Accepted Manuscript Title: Development of a new microextraction method based on elevated temperature dispersive liquid-liquid microextraction for determination of triazole pesticides residues in honey by gas chromatography-nitrogen phosphorus detection Author: Mir Ali Farajzadeh Mohammad Reza Afshar Mogaddam Houshang Ghorbanpour PII: DOI: Reference:
S0021-9673(14)00657-8 http://dx.doi.org/doi:10.1016/j.chroma.2014.04.067 CHROMA 355359
To appear in:
Journal of Chromatography A
Received date: Revised date: Accepted date:
9-12-2013 20-4-2014 22-4-2014
Please cite this article as: M.A. Farajzadeh, M.R.A. Mogaddam, H. Ghorbanpour, Development of a new microextraction method based on elevated temperature dispersive liquid-liquid microextraction for determination of triazole pesticides residues in honey by gas chromatography-nitrogen phosphorus detection, Journal of Chromatography A (2014), http://dx.doi.org/10.1016/j.chroma.2014.04.067 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Development of a new microextraction method based on elevated temperature dispersive liquid-
2
liquid microextraction for determination of triazole pesticides residues in honey by gas
3
chromatography-nitrogen phosphorus detection
ip t
4
Mir Ali Farajzadeh*a, Mohammad Reza Afshar Mogaddama, Houshang Ghorbanpourb
6
a
7
b
cr
5
us
Department of Analytical Chemistry, Faculty of Chemistry, University of Tabriz, Tabriz, Iran
Food and Drug Laboratories, Tabriz University of Medical Sciences, Tabriz, Iran
*Corresponding author: M. A. Farajzadeh
M
9
an
8
Tel.: +98 411 3393084
11
Fax: +98 411 3340191
12
E-mail address:
[email protected];
[email protected] 14 15 16 17
te
Ac ce p
13
d
10
18 19 20
Abstract 1 Page 1 of 30
In the present study, a rapid, highly efficient, and reliable sample preparation method named “elevated
22
temperature dispersive liquid-liquid microextraction” followed by gas chromatography-nitrogen-
23
phosphorus detection was developed for the extraction, preconcentration, and determination of five
24
triazole pesticides (penconazole, hexaconazole, diniconazole, tebuconazole, and difenoconazole) in honey
25
samples. In this method the temperature of high-volume aqueous phase was adjusted at an elevated
26
temperature and then a disperser solvent containing an extraction solvent was rapidly injected into the
27
aqueous phase. After cooling to room temperature, the phase separation was accelerated by
28
centrifugation. Various parameters affecting the extraction efficiency such as type and volume of the
29
extraction and disperser solvents, temperature, salt addition, and pH were evaluated. Under the optimum
30
extraction conditions, the method resulted in low limits of detection and quantification within the range
31
0.05 – 0.21 ng g-1 in honey (15 - 70 ng L-1 in solution) and 0.15–1.1 ng g−1 in honey (45 - 210 ng L-1 in
32
solution), respectively. Enrichment factors and extraction recoveries were in the ranges of 1943 – 1994
33
and 97 – 100 %, respectively. The method precision was evaluated at 1.5 ng g-1 of each analyte, and the
34
relative standard deviations were found to be less than 4 % for intra-day (n = 6) and less than 6 % for
35
inter-days. The method was successfully applied to the analysis of honey samples and difenoconazole was
36
determined at ng g−1 levels.
38 39 40
cr
us
an
M
d
te
Ac ce p
37
ip t
21
41
Keywords: Elevated-temperature dispersive liquid-liquid microextraction; Honey; Triazole pesticides;
42
Gas chromatography; Nitrogen-phosphorus detector
43 44
1. Introduction
2 Page 2 of 30
Honey is a wholesome natural product consumed worldwide. The nutritional and quality aspects of honey
46
are important since they are among the significant attributes that affect consumer acceptance. Of even
47
more significance is chemical safety of honey as it affects human health. So there is an increasing interest
48
in monitoring honey for the presence of pesticides and other harmful chemical compounds. According to
49
European Union (EU) regulations, honey must be free of chemical contamination, particularly those due
50
to the presence of pesticides.
51
Maximum residue limits (MRLs) for some triazole pesticides in royal, jelly, pollen, and honeycomb are
52
under statutory regulations by the EU Council (diniconazole, 0.01 mg kg-1, Reg. (EU) No. 899/2012;
53
tebuconazole, 0.05mg kg-1, Reg. (EU) No. 500/2013; and difenoconazole, 0.05 mg kg-1, Reg. (EU) No.
54
834/2013) [1]. Considering the fact that beehives are frequently pastured on plants and agricultural crops
55
contaminated by pesticides, there is a need for accurate and reliable determination of pesticide residues in
56
honey products.
57
Triazole fungicides are among the flourishing new generations of pesticides applied to fruits, vegetables,
58
and grain crops [2]. Besides their antifungal activity, they are also of concern as a group of compounds
59
that disturb endocrine activity in human beings. Due to their lipophilic nature, these compounds can be
60
bio-accumulated in various tissues of living organisms and they can be transported between various
61
compartments of ecosystems and contaminate food chains.
62
Sample preparation plays a key role in the analysis of pesticide residues in complex matrices such as
63
those found in honey samples [3]. The main objective of this challenging critical step is to transfer the
64
analytes into a phase in which they are pre-purified, concentrated, and compatible with the analytical
65
system [4, 5]. Traditionally, the extraction and enrichment of analytes from the sample matrix are often
66
accomplished by procedures such as liquid-liquid extraction (LLE) [6, 7] and solid-phase extraction
67
(SPE) [8, 9]. However, these traditional pretreatment methods suffer from disadvantages such as
68
demanding intensive labor, being time-consuming, resulting in unsatisfactory enrichment factors (EFs),
69
and consuming large quantities of toxic solvent(s), which compel analysts to limit their application.
70
Recent research on sample pretreatment and preparation methods have being oriented toward the
Ac ce p
te
d
M
an
us
cr
ip t
45
3 Page 3 of 30
development of efficient, economical, and miniaturized methods. As a result of this, solid-phase
72
microextraction (SPME) [10, 11], and liquid-phase microextraction (LPME) [12, 13] were developed.
73
SPME, a technique introduced in 1990 by Pawliszyn [14, 15], was based on equilibration of analyte(s)
74
between the sample matrix and a fused silica fiber coated with an adsorbent [16-21]. However, most of
75
the commercial extractive fibers used in SPME were relatively expensive and fragile and occurrence of
76
sample carry-over further complicated the problem with them [22]. LPME methods such as single-drop
77
microextraction (SDME) and hollow-fiber supported LPME (HF-HPME) were developed as solvent-
78
minimized sample pretreatment techniques that were inexpensive and caused minimal exposure to toxic
79
organic solvents [23-26].
80
In 2006, a microextraction technique termed dispersive liquid–liquid microextraction (DLLME) was
81
developed by Rezaee et al [27]. Similar to homogeneous liquid–liquid extraction and cloud-point
82
extraction, it was based on a ternary component solvent system. In this method, an appropriate mixture of
83
an extraction solvent and a dispersive solvent was rapidly injected by a syringe into aqueous sample
84
which resulted in the formation of a cloudy solution. Then the analytes were rapidly extracted into the
85
fine droplets of extraction solvent. After extraction, phase separation was performed by centrifugation and
86
the analytes were enriched in the organic phase and determined by a chromatographic or
87
spectrophotometric method. The advantages of the DLLME method were simplicity of operation,
88
rapidity, low cost, and high extraction recoveries (ERs) and EFs [28-31]. However, the relatively high
89
volumes (in the mL range) of polar solvent (e.g. methanol or acetonitrile) consumed as dispersive solvent
90
lead to lower extraction efficiencies because of increased solubility of the analytes in the solution. In
91
2008, Zhou et al [32] developed a novel ionic liquid (IL) LPME method termed temperature-controlled
92
ionic liquid dispersive LPME. The method was based on the dispersion of IL into aqueous phase by
93
changing the temperature. Many analytical methods have been applied to measure pesticides in honey
94
samples, mainly including high-performance liquid chromatography (HPLC) with different detectors such
95
as variable wavelength detector (VWD) [33], diode-array detector (DAD) [34], and tandem mass
Ac ce p
te
d
M
an
us
cr
ip t
71
4 Page 4 of 30
spectrometer (MS-MS) [35], and gas chromatography (GC) with detectors such as MS [36], flame
97
ionization detector (FID) [37], and electron capture detector (ECD) [38].
98
The goal of this study was to develop a sensitive procedure for the trace determination of triazole
99
pesticides in honey samples using elevated-temperature dispersive liquid-liquid microextraction (ET-
100
DLLME) combined with GC-nitrogen-phosphorous detection (NPD) method. In this method, large
101
volumes of aqueous phase, along with small volumes of extracting phase improved ERs. Temperature can
102
have an important effect in DLLME method and help reaching higher EFs and ERs in spite of very large
103
volume ratio of aqueous phase to organic phase, because higher temperatures can be a driving force for
104
better dispersion of extraction solvent in the aqueous phase. The main disadvantage of the DLLME
105
technique lies in its extractant solvent which is usually a halogenated solvent of highly toxic nature that is
106
difficult to handle in the laboratory. Furthermore, 1,1,2,2-tetrachloroethane (1,1,2,2-TCE) has
107
considerable hepatotoxicity and 1,2-dibromoethane (1,2-DBE) is classified by IRAC as Group 2A,
108
suspected carcinogen to humans with evidence of carcinogenicity in animals [39]. To the best of our
109
knowledge, this is the first report on application of ET-DLLME to the determination of triazole pesticides
110
using large-volume aqueous sample. The proposed method was successfully applied to the quantification
111
of residues of some selected triazole pesticides in honey samples of different floral origins.
cr
us
an
M
d
te
Ac ce p
112
ip t
96
113
2. Experimental
114
2.1 Chemicals and solutions
115
All triazole pesticides used (penconazole, hexaconazole, diniconazole, tebuconazole, and difenconazole)
116
with purity of >98% were kindly provided by GYAH Corporation (Karadj, Iran). The tested extraction
117
solvents were supplied by the following sources: 1,2-DBE was from Merck (Darmstadt, Germany),
118
1,1,2,2-TCE, and 1,1,2,2-tetrabromoethane (1,1,2,2-TBE) were from Janssen Chimica (Beerse, Belgium).
119
Acetonitrile, acetone, methanol, dimethylformamide (DMF), dimethyl sulfoxide (DMSO), and n-propanol
120
tested as disperser solvents were from Merck. Analytical-reagent grade sodium chloride, hydrochloric 5 Page 5 of 30
acid, and sodium hydroxide were also obtained from Merck. De-ionized water (Ghazi Company, Tabriz,
122
Iran) was used for the preparation of aqueous solutions.
123
A stock solution of pesticides (1000 mg L-1 of each pesticide) was prepared by dissolving an appropriate
124
amount of each pesticide in acetone. Working solutions were prepared daily by appropriate dilutions of
125
the stock solution with de-ionized water. Another standard solution of analytes was prepared in 1,2-DBE
126
at a concentration of 100 mg L-1 (each pesticide). This solution was directly injected into the
127
chromatographic system three times a day for quality control and areas of the obtained peaks were used in
128
calculation of EFs and ERs.
129
2.2 Samples
130
Four honey samples of different floral origins were purchased from local vendors (East Azarbaijan
131
Province, Iran). One further honey sample was obtained from beehives located in virgin mountainous
132
lands which are far away from the agricultural areas (Kaleybar, East Azarbaijan Province, Iran). It seems
133
plausible to assume such honey to be free of any pesticides. Some preliminary tests performed on the
134
basis of our previous works confirmed plausibility of this assumption. So it was used as a pesticide-free
135
sample in optimization of the proposed method. All samples were stored in their original containers at
136
ambient temperature just like normal storage conditions in their everyday use. To prepare aqueous
137
samples, 15.0 g honey was dissolved in de-ionized water and the obtained homogeneous solution was
138
brought to 50 mL by water. This solution was left to equilibrate for at least 15 min prior to performing the
139
proposed extraction method. This solution was directly subjected to the extraction procedure without
140
filtration or any other pretreatment.
141
2.3 Apparatus
142
Chromatographic analyses were performed on a gas chromatograph (GC-1000, Dani, Italy) equipped with
143
a splitless/split injector operated at 290 ◦C in splitless mode (sampling time 1 min) and an NPD. Helium
144
(99.999%, Gulf Cryo, United Arabic Emirates) was used as the carrier gas (at a constant linear velocity of
Ac ce p
te
d
M
an
us
cr
ip t
121
6 Page 6 of 30
30 cm s-1) and make-up gas (25 mL min-1). Chromatographic separations were achieved on a BPX-5
146
capillary column (5% phenyl methyl siloxane, 95% dimethyl siloxane, 30 m × 0.25 mm i.d., and film
147
thickness 0.25 μm) (SGE, Australia) with the following oven temperature programming: initial
148
temperature 100 ◦C (held 2 min), ramped at 10 ◦C min-1 to 300 ◦C and held at 300 ◦C for 5 min. The NPD
149
temperature was maintained at 300 ◦C. For NPD, hydrogen gas was generated with a hydrogen generator
150
(GLAIND-2200, Dani, Italy) at a flow rate of 5 mL min-1. Air was used at a flow rate of 95 mL min-1. Gas
151
chromatography-mass spectrometry (GC-MS) analyses were carried out on an Agilent 7890A gas
152
chromatograph with a 5975C mass-selective detector (Agilent Technologies, CA, USA) and a splitless
153
injector operated at 300 ◦C in a splitless/split mode (sampling time 1 min). The separation was carried out
154
on an HP-5 MS capillary column (30 m × 0.25 mm i.d. and film thickness 0.25 μm) (Hewelett-Packard,
155
Santa Clara, USA). The injector temperature and column oven temperature programming were the same
156
as that of GC-NPD analyses mentioned above. Helium was used as carrier gas at a flow rate of 1.0 mL
157
min-1. For GC-MS, the following ions were selected: m/z 159, 213, and 247 for penconazole; m/z 83, 214,
158
and 231 for hexaconazole; m/z 70, 232, and 268 for diniconazole; m/z 85, 125, and 250 for tebuconazole;
159
and m/z 207, 265, and 323 for difenoconazole. Hettich centrifuge (Model ROTOFIX 32A, Germany) was
160
used for accelerating phase separation.
cr
us
an
M
d
te
Ac ce p
161
ip t
145
162
2.4. Procedure
163
For this extraction procedure, 50 mL diluted pesticide-free honey sample spiked with 25 ng g-1 of each
164
pesticide or 50 mL diluted sample of potentially contaminated honey was transferred into a 70-mL glass
165
conical-bottom centrifuge tube. The tube was held in a water bath at 75 ◦C for 4 min. Then, the mixture of
166
extraction solvent and disperser, consisting of 1.5 mL DMF (as a disperser solvent) and 130 µL 1,2-DBE
167
(as an extractant), was rapidly injected in one step into the solution using a 5-mL glass syringe.
168
Consequently, 1,2-DBE was dispersed completely in all parts of aqueous solution without any need for
169
agitation. The solution was cooled with tap water for 3 min. By cooling, turbidity of the solution
170
increased gradually. Then the solution was centrifuged at a rate of 4000 rpm for 5 min. During this 7 Page 7 of 30
process, 1,2-DBE was settled down at the bottom of conical test tube (25 ± 1 µL). It should be noted that
172
major proportion of the extraction solvent (105 µL) was lost due to dissolution in the aqueous phase and
173
adherence of droplets onto the inner walls of the tube. Finally, aqueous phase was completely removed
174
using a 50-mL syringe and then 1 µL of the extractant was taken and injected into the GC system for
175
analysis.
ip t
171
cr
176 2.5 Calculation of EFs and ERs
178
EF was defined as the ratio of the analyte concentration in the sedimented phase (Csed) to the initial
179
concentration of analyte (C0) in the sample:
an
us
177
M
180
In practice, Csed was determined by comparing areas of peaks obtained from two distinct injections into
182
chromatographic system: (i) direct injection of pesticides standard solution prepared in the extraction
183
solvent, and (ii) injection of the sedimented phase into GC. ER was defined as the percentage of the total
184
amount of analyte (n0) which was extracted into the sedimented phase:
te
Ac ce p
185
d
181
(2)
186
where nsed is the amount of the analyte which was extracted into the sedimented phase, and Vsed and Vaq
187
are the volumes of sedimented phase and sample solution, respectively.
188 189
3. Results and discussion
190 191
In this work, a DLLME method performed at an elevated temperature to determine triazole pesticides in
192
honey samples. For obtaining the maximum extraction efficiency, some important experimental
193
parameters that would influence the performance of DLLME method was investigated in detail. The 8 Page 8 of 30
194
studied parameters included type and volume of extraction and dispersive solvents, temperature, sample
195
pH, and ionic strength of aqueous phase.
196 3.1. Selection of extraction solvent
198
In order to obtain high extraction efficiencies in DLLME, selection of an appropriate extraction solvent is
199
of vital importance. Generally, the extraction solvent used in DLLME must fulfill some requirements: it
200
should have a density higher or lower than water, be sparingly soluble in water, have high extraction
201
capability for the target analytes, have good chromatographic behavior, and finally it should be easily
202
dispersed into aqueous phase during dispersion. Based on the above requirements and by considering the
203
fact that in this study the DLLME was to be performed at an elevated temperature, some organic solvents
204
having relatively high boiling points, namely 1,1,2,2-TBE (b.p. 244 ◦C), 1,2-DBE (b.p. 133 ◦C), and
205
1,1,2,2-TCE (b.p. 146 ◦C), were tested as potential extraction solvents. A series of experiments were
206
carried out with different volumes of extraction solvents and a constant volume of aqueous sample
207
solution (50 mL) to achieve a similar volume of the sedimented phase (25 ± 1 µL). The obtained volumes
208
were 140 µL of 1,1,2,2-TCE, 60 µL of 1,1,2,2-TBE, and 130 µL of 1,2-DBE. The extraction efficiencies
209
for different extraction solvents were shown in Fig. 1. The results revealed that 1,1,2,2-TCE and 1,2-DBE
210
gave higher extraction efficiencies than those of 1,1,2,2-TBE. By considering the relatively lower solvent
211
consumption by using 1,2-DBE rather than 1,1,2,2-TCE (130 vs. 140 µL) and its relatively higher
212
extraction efficiencies in the case of some analytes, it was selected as the extraction solvent for the
213
subsequent experiments.
Ac ce p
te
d
M
an
us
cr
ip t
197
214
Fig. 1
215
3.2 Selection of disperser solvent
216
In order to achieve high preconcentration of analytes, type of the disperser solvent is very important. The
217
potentially applicable solvents must have appropriate miscibility with both extraction solvent and sample 9 Page 9 of 30
solution and produce a distinct cloudy solution as a result of dispersion of extraction solvent into the
219
aqueous phase. According to these requirements and because of the high temperatures involved in ET-
220
DLLME, three disperser solvents of high boiling point including DMF, DMSO, and n-propanol were
221
tested for their efficiencies (using 1.5 mL of each disperser). Volume of the sedimented phase for all
222
dispersers was 25 ± 1 µL. The extraction efficiencies for the analytes obtained by using different disperser
223
solvents were given in Fig. 2. The results showed that among the tested solvents, DMF displayed the
224
highest ERs for all analytes. So DMF was selected for the subsequent studies because of its capability of
225
forming better cloudy state with very fine droplets which resulted in higher ERs for the analytes.
us
cr
ip t
218
226
Fig. 2
3.3 Effect of extraction solvent volume
228
Volume of the extraction solvent used can affect volume of the sedimented organic phase, repeatability of
229
results, and extraction efficiencies. By changing volume of the extraction solvent (with keeping sample
230
size constant), the volume ratio of sample to extractive phase varies and hence ERs of the analytes may
231
also change. To evaluate the effect of extraction solvent volume, different volumes of 1,2-DBE (115, 125,
232
130, 145, and 165 µL) were dissolved in a constant volume of DMF (1.5 mL) and the obtained mixtures
233
were used in exactly the same DLLME procedure. Figure 3 shows the variations of ERs vs. volume of the
234
extraction solvent. It should be noted that volume of the sedimented phase increased from 10 to 80 µL by
235
increasing volume of 1,2-DBE from 115 to 165 µL. According to the figure, by increasing volume of 1,2-
236
DBE, the ERs increased till 130 µL and then remained nearly constant (~100 %) for all pesticides.
237
Therefore, 130 µL was selected as the optimum volume of 1,2-DBE.
M
d
te
Ac ce p
238
an
227
Fig. 3
239
3.4 Effect of disperser solvent volume
240
In a DLLME technique, volume of disperser solvent should also be optimized. Volume of disperser
241
solvent affects solubility of extraction solvent in aqueous phase and volume of the settled phase which in
242
turn directly affect the ERs and EFs. On the other hand, upon using small volumes of the disperser
243
solvent, the cloudy state is not formed satisfactorily and thereby the DLLME procedure is disturbed. By 10 Page 10 of 30
using large volumes of the disperser, solubility of analytes in aqueous phase increases and hence the
245
extraction efficiency decreases. To acquire the optimal volume of disperser solvent, the volumes of DMF
246
and 1,2-DBE were changed simultaneously to obtain a fixed volume of the sedimented phase. While the
247
other experimental conditions were kept constant, different volumes of DMF (0.5, 1.0, 1.5, 2.0, and 3.0
248
mL containing 105, 115, 130, 145, and 155 µL of 1,2-DBE, respectively) were examined. Under these
249
conditions, the volume of the sedimented phase was constant at 25 ± 1 µL. The results showed that (Fig.
250
4) for all selected pesticides ERs increased rapidly till 1.5 mL and then remained nearly constant by
251
increasing the volume of DMF. By using lower volumes of DMF, the cloudy state was not formed well;
252
thereby the ERs were low, so 1.5 mL was selected as the optimum volume for the disperser.
us
cr
ip t
244
an
253
Fig. 4
3.5 Effect of elevating the temperature
255
Modifying the temperature can affect the extraction rate in extractive methods by changing diffusion
256
coefficients. Moreover, heating can be a driving force for better dispersion of extraction solvent in the
257
aqueous solution so that the contact area between extractant and sample increases and hence mass transfer
258
rates of analytes are improved. The effect of temerature was studied within the range of 23-80 ◦C. The
259
results shown in Fig. 5 revealed that by increasing temperature, the ERs increased till 70 ◦C and then
260
remained almost constant. At lower temperatures the dispersion of 1,2-DBE in aqueous phase was poor
261
and the diffusion rate of analytes could be low so that mass transfer resistance could occur, but at higher
262
temperatures the diffusion coefficients and mass transfer rates increased. So the amounts of extracted
263
analytes increased with the rise of temperature. However, as it is well known, the rise of temperature also
264
would increase the migration of analytes from the extractant into aqueous phase, i.e. the rise of
265
temperature would have a dual function: increasing the transfer of analyte(s) into the extractant and at the
266
same time enhancing the migration of analyte(s) out of the extractant. The migration rates and solubility
267
variations of analytes in organic phase and aqueous phase by temperature can vary the distribution
268
coefficients of analytes. Therefore different ERs were obtained at different temperatures. Also it should
269
be noted that cooling of solution after dispersion of extractant into aqueous phase played an important
Ac ce p
te
d
M
254
11 Page 11 of 30
role in obtaining high efficiencies for the proposed method at high temperatures. Upon cooling, turbidity
271
of the solution increased which showed that new droplets of extractant were produced. This caused the
272
extraction process to perform similar to the continuous extraction. This would be a reason for the
273
extraction during the proposed DLLME procedure in spite of the very high volume ratio of sample to
274
extractant (50 mL vs. 25 µL). Based on the above mentioned facts, 75 ◦C was selected for the next
275
experiments. On the other hand, an elevated temperature (75 ◦C) for a relatively long time (4 min) was
276
likely to facilitate hydrolysis of the pesticides. In order to investigate hydrolysis of the analytes, heating
277
time was studied in the range of 0-10 min at 75 ◦C. With 0 min the solution was centrifuged immediately
278
after injecting mixture of disperser and extraction solvents into aqueous solution. By increasing heating
279
time the analytical signals increased till 3 min and then remained constant up to 10 min. The results (not
280
given here) showed that hydrolysis of the pesticides are negligible during the extraction procedure.
an
us
cr
ip t
270
M
281
Fig. 5
3.6 Optimization of solution pH
283
The pH of aqueous sample solution usually affects the extraction performance. The effect of sample pH
284
was evaluated over the range of 2-12 (at 2-unit intervals) with adjusting pH by 1 M solutions of HCl or
285
NaOH. The efficiency of the method was pH-independent in the pH range of 4-8, whereas a decrease in
286
ERs was observed at pH 2, 10, and 12. This decrease can be attributed to hydrolysis of the pesticides at
287
highly acidic or alkaline pH. It should be noted that the pH of all samples used in this study was between
288
4 and 8, therefore no attempts were made to adjust pH.
te
Ac ce p
289
d
282
290
3.7 Optimization of centrifugation time and speed
291
Centrifugation is a mandatory step in DLLME to accelerate the collection of extractant droplets. Time and
292
speed of centrifugation were optimized by investigating their effect in the ranges of 2-7 min and 2000-
293
5000 rpm, respectively. The obtained results showed that the efficiency of the method increased with the
294
increasing centrifugation time and speed up to 5 min and 4000 rpm, respectively. The results were normal
295
and reasonable because after the complete separation of organic phase from the sample solution, longer 12 Page 12 of 30
296
centrifugation could not play any further role. Therefore, 5 min and 4000 rpm were selected in the further
297
experiments for centrifuging time and speed, respectively,
298 3.8. Analytical features of the proposed method
300
Analytical characteristics of the method in determination of the target analytes under the optimized
301
conditions were evaluated according to the recommended procedure for estimating figures of merit. The
302
analytical performance of the proposed method was evaluated in terms of linear range, correlation
303
coefficient (r), repeatability, EF, ER, and limit of detection (LOD) and quantification (LOQ) which were
304
calculated on the basis of signal to noise ratio (S/N) of 3 and 10, respectively. The obtained results were
305
summarized in Table 1. Good linearity ranges were obtained for the calibration graphs, with correlation
306
coefficients higher than 0.993. The LODs ranged from 0.05 to 0.21 ng g-1 and 15 to 70 ng L-1 in honey
307
and solution, respectively, and the LOQs ranged from 0.15 to 1.1 ng g-1 and 45 to 210 ng L-1 in honey and
308
solution, respectively, which are of significantly low values. The EFs and ERs for the five pesticides
309
ranged from 1943 to 1994 and from 97 to 100%, respectively. It should be noted that the maximum EF
310
which could be theoretically achieved in this study was 2000. All obtained EFs were completely near to
311
the maximum theoretically calculated values. Precision of the method was determined by analyzing the
312
spiked samples contained with 1.5 ng g-1 of the analytes. Relative standard deviations (RSDs) were in the
313
ranges 3 - 4 % and 4 - 6 % for intra-day (n=6) and inter-day (n=4) determinations, respectively, which
314
indicated that the method was satisfactorily repeatable. These excellent results showed the proposed
315
method has very high sensitivity and stability and would have a tremendous potential to be widely used
316
for the analysis of such pesticides at trace levels in honey.
Ac ce p
te
d
M
an
us
cr
ip t
299
317
Table 1
318
3.9 Real samples analysis
319
The developed method was applied to the analysis of four honey samples of different floral origins being
320
commercially available in East Azarbaijan Province (Iran). Fig. 6 depicts the typical GC-NPD
13 Page 13 of 30
chromatograms of honey sample, honey sample spiked with 10 ng g! 1 of each pesticide, and standard
322
solution (100 mg L−1 of each pesticide in 1,2-DBE). In the chromatogram of honey sample a peak
323
emerged at the retention time corresponding to difenoconazole. It should be noted that difenoconazole has
324
both cis and trans isomers, therefore its peak in chromatograms was appeared as a split peak. Although
325
the detector used, i.e. NPD, is a selective detector and the peak shape (spilt) in honey sample
326
chromatogram was similar to that of difenoconazole in standard solution, for further confirmation, the
327
honey sample was also injected into GC-MS after performing the proposed sample preparation method
328
(Fig. 7). The presence of difenoconazole in the studied honey sample was confirmed by comparison of
329
mass data for scan 3302 (retention time 28.5 min) with those of the studied pesticide. Four ions, observed
330
at m/z 265, 267, 323, and 325 in both mass spectrum of the compound eluted in the retention time of 28.5
331
min and in mass spectrum of difenoconazole, confirmed presence of this pesticide in the studied samples.
332
Also the same retention times which were observed for difenoconazole in standard solution and in
333
samples injected into the GC-MS apparatus were considered as further evidence for the presence of
334
difenoconazole in the studied samples. Moreover, using a microextraction method along with GC-NPD or
335
GC-MS provided a relatively selective method for the determination of the target analytes in
336
honey.Difenoconazole was detected and determined on the basis of GC-NPD data in two of the honey
337
samples. The obtained mean concentrations along with their standard deviations were 8.0 ± 0.35 and 19 ±
338
0.82 ng g-1 (n = 3). The other two honey samples were free of pesticides.
340
cr
us
an
M
d
te
Ac ce p
339
ip t
321
Fig. 6 Fig. 7
341
3.10 Matrix effects
342
The influence of matrix effect on the detection response of analytes is a well-known subject in pesticide
343
residue analysis. This can result in an enhanced or decreased analyte signal from extracts obtained in the
344
presence of matrix compared to those obtained in the absence of the matrix. In order to evaluate the
345
matrix effect, the samples were spiked with the analytes at three levels (10, 25, and 50 ng g−1 of each
346
pesticide) and the proposed method was applied to them (three replications for each concentration). The 14 Page 14 of 30
recoveries obtained for the analytes in honey samples in comparison with those obtained from the de-
348
ionized water samples spiked at the same three concentration levels were given in Table 2. The results
349
demonstrated that all obtained recoveries were between 96-99 % and the matrices of real samples had no
350
effect on the efficiency of the proposed method. Therefore there was no need to perform any additional
351
treatments.
ip t
347
352
cr
Table 2
3.11 Comparison of the proposed method with other approaches
354
The efficiency of the presented ET-DLLME method for the selected analytes was compared with those of
355
other methods reported in the literature, in terms of features such as RSD, LOQ, LOD, and EF and the
356
results were summarized in Table 3. The RSD values for the proposed method were lower than those of
357
the other methods. It was found that the LODs of the developed method were completely lower than or
358
comparable to those of the other methodologies such as GC-ECD or GC-MS. EFs near to 2000 were
359
achievable by the present method which was very higher than EFs of the other methods, except with
360
respect to penconazole when compared to the second method. By considering these results, the newly
361
developed method can be considered to be a rapid, sensitive, efficient, reliable, and easy to use technique
362
for the extraction and highly efficient preconcentration of the triazole pesticides from the honey samples.
363
Table 3
Ac ce p
te
d
M
an
us
353
364
4. Conclusion
365
In this paper, for the first time, an improved microextraction method based on ET-DLLME was
366
developed for the preconcentration of triazole pesticides found in honey samples and the method was
367
combined with GC-NPD for quantitative analysis of the pesticides. The experimental results
368
demonstrated that this technique exhibits many merits such as excellent ERs and EFs, very low LODs,
369
excellent sensitivity, shorter extraction time, and better repeatability and reproducibility. The excellent
370
performance of the method for the analysis of the honey samples showed that it can be successfully
371
applied to relatively complex matrices. By considering these advantages, the developed method can be 15 Page 15 of 30
considered as a high-performance technique for the determination of ultra-trace triazole pesticides in
373
honey samples.
374
Acknowledgments
375
The authors thank the Research Council of University of Tabriz for financial support.
ip t
372
cr
376 References
378
[1] http://ec.europa.eu/sanco_pesticides/public/?event=homepage, access date, 2/3/2014.
379
[2] B. Kmell´ar, L. Abrank´o´o, P. Fodor and S.J. Lehotay, Food Addit. Contam., Part A, 27 (2010) 1415.
380
[3] M. Kahle, I.J. Buerge, A. Hauser, M.D. Muller, T. Poiger, Environ. Sci. Technol. 42 (2008) 7193.
381
[4] R. Rial-Otero, E.M. Gaspar, I. Moura, J.L. Capelo, Talanta 71 (2007) 503.
382
[5] S. Ulrich, J. Chromatogr. A 902 (2000) 167.
383
[6] M.B. Melwanki, M.R. Fuh, J. Chromatogr. A 1198-1199 (2008) 1.
384
[7] K.B. Borges, E.F. Freire, I. Martins, M.E. Siqueira, Talanta 78 (2009) 233.
385
[8] Y.M. Park, H. Pyo, S.J. Park, S.K. Park, Anal. Chim. Acta 548 (2005) 109.
386
[9] A.L. Saber, Talanta 78 (2009) 295.
387
[10] C.E. Banos, M. Silva, Talanta 77 (2009) 1597.
388
[11] M.F. Alpendurada, J. Chromatogr. A 889 (2000) 3.
389
[12] A. Penalver, E. Pocurull, F. Borrull, R.M. Marce, Trends Anal. Chem. 18 (1999) 557.
390
[13] E. Psillakis, N. Kalogerakis, Trends Anal. Chem. 22 (2003) 565.
391
[14] K.E. Rasmussen, S. Pedersen-Bjergaard, Trends Anal. Chem. 23 (2004) 1.
392
[15] C.L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145.
393
[16] Z. Zhang, M.J. Yang, J. Pawliszyn, Anal. Chem. 66 (1994) 844 A.
394
[17] D.A. Lambropoulou, T. Sakellarides, T. Albanis, Fresenius J. Anal. Chem. 368 (2000) 616.
395
[18] Dj. Djozan, Y. Assadi, S. Hosseinzadeh Haddadi, Anal. Chem. 73 (2001) 4054.
Ac ce p
te
d
M
an
us
377
16 Page 16 of 30
[19] Dj. Djozan, Y. Assadi, Chromatographia 60 (2004) 313.
397
[20] H.L. Lord, J. Pawliszyn, Anal. Chem. 69 (1997) 3899.
398
[21] C.C. Chou, M.R. Lee, Anal. Chim. Acta 538 (2005) 49.
399
[22] B.C. Blount, R.J. Kobelski, D.O. McElprang, D.L. Ashley, J.C. Morrow, D.M. Chambers, F.L.
400
ip t
396
Cardinali, J. Chromatogr. B 832 (2006) 292. [23] P. Helena, Z.K. Lucija, Trends Anal. Chem. 18 (1999) 272.
402
[24] S. Pedersen-Bjergaard, K.E. Rasmussen, Anal. Chem. 71 (1999) 2650.
403
[25] L. Zhao, H.K. Lee, J. Chromatogr. A 919 (2001) 381.
404
[26] G. Shen, H.K. Lee, Anal. Chem. 74 (2002) 648.
405
[27] M. Rezaee, Y. Assadi, M.R.M. Hosseini, E. Aghaee, F. Ahmadi, S. Berijani, J. Chromatogr. A 1116
us
an
406
cr
401
(2006) 1.
[28] F. Ahmadi, Y. Assadi, M.R.M. Hosseini, M. Rezaee, J. Chromatogr. A 1101 (2006) 307.
408
[29] S. Berijani, Y. Assadi, M. Anbia, M.R. Milani Hosseini, E. Aghaee, J. Chromatogr. A 1123 (2006) 1.
409
[30] R.R. Kozani, Y. Assadi, F. Shemirani, M.R.M. Hosseini, M.R. Jamali, Talanta 72 (2007) 387.
410
[31] M.A. Farajzadeh, M. Bahram, J.Å. Jonsson, Anal. Chim. Acta 591 (2007) 69.
411
[32] Q. Zhou, H. Bai, G. Xie, J. Xiao, J. Chromatogr. A 1177 (2008) 43.
412
[33] M.A. Farajzadeh, Dj. Djozan, M.R. Afshar Mogaddam, J. Norouzi, J. Sep. Sci. 35 (2012) 742.
413
[34] Y. Wang, J. You, R. Ren, Y. Xiao, S. Gao, H. Zhang, A. Yu, J. Chromatogr. A 1217 (2010) 4241.
414
[35] J. Zhang, H. Gao, B. Peng, S. Li, Z. Zhou, J. Chromatogr. A 1218 (2011) 6621.
415
[36] P. Jovanov, V. Guzsvany, M. Franko, S. Lazic, M. Sakac, B. Saric, V. Banjac, Talanta 111 (2013)
d
te
Ac ce p
416
M
407
125.
417
[37] C. Zacharis, I. Rotsias, P. Zachariadis, A. Zotos, Food Chem. 134 (2012) 1665.
418
[38]M. Bashiri, A.Mehdinia, A. Jabbari, Y. Yamini, AJAC. 2 (2011) 632.
419
[39] http://monographs.iarc.fr/ENG/Classification.
420
[40] I.R. Pizzutti, A. Kok, R. Zanella, M.B. Adaimea, M. Hiemstra, C. Wickert, O.D. Prestes, J.
421
Chromatogr. A 1142 (2007) 123. 17 Page 17 of 30
[41] M.A. Farajzadeh, M. Bahram, F. Jafari, M. Bamorowat, Chromatographia 73 (2011) 393.
423
[42] A. Bordagaray, R. Garcia-Arrona, E. Millan, Food Anal. Methods 4 (2011) 293-299.
424
[43] Y. Li, J. Zhang, B. Peng, S. Li, H. Gao, W. Zhou, Anal. Methods 5 (2013) 2241.
425
[44] L. Wang, X. Zang, Q. Chang, G. Zhang, C. Wang, Z. Wang, Food Anal. Methods, 7 (2014) 318.
426
[45] M.A. Farajzadeh, Dj. Djozan, P. Khorram, Anal. Chim. Acta 713 (2012) 70.
427
Figure captions
428
Fig. 1 Effect of the chemical identity of extraction solvent in ERs of the selected pesticides in DLLME.
429
Extraction conditions: sample, 50 mL diluted pesticides-free honey sample solution spiked with analytes
430
(each pesticide, 25 ng g-1); extraction solvent, 1,2-DBE (130 µL), 1,1,2,2-TCE (140 µL), and 1,1,2,2-TBE
431
(60 µL); disperser solvent, DMSO (1.5 mL); temperature, 70 ◦C; centrifuge rate, 4000 rpm; and centrifuge
432
time, 5 min. The error bars indicate the minimum and maximum of three independent determinations.
M
an
us
cr
ip t
422
d
433
Fig. 2 Effect of the chemical identity of disperser solvent.
435
Extraction conditions: extraction solvent, 1,2-DBE(130 µL). Other conditions are the same as Fig. 1. The
436
error bars indicate the minimum and maximum of three independent determinations.
Ac ce p
437
te
434
438
Fig. 3 Effect of extraction solvent (1,2-DBE) volume on the ERs of pesticides.
439
Extraction conditions: the same as Fig. 2. The error bars indicate the minimum and maximum of three
440
independent determinations.
441 442
Fig. 4 Effect of disperser (DMF) volume.
18 Page 18 of 30
443
Extraction conditions: disperser (extraction) solvents volumes, 0.5 (105), 1.0 (115), 1.5 (130), 2.0 (145),
444
and 3.0 (155) mL (µL). Other conditions are the same as Fig. 3. The error bars indicate the minimum and
445
maximum of three independent determinations.
ip t
446 Fig. 5 Effect of temperature on the extraction efficiency.
448
Extraction conditions: 130 µL extraction solvent (1,2-DBE); 1.5 mL disperser solvent (DMF),; the other
449
condition are the same as Fig. 4. The error bars indicate the minimum and maximum of three independent
450
determinations.
an
us
cr
447
451
Fig. 6 Typical GC-NPD chromatograms of (I) standard solution (100 mg L-1 of each pesticide in 1,2-
453
DBE) (direct injection), (II) honey sample spiked with the selected pesticides (each pesticide 10 ng g-1),
454
and (III) un-spiked honey sample. In the cases of (II) and (III), the proposed method was performed and 1
455
µL of the final sedimented phase was injected into the separation system. Peak identification: 1,
456
penconazole; 2, hexaconazole; 3, diniconazole; 4, tebuconazole; and 5, difenoconazole.
457
Fig. 7Typical total ions current GC–MS chromatogram of the honey sample and mass spectra of
458
difenoconazole, scan 3302 (retention time 28.5 min).
460
d
te
Ac ce p
459
M
452
461 462
19 Page 19 of 30
464 465 466
Highlights
467
volume dispersive liquid-liquid microextraction was performed for the first time. ►Very high EFs
468
were obtained. ►The method was applied to determine some triazole pesticides in honey. ►LODs
469
are achievable at ng g-1 using GC-NPD.
efficient and reliable ET-DLLME method has been developed for the first time.► High
ip t
►An
cr
470
us
471
Ac ce p
te
d
M
an
472
21 Page 20 of 30
Ac
ce
pt
ed
M
an
us
cr
i
Figure 1
Page 21 of 30
Ac
ce
pt
ed
M
an
us
cr
i
Figure 2
Page 22 of 30
Ac
ce
pt
ed
M
an
us
cr
i
Figure 3
Page 23 of 30
Ac
ce
pt
ed
M
an
us
cr
i
Figure 4
Page 24 of 30
Ac
ce
pt
ed
M
an
us
cr
i
Figure 5
Page 25 of 30
Ac ce p
te
d
M
an
us
cr
ip t
Figure 6
Page 26 of 30
Ac ce p
te
d
M
an
us
cr
ip t
Figure 7
Page 27 of 30
Table 1. Quantitative features of the method for the selected pesticides. LR c)
r d)
In solution (ng L-1)
In honey (ng g-1)
In solution (ng L-1)
In honey (ng g-1)
In solution (ng L-1)
Penconazole
0.05
15
0. 15
45
0.15 - 100
45 - 30000
Hexaconazole
0.05
15
0. 15
Diniconazole
0.10
30
0. 35
Tebuconazole
0.15
45
0. 45
Difenconazole
0.21
70
1.1
M an
In honey (ng g-1)
ed
LOQ b)
us
LOD a)
Analyte
cr
ip t
Table 1
RSD %e)
EF ± SD f)
ER ± SD g)
Intra-day
Inter-days
0.995
3
5
1994 ± 39
100 ± 2
45
0.15- 100
45 - 30000
0.996
3
4
1960 ± 40
98 ± 2
100
0.35 - 100
100 - 30000
0.997
4
5
1988 ± 21
99 ± 1
150
0.45 - 150
150 - 45000
0.995
4
6
1943 ± 25
97 ± 1
1.1- 150
210 - 45000
0.993
3
4
1978 ± 28
99 ± 1
210
Limit of detection (S/N=3).
b)
Limit of quantification (S/N=10).
c)
Linear range.
d)
Correlation coefficient.
e)
Relative standard deviation (n=6, C =0.15 ng g-1) for intra-day and (n= 4, C=0.15 ng g-1) for inter-days
f)
Enrichment factor ± standard deviation (n=3).
g)
Extraction recovery ± standard deviation (n= 3).
Ac
ce pt
a)
Page 28 of 30
Table 2
Table 2. Study of matrix effect in samples spiked at different concentrations. Difenconazole’s content of samples was subtracted. Analyte
98 ± 3
99 ± 4
98 ± 1
cr
Difenconazole
Sample 4
-1
99 ± 3 97 ± 3 97 ± 4 99 ± 2
ip t
Mean recovery ± standard deviation (n=3) Sample 1 Sample 2 Sample 3 All samples were spiked with each analyte at a concentration of 10 ng g-1 96 ± 3 98 ± 2 97 ± 3 Penconazole 97 ± 2 99 ± 2 99 ± 3 Hexaconazole 98 ± 1 97± 3 97 ± 3 Diniconazole 99 ± 3 98 ± 2 97 ± 4 Tebuconazole
97 ± 4
98 ± 3 98 ± 4 97 ± 3 98 ± 4
99 ± 3 99 ± 3 98 ± 2 97 ± 4
Difenconazole
99 ± 3
96 ± 4
98 ± 3
98 ± 4
us
All samples were spiked with each analyte at a concentration of 25 ng g 98 ± 3 97 ± 3 Penconazole 99 ± 3 97 ± 3 Hexaconazole 97 ± 3 98 ± 4 Diniconazole 99 ± 3 97 ± 3 Tebuconazole
an
-1
97 ± 3 98 ± 3 97 ± 2 98 ± 3 97 ± 4
98 ± 4 98 ± 2 97 ± 3 98 ± 3 99 ± 4
Ac ce p
te
d
M
All samples were spiked with each analyte at a concentration of 50 ng g 97 ± 3 97 ± 3 Penconazole 98 ± 3 98 ± 3 Hexaconazole 98 ± 3 98 ± 2 Diniconazole 98 ± 4 97 ± 4 Tebuconazole 99 ± 3 98 ± 3 Difenconazole
Page 29 of 30
Table 3
Table 3. Comparison of TA-DLLME with other methods used in preconcentration and determination of the target analytes. Analytes Penconazole
Sample
LOD a)
Soya grain
LOQ b) -1
RSD c)
EF d)
-
100 ng g
2.2
-
Hexaconazole
-
-
12.1
-
Diniconazole
-
50 ng g-1
7.2
-
Tebuconazole
-
100 ng g-1
5.4
-
-
-1
6.3
-
Method
Ref. [40]
Hexaconazole
samples
0.11 ng mL-1
0.37 ng mL-1
4.8
2738
-1
-1
5.26
1580
-1
482
0.09 ng mL
0.30 ng mL
-1
Tebuconazole
0.14 ng mL
0.47 ng mL
3.43
Difenconazole
1.04 ng mL-1
3.47 ng mL-1
4.97
Penconazole
Rat blood
106, 102 ng mL-1
-1
0.23, 0.29 ng mL
Hexaconazole Tebuconazole
18
-
-
6.0 ng mL-1
-
6.0 ng mL-1
-
-1
6.0 ng mL
Vegetable
0.05 ng g-1
-
Diniconazole
samples
0.01 ng g-1
Penconazole
0.03 ng g Aqueous
5.45
189
6.45
178
4.53
185
3.9
263
Ac ce p
545
0.4 ng mL-1
1.1 ng mL-1
3.2
275
Diniconazole
0.5 ng mL-1
1.5 ng mL-1
3.4
306
Tebuconazole
2.0 ng mL-1
4.2 ng mL-1
5.2
380
0.05 ng g-1
0.15 ng g-1
3
1994
Hexaconazole
0.05 ng g-1
0.15 ng g-1
3
1960
Diniconazole
0.10 ng g-1
0.21 ng g-1
4
1988
Tebuconazole
0.15 ng g-1
0.45 ng g-1
4
1943
Difenconazole
0.21 ng g-1
1.1 ng g-1
3
1978
Honey
[42]
[43] UETC-IL-DLLME-HPLCDAD h)
MSPE-GC-MS i)
[44]
611
Hexaconazole
Penconazole
SPME-GC-ECD g)
509
0.9 ng mL-1
0.3 ng mL-1
samples
-
-
te
Tebuconazole
-1
5.6
< 11.7
-
[41]
-
d
Hexaconazole
SVE-DLLME-GC-FID f)
1142
an
Diniconazole
Grape and apple juices
M
Tebuconazole
ip t
Aqueous
us
Penconazole
100 ng g
cr
Difenconazole
QuEChERS–LC–MS/MS e)
DLLME-NBT-GC–FID j) [45]
ET-DLLME-GC-NPDk)
This method
a)
Limit of detection. b) Limit of quantification. c) Relative standard deviation. d) Enrichment factor. e) Quick, easy, cheap, effective, rugged and safe-liquid chromatography–tandem mass spectrometry. f) Silylated vessel extraction-dispersive liquid-liquid microextraction-gas chromatography-flame ionization detection. g) Solid-phase microextraction-gas chromatography-electron capture detector. h) Ultrasound-enhanced temperature-controlled ionic liquid-dispersive liquid-liquid microextraction-high performance liquid chromatographydiod array detector. i) Magnetic solid-phase extraction-gas chromatography-mass spectrometry. j) Dispersive liquid–liquid microextraction method in a narrow-bore tube-gas chromatography-flame ionization detection. k) Elevated temperature-dispersive liquid-liquid microextraction-gas chromatography-nitrogen-phosphorus detection.
Page 30 of 30