Full text of DOI:10.1093/chromsci/40.9.523

Journal of Chromatographic Science, Vol. 40, October 2002
Determination of Triazines in Water Samples by
High-Performance Liquid Chromatography with
Diode-Array Detection
M.S. Dopico, M.V. González*, J.M. Castro, E. González, J. Pérez, M. Rodríguez, and A. Calleja
Departamento de Química Analítica, Escuela Universitaria Politécnica, Universidade da Coruña, Avenida 19 de Febrero,15405 – Ferrol,
A Coruña, Spain
liquid–liquid or liquid–solid procedures. Liquid–liquid extraction
(LLE) has been the most employed procedure, but today, solid-
Abstract
phase extraction (SPE) is being applied as an alternative to LLE in
many official methods of analysis established by the regulatory
Triazines are widely used herbicides that can be detected in the
environment at trace level. A preconcentration step is necessary to
agencies from the U.S. and Europe. SPE has some advantages
over LLE, such as the availability of different solid sorbents and
the need for smaller volumes of organic solvents. However, the
highly polar water-soluble degradation products are more diffi-
cult to extract in many organic solvents (5).
determinate them before analysis. In this study, carbonaceous and
polymeric adsorbents are compared with C18 for the solid-phase
extraction of simazine, atrazine, and propazine in water samples in
order to quantitate their levels by high-performance liquid
chromatography using photodiode-array detection.
High-performance liquid chromatography (HPLC) has been
used these last few years because of its ability to analyze both
polar and nonpolar thermodegradable compounds without the
need for a derivatization step (5). Triazines are easily determined
by liquid chromatography (LC) using a diode-array–UV detector
because of their strong absorbance at the 220-nm line, which
gives detection limits almost equivalent to those obtained by gas
chromatography (GC)–nitrogen–phosphorus detection, the ana-
lytical method initially used (6).
Introduction
The use of pesticides in agriculture has increased over the last
several years. Such different products can be damaging to the
human health and environment. The presence of pesticides in the
environment has forced official international institutions to
establish maximum concentration levels allowed in drinking
water and foods (1). One of the most frequently employed herbi-
cides is triazine atrazine (2). Most triazines are derived from s-tri-
azine, a six-atom heterocyclic compound characterized by the
presence of nitrogen atoms symmetrically arranged in the 1, 3,
and 5 positions and various substituents in the 2, 4, and 6 posi-
tions. The persistence of triazines and their degradation products
in soil, water, plant matter, and animals can be significant (3).
The 98/83/CE Directive (4), relating to the quality of waters for
human consumption, establishes 0.1 µg/L as the highest concen-
tration allowed for individual plaguicides and 0.5 µg/L for the
total of plaguicides. These levels are not detected by the usual ana-
lytical techniques, thus the combination of analytical methods
and preconcentration procedures that will achieve the enrich-
ment of the analyte necessary for their determination in the envi-
ronment.
In this study, the determination of simazine, atrazine, and
propazine (Figure 1) in water samples are concentrated by an SPE
The enrichment of trace triazine can be attained by
Figure 1. Chlorotriazine structures.
* Author to whom correspondence should be addressed: email victoria@udc.es.
Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
523

Journal of Chromatographic Science, Vol. 40, October 2002
step, offline, and followed by subsequent chromatographic anal-
ysis. V. Pichon in 1998 (7) recognized three kinds of sorbents for
SPE. N-alkyl silica sorbents have been the universal extraction
sorbents, being particularly appropriate for polar compounds in
having log Kow < 2.5–3. Additional sorbents showing a high capa-
bility to extract polar analytes are highly cross-linked styrene-
divinylbenzenes (SDB) as well as carbon-based sorbents. The
effectiveness of three kinds of sorbents in the SPE step has been
studied and a reversed-phase HPLC with a diode-array detection
method has been developed.
Experimental
Materials and reagents
Acetonitrile (ACN) (HPLC grade) was obtained from Panreac
Química S.A. (Barcelona, Spain). Potassium dihydrogen phos-
phate (KH2PO4) was from Merck (Darmstadt, Germany). Toluene
for analysis was from Vorquímica S.L. (Vigo, Spain). Helium (C50)
Figure 2. SPE with C18 sorbent.
Figure 3. SPE with carbonaceous sorbent.
524

Journal of Chromatographic Science, Vol. 40, October 2002
was from Carburos Metálicos (Barcelona, Spain) and was used for
degassing the mobile phases.
used to stabilize with (~20 mg/L amylene for LC), and ethyl
acetate used for UV–IR–HPLC–HPLC prep-PAR instrumental
analysis were obtained from Panreac Química S.A. Methanol for
LC was obtained from Merck. Tartaric acid crystallized pure was
obtained from Probus-Merck (Darmstadt, Germany), and N2 (C55)
was from Carburos Metálicos.
C18 columns (900-mg w/2 frits) and Carbograph columns (car-
bonaceous sorbent, 12 cc, 1-g w/2 frits) were obtained from Lida
Manufacturing Corp. (Kenosha, WI). Polymeric Resins HR-P
(highly porous sorbent resin on the basis of polystyrene– divinyl-
benzene, 6 mL, 1 Kg Chromabond) was from Macherely-Nagel
(Postfach, Düren, Germany). ACN used for UV–infrared (IR)–
HPLC–HPLC prep-PAR instrumental analysis, dichloromethane
Instrumentation
Chromatography was performed with a modular HPLC system
(Waters Corporation, SA, Milford, MA) equipped with a Waters
600 quaternary pump for HPLC, a Waters 600 solvents distribu-
tion unit, a Waters U6K injector, a Waters 996 diode-array
detector, and a Millennium system Version 2.15 chromatography
manager for acquisition and treatment of data. Column
Symmetry C18 (5 µm, 150 mm) and precolumn Symmetry C18
were from Waters Corporation, SA. A Crison pH-meter and a
Sartorius 1 × 10–4 analytical balance were used.
Chromatographic determination
The chromatographic conditions used were as follows. The
mobile phase was water–phosphate buffer (1µM, pH 6–7, 55:45).
The flow rate was 1.5 mL/min, the injection volume 70 µL, and
the internal standard toluene. Chromatograms were obtained at
the 220-nm line and, when necessary, at the 260-nm line.
A 80-mg/L stock solution was obtained by dissolving 4.0 0.1
mg of each triazine standard (Supelco, Bellefonte, PA) in 50 mL of
ACN. A 2M HCl sample was added in order to improve the disso-
lution of the triazines in ACN (8).
The working standards were prepared from the 80-mg/L stock
solution by dilution with ACN in 10-mL flasks. Ten microliters of
0.001% (v/v) toluene were added as the internal standard.
Aqueous samples were spiked by adding known volumes of the
triazine standards, resulting in 2- and 20-µg/L levels.
All of the solutions were maintained at 4°C in glass flasks pro-
tected from light (9).
SPE conditions
Three kinds of sorbents were employed: C18 (Junker-Bucheit et
al. [10], Martin-Esteban et al. [11], and Ferrer et al. [8]); poly-
meric (Junker-Bucheit et al. [10], Aguilar et al. [12], and Lacorte
et al. [13]); and carbonaceous sorbents (Berg et al. [14], Pichon et
al. [15], and Di Corcia et al. [16]). The different steps of the pro-
cess (passage of the sample through the sorbent, washing, and
elution) were optimized. In order to improve the elution step and
obtain the smallest volume of eluent, two kinds of experiments
were performed. In the early investigations, mixtures of the elu-
tion solvents at different percentages were used to optimize their
polarity. Then, the solvent volume was studied in order to settle
the smallest solvent volume to desorb analytes completely. The
optimized concentration step in the solid phase for each kind of
sorbent is shown in Figures 2, 3, and 4.
Sample preparation
Samples were collected according to ISO 5667-3:1994 (17)
using dark glass bottles previously washed with water and deter-
gent, rinsed with distilled water, dried for 2 h at 105°C in a heater,
and finally rinsed with ACN (the solvent used in the extraction
step). According to Martín-Esteban et al. (1), the pH of the sample
Figure 4. SPE with polymeric sorbent.
525

Journal of Chromatographic Science, Vol. 40, October 2002
must be adjusted to pH 5–9, then it must be stored at 4°C,
extracted before 7 days, and the extract analyzed before 40 days.
The pH of the samples was determined showing that the values
for all of them were approximately 7–7.5. Samples were preserved
in the refrigerator and filtered by using 0.45-µm Millipore
(Bedford, MA) filters prior to the preconcentration step.
According to Sabik et al. (2), the previous filtration of the sample
does not affect the determination of triazines.
Results and Discussion
Chromatographic method
The chromatogram of Figure 5 (carried out at a 1-mL/min flow
rate and with an injection volume of 50 µL) shows that the elu-
tion order for the studied triazines is simazine (k' = 1.17), atrazine
(k' = 1.72), and propazine (k' = 2.59). Noble (18) gives log Kow
=
1.51–2.26, 2.0–2.5, and 2.91–3.02 for simazine, atrazine, and
propazine, respectively. This shows that the apolar character of
the triazines increases with the length of the alkylic chain, which
agrees with the elution order observed. The spectra corre-
sponding with each peak showed a first maximum at 220 nm and
a second one of less intensity at 260 nm. Therefore, the individual
triazines must be distinguished by their retention times. The
presence of a peak coeluting with the atrazine peak and inter-
fering with its response was observed when standard solutions of
small concentrations were injected. This peak seems to be a result
of some impurity present in the ACN. Thus, the evaluation of the
response of the atrazine must be made at the 260-nm line in order
to avoid their interference.
Chromatographic parameters associated with the optimized
conditions (1 < k' < 5, α < 1.14) (19) show that they are suitable
for the chromatographic analysis with regard to resolution and
efficiency of the separation and commensurate with a short anal-
ysis time. In order to validate the chromatographic method, a
study of the precision was carried out evaluating its repeatability
and reproducibility. The evaluation of the repeatability of the
chromatographic method was performed from the values
Figure 5. Chromatogram obtained with 0.2 mg/L dissolution and an injection
volume of 70 µL. The mobile phase was 55:45 ACN–water–phosphate buffer
(1 × 10–6M) with a 1.5-mL/min flow rate. The stationary phase was Symmetry
C18 (5 µm, 150 mm, 220 nm).
Table II. Parameters that Define the Graphic Chart
Table I. Repeatability of the Chromatographic Method*
Simazine Atrazine Propazine
(mg/L)
(mg/L)
(mg/L)
1 mg/L
0.1 mg/L
%A/AIS
Concentration
RSD (%)
Nominal level
Warning limits
Control limits
Mean (A/AIS)
SD*
Mean + 2SD
Mean – 2SD
Mean + 3SD
Mean – 3SD
0.50
0.10
0.71
0.30
0.81
0.20
0.43
0.09
0.61
0.24
0.70
0.15
0.50
0.08
0.65
0.34
0.73
0.26
†
%tR
%A/AIS
%tR
Simazine
Atrazine
Propazine
0.24
0.18
0.25
5.7
5.6
5.8
0.22
0.33
0.18
15
11
16
* n = 10.
† tR, retention time.
* SD, standard deviation.
Table III. Calibration Curve (10, 25, 50, 160, and 200 µg/L); Injection Volume (70 µL); and Flow Rate (1.5 mL/min) Used to
Obtain the Calibration Graph by Linear Regression*
xLD§
(µg/L)
xLC**
(µg/L)
‡
a
b
r†
r2
Sy/x
Simazine
34
Atrazine
41
–6.1 × 10–2 2.8 × 10–3
2.2 × 10–1 1.6 × 10–3
–7.3 × 10–3 2.1 × 10–3
16 0.91
7.4 0.51
15 0.68
0.9991
0.9988
0.9995
0.9983
0.9976
0.9989
0.0029
0.0303
0.0407
10
12
8.2
Propazine
27
* y = a + bx; a = intercept; and b = slope.
† r, correlation coefficient.
526

Journal of Chromatographic Science, Vol. 40, October 2002
obtained for 10 injections of a standard of the same solution using
two different concentrations (1.0 and 0.1 mg/L) containing an
internal standard. Table I shows a diminution in the variation
coefficient between 76% and 85% for the 1.0-mg/L standard solu-
tion using toluene as the internal standard. For simazine and
propazine, a diminution between 63% and 72% in the 0.1-mg/L
standard solution was observed.
carbonaceous sorbent had also the advantage of a rapid analysis
because it allows a high flow rate that cannot be achieved with the
C18 or polymeric sorbents used (16).
Comparing these results with that of other authors, it can be
concluded that similar and even higher recoveries have been
obtained. Sabik et al. (2) obtained a 51–84% recovery by using
carbonaceous cartridges in the analysis of a sequence of pesticides
including simazine, atrazine, and propazine. Nouri et al. obtained
a 94–109% recovery by using SDB cartridges (2). Barceló et al.
(20) obtained a 80–125% recovery by using C18 silica disks in the
analysis of 22 pesticides including simazine and atrazine.
Peak-area reproducibility was evaluated by means of 10 injec-
tions of a 1.0-mg/L standard solution carried out on different
dates over a month, giving the same relative standard deviation
(RSD) of 17% for simazine, atrazine, and propazine. The values
show the advantage of using an internal standard. In order to
complete this reproducibility study, Shewmart graphics were
obtained representing the variable of concern (A/AIS) for each of
the triazines versus time. Twenty determinations of the 1.0-mg/L
standard solution (n = 3) were performed to define the middle
value, the warning, and the action limits, taking into account the
parameters that will affect the response of the instrument: dif-
ferent day, different operator, and different standard (two samples
prepared with a month of difference were used) (Table II).
The calibration graph was obtained by linear regression (y = a
+ bx) for 10, 25, 50, 160, and 200 µg/L (Table III). The mean value
of three measurements was calculated for each point. In order to
improve the detection and quantitation limits, the initial condi-
tions were changed (70-µL injection volume and 1.5-mL/min
flow rate).
Sample analysis
The preconcentration and chromatographic analysis methods
described were applied in the determination of the presence of
simazine, atrazine, and propazine in the Grande of Xubia River
(northwestern Spain). Five sampling points were established
along the river. Samples were taken in July in duplicate with the
purpose of making the analysis of each sampling point by using
carbonaceous and C18 columns. None of the analyses showed the
presence of the triazines studied (Figure 6). For C18 experiments,
500 mL of the sample was used and 1000 mL was used for the car-
bonaceous analysis. Mills et al. (21) showed that a 100% retention
can be obtained by using C18 cartridges containing 360 mg of 40-
µm bonded silica, with a breakthrough volume starting at 750,
1250, and 2500 mL, respectively, for pure water samples. In
regards to carbonaceous sorbents, Di Corcia et al. (16) analyzed
2 L of water samples. Therefore, sample volumes of 500 and 1000
mL for C18 and carbonaceous columns will be below the break-
through volume.
The values obtained for the chromatographic parameters and
the precision of the chromatographic method were suitable. The
detection and quantitation limits (8–12 µg/L and 27–41 µg/L,
respectively) (Table III) were higher than those established by reg-
ulatory limits (0.1 µg/L), which confirms the need for a precon-
centration step previous to the chromatographic analysis.
Conclusion
Solid-phase preconcentration
A study of the preconcentration for the three triazines was per-
formed by evaluating the recovery for aqueous standard solu-
tions. Table IV shows the results obtained for the precon-
centration of 100 mL ultrapure water spiked with 2 µg/L each of
the three triazines with the kinds of sorbents used.
The higher recovery percentages for the three triazines were
obtained by using C18 or carbonaceous sorbent except for
propazine, which showed a better recovery by using a polymeric
sorbent. However, the smaller values of standard deviation and
variation coefficient corresponding with the carbonaceous
column were very similar for C18 and polymeric columns. The
By introducing a solid-phase preconcentration step, the initial
volume passes from 1 L for a carbonaceous sorbent to a final
volume of 1 mL, obtaining an enrichment factor of 1000. Thus,
with a 100% recovery, the detection and quantitation limits of the
method (8–12 µg/L and 27–41 µg/L, respectively) will improve
Table IV. Recovery (%) of 100 mL of Ultrapure Water
Spiked Until 2 µg/L of Simazine, Atrazine, and Propazine
Results After SPE With C18, Carbonaceous, and Polymeric
Sorbents*
C18
Carbonaceous
Polymeric
75
87 10
100 10
Simazine
Atrazine
Propazine
104
104 11
105 10
7
99
95
92
1
6
3
8
Figure 6. Chromatogram obtained after SPE with a carbonaceous sorbent of
1 L of the sample. The injection volume was 70 µL, and the mobile phase
was 55:45 ACN–water–phosphate buffer (1 × 10–6M). The flow rate was
1.5 mL/min, and the stationary phase was Symmetry C18 (5 µm, 150 mm,
220 nm).
* n = 6.
527

Journal of Chromatographic Science, Vol. 40, October 2002
8. I. Ferrer and D. Barceló. Determination and stability of pesticides in
freeze-dried water samples by automated on-line solid-phase fol-
lowed by liquid chromatography. J. Chromatogr. A 737: 93–99
(1996).
9. G. Sacchero, C. Sarzanini, and E. Mentasti. On-line preconcentration
and ion chromatography of triazine compounds. J. Chromatogr. A
671: 151–57 (1994).
10. A. Junker-Buchheit and M. Witzenbacher. Pesticide monitoring of
drinking water with the help of solid-phase extraction and high-per-
formance liquid chromatography. J. Chromatogr. A 737: 67–74
(1996).
11. A. Martin-Esteban, P. Fernández, and C. Cámara. New design for the
on-line solid-phase extraction of pesticides using membrane extrac-
tion disk material and liquid chromatography. J. Chromatogr. 752:
291–97 (1996).
12. C. Aguilar, I. Ferrer, F. Borrull, R.M. Marcé, and D. Barceló.
Monitoring of pesticides in river water based on samples previously
stored in polymeric cartridges followed by on-line solid-phase
extraction. Anal. Chim. Acta. 386: 237–48 (1999).
13. S. Lacorte, I. Guiffard, D. Fraisse, and D. Barceló. Broad spectrum
analysis of 109 priority compounds listed in the 76/464/CEE Council
Directive using solid-phase extraction and GC/EI/MS. Anal. Chem.
72: 1430–40 (2000).
14. M. Berg, S.R. Müller, and R.P. Schwarzenbach. Simultaneous deter-
mination of triazines including atrazine and their major metabolites
hydroxiatrazine, desethylatrazine and deisopropylatrazine in natural
waters. Anal. Chem. 67: 1860–65 (1995).
15. V. Pichon, L. Chen, S. Guenus, and M.C. Hennion. Comparison of
sorbents for the solid-phase extraction of the highly polar degrada-
tion products of atrazine (including ammeline, ammelide and cya-
nuric acid). J. Chromatogr. A 711: 257–67 (1995).
between 0.008–0.012 µg/L and 0.027–0.041 µg/L, respectively.
Because of this, the method could be used for the determination
of samples with a concentration lower than the limit established
by the legislation (0.1 µg/L for individual pesticides). Compared
with other authors, it can be noted that the detection and quanti-
tation limits are similar and even better than those obtained with
other SPE–LC–UV detection analyses. Aguilar et al. (12) deter-
mined a sequence of pesticides including simazine and atrazine
by SPE–LC–diode array coupled online, reaching a detection
limit of 0.1 µg/L for both pesticides. Pinto et al. (22) determined
simazine and atrazine by SPE–LC–UV with detection limits of
0.012 and 0.018 µg/L, respectively. The results obtained are even
comparable with those obtained by GC. Choudhury et al. (2)
determined 39 pesticides, including the studied chlorotriazines,
by solid-phase microextraction–GC–MS obtaining detection
limits ranging from 0.010 to 0.030 µg/L (2).
This method offers an important advantage with respect to
multiresidual analysis such as its lower analysis time. For a
1-mL/min flow rate the chromatographic determinations take
8 min (and 6 min using a 1.5-mL/min flow rate). They contrast
with their determination part of a multiresidual analysis, as for
example in Junker-Buchleit et al. (10) in which propazine elutes
at 43 min. Di Corcia et al. (16) reported atrazine elutes at 14.4
min. These results also show a faster elution than that indicated
by Pinto et al. (22), who determined elution times for simazine
and atrazine at 3.24 and 4.62 min, respectively.
16. A. Di Corcia and M. Marchetti. Multiresidue method for pesticides in
drinking water using a graphitised carbon black cartridge extraction
and liquid chromatography. Anal. Chem. 63: 580–85 (1991).
17. Calidad del Agua-Medioambiente. Tomo 1. Asociación Española de
Normalización y Certificación (AENOR), Ed. 1997. ISO 5667-
3:1994.
References
18. A. Noble. Partition coefficients (n-octanol–water) for pesticides.
J. Chromatogr. 642: 3–14 (1993).
19. M.V. Dabrio, G.P. Blanch, A. Cienfuentes, J.A. Díez-Masa, M. de
Frutos, M. Herraiz, I. Martínez Castro, and J. Sanz Perucha.
Cromatografía y Electroforesis en Columna. Springer-Verlag Ibérica,
Barcelona, Spain, 2000.
20. D. Barceló, G. Durand, V. Bouvot, and M. Nielen. Use of extraction
disks for trace enrichment of various pesticides from river water and
simulated seawater samples followed by liquid chromatography–
rapid-scanning UV–visible and thermospray-mass spectrometry
detection. Environ. Sci. Technol. 27: 271–77 (1993).
21. M.S. Mills and E.M. Thurman. Mixed-mode isolation of triazine
metabolites form soil and aquifer sediments using automated solid-
phase extraction. Anal. Chem. 64: 1985–90 (1992).
1. A. Martin-Esteban, P. Fernández, A. Fernández-Alba, and C. Cámara.
Analysis of polar pesticides in environmental waters: a review.
Quím. Anal. 17: 51–66 (1998).
2. H. Sabik, R. Jeannot, and B. Rondeau. Multiresidue methods using
solid-phase extraction techniques for monitoring priority pesticides,
including triazines and degradation products. J. Chromatogr. A 885:
217–36 (2000).
3. V. Pacakova, K. Stulik, and J. Jiskra. High-performance separations in
the determination of triazine herbicides and their residues.
J. Chromatogr. A 754: 17–31 (1996).
4. Council Directive 98/83/EC of 3 November 1998 on the quality of
water intended for human consumption. Off. J. L 330: 32–54 (1998).
5. M.C. Hennion. Sample Handling and Trace Analysis of Pollutants
Techniques, Applications and Quality Assurance. D. Barceló, Ed.
Elsevier, The Netherlands, 2000, Chapter 1, pp. 3–73.
6. D. Barceló. Sample Handling and Trace Analysis of Pollutants
Techniques, Applications and Quality Assurance. D. Barceló, Ed.
Elsevier, The Netherlands, 2000, Chapter 1, pp. 155–207.
7. V. Pichon. Multiresidue solid-phase extraction for analysis for trace-
analysis of pesticides and their metabolites in environmental water.
Pest. Anal. 26(6): M91–M98 (1998).
22. G.M.F. Pinto and I.C.S.F. Jardim. Use of solid-phase extraction and
high-performance liquid chromatography for the determination of
triazine residues in water: validation of the method. J. Chromatogr. A
869: 463–69 (2000).
Manuscript accepted June 28, 2002.
528