Effect of equilibrium pH on the structure and properties of bleach‐damaged human hair fibers

Abstract Hair proteins are significantly affected by environmental pH. This impact tends to increase with prior hair damage. To understand how pH affects bleached hair properties, we utilized a number of techniques allowing for the determination of hair thermal properties, swelling and water sorption, and dry and wet tensile properties. At pH 5, hair proteins had the best structural integrity, as determined by differential scanning calorimetry and the highest tensile modulus. At pH 10, protein cross‐linking density decreased, water content and hair cross‐sectional diameter increased. Alkaline treatment, when compared with pH 5, did not reduce intermediate filament conditions (evaluated via enthalpy measurement) nor mechanical property performance in the wet state. In contrast to alkaline‐treated hair, bleached hair equilibrated at pH 3 behaved very differently: it contained two different crosslink density zones, was the least stiff in dry and stiffest in wet conditions. Additionally, it absorbed less water and had the lowest diameter because of reduced water binding by protonated carboxyl groups. The pH 3 to 10 did not affect the mechanical strength of bleached hair in dry or wet conditions.


| INTRODUCTION
Human hair is a complex, multi-compartmented biological substrate. It comprises an inner cortex encased in a protective cuticle layer. The cortex makes up most of the hair mass and contains keratin and keratin-associated proteins (KAP) that provide much of the fiber's mechanical strength. Thicker hair may also have a medulla, which consists of air-filled voids and runs, laterally, along the center of the fiber.
The key structural proteins in the hair cortex are the keratins.
Keratins are sub-classified as acidic type I keratins and neutral type II keratins. [1] Types I and II coil together in pairs, in α-helical structures called protofilaments. [2] Seven to ten pairs of protofilaments then cluster together to form intermediate filaments (IF). [2] IF's are embedded in an amorphous, interfilamentous matrix comprised of KAP. [2] The KAPs are cross-linked with each other and with the keratins through disulfide bonds; these play a key role in determining the physical properties of the fiber. [3] The combination of filaments embedded in a continuous matrix phase is commonly used to explain the mechanical properties of hair fibers and is termed the "two-phase" model of hair. [4] A modification introducing the third phase-KAP crosslinks with IF affecting thermal and mechanical properties was proposed as well. [5,6] The proteins are amphoteric in nature and contain both acidic and basic side chains. At neutral pH's, away from proteins' isoelectric point, these groups attract each other through electrostatic forces and form cross-links commonly referred to as salt-links. These saltlinks contribute to the strength of the fiber. Titration experiments show that when hair is placed in an acid solution, hydrogen ions are absorbed by the hair. Between pH 7 and approximately pH 4 hydrogen ions concentration is low, thus relatively few hydrogen ions are absorbed by hair. Below pH 4 hydrogen ion absorption increases substantially. [7] Studies on wool fibers show that below pH 4 there is an increase in the ease of fiber extension in a wet state. [8] Similar studies on human hair fibers also show that below pH 4 (65% RH) there is a reduction in fiber elastic modulus, [8] and that there is a small amount of fiber swelling. [9] The protonation of carboxyl groups at low pH's replaces the salt-linkages with hydrogen bonds thus weakening hair structures (much weaker bonds). Below pH 2 there may also be acid hydrolysis of peptide bonds that causes irreversible structural changes with concomitant fiber swelling.
In titration experiments with alkaline solutions, OH ions are absorbed by the hair. Between pH 7 and pH 10, few hydroxide ions are absorbed. [7] Above pH 10, hydroxide sorption increases markedly as the hydroxyl ion concentration is increased. [7] As with low pH studies, high pH (more than pH 10) break salt-bridges and weaken both hair [7] and wool. [8] In addition, hair undergoes considerable swelling at high pH, becoming more porous. Alkali hydrolysis of peptide and disulfide bonds above pH 10 weakens the hair further and causes hair damage.
Thermal analysis methods, such as differential scanning calorimetry (DSC), enable us to investigate the effects of treatments on the structural properties of hair proteins. [10] Wet DSC techniques have been used to investigate the effects of various hair treatments. Data from these studies suggest that, broadly speaking, the keratin denaturation enthalpy (ΔH) is related to the structural integrity of the IFs, whereas the keratin denaturation temperature correlates with the cross-linking density of the KAPs, [10] some authors also include the impact of cross-linking between KAP and IF in the interpretation of this parameter. [6] The experiments combining short-term hair soaking in acidic and alkaline conditions with wet DSC have shown that denaturation temperature and enthalpy vary markedly with different pH treatments. [6] Data show that both measures increase as the pH is reduced from 7 to 1 and decrease as the pH is increased from 7 to 13. The same study also showed no change in virgin hair mechanical properties between pH 1 and pH 13. To some extent, this contradicts the historical evidence on wool and hair, which was discussed already above, but the differences could possibly be explained by the short duration of treatments.
The present study is part of a larger study that aims to modify environmental treatment conditions in order to improve actives penetration into hair. To achieve this, we aimed at exploring the effects of

| Other materials
HCl (2 N) and 2 N NaOH solutions were obtained from Sigma-Aldrich, and diluted to 0.5 and 0.01 N to adjust pH of solutions.

| pH measurements
pH measurements were performed using a Sper Scientific benchtop pH-meter. The pH meter was calibrated before measurements using commercial pH 4, pH 7 and pH 10 buffers.

| Soaking procedure
All of the soaking solutions were prepared by titrating deionized water with HCL and NaOH. The hair to liquid ratio was 1 g:250 g. Bleached hair was soaked for 24 hours in sealed glass containers to avoid impact of CO 2 or evaporation of HCl while constantly shaking (HY-5 orbital shaker, Zenith Lab Inc., Brea, California) at 25 ± 2 C. After 24 hours, all hair samples were rinsed with deionized water for 30 seconds and air-dried at 60% relative humidity (RH).

| Differential scanning calorimetry
Thermal properties of hair were tested using DSC2500 instrument (TA Instruments, New Castle, Delaware). Three to five replicates were prepared. Prior to the testing, hair was equilibrated at 60 ± 2% RH, 20 ± 2 C for at least 24 hours and the instrument was calibrated using an indium standard. Approximately 10 mg of finely chopped hair was placed in a high-pressure hermetic DSC pan together with 50 μL of DI water. The heating rate was 5 C min −1 .

| Hair diameter and single fiber tensile testing studies
Prior to the testing, hair was equilibrated at 60 ± 2% RH, 20 ± 2 C for at least 24 hours. Fibers were mounted in brass crimps and dimension measurements were performed at 60 ± 2% RH at five locations using a laser scanning micrometer (FDAS770 Fiber Dimensional Analysis System, Diastron Ltd, Andover, UK). Fifty fibers from each group were extended at 40 mm min −1 rate. Stressstrain curves of fibers were recorded at 60 ± 2% RH and in a wet state using a MTT690 Miniature Tensile Tester (Diastron Ltd). Tensile parameters were calculated using the UVWin 3.6 build 3 software (Diastron Ltd, Andover, UK). Young's modulus was calculated by analyzing the linear slope.

| Swelling experiment
This part of the study had two testing groups (n = 30). Group 1-Hair equilibrated first at pH 5 for 24 hours, dried and equilibrated at 60 ± 2% RH, measured, then equilibrated at pH 3 for 24 hours, dried and equilibrated at 60 ± 2% RH and measured again. Group 2-Hair equilibrated first at pH 5 for 24 hours, dried and equilibrated at 60 ± 2% RH, measured, then equilibrated at pH 10 for 24 hours, dried and equilibrated at 60 ± 2% RH and measured again. Dimension measurements were performed the same as described in hair diameter and single fiber tensile testing studies section 2.6: however, eight fiber locations were measured per hair.

| Statistical analysis
The tensile results were statistically analyzed using the JMP 14.0.0 analytical software. The outliers were removed using the Tukey Quartile Method (includes median). The statistically significant differences between results were evaluated using a Student t test at 95% confidence level. The results are presented as means ± the respective SEM.
The fiber swelling results were analyzed performing a two-tailed paired t test (MS Excel, 2016). The paired sample t test is a statistical procedure used to determine whether the mean difference between two sets of observations is zero. The direction of the difference did not matter, thus a two-tailed hypothesis was used for the analysis.
The Pearson correlation coefficient that measures the strength of the linear relationship between two variables was performed using MS Excel PEARSON function (2016). Results suggest that hair absorbs slightly more acid at pH 3 than base at pH 10 if we use the isoionic point as a reference. The isoionic point is the pH range at which the entire fiber's acid-base side chains are at equilibrium (pH 6.2-6.9 from our data). The starting pH of the hair after dialysis, in this study, was the isoionic point. Hair, including bleached hair, contains more carboxylic acid side chains than the basic (lysine and arginine). [6] Additionally, bleached hair contains a significant amount of cysteic acid. [6] The dissociation constants may help explain our findings. pK a of aspartic and glutamic acids in peptides is 4.4 to 4.6. [11] pK b of lysine is 10 to 10.2 [11] and that of arginine higher than 12. [11] The pK a of cysteic acid is 1.3. [12] We propose that more hydrogen ions were absorbed at pH 3 than hydroxide ions at pH 10 F I G U R E 1 Hydrogen and hydroxide ions absorption by bleached hair treated with unbuffered HCl/NaOH solutions. HCl, hydrochloric acid; NaOH, sodium hydroxide because the protonation process of carboxylic side chains and possibly the beginning of cysteic acid protonation consumed more ions than the deprotonation process of lysine side chains.

| Hydrogen and hydroxide ions sorption by bleached hair
After increasing NaOH concentration to similar levels found in relaxing treatments, the absorption increased drastically. Hair at pH 3 absorbed slightly more than twice the amount of hydrogen ions that were absorbed by hair at pH 5. Bhat et al., found that the rate of HCl and NaOH absorption by bleached hair is significantly higher than that of virgin hair; however, the amount of acid that bleached hair absorbs is comparable to virgin hair. [7] They also found that a significant increase in the amount of absorbed hydroxide ions (NaOH) starts at a lower pH and is significantly higher than that of virgin hair. When compared to our results, the absolute values of Bhat et al. [7] differ most likely due to different levels of bleach and the ratio between liquid volume and amount of hair, but the trends of ion absorption are comparable.

| pH effects on thermal properties of hair
The denaturation temperature marks the peak of the denaturation process for the fiber's proteins. It reflects the cross-link density of the hair matrix (amorphous protein) in which the ordered proteins are embedded. [10] A summary of denaturation temperature results is presented in Figure 2.
We observed a clear doublet of denaturation transition while testing pH 3 hair. An example of this is visualized in Figure 3 along with the single denaturation transitions of hair at pH 4, pH 5 and pH 10. Although not visible at pH 10 or pH 5, there is a clear suggestion of the second denaturation temperature peak (labeled with ellipse) at pH 4 which develops further at pH 3.
Denaturation enthalpy values reflect the relative amount and structural integrity of crystalline or ordered proteins (α helix) within a fiber. [10] A summary of denaturation enthalpy results is presented in F I G U R E 3 Examples of DSC traces of hair treated with unbuffered HCl/NaOH solutions. The ellipse labels the suggestion of a higher denaturation temperature peak that is already visible at pH 4. DSC, differential scanning calorimetry; HCl, hydrochloric acid; NaOH, sodium hydroxide F I G U R E 2 Denaturation temperature of hair treated with unbuffered HCl/NaOH solutions. HCl, hydrochloric acid; NaOH, sodium hydroxide In both denaturation temperature and enthalpy graphs, the trends can be represented by three straight lines between the respective pH values of lower than pH 5, between pH 5 and 10 and higher than pH 10 (Figures 2 and 4).

| pH 5 region
pH 5 region will be discussed separately as this is the pH of a typical shampoo and we chose to consider it as a typical condition of hair.
At approximately pH 5, the hair proteins appear to be in the best condition: (a) denaturation temperature of 147.6 ± 0.3 C suggests that matrix cross-linking density is much higher than that of "neutral" hair (denaturation temperature of dialyzed hair was 142.1 ± 0.5 C), or treated with neutral to alkaline treatments and (b) denaturation enthalpy of 13.4 ± 1.4 J g −1 implies that keratin crystallite condition is preserved (denaturation enthalpy of dialyzed hair was 13.6 ± 1.4 J g −1 ).

| 10 > pH > 5 region (green ellipse)
In this region, a linear relationship between denaturation temperature and pH is observed. Denaturation temperature decreased with increasing pH (R 2 = 0.98). A continuous reduction in crosslinking density can be explained via the occurrence of protonation and deprotonation processes close to dissociation constants of amino acid side chains. At pH higher than 5, systematic carboxyl group's deprotonation increased the repulsion between protein chains.
Closer to pH 9, deprotonation of the lysine group possibly began.
As a consequence, salt-bridges, which had previously included lysine, were broken, which was followed by an even further increase in repulsion by available carboxyl groups.
Denaturation enthalpy results revealed that, despite a decrease in matrix stability in this region, the intermediate filament stability seemed to be preserved across the entire pH range (R 2 = 0.076 suggests denaturation enthalpy independence from pH).

| 12 > pH > 10 region (red ellipse)
At pH 10, there was a shift in the steepnes os of the denaturation temperature slope. At pH 10, the basic amino acid, lysine, was neutralized to approximately 50% (according to Henderson-Hasselbalch equation [11] ). The results suggest this is a critical point and further deprotonation of lysine generated a "snowball" effect in repulsion between freed glutamic, aspartic and cysteic acid side chains that resulted in increasing destabilization of the matrix structure (steep slope in this region). There is a possibility that highly concentrated hydroxyl ions caused some disulfide/peptide bond cleavage as well (permanent). However, to confirm this, further research is required.
In the range of pH 10 to 12, a sudden, yet slight increase in denaturation enthalpy is observed between approximately pH 9.8 and 10.
Then, above pH 10, the denaturation enthalpies plateaued at their increased position. It is quite difficult to explain the slight increase in denaturation enthalpy with the matrix being so destabilized. Similar denaturation enthalpy behavior was noticed during unrelated studies performed at TRI when bleached hair was significantly affected by UV: denaturation temperature decreased to approximately 120 C; however, denaturation enthalpy was as high as that of UV untreated bleached hair. This may be influenced by other thermal process(es) that happen from 95 C to 120 C which contribute to the energy associated with protein denaturation.

| 5 > pH ≥ 3 region (yellow ellipse)
In the range of pH 3 to 5, with increasing solution acidity, the single denaturation peak gradually changed into two partially separated peaks with joint denaturation enthalpy decreasing (flattening and widening of denaturation peaks was observable as well). This would suggest that part of the matrix in higher acidity samples appeared to be less dense, while the other part was significantly denser than that of hair at pH 5. We postulate that more intense matrix changes should happen closer to the surface where the hair became more compact. This possibly slowed down further penetration of HCl, which would leave the middle of the cortex less affected. In trying to explain the increasing crosslinking density of the matrix, again we hypothesize that at pH lower than 5 protonation of glutamic and aspartic acids occurred resulting in a reduction of the repulsion between negatively charged groups as well as salt-bridges being replaced by hydrogen bonds. These processes enabled more compact packing within areas of hair represented by the higher denaturation temperature peak. Our data show that the appearance of a second denaturation temperature peak was already visible at pH 4. At pH 3, the matrix still consisted of two peaks, implying that the transition was not completed.
The combined denaturation enthalpies of both peaks are plotted in Figure 3. Deconvolution of these peaks could help to clarify the ratio of peak one and peak two energetics, although, because the enthalpies are so low it may not change the result of parameters correlation. Therefore, we will leave this task for future investigation.
F I G U R E 4 Denaturation enthalpy of hair treated with unbuffered HCl/NaOH solutions. HCl, hydrochloric acid; NaOH, sodium hydroxide Although the matrix cross-linking density increased, it seems that the lower energy of hydrogen bonding made it easier to dismember the matrix, thus, the total energy of denaturation diminished. Additionally, because bleached hair before pH treatments are weakened further by the bleach, acidic peptide hydrolysis is also possible. We expect signs of hydrolysis to be expressed in the wet state of tensile performance where water could help to reveal such effects.
The literature analysis showed very limited research being done on pH effect on hair thermal properties. Istrate et al. concentrated researching on virgin hair property changes and only few acidic pH points were dedicated to the bleached hair. The authors found an increase in denaturation temperature and enthalpy [6] which partially contradicts our findings. The differences in the acidic region could be explained by the following differences in the treatment protocol

| pH effects on hair water sorption properties
All mechanical and dimensional measurements were performed after drying fibers naturally from the wet state, thus, for this analysis we selected the water desorption isotherm. Water desorption results are presented in Figure 5.
From the 60% RH data, it is clear that the differences between pH 3 and pH 5; and pH 5 and pH 10 are rather small. At 95% RH, these differences increased significantly. Interestingly, pH 3 hair absorbed approximately 26% less water than pH 10 hair. The literature is quite limited regarding pH effect on water sorption of hair.
Breauer and Prichard found that at pH lower than 2, the absorption of reducing agents by hair decreased, [13] which agrees with our observations using water as a penetrant. The authors investigated whether abnormalities in the sorption of reducing agents could be caused by peptide or disulfide bond hydrolysis. Amino acid end groups and amino acid analysis did not show evidence that such structural changes would have occurred while virgin hair was soaked at pH lower than 2 solution. [13] However, observed differences in water sorption at different pH can be explained via differences in side chain solvation by water or hydration. The hydration number is defined as the average number of water molecules that are bound stronger to the compound of interest than to other water molecules. The papers that studied hydration of amino acids at different pH suggest the following: (a) hydration number at pH 4 (below pK a ) for aspartic acid and glutamic acid is approximately two [14] to three [15] times lower than at pH 6 to 8 (above pK a ) [16] ; (b) hydration numbers of protonated glutamic and aspartic acids are lower than that of lysine and arginine; [14,15] (c) hydration numbers of charged carboxylic side chains of glutamic and aspartic amino acids are higher than those of lysine and arginine. [15,16] These findings fit our results and help to explain the trends we observed.
Additionally, these also can be used to explain the effect of other acids and bases on hair.

| pH effects on hair swelling properties
Cross-sectional area with the measured minimum and maximum diameter of hair soaked in different pH solutions are presented in Tables 1 and 2, respectively. All measurements were performed at 60% RH.
The average change in fiber dimensions after soaking pH 5 fibers in pH 3 solution is presented in the "Total" row. The difference in cross-sectional area and mean diameter is rather small (−2.9% and −1.6%, respectively). Two-tailed paired t test was performed to evaluate whether there is a difference, on average, in the fiber dimensions between "before" pH 5 and "after" pH 3. The null hypothesis (H 0 ) assumed that the true mean difference is equal to zero. The two-tailed alternative hypothesis (H 1 ) assumed that true mean difference is not equal to zero. The calculation revealed that P = .002.
Consequently, the decrease in fiber dimensions is significant at the α = 0.01 level.
To observe if the initial fiber thickness was important to the swelling of the hair, we sorted the fibers according to their crosssectional area in three groups. Grouped by size of cross-sectional area, results ( Table 1) Table 2).
The average change in fiber dimensions after soaking pH 5 fibers in pH 10 solution is presented in the "Total" row. The difference in cross-sectional area and mean diameter is rather small (1.3% and 1.4% respectively). Two-tailed paired t test was performed to evaluate whether there is a difference, on average, in the fiber dimensions between "before" pH 5 and "after" pH 10. The calculation revealed that P = .00003. The increase in fiber dimensions is significant at the α = 0.01.
Thinner hair swelled more than thicker hair. The cross-sectional area of the finest hair (2525 μm 2 ) increased by little over 2%, while the cross-sectional area of both thicker hair groups increased by 1%.
Besides general knowledge that hair swells above the isoelectric point and allows reducing agents/dye/bleach penetration into hair, [17] we could not find published data regarding bleached hair swelling due to exposure to different pH solutions. The reduction in fiber diameter at pH 3 and its increase at pH 10, in comparison to pH 5 hair, was similar to that of virgin hair reported previously. [18] 3.5 | pH effects on mechanical properties of hair The literature suggests that DSC and tensile parameters correlated well while studying bleached and thermal damage levels. [19] These treatments permanently changed the hair structure. pH treatment, excluding extreme values, cosmetically modified hair proteins (recoverable changes), thus, it is useful to understand if (a) large observed changes in hair matrix crosslinking and (b) integrity of IF across the pH range, can translate into significant changes in the mechanical properties of hair. Due to tensile experiments being labor intensive and costly, we decided not to continue with the property investigation across the entire pH range, but to select consumer relevant pH representatives from green, yellow and red ellipse zones. pH 3 corresponds to acidic treatments such as lemon juice, vinegars or certain patented hair care formulations. [20] pH 5 is a conventional pH for most hair care products (for this reason, pH 5 hair will be treated as the baseline). pH 10 is a common pH in hair coloring systems.

| Dry tensile test results
Dry Young's modulus, turnover point and its stress, post-yield extension, break stress and break extension data are presented in Figure 6.
The majority of models explaining hair stress/strain behavior agree that, in the linear region, the stretching of chemical bonds takes place. [4,21,22] This means that, depending on the dominance of particular bonds, we may observe different behavior. By controlling the denaturation temperature of the samples prior to pH treatment, we are safe to assume that the level of disulfide bonds is similar among different treatment groups. Therefore, by varying environmental pH, we aim to evaluate the importance of the ratio between salt-bridges and hydrogen bonds, and repulsion between free negatively charged groups on the mechanical properties of bleached hair. The highest Young's modulus was obtained when measuring pH 5 hair, followed by pH 10, then pH 3. These differences, at the 60% RH, were found T A B L E 1 Dimensional parameters' summary of fibers that were soaked in pH 5 and pH 3 solutions  virgin hair. [7] Similar to our results, they found that the dry state elastic modulus decreased at both pH extremes. However, a larger decrease in modulus was found in alkaline treated hair. [7] The differences could possibly be attributed to differences in initial hair condition and their pH range was one pH unit more extreme.
Results suggest that pH affected turnover point stress and strain of bleached hair fibers as well. Both parameters of pH 3 soaked hair were the lowest, and those of pH 10 treated hair were the highest.
pK a of the thiol group is 8.5 to 8.8. [11] We anticipate that further research of thiol/disulfide interchange (happens above pK a ) under these pH could help explain the observed differences in these results. The analysis revealed no pH impact on dry hair break stress.

| Wet tensile test results
Chemically damaged hair was especially weakened by water. Plasticization by water helps to amplify the level of damage, which would be hardly measurable in dry conditions. By completing this experiment, we hoped to not only assess the IF condition but also to investigate if potential peptide bond hydrolysis is measurable.
While comparing wet and dry tensile profiles, we found a 4-fold reduction in stresses that was required to reach the turnover point in a wet state, and the break-stress in the wet state was approximately 35% lower than that in a dry state. A truly linear region was found at less than 1% strain extension. This suggests that the matrix contribution to the hair's mechanical properties diminished.
Wet Young's modulus, break stress and break extension data are presented in Figure 7.
Clearly, hair equilibrated at pH 5 shared similarities with hair that was equilibrated at pH 3 or pH 10 due to water breaking the majority of salt-bridges and hydrogen bonds that were responsible for differ-

ACKNOWLEDGMENT
We wish to acknowledge the expert technical assistance by Katerin Mateo of TRI Princeton (DSC and tensile data collection). This work was supported by funding from Lonza and TRI Princeton.

CONFLICT OF INTEREST
The authors declare no competing interests. Lonza and TRI jointly agreed the design of the study; in the collection, analyses, and interpretation of the data, in writing and reviewing of the manuscript, and in the decision to publish the results.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in TRI Library at https://library.triprinceton.org/1erqg3t/ (bottom of the page), reference ID 1erqg3t.
T A B L E 3 Dimensional parameters' summary of fibers that were soaked in pH 5 and pH 10 solutions