8
6
Anal. Methods Environ. Chem. J. 4 (2) (2021) 86-98  
Research Article, Issue 2  
Analytical Methods in Environmental Chemistry Journal  
Journal home page: www.amecj.com/ir  
AMECJ  
Determination of fenthion in environmental water samples  
by dispersive liquid–liquid microextraction coupled with  
spectrofluorimetric and chemometrics methods  
a
b,  
c
c
Tahereh Eskandari , Ali Niazi , Mohammad Hossein Fatemi and Mohammad Javad Chaichi  
a
Department of Chemistry, Arak Branch, Islamic Azad University, Arak, Iran  
Department of Chemistry, Central Tehran Branch, Islamic Azad University, Tehran. Iran  
Department of Analytical Chemistry, Faculty of Chemistry, Mazandaran University, Babolsar, Iran  
b
c
A R T I C L E I N F O :  
Received 15 Feb 2021  
A B S T R A C T  
In the present study, a simple, rapid and efficient dispersive liquid–  
liquid microextraction (DLLME) coupled with spectrofluorimetry  
Revised form 24 Apr 2021  
Accepted 20 May 2021  
Available online 30 Jun 2021  
(SFM) and chemometrics methods have been proposed for the  
preconcentration and determination of fenthion in water samples.  
Box–Behnken design was applied for multivariate optimization  
of the extraction conditions (sample pH, the volume of dispersive  
solvent and volume of extraction solvent). Analysis of variance  
was performed to study the statistical significance of the variables,  
their interactions and the model. Under the optimum conditions, the  
------------------------  
Keywords:  
Fenthion,  
Pesticides,  
Organophosphoruse pesticides,  
Dispersive liquid–liquid icroextraction,  
Box–Behnken design,  
Spectrofluorimetry  
-1  
calibration graph was linear in the range of 5.0–110 ng mL with  
the detection limit of 1.23 ng mL (3S /m). Parallel factor analysis  
-1  
b
(PARAFAC) and partial least square (PLS) modelling were applied  
for the multivariate calibration of the spectrofluorimetric data. The  
orthogonal signal correction (OSC) was applied for preprocessing of  
data matrices and the prediction results of model, and the analysis  
results were statistically compared. The accuracy of the methods,  
evaluated by the root mean square error of prediction (RMSEP) for  
fenthion by OSC-PARAFAC and OSC-PLS models were 0.37 and  
0
.78, respectively. The proposed procedure could be successfully  
applied for the determination of fenthion in water samples.  
1
. Introduction  
phosphorothioate ) is a contact and stomach or-  
The organophosphorous pesticides (OPPs) have  
been widely used in agriculture for crop production  
and fruit tree treatment, but many of them are iden-  
tified as highly toxic compounds [1–3]. They are  
released into the environment from manufacturing,  
transportation and agriculture applications. OPPs  
have been found in ground waters, surface waters,  
lagoons and drinking water. Fenthion (O,O-Di-  
ganophosphorous pesticide widely used in the con-  
trol of many sucking, biting pests, especially fruit  
flies, stem borers and mosquitoes on crops such as  
alfalfa, rice, sugar, vegetables and forests. Fenthion  
is toxic for the human and animal health [4–6]. The  
toxicological effect of fenthion, is almost entirely  
due to the inhibition of acetylcholinesterase in the  
nervous system, resulting in respiratory, myocar-  
dial and neuromuscular transmission impairment  
methyl  
O-[3-methyl-4-(methylsulfanyl)phenyl]  
[5, 7]. Due to the low concentration of the analytes  
*
Corresponding Author: Ali Niazi  
and the complex matrix of the samples, a prelim-  
inary sample preconcentration and a separation  
Email: ali.niazi@gmail.com, ali.niazi@iauctb.ac.ir  
https://doi.org/10.24200/amecj.v4.i02.138  
Determination of fenthion by DLLME-SFM  
Tahereh Eskandari et al  
87  
technique are required. Thus, different extraction  
processes have been used for separation and  
pre-concentration of trace pesticide residues, such  
as solid phase extraction method (SPE) [8–11], sol-  
id phase microextraction (SPME) [12], [13], single  
drop microextraction (SDME) [14] and dispersive  
liquid–liquid microextraction (DLLME) [15–17].  
In the last decades, liquid–phase microextraction  
matical and statistical method. The main advantage  
of RSM is that it reduces the number of experiment  
because several factors can be varied simultaneous-  
ly for optimization and as a result saves time, en-  
ergy, and chemicals [16,37,38]. Box–Behnken de-  
sign is the most common and efficient design used  
in RSM. Box–Behnken design is a second order  
multivariate technique based on three level partial  
factorial designs. Box–Behnken is a spherical, ro-  
tatable or nearly rotatable that consists of a central  
point and with the midpoints of the edges of the  
variable space [15–17], [33, 34]. Two dimensional  
excitation emission (EEM) fluorescence data can  
be obtained by measuring the emission spectra at  
various excitation wavelengths. In recent years, ap-  
plication of multi-way data analysis techniques has  
increased significantly in the analytical chemistry.  
There are several multivariate calibration proce-  
dures that can be used for the treatment of EEM flu-  
orescence data, in order to quantify the compounds,  
present in a mixture [39]. In fluorescence analysis,  
parallel factor analysis (PARAFAC) [28], [40–44]  
and partial least-squares regression (PLS) [34, 43],  
[45–47] has been mostly applied for the analyses  
of three-way data obtained as series of emission  
spectra measured for different excitations. PLS is a  
factor analysis method that has been used in multi-  
component quantitative analysis from several spec-  
tral data, such as IR, UV-visible or fluorescence  
[47]. Partial Least Squares (PLS) regression is a  
method to predict the response variable based on  
predictor variables and to describe their common  
structure. The main advantage of PLS calibration  
procedures is that they can model a system even  
in the presence of interfering signals, provided that  
they are included in the calibration step. PARA-  
FAC is a multi-way decomposition method that  
has investigated to be useful for the analysis of  
second-order calibration. The main advantages of  
the PARAFAC model are the uniqueness, simplic-  
ity of its solutions and quantification of an analyte,  
even in the presence of unknown interferences (the  
second-order advantage) [40, 44].The orthogonal  
signal correction (OSC) is a useful pre-process-  
ing step that improves the chemometrics model  
(LPME), based on the miniaturization of traditional  
LLE technique by greatly reducing the use of or-  
ganic solvent has been reported as an alternative  
for sample preparations. One of the most popular  
LPME techniques is dispersive liquid-liquid mi-  
croextraction (DLLME) which is widely used as  
a preconcentration method [18-21]. DLLME was  
developed by Assadi and co-workers [16]. By con-  
sisting of the formation of a cloudy solution pro-  
moted by the fast addition in the aqueous sample  
of a mixture of extractor and dispersive solvents.  
The tiny droplets formed and dispersed among the  
aqueous sample solution are further joined and  
sedimented in the bottom of a conical test tube by  
centrifugation. This method provides many advan-  
tages including rapidity, simplicity of operation,  
high recovery and enrichment factor. After sample  
preparation, the determination of OPPs in different  
sample matrices was carried out by using gas chro-  
matography mass spectrometry (GC-MS) [9, 22],  
gas chromatography (GC) [23–25] and high-per-  
formance liquid chromatography (HPLC) [26,27].  
Fluorescence spectrometry is a sensitive, selective  
and relatively low cost method for the quantitative  
analysis of pesticides and other pollutants [28–30].  
Different experimental variables can affect the ex-  
traction yield in the DLLME procedure; therefore,  
a multivariate approach has been widely used for  
their optimization. Statistical methods are useful to  
determine the effects of variables on the extraction  
procedure. The response surface methodology  
(
RSM) based on statistical design of experiments  
DOEs) has been extensively used for modelling  
(
and optimization in various analytical procedures  
31–36]. Response surface methodology (RSM)  
[
is powerful multivariate technique that used for  
building empirical model via collection of mathe-