect to that of the pure acid (1720 cm
1 ). This result, which appears even more evident in the ATR/IR spectra (see below), may also be explained by hypothesizing that the COOH group interacts with the protein chains by a hydrogen-bonding pattern different with respect to that occurring in the pure glyoxylic acid. The formation of hydrogen bonds with tyrosine residues may change their environment towards a more exposed state. Alternatively, this shift might be also due to the possible formation of a small amount of imine species, bearing the COOH group. Although no significant changes in the AS-S/A1450 ratio were detected (Fig. 3B, case c), the spectral profile in the mSS region would suggest a certain change in the distribution of the CaCb-S-SCbCa linkage (Fig. 1, trace b, down right corner): the relative content of the gauche-gauche-trans conformation (band at about 520 cm
1 ) seems to increase. The weakening of the component at about 1400 cm
1 (Fig. 1, trace b, down left corner), assignable to the dOH bending mode of serine, would suggest the involvement of this amino acid in the reaction with glyoxylic acid, in agreement with the IR spectra (see below). The spectrum c of Fig. 1, corresponding to the sample treated with glyoxylic acid and ironing with the KOH pre-treatment, showed, as reported for the previous sample, a shoulder at about 1730 cm
1 (due to the reacted glyoxylic acid), a weakening of the component at about 1400 cm
1 (assignable to serine) and an enrichment in the gauche-gauche-trans S-S conformation. If compared with the sample treated with KOH and glyoxylic acid (Fig. 1, trace b), the fibres that underwent the additional step of ironing (Fig. 1, trace c) showed a further change in the Amide I range; the shoulder at about 1670 cm
1 (due to b-sheet conformation) increased in intensity with respect to the component at about 1655 cm
1 . The curve-fitting analysis (Fig. 2, compare data sets c and d) was in agreement with this qualitative result, indicating a decrease of the a-helix content (from 50% to 39%) and an increase of the b-sheet content (from 28% to 38%) induced by ironing. Accordingly, the A935/A1450 ratio decreased (Fig. 3A, cases c and d). Ironing seemed to affect also the tyrosine environment. In fact, the I850/I830 intensity ratio decreased from 1.46 (KOH pre-treatment + glyoxylic acid) to 1.23 (KOH pre-treatment + glyoxylic acid + ironing), suggesting a change towards a more buried state for tyrosine residues upon ironing. Actually, according to the above reported classification by Siamwiza [50], lower I850/I830 values are characteristic of less exposed tyrosine residues. This result may be explained also in relation to the decrease in intensity observed in the IR spectra for the mCOOH band at about 1730 cm
1 upon ironing (see below): the presence of COOH groups, as above observed, may change the tyrosine environment towards a more exposed state while their loss to a more buried state.

ATR/IR investigation on yak fibres

ATR/IR spectra of yak fibres before and after each step of the usual straightening treatment are shown in Fig. 4. According to the Raman findings, no significant changes were observed upon the simple KOH treatment (spectrum not shown). With regard to the control yak sample, the wavenumber position of the Amide I, II and III modes (at 1622, 1515 and 1233 cm
1 , respectively) suggests a prevailing b-sheet and/or unordered bsheet/b-strand conformation [22]. An analogous result has been reported for wool fibres [45,53]: the cuticle cells, which constitute the outer layer of the fibre (i.e. that analysable with the ATR technique), are composed of keratin in b-sheet and disordered conformations, whereas the more internal cortical cells are constituted by keratin in a-helix conformation. As previously reported [46], the latter structure is absent in the hair cuticle. Some spectral features in Fig. 4 are ascribable to the relatively high serine content in cuticle cells: the bands at 1386 and 1075 cm
1 can be attributed to dOH [55] and mCO [56] vibrations of serine, respectively. Upon the treatment of fibres with KOH and glyoxylic acid as well as after the subsequent ironing, the most interesting spectral feature was the significant decrease in intensity of the band at 1386 cm
1 , ascribable to the dOH bending mode of serine [55] (Fig. 4, spectra b and c). This result suggests the involvement of this amino acid in the reaction. An analogous finding has been obtained upon sulphation of the serine residues of wool fibres [53]. The marker bands of the aldehyde group of glyoxylic acid at 2790 and 2710 cm
1 disappeared in both the spectra. The band at about 1730 cm
1 , ascribable to the mC=O stretching mode of the COOH group of incorporated glyoxylic acid, was detected with a higher intensity in the spectrum of the sample treated solely with glyoxylic acid (Fig. 4, spectrum b) than upon the subsequent ironing (Fig. 4, spectrum c). Accordingly, the bands at about 1230 and 1050 cm
1 showed the highest shift and increase in intensity in the former sample, due to the highest contribution of the glyoxylic acid bands at 1220 and 1050 cm
1. These trends suggest that ironing caused a decrease in the incorporated glyoxylic acid content, even if the occurrence of other reactions (favoured by the drastic temperature reached with ironing) [57], including that of glyoxylic acid cannot be ruled out. It is interesting to note that the mC=O stretching band of the COOH group appeared shifted with respect to the pure acid (1724 cm
1 ), as above reported for the Raman spectra; this result suggests that glyoxylic acid reacted (actually, the aldehyde marker bands at 2790 and 2710 cm
1 disappeared) or, more in general, that the COOH group was involved in a hydrogen-bonding pattern different with respect to the pure acid. Only slight changes in the Amide I and II regions upon the treatments were observed, suggesting that conformational rearrangements into the first 2 lm of sample thickness (i.e. those analysable by the ATR technique) were negligible.

Scanning electron microscope investigation on yak hair

To investigate the morphological modifications of the hair cuticle during the straightening process, we analysed the fibres longitudinally through a SEM in variable pressure mode in order to avoid the sample pre-treatment (gold sputtering) that might have caused artefacts. We analysed both yak and curly human hair subjected to different treatments. For each sample, SEM images were recorded on a central region of the fibre belonging to the same lock; two analyses on two different, near, regions along the same fibre were made. Figure 5(A,B,D,E) show the SEM images obtained on yak hair after each step of the usual straightening procedure. Figure 5(C) is obtained after ironing of hair treated with base. Other images have been added in Appendix S1. To the best of our knowledge, these are the first reported SEM images related to the semipermanent hair straightening with aldehydes. Although a fair comparison should be performed on the same fibre at exactly the same region before and after the treatment, at a first glance the images of Fig. 5 show that hair treated with a basic solution and then ironed (Fig. 5C) seems to have a surface less regular with respect to samples A (Fig. 5A) and B (Fig. 5B), thus indicating a possible damage of the cuticle after the thermal stress, whereas the sample subjected to the straightening procedure with glyoxylic acid (Fig. 5E) shows a surface almost regular, similar to that of the control sample (Fig. 5A). Similar images have been obtained by SEM analyses of human hair subjected to the same procedure (Figs S3–S17), and this corroborates the choice of yak fibres as a model for human hair for the Raman and ATR/IR analyses. It has also previously been reported that yak hair fibres are suitable models for studies on human hair [26].

Conclusions

This study represents the first multidisciplinary investigation on the hair modifications and reactions involved in the semipermanent hair straightening treatment with carbonyl-based compounds. First, the attention was focused on the possible reversible reactions that can occur between carbonyl groups belonging to hair straightener such as formaldehyde and glyoxylic acid and nucleophilic sites on the side chain of amino acids belonging to the keratin. To this end, Nacetyl-L-cysteine and N-a-acetyl-L-lysine were chosen as model compounds, and it was shown that imine formation easily occurred between the e-amino group of the latter and very electrophilic aldehydes such as formaldehyde and glyoxylic acid. The importance of water removal for imine formation, as is already well known, was also demonstrated. The new imines formed were used as reference compounds in the Raman and IR investigations. These latter techniques indicated that the hair straightening with glyoxylic acid causes some rearrangements in the secondary structure distribution as well as some conformational changes at the level of disulphide bridges. The use of the Amide I curve-fitting data together with the a-helix A935/A1450 marker ratio could be used as an indicator for the formation of imines, whose mC=N band was superimposed on the 1655 cm
1 band of a-helix. Taking into account that the hair fibres were dried and thus most of the water was removed, it is clear that the common, high-temperature straightening conditions are suitable for the formation of imines. Regarding the side chains of amino acids such as serine and threonine that may behave as oxygen nucleophiles giving hemi-acetals, the Raman and IR spectra showed a decrease of the dOH bending mode of serine suggesting a possible involvement of this amino acid in the reaction with glyoxylic acid. Furthermore, the IR spectra indicated that the treatment with glyoxylic acid produced the major conformational rearrangements within the hair rather than in the cuticle. In conclusion, although the mechanism of interaction between the components of the straightening formulations and the hair could be more complex involving not only the formation of imines or hemi-acetals (e.g. the formation of cross-links involving two or more polypeptide chains or other effects might also be possible), the known reversibility of the treatment (semipermanent) with aldehydic compounds supports the occurrence of the reactions proposed in this study suggesting that the search for new hair-smoothening agents might take in account the findings herein reported.

Acknowledgements

Work supported by Alma Mater Studiorum – University of Bologna (RFO funds) and by Ilios Srl. Many thanks to Dr. Iuri Boromei for his technical support in SEM analyses.

References

1. Robbins, C.R. (ed.) Chemical and Physical Behavior of Human Hair, 5th edn. SpringerVerlag, Berlin, Heidelberg (2012). 2. Zviak, C. (ed.) The Science of Hair, pp. 149– 182, 183–212. Marcel Dekker, New York (1986). 3. Miranda-Vilela, A.L., Botelho, A.J. and Muehlmann, L.A. An overview of chemical straightening of human hair: technical aspects, potential risks to hair fibre and health and legal issues. Int. J. Cosmet. Sci. 36, 2–11 (2014). 4. Malle, G., Barbarat, P. and Pasini, I. Procede de defrisage des fibres keratiniques avec un moyen de chauffage et un derive d’acide. WO 2007/135299 A1, L’OREAL, Paris (2007) 5. Tolgyesi, E. and Fang, F. Hair Research, pp. 11–122. Springer, Berlin (1981). 6. McMullen, R. and Jachowicz, J. Thermal degradation of hair. I. Effect of curling irons. J. Cosmet. Sci. 49, 223–244 (1998). 7. Monakhova, Y.B., Kuballa, T., Mildau, G., Kratz, E., Keck-Wilhelm, A., Tschiersch, C. and Lachenmeier, D.W. Formaldehyde in hair straightening products: Rapid 1 H NMR determination and risk assessment. Int. J. Cosmet. Sci. 35, 201–206 (2013). 8. Pierce, J., Abelmann, A. and Spicer, L. Characterization of formaldehyde exposure Resulting from the use of four professional hair straightening products. J. Occup. Environ. Hyg. 8, 86–699 (2011). 9. IARC. Formaldehyde. IARC Monogr. Eval. Carcinog. Risks Hum. 100F, 401–435 (2012). 10. European Commission. Council directive 76/ 768/eec of 27 July 1976 on the approximation of the laws of the member states relating to cosmetic products. Off. J. Europ. Comm. L 149, 38 (1976). 11. Malle, G., Barbarat, P. and Pasini, I. Method for straightening keratinous fibres using heating means and an acid derivative. US Patent 0 300 471 A1 (2010). 12. Coppola, P. and Bucario, V.C. Reactive keratin protein formulations and methods of using for revitalizing hair. US 0 211 593 A1, 2009. 13. Mannozzi, A. Process for semi-permanent straightening of curly, frizzy or wavy hair. WO 2011104282 A1 and A2, Alderan S.A.S. di D’Ottavi Adele, Ascoli Piceno (IT) (2011). 14. Daubresse, N., Samain, H. and Plos, G. Cosmetic treatment method, and composition comprising a glyoxylic acid derivative. WO 2013117771 A2, L’OREAL, Paris (2013). 15. Mannozzi, A. Process for semi-permanent straightening of curly, frizzy or wavy hair. WO 2012010351 A2, Alderan S.A.S. di D’Ottavi Adele, Ascoli Piceno (IT) (2012). 16. Mannozzi, A. Process for semipermanent straightening of curly, frizzy or wavy hair. US 0 118 520 A1, KAO Corp. (2013). 17. Mannozzi, A. Process for semipermanent straightening of curly, frizzy or wavy hair. IT 1398503 B1 (2013). 18. Tu, A.T. Raman Spectroscopy in Biology: Principles and Applications. John Wiley, New York (1982). 19. Parker, F.S. Applications of Infrared, Raman, and Resonance Raman Spectroscopy in Biochemistry. Plenum Press, New York (1983). 20. Ackermann, K.R., Koster, J. and Schlucker, € S. Polarized Raman microspectroscopy on intact human hair. J. Biophotonics 1, 419– 424 (2008). 21. Lefevre, T., Rousseau, M.-E. and P
ezolet, M. Protein secondary structure and orientation in silk as revealed by Raman spectromicroscopy. Biophys. J . 92, 2885–2895 (2007). 22. Kuzuhara, A. Analysis of structural change in keratin fibres resulting from chemical treatments using Raman spectroscopy. Biopolymers 77, 335–344 (2005). 23. Kuzuhara, A. Protein structural changes in keratin fibres induced by chemical modification using 2-iminothiolane hydrochloride: a Raman spectroscopic investigation. Biopolymers 79, 173–184 (2005). 24. Kireche, M., Peiffer, J.L., Antonios, D. et al. Evidence for chemical and cellular reactivities of the formaldehyde releaser bronopol, independent of formaldehyde release. Chem. Res. Toxicol. 24, 2115–2128 (2011). 25. Tome, D. and Naulet, N. Carbon 13 nuclear magnetic resonance studies on formaldehyde reactions with polyfunctional amino-acids. Int. J. Pept. Protein Res. 17, 501–507 (1981). 26. Boga, C., Delpivo, C., Ballarin, B., Morigi, M., Galli, S., Micheletti, G. and Tozzi, S. Investigation on the dyeing power of some organic natural compounds for a green approach to hair dyeing. Dyes Pigm. 97, 9– 18 (2013). 27. Earland, C. and Raven, D.J. Lanthionine formation in keratin. Nature 191, 384 (1961). 28. Martuscelli, E. Degradazione delle fibre naturali e dei tessuti antichi, Aspetti Chimici, Molecolari, Strutturali e Fenomenologici. Paideia, Firenze (2006). 29. Menefee, E. and Yee, G. Crosslinking in keratins. IV. Thermal cleavage of crosslinks. J. Appl Polym. Sci. 9, 2847–2853 (1965). 30. Kallen, R.G. Mechanism of reactions involving Schiff base intermediates. Thiazolidine formation from L-cysteine and formalde-hyde. J. Am. Chem. Soc. 93, 6236–6248 (1971). 31. McMurry, J.E. and Begley, P.T. The Organic Chemistry of Biological Pathways. Roberts and Company, Englewood (2005). 32. Poulsen, M.V., Hedegaard, R.V., Andersen, J.M. et al. Advanced glycation endproducts in food and their effects on health. Food Chem. Toxicol. 60, 10–37 (2013). 33. Belowich, M.E. and Stoddart, F. Dynamic imine chemistry. Chem. Soc. Rev. 41, 2003– 2024 (2012). 34. Humeny, A., Kislinger, T., Becker, C.M. and Pischetsrieder, M. Qualitative determination of specific protein glycation products by matrix-assisted laser desorption/ ionization mass spectrometry peptide mapping. J. Agric. Food Chem. 50, 2153–2160 (2002). 35. Bocoum, A., Boga, C., Savoia, D. and Umani-Ronchi, A. Diastereoselective allylation of chiral imines. Novel application of allylcopper reagents to the enantioselective synthesis of homoallyl amines. Tetrahedron Lett. 32, 1367–1370 (1991). 36. Hansch, C. and Leo, A. Substituent Constants for Correlation Analysis in Chemistry and Biology. Wiley-Interscience, New York (1979). 37. Layer, R.W. The chemistry of imines. Chem. Rev. 63, 489–510 (1963). 38. Csuk, R., Stark, S., Barthel, A., Kluge, R. and Strohl, D. A robust synthesis of € Ne-(carboxymethyl)-l-lysine (CML). Synthesis 1933–1934 (2009). 39. Chastrette, F., Bracoud, C., Chastrette, M., Mattioda, G. and Christidis, Y. Study of aqueous glyoxylic-acid solutions by C-13 NMR. Bull. Soc. Chim. Fr. 1, II- 66–74 (1985). 40. Dutta, U., Cohenford, M.A., Guha, M. and Dain, J.A. Non-enzymatic interactions of glyoxylate with lysine, arginine, and glucosamine: a study of advanced non-enzymatic glycation like compounds. Bioorg. Chem. 35, 11–24 (2007). 41. Schmitt, J., Schmitt, G., Munch, G.-M. and € Milencovic, J. Characterization of advanced glycation end products for biochemical studies: side chain modifications and fluorescence characteristics. Anal. Biochem. 338, 201–215 (2005). 42. Private communication from the coauthors of Ilios srl. 43. Weathersby, C. and McMichael, A. Brazilian keratin hair treatment: a review. J. Cosmet. Dermatol. 12, 144–148 (2013). 44. Colthup, N.B., Daly, L.H. and Wiberley, S.E. Introduction to Infrared and Raman spectroscopy, 3rd edn. Academic Press, San Diego (1990). 45. Church, J.S., Corino, G.L. and Woodhead, A.L. The analysis of merino wool cuticle and cortical cells by Fourier transform Raman spectroscopy. Biopolymers 42, 7–17 (1997). 46. Kuzuhara, A. and Hori, T. Reduction mechanism of tioglycolic acid on keratin fibres using microspectrophotometry and FTRaman spectroscopy. Polymer 44, 7963– 7970 (2003). 47. Fraser, R.D.B., Macrae, T.P. and Rogers, G.E. Molecular organization in alpha-keratin. Nature 193, 1052–1054 (1962). 48. Frushour, B.G. and Koenig, J.L. Raman spectroscopic study of poly (b-benzyl-L-aspartate) and sequential polypeptides. Biopolymers 14, 2115–2135 (1975). 49. Pielesz, A. and Wlochowicz, A. Semiempirical infrared spectra simulations for some aromatic amines of interest for azo dye chemistry. Spectrochim. Acta, Part A 57, 2637–2646 (2001). 50. Siamwiza, M.N., Lord, R.C., Chen, M.C., Takamatsu, T., Harada, I., Matsuura, H. and Shimanouchi, T. Interpretation of the doublet at 850 and 830 cm-1 in the Raman spectra of tyrosyl residues in proteins and certain model compounds. Biochemistry 14, 4870–4876 (1975). 51. Thomas, G.J. Jr New structural insights from Raman spectroscopy of proteins and their assemblies. Biopolymers 67, 214–225 (2002). 52. Taddei, P., Asakura, T., Yao, J. and Monti, P. Raman study of poly(alanine-glycine)- based peptides containing tyrosine, valine, and serine as model for the semicrystalline domains of Bombyx mori silk fibroin. Biopolymers 75, 314–324 (2004). 53. Taddei, P., Tsukada, M. and Freddi, G. Affinity of protein fibres towards sulfation. J. Raman Spectrosc. 44, 190–197 (2013). 54. Yan, K., H€ocker, H. and Sch€afer, K. Handle of bleached knitted fabric made from fine yak hair. Text. Res. J. 70, 734–738 (2000). 55. Koenig, J.L. and Sutton, P.L. Raman scattering of some synthetic polypeptides: Poly (©-benzyl L-glutamate), poly-L-leucine, polyL-valine, and poly-L-serine. Biopolymers 10, 89–106 (1971). 56. Gupta, A., Tandon, P., Gupta, V.D. and Rastogi, S. Vibrational dynamics and heat capacity of
-poly(l-serine). Polymer 38, 2389–2397 (1997). 57. Back, R.A. and Yamamoto, S. The gas-phase photochemistry and thermal decomposition of glyoxylic acid. Can. J. Chem. 63, 542–548 (1985).

HISTORY

•HAIR STRAIGHTENING/SMOOTHING WITH FORMALDEHYDE WAS STARTED IN RIO DE JANEIRO, BRASIL IN AND AROUND 2003. • THE SOLUTIONS OF FORMALDEHYDE @ 1.00% TO 4.5% LEVEL WERE USED TO STRAIGHTEN WAVY AND CURLY HAIR. • FORMALDEHYDE WAS APPLIED TO HAIR AND HAIR BLOW-DRIED AND FLAT IRONED AT 450°F. •IT RESULTED IN SMOOTH SHINY AND STRAIGHT HAIR.

HUMAN HEALTH & FORMALDEHYDE

• FUMES EMITTED DURING THE STRAIGHTENING/SMOOTHING TREATMENTS CONTAINING FORMALDEHYDE CAN CAUSE: –EYE IRRITATION, RESPIRATORY IRRITATION, AND IRRITATION TO NOSE AND THROAT. (TOXICOLOGICAL PROFILES, 2010) + (AGENCY FOR TOXIC SUBSTANCES AND DISEASE REGISTRY: TOXICOLOGICAL PROFILE FOR FORMALDEHYDE. IN TOXICOLOGICAL PROFILES. ATLANTA, GA.: DEPARTMENT OF HEALTH AND HUMAN SERVICES, 1999) –BURNING OF EYES AND THROAT, WATERING OF EYES, DRY MOUTH, LOSS OF SMELL, HEADACHE AND FEELING OF GROGGINESS (OSHA OREGON STUDY, 2010) • KNOWN HUMAN CARCINOGEN (NATIONAL TOXICOLOGY PROGRAM. SUBSTANCE PROFILE: FORMALDEHYDE. 12TH REPORT ON CARCINOGENS, 2011)

Formaldehyde & Government Regulations

• ANVISA (Brazil) + FDA (USA) + REACH (EU): allow the use of Formaldehyde from 0.02% to 0.10% active level as a preservative • OSHA has three airborne exposure levels for formaldehyde: 1. Permissible Exposure Limit (PEL) IS 0.75 ppm of formaldehyde in 8 hours. 2. The action level (AL) for 8 hours is 0.50 ppm Short Term Exposure Limit (STEL) of 15 minutes IS 2 ppm NIOSH (National Institute of Safety and Health) • 8 - hour recommended exposure limit (REL) is 0.016 ppm • 15 - minute recommended exposure limit (REL-C) is 0.1 ppm ACGIH (American Conference of Governmental Industrial Hygienists) • threshold limit value (TLV-C) ceiling is 0.3 ppm

GLYOXYLOL CARBOCYSTEINE

GLYOXYLOL CARBOCYSTEINE WAS USED AS A SMOOTHING/ STRAIGHTENING AGENT

Hypotheses

Hypothesis 1 -The hair volume is reduced significantly. Hypothesis 2 The properties of hair such as ease of combing, shine, and humidity resistance are significantly improved Hypothesis 3 The bulk samples DO not contain formaldehyde. Hypothesis 4 The air analysis during Restructuring liberates a small amount of formaldehyde

Literature Review

The use of formaldehyde in Industry and Chemical Reactions of Wool with formaldehyde. Analysis of fumes emitted during the treatment of hair fibers with formaldehyde based products OSHA Oregon Study, Pierce et al Study

REACTIONS OF FORMALDEHYDE WITH WOOL KERATIN

Formaldehyde can form crosslinks with keratin (1), such as NH CH2 NH bridges, and between N-terminal cysteine and amine groups S CH2 NH The best pH for the reaction is between 6.0 to 7.0. 1. Hinton, E. (1974). A Survey and Critique of the Literature on Crosslinking Agents and Mechanisms as Related to Wool Keratin. Textile Research Journal, p. 256.

REACTION OF FORMALDEHYDE WITH WOOL KERATIN - continued

According to Simpson (2002), side chains of amino acids of keratin such as ARG, LYS, TYR, HIS, AND amide derivatives of ASP, and GLU react with formaldehyde and some of these reactions can be bi-functional and mono-functional. The simple crosslinks are CH2 . It is very difficult to verify the sites and the extent of formaldehyde keratin reactions as most of the modified side chains of amino acids are not stable to hydrolytic reaction conditions (2). 2. Simpson, W.S. (2002). Wool Chemistry. Ed: W.S. Simpson and G.H. Crawshaw; Woodhead Publishing, p. 153.

OSHA OREGON STUDY

The bulk sample analysis of smoothing/ straightening products with respect to formaldehyde contents was conducted by using EPA Method 8315. The exposures to the hairstylists, any bystanders in the salon, were conducted by using NIOSH 2016 Method. 3. McCarthy, K., McLaughlin, D., Montgomery, D., Munsell, P., Schuster, M., and Wood, M. (2010). Keratin Based Hair Smoothing Products And the Presence of Formaldehyde. OSHA Oregon and CROET. www.orosha.org/pdf/Final_Smoothing

Pierce, et al Study

The recent 2011 study of Pierce, et al has reported formaldehyde exposure of four professional hair straightening products where they have used similar methods of analysis: The study used EPA 8315 METHOD for bulk analysis of various brands. For hairstylists exposure the NIOSH 2016 method was employed (6). This study utilized both passive and active sampling techniques during sample collections at various stages of the process. 6. Pierce, J.S., Abelmann, A., Spicer, L.J., Adams, R.E., Glynn, M.E., Neier, K., Finley, B.L., and Gaffney, S.H. (2011). Characterization of Formaldehyde Exposure Resulting from the use of Four Hair Straightening Products. Journal of Occupational and Environmental Hygiene, 8:11, 686-699.

METHODOLOGY

Hair Smoothing/Straightening by Visual Method Fiber Elasticity Measurements Hair Combing Using Instron Moisture Contents Using MicroWave Resonance Technique Hair Shine Using Digital CAMERA Analysis of Bulk Sample for Formaldehyde EPA 8315 Method Analysis of Air During RestructurING NIOSH 2016 METHOD

Hair Smoothing By Visual Method

The treated tresses are visually compared against untreated tress for the degree of straightening. The degree of smoothing is measured on a Likert Scale of 1 to 5. The 1 being very poor smoothing/straightening to 5 being very good smoothing/straightening.

Fiber Elasticity Index by Using Dynamic Mechanical Analyzer

In this test, each single fiber (gauge length = 14.82 mm) was mounted to the submersible fiber specimen clamp containing water. The fiber was stretched to a constant strain or 0.5% of its length for 0.1 minute and allowed to recover for 0.9 minute. This process of imposing the strain and allowing it to recover was repeated for a total of 10 cycles. Index = After / Before treatment Index = 1.0 (no change); > 1.0 (strengthening); < 1.0 (loss in strength)

Ease of Hair Combing

The combing test was performed on hair tress using an Instron Materials Testing System model 5542, hooked-up to a computer and equipped with Bluehill software Load cell capacity of 50 Newton or 5.1Kg. Each hair tress was combed before and after treatment. The amount of total energy (milli Joules) required to comb each hair tress was measured. the highest force or peak load (gmf) was used to measure the relative ease or difficulty to detangle . Five consecutive combing readings were taken from each tress. The combing Index (After Treatment / Before Treatment) was calculated FOR 5 TRESSES weighing approximately 4 gm and 7 inches in length Combing Energies were noted and compared.

MOISTURE CONTENTS

MICROWAVE RESONANCE

•MICROWAVES RESONATE IN AN EMPTY CHAMBER. •INSERTING EACH HAIR SPECIMEN ONTO THE APPLICATOR TUBE SHIFTS THE RESONANCE DOWN AND INCREASES THE BANDWIDTH THAT ENABLE US TO OBTAIN THE MICROWAVE RESONANCE VALUES. •THESE RESONANCE VALUES WERE CALIBRATED AGAINST A DIRECT LABORATORY REFERENCE GRAVIMETRIC ANALYSIS FROM HUMIDITY RANGE OF 35% TO 80% RH. THE R-SQUARE VALUE OBTAINED FROM THE MEASURED RESONANCE VALUES AND MOISTURE CONTENT OF HAIR WAS 0.96.

Hair Shine by Digital Camera

•A LEVEL OF SHINE IS REFLECTED FROM A SURFACE OF THE HAIR TRESS. •THE DIGITAL CAMERA CAPTURES THE IMAGE THAT IS TRANSFERRED TO AN IMAGE ANALYSIS SOFTWARE (IMAGE PRO 5.1), WHICH CAN PRECISELY ESTIMATE THE BRIGHTNESS AND THE LIGHT INTENSITY OF DIGITAL IMAGE. •A SELECTED RECTANGULAR AREA OF THE IMAGE IS CROPPED AND THE LIGHT INTENSITY ALONG THIS RECTANGULAR AREA IS MEASURED. •THE MEASUREMENT PRODUCES THE LIGHT INTENSITY CURVE WITH ONE SIGNIFICANT MAXIMUM. THE HEIGHT H AND THE HALF-WIDTH OF MAXIMUM W ARE USED IN CALCULATING A FACTOR FS = H/W, WHICH IS ADDITIONALLY MULTIPLIED BY 0.729. •THE SHINE FACTOR FOR THE PERFECT TRESS WOULD BE EQUAL TO 1.

Analysis of Bulk Sample for Formaldehyde EPA 8315 Method

In this method, the aldehyde present in a product forms its derivatives with 2, 4- dinitrophenyl hydrazine and the RESULTANT derivative is a hydrazone (3). One drop of the bulk is weighed and diluted to 10 milliliters in water. Then 200 microliters of this solution are added to 2 milliliters of 2, 4-dinitrophenyl hydrazine solution in acetonitrile. The resultant mixture is analyzed by reverse phase High Performance Liquid Chromatography (HPLC) using a methanol/water eluent with a C 18 column and a diode array detector. 4. Kazuhiro Kuwata, et. al (1979). Determination of Aliphatic and Aromatic Aldehydes in Polluted Airs as 2,4-Dinitrophenyl Hydrazones by High Pressure Liquid Chromatography. Journal of Chromatic Sciences, Vol17, p. 264.

Analysis of Air During Processing of Hair with Restructure NIOSH 2016

THE UMAX 100 PASSIVE SAMPLERS (CATELOG NO. 500-100 BY SKC INC.)were utilized. This sampler comes in the form of a small BADGE filled with sorbent material that can be easily attached to the collar of the person who is being tested for formaldehyde or can be placed in an area under test. More details can be found on web site www.skcinc.com. The COLLAR BADGE containeS 2, 4-dinitrophenyl hydrazine on silica gel and reacts with any aldehydes present in the air or the fumes from the product under use. The time of exposure is noted. The contents of the BADGE are placed in auto sampler vials and desorbed in 2 milliliters acetonitrile. The tubes are then analyzed by reverse phase High Performance Liquid chromatography on a C 18 column with a methanol/water eluent. The detection is made at 353 nanometers with a diode array detector.

GLYOXYLOL Carbocysteine Based Smoothing System

I.Shampoo II.Fiber Restructuring Lotion III.Conditioning Mask

ii. Fiber Smoothing Treatment Lotion

It is based upon Glyoxylol Carbocysteine and Amino Acids at 15% active level • pH of the product is 1.50 1.70 • It is left on the hair for 20 to 30 minutes, blow dried in and then flat ironed at 230°C. • Hair is rinsed with water, shampooed gently, conditioned, and blow dried and flat ironed for final style.

Results

Smoothing System Summary

• Less damaging straightening systems • Hair volume is reduced significantly for Type 2A, 2B, 2C, and 3A hair types • Hair is very easy to comb during wet and dry combing • Hair is resistant to humidity and less prone to frizz • Hair has radiant shine • Daily hair styling and manageability of hair is remarkably easy

Bulk Sample Analysis for CarbonYL Compounds via HPLC

.ANALYTE NAME QUANTITY MRL (METHOD REPORT LIMIT) (PARTS PER BILLION) FORMALDEHYDE UNDETECTED 80,000 PPB ACETALDEHYDE UNDETECTED 80,000 PPB

FORMALDEHYDE ANALYSIS x VARIOUS STAGES

VARIOUS STAGES RESULTS IN PPM RESTRUCTURING LOTION APPLICATION (AVE: 19 MINS) NOT DETECTED DRYING RESTRUCTURING LOTION ON HAIR (AVE: 18.25 MINS) NOT DETECTED FLAT IRONING: RESTRUCTURING LOTION (AVE: 65.4 MINS) 0.14 TOTAL EXPOSURE (8 HOURS TWA) 0.05

CONCLUSION

Glyoxylol Carbocysteine is very effective in straightening Type 2 European and Hispanic hair. The elasticity after straightening with Glyoxylol Carbocysteine is less than formaldehyde but formaldehyde is not as good OF A straightening agent. Also Formaldehyde products have to be left in the hair for 72 hours for better straightening The ease of combing of Type 2 hair increases significantly after the treatment. The hair shine is significantly higher than formaldehyde treated hair. The hair is more resistant to humidity absorption when treated with Glyoxylol Carbocysteine. The presence of formaldehyde is not detected in the bulk sample of Restructuring Lotion containing Gloxylol Carbocysteine The formaldehyde emitted is well within the OSHA guidelines during heat treatment with Restructuring Lotion containing Glyoxylol Carbocysteine.

ACKNOWLEDGEMENTS

Thanks to following R&D Team Members: Thomas Ventura Dr. P. Milzarick Dr. Ali Khan Dr. RajEn Gandhi Hasan Syed Our Hair Stylists in the our Test Salon

REFERENCES

1. Hinton, E. (1974). A Survey and Critique of the Literature on Crosslinking Agents and Mechanisms as Related to Wool Keratin. Textile Research Journal, p. 256.2. Simpson, W.S. (2002). Wool Chemistry. Ed: W.S. Simpson and G.H. Crawshaw; Woodhead Publishing, p. 153. 2. Simpson, W.S. (2002). Wool Chemistry. Ed: W.S. Simpson and G.H. Crawshaw; Woodhead Publishing, p. 153. 3. McCarthy, K., McLaughlin, D., Montgomery, D., Munsell, P., Schuster, M., and Wood, M. (2010). Keratin Based Hair Smoothing Products And the Presence of Formaldehyde. OSHA Oregon and CROET. www.orosha.org/pdf/Final_Smoothing 4. Kazuhiro Kuwata, et. al (1979). Determination of Aliphatic and Aromatic Aldehydes in Polluted Airs as 2,4-Dinitrophenyl Hydrazones by High Pressure Liquid Chromatography. Journal of Chromatic Sciences, Vol17, p. 264. 5. National Institute of Occupational Safety and Health (NIOSH): Formaldehyde. Method 2016. In NIOSH Manual of Analytical Methods, 4 th ed. S.P. Tucker (ed), 2003. 6. Pierce, J.S., Abelmann, A., Spicer, L.J., Adams, R.E., Glynn, M.E., Neier, K., Finley, B.L., and Gaffney, S.H. (2011). Characterization of Formaldehyde Exposure Resulting from the use of Four Hair Straightening Products. Journal of Occupational and Environmental Hygiene, 8:11, 686-699.