International Journal of Environmental Analytical

Chemistry

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/geac20

Determination of copper from environmental

samples by solvent microextraction method using
AAS. Multivariate modelling and factorial design

Waheed Ali Soomro, Muhammad Yar Khuhawar, Taj Muhammad Jahangir,

Muhammad Farooque Lanjwani & Imran Khan Rind

To cite this article: Waheed Ali Soomro, Muhammad Yar Khuhawar, Taj Muhammad Jahangir,
Muhammad Farooque Lanjwani & Imran Khan Rind (2023): Determination of copper from
environmental samples by solvent microextraction method using AAS. Multivariate modelling
and factorial design, International Journal of Environmental Analytical Chemistry, DOI:
10.1080/03067319.2023.2183360

To link to this article: https://doi.org/10.1080/03067319.2023.2183360

View supplementary material

Published online: 27 Feb 2023.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=geac20
INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY
https://doi.org/10.1080/03067319.2023.2183360

Determination of copper from environmental samples by

solvent microextraction method using AAS. Multivariate
modelling and factorial design
Waheed Ali Soomroa, Muhammad Yar Khuhawara, Taj Muhammad Jahangira,
Muhammad Farooque Lanjwanib and Imran Khan Rindc
a
Institute of Advanced Research Studies Chemical Sciences University of Sindh, Jamshoro, Pakistan; bDr. M.A.
Kazi Institute of Chemistry, University of Sindh, Jamshoro, Pakistan; cNational Centre for Excellence in
Analytical Chemistry University of Sindh, Jamshoro, Pakistan

ABSTRACT ARTICLE HISTORY

The dispersive liquid–liquid microextraction methods were devel­ Received 26 December 2022
oped, based on the enrichment and separation of copper after Accepted 14 February 2023
complexation with bis(salicylaldehyde)ethylenediamine (H2SA2en) KEYWORDS
from environmental samples. The quantitation was made by the DLLME; copper; flame atomic
flame atomic absorption spectrometric methods. The effect of para­ absorption
meters pH, derivatising reagent (H2SA2en) concentration, the spectrophotometry;
volume, extracting solvent, disperser solvent, centrifugation time environmental samples
and temperature on analytical signals and the extraction efficiency
of metal ion was examined and optimised using univalent and
multivalent (factorial design) techniques. The linearity of calibration
curve was in the range of 5–30 ng/ml with R2 = 0.9907. Limit of
detection was 2.0 ng/ml and limit of quantification was 6.0 ng/ml,
and the enrichment factor (21.5), repeatability (n = 4) and robust­
ness of the determination were also evaluated. Method was effec­
tively employed for the estimation of copper from the samples of
water and sediments. Samples were collected from Darawat Dam
and River Indus. The results obtained were matched with reported
methods, and comparable sensitivity and selectivity was indicated.

1. Introduction
The analysis of metal in food, water and different substances is of great consequence.
The extremely common analytical techniques for the analysis are flame atomic absorp­
tion spectrophotometry (FAAS) and inductively coupled plasma-mass spectrometry
[1,2]. The trace elements may be present in the samples near or below the limit of
detection (LOD) by the cited techniques and their quantitative determination may be
subjected to uncertainty [3]. To solve the problem, preconcentration and separation
from the matrix are required. A number of different procedures have been carried out
such as co-precipitation [4], liquid–liquid extraction [5] and solid phase extraction [6].
These procedures require time and efforts on larger amount of solvents [7]. To over­
come these problems, DLLME procedure has been established by Assadi and his co-

CONTACT Waheed Ali Soomro [email protected]

Supplemental data for this article can be accessed online at https://doi.org/10.1080/03067319.2023.2183360.
© 2023 Informa UK Limited, trading as Taylor & Francis Group
2 W. A. SOOMRO ET AL.

workers [8,9]. This method consists of a mixture of organic solvents, where highly
miscible mixture of solvents (extracting and dispersive solvents) is quickly injected into
aqueous solution containing the analyte. The process results into the development of
dispersive droplets of extracting solvent and efficient transfer of the analyte from the
aqueous to organic phases within short time [10]. The droplets are located at bottom
on centrifugation of cloudy solution. A syringe is used to remove droplets. The droplet
containing enriched analyte can be determined by GFAAS, FAAS, GC or HPLC [11,12].
The main benefits of DLLME include the simplicity of the procedure, low amount of the
solvent used, short time of extraction, high recovery, low cost of operation involved
and a green method, with less consumption of toxic solvents [13,14]. Another advan­
tage of the method is that total of the extracted solvent is introduced into the
detection system [15]. Before the analysis of trace metals, preconcentration and
separation steps are necessary, because of their matrix effects and small concentra­
tions in the environmental samples. The solvent extraction of metal complex is suitable
for metal preconcentration, preceding to FAAS [16].
The determination of Cu metal in water and sediment samples [1] has important
function in metabolism, antioxidant effects and cofactor in incorporation of iron in
haemoglobin [17].
Copper is a heavy metal, essential for the individual. In extra quantity, Cu
produces major problems in human body. It comes into the nature by various activities
of human beings.
Trindade et al. (2021) determined the Cu in coconut water using 2-(5-bromo-
2-pyridylazo)-5-(diethyl amino)phenol (5-Br-PADAP) and dithizone chelating reagents
[18], and Adhami et al. (2020) determined Cu2+ in vegetable oil at trace levels by
DLLME using ionic liquid [TBP] [PO4] [19]. Yilmaz and Soylak (2014) determined copper
by micro sampling (FAAS), by chelating with dimethyl dithiocarbamate and extracted
into supramolecular solvent phase [20]. Kocot et al. (2012) determined Cu in water
samples by DLLME and sodium diethyldithiocarbamate as a chelating agent [21]. Arain
et al. (2016) determined the copper in the samples of drinking water and serum based
on ionic liquid assisted micro emulsion (IL-mE-DLLME) [22]. Mortada et al. [23] devel­
oped simple preconcentration of Yb, Y, Gd, Sm, Er and Eu by inductively coupled
plasma optical emission spectrometry. Xie et al. [24] prepared molecularly imprinted
polymer for rapid removal and selective determination of alkylphenols from water. Gao
and Ma [25] determined the mercury from water using dispersive LLME combined by
HPLC technique. Li et al. [26] determined the bisphenol A from environmental water
using HPLC with the help of magnetic reduced graphene-oxide-based SPE coupled
with DLLME.
Therefore, the analysis of Cu2+ in water and different ecological samples is of great
importance. Proposed work is also considered for preconcentration and analysis of Cu2+
[1,13]. DLLME method is proposed for the separation of copper after complex formation
with bis(salicylaldehyde)ethylenediamine (H2SA2en) [27] from environmental sample
using FAAS method. A number of factors are examined and optimised. The method
indicated sensitivities at µg/l in the original samples. The copper is determined from the
samples collected from River Indus before and after Kotri Barrage and finally Indus is
draining water down to the sea. The samples of water and sediments were also analysed
from Darawat Dame used for irrigation and human consumption.
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 3

2. Experimental
2.1. Reagents and materials
Ethylenediamine, 2-hydroxybenzaldehyde (salicylaldehyde), chlorobenzene, chloroform (tri­
chloromethane), dichloroethane, nitric acid, hydrogen peroxide, carbon tetrachloride, acetic
acid, acetone, ethanol, acetonitrile, hydrochloric acid (37%), sodium acetate, methanol,
ammonium chloride, boric acid, ammonium acetate, sodium tetraborate and ammonia
solution were obtained from Merck Germany. Standard solution of Cu2+ ions (1000 μg/ml)
was made ready from copper sulphate (Merck, Germany).

2.2. Synthesis of reagent

Bis(salicylaldehyde)-ethylenediamine was synthesised and decontaminated in ethanol as
reported [28]. Salicylaldehyde (2.08 ml) and ethylenediamine (0.7 ml) were dissolved
distinctly in ethanol (25 ml). Addition of ethylenediamine solution was combined slowly
in to the salicylaldehyde solution. Solution was refluxed for about 1 h and cooled down at
room temperature. The yellow precipitates were filtered, dried and recrystalised from
ethanol. The melting point and FT-IR of the compound agreed with authenticated sample
[28]. H2SA2en solution (0.01 M) was prepared via dissolving the 0.067 g of the reagent in
ethanol and diluted up to 25 ml with ethanol.

2.3. Factorial design

The factorial design was built by Minitab software applying three coded level, low (−1),
middle (0) and high (+1), using Box-Behnken design (BBD) method [29]. The variables pH,
solvent volume (ml) and working time (min) were used to draw the factorial design
(Table 1). Sixteen experimental runs were used to prepare the BBD model and compared
with the variance of each run (Table 2). The factorial design is a powerful tool use for
estimating the optimal level and significance level of variables used in the model.

2.4. Analytical procedure

A 10.0-ml solution of Cu(II) at pH = 7 was shifted into screw-capped conical bottom and
test tube, 25.0 μl of H2SA2en reagent solution (0.01 M) was added as the complexing
reagent, and reaction solution was warmed at 90°C in water bath for 5 min and cooled.
Then, a mixture of acetonitrile (1.3 ml) and chloroform (130 μl) was quickly injected by the
Hamilton syringe into the solution, and the resulting solution appeared cloudy. The
cloudiness happened by admixtures of water, acetonitrile and chloroform. The resulting
Cu–SA2en complex was separated into fine-sized droplets of trichloromethane. The solu­
tion was centrifuged for 5 min at 5000 rpm, around 20°C, extracting solvent appeared in

Table 1. Levels of factors used in experimental design.

Variables Symbol Low (−1) Middle (0) High (+1)
pH A 3 7 11
B 0.8 1.2 1.6
Working time (min) C 5 20 30
4 W. A. SOOMRO ET AL.

Table 2. Experiments design for factorial design (BBD).

Run A B C Recovery%
1 1 1 0 91.25
2 0 −1 0 90.75
3 0 0 −1 83.0
4 −1 0 0 79.75
5 0 −1 0 83.0
6 −1 0 0 87.5
7 1 0 0 86.25
8 0 −1 −1 82.0
9 0 −1 1 89.0
10 0 0 0 97.5
11 1 0 0 94.0
12 0 0 0 97.5
13 0 1 0 89.0
14 1 0 −1 85.25
15 −1 0 −1 78.75
16 1 −1 0 85.25

tube with sedimentation, and approximately 75 ± 3 μl sediment solution was separated

by the help a syringe, shifted into vial and diluted through ethanol (140 μl). The concen­
tration was measured using FAAS. The absorbance of reagent blank was also measured.
A calibration curve was made via calibration graph through analytical signal against the
Cu(II) concentration in sequences of working standard solutions. The results were
obtained in triplicate (n = 3) analysis with 4-s delay integration time, wavelength 324.5,
lamp current 4, slit width 0.5 nm and acetylene flow 1.5. The instrument was controlled
through Winlab software.

2.5. Preparation of samples and analysis

All samples of water and sediments were selected from Indus River at Kotri Baradge and
Darawat Dame, Thano Bula Khan, Sindh, Pakistan. Samples were taken from different
locations of upstream and downstream. The water samples were filtered from Whatman
filter papers 42, and 10 ml in triplicate (n = 3) was taken from each samples and analysed
using standard procedures for analysis [30]. Three samples from each sampling station
were added 5 ng Cu solution and the general procedure was again carried out. The
quantitation was made from linear regression equation of external calibration curve and
rise in response with added standard. Dried sediment samples in oven at 60°C, 0.5 g of
dried sample was weighed and dissolved in 5 ml aqua regia (HNO3 and HCl in the 1:3
ratio). Solution was heated slowly near to dryness, and again 2 ml HCl was added and
heated slowly to near dryness. The residues were diluted in distilled water and then
filtered and final volume adjusted to 25 ml. Each sample (10 ml) was taken in duplicate (n
= 2) and general procedure of analysis was carried out.

2.6. Process for achieving extraction efficiency

For the standard solutions of the Cu(II), absorbance was measured through FAAS, orga­
nised in the series of 5.0–25.0 ng/ml following analytical procedure. The straight line
calibration graph appeared through the analytical signal gained from the solutions.
Cu(II) concentration was estimated from linear regression equation of calibration curve.
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 5

The extraction efficiency was determined from an aqueous solution (10 ml) utilising the
different masses of Cu following the procedure of DLLME.

2.7. Computational modelling study

The computational study was conducted to observe the theoretical interactions of Cu (II)
metal with bis (salicylaldehyde)ethylenediamine (Schiff base). The molecular structures
were drawn using Gauss view program of Gaussian 09W package. The optimisation of
molecular models was done by UFF level of mechanics method. The optimised structures
indicated chemical and physical interactions of Cu metal at different functional groups of
Schiff base.

3. Results and discussion

Cu(II) ions reacted with bis(salicylaldehyde))-ethylenediamine (H2SA2en) reagent to form
(1:1) complex and was examined at the experimental conditions. The Cu complex was
extracted into the organic solvent. To optimise the various factors like pH, types and
volumes of the extracting solvent, as well as the disperser solvent, concentration of the
complex reagent and signal for higher enrichment factor were examined. The univariate
and multivariate strategies were used for optimization optimisation.

3.1. Analysis of variance (ANOVA)

The ANOVA was calculated on the basis of P-values and F-values [31]. The F-values were
2.54 to 752.42 and P-values from 0.0 to 0.112. The results showed that P-values of
variables pH (A), solvent volume (B) and working time (C) and combined variables AA
and BB showed significant effects because their values were less than 0.05 at 95%
confidence interval. The variables CC, AB and AC had P-value higher than 0.05 and did
not exhibit a significant effect on the recovery of copper. It was revealed that when
P-value was higher F-value was lower (Table 3).

Table 3. Analysis of variance (ANOVA).

Source DF Adj SS Adj MS F-Value P-Value
Model 13 2.476 0.1456 253.02 0.000
Linear 3 1.345 0.423 453.54 0.000
A 1 0.432 0.425 345.43 0.040
B 1 0.312 0.354 261.34 0.004
C 1 0.432 0.398 389.45 0.003
D 1 0.523 0.497 487.60 0.050
A*A 1 0.678 0.906 782.42 0.016
B*B 1 0.268 0.361 10.63 0.043
C*C 1 0.134 0.325 43.52 0.068
A*B 1 0.231 0.135 95.45 0.056
A*C 1 0.0005 0.002 2.54 0.112
6 W. A. SOOMRO ET AL.

Figure 1. Response surface plot pH versus solvent volume.

3.2. Response surface plots

The response surface plots were drawn with the help of Minitab software version 22. The
3D response plots showed the joint effects of two factors versus recovery % [32]. The 3D
response surface (Figure 1) presented that maximum response was observed at neutral
pH 7 and also surface height was higher when increased the solvent volume up to 1.2 ml.
The 3D surface plot (Figure 2) indicated the highest response was also observed at
working time 20 min and response was increased when increased the solvent volume

Figure 2. Response surface plot solvent volume versus working time.

 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 7

up to 1.2 ml, then decreased with increased solvent volume. The highest response was
also observed at working time 20 min for recovery of copper from different samples using
DLLME method.

3.3. Pareto chart

Pareto chart was utilised to examine the significant and insignificant effects of the
variables used in factorial design [29]. The results indicated the variables pH (A), solvent
volume (B) and working time (C) and combined variables AA, crossed the reference line
3.52 and were considered as significant. Other variables, BB, CC, AB and AC did not cross
the reference line and indicated no significant effects on the recovery of Cu (Figure S1).

3.4. Normal and half normal plots

The normal probability plot was built by using Minitab software using three variables and
significant level of factors was assessed [33]. The normal probability plot displayed that
the majority of variables were significant such as A, B and C and combined variables CC,
AA and BB. However, variables AB and AC did not show significant effects on the recovery
of the copper from different samples (Figure S2).

3.5. Effect of pH
The pH value was optimised in a formation of the complex (Cu-SA2en) and its successive
extraction. For this purpose, the series of pH from 3 to 10 was examined. The results found
(Figure 3(a)) showed maximum signal for copper at pH 7 and was selected. The ionic
strength of the solution was calculated as 0.005 Mol/L l l.

3.6. Type of extracting solvent

The nature of the extracting solvent is a factor, which influences on the extraction
proficiency and the capability to extract the compound. Therefore, different solvents
chloroform (CHCl3), 1,3-dichloroethane (C2H4Cl2), carbon tetrachloride (CCl4) and
chlorobenzene(C6H5Cl) at 130 μl were investigated with 1.30 ml of disperser solvent
(acetonitrile) and were distinctly injected into solutions. The solutions were warmed at
90°C in water bath before addition of organic solvents. Subsequently, the water solubility
of extracting solvents is not the same, therefore, solutions were cooled at 20°C to lower
the solubility of solvents. The results obtained (Figure 3(b)) presented that extraction
efficiencies of chloroform, 1,2-dichloroethane, carbon tetrachloride and chlorobenzene
were 98%, 87%, 72% and 56%, respectively. Therefore, chloroform was preferred as
suitable extracting solvent.

3.7. Volume of extracting solvent

The volume of extracting solvent was similarly investigated by making combinations of
1.3 ml of acetonitrile and various volumes of chloroform as shown in Table 4. The
efficiency of extraction gradually raised from 30 to 130 μl volume, then regularly
8 W. A. SOOMRO ET AL.

Figure 3. (a) pH versus concentration, (b) extracting solvents versus concentration, (c) dispersive
solvents versus concentration, (d) volume of solvents versus absorbance and (e) reagent concentration
versus absorbance.

Table 4. Effect of extraction solvent (chloroform) volume on the analytical signal and extraction
efficiency of copper obtained from DLLME.
Sediment
phase Ethanol as diluting solvent Extraction efficiency
Chloroform (µl) (µl) (µl) Recovery of Cu in ng/ml (%)
30 10 170 8.33 41
50 20 160 12.66 63
70 30 150 15.85 79
90 40 140 17.23 86
110 50 130 18.89 94
130 60 120 19.35 96
150 70 110 18.82 94
170 80 100 17.77 88

decreased. The reason behind it that the volume of extracting solvent greater than 130 μl,
the formation of turbid solution was unstable due to larger droplets produced. Thus,
surface area was reduced between the aqueous and extracting solvent phases, resulting
in decreasing in extraction efficiency as well as the mass transfer of complex Cu-SA2en
from aqueous phase into chloroform. Thus, the reaction mixture was optimised at 130 μl
volume of chloroform and 75 ± 3 μl volume of the sedimented phase was estimated.
Ethanol was used as diluting solvent for sediment phase. Due to variable volume of
sediment, the solutions were therefore diluted up to 180 μl of ethanol applying different
volumes. Hence, it was required to study the impact of volume of ethanol on efficiency.
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 9

The sequence of 10.0 ml water sample at pH = 7.0, contents of Cu(II) 20.0 ng/ml, the
mixture of 1.30 ml of acetonitrile and 30, 50, 70, 90, 110, 130, 150 and 170 μl of chloroform
was injected and the Cu-SA2en complex was extracted. The sedimented phase was
shifted into a vial after centrifugation. Then diluted with 170, 160, 150, 140, 130,
120,110 and 100 μl of ethanol respectively to obtain total volumes of 180 μl. The solutions
were analysed by FAAS. The best efficiency of extraction was observed 96% using 130 μl
chloroform sedimented to 60 μl diluted with 120 μl of ethanol as shown in Table 4.

3.8. Nature of disperser solvent

Disperser solvent having capability to mix with water and extracting solvent.
Consequently, the effect of disperser solvent (acetone, methyl alcohol, ethanol and
acetonitrile 1.30 ml volume) on signal of Cu(II) ions was examined with 130.0 μl of
extracting solvent (chloroform). The extraction efficacies were found for acetone 18 ng/
ml (90%), methanol 6.23 ng/ml (31%), ethanol 9.95 ng/ml (50%) and acetonitrile 19.21
(96%) (Figure 3(c)). The best results were achieved utilising acetonitrile; therefore, it was
utilised as disperser solvent in all consequent experiments.

3.9. Volume of disperser solvent

After selecting acetonitrile as the disperser solvent, it was required to adjust its volume.
The solutions with different volumes of acetonitrile (0.8–1.6 ml) were mixed with 130 μl of
CHCl3 and procedure was followed. As the volume of acetonitrile increased from 1.20 up
to 1.30 ml, product efficiency was increased, later, no any change occurred (Figure 3(d)).
When a lower volume of acetonitrile was employed, it might not diffuse chloroform
appropriately, and a unstable cloudy solution was produced, thus, the extraction ability
was lowered. Though, with a continuing increase in volume, owing to the formation of
slighter sized droplets of trichloromethane, surface area between extracting solvent and
aqueous phase increased. This phenomenon preceded to rise in extraction efficiency.
Based on obtained results and to generate stable cloudy solution, 1.3 ml of acetonitrile
was chosen as optimal volume.

3.10. Reagent concentration

The concentration of reagent (H2SA2en) as the complexing agent, has effect on the
extraction. It was analysed in the range of 1.5 × 10−5 to 4.0 × 10−5 M. Thus reagent
concentration of 2.5 × 10−5M was optimised and selected (Figure 3(e)).

3.11. Effect of centrifugation

The centrifugation time and temperature was estimated between 1 and 10 min at 30 to
18°C for five thousands revolve per minutes (5000 rpm). The results indicated that
analytical signal was directly proportional to centrifugation period up to 5 min at 20°C,
then no change was observed. In order to confirm sedimentation of extracting solvent,
a time period of 7 min was chosen as centrifugation time at 20 ̊C during the experiments.
10 W. A. SOOMRO ET AL.

Table 5. Determination of Copper under the following optimum

conditions.
Optimal factor Optimised conditions
Effect of pH 07
Selection of extracting solvent Chloroform
Selection of extracting solvent volume (µl) 130
Selection of disperser solvent Acetonitrile
Selection of disperser solvent volume (ml) 1.3
Effect of reagent concentration 2.5 × 10−5 M

3.12. Optimised condition for the determination of copper

All the parameters of the recommended technique were designed under the following
optimised conditions (Table 5).

3.13. Analytical characteristics for the determination of copper

The analytical features of the suggested method were considered applying the opti­
mised circumstances. Volume of the sample was 10.0 ml, and the standardisation
curve showed a linearity range of 0.005–0.03 μg/ml in the optimum circumstances.
The calculated regression equation was y = 0.1925x + 0.0337, where Y is analytical
signal (the disparity between the absorbance of sample and blank sample at 324.8
nm) and x represent the concentration of Cu(II) (μg/ml). The coefficient of determina­
tion of the calibration curve was R 2 = 0.9907, which showed a good linearity in
revealed concentration range (Figure 4). The LOD 0.002 μg/ml and LOQ 0.006 μg/ml.
Repeatability RSD % (n = 4) ± 0.56–5.41 with recovery% 97–104. A review of analytical
features of DLLME method for Cu+2 detection is described in Table 6.

0.7

0.6 y = 0.1925x + 0.0337

R² = 0.9907
0.5
Absorbance

0.4

0.3

0.2

0.1

0
0 1 2 3 4
Concentraon ug/ml

Figure 4. The linear calibration curve of Cu (II) by DLLME-FAAS method conditions: sample volume,
10.0 ml; Cu(II), 20 ng/ml; extraction solvent (chloroform) volume, 130 μl; disperser solvent (acetone)
volume, 1.3 ml; diluting solvent (ethanol) volume, 140 μl; centrifuge time, 5 min (at 5000 rpm).
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 11

Table 6. Analytical characteristics of the DLLME-FAAS

method for copper determination.
Parameter Analytical feature
Sample (ml) 10
Linear range (µg/ml) 0.005–0.03
Regression equation Y = 0.1925× + 0.0337
Limit of detection (µg/ml) 0.002
Limit of quantification (µg/ml) 0.006
Recovery % 97–104
Reproducibility RSD% (n = 4) 3.55–6.59
R2 0.9907

3.14. Analysis of real samples of water and sediment

The recommended method was employed to determine Cu(II) in water and sediment
samples. Four water samples were collected from Indus River Kotri Barrage, two upstream
and two downstream. The Cu contents upstream were 30.35–95.23 ng/ml, but the down­
stream Cu contents were slightly higher within 87.86–258.46 ng/ml, due to the extraction of
Cu from the sediments owing to the flow of water under pressure. Four water samples were
collected from Darawat Dam and Cu contents were within 53.24–95.13 ng/ml. The Cu con­
tents in water samples were within World Health Organization (WHO) limits is 2 mg/l for
drinking [WHO guidelines for drinking water quality 4th edition 2011 p.224–225]. Four samples
for total Cu contents in the bottom sediments at Indus River Kotri Barrage were analysed and
results indicated within 2612.5–10067.0 ng/g. Four samples of sediment were also analysed
from Darawat Dam and the Cu contents and were observed within 2508.6–2638.0 ng/g. The
Cu contents in sediments samples were within WHO limits is (100 mg/kg) [34]. All the samples
were analysed directly and by the standard addition process for verification of results. For the
analysis of water, 5 ng/10 ml of copper standard solution in one sample was added, while
another was without addition of standard (Table 7). For the samples of sediment, 5 ng/200 mg
of sediment was added as standard solution in one sample and an another sample was
without standard addition (Table 8). The calculation was carried out per gram while the

Table 7. Determination of copper from water samples of Indus River at Kotri Barrage and Darawat
Dam by DLLME-FAAS method.
Copper added Copper found
Sample ng/10 ml ng /ml or µg/l Recovery % RSD % Error %
Upstream water sample #1 0 30.353 99.919 ± 1.05 −0.081
5 35.226
Upstream water sample #2 0 95.238 98.691 ± 1.25 −1.308
5 98.926
Downstream water sample #1 0 258.46 100.39 ± 1.68 0.381
5 264.48
Downstream water sample # 2 0 87.861 97.469 ± 1.26 −2.531
5 90.511
Dam water sample #1 0 53.211 99.174 ± 0.564 −0.825
5 57.731
Dam water sample #2 0 95.134 101.23 ± 2.78 1.284
5 101.42
Dam water sample #3 0 73.671 97.144 ± 2.35 −2.855
5 76.433
Dam water sample #4 0 80.640 99.742 ± 1.29 −0.258
5 85.419
12 W. A. SOOMRO ET AL.

Table 8. Determination of copper from sediment samples Indus River Kotri Barrage and Drawat Dam
by DLLME-FAAS method.
Copper added Copper found
Sample ng/200 mg ng/g Recovery % RSD % Error %
Upstream sample #1 0 2872.27 101.58 ± 0.56 1.582
5 2917.66
Upstream sample #2 0 3547.53 97.620 ± 3.29 −2.379
5 3463.12
Downstream sample #1 0 10067.0 99.161 ± 0.63 −0.838
5 9982.51
Downstream sample # 2 0 2612.46 104.72 ± 5.41 4.722
5 2735.84
Dam sample #1 0 2508.57 98.705 ± 3.19 −1.294
5 2476.10
Dam sample #2 0 2638.44 99.753 ± 1.21 −0.246
5 2631.94
Dam sample #3 0 2586.49 102.76 ± 3.32 2.761
5 2657.92
Dam sample #4 0 2612.46 102.73 ± 1.64 2.734
5 2683.89

analysis was carried out per 200 mg. The following equation was used for the calculation of
percentage recovery (R%).

Cf
R% ¼ � 100
C0 þ CA

where Cf is the concentration of copper found for analyte in spiked sample, Co is the
concentration originate for analyte without addition and CA is the added concentration.
Acceptable recoveries were found for the spiked copper(II) ions by DLLME-FAAS method
from water samples in the range of 97–101% with RSD% ±0.56–2.78 (Table 7) and 97–
104% with RSD% ± 0.56–5.41 for sediment (Table 8).

3.15. Reproducibility of inter-day and intra-day

Intra-day accuracy was calculated based on the analysis of constant concentration (20 ng/
ml of copper) and four replicates (n = 4) were analysed on a same day. The RSD% was
found to be 3.55%. The inter-day precision was determined on three repeated days for 20
ng/ml copper concentration and was found with RSD% 6.49%.

3.16. Comparison of the work

The results obtained were matched with reported methods in terms of sensitivity,
selectivity and ease of analysis given in Table 9. The DLLME-FAAS method has several
benefits consisting of ease, high sensitivity, quickness, sharp extraction time, utilised the
little volume of organic solvents low LOD (0.002 µg/ml) and good the enrichment factor
(21.5) and is comparable with reported procedures using FAAS, but is somewhat less
sensitive than using GF-AAS, because both the procedures have different requirements.
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 13

Table 9. Comparison of the present method with reported methods for determination of copper.
LOD LOQ LR
Name of instruments Method µg l−1 µg l−1 µg l−1 ER (time min) PF, EF RSD% References
FAAS DLLME 3.32 4.83 0–30 5.0 11 6.6 [18]
GFAAS SBME 0.03 0.09 0.1–20 60 51.6 2.3 [35]
FAAS MF-LLME 100 300 300–4000 5.0 33 5.3 [36]
FAAS SS-LPME 5.8 19 2.6–200 0.015 23 9.8 [37]
SQT-FAAS DLLME 45 149 150–2000 2.0 2.0 6.1 [38]
SDIC RP-SHS-LLME 100 300 0.3–1000 5.0 4.6 6.9 [39]
FAAS DLLME 2.0 6.0 5.0–30 5.0 21.5 5.41 Present work
*Dispersive liquid–liquid microextraction (DLLME), Flame atomic absorption spectrometry (FAAS), Microfluidic-based
liquid–liquid microextraction (MF-LLME), Supra solvent-liquid phase microextraction (SS-LPME), Slotted quartz tube-
fame atomic absorption spectrometry (SQT-FAAS), Reversed-phase switchable-hydrophilicity solvent liquid–liquid
microextraction (RP-SHS-LLME), Smartphone digital image colorimetry (SDIC), Graphite furnace atomic absorption
spectrometry (GFAAS), Solvent bar micro-extraction (SBME), Limit of detection (LOD), Limit of quantification (LOQ),
Linear range (LR), Extraction recovery time (ER), Pre-concentration factor (PF), Enrichment factor (EF), Relative standard
deviation (RSD).

Table 10. Physical quantities of optimised molecular models.

Total Energy (a.u.) Bond Lengths (A°)
Cu 0.1289
Schiff base 0.0538
Cu–O (chem)-N (chem) 0.1058 O = 1.8851 N = 1.8719
Cu–O (chem)-N (phys) 0.1058 O = 1.8851 N = 1.8723
Cu–O (phys)-N (chem) 0.1058 O = 1.8850 N = 1.8719
Cu–O (phys)-N (phys) 0.1058 O = 1.8848 N = 1.8720

3.17. Computational modelling results

The results of optimised models were included for total energy values and bond lengths of
Cu interactions with different groups of Schiff base. The models studied are shown
(Table 10), and these include Cu metal, Schiff base and Cu interaction with various atoms
of Schiff base such as Cu bonds with O and N atoms. Structure such as Cu–O (chem)-N
(chem) indicates Cu interaction chemically with O and N atoms, Cu–O (chem)-N (phys)
indicates Cu interaction chemically with O and physically with N. Similarly, Cu–O (phys)-N
(chem) presents Cu physical interaction with O and chemical with N atoms and Cu–O (phys)-
N (phys) indicates Cu physical interaction with both O and N atoms as shown in Figure S3.
The total energy values measured for structures that resulted after Cu interactions were
almost same (0.1058 a.u.), whereas the variations computed in bond lengths were also not
significant. The average bond lengths measured for O chemical interaction were 1.8851
and physical 1.8849 whereas average bond lengths of N atoms were 1.8719 A for chemical
interaction and physical was 1.8721 A. Overall, it is concluded that the average bonds of
N atom with Cu are slightly shorter than O atoms so that the N atoms of Schiff base
interact more strongly with Cu than O atoms. Hence, according to these theoretical
computational models, Cu interaction with N functional groups is more favourable than
O containing functional groups.

4. Conclusion
Dispersive liquid–liquid microextraction method was established for preconcentration of
Cu2+ preceding to estimate by flame AAS technique. Recovery for the objective analyte
14 W. A. SOOMRO ET AL.

was found about 97–104%. The suggested method was employed for the quantity of Cu2+
in sediment and water samples. The acquired results were matched with those found by
a standard addition method, and comparatively good agreement was examined. The
method is fast, effective, easy and employs at a micro level (μl).

Disclosure statement
No potential conflict of interest was reported by the authors.

References
[1] G. Bagherian, M. Arab Chamjangali, H. Shariati Evari and M. Ashrafi, J. Anal. Sci. Technol. 10,
1–11 (2019). doi:10.1186/s40543-019-0164-6.
[2] M.N.M. Reyes and R.C. Campos, Talanta 70, 929–932 (2006).
[3] Y. Fan, C. Xu, R. Wang, G. Hu, J. Miao, K. Hai and C. Lin, J. Food Compos. Anal. 62, 63–68 (2017).
doi:10.1016/j.jfca.2017.05.003.
[4] N. Bader, A.A. Benkhayal and B. Zimmermann, Int. J. Chem. Sci. 12, 519–525 (2014).
[5] C.I. Silvestre, J.L. Santos, J.L. Lima and E.A. Zagatto, Anal. Chim. Acta 652, 54–65 (2009). doi:10.
1016/j.aca.2009.05.042.
[6] F. Khalilian and M. Rezaee, Food Anal. Methods 10, 885–891 (2017). doi:10.1007/s12161-016-
0653-9.
[7] A. Quigley, W. Cummins and D. Connolly, J. Chem. 2016 (2016). doi:10.1155/2016/4040165
[8] M. Rezaee, Y. Assadi, M.-R.M. Hosseini, E. Aghaee, F. Ahmadi and S. Berijani, J. Chromatogr. A.
1116, 1–9 (2006). doi:10.1016/j.chroma.2006.03.007.
[9] Z. Bahadır, V.N. Bulut, A. Mermer, N. Demirbaş, C. Duran and M. Soylak, J. Iran. Chem. Soc. 15,
1347–1354 (2018). doi:10.1007/s13738-018-1333-z.
[10] P. Viñas, N. Campillo, I. López-García and M. Hernández-Córdoba, Anal. Bioanal. Chem. 406,
2067–2099 (2014).
[11] L. Ming-Jie, H.-Y. Zhang, L. Xiao-Zhe, C. Chun-Yan and S. Zhi-Hong, Chin. J. Anal. Chem. 43,
1231–1240 (2015). doi:10.1016/S1872-2040(15)60851-9.
[12] W.A. Soomro, M.Y. Khuhawar, T.M.J. Khuhawar, M.F. Lanjwani and I.K. Rind, Int. Res. J. Med. Sci.
2, 10–14 (2020).
[13] G. Özzeybek, S. Erarpat, D.S. Chormey, M. Fırat, Ç. Büyükpınar, F. Turak and S. Bakırdere,
Microchem. J. 132, 406–410 (2017). doi:10.1016/j.microc.2017.02.031.
[14] M. Carabajal, C.M. Teglia, M.A. Maine and H.C. Goicoechea, Talanta 224, 121929 (2021). doi:10.
1016/j.talanta.2020.121929.
[15] E. Yilmaz and M. Soylak, Turk. J. Chem. 40, 868–893 (2016). doi:10.3906/kim-1605-26.
[16] M.T. Naseri, P. Hemmatkhah, M.R.M. Hosseini and Y. Assadi, Anal. Chim. Acta 610, 135–141
(2008). doi:10.1016/j.aca.2008.01.020.
[17] L. Banci, I. Bertini, F. Cantini and S. Ciofi-Baffoni, Cell. Mol. Life Sci. 67, 2563–2589 (2010).
doi:10.1007/s00018-010-0330-x.
[18] J.S. Trindade, V.A. Lemos, U.M.F.M. Cerqueira, C.G. Novaes, S.A. Araujo and M.A. Bezerra, Food
Chem. 365, 130473 (2021). doi:10.1016/j.foodchem.2021.130473.
[19] K. Adhami, H. Asadollahzadeh and M. Ghazizadeh, J. Food Compos. Anal. 89, 103457 (2020).
doi:10.1016/j.jfca.2020.103457.
[20] E. Yilmaz and M. Soylak, Talanta 126, 191–195 (2014). doi:10.1016/j.talanta.2014.03.053.
[21] K. Kocot, B. Zawisza and R. Sitko, Spectrochimica Acta Part B. 73, 79–83 (2012). doi:10.1016/j.
sab.2012.05.003.
[22] S.A. Arain, T.G. Kazi, H.I. Afridi, M.S. Arain, A.H. Panhwar, N. Khan, J.A. Baig and F. Shah,
Ecotoxicol. Environ. Saf. 126, 186–192 (2016). doi:10.1016/j.ecoenv.2015.12.035.
[23] W.I. Mortada, G.G. El-Gamal, M.M. Hassanien, A.A. Ibrahim and Y.G. Abou El-Reash, Int.
J. Environ. Anal. Chem. 1–15 (2022). doi:10.1080/03067319.2022.2100259.
 INTERNATIONAL JOURNAL OF ENVIRONMENTAL ANALYTICAL CHEMISTRY 15

[24] X. Xie, X. Ma, L. Guo, Y. Fan, G. Zeng, M. Zhang and J. Li, Chem. Eng. J. 357, 56–65 (2019).
doi:10.1016/j.cej.2018.09.080.
[25] Z. Gao and X. Ma, Anal. Chim. Acta 702 (1), 50–55 (2011). doi:10.1016/j.aca.2011.06.019.
[26] D. Li, X. Ma, R. Wang and Y. Yu, Anal. Bioanal. Chem. 409 (5), 1165–1172 (2017). doi:10.1007/
s00216-016-0087-7.
[27] M. Khuhawar and S. Lanjwani, J. Chromatogr. A. 740, 296–301 (1996). doi:10.1016/0021-
9673(96)00227-0.
[28] M.A. Mirza, M.Y. Khuhawar and R. Arain, Electrophoresis 29, 597–603 (2008). doi:10.1002/elps.
200700414.
[29] M.F. Lanjwani, N. Altunay and M. Tuzen, Food Chem. 400, 134085 (2023). doi:10.1016/j.
foodchem.2022.134085.
[30] T.M. Jahangir, M.Y. Khuhawar, S.M. Leghari, M.T. Mahar and K.P. Mahar, Arab. J. Geosci. 8,
3259–3283 (2015). doi:10.1007/s12517-014-1395-x.
[31] M.F. Lanjwani, M.Y. Khuhawar, T.M.J. Khuhawar, A.H. Lanjwani, S.Q. Memon, W. Soomro and I.
K. Rind, J. Cluster Sci. 1–12 (2022). doi:10.1080/03067319.2021.1884241.
[32] N. Altunay, M. Tuzen, M.F. Lanjwani and M.R.A. Mogaddam, J. Food Compos. Anal. 114,
104791 (2022). doi:10.1016/j.jfca.2022.104791.
[33] N. Altunay, A. Elik, M. Tuzen, M.F. Lanjwani and M.R.A. Mogaddam, J. Food Compos. Anal. 115,
105023 (2023). doi:10.1016/j.jfca.2022.105023.
[34] S.H. Abd El-Aziz, J. Saudi Soc. Agric. Sci. 20, 359–370 (2021). doi:10.1016/j.jssas.2021.04.003.
[35] I. Belbachir, J.A. Lopez-Lopez, B. Herce-Sesa and C. Moreno, J. Hazard. Mater. 430, 128505
(2022). doi:10.1016/j.jhazmat.2022.128505.
[36] A. Shahvar, D. Shamsaei, M. Saraji, N. Arab and S. Alijani, Microchem. J. 160, 105655 (2021).
doi:10.1016/j.microc.2020.105655.
[37] M.S.F. Ercan, M.F. Ayyıldız, D.S. Chormey and S. Bakırdere, Environ. Monit. Assess. 193, 1–7
(2021). doi:10.1007/s10661-020-08729-w.
[38] N. Bakaraki Turan, B.T. Zaman, B. Arvas, Ç. Yolaçan and S. Bakirdere, Chem. Pap. 75, 2929–2935
(2021). doi:10.1007/s11696-021-01538-6.
[39] M. Al-Nidawi and U. Alshana, J. Food Compos. Anal. 104, 104140 (2021). doi:10.1016/j.jfca.
2021.104140.