Journal of Pharmaceutical and Biomedical Analysis 88 (2014) 509512

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis
j our nal homepage: www. el sevi er . com/ l ocat e/ j pba
Short communication
Determination of thioglycolic acid in cosmetics by
capillary electrophoresis
Na Xie
a,b
, Xiaojing Ding
a,c,
, Xinyu Wang
a
, Ping Wang
a
, Shan Zhao
a
, Zhi Wang
b,
a
Beijing Key Laboratory of Diagnostic and Traceability Technologies for Food Poisoning, Beijing Center for Disease Control and Prevention,
Beijing 100013, China
b
College of Science, Agricultural University of Hebei, Baoding 071001, China
c
Department of Public Health, Capital Medical University, Beijing 100069, China
a r t i c l e i n f o
Article history:
Received 5 July 2013
Received in revised form
30 September 2013
Accepted 2 October 2013
Available online 14 October 2013
Keywords:
Thioglycolic acid
Capillary electrophoresis
Cosmetics
Depilatory creams
Hair-treatment products
a b s t r a c t
A new and simple method for the accurate determination of thioglycolic acid (TGA) in cosmetics was
developed using capillary electrophoresis (CE) with diode array detection at 236nm. The CE separa-
tion was performed on an uncoated fused silica capillary with a separation buffer solution containing
300mmol L
1
tri-sodium phosphate and 0.5mmol L
1
cetyltrimethylammonium bromide at a voltage of
5kV. Both the intra- and inter-day precisions of the method were 1.4%. The calibration curve between
the corrected peak areas and the concentrations of the TGA was linear within the concentration range of
0.0061.0mgmL
1
with a correlation coefcient (r) of 0.9998. The limit of detection and limit of quan-
titation were 0.002mgmL
1
(S/N=3) and 0.006mgmL
1
(S/N=10), respectively. The average recoveries
at the spiked levels of 0.125, 0.250 and 0.500mgmL
1
were 96.9%, 102.3% and 94.0% with the rela-
tive standard derivations of 2.1%, 3.9% and 2.2%, respectively. The method was cross-validated by both
high performance liquid chromatographic and ion chromatographic method. Eighty-ve commercial
depilatory creams and hair-treatment products were analyzed with satisfactory results.
Crown Copyright 2013 Published by Elsevier B.V. All rights reserved.
1. Introduction
Thioglycolic acid (TGA), thioglycolates and thiol esters have
widespread applications in cosmetics, especially in hair-treatment
products and depilatory creams due to their reducing character-
istics. TGA can diffuse into and react with the hair keratin bers
and give rise to the disconnection of the disulde ( SS ) groups
[1,2]. However, the exposure to TGA through cosmetic products
has a potential adverse effect on human health [3], especially due
to its reproductive toxicity [4,5]. The Ministry of Public Health of
China has set four safety levels for TGA, namely 8% for home-use
hair-treatment products, 11% for professional-use hair-treatment
products, 5% for depilatory creams, and 2% for rinse-off hair care
products [6]. Therefore, straightforward and effective analytical
methods for the accurate determination of TGA are needed for
supervising purpose.

Corresponding author at: Beijing Center for Disease Control and Prevention,
Beijing 100013, China. Tel.: +86 10 64407087; fax: +86 10 64407210.

Corresponding author. College of Science, Agricultural University of Hebei, Baod-

ing 071001, China. Tel.: +86 312 7521513; fax: +86 312 7521513.
E-mail addresses: dingxiaojing@gmail.com, dxj wry@yahoo.com (X. Ding),
wangzhi@hebau.edu.cn (Z. Wang).
At present, the mainly used analytical techniques include ion
chromatography (IC) [6,7], chemical titration[6], highperformance
liquid chromatography (HPLC) [810], differential pulse voltam-
metry [11] and spectrophotometry [12,13]. All of these methods
have some limitations such as equipment cost, complex sample
pretreatment or sample matrix interferences. Interestingly, cap-
illary electrophoresis (CE) has shown tremendous potentials in
cosmetics analysis [14,15]. To the best of our knowledge, there has
been no report yet on the use of CE for the determination of TGA in
cosmetics.
The aim of the current study was to develop a new and effec-
tive CE method for the determination of TGA in cosmetics. The new
method was cross-validated by both an established IC method [6]
proposed by the Ministry of Public Health of China and a HPLC
method issued by Chinas Food and Drug Administration [10].
Eighty-ve commercially available depilatory creams and hair-
treatment products were analyzed.
2. Experimental
2.1. Reagents and materials
All chemicals used were of analytical grade or higher. The ultra-
pure water used throughout the experiment was prepared by a
Millipore Milli-QRG ultra-pure water system (Bedford, MA, USA).
0731-7085/$ see front matter. Crown Copyright 2013 Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jpba.2013.10.003
510 N. Xie et al. / Journal of Pharmaceutical and Biomedical Analysis 88 (2014) 509512
Tri-sodium phosphate dodecahydrate (Na
3
PO
4
12H
2
O, 98%) was
bought from Beijing Chemical Reagent Co. (Beijing, China). Dode-
cyltrimethylammoniumbromide (DTAB, 99%) was purchased from
J.K. Scientic Ltd. (Beijing, China). Tetradecyl trimethyl ammonium
chloride (TTAC, >98%) was obtained from Tokyo Chemical Industry
Co. Ltd. (Tokyo, Japan). Cetyltrimethylammonium bromide (CTAB,
99%) andcalciumthioglycolate trihydrate (98%) were purchased
from SigmaAldrich, (St. Louis, MO, USA). Eighty-ve samples
including fty permanent wavings, 16 hair straighteners and 19
depilatory creams were bought from the local market.
2.2. Apparatus and conditions
All CE experiments were performed on a P/ACE MDQ sys-
tem (Beckman Instruments, Fullerton, CA, USA) equipped with
an auto-sampler and a photodiode array (PDA) detector. Data
acquisition and treatment was controlled using 32 Karat soft-
ware, version 8.0 (Beckman Coulter, Brea, CA, USA). The running
buffer was 300mmol L
1
Na
3
PO
4
(without pH adjustment) con-
taining 0.5mmol L
1
CTAB. A 1:10 dilution of the running buffer
was used as the sample buffer. An uncoated fused-silica capillary
(Yongnian Ruifeng Sepu Peijian Plant, Heibei Province, China) of
50mi.d. 40.2cm (375mo.d., effective length to the detection
window was 30cm) was used. The new capillary was conditioned
by successively ushing with 1mol L
1
NaOH at 20psi for 20min,
ultrapure water for 5min, and the running buffer for 5min. To
ensure the repeatability, before each subsequent run, the capillary
was rinsed with 1mol L
1
NaOH, ultrapure water, and the running
buffer for 3, 2, and 2min, respectively. The separation was car-
ried out at a constant voltage of 5kV (current about 124A).
The detection wavelength was set at 236nm. The sample was
introduced into the capillary by hydrodynamic injection at 0.5psi
for 10s. All electrophoresis runs were performed at a temperature
of 25

C.
2.3. Preparation of calibration curve
A TGA standard stock solution (10mgmL
1
) was prepared daily
by accurately weighing 20mg of calcium thioglycolate trihydrate
into a 1.5mL of sample vial. 1mL of water was added and then
vortex-mixed. A series of standard solutions at the nal concentra-
tions of 0.006, 0.025, 0.125, 0.250, 0.500 and 1.000mgmL
1
were
made and kept in dark to avoid possible photodissociations [16].
2.4. Sample pretreatment
An appropriate amount of samples were weighed into a 15mL
centrifuge tube with a cap and then diluted with sample buffer
to the mark of 10mL. The solutions were thoroughly vortexed to
make the samples homogeneous. The mixed water-based solutions
could be injected directly, while the mixed oil-based samples must
be centrifuged at 9000rpm for 10min and then the resultant clear
supernatant was directly injected.
3. Results and discussion
3.1. Optimization of CE conditions
3.1.1. Selection of the running buffer
TGA is a weak organic acid with two pKa values, pKa
1
=4.32 and
pKa
2
=10.20 which comes fromthe dissociation of its carboxyl and
thiol groups, respectively [17]. As a rule of thumb, the pHof the run-
ning buffer should be adjusted to at least two units belowthe pKa
1
or above the pKa
2
of TGA to ensure its complete ionization [18].
Phosphate, with low UV absorption and wide pH buffer capacity
Table 1
Comparison of the results of the TGA in 16 cosmetic samples determined by CE,
HPLC and IC.
Sample CE HPLC IC
Content (%, w/v) Content (%, w/v) Content (%, w/v)
1 / /
2 / /
3 0.7
4 / / /
5 / /
6 4.1 4.0 4.3
7 1.3
8 1.1
9 8.7 8.4 8.2
10 8.1 7.7 8.0
11
a
2.7 2.6 2.8
12
a
3.4 3.3 2.9
13
a
2.3 2.4 2.3
14
a
3.7 4.0 3.7
15
a
2.7 2.7 2.6
16
a
3.4 3.4 3.4
/, Not detected; , cannot be quantied due to interferences.
a
Oil-based sample.
(pKa
1
=2.12, pKa
2
=7.20 and pKa
3
=12.36), is a popular buffer sys-
tem for CE. Considering that TGA salts at pH 7.09.5 or 12.7 and
TGA esters at pH6.09.5 are usually used in cosmetic formulations
[6] and only phosphate can provide a good buffer capacity at pH
higher than (pKa
2
(of TGA) +2), phosphate was chosen as a priority
running buffer for the subsequent optimizations.
3.1.2. Optimization of the concentration of Na
3
PO
4
and the pH of
the separation buffer
Although the analysis of only a single analyte TGA in cosmetic
samples was studied in this work, however, in such a case, the
researchers were often dismayed by the interferences coming from
the sample matrix. In the present work, a real sample 12 (listed in
Table 1) was used to optimize the following conditions of the CE
method.
The concentration of Na
3
PO
4
played a key role for the sep-
aration of TGA. Keeping the concentration of CTAB (0.5mmol/L)
unchanged, the concentration of Na
3
PO
4
(without pH adjustment)
increasing from100mmol L
1
to 350mmol L
1
, not only decreased
the EOF, but also suppressed the adsorption of TGA to the capillary
wall. As shown in Fig. 1, the detection sensitivity was enhanced and
the separation efciency was increased. However, high concentra-
tions of phosphate would produce large amount of Joule heat due
to its inherently high conductivity, which could make the TGApeak
broadened. As a result, 300mmol L
1
Na
3
PO
4
was chosen.
The pHof running buffer is another key factor for the separation
of TGA from the cosmetic sample matrix. It controls both the TGA
charge state and the level of EOF. Since the running buffer with pH
close to the pKa
3
value of H
3
PO
4
(12.36) can provide better buffer-
ing capacity [19], the effect of the running buffer at pH12.35, 12.60
(without pH adjustment) and 12.85 were investigated while the
concentration of 300mmol L
1
Na
3
PO
4
was kept constant. It was
found that under all the three pHvalues, the TGA could be baseline
separated from the interfering peaks coming from sample matrix.
Considering that the running buffer at pH12.60 could be easily pre-
paredandhadahighbuffer capacity, 300mmol L
1
Na
3
PO
4
without
pH adjustment was chosen for the following experiments.
3.1.3. The selection of the EOF-reversing agent and its
concentration
Samples are normally injected at the anode and detected at the
cathode. At pH>7, the EOF is sufcient to ensure most species
to move from the anode to the cathode in the order of cations,
N. Xie et al. / Journal of Pharmaceutical and Biomedical Analysis 88 (2014) 509512 511
Fig. 1. Effect of Na
3
PO
4
concentration on separation. Na
3
PO
4
concentration
(mmol L
1
): (a) 100, (b) 200, (c) 300, and (d) 350. (1) TGA and (2) unknown peak.
Running buffer: x mmol L
1
Na
3
PO
4
and 0.5mmol L
1
CTAB. Uncoated capillary:
40.2cm50m, i.d. (30cmto the detector). Injectiontime: 10s. Injectionpressure:
0.5psi. Separation voltage: 5kV, detection: UV 236nm, temperature: 25

C.
neutrals and anions. However, the negatively charged TGA
migrated against the EOF and moved to the anode under the opti-
mized running buffer pH 12.60, which made it difcult to migrate
out of the capillary in a reasonable time (within 30min). To ensure
that the TGA could rapidly move to the detection window, an
EOF-reversing agent was needed [20]. Long-chain alkyl trimethyl
quaternary salts are frequently used as EOF modiers for anionic
analysis. In this study, three EOF modiers, namely DTAB, TTAC and
CTAB were investigated. The migration time of TGA was reduced
along withthe increased lengthof the alkyl chainas showninFig. 2.
It might be relatedtothe surfactant structures, the longer the chain,
the stronger the adsorption and the ability to control EOF [21]. As
a result, CTAB was selected.
Once EOF is reversed and reaches a stable value, the EOF no
longer depends on the CTAB concentration when its concentration
is 0.1mmol L
1
which corresponds to 10% of its standard critical
micelle concentration. On this occasion, the surface of capillary is
a dynamic structure represented by the bilayer assembly instead
Fig. 2. Effect of the EOF-reversing agent types on migration time. (a) CTAB, (b) TTAC
and (c) DTAB. (1) TGA, other CE conditions were the same as in Fig. 1.
Fig. 3. Effect of sample buffer on separation: (a) pure water as the sample medium,
(b) one-tenth dilution of the running buffer as the sample buffer, and (c) running
buffer as the sample buffer. (1) TGA and (2) unknown peak. Other CE conditions
were the same as in Fig. 1.
of the solid backbone of silica [21]. In order to further verify this
observation, CTAB concentrations at 0.2, 0.5 and 1.0mmol L
1
were
studied, respectively, with no differences between the separation
efciency and the migration time observed. Thus, CTAB with con-
centration of 0.5mmol L
1
was chosen because the EOF is stable at
this concentration [14,21].
3.1.4. Choice of sample buffer
For efcient sample stacking, the conductivity of the sample
buffer should be relatively lower than that of the running buffer.
In this study, pure water, one-tenth dilution of the running buffer
and the running buffer were tried as the sample buffer (Fig. 3). As
expected, one-tenth dilution of the running buffer showed the best
result.
3.2. CE method validation
3.2.1. Linearity, limit of detection, limit of quantitation, precision
and recovery
The corrected peak areas (Ac) versus the concentrations (,
mgmL
1
) of TGA showed a good linear relationship (Ac =50.33
+40.174) with a correlation coefcient (r) of 0.9998. The limit
of detection (S/N=3) and limit of quantitation (S/N=10) were
0.002mgmL
1
and 0.006mgmL
1
, respectively.
Sample 12 was prepared seven times within a day to evaluate
the intra-day precision of the CE method. Samples were accurately
weighed (0.1g for each) and treated as described in Section 2.4. To
evaluate the inter-day precision of the CE method, sample 12 was
pretreated in triplicate each day on seven consecutive days. Both
the intra- and inter-day precisions of the method were 1.4% and
the average content of TGA in sample 12 was 3.4%.
The average recoveries at the three spiked levels (0.125, 0.250
and 0.500mgmL
1
) were 96.9%, 102.3% and 94.0% with relative
standard derivations 2.1%, 3.9% and 2.2%, respectively.
3.2.2. Comparison of CE, HPLC and IC results
Sixteen samples of hair-treatment products and depilatory
creams were analyzed by CE, HPLC and IC, respectively. The
results are listed in Table 1. For most of the samples, the
results of the TGA assayed by CE were in good agreement
with the values obtained by HPLC and IC. However, for the
HPLC method, six out of 16 cosmetic samples could not be
512 N. Xie et al. / Journal of Pharmaceutical and Biomedical Analysis 88 (2014) 509512
Fig. 4. Chromatogram and electropherogram of the sample 3. (A) Chromatogram, (B) electropherogram. (1) TGA and (2) interfering peak. Other CE conditions were the
same as in Fig. 2. HPLC conditions: Mobile phase: acetonitrile and 0.01mol L
1
potassium dihydrogen phosphate solution (pH 2.5) with volume ratio of 10:90. Flow rate:
1.0mL min
1
. Detection wavelength: 215nm. Column temperature: 30

C. The sample injected was 20L.

quantied due to the sample matrix interferences. The chro-
matogram and electropherogram of the sample 3 are shown in
Fig. 4. For the IC method, the TGA could not be quantied due
to the serious interferences suffering from the sample matrix
insamples 1, 2, 3, 5, 7and8. All thedatademonstratedthereliability
and practicality of the new CE method.
3.3. Real sample analysis
Sixty-nine commercially available samples, including forty per-
manent wavings, sixteen hair straighteners and thirteen depilatory
creams were successfully analyzed. None of these samples con-
tained TGA exceed the permitted level.
4. Conclusions
The developed CE method demonstrated the straightforward
preparation, robustness, and accuracy for TGA analysis, and has
been recommended for routine quality control analysis of cosmet-
ics.
Acknowledgement
The project for training high-level medical technical personnel
in health system in Beijing City is gratefully acknowledged.
References
[1] Y.F. Zabashta, A.V. Kasprova, S.P. Senchurov, Y.E. Grabovskii, The location of the
thioglycolic acid molecules in intrabrillar unordered areas of the human hair
keratin structure, Int. J. Cosmet. Sci. 34 (2012) 223225.
[2] K. Suzuta, S. Ogawa, Y. Takeda, K. Kaneyama, K. Arai, Intermolecular disul-
de cross-linked structural change induced by permanent wave treatment of
human hair with thioglycolic acid, J. Cosmet. Sci. 63 (2012) 177196.
[3] D. Oikawa, W. Takeuchi, S. Murata, K. Takahashi, Y. Sekine, Measurement of
concentrations of thioglycolic acid, dithiodiglycolic acidandammoniainindoor
air of a beauty salon, J. Occup. Health 54 (2013) 370375.
[4] Z. Wang, X. Ren, D. Wang, Y. Guan, L. Xia, Effects of thioglycolic acidonpartheno-
genetic activation of xenopus oocytes, PLoS ONE 6 (2011) e16220.
[5] L. Xia, S. Hou, X. Ren, Z. Wang, Effects of thioglycolic acid on in vivo oocytes
maturation in mice, PLoS ONE 6 (2011) e23996.
[6] T. Zhao, Hygienic Standardfor Cosmetics Health, MilitaryMedical Press, Beijing,
China, 2007, pp. 239241.
[7] Z. Zhong, D. Du, C. Liang, J. Yao, Study on analysis of thioglycolic acid in cos-
metics by ion chromatography, Wei Sheng Yan Jiu 33 (2004) 491493.
[8] V. Cavrini, V. Andrisano, R. Gatti, G. Scapini, HPLC determination of thiogly-
colic acid and other aliphatic thiols in cosmetic formulations using ethacrynic
acid as precolumn derivatization reagent, Int. J. Cosmet. Sci. 12 (1990)
141150.
[9] J.M. Zen, H.H. Yang, M.H. Chiu, Y.J. Chen, Y. Shih, Determination of thioglycolic
acid in hair-waving products by disposable electrochemical sensor coupled
withhigh-performanceliquidchromatography, J. AOACInt. 92(2009) 574579.
[10] Methods for Determination of Thioglycolic Acid in Cosmetics, Documents of
Chinas Food and Drug Administration, Accessory Four, 2012, pp. 1719.
[11] L. Wang, Z. Chen, Determination of thioglycolic acid and cysteine in hair-
treatment products at ceramic carbon composite electrodes, Electroanalysis
9 (1997) 12941297.
[12] N. Bansal, A kinetic method for the determination of organosulfur compounds
by inhibition: determination of cysteine, 2,3-dimercaptopropanol and thiogly-
colic acid, Transition Met. Chem. 35 (2010) 483490.
[13] M. Sierra-Rodero, J.M. Fernndez-Romero, A. Gmez-Hens, Photometric deter-
mination of thioglycolic acid in cosmetics by using a stopped-ow reverse
ow-injection system and the formation of gold nanoparticles, Microchem. J.
97 (2011) 243248.
[14] E.A. Dutra, M.I.R.M. Santoro, G.A. Micke, M.F.M. Tavares, E.R.M. Kedor-
Hackmann, Determinationof -hydroxyacids incosmetic products bycapillary
electrophoresis, J. Pharm. Biomed. Anal. 40 (2006) 242248.
[15] P. Wang, X. Ding, Y. Li, Y. Yang, Simultaneous determination of eleven preser-
vatives in cosmetics by micellar electrokinetic chromatography, J. AOAC Int. 95
(2012) 10691073.
[16] A.R. Attar, D.E. Blumling, K.J. Knappenberger, Photodissociation of thioglycolic
acid studied by femtosecond time-resolved transient absorption spectroscopy,
J. Chem. Phys. 134 (2011) 024514.
[17] P. Ke, X. Sun, Preparation of CdS self-assembled monolayers and interfacial
sensing for DNA, Chin. J. Huaqiao Univ. (Nat. Sci.) 32 (1) (2011) 4852.
[18] Beckman Coulter, Introduction to Capillary Electrophoresis, Beckman Instru-
ment Inc., 1996, pp. 18.
[19] K. Altria, Buffer preparation hints, tips and common errors, LCGC Asia Pacic
10 (2) (2007) 2529.
[20] J.E. Melanson, N.E. Baryla, C.A. Lucy, Dynamic capillary coatings for electroos-
motic ow control in capillary electrophoresis, Trends Anal. Chem. 20 (2001)
365374.
[21] M.F.M. Tavares, R. Colombara, S. Massaro, Modied electroosmotic ow by
cationic surfactant additives in capillary electrophoresis: evaluation of elec-
trolyte systems for anion analysis, J. Chromatogr. A 772 (1997) 171178.