J. $oc. Cosmet. Chem. 27, 155-161 (1976) ¸ 1976 Society of Cosmetic Chemists of Great Britain
A study of damaged hair
V. N. E. ROBINSON*
Synopsis--A SCANNING ELECTRON MICROSCOPE study of the morphological changes
in human HAIR has revealed that these changes may be entirely attributable to the ABRASION
normally associated with hair grooming. In particular, SHAMPOOING and BRUSHING
can produce all of the observed changes detected in virgin hair. These changes are more
deleterious in some cosmetically altered hairs. This appears to be attributable to a decrease in
the disulphide bond content of these fibres, this decrease being brought about by the cosmetic
In a study of fine changes in the surface architecture of human hair due to
cosmetic treatment, Swift and Brown (1) have illustrated the stages of breakdown
of human hair during cosmetic treatments. The natural appearance of hair, when
first formed, displays a scale structure with smooth edges. This is subsequently
transformed into stages they have characterized by: jagged scale edges, partial
scale erosion, complete scale erosion, fibre splitting into two roughly hemi-
cylindrical components and gross longitudinal splitting into many fibrous
elements. These changes are generally referred to as weathering and the pre-
dominant cause is thought to be exposure to sunlight (2), which may result in a
variation of the chemical and physical properties from the root to the tip of the
fibres (3, 4).
By means of before and after treatments and the examination of the same hair
fibres in the SEM, Swift and Brown (1) have established that some of these
* School of Textile Technology, University of New South Wales, P.O. Box 1, Kensington,
NSW 2033, Australia.
156 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
variations are due to combing of the hair and concluded that some may be due
to natural weathering processes. They have also briefly mentioned some of the
changes introduced into hair by perming and bleaching treatments. In a series
of dynamic in situ experiments in a scanning electron microscope, Brown and
Swift (2) have further demonstrated the deleterious effects of combing out tangles
in hair. Such combing can cause cuticle cell lifting in tightly-looped fibres, as well
as snapping fibres transversely or through the initial formation of a longitudinal
fracture, when excess tension is applied during the combing.
In a study of the structure and properties of normal adult hair, Wall and
Hunter (5) have similarly illustrated the production of jagged edges in the cuticle,
and have shown features on the cuticle that they have attributed to sun and
atmosphere exposure. DiBianca (6) has demonstrated many different types of
damaged hair ends, but has not studied how these different types of damage
occurred. Robbins and Kelly (7) have analysed the amino acid content of cos-
metically-altered hair. They found that bleached and permanent-waved hair
contained less cysfine than unaltered hair, and correspondingly more cysteic acid
residues. Bleached hair also contained slightly less tyrosine and methionine than
unbleached hair. Miyazawa, Nozaki and Tamura (8) have made a similar study of
the amino acid composition of hair damaged by treatment with cold-waving and
hair bleaching lotions. They have also observed a reduction in the cysfine content
of human hair, with one bleaching condition reducing the cystinc content from
the normal of 13.9•o to 5.0•o, corresponding to an approximately 64•o reduction
in the disulphide bond cross-linking of the protein molecules.
These studies have illustrated the extent of chemical changes and the mag-
nitude of physical changes introduced to hair, but have not yet confirmed how
these observed physical changes were introduced. There is a need to understand
how and why these deleterious physical changes to hair fibres occur in order to be
able to prevent the damage. Some of this information is presented in this study.
Samples of Caucasian hair were used in this study. This hair included many
fibres displaying various degrees of splitting and hair damage, and samples that
displayed no damage visible to the unaided eye. No attempt was made to differen-
tiate between hair of different texture or condition.
The fibres were mounted on stubs so that the tip, root and portions of the
mid shaft of each fibre could be examined in a scanning electron microscope
(JSM2). The mounted fibres were coated with approximately 500 3, of gold prior
to examination. They were examined for variation of scale structure along and
between fibres, extent of scale damage and removal and degree of splitting of
ends. The findings were correlated with the known history of the hair, and
A STUDY OF DAMAGED HAIR 157
attempts were made to ascertain what caused the observed changes by reproducing
the damage in controlled laboratory experiments.
The study was divided into two sections: the examination of virgin hair, that
is, hair that had not been dyed, bleached, permed or otherwise treated, and
cosmetically altered hair, hair that had received one or more of these types of
In common with the findings of Wall and Hunter (5) this study has shown that
there was very little difference in fibre appearance between fibres from different
people. The great variation detected was along fibres from root to tip and this
variation was similar for all the fibres studied. This variation is reported below.
Fig. 1 shows a micrograph of a hair fibre, taken near the root, indicating the
natural appearance of hair that is free from externally promoted defects (1). A
few millimetres from the root the scale edges became jagged. For most of the
fibres examined this jagged scale edge appearance, Fig. 2, represents the typical
appearance of most of the fibre length.
Previous researchers (1, 2) have concluded that some of this change is due to
the mechanical damage caused by brushing, combing and handling, and that
some contribution to this deterioration may be due to weathering by exposure to
rain, sunlight and dirt. Some of the fibres examined in this study had been
shampooed and towel-dried three or four times per week, given minimal combing,
approximately five comb strokes per day, were hardly ever exposed to sun or rain
and had never been brushed. This abrasion of the scale edges still occurred in this
hair, see Fig. 2. It seems probable that this deterioration of scale edge appearance
was, in this case, due almost entirely to the wet abrasion associated with sham-
pooing and towel drying.
To check the effect of wet abrasion on the cuticle, two experiments were per-
formed. In one study, a child's hair was lightly shampooed, approximately once
every week to minimize wet abrasion, and given minimal brushing and combing.
The majority of the fibres examined from this child displayed a fibre appearance
similar to that shown in Fig. 1. In the second experiment, a group of fibres from
the above study were wet and rubbed vigorously between the hands to simulate
shampooing and towel-drying. It was found that this action produced deteriora-
tion similar to that displayed in Fig. 2. From these observations it was concluded
that the wet abrasion associated with shampooing and towel-drying was a
dominant factor in the deterioration of scale edge appearance, at least amongst
the persons studied.
158 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
With the exception of hair that had only been lightly shampooed, all other
hair studied always displayed this deterioration of scale edges. However, some
hair fibres displayed damage beyond that shown above. As reported by Swift and
Brown (1), the next stages of fibre damage that were observed were partial,
followed by complete scale removal, see Fig. 3. Again, by rubbing wet hair fibres
in a manner that simulated shampooing, it was found that it was possible to
remove completely the scale structure from the fibres. Similar rubbing of dry
fibres did not produce the same degree of scale damage. Thus it seems probable
that shampooing is a contributing factor to complete scale removal.
At this stage no assessment has been made of the role of hair brushing and
combing in the damage to and removal of scales, although it seems probable (2)
that these actions could also produce deterioration in the cuticle.
Amongst the hair fibres studied, it was found that complete scale removal
only occurred in the last few centimetres. Once complete scale removal had
occurred splitting of the fibre end into two, three or more longitudinal sections
usually occ•trred, see Fig. 4.
Attempts were made to reproduce this type of damage. It was found that
brushing a group of fibres that had been subjected to simulated shampooing
caused the ends of some of the fibres to split, indicating that shampooing and
brushing alone is sufficient to cause the ends of hair fibres to split. At this stage,
no attempt has been made to ascertain which of these two actions is most
responsible for the production of splitting, or what other procedures can cause
It was noted during the brushing trials that when a fibre had split longi-
tudinally, it was common for the individual splits to fibrillate, see Fig. 5. Continued
brushing caused these fibrillated regions to break, Fig. 6, thus producing a
fibrillated end fibre.
The appearance of split ends could vary quite considerably from the simplified
breakdown pattern shown in Figs 4, 5 and 6. Some of these variations have been
shown by Swift and Brown (1) and DiBianca (6). There is a shortening of fibres
associated with the rupture of these fibrillated ends and this shortening appears
to keep this type of damage restricted to the last few centimetres of a fibre shaft.
These observations have illustrated that shampooing and brushing alone are
sufficient to transform the hair from its natural state, Fig. 1, to split ends, Figs.
5 and 6. It seems probable that these grooming treatments of shampooing and
brushing are the dominant factors in the physical deterioration of virgin hair.
Weathering and exposure to sunlight would appear to be minor effects. If a sample
of hair is just stored for hundreds of years, this type of damage does not occur (1),
indicating that the deterioration is not an ageing process.
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Figure 1. Typical appearance of human hair within a few millimetres of
root. x 640.
•' •"5 '• • • -• •-
Figure 2. Typical appearance of human hair over most of the mid-shaft. •e
deterioration of scale edge appearance can be caused by shampooing alone.
X 64O. (Facing p. 158)
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Figure 3. Complete cuticle removal near the ends of long hair. x 640.
Figure 4. A longitudinal split near the end of a fibre. x 480.
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Figure 5. A hemicylindrically split fibre segment, fibrillating at cortical cell
boundaries. x 480.
Figure 6. A fibrillated end fibre caused by the rupture of a fibrillating region.
JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Figure 7. Fibre splitting in permanently-waved hair. X 640.
Figure 8. Shaft damage in permanently-waved hair. This damage was
detected about 10 cm from the end of the fibre. X 400.
A STUDY OF DAMAGED HAIR 159
Cosmetically-altered hair had an appearance very similar to virgin hair, when
no split ends were detected. The hair appearance shown in Figs 1-6 for virgin
hair, can also be seen in cosmetically-altered hair. However, there are also other
types of breakdown associated only with cosmetically-altered hair.
Fig. 7 shows a split end from hair that has had one permanent-wave applica-
tion. Unlike virgin hair, where splitting was only observed after almost all of the
cuticle was removed, hair that had been permed or bleached was often observed to
split whilst the cuticle was still clearly visible.
It was observed that splits and general shaft damage could also occur well
away from the end. Fig. 8 shows shaft damage detected about 10 cm from the end
of a fibre that had been permed once and the perm did not 'take'.
The third important factor of cosmetically-altered hair was the extent of
splitting. Virgin hair less than 15 cm long was hardly ever observed to split, whilst
cosmetically-altered hair less than 15 cm long was often observed to split. This
was detected as most likely to occur when a perm 'went wrong' or did not 'take',
and on some bleached hair. Also, the splits and shaft damage associated with hair
that had been permed or bleached, was generally observed to be far more severe
than for virgin hair. That is, cosmetically-altered hair was observed to contain
far more of the hair categorized as fly-away fibre (6) than virgin hair, for hair
fibres of the same length.
The reason for this additional deterioration associated with cosmetically-
altered hair is not immediately apparent. Swift and Brown (1) have shown that
very minor changes in the surface structure of hair occur immediately after
perming and bleaching, but these are not likely to be directly responsible for the
observed additional damage. It is probable that these processes have chemically
altered and weakened the hair and then during subsequent physical treatments,
the altered hair is unable to withstand the abrasion normally associated with hair
grooming to the same extent as can the virgin hair, and a greater degree of fibre
Robbins and Kelly (7) and Miyazawa et al. (8) have found a reduction in the
cystine content of bleached and permanent-waved hair. It is primarily the cystine
content of keratin fibres, through the cross-linking associated with the disulphide
bonds, that gives these protein fibres their high mechanical stability. It seems
probable that the reduced fibre cross-link content associated with these treatments
is responsible for loss of physical strength associated with these fibres.
Feughelman and Chapman (9) have shown that the relative cross-link density
of keratin fibres can be ascertained by a determination of the diametral swelling
of the fibres in 98• formic acid. This provides a convenient method for the
determination of a relationship between fibre cross-link and splitting caused by
the mechanical abrasion associated with hair grooming.
160 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS
Virgin hair, when immersed in formic acid, swelled diametrally approximately
35•o above the diameter of the dry fibre. Cosmetically-altered hair samples
swelled to different amounts. Some fibres swelled the same as or only a little more
than virgin hair, approximately 35-405/o above the diameter of dry fibres. Other
fibres were observed to swell to over 100•o more than their dry diameter, indi-
cating (9) an approximate 50•o loss of disulphide cross-link content. Miyazawa
et al. (8) have also observed similar reductions in the cross-link content of
As a general observation, it was noticed that the fibres displaying greatest
diametral swelling in formic acid also displayed the greatest tendency to split. No
direct correlation was obtained between diametral swelling in formic acid and
degree of splitting because it appears that the degree of splitting depends upon
grooming conditions as well as disulphide bond content, and therefore no direct
correlation could be expected to exist. It was also noticed that persons with hair
that swelled more in formic acid were less satisfied with the appearance and
condition of their hair. It seems likely that the cross-link content of the hair is at
least partially responsible for the condition and manageability of the hair. That is,
a reduction of the cross-link content of the hair, which may be introduced by
some cosmetic treatments, results in a reduction of the ability of fibres to with-
stand the abrasive forces normally associated with hair grooming, causing an
increase in hair damage during grooming, and appears to result in a loss of
manageability of the hair.
The results presented above indicate that the structural changes observed in
hair fibres can be due entirely to the mechanical abrasion associated with normal
hair grooming. This study has demonstrated that shampooing of virgin hair is
sufficient in itself to damage and completely remove the cuticle. it is, of course,
probable that some other treatments may also produce the same effect, but these
have not yet been exhaustively investigated. It also appears that brushing of the
hair contributes to damage to the cortex, and is at least partially responsible for
the production of split ends in hair. No attempt has yet been made to determine
the relative roles played by all of the hair grooming processes in breakdown of
Some cosmetically-altered hair is more prone to mechanical breakdown during
grooming than virgin hair. This hair exhibits a greater tendency to swell in formic
acid, indicating a loss of cross-link content resulting from the cosmetic treatment.
It appears that the cosmetic treatment has reduced the cross-link content of the
hair and that this has rendered the hair less able to withstand the mechanical
abrasion associated with hair grooming, resulting in earlier splitting of the fibres.
A STUDY OF DAMAGED HAIR 161
Hydrogen bonds represent another useful form of cross-linking. Hydrogen
bonds are reversibly broken every time hair is wet and are reformed again when
it dries. Because of the decrease in hydrogen bonding, wet hair is completely
unmanageable. Observations made during this study have indicated that wet hair is
far more susceptible to damage during grooming than dry hair, and it would thus
appear that minimal handling and brushing of wet hair could reduce the extent of
physical damage to the hair.
This study has shown that the mechanical deterioration of hair which results
in the production of split ends can be entirely attributable to the abrasion
associated with hair grooming. It has also indicated that the increased tendency of
hair to split after cosmetic treatments, particularly perming and bleaching,
appears to be due to a reduction in the cross-link content of the fibres.
To prevent this mechanical damage minimum brushing and shampooing
would appear to be desirable. Prevention of damage to the disulphide bond
content of the hair fibres would also appear to be desirable to reduce splitting
and improve the manageability of hair. There would also appear to be some
advantages in increasing the cross-link content of hair by the introduction of
cross-linking molecules. The additional cross-linking would retard fibre swelling
when wet and possibly increase the wet abrasion resistance of the hair. It is also
possible that such a treatment would improve the manageability and condition of
the hair as well as being able to impart a permanent change of shape to the fibres.
Studies in these directions are continuing. (Received: 24th February 1975)
(1) Swift, J. A. and Brown, A. C. The critical determination of fine changes in the surface
architecture of human hair due to cosmetic treatment. J. Soc. Cosmet. Chern. 23 695 (1972).
(2) Brown, A. C. and Swift, J. A. Hair breakage: the scanning electron microscope as a diag-
nostic tool. J. Soc. Cosmet. Chern. 26 289 (1975).
(3) Robbins, C. Weathering in human hair. Text. Res. J. 37 337 (1967).
4) Robbins, C. R. and Kelly, C. H. Amino acid composition of human hair. Text. Res. J. 40
(5) Wall, R. A. and Hunter, L. D. Normal adult hair--structure and properties. Cosmet.
Perrum. 893 (1974).
(6) DiBianca, S. P. Innovative scanning electron microscopic techniques for evaluating hair
care products. J. Soc. Cosmet. Chern. 24 609 (1973).
(7) Robbins, C. R. and Kelly, C. H. Amino acid analysis of cosmetically-altered hair. J. Soc.
Cosmet. Chern. 20 555 (1969).
(8) Miyazawa, F., Nozaki, F. and Tamura, T. Studies on damaged hair. Proceedings 8th Inter-
national Congress IFSCC, London. C8 (1974) (Blackwell Scientific Publications, Oxford).
(9) Feughelman, M. and Chapman, B. M. The swelling of wool fibres with reduced disulphide
content in 98 % formic acid. Text Res. J. 36 1110 (1966).