microscopy as a more sensitive technique to investigate small differ- ences in these actives penetrating into hair. Penetration of labelled keratin peptides into hair fibres Figures 2, 3 and 4 show the sections obtained from relaxed control hair and relaxed hair treated with labelled keratin peptides for 3 h. The images taken using the green channel show natural autofluo- rescence of the hair and help display the positioning of the cross- sections in the images. The images of the same samples taken using the red channel show only the fluorescence from the labelled hydrolysed keratin. It is interesting to note that, as these were tex- tured hair samples, the cross-sections were especially flat and ellip- tical. This is characteristic for this hair type. There was no presence of fluorescence in the red channel with relaxed hair which had been soaked in filtered dye solution for 3 h (Fig. 1), indicating that filtration using the centrifugal filters removes any free dye. Fig. 2 shows that after 3 h of soaking, the mid-range MW keratin peptides had penetrated deep into the cortex. However, after 3 h soaking, the high-MW peptide had only just entered the outer layers of the cortex (Fig. 3). Pre-existing mechanical damage of hair via SEM Literature suggests that textured hair contains higher levels of mechanical damage as a result of intense grooming regimens [21]. The preliminary evaluation of hair surface condition was performed using SEM. Examples of these cracks in virgin and relaxed hair controls are presented in Fig. 5. All selected fibres were thoroughly investigated from the top of the stud to the bottom. The screening revealed that some fibres contained long longitudinal cracks in cer- tain areas of the cortex: five fibres in virgin group and six fibres in relaxed hair group contained the aforementioned flaws. Some of the fibres had more than one longitudinal crack. Some cracks were widely open (659 magnification), whereas others showed clear signs of cracking however, the cracks’ depth was unknown (5009 magnification). Because of a relatively small sample size (15 fibres per group), we consider these results preliminary and further research is required in order to study the relationship between lon- gitudinal cracks and mechanical properties of hair. Surface coverage by mid- and high-MW peptide treatments via SEM Figure 6 presents SEM images of randomly chosen fibres from the relaxed hair control, mid- and high-MW peptide treatment groups. Only the presence of high-MW peptide film is visible at 10 0009 magnification. This suggests that mid-MW peptide size is not suffi- cient to form relatively thick film. Effects of peptide treatments on hair stiffness Figure 7 summarizes Young’s moduli data obtained at 20% and 80% RH. Dunnett’s test showed that, at 20% RH, Young’s moduli of hair treated with mid- and high-MW peptides were significantly higher than those of relaxed hair control. Young’s moduli of vir- gin hair and relaxed hair treated with low-MW peptide actives were found to be no different than the relaxed hair control. Young’s moduli of low-MW keratin peptides and leucine-treated relaxed hair were significantly lower than those of virgin hair control. An increase in humidity caused a major shift in elastic properties of the hair samples. This was especially visible when comparing virgin with relaxed hair. At 80% RH, Young’s modulus of virgin hair was significantly higher than that of differently treated relaxed hair (14.0% higher than relaxed hair control). When comparing Figure 4 Cross-sections of Afro-textured relaxed hair soaked in labelled high-MW keratin peptides for 3 h at low (209, top) and high (409, bottom) magnifi- cation. © 2020 The Authors. International Journal of Cosmetic Science published by John Wiley & Sons Ltd on behalf of Society of Cosmetic Scientists and Societe Francaise de Cosmetologie 5 International Journal of Cosmetic Science, 1–12 Larger hydrolyzed keratins reduce hair breakage E. Malinauskyte et al.