Preconcentration and simultaneous spectrophotometric determination of copper and mercury by dispersive liquid–liquid microextraction and orthogonal signal correction–partial least squares

A. Niazi; S. Habibi; M. Ramezani

doi:10.1016/j.arabjc.2013.12.014

View/Download PDF

Buy Reprints

PDF

Translate this page into:

Original article

8 (

); 706-714

doi:

10.1016/j.arabjc.2013.12.014

Preconcentration and simultaneous spectrophotometric determination of copper and mercury by dispersive liquid–liquid microextraction and orthogonal signal correction–partial least squares

A. Niazi^⁎, S. Habibi, M. Ramezani

Department of Chemistry, Faculty of Science, Arak Branch, Islamic Azad University, Arak, Iran

⁎Corresponding author. Tel./fax: +98 861 3670017. a-niazi@iau-arak.ac.ir (A. Niazi) ali.niazi@gmail.com (A. Niazi)

Received: 2012-7-20, Accepted: 2013-12-17, Published: 2013-09-01

Disclaimer:
This article was originally published by Elsevier and was migrated to Scientific Scholar after the change of Publisher.

Peer review under responsibility of King Saud University.

Abstract

A new method for spectrophotometric simultaneous determination of copper and mercury was developed by the dispersive liquid–liquid microextraction (DLLME) preconcentration and orthogonal signal correction–partial least squares (OSC–PLS). In the proposed method, dithizone was used as a chelating agent, and carbon tetrachloride and acetonitrile were selected as extraction and dispersive solvents. All factors affecting the sensitivity were optimized by the Box–Behnken design and the linear dynamic range for determination of copper and mercury was found. Under the optimum conditions, the calibration graphs were linear in the range of 10.0–250.0 and 10–300 ng mL⁻¹ with detection limit of 2.6 and 2.8 ng mL⁻¹ (3δ_B/m) and the enrichment factor of this method for copper and mercury, reached 180 and 175, respectively. The simultaneous determination of copper and mercury by using spectrophotometric methods is a difficult problem, due to the spectral interferences. The PLS modeling was used for the multivariate calibration of the spectrophotometric data. The OSC was used for preprocessing of data matrices and the prediction results of model, with and without using OSC, were statistically compared. The experimental calibration matrix was designed by measuring the absorbance over the range of 400–700 nm for 25 samples. The root mean squares error of prediction for copper and mercury with and without OSC was 0.010, 0.026 and 0.055, 0.086, respectively. The proposed method was successfully applied for the simultaneous determination of copper and mercury in spiked water and synthesis samples.

Keywords

Copper

Mercury

Determination

Box–Behnken design

DLLME

OSC–PLS

Show Related Articles from PubMed

1 Introduction

Nowadays the pollution of different natural waters by heavy metals is a great concern because of the toxic effects on living organisms. Urbanization, industrial development, and heavy traffic lead to contamination of water bodies by heavy metals. Copper is a widespread anthropogentic pollutant of environment and the determination of this metal is the actual problem. However, high amounts of copper can be harmful, causing irritation of nose and throat, nausea, vomiting, and diarrhea. Mercury is one of the toxic heavy metals and as an introduced contaminant in the environment. The toxicity of mercury depends on its chemical species and it is found that organomercurials are more toxic than inorganic mercury compounds. Thus the development of new methods for separation, preconcentration and determination of these metal ions at trace levels in environmental is one of the targets of analytical chemists, due to their important roles in our life.

Several procedures have been developed for the separation and preconcentration of copper and mercury from environmental matrices such as: liquid–liquid extraction (Kara and Alkan, 2002), co-precipitation (Doner and Ege, 2005 and Zhang et al., 2004), solid phase extraction (Starvin and Prasada Rao, 2004; Tobiasz et al., 2012; Walas et al., 2008 ) and cloud point extraction (Gao et al., 2010; Niazi et al., 2009; Shoaee et al., 2012 ). However, liquid–liquid extraction (LLE) is time-consuming and requires large amounts of organic solvents that are potentially toxic. Solid phase extraction (SPE) uses much less solvent than LLE but can be relatively expensive. Additionally, evaporation of the final organic extract into a small volume is necessary to achieve high enrichment of the analytes. Batch-to-batch reproducibility continues to be the major concern for analysts in selecting SPE devices. Elution of sorbed solute must be performed after sample loading. Solvent evaporation and redissolution are often required (Nagaraju and Huang, 2007).

A novel microextraction technique as a high performance and powerful preconcentration method termed as dispersive liquid–liquid microextraction (DLLME) was demonstrated by Rezaee et al. (2006). In this method, an appropriate mixture of the extraction solvent and the dispersive solvent are injected into aqueous sample by a syringe and forms a cloudy solution. The cloudy state results from the formation of fine droplets of the extraction solvent which disperse in the sample solution. The cloudy solution shall be centrifuged and the fine droplets sediment at the bottom of the conical test tube. Determination of analytes in the remained phase can be performed by instrumental techniques. Recently, several papers have been published about the dispersive liquid–liquid microextraction in preconcentration and determination of metals (Bidari et al., 2007; Fan, 2007; Gharehbaghi et al., 2008; Liang and Sang, 2008; Liang et al., 2009; Naseri et al., 2008; Rivas et al., 2009; Xiao-Huan et al., 2009 ).

Spectrophotometric methods are the most commonly used techniques and continue to enjoy wide popularity. The common availability of the instrumentation, the simplicity of procedures, speed, precision, accuracy and low-operating costs of the technique still make spectrophotometric methods attractive. In the present work, the DLLME was combined with UV–Visible spectrophotometry by using micro-sample introduction system for the simultaneous determination of copper and mercury was proposed. In this method, dithizone, which forms complexes with copper and mercury, was selected as chelating reagent. Dithizone contains azo and hydrosulfide groups, which are good electron donors. Dithizone is widely used in extraction spectrophotometry because it can form stable complexes with many metal ions under proper conditions. The factors influencing the efficiency of DLLME and determination of copper and mercury were systematically studied and optimized by the Box–Behnken design (Khajeh, 2009). Box–Behnken is a second-order multivariate design technique based on three-level incomplete factorial designs that received widespread application for evaluation of critical experimental conditions, that is, maximum or minimum of response functions (Macedo et al., 2009).

The simultaneous determination of several compounds in a mixture can be a difficult problem, especially for components that have similar analytical characteristics. The problem is how to distinguish overlapped signals that are often encountered in analytical experiments. Quantitative simultaneous spectrophotometry has been greatly improved by the use of multivariate statistical methods such as the PLS method. PLS modeling is a powerful multivariate statistical tool and can be performed with easily accessible statistical software (Brereton, 2000; Geladi and Kowalski, 1986). The basic concept of PLS was originally described by Gerlach et al. (1979) and Joreskog and Wold (1982), and consequently different applications for PLS modeling were reported (Ni et al., 2008; Niazi et al., 2005, 2007; Zapata-Urzua et al., 2010 ). Wold et al. (1988) introduced OSC as a pre-processing step that improves the calibration model by filtering strong structured (i.e. systematic) variation in X that is not correlated to Y. Therefore, one can be certain that important information regarding the analyte is retained. Since then, several groups (Andersson, 1999; Fearn, 2000; Pierna et al., 2001; Sjoblom et al., 1998; Westerhuis et al., 2001; Wold et al., 2001 ) have published various OSC algorithms in an attempt to reduce model complexity by removing orthogonal components from the signal. Recently, application of OSC in UV–Visible spectrophotometry simultaneous determination by PLS is reported (Khajehsharifi et al., 2009; Niazi et al., 2008; Niazi and Goodarzi, 2008; Niazi and Yazdanipour, 2007 ).

The aim of this study was to investigate, the possibility of using DLLME for simultaneous spectrophotometric determination of copper and mercury, in synthetic and real matrix samples such as different water samples, for the first time. The results obtained, with and without using OSC algorithm as a preprocessing treatment of original data, were compared. To our knowledge this is the first spectrophotometric report with DLLME extraction on the simultaneous determination of copper and mercury.

2 Experimental

2.1

2.1 Instrument and software

A Hewlett–Packard 8453 diode array spectrometer controlled by a Hewlett–Packard computer and equipped with a 100 μL quartz cell was used for recording the spectra. A centrifuge (Behdad Universal Centrifuge) was used to accelerate the phase separation process. The pH was determined with a model 780 Metrohm pH-meter with combined glass-calomel electrode.

PLS and OSC programs were written in MATLAB Version 6.5 (Math works Inc.). All programs were run on a personal computer (CPU 3.0 GHz and RAM 4 GB) with Windows XP operation system. Box–Behnken design was accomplished with Minitab Version 15. The OSC version applied here is based on the Wold algorithm (Wold et al., 1988).

2.2

2.2 Reagents and materials

All reagents were of analytical reagent grade. The water utilized in all studies was double-distilled and deionized. Stock solution of Cu²⁺ and Hg²⁺ (1000 μg mL⁻¹) was prepared by dissolving appropriate amounts of their commercial nitrate salts in deionized water and standardized titrimetrically (Mendham et al., 1998). Standards of working solutions were made by appropriate dilution daily as required. The dithizone and all solvents, such as methanol, ethanol, chloroform, carbon tetrachloride, acetonitrile and acetone were obtained from Merck. Universal buffer solutions were prepared from boric acid, citric acid and phosphoric acid (0.04 mol L⁻¹). The final pH was adjusted by the addition of 0.2 mol L⁻¹ sodium hydroxide.

2.3

2.3 Dispersive liquid–liquid microextraction procedures

A 10 mL of sample of standard solution containing 10.0–250.0 and 10.0–300.0 ng mL⁻¹ of copper and mercury, respectively, and potassium nitrate (5%) was poured in a test tube with a conical bottom and 1 mL of dithizone (1.6 × 10⁻⁴ mol L⁻¹) as chelating agent was added to the solution. After a few minutes the complex of copper and mercury was formed and the solution pH was adjusted to 3.4 by universal buffer. Then a binary solution containing 800 μL of acetonitrile (disperser solvent) and 200 μL of carbon tetrachloride (extraction solvent) was injected rapidly into the sample using a syringe and a stable cloud solution were obtained. Then, Cu(Dithizone)₂ and Hg(Dithizone)₂ complexes were extracted into fine droplets of carbon tetrachloride. After that the mixture was centrifuged for 5 min at 3000 rpm. After this process the fine droplets of carbon tetrachloride were joined together and separated at the bottom of the conical test tube. After removing the whole aqueous solution, and transferring to a 100 μL cell, the absorbance was measured at 400–700 nm. The volume of the separated phase was determined to be about 150 μL.

2.4

2.4 Sample preparation before DLLME

Tap, mineral, river and waste water samples were collected from Kermanshah, Taq-e-Bostan (Mineral water, river and waste water). Prior to the preconcentration procedure, all the water samples were filtered through a 0.45 μm pore size membrane filter to remove suspended particulate matter, and adjusted to approximately pH 3.4 by universal buffer and then were stored at 4 °C in the dark.

3 Results and discussion

In order to select the optimum DLLME conditions for the determination of copper and mercury, a procedure was required to optimize the different parameters that affect the DLLME extraction process. Some of these parameters are selection of pH, suitable concentration of dithizone, extraction solvent, disperser solvent, volume of extraction solvent, volume of disperser solvent, ionic strength of aqueous phase, and extraction time. It is very important to optimize them in order to obtain good recovery.

3.1

3.1 Experimental design optimization

A two level factorial 2⁴ design with two replicates of center point was performed in order to determine the influence of these factors and their interactions. The factorial design was evaluated using analytical response (Absorbance in 550 and 490 nm for copper and mercury complexes, respectively). An analysis of the variance (ANOVA) demonstrated that, within the experimental range, pH, concentration of dithizone, volume of extraction solvent and disperser solvent were statistically significant. The significant variables like pH; concentration of dithizone, volume of extraction solvent and disperser solvent were chosen as the critical variables and designated as pH, C_Ligand, V_disperser and V_extraction, respectively. The low, middle and high levels of each variable were designated as −, 0 and +, respectively, are given in Table 1.

Table 1 The levels of variables chosen for the trials.

Variable	Low level (−)	Middle level (0)	High level (+)
pH	1.0	2.5	4.0
C_Ligand (mol L⁻¹)	1.0 × 10⁻⁵	1.5 × 10⁻⁵	2.0 × 10⁻⁵
V_disperser (μL)	700	800	900
V_extraction (μL)	100	200	300

In a system involving four significant independent variables, the mathematical relationship of the response (Absorbance in 550 and 490 nm for copper and mercury complexes, respectively) of these variables can be approximated by the quadratic polynomial equation:

(1)

Y = β_{0} + β_{1} pH + β_{2} C_{Ligand} + β_{3} V_{disperser} + β_{4} V_{extraction} + β_{12} pH \times C_{Ligand} + β_{13} pH \times V_{disperser} + β_{14} pH \times V_{extraction} + β_{23} C_{Ligand} \times C_{disperser} + β_{24} C_{Ligand} \times V_{extraction} + β_{34} V_{disperser} \times V_{extraction} + β_{11} {pH}^{2} + β_{22} C_{Ligand}^{2} + β_{33} V_{disperser}^{2} + β_{44} V_{extraction}^{2}

A multiple regression analysis is done to obtain the coefficients and the equation can be used to predict the response. The design of experiments chosen for this study was Box–Behnken, a fractional factorial design for four independent variables. It is applicable for the critical variables that have been identified. In the model given in Eq. (1), interactions higher than second order have been neglected. A total of 27 experiments for each cation were necessary to estimate the full model. The results from this experimental design provided a statistical process, which was used to identify high yield trends for the DLLME process. Analysis of variance was applied to the statistical significance of the models. According to the results shown in Table 2, the four factors pH, C_Ligand, V_disperser and V_extraction are significant and also two parameters pH × C_Ligand and pH² significantly influenced on the DLLME method. The critical points in the surface response are found by solving the derivation obtained from Eq. (1) for the condition of δ(A)/δ(pH) = 0, δ(A)/δ(C_Ligand) = 0, δ(A)/δ(V_disperser) = 0 and δ(A)/δ(V_extraction) = 0 for each models. The calculated values for the critical point for copper and mercury are: pH = 3.4, C_Ligand = 1.6 × 10⁻⁵ mol L⁻¹, V_disperser = 800 μL and V_extracion = 200 μL.

Table 2 Coefficients and p-values of Box–Behnken design for copper and mercury determination after DLLME.

Variables	Coefficient		p-Value
Variables	Copper	Mercury	Copper	Mercury
β₀	−3.5209	−3.8824	0.012	0.014
β₁	0.5723	0.5926	0.016	0.012
β₂	5.5203	5.3991	0.029	0.023
β₃	0.0023	−0.0011	0.037	0.032
β₄	0.0038	0.0014	0.015	0.030
β₁₂	0.2136	0.2526	0.018	0.027
β₁₃	−0.0004	−0.0008	0.073	0.062
β₁₄	0.0003	0.0005	0.083	0.089
β₂₃	−0.0036	−0.0060	0.145	0.106
β₂₄	0.0004	0.0003	0.066	0.053
β₃₄	0.0002	0.0000	0.186	0.165
β₁₁	−0.0712	−0.0634	0.046	0.042
β₂₂	−0.0975	−0.1446	0.107	0.098
β₃₃	0.0000	0.0000	0.058	0.062
β₄₄	0.0000	0.0000	0.142	0.133

Selecting the extraction solvent by paying attention to its characteristic properties is very important. It must have a higher density than water, be capable of extracting the compounds of interest, and have low solubility in water. Chloroform, carbon dichloride and carbon tetrachloride were compared in this extraction and obtained recoveries were higher for carbon tetrachloride. The main criterion for the selection of the disperser solvent is its miscibility in the extraction solvent and aqueous solution. In addition, the type of disperser directly influences the viscosity of the binary solvent. Thus, this solvent can control droplet production and extraction efficiency. To study this effect, four different solvents such as acetone, acetonitrile, ethanol and methanol were tested and obtained recoveries were higher for acetonitrile.

In DLLME, extraction time is defined as the time between the injection of the binary solvent and starting to centrifuge. The effect of extraction time was examined in the range of 1–10 min with constant experimental conditions. The surface area between extraction solvent and aqueous phase is infinitely large. Thereby, transferring the complex from aqueous phase to extraction solvent is fast. Subsequently, equilibrium state is achieved quickly; therefore, the extraction time is very short, which is the advantage of DLLME technique. In this method, the most time-consuming step is the centrifuging of sample solution, which is about 1 min. A series of same solutions were tested at various rates of centrifugation. The rate of centrifugation was adjusted between 1000 and 6000 rpm for 5 min. The absorbance slowly increases with increasing the rate to 3000 rpm and after that, it approximately stays constant. 3000 rpm was selected as the optimum rate for centrifuging.

For studying the influence of ionic strength on the performance of DLLME, we investigated KNO₃ concentration in the range of 0–10% (w/v) while other experimental conditions were kept constant. By increasing KNO₃ concentration, extraction efficiency is slowly increased due to salting-out effect and then approximately stays constant. These studies showed the possibility of DLLME for separation of copper and mercury from saline solution to 10% (w/v). As the three parameters, extraction time, rate of centrifugation and ionic strength had approximately constant influence at studied ranges; they were not investigated by experimental design.

3.2

3.2 Interferences study

The potential interference in the present method was investigated. The interference was due to the competition of other heavy metal ions for the chelating agent and their subsequent co-extraction with copper and mercury. In these experiments, solutions containing 10 ng mL⁻¹ of copper and mercury, and the interfering ions were treated according to the recommended procedure. The tolerance limits of the co-existing ions, defined as the largest amount making the recovery of copper and mercury less than 95%, are given in Table 3.

Table 3 Tolerance limits of co-existing ions in DLLME of copper and mercury.

Co-existing ions	Tolerance limits (ng mL⁻¹)
Ca²⁺, Mg²⁺, Ba²⁺, Sr²⁺, As³⁺, Mn²⁺, Li⁺, Rb⁺, K⁺, Na⁺, NH₄⁺, Cr³⁺	1000
${CO}_{3}^{2 -}$ , ${NO}_{3}^{-}$ , Br⁻, $S_{2} O_{3}^{2 -}$ , ${SO}_{3}^{2 -}$ , I⁻, ${CH}_{3} {COO}^{-}$ , Cl⁻, SCN⁻, ${BrO}_{3}^{-}$ , $C_{2} O_{4}^{2 -}$ , ${PO}_{4}^{3 -}$	1000
Cs⁺, Cd²⁺	500
Zr⁴⁺, Ga²⁺, Pd²⁺	400
Al³⁺	150
Ni²⁺, Co²⁺, In³⁺	120
Tl⁺	100

3.3

3.3 Analytical performance

Table 4 summarizes the analytical characteristics of the optimized method, including optimization conditions, linear range, limit of detection, reproducibility and enhancement factor. The calibration graph was linear in the range of 10.0–250.0 and 10–300 ng mL⁻¹ of copper and mercury, respectively. The limit of detection, defined as C_DL = 3 S_B/m (where C_DL, S_B and m are the limit of detection, standard deviation of the blank and slope of the calibration graph, respectively), was as 2.6 and 2.8 ng mL⁻¹ for copper and mercury, respectively. The relative standard deviation (R.S.D.) for five replicate measurements of 50 ng mL⁻¹ of each Cu²⁺ and Hg²⁺ were 1.6% and 1.9%. The enhancement factors were obtained from the slope ratio of the calibration graph after and before extraction, which were about 180 and 175 for copper and mercury, respectively. The equations of calibration graphs after and before extraction are summarized in Table 4.

Table 4 Analytical characteristics of DLLME for determination of copper and mercury.

Parameter	Analytical feature
Parameter	Copper	Mercury
Calibration curve before extraction	A = 3.501C_Cu + 0.0605 (R² = 0.9920)	A = 3.551C_Hg + 0.0536 (R² = 0.9912)
Calibration curve after extraction	A = 0.0195C_Cu + 0.1409 (R² = 0.9956)	A = 0.0201C_Hg + 0.1264 (R² = 0.9945)
Linear range (ng mL⁻¹)	10.0–250.0	10.0–300.0
Limit of detection (ng mL⁻¹)	2.6	2.8
Repeatability (R.S.D.,%) (n = 10)	1.6	1.9
Enhancement factor (mL)	∼180	∼175

3.4

3.4 Multivariate calibration

Fig. 1 shows the absorption spectra after DLLME of dithizone and the individual copper and mercury complexes at optimum conditions.

Absorption spectra of (a) dithizone, (b) 190 ng mL−1 of copper and (c) 210 ng mL−1 of mercury under optimum conditions.

As Fig. 1 shows, there is a clear overlapping of the two complexes’ spectra. This prevents the simultaneous determination of the copper and mercury by direct UV–Visible absorbance measurements. To overcome this problem a suitable and simple technique, which presents a good recovery, is PLS regression. A mixture design was used to maximize statistically the information content in the spectra. A training set of 25 samples was taken. The concentrations of copper and mercury varied between 10.0–250.0 and 10.0–300.0 ng mL⁻¹, respectively.

In Table 5 the compositions of the binary mixtures used in the calibration matrices are summarized. For prediction set, five prepared mixtures that were not included in the previous set were employed as independent test (see Table 5). To ensure that the prediction and real samples are in the subspace of training set, the score plot of first principal component vs. second was sketched and all the samples are spanned with the training set scores. The spectral region between 400 and 700 nm, which implies working with 300 experimental points per spectra, was selected for analysis, because this is the zone with the maximum spectral information from the mixture components of interest. All absorption data are pretreated by mean-centering.

Table 5 Concentration data of the different mixtures used in the calibration and prediction sets for the determination of copper and mercury (ng mL⁻¹).

Mixture^a	Copper	Mercury	Mixture	Copper	Mercury
C1	10.0	10.0	C16	190.0	10.0
C2	10.0	80.0	C17	190.0	80.0
C3	10.0	150.0	C18	190.0	150.0
C4	10.0	220.0	C19	190.0	220.0
C5	10.0	300.0	C20	190.0	300.0
C6	70.0	10.0	C21	250.0	10.0
C7	70.0	80.0	C22	250.0	80.0
C8	70.0	150.0	C23	250.0	150.0
C9	70.0	220.0	C24	250.0	220.0
C10	70.0	300.0	C25	250.0	300.0
C11	130.0	10.0	P1	240.0	30.0
C12	130.0	80.0	P2	40.0	110.0
C13	130.0	150.0	P3	50.0	50.0
C14	130.0	220.0	P4	95.0	90.0
C15	130.0	300.0	P5	110.0	50.0

a C: calibration, P: prediction.

3.5

3.5 Preprocessing by orthogonal signal correction

Orthogonal signal correction (OSC) is a preprocessing technique used for removing the information unrelated to the target variables based on constrained principal component analysis. OSC is a suitable preprocessing method for PLS calibration of mixtures without loss of prediction capacity using the spectrophotometric method. For calibration set two OSC components were used for filtering. Evaluation of the prediction errors for the validation set reveals that the OSC treated data give substantially lower root mean squares error of prediction values than original data. Also, the OSC-filtered data give much simpler calibration models with fewer components than the ones based on original data (Table 6).

Table 6 Composition of synthetic mixtures and predicted values for simultaneous determination of copper and mercury (ng mL⁻¹).

Added		Found (PLS)		Error (%)		Found (OSC–PLS)		Error (%)
Copper	Mercury	Copper	Mercury	Copper	Mercury	Copper	Mercury	Copper	Mercury
240.0	30.0	229.1	33.4	−4.5	11.3	238.6	30.4	−0.6	1.3
40.0	110.0	36.5	119.8	8.7	8.9	39.6	113.6	−1.0	3.3
50.0	50.0	44.3	55.4	−11.4	10.8	49.2	50.7	−1.6	−1.4
95.0	90.0	98.7	96.4	3.9	7.1	97.4	91.3	2.5	1.4
110.0	50.0	101.6	54.1	−7.6	8.2	109.3	51.7	−0.6	3.4

NF^a		5	6			2	2
RMSEP^b		0.055	0.086			0.010	0.026
RMSEC^c		0.032	0.064			0.008	0.018
RSEP (%)^d		5.467	8.633			1.043	2.640
γ (ng⁻¹ mL⁻¹)^e		153	162			89	96
LOD (ng mL⁻¹)^e		4.6	5.6			1.8	2.1

a Number of factors.

b Root mean squares error of prediction.

c Root mean squares error of calibration.

d Relative standard error of prediction.

e γ (analytical sensitivity) = SEN/[V(R)]^1/2 where SEN is the sensitivity (estimated as the net analyte signal) and V(R) is the variance of the instrumental signal and LOD (limit of detection) = 3.3s(0) where s(0) is the S.D. in the predicted concentration of copper and mercury in a blank sample (Lorber et al., 1997).

The results imply that the OSC method indeed removes information from UV–Visible data that is not necessary for fitting the Y-variables. In some cases the OSC method also removes non-linear relationships between X and Y. Fig. 2 shows the score plot when the PLS and OSC–PLS models are used. The score plots are shown for comparison of the results obtained from PLS and OSC–PLS. The results show that, score plots have better results when OSC–PLS is used. Score plots reveal the geometrical placement of solutions in principal component space. The experimental noise can destroy this relation (Fig. 2a) but by removing the noise using OSC filtering (Fig. 2b), the suitable geometrical placement is depicted in a more clear way.

Plots of first principal component against second principal component for simultaneous determination of copper and mercury (a) by PLS and (b) by OSC–PLS model.

3.6

3.6 Selection of the optimum number of factors

The number of latent variables (factors) for each cation was determined by the cross-validation method. The prediction error sum of squares (PRESS) for cross-validated models was calculated. The cross-validation method employed was to eliminate only one sample at a time and then PLS calibrates the remaining standard spectra. By using this calibration the concentration of the sample that is left out, was predicted. This process was repeated until each standard had been left out once. One reasonable choice for the optimum number of factors would be that number which yielded the minimum PRESS. Since there are a finite number of samples in the training set, in many cases the minimum PRESS value causes overfitting for unknown samples that were not included in the model. A solution to this problem has been suggested by Haaland and Thomas (1988) and Haaland and Thomas (1990) in which the PRESS values for all previous factors are compared to the PRESS value at the minimum. The F-statistical test can be used to determine the significance of PRESS values greater than the minimum. The maximum number of factors used to calculate the optimum PRESS was selected as 13 and the optimum number of factors obtained by the application of PLS and OSC–PLS models is summarized in Table 5. In all instances, the number of factors for the first PRESS values whose F-ratio probability drops below 0.75 was selected as the optimum. In Fig. 3, the PRESS obtained by optimizing the calibration matrix of the absorbance data with PLS and OSC–PLS models is shown.

Plots of PRESS vs. number of factors by PLS and OSC–PLS, (a) copper and (b) mercury.

3.7

3.7 Simultaneous determination of copper and mercury in synthetic and real matrix samples

The predictive ability of method was determined using five binary mixtures (their compositions are given in Table 6). The results obtained by applying PLS and OSC–PLS algorithm to five synthetic samples are listed in Table 6.

Table 6 also shows the error for prediction series of copper and mercury mixtures. As can be seen, the errors were also quite acceptable for the OSC–PLS method. In chemometrics, the root mean squares error of prediction (RMSEP), calibration (RMSEC) and relative standard error of prediction (RSEP) generally express the accuracy of the model (Valderrama and Poppi, 2008). The values of RMSEP, RMSEC and RSEP (%) for copper and mercury are summarized in Table 6.

The proposed DLLME method was validated by extraction and simultaneous determination of copper and mercury from spiked tap, mineral, river and waste water and the results are summarized in Table 7. To this end, aliquots of solutions of the copper and mercury were supplied with variable amounts, between 10.0–250.0 and 10.0–300.0 ng mL⁻¹ for copper and mercury, respectively, and their absorbances were recorded.

Table 7 Simultaneous determination of copper and mercury using the OSC–PLS method in synthesis and water samples.

Sample^a	Spiked (ng mL⁻¹)		Found^b (ng mL⁻¹)		Recovery (%)
Sample^a	Copper	Mercury	Copper	Mercury	Copper	Mercury
Tap	0	0	N.D.^c	N.D.	–	–
Tap	100.0	50.0	105.6 ± 0.5	51.3 ± 0.2	105.6	102.6

Mineral	0	0	23.8 ± 0.7	N.D.	–	–
Mineral	50.0	100.0	74.2 ± 0.5	101.7 ± 0.4	100.8	101.7

River	0	0	88.2 ± 0.7	13.7 ± 0.2	–	–
River	50.0	20.0	137.4 ± 0.6	33.5 ± 0.4	98.4	99.0

Waste	0	0	102.3 ± 0.8	42.3 ± 0.3	–	–
Waste	100.0	100.0	206.8 ± 0.9	146.3 ± 0.6	104.5	104.0

Sample 1^d	100.0	50.0	93.3 ± 0.6	46.8 ± 0.7	93.3	93.6

Sample 2^e	50.0	75.0	47.9 ± 0.7	73.8 ± 0.8	95.8	98.4

a Appropriate dilution before analysis was performed and the source of samples is described in the text.

b Mean value of three replicate determination ± standard deviation (n = 3).

c Not detected.

d Co²⁺ (30), Zn²⁺ (30), Ni2⁺ (30), and Al³⁺ (20).

e Co²⁺ (50), Zn²⁺ (30), Fe²⁺ (50), and Fe²⁺ (30).

Also, two synthesis samples and their compositions listed in Table 7 are used for validation of the proposed method. As can be seen from Table 7, the results obtained by the OSC–PLS method in the simultaneous determination of copper and mercury in water samples were quite good. In fact, the recoveries ranged from 93.3% to 105.6% for copper and 93.6% to 104.0% for mercury. Therefore, the OSC–PLS model is able to predict the concentration of each cation in the real matrix samples.

4 Conclusion

A new method of DLLME combined with spectrophotometry has been proposed for the simultaneous determination of copper and mercury in water samples. The proposed DLLME method has numerous advantages such as; rapidness, simplicity, low-cost, ease of operation, low toxic, high efficiency and low organic solvent-consumption. The copper and mercury reacted sensitivity with dithizone to form the colored complexes in the non-aqueous media. Their absorption spectra of these complexes completely overlap with the spectrum of dithizone and with each other. In order to overcome the drawback, PLS and OSC–PLS multivariate calibration approaches were applied and compared. Analysis of the results for binary mixtures showed that the use of PLS leads to significantly less-accurate prediction. The predicted values obtained by application of the OSC–PLS model for absorbance data show the high prediction ability of the OSC–PLS method. The proposed method provides a good reproducibility and gives a precise, highly sensitive and selective procedure with good LODs. The enrichment factors for 10 mL sample solution for copper and mercury were obtained ∼180 and ∼175, respectively. Also, the application of a Box–Behnken matrix was a possible, rapid, economical and efficient way of an optimization strategy of the proposed procedure. The method was successfully applied to simultaneous determination of copper and mercury in environmental water samples; satisfied recovery and reproducibilities of the proposed method were also obtained.

References

Andersson C.A., . Direct orthogonalization. Chemom. Intell. Lab. Syst.. 1999;47:51-63.
[Google Scholar]
Bidari A., Zeini Jahromi E., Assadi Y., Milani Hosseini M.R., . Monitoring of selenium in water samples using dispersive liquid–liquid microextraction followed by iridium-modified tube graphite furnace atomic absorption spectrometry. Microchem. J.. 2007;87:6-12.
[Google Scholar]
Brereton R.G., . Introduction to multivariate calibration in analytical chemistry. Analyst. 2000;125:2125-2154.
[Google Scholar]
Doner G., Ege A., . Determination of copper, cadmium and lead in seawater and mineral water by flame atomic absorption spectrometry after co-precipitation with aluminum hydroxide. Anal. Chim. Acta. 2005;547:14-17.
[Google Scholar]
Fan Z., . Determination of antimony(III) and total antimony by single-drop microextraction combined with electrothermal atomic absorption spectrometry. Anal. Chim. Acta. 2007;585:300-304.
[Google Scholar]
Fearn T., . On orthogonal signal correction. Chemom. Intell. Lab. Syst.. 2000;50:47-52.
[Google Scholar]
Gao Y., Wu P., Li W., Xuan Y., Hou X., . Simultaneous and selective preconcentration of trace Cu and Ag by one-step displacement cloud point extraction for FAAS determination. Talanta. 2010;81:586-590.
[Google Scholar]
Geladi O., Kowalski B.R., . Partial least-squares regression: a tutorial. Anal. Chim. Acta. 1986;185:1-17.
[Google Scholar]
Gerlach R.W., Kowalski B.R., Wold H., . Partial least-squares path modelling with latent variables. Anal. Chim. Acta. 1979;112:417-421.
[Google Scholar]
Gharehbaghi M., Shemirani F., Baghdadi M., . Dispersive liquid–liquid microextraction and spectrophotometric determination of cobalt in water samples. Int. J. Environ. Anal. Chem.. 2008;88:513-523.
[Google Scholar]
Haaland D.M., Thomas E.V., . Partial least-squares methods for spectral analyses. 1. Relation to other quantitative calibration methods and the extraction of qualitative information. Anal. Chem.. 1988;60:1193-1202.
[Google Scholar]
Haaland D.M., Thomas E.V., . Comparison of multivariate calibration methods for quantitative spectral analysis. Anal. Chem.. 1990;62:1091-1099.
[Google Scholar]
Joreskog K.G., Wold H., . System Under Indirect Observation. Amsterdam: North Holland; 1982.
Kara D., Alkan M., . Preconcentration and separation of copper (II) with solvent extraction using N, N′-bis(2-hydroxy-5-bromo-benzyl)1,2 diaminopropane. Microchem. J.. 2002;71:29-39.
[Google Scholar]
Khajeh M., . Application of Box–Behnken design in the optimization of a magnetic nanoparticle procedure for zinc determination in analytical samples by inductively coupled plasma optical emission spectrometry. J. Hazard. Mater.. 2009;172:385-389.
[Google Scholar]
Khajehsharifi H., Sadeghi M., Pourbasheer E., . Spectrophotometric simultaneous determination of creatine, creatinine, and uric acid in real samples by orthogonal signal correction–partial least squares regression. Monatsh. Chem.. 2009;140:685-691.
[Google Scholar]
Liang P., Sang H., . Determination of trace lead in biological and water samples with dispersive liquid-liquid microextraction preconcentration. Anal. Biochem.. 2008;380:21-25.
[Google Scholar]
Liang P., Peng L., Yan P., . Speciation of As(III) and As(V) in water samples by dispersive liquid-liquid microextraction separation and determination by graphite furnace atomic absorption spectrometry. Microchim. Acta. 2009;166:47-52.
[Google Scholar]
Lorber A., Faber K., Kowalski B.R., . Net analyte signal calculation in multivariate calibration. Anal. Chem.. 1997;69:1620-1626.
[Google Scholar]
Macedo S.M., de Jesus R.M., Garcia K.S., Hatje V., de Queiroz A.F., Ferreira S.L.C., . Determination of total arsenic and arsenic (III) in phosphate fertilizers and phosphate rocks by HG-AAS after multivariate optimization based on Box–Behnken design. Talanta. 2009;80:974-979.
[Google Scholar]
Mendham J., Denney R.C., Barnes J.D., Thomas M.J.K., . Vogels’s Quantative Chemical Analysis (sixth ed.). London: Longmann; 1998.
Nagaraju D., Huang Sh.D., . Determination of triazine herbicides in aqueous samples by dispersive liquid–liquid microextraction with gas chromatography–ion trap mass spectrometry. J. Chromatogr. A. 2007;1161:89-97.
[Google Scholar]
Naseri M.T., Hemmatkak P., Milani Hosseini M.R., Assadi Y., . Combination of dispersive liquid–liquid microextraction with flame atomic absorption spectrometry using microsample introduction for determination of lead in water samples. Anal. Chim. Acta. 2008;610:135-141.
[Google Scholar]
Ni Y., Wang Y., Kobot S., . Simultaneous kinetic-spectrophotometric determination of maltol and ethyl maltol in food samples by using chemometrics. Food Chem.. 2008;109:431-438.
[Google Scholar]
Niazi A., Goodarzi M., . Orthogonal signal correction–partial least squares method for simultaneous spectrophotometric determination of cypermethrin and tetramethrin. Spectrochim. Acta Part A. 2008;69:1165-1169.
[Google Scholar]
Niazi A., Yazdanipour A., . Spectrophotometric simultaneous determination of nitrophenol isomers by orthogonal signal correction and partial least squares. J. Hazard. Mater.. 2007;146:421-427.
[Google Scholar]
Niazi A., Ghasemi J., Yazdanipour A., . PARAFAC decomposition of three-way kinetic-spectrophotometric spectral matrices based on phosphomolybdenum blue complex chemistry for nitrite determination in water and meat samples. Anal. Lett.. 2005;38:2377-2392.
[Google Scholar]
Niazi A., Ghasemi J., Zendehdel M., . Simultaneous voltammetric determination of morphine and noscapine by adsorptive differential pulse stripping method and least-squares support vector machines. Talanta. 2007;74:247-254.
[Google Scholar]
Niazi A., Azizi A., Ramezani M., . Simultaneous spectrophotometric determination of mercury and palladium with Thio-Michler’s Ketone using partial least squares regression and orthogonal signal correction. Spectrochim. Acta Part A. 2008;71:1172-1177.
[Google Scholar]
Niazi A., Momeni-Isfahani T., Ahmari Z., . Spectrophotometric determination of mercury in water samples after cloud point extraction using nonionic surfactant Triton X-114. J. Hazard. Mater.. 2009;165:1200-1203.
[Google Scholar]
Pierna J.A.F., Massart D.L., de Noord O.E., Ricoux Ph., . Direct orthogonalization: some case studies. Chemom. Intell. Lab. Syst.. 2001;55:101-108.
[Google Scholar]
Rezaee M., Assadi Y., Milani Hosseini S.M.R., Aghaee E., Ahmadi F., Berijani S., . Determination of organic compounds in water using dispersive liquid–liquid microextraction. J. Chromatogr. A. 2006;1116:1-9.
[Google Scholar]
Rivas R.E., Lopez-Garcia L., Hernandez-Cordoba M., . Speciation of very low amounts of arsenic and antimony in waters using dispersive liquid–liquid microextraction and electrothermal atomic absorption spectrometry. Spectrochim. Acta Part B. 2009;64:329-333.
[Google Scholar]
Shoaee H., Rosgdi M., Khanlarzadeh N., Beiraghi A., . Simultaneous preconcentration of copper and mercury in water samples by cloud point extraction and their determination by inductively coupled plasma atomic emission spectrometry. Spectrochim. Acta Part A. 2012;98:70-75.
[Google Scholar]
Sjoblom J., Svensson O., Josefson M., Kullberg H., Wold S., . An evaluation of orthogonal signal correction applied to calibration transfer of near infrared spectra. Chemom. Intell. Lab. Syst.. 1998;44:229-244.
[Google Scholar]
Starvin A.M., Prasada Rao T., . Removal and recovery of mercury(II) from hazardous wastes using 1-(2-thiazolylazo)-2-naphthol functionalized activated carbon as solid phase extractant. J. Hazard. Mater.. 2004;113:75-79.
[Google Scholar]
Tobiasz A., Walas S., Soto Hernandez A., Mrowiec H., . Application of multiwall carbon nanotubes impregnated with 5-dodecylsalicylaldoxime for on-line copper preconcentration and determination in water samples by flame atomic absorption spectrometry. Talanta. 2012;96:89-95.
[Google Scholar]
Valderrama P., Poppi R.J., . Bilinear least squares and molecular fluorescence in the quantification of the propranolol enantiomers. Anal. Chim. Acta. 2008;623:38-45.
[Google Scholar]
Walas S., Tobiasz A., Gawin M., Trzewik B., Stronjny M., Mrowiec H., . Application of a metal ion-imprinted polymer based on salen-Cu complex to flow injection preconcentration and FAAS determination of copper. Talanta. 2008;76:96-101.
[Google Scholar]
Westerhuis J.A., de Jong S., Smilde A.K., . Direct orthogonal signal correction. Chemom. Intell. Lab. Syst.. 2001;56:13-25.
[Google Scholar]
Wold S., Antii H., Lindgren F., Ohman J., . Orthogonal signal correction of near-infrared spectra. Chemom. Intell. Lab. Syst.. 1988;44:175-185.
[Google Scholar]
Wold S., Trygg J., Berglund A., Antii H., . Some recent developments in PLS modeling. Chemom. Intell. Lab. Syst.. 2001;58:131-150.
[Google Scholar]
Xiao-Huan Z., Qiu-Hua W., Mei-Yue Z., Guo-Hong Z., Zhi W., . Developments of dispersive liquid–liquid microextraction technique. Chin. J. Anal. Chem.. 2009;37:161-168.
[Google Scholar]
Zapata-Urzua C., Perez-Ortiz M., Bravo M., Olivieri A.C., Alvarez-Lueje A., . Simultaneous voltammetric determination of levodopa, carbidopa and benserazide in pharmaceuticals using multivariate calibration. Talanta. 2010;82:962-968.
[Google Scholar]
Zhang Q., Minami H., Inoue S., Astuya I., . Differential determination of trace amounts of arsenic(III) and arsenic(V) in seawater by solid sampling atomic absorption spectrometry after preconcentration by co-precipitation with a nickel–pyrrolidine dithiocarbamate complex. Anal. Chim. Acta. 2004;508:99-105.
[Google Scholar]

Show Sections

[1] Andersson C.A., . Direct orthogonalization. Chemom. Intell. Lab. Syst.. 1999;47:51-63.
[Google Scholar]

[2] Bidari A., Zeini Jahromi E., Assadi Y., Milani Hosseini M.R., . Monitoring of selenium in water samples using dispersive liquid–liquid microextraction followed by iridium-modified tube graphite furnace atomic absorption spectrometry. Microchem. J.. 2007;87:6-12.
[Google Scholar]

[3] Brereton R.G., . Introduction to multivariate calibration in analytical chemistry. Analyst. 2000;125:2125-2154.
[Google Scholar]

[4] Doner G., Ege A., . Determination of copper, cadmium and lead in seawater and mineral water by flame atomic absorption spectrometry after co-precipitation with aluminum hydroxide. Anal. Chim. Acta. 2005;547:14-17.
[Google Scholar]

[5] Fan Z., . Determination of antimony(III) and total antimony by single-drop microextraction combined with electrothermal atomic absorption spectrometry. Anal. Chim. Acta. 2007;585:300-304.
[Google Scholar]

[6] Fearn T., . On orthogonal signal correction. Chemom. Intell. Lab. Syst.. 2000;50:47-52.
[Google Scholar]

[7] Gao Y., Wu P., Li W., Xuan Y., Hou X., . Simultaneous and selective preconcentration of trace Cu and Ag by one-step displacement cloud point extraction for FAAS determination. Talanta. 2010;81:586-590.
[Google Scholar]

[8] Geladi O., Kowalski B.R., . Partial least-squares regression: a tutorial. Anal. Chim. Acta. 1986;185:1-17.
[Google Scholar]

[9] Gerlach R.W., Kowalski B.R., Wold H., . Partial least-squares path modelling with latent variables. Anal. Chim. Acta. 1979;112:417-421.
[Google Scholar]

[10] Gharehbaghi M., Shemirani F., Baghdadi M., . Dispersive liquid–liquid microextraction and spectrophotometric determination of cobalt in water samples. Int. J. Environ. Anal. Chem.. 2008;88:513-523.
[Google Scholar]

[11] Haaland D.M., Thomas E.V., . Partial least-squares methods for spectral analyses. 1. Relation to other quantitative calibration methods and the extraction of qualitative information. Anal. Chem.. 1988;60:1193-1202.
[Google Scholar]

[12] Haaland D.M., Thomas E.V., . Comparison of multivariate calibration methods for quantitative spectral analysis. Anal. Chem.. 1990;62:1091-1099.
[Google Scholar]

[13] Joreskog K.G., Wold H., . System Under Indirect Observation. Amsterdam: North Holland; 1982.

[14] Kara D., Alkan M., . Preconcentration and separation of copper (II) with solvent extraction using N, N′-bis(2-hydroxy-5-bromo-benzyl)1,2 diaminopropane. Microchem. J.. 2002;71:29-39.
[Google Scholar]

[15] Khajeh M., . Application of Box–Behnken design in the optimization of a magnetic nanoparticle procedure for zinc determination in analytical samples by inductively coupled plasma optical emission spectrometry. J. Hazard. Mater.. 2009;172:385-389.
[Google Scholar]

[16] Khajehsharifi H., Sadeghi M., Pourbasheer E., . Spectrophotometric simultaneous determination of creatine, creatinine, and uric acid in real samples by orthogonal signal correction–partial least squares regression. Monatsh. Chem.. 2009;140:685-691.
[Google Scholar]

[17] Liang P., Sang H., . Determination of trace lead in biological and water samples with dispersive liquid-liquid microextraction preconcentration. Anal. Biochem.. 2008;380:21-25.
[Google Scholar]

[18] Liang P., Peng L., Yan P., . Speciation of As(III) and As(V) in water samples by dispersive liquid-liquid microextraction separation and determination by graphite furnace atomic absorption spectrometry. Microchim. Acta. 2009;166:47-52.
[Google Scholar]

[19] Lorber A., Faber K., Kowalski B.R., . Net analyte signal calculation in multivariate calibration. Anal. Chem.. 1997;69:1620-1626.
[Google Scholar]

[20] Macedo S.M., de Jesus R.M., Garcia K.S., Hatje V., de Queiroz A.F., Ferreira S.L.C., . Determination of total arsenic and arsenic (III) in phosphate fertilizers and phosphate rocks by HG-AAS after multivariate optimization based on Box–Behnken design. Talanta. 2009;80:974-979.
[Google Scholar]

[21] Mendham J., Denney R.C., Barnes J.D., Thomas M.J.K., . Vogels’s Quantative Chemical Analysis (sixth ed.). London: Longmann; 1998.

[22] Nagaraju D., Huang Sh.D., . Determination of triazine herbicides in aqueous samples by dispersive liquid–liquid microextraction with gas chromatography–ion trap mass spectrometry. J. Chromatogr. A. 2007;1161:89-97.
[Google Scholar]

[23] Naseri M.T., Hemmatkak P., Milani Hosseini M.R., Assadi Y., . Combination of dispersive liquid–liquid microextraction with flame atomic absorption spectrometry using microsample introduction for determination of lead in water samples. Anal. Chim. Acta. 2008;610:135-141.
[Google Scholar]

[24] Ni Y., Wang Y., Kobot S., . Simultaneous kinetic-spectrophotometric determination of maltol and ethyl maltol in food samples by using chemometrics. Food Chem.. 2008;109:431-438.
[Google Scholar]

[25] Niazi A., Goodarzi M., . Orthogonal signal correction–partial least squares method for simultaneous spectrophotometric determination of cypermethrin and tetramethrin. Spectrochim. Acta Part A. 2008;69:1165-1169.
[Google Scholar]

[26] Niazi A., Yazdanipour A., . Spectrophotometric simultaneous determination of nitrophenol isomers by orthogonal signal correction and partial least squares. J. Hazard. Mater.. 2007;146:421-427.
[Google Scholar]

[27] Niazi A., Ghasemi J., Yazdanipour A., . PARAFAC decomposition of three-way kinetic-spectrophotometric spectral matrices based on phosphomolybdenum blue complex chemistry for nitrite determination in water and meat samples. Anal. Lett.. 2005;38:2377-2392.
[Google Scholar]

[28] Niazi A., Ghasemi J., Zendehdel M., . Simultaneous voltammetric determination of morphine and noscapine by adsorptive differential pulse stripping method and least-squares support vector machines. Talanta. 2007;74:247-254.
[Google Scholar]

[29] Niazi A., Azizi A., Ramezani M., . Simultaneous spectrophotometric determination of mercury and palladium with Thio-Michler’s Ketone using partial least squares regression and orthogonal signal correction. Spectrochim. Acta Part A. 2008;71:1172-1177.
[Google Scholar]

[30] Niazi A., Momeni-Isfahani T., Ahmari Z., . Spectrophotometric determination of mercury in water samples after cloud point extraction using nonionic surfactant Triton X-114. J. Hazard. Mater.. 2009;165:1200-1203.
[Google Scholar]

[31] Pierna J.A.F., Massart D.L., de Noord O.E., Ricoux Ph., . Direct orthogonalization: some case studies. Chemom. Intell. Lab. Syst.. 2001;55:101-108.
[Google Scholar]

[32] Rezaee M., Assadi Y., Milani Hosseini S.M.R., Aghaee E., Ahmadi F., Berijani S., . Determination of organic compounds in water using dispersive liquid–liquid microextraction. J. Chromatogr. A. 2006;1116:1-9.
[Google Scholar]

[33] Rivas R.E., Lopez-Garcia L., Hernandez-Cordoba M., . Speciation of very low amounts of arsenic and antimony in waters using dispersive liquid–liquid microextraction and electrothermal atomic absorption spectrometry. Spectrochim. Acta Part B. 2009;64:329-333.
[Google Scholar]

[34] Shoaee H., Rosgdi M., Khanlarzadeh N., Beiraghi A., . Simultaneous preconcentration of copper and mercury in water samples by cloud point extraction and their determination by inductively coupled plasma atomic emission spectrometry. Spectrochim. Acta Part A. 2012;98:70-75.
[Google Scholar]

[35] Sjoblom J., Svensson O., Josefson M., Kullberg H., Wold S., . An evaluation of orthogonal signal correction applied to calibration transfer of near infrared spectra. Chemom. Intell. Lab. Syst.. 1998;44:229-244.
[Google Scholar]

[36] Starvin A.M., Prasada Rao T., . Removal and recovery of mercury(II) from hazardous wastes using 1-(2-thiazolylazo)-2-naphthol functionalized activated carbon as solid phase extractant. J. Hazard. Mater.. 2004;113:75-79.
[Google Scholar]

[37] Tobiasz A., Walas S., Soto Hernandez A., Mrowiec H., . Application of multiwall carbon nanotubes impregnated with 5-dodecylsalicylaldoxime for on-line copper preconcentration and determination in water samples by flame atomic absorption spectrometry. Talanta. 2012;96:89-95.
[Google Scholar]

[38] Valderrama P., Poppi R.J., . Bilinear least squares and molecular fluorescence in the quantification of the propranolol enantiomers. Anal. Chim. Acta. 2008;623:38-45.
[Google Scholar]

[39] Walas S., Tobiasz A., Gawin M., Trzewik B., Stronjny M., Mrowiec H., . Application of a metal ion-imprinted polymer based on salen-Cu complex to flow injection preconcentration and FAAS determination of copper. Talanta. 2008;76:96-101.
[Google Scholar]

[40] Westerhuis J.A., de Jong S., Smilde A.K., . Direct orthogonal signal correction. Chemom. Intell. Lab. Syst.. 2001;56:13-25.
[Google Scholar]

[41] Wold S., Antii H., Lindgren F., Ohman J., . Orthogonal signal correction of near-infrared spectra. Chemom. Intell. Lab. Syst.. 1988;44:175-185.
[Google Scholar]

[42] Wold S., Trygg J., Berglund A., Antii H., . Some recent developments in PLS modeling. Chemom. Intell. Lab. Syst.. 2001;58:131-150.
[Google Scholar]

[43] Xiao-Huan Z., Qiu-Hua W., Mei-Yue Z., Guo-Hong Z., Zhi W., . Developments of dispersive liquid–liquid microextraction technique. Chin. J. Anal. Chem.. 2009;37:161-168.
[Google Scholar]

[44] Zapata-Urzua C., Perez-Ortiz M., Bravo M., Olivieri A.C., Alvarez-Lueje A., . Simultaneous voltammetric determination of levodopa, carbidopa and benserazide in pharmaceuticals using multivariate calibration. Talanta. 2010;82:962-968.
[Google Scholar]

[45] Zhang Q., Minami H., Inoue S., Astuya I., . Differential determination of trace amounts of arsenic(III) and arsenic(V) in seawater by solid sampling atomic absorption spectrometry after preconcentration by co-precipitation with a nickel–pyrrolidine dithiocarbamate complex. Anal. Chim. Acta. 2004;508:99-105.
[Google Scholar]

Preconcentration and simultaneous spectrophotometric determination of copper and mercury by dispersive liquid–liquid microextraction and orthogonal signal correction–partial least squares

Abstract

Keywords

Copper

Mercury

Determination

Box–Behnken design

DLLME

OSC–PLS

1 Introduction

2 Experimental

2.1 Instrument and software

2.2 Reagents and materials

2.3 Dispersive liquid–liquid microextraction procedures

2.4 Sample preparation before DLLME

3 Results and discussion

3.1 Experimental design optimization

3.2 Interferences study

3.3 Analytical performance

3.4 Multivariate calibration

3.5 Preprocessing by orthogonal signal correction

3.6 Selection of the optimum number of factors

3.7 Simultaneous determination of copper and mercury in synthetic and real matrix samples

4 Conclusion

References

Suggested read for related articles:

We value your privacy