+Model

TAL-9613; No. of Pages 7 ARTICLE IN PRESS

Available online at www.sciencedirect.com

Talanta xxx (2008) xxx–xxx

Determination of textile dyes by means of non-aqueous capillary

electrophoresis with electrochemical detection
Alejandra-Alicia Peláez-Cid a,b , Sonia Blasco-Sancho a , Frank-Michael Matysik a,∗
a Institute of Analytical Chemistry, University of Leipzig, Linnéstr. 3, D-04103 Leipzig, Germany
b Centro de Quı́mica, Instituto de Ciencias, Universidad Autónoma de Puebla, Complejo de Ciencias Edificio 192 CU, 72570 Puebla, Mexico
Received 6 August 2007; received in revised form 15 January 2008; accepted 24 January 2008

Abstract
Eight textile dye compounds including five cationic dyes, namely, basic blue 41, basic blue 9, basic green 4, basic violet 16 and basic violet 3,
and three anionic dyes, acid green 25, acid red 1 and acid blue 324, were separated and detected by non-aqueous capillary electrophoresis (NACE)
with electrochemical detection. Simultaneous separations of acid and basic dyes were performed using an acetonitrile-based buffer. Particular
attention was paid to the determination of basic textile dyes. The optimized electrophoresis buffer for the separation of basic dyes was a solvent
mixture of acetonitrile/methanol (75:25, v/v) containing 1 M acetic acid and 10 mM sodium acetate. The limits of detection for the basic dyes were
in the range of 0.1–0.7 ␮g mL−1 . An appropriate solid-phase extraction procedure was developed for the pre-treatment of aqueous samples with
different matrices. This analytical approach was successfully applied to various water samples including river and lake water which were spiked
with textile dyes.
© 2008 Elsevier B.V. All rights reserved.

Keywords: Textile dyes; Non-aqueous capillary electrophoresis; Electrochemical detection; Environmental analysis; Solid-phase extraction

1. Introduction defined in both its inherent ecotoxicology and the concentra-

tions found in the environmental compartments. Thus, reliable
Dyes represent an important class of compounds produced analytical methods are needed to detect critical concentrations
worldwide in large quantities of more than 7 × 108 kg per annum of dyestuff in the environment.
[1] and are used in many industries such as nutritional, cos- A variety of analytical approaches for the determination of
metics, paper, pharmaceutical, printing inks, textile, tannery, dye compounds has been reported in the literature. These meth-
and others. Among these industrial applications textile indus- ods include thin-layer chromatography [5], HPLC [2,3,6–8],
try consumes 50% of the overall dye production. Some dyes, pyrolysis gas chromatography [9], capillary electrophoresis
like malachite green (basic green 4), have found applications as in aqueous [5,10–15], and in non-aqueous media [16–19].
fungicide and antiseptic in aquaculture [2,3]. Dye compounds Common detection techniques are UV [8,11–13] and diode
can be classified into cationic, non-ionic, and anionic subgroups. array detection [5–7,10], fluorescence [15], mass spectrometry
Anionic dyes can further be divided into direct, acid, and reac- [2,3,9,14], and electrochemical detection (ED) [16–20]. Cap-
tive type while cationic dyes represent the basic ones. Non-ionic illary electrophoresis (CE) techniques are particularly suited
dyes refer to disperse dyes because they do not form ions in an to achieve highly efficient separations of ionic dye species.
aqueous medium. For example an aqueous CE method with diode array detec-
A number of dyes, in particular azo dyes and their metabo- tion enabled the determination of eight food colorants in milk
lites, are known to have toxic and even carcinogenic character beverages [10] with detection limits lower than 0.5 ␮g mL−1 .
[4]. The environmental risk posed by a dye compound can be In addition, the attractive performance of CE separations in
the context of ink analysis was demonstrated by Vogt et al.
[21] and reviewed by Zlotnick and Smith [22]. However, in
∗ Corresponding author. Tel.: +49 341 97 36 107; fax: +49 341 97 36 115. aqueous systems various dye compounds can form hydropho-
E-mail address: [email protected] (F.-M. Matysik). bic interactions with the capillary wall which results in poor

0039-9140/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.talanta.2008.01.049

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
2 A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx

peak shapes. Non-aqueous capillary electrophoresis (NACE) is 2.2. Detector configuration and capillary specifications
an attractive alternative which can diminish these problems. In
NACE organic solvent systems are used instead of conventional The electrochemical detector cell used for the present work
aqueous buffers. In particular, the manipulation of selectivity has been described in detail elsewhere [24]. Briefly, the detec-
by varying the organic solvents choices and ratios can help to tor cell consists of an inert PTFE cell body with two stainless
solve complicated separation problems. This forms the basis steel tubes which guide the separation capillary and the work-
for interesting possibilities to adjust performance characteristics ing electrode to the right axial position. The working electrode
like selectivity, resolution, and analysis time. Among the sol- was a 25-␮m platinum microdisk electrode. Fused-silica capil-
vents that can be used for NACE acetonitrile exhibits the largest laries with an i.d. of 75 ␮m and an o.d. of 360 ␮m (Polymicro
ratio of dielectric constant to viscosity which leads to a high Technologies, Phoenix, AZ, USA) were used throughout this
electro-osmotic velocity and fast separations. It has been demon- work. Before initial use, new capillaries were washed with 0.1 M
strated that electrochemical methods are well suited for analyte NaOH for 10 min, with distilled water for 5 min, for 15 min with
detection in conjunction with NACE [23]. In acetonitrile-based pure acetonitrile and for 25 min with the buffer used. After use,
buffers electrochemical detection exhibits a very good stability the capillaries were flushed with pure acetonitrile. The capillary
of the detection response within an extended accessible poten- and working electrode are kept in place by PTFE adapters fitted
tial range and low limits of detection (LOD) can be achieved to the stainless steel tubes. The capillary-to-electrode distance of
[18]. 75 ± 5 ␮m was adjusted under a stereomicroscope by carefully
The aim of this work was to investigate the applicability of pushing the separation capillary towards the tip of the working
NACE–ED to the determination of cationic and anionic textile electrode. One of the stainless steel tubes served as counter elec-
dyes which represent common dye classes in textile effluents. trode and high-voltage ground. The cell was ready for operation
Particular attention was paid to the adaptation of the analyti- after filling it with 1.5 mL of the separation buffer and plac-
cal approach to the necessary matrix change from an aqueous ing a PTFE cap with an attached silver/silver chloride reference
sample to the NACE buffer system. electrode on top of it. As internal solution of the reference elec-
trode a non-aqueous acetonitrile-based buffer containing 1 M
2. Experimental acetic acid and 10 mM sodium acetate was used which ensured
that the drift of the reference electrode potential was less than
2.1. Electrophoretic and electrochemical equipment 1 mV h−1 . All potentials given in this work were measured with
respect to this reference system. In the presence of a high volt-
A high-voltage supply (Model HCN 7E-35000, F.u.G. Elek- age of 20 kV a shift of the working electrode potential of about
tronik, Rosenheim–Langenpfunzen, Germany) was used for 300 mV towards more negative potentials occurred [25] that had
measurements in conjunction with a home-made CE system. to be taken into consideration for a suitable potential setting.
Care was taken to ensure that the hydrostatic levels of the input
and output reservoirs were identical. Sample injection was done 2.3. Chemicals
by elevating the sample vial up to a difference in height of 10 cm
above the detector cell electrolyte level for a duration of 10 s. A Acetonitrile (99.9% HPLC grade, water <0.02%) and acetic
commercial CE system PrinCE model 255 (Prince Technologies, acid (99.99%) were obtained from Sigma–Aldrich (Stein-
Emmen, The Netherlands) equipped with a UV–vis detector heim, Germany), methanol (LiChrosolv) and sodium acetate
(Bischoff model lambda 1010, Leonberg, Germany) operated (Suprapur) were purchased from Merck (Darmstadt, Germany).
at 200 nm, was used for comparative measurements with UV Ferrocene was from Riedel-de-Haën AG (Seelze–Hannover,
detection. Germany) and was purified by sublimation. For solid-phase
In case of electrochemical detection the implementation of an extraction (SPE), the cartridges ENVI-Chrom P (0.25 and
isolating transformer ensured galvanic separation between the 0.50 g—6 ml) and ENVI-18 (0.50 g—6 ml) were purchased
high-voltage supply and the potentiostat. The electrochemical from Supelco (Bellefonte, PA, USA). The chemicals used for
detector cell was placed in a Faraday cage so as to minimize inter- SPE were of analytical-reagent grade and were used as received.
ference from external noise. All electrochemical measurements The following textile dyes were used in this work: San-
were performed in the three-electrode mode with the aid of a docryl blue (basic blue 41), sandocryl red (basic violet 16),
voltammetric analyzer Model Autolab PGSTAT10 (Eco Chemie, sandocryl green (basic green 4), sandolan green (acid green
Utrecht, The Netherlands) equipped with a low current amplifier 25), sandolan red (acid red 1) and nylosan blue (acid blue 324)
module ECD system. The current signal was filtered through a which were kindly provided by Clariant (México). The dye com-
third-order Sallen–Key filter having a time constant of 100 ms. pounds crystal violet (basic violet 3) and methylene blue (basic
The interval time of the current measurements for amperometric blue 9) were analytical-reagent grade and were obtained from
recordings of the electropherograms was 0.3 s when the analy- Feinchemie K.-H., Kallies KG (Sebnitz, Germany) and Riedel-
sis time was shorter than 1000 s and when the analysis time was de-Haën (Seelze, Germany), respectively. All dye substances
longer, it was 0.5 s. Before running a new electropherogram, a were used as received without further purification. Table 1 sum-
working electrode potential of 3.0 V was applied for 5 s followed marizes specifications of the dye compounds used.
by −1.0 V for 5 s to ensure long-term stability of the detector Sample pre-treatment for aqueous solutions was based on
response. SPE using ENVI-Chrom P cartridges. The SPE cartridges were

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx 3

Table 1
Overview of basic and acid dyes used in this work
Name Color index (code) Chemical structure Chemical classification

Sandocryl blue CI. 11105 basic blue 41 (BB41) Monoazo

Methylene blue CI. 52015 basic blue 9 (BB9) Thiazine

Sandocryl red CI. 48013 basic violet 16 (BV16) Methine

Sandocryl green CI. 42000 basic green 4 (BG4) Triarylmethane

Crystal violet CI. 42555 basic violet 3 (BV3) Triarylmethane

Nyosan blue Acid blue 324 (AB324) Anthraquinone

Sandolan red CI. 18050 acid red 1 (AR1) Monoazo

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
4 A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx

Table 1 (Continued )
Name Color index (code) Chemical structure Chemical classification

Sandolan green CI. 61570 acid green 25 (AG25) Anthraquinone

activated immediately prior to use by passing through 3 ml shows the cyclic voltammograms for the basic and acid dyes in
methanol followed by 6 ml of distilled water. The sample was acetonitrile solution. All target analytes can be oxidized within
applied to the cartridge and drawn through at a flow-rate of the accessible potential range. However, the oxidation mech-
3 mL min−1 . The cartridges were then rinsed with 12 mL of dis- anisms are rather complex involving typically more than one
tilled water and 2 mL of pure acetonitrile. Thereafter the dyes oxidation step and following reactions. In the context of this
were eluted with 3–5 mL of acetonitrile/tetrahydrofuran (2:1, work no further efforts were undertaken to investigate details of
v/v) solution at a flow-rate of 1 mL min−1 and investigated by the oxidation processes. All dye compounds allow the recording
NACE–ED by injecting directly the extracted solution. of mass transport controlled detection signals within the poten-
tial window from 1.5 to 1.7 V. In order to obtain information
3. Results and discussion regarding problems due to electrode fouling five consecutive
cyclic voltammograms were recorded for the respective dye
3.1. Voltammetric behaviour of textile dyes in non-aqueous compounds. Without an electrochemical pre-treatment between
solutions the successive recordings decreasing and changing signals were
typically obtained. This indicates the presence of reaction prod-
In order to investigate the applicability of ED in non-aqueous ucts on the electrode surface causing changes in the electrode
media for the determination of textile dyes the voltammetric response. However, an electrochemical activation procedure
behaviour of the individual dye compounds was studied. Fig. 1 which applied an anodic pulse of 3.0 V for 5 s and a cathodic
pulse of −1.0 V for 5 s was sufficient to restore the initial
electrode behaviour. Fig. 2 exemplifies this situation for basic
blue 9 illustrating the positive effect of an electrochemical pre-

Fig. 1. (a) Cyclic voltammograms of basic dyes: 1, basic violet 3 (408 ␮g mL−1 );
2, basic green 4 (418.5 ␮g mL−1 ); 3, basic blue 9 (320 ␮g mL−1 ); 4, basic blue 41
(451 ␮g mL−1 ); 5, basic violet 16 (369 ␮g mL−1 ) and acetonitrile-based back-
ground electrolyte (BGE) containing 1 M acetic acid and 10 mM sodium acetate. Fig. 2. Repetitive cyclic voltammograms of basic blue 9 (320 ␮g mL−1 ) in
(b) Cyclic voltammograms of saturated solutions of acid dyes 6, acid blue 324; acetonitrile containing 1 M acetic acid and 10 mM sodium acetate (a) without
7, acid green 25; 8, acid red 1 and acetonitrile-based background electrolyte con- pre-treatment between successive runs and (b) with an electrochemical pre-
taining 1 M acetic acid and 10 mM sodium acetate. A 25-␮m platinum microdisk treatment (3 V and −1 V for 5 s) between successive runs. Other conditions
electrode was used as working electrode and the scan rate was 25 mV/s. were as in Fig. 1.

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx 5

treatment between the runs. It should be noted that the effects

of deactivation of the sensing electrode is much smaller for typ-
ical concentrations (␮g mL−1 ) used in the NACE experiments.
In addition, in contrast to cyclic voltammetry experiments in
NACE the electrode is just a short period of time in contact with
the analyte zones.
As a conclusion of the above voltammetric studies and taking
into account the shift of the signal due to the presence a high
voltage during NACE separations [25], the detection potential of
1.8 V was selected for NACE–ED determinations combined with
an electrochemical pre-treatment between consecutive electro-
pherograms.

3.2. Non-aqueous capillary electrophoresis studies

Fig. 3. Non-aqueous capillary electrophoretic separations with electrochemical
detection (NACE–ED) for a mixture of acid and basic textile dyes: 1, basic blue 9
The electrolyte system used for NACE–ED determinations of
(3.2 ␮g mL−1 ); 2, basic green 4 (4.2 ␮g mL−1 ); 3, basic violet 16 (3.7 ␮g mL−1 );
textile dye compounds consisted of 1 M acetic acid and 10 mM 4, basic violet 3 (4.1 ␮g mL−1 ); 5, basic blue 41 (4.5 ␮g mL−1 ); 6, acid green
sodium acetate in acetonitrile which was previously proofed 25 (31 ␮g mL−1 ); 7, acid blue 324 (25 ␮g mL−1 ); 8, acid red 1 (28 ␮g mL−1 ).
to yield very reliable NACE–ED results [16]. Basic dye com- The letters a–d represent several separated species present in the respective dye
pounds form cationic species in this buffer medium and acid dyes sample, EOF corresponds to the electro-osmotic flow. Experimental conditions:
capillary dimensions, 60 cm × 75 ␮m i.d.; running electrolyte, 1 M acetic acid
migrate as anionic species. Consequently, the migration times
and 10 mM of sodium acetate in acetonitrile; hydrodynamic injection, 10 s at
for basic dyes were shorter than that of an electro-osmotic flow a height difference of 10 cm; separation voltage of 20 kV; sensing electrode,
(EOF) marker (ferrocene) and acid dyes showed longer migra- 25-␮m diameter platinum microdisk electrode; detection potential, 1.8 V.
tion times than ferrocene. The acid dye compounds were only
slightly soluble in the electrolyte system used. Therefore, stock
solutions of known concentration were prepared in methanol baseline separation could be established. Fig. 4 illustrates these
and diluted with the buffer used for NACE separations. For results in detail. The composition of the electrophoresis buffer
some of the dye substances, in particular acid green 25 and acid used in Fig. 4(b) was chosen for all subsequent studies of basic
red 1, more than one component were found in the NACE–ED textile dyes.
recordings. This observation reflected the limited purity of the The precision of migration times and peak heights was very
respective dye substances. All concentrations given in this study good as specified in Table 2. Comparative measurements were
refer to the total amount of the respective dye preparations. The made using a commercial CE system equipped with a UV detec-
actual concentrations of the main dye components are corre- tor. The limits of detection obtained with ED and UV detection
spondingly lower in these cases. Fig. 3 illustrates a simultaneous were compared. The results are summarized in Table 2. In all
separation of acid and basic textile dyes present in a mixture. cases ED led to lower LODs than UV detection.
Obviously, the basic dyes exhibit similar migration behaviour
and some of them co-migrate (basic violet 3 and basic blue 41)
or are not completely resolved from other signals (basic vio-
let 16). On the other hand a good separation performance was
achieved for the acid dyes and some of their impurities.
The following studies were focused on the separation and
detection of basic dyes because they were available as pure
substances and the reliability of the ED response is better if
the analytes migrate before the EOF. The latter result can be
attributed to water transported by the EOF and corresponding
oxidation reactions at the platinum sensing electrode leading to
changes of the electrode response characteristics. However, a
well-defined detection response could be restored by the elec-
trochemical pre-treatment protocol described under Section 3.1
which was applied between consecutive electrophoretic runs.
In order to improve the separation of basic dyes methanol was
added to the separation buffer. It was found that an increasing
content of methanol led to an improved resolution for the sep- Fig. 4. NACE–ED for a mixture of five basic dyes: 1, basic blue 9 (16 ␮g mL−1 );
2, basic green 4 (21 ␮g mL−1 ); 3, basic violet 16 (18 ␮g mL−1 ); 4, basic violet 3
aration of basic dyes, however, the peak width for basic blue 9
(20 ␮g mL−1 ); 5, basic blue 41 (23 ␮g mL−1 ) (a) in acetonitrile containing 1 M
increased slightly. A methanol content of 25% resulted in well- acetic acid and 10 mM sodium acetate and (b) in acetonitrile–methanol (75:25,
separated signals for all dye compounds. In particular, for the v/v) containing 1 M acetic acid and 10 mM of sodium acetate. Experimental
signals of basic violet 3 (peak 4) and basic blue 41 (peak 5) a conditions were as in Fig. 3.

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
6 A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx

Table 2
Analytical characteristics of determinations of basic textile dyes by means of non-aqueous capillary electrophoresis with electrochemical detection (ED)/ultraviolet
detection (UV)

Compound Concentration (␮g mL−1 ) Migration time (s) (RSDa , %) Peak height (nA) (RSDa , %) LODb (ED) (␮g mL−1 ) LODb (UVc ) (␮g mL−1 )

BB9 16 201.5 (0.43) 3.80 (2.55) 0.13 1.70

BG4 21 213.8 (0.41) 4.07 (0.88) 0.16 0.86
BV16 18 218.5 (0.45) 1.28 (2.30) 0.43 2.39
BV3 20 222.1 (0.45) 5.64 (1.68) 0.11 0.68
BB41 23 225.5 (0.42) 0.98 (4.02) 0.72 2.16

Experimental conditions were as in Fig. 4.

a Relative standard deviation (number of measurements, 6).
b The limits of detection (LOD) were calculated on the basis of the peak-to-peak noise and a signal-to-noise ratio of 2.
c Detection wavelength, 200 nm; capillary dimensions, 85 cm × 75 ␮m i.d. (65-cm effective length); separation voltage, 25 kV.

3.3. Studies of aqueous sample solutions

The above results demonstrate the attractive performance of

NACE–ED for mixtures of cationic textile dyes prepared in non-
aqueous media. However, for the determination of cationic dyes
in real aqueous samples additional challenges have to be consid-
ered. The main problem is the necessity of matrix change from
the aqueous to the non-aqueous medium. Initial attempts were
made to inject aqueous samples directly into the NACE–ED
system. In this case noisy signals and distorted peak shapes
were obtained. In addition, precipitation of inorganic salts could
complicate the determination. Therefore, sample pre-treatment
methods such as liquid–liquid extraction (LLE) and solid-phase
extraction were tested. A LLE protocol was adapted from a
report by Pandit and Basu [26]. However, this approach was only
Fig. 5. NACE–ED of (a) a lake water sample after solid-phase extraction and
applicable for spiked samples prepared from distilled water but (b) of a lake water sample spiked with a mixture of four basic dyes: 1, basic
failed in case of more complex aqueous matrices. Consequently, blue 9 (0.013 ␮g mL−1 ); 2, basic green 4 (0.017 ␮g mL−1 ); 3, basic violet 16
experiments based on SPE were conducted using various SPE (0.015 ␮g mL−1 ); 4, basic violet 3 (0.016 ␮g mL−1 ) after SPE. Experimental
cartridges and elution procedures. Finally, a SPE protocol was conditions were as in Fig. 3.
elaborated which is based on ENVI-Chrom P cartridges and
described in Section 2. The recoveries of extractions of cationic 10 and 7000 ␮g mL−1 [27–30] to which the above protocol could
textile dyes were determined for different aqueous sample matri- easily be adapted. For environmental samples with considerably
ces including distilled, river and lake water. Table 3 summarizes lower concentrations of the target analytes the sensitivity can
the recoveries obtained for various dye components in differ- further be increased by the application of large sample volumes
ent aqueous samples. In general, the recoveries were reduced in to the SPE extraction cartridges.
more complex aqueous matrices.
Fig. 5 illustrates results of NACE–ED measurements after 4. Conclusions
SPE of lake water (5a) and lake water spiked with textile dyes
(5b). No interfering substances were found which suggests that This work demonstrated the applicability of NACE–ED for
this analytical approach has the potential to be applied to a wide the determination of acid and basic textile dyes. A compromise
range of real samples. According to literature reports typical between separation and detection aspects was found using a
dyestuff concentrations found in textile effluents range between mixed acetonitrile/methanol (75:25, v/v) solvent system con-
taining 1 M acetic acid and 10 mM sodium acetate. In this
Table 3 medium baseline separations could be achieved for all dye
Recoveries of solid-phase extraction experiments of basic textile dyes based on compounds studied with limits of detection between 0.1 and
ENVI-ChromP cartridges (see Section 2 for details) 0.7 ␮g mL−1 which were clearly lower than obtained with UV
Type of Recovery (%) detection.
water Based on solid-phase extraction an effective sample pre-
Basic blue 9 Basic green 4 Basic violet 16 Basic violet 3
treatment protocol could be established to investigate aqueous
Distilled 93 81 Not determined 98 dye samples for which a matrix change from the aqueous to
River 87 84 85 82 the non-aqueous medium has to be carried out. The application
Lake 62 48 83 43
of the elaborated analytical approach enables determinations of
Determinations were done by NACE–ED. textile dyes in real samples such as river or lake water.

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049
+Model
TAL-9613; No. of Pages 7 ARTICLE IN PRESS
A.-A. Peláez-Cid et al. / Talanta xxx (2008) xxx–xxx 7

Acknowledgements [12] P. Blatny, C.-H. Fischer, A. Rizzi, E. Kenndler, J. Chromatogr. A 717 (1995)
157.
[13] S. Hamai, K. Sato, Dyes Pigments 57 (2003) 15.
A.A. Peláez-Cid is grateful for the Alßan scholarship no.
[14] E.D. Lee, W. Mück, J.D. Henion, T.R. Covey, Med. Environ. Mass Spec-
E04E044054MX within the European Union Programme of trom. 18 (1989) 253.
High Level Scholarships for Latin America. The authors wish [15] P.K. Dasgupta, Z. Genfa, J. Li, C.B. Boring, S. Jambunathan, R. Al-Horr,
to thank M.P. Elizalde-Gonzáles for support and discussion. Anal. Chem. 71 (1999) 1400.
[16] F.-M. Matysik, J. Chromatogr. A 802 (1998) 349.
[17] F.-M. Matysik, Electrochim. Acta 43 (1998) 3475.
References [18] F.-M. Matysik, Electroanalysis 11 (1999) 1017.
[19] A.R. Fakhari, M.C. Breadmore, M. Macka, P.R. Haddad, Anal. Chim. Acta
[1] T. Robinson, G. McMullan, R. Marchant, P. Nigam, Bioresour. Technol. 580 (2006) 188.
77 (2001) 247. [20] N. Dossi, E. Piccin, G. Bontempelli, E. Carrilho, J. Wang, Electrophoresis
[2] P. Scherpenisse, A.A. Bergwerff, Anal. Chim. Acta 529 (2005) 173. 28 (2007) 4240.
[3] D.R. Doerge, M.I. Churchwell, T.A. Gehring, Y.M. Pu, S.M. Plakas, Rapid [21] E. Rohde, A.C. Mcmanus, C. Vogt, W.R. Heinemann, J. Forensic Sci. 42
Commun. Mass Spectrom. 12 (1998) 1625. (1997) 1004.
[4] M. Inoue, F. Okada, A. Sakurai, M. Sakakibara, Ultrason. Sonochem. 13 [22] J.A. Zlotnick, F.P. Smith, J. Chromatogr. B 733 (1999) 265.
(2006) 313. [23] F.-M. Matysik, Electrophoresis 23 (2002) 3711.
[5] T. Goldmann, F. Taroni, P. Margot, J. Forensic Sci. 49 (2004) 716. [24] F.-M. Matysik, J. Chromatogr. A 853 (1999) 27.
[6] M.P. Elizalde-González, A.A. Peláez-Cid, Environ. Technol. 24 (2003) 821. [25] F.-M. Matysik, Anal. Chem. 72 (2000) 2581.
[7] M.M. Dávila-Jiménez, M.P. Elizalde-González, A.A. Peláez-Cid, Colloids [26] P. Pandit, S. Basu, J. Colloid Interface Sci. 245 (2002) 208.
Surf. A 254 (2005) 107. [27] I.G. Laing, Rev. Prog. Coloration 21 (1991) 56.
[8] L.-H. Ahlström, E. Björklund, L. Mathiasson, Anal. Bional. Chem. 382 [28] C. O’Neill, F.R. Hawkes, D.L. Hawkes, N.D. Laurenco, H.M. Pinheiro, W.
(2005) 1320. Delée, J. Chem. Technol. Biotechnol. 74 (1999) 1009.
[9] A. Plum, W. Engewald, A. Rehorek, Chromatographia 57 (2003) S-243. [29] P.C. Vandevivere, R. Bianchi, W. Verstraete, J. Technol. Biotechnol. 72
[10] H.-Y. Huang, Y.-C. Shih, Y.-C. Chen, J. Chromatogr. A 959 (2002) 317. (1998) 289.
[11] P. Blatny, C.-H. Fischer, E. Kenndler, Fresenius J. Anal. Chem. 352 (1995) [30] O.J. Hao, H. Kim, P.-C. Chiang, Crit. Rev. Environ. Sci. Technol. 30 (1999)
712. 449.

Please cite this article in press as: A.-A. Peláez-Cid, et al., Talanta (2008), doi:10.1016/j.talanta.2008.01.049