Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
Research Article, Issue 1
Analytical Methods in Environmental Chemistry Journal
Journal home page: www.amecj.com/ir
AMECJ
Application of experimental design methodology to
optimize acetaminophen removal from aqueous environment
by magnetic chitosan@multi-walled carbon nanotube
composite: Isotherm, kinetic, and regeneration studies
Ebrahim Nabatian a, b, Maryam Dolatabadi c, Saeid Ahmadzadeh d, e*
a Student Research Committee, Kerman University of Medical Sciences, Kerman, Iran.
b Department of Chemistry, Faculty of Sciences, Shahid Bahonar University of Kerman, Kerman, Iran.
c Environmental Science and Technology Research Center, Department of Environmental Health Engineering, School of Public
Health, Shahid Sadoughi University of Medical Sciences, Yazd, Iran.
d Pharmaceutics Research Center, Institute of Neuropharmacology, Kerman University of Medical Sciences, Kerman, Iran.
e Pharmaceutical Sciences and Cosmetic Products Research Center, Kerman University of Medical Sciences, Kerman, Iran
ABSTRACT
Acetaminophen is a widely used drug worldwide and is frequently
detected in water and wastewater as a high-priority trace pollutant.
This study investigated the applicability of the adsorption processes
using a composite of magnetic chitosan and multi-walled carbon
nanotubes (MCS@MWCNTs) as an adsorbent in the treatment of
acetaminophen. The model was well tted to the actual data, and the
correlation coefcients of R2 were 0.9270 and 0.8885, respectively.
The maximum ACT removal efciency of 98.1% was achieved at
ACT concentration of 45 mg L-1, pH of 6.5, MCS@MWCNTs dosage
of 400 mg L-1, and the reaction time of 23 min. The result shows that
BET specic surface area of 640 m2 g-1. The adsorption isotherms
were well tted with the Langmuir Model (R2 =0.9961), depicting
the formation of monolayer adsorbate onto the surface of MCS@
MWCNTs. The maximum monolayer adsorption capacity of 256.4
mg g-1 was observed for MCS@MWCNTs. The pseudo-second-
order kinetic model well depicted the kinetics of ACT adsorption on
MCS@MWCNTs (R2=0.9972). Desorption studies showed that the
desorption process is favored at high pH under Alkaline conditions.
The results demonstrate that the MCS@MWCNTs is an efcient,
durable, and sustainable adsorbent in water purication treatment.
Keywords:
Adsorption,
Acetaminophen,
Experimental design,
Isotherm, Kinetic,
Regeneration
ARTICLE INFO:
Received 23 Dec 2021
Revised form 20 Feb 2022
Accepted 2 Mar 2022
Available online 29 Mar 2022
*Corresponding Author: Saeid Ahmadzadeh
Email: chem_ahmadzadeh@yahoo.com
https://doi.org/10.24200/amecj.v5.i01.168
------------------------
1. Introduction
Pharmaceutical pollutants (PPs) are a group of
emerging anthropogenic hazard contaminants that
contain different groups of human and veterinary
medicinal compounds that are used widely all over
the globe [1, 2]. Acetaminophen (ACT) is one of
the most frequently used drugs worldwide. ACT is
a type of analgesic and antipyretic drug commonly
used as a fever reducer and pain reliever. Because
of high consumption worldwide, it is one of the
most frequently detected drugs in bodies of water
and wastewater [3-5]. Overdoses of ACT produce
the accumulation of toxic metabolites, which may
cause severe and sometimes fatal hepatoxicity,
and nephrotoxicity; generally, due to their bio-
accumulation, they pose a potential long-term risk
62
for aquatic and terrestrial organisms. To improve
the water quality and protect human health, the
water contaminated with emerging contaminants,
including pharmaceuticals, must be efciently
treated using an appropriate technique before being
supplied for consumption [6, 7]. Various physical,
chemical, and biological technologies can be
employed to treat ATC from water and wastewater.
Among the different treatments, adsorption
technology is attractive due to its effectiveness,
efciency, and economy. The common adsorbents
primarily include activated carbons, zeolites, clays,
industrial by-products, agricultural wastes, biomass,
and polymeric materials. AC is characterized
by high porosity and an extensive surface area,
enabling it to adsorb many kinds of pollutants
efciently. Despite its high adsorption capacity, the
use of AC on a large scale is limited by process
engineering difculties such as the dispersion of
the AC powder and the cost of its regeneration.
However, these adsorbents described above suffer
from low adsorption capacities and separation
inconvenience. Therefore, efforts are still needed to
investigate new promising adsorbents [8]. Chitosan
(CS) has gained considerable attention as a non-
conventional adsorbent in water decontamination
research due to its favorable properties such as
non-toxicity, eco-friendliness, high availability,
biodegradability, good biocompatibility, low cost,
and good adsorption properties. However, the high
solubility of CS at lower pH (i.e., below 4) and
poor mechanic properties are the limiting problems
for taking advantage of the interaction ability of
CS with Pollutant molecules. Thus, it might be
not advisable to use untreated CS as an adsorbent
in aqueous media [9-12]. One good strategy is to
immobilize CS on solid matrixes that can stabilize
CS in acid solutions and improve its mechanical
strength to overcome these problems. Different
kinds of solid organic or inorganic matrixes have
been used to form composites with CS, such as
glass plates, activated clay, silica, and polymer
spheres.
Recently, carbon nanotubes (CNTs) have also been
used as a matrix to prepare CS/CNTs composites.
CNTs, a fascinating new member of the carbon
family, have attracted strong research interest
since their discovery because of their unique
morphologies and various potential applications,
as well as their remarkable mechanical properties
[13]. CNTs have been proven to possess excellent
adsorption capacity in removing organic and
inorganic pollutants because of their hollow and
layered nanosized structures with a large specic
surface area. Also, CNTs can provide improved
mechanical strength and better structural integrity
conditions. However, the difculty in collecting
these adsorbents from treated efuents can cause
inconveniences in their practical application.
As an efcient, fast, and economical method for
separating magnetic materials, Magnetic separation
technology has received considerable attention
in recent years. Imparting magnetic properties to
adsorbents can isolate them from the medium using
an external magnetic eld without the need for
complicated centrifugation or ltration steps [11,
14].
The current work aimed to investigate the
efcacy of magnetic CS and multi-walled carbon
nanotubes (MCS@MWCNTs) as the adsorbent in
ACT removal under various operating conditions.
Effective parameters on ACT removal such as
solution pH, reaction time, ACT concentration, and
adsorbent dosage were optimized with response
surface method (RSM) using central composite
design (CCD). In addition, some extra experiments
were performed to study adsorption kinetics and
isotherms, and adsorbent reusability.
2. Experimental
2.1. Chemicals
Chitosan (Merck), Multi-walled carbon nanotubes
(Sigma-Aldrich), Acetaminophen (C8H9NO2,
Merck), Ferric chloride (FeCl3.6H2O, Merck),
Sodium acetate (C2H3NaO2 Merck), Ethylene
glycol (C2H6O2, Merck), Acetone (C3H6O, Merck),
Parafn (CnH2n+2, Merck), Sodium hydroxide
(NaOH, Merck), Hydrochloric acid (HCl, Merck),
were of analytical grade. All solutions used in the
experiments were prepared with distilled water.
Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
63
2.2. Preparation of Fe3O4 nanoparticles
Typically, 1.35 g of FeCl3.6H2O and 3.6 g of
sodium acetate were dissolved in 40 mL ethylene
glycol with stirring and heating simultaneously.
The temperature has risen to 80-100 °C. After
stirring for 30 min, the yellow-brown color
solution was transferred to a Teon-lined stainless-
steel autoclave and heated in the oven at 180°C
for 12 h. Then, the autoclave was allowed to cool
down to room temperature naturally. The black
magnetite particles were washed with acetone and
water several times and dried in the oven at 60°C
overnight [15].
2.3. Preparation of MCS@MWCNTs
The composites of MCS@MWCNTs were
synthesized by a suspension cross-linking approach
with some modication. Typically, 0.1 g of
chitosan was dissolved in 20 mL of 2% (v/v) acetic
acid solution under ultrasonication. Subsequently,
0.2 g of Fe3O4 and 0.2 g of MWCNTs were added
into the chitosan solution, and the reaction system
was further ultrasonicated for 20 min. Then, the
above mixture was slowly dispersed in 40 mL of
parafn containing 2 mL of span-80 under stirring.
After 30 min of emulsication, 1 mL of 25% (v/v)
glutaraldehyde was introduced into the system for
the cross-linking of chitosan. Then the mixture was
stirred continuously for 1 h at 70 °C. Afterward,
the pH value of the reaction solution was adjusted
to 9–10 with 1 mol L-1 NaOH and the reaction
system was allowed to stir for another 1 h at 80
°C. The particles were washed with petroleum
ether, ethanol, and ultrapure water three times,
respectively. Finally, MCS@MWCNTs were
obtained by magnetic separation and freeze-dried
for 12 h [16].
2.4. Characterization of MCS@MWCNTs
The standard BET equation was employed to
calculate the Brunauer-Emmert-Teller (BET)
surface area from the desorption isotherms. The pore
size distribution was determined from desorption
isotherms using the Barrett, Joyner, and Halenda
(BJH) method. All calculations were performed
automatically by an Accelerated Surface Area and
Porosimeter system (ASAP 2010, Micromeritics,
U.S.A.). The pH of zero point charge (pHzpc)
describes the condition when the charge density
on the surface is zero. It is usually determined
concerning the pH of the mixtures. Researchers
proposed a mass titration method to determine the
values of pHzpc: portions of 20 mL NaCl (0.01 M)
solution were added into different asks. The initial
pH was adjusted with NaOH or HCl to the desired
values between 2 and 12 (metrohm 827 pH/mV lab
pH meter). Then, 20 mg of the MCS@MWCNTs
sample was suspended into each ask. The asks
with caps were placed in a shaker. After shaking
for 24 h, the pH of the solutions was measured and
designated as pHnal. The pHzpc value is the point
where the pHinitial = pHnal [17].
2.5. Experimental design and statistical analysis
The response surface methodology (RSM) is a
set of statistical and mathematical methods used
to analyze experimental results. RSM is used in
conditions that many input variables affect the
performance and response characteristics of the
process. A complete description of a process
requires that it be modeled as a polynomial function
generally of degree 2 or higher. Since operational
conditions may be associated with changes, the
nonlinear second-order model can describe it. The
quadratic regression model was considered in the
form of Equation 1[18-20]:
(Eq. 1)
where Y represents the response of process, i and j
index numbers for factors, Xi and Xj are the design
variables, βi and βj represents the coefcient of rst-
order effect, βij is the coefcient of interaction,β0 is
constant-coefcient, k is the number of factors
and ε is the model error. In this study, CCD is
based on a four-factor design, including ACT
concentration (X1), pH (X2), adsorbent dosage
(X3), and reaction time (X4) were discussed. In the
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
64
design of the experiments, each variable with ve
levels was considered, in accordance with Table
1. The statistical signicance of CCD modeling
parameters and their combined interactions at
certain levels were examined based on their p
values. Analysis of variance (ANOVA) was used
to check the experimental data and accuracy of
the response surface model. The coefcient of
variation (C.V. %) and R2 values was calculated
to evaluate the goodness of t of the regression
model. The model precision associated with the
range of predicted values at the given points was
also elucidated.
2.6. Adsorption experiments
All adsorption experiments were carried out in
100 mL of pyrex reactor by mixing a given dose of
MCS@MWCNTs with a certain concentration of
ATC solution in a thermostatic shaker. The initial
pH of the ACT solution was adjusted to a certain
value using NaOH and HCl solution by pH meter.
After adsorption, the mixture was immediately
centrifuged, and the supernatant was analyzed for
the concentration of ATC was measured using a UV/
Vis spectrophotometer (Optizen). The wavelength
corresponding to the maximum absorbance of
ATC was 242 nm. Each experiment was repeated
at least three times, and the average value was
recorded. The ACT removal (qe) and adsorption
capacity were calculated by Equations (2) and (3),
respectively [15, 21]:
(Eq. 2)
(Eq. 3)
Where C0 is the initial ACT concentration (mg L-1),
Ce is the ACT concentration (mg L-1) after the batch
adsorption procedure, m is the adsorbent dosage (g
L-1), and qe is the amount of ACT adsorbed by the
adsorbent (mg g-1).
2.7. Kinetic and isotherm models
Two widely used kinetics models, pseudo-rst-
order and pseudo-second-order models were
examined to t the experimental data. The linear
expression of pseudo-rst-order and pseudo-
second-order models are expressed as Equation 4
and 5 [15, 22].
(Eq. 4)
(Eq. 5)
where k1 (min-1) is the rate constant of the pseudo-
rst-order, k2 (g mg-1 min-1) is the second-order
rate constant. qe and qt (mg g-1) are the adsorption
capacities at equilibrium and time t (min),
respectively.
Three commonly used isotherm models, the
Langmuir and Freundlich isotherms, are selected
to analyze the equilibrium experimental data for
the adsorption of ACT onto MCS@MWCNTs. The
two models are given as Equation 6 and 7 [23].
(Eq. 6)
(Eq. 7)
where Ce (mg L-1) is the equilibrium concentration
of the ACT, qe (mg g-1) is the amount of ACT
Table 1. Coded and actual values of independent process variables used in the design
of an experimental matrix using the RSM-CCD framework.
Coded Variables (Xi) Factors
Coded Level
-α -1 0 +1 +α
X1A= ACT Concentration (mg L-1) 20.0 40.0 60.0 80.0 100.0
X2B= pH 4.00 5.50 7.00 8.50 10.00
X3C= Adsorbent dosage (mg L-1)100 200 300 400 500
X4D= Reaction time (min) 5.00 11.25 17.50 23.75 30.00
Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
65
adsorbed under equilibrium, b (L mg-1) is the
Langmuir adsorption constant, and qm (mg g-1) is
the maximum adsorption amount. KF and n are
Freundlich constants. n presents the adsorption
intensity, assessing the unfavorable adsorption
(n<1) or preferential adsorption (n > 1), and KF
((mg g-1 (L mg-1)1/n) is the adsorption capacity of
the adsorbent.
The isotherm can predict if an adsorption system
is favorable or unfavorable. Researchers pointed
out that the essential characteristics of the
Langmuir isotherm can be expressed in terms
of a dimensionless constant, RL, as presented in
Equation 8.
(Eq. 8)
where, RL is the separation factor or equilibrium
parameter, which is a direct function of the
Langmuir constant b. The values of RL indicate
the shape of the isotherm: RL>1 (unfavorable),
RL =1 (linear), 1 > RL > 0 (favorable) and RL = 0
(irreversible).
3. Results and discussions
3.1. Characterization of adsorbent
Figure 1 shows the N2 adsorption-desorption isotherms
of MCS@MWCNTs. According to Figure 1, the N2
adsorption nearly completed at a lower relative pressure
of P/P0 < 0.1, suggesting that the sample has the
micropore size distribution. In addition, the N2 hysteresis
loop at P/P0 > 0.5 was observed, indicating the presence
of mesopores structure. Accordingly, the micropore
volume fraction is higher than 79%, conrming the
majority of micropores. The MCS@MWCNTs had a
high BET-specic surface area of 640 m2 g-1.
The adsorbent’s pHzpc value was determined to
explain the adsorption behavior. This parameter
reveals the characteristics of the surface’s active
sites in a linear range of solution pH. The pHzpc
of MCS@MWCNTs was found to be around 6.8,
implying that the surface of the adsorbent would
be positively charged at solution pH below 6.8
and negatively charged at solution pH above 6.8.
Hence, one can conclude that adsorption of cationic
molecules is favored at pH>6.8, while anionic
molecules adsorption is favored at pH < 6.8.
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
Fig. 1. N2 adsorption–desorption isotherms of MCS@MWCNTs.
66
3.2. Development and analysis of regression
model equation
CCD is considered a reliable method to analyze
diagnostic plots, such as the normal probability
plot of residuals and to predict versus actual values,
to validate the adequacy of the model. The normal
probability plot of the studentized residuals is an
excellent graphical representation for the diagnosis
of data normality (Fig. 2). The data were well tted
with the line. In addition, the residuals followed a
random distribution around zero with a variation of
±3.0 (Fig. 3).
Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
Normal % Probability
-3.0 -1.5 0.0 1.5 3.0
Exte rnally Studentized R e s iduals
Desi g n-Expert® Software
removal
Col or points by value of
removal :
98
60.3096
Run Number
Internally Studentized Residuals
Residuals vs. Run
-3.00
-1.50
0.00
1.50
3.00
1 5 9 13 17 21 25 29
1 5 9 13 17 21 25 29
Run Number
3.0
1.5
0.0
-1.5
-3.0
Inte rnally Stude ntized Res iduals
Fig. 2. Normal probability plot of the internally studentized
residuals for ACT removal.
Fig. 3. Run number versus residual data for ACT removal.
67
The result indicates that the data were normally
distributed in the model response. In addition, the
corresponding relationship between the residual
and the predicted value of the equation also could
reect the reliability of the model. When the points
are distributed discretely in the Figure, it could
represent the higher reliability of the model.
ANOVA carried out the analysis of obtained
experimental data. Table 2 shows ANOVA data
for the removal efciency (Y). F-value in ANOVA
implies that the model is signicant for the
dependent variable. As shown in Table 2, F-value
is 87.47 for removal efciency, which indicates
the model is signicant. Prob>F or p-value less
than 0.05 demonstrates that these model terms
are important. The lack of t F-value of 4.13 with
the associated p-value of 0.0612 for the response
was insignicant due to relative pure error. The
validity of the model is checked by some statistical
parameters, including the determination coefcient
(R2), adjusted R2, predicted R2, adequate precision
(AP), and coefcient of variation (CV). R2 is
dened as a measure of the degree of t. As R2
approaches unity, the degree of t increases.
The similarity between R2 and adjusted R2 shows
the model’s compatibility to predict the dependent
variable. The difference between predicted R2 and
adjusted R2 must be less than 0.2. It is indicated
that predicted R2 is in acceptable agreement with
adjusted R2. AP is dened as a measure of the
ratio of the signal to noise. A ratio greater than 4
is desirable. In our current study, all ndings in
Table 2 were acceptable, which conrmed the
model’s tness with the experimental results.
Overall, the ANOVA analysis was reliable to
optimize and determine the level of each factor for
removal ACT. Therefore, it is pretty satisfactory
to predict the ACT removal efciency by the
second-order polynomial model. In the present
study, ve model terms (X1(ACT concentration),
X2(pH), X3(Adsorbent dosage), X4(Reaction time),
X2
2(quadratic effect of pH)) was most important
because of their p-values less than 0.05. These
model terms are shown in the model equation.
Based on ANOVA results, the empirical model
equation in terms of coded variables was developed
for the removal efciency as Equation 9.
(Eq. 9)
The positive sign in the equation indicates the
synergistic effect of the corresponding factor on
the response, and the negative sign represents
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
Table 2. Analysis of variance (ANOVA) for regression model.
Source Sum of Squares Degree of
freedom (df)
Mean
Squares F-Value Probability
P-value > F
Model 2775.21 5 555.04 87.47 < 0.0001
561.89 1 561.89 88.55 < 0.0001
256.30 1256.30 40.39 < 0.0001
781.29 1 781.29 123.12 < 0.0001
744.41 1 744.41 117.31 < 0.0001
431.33 1 431.33 67.97 < 0.0001
Residual 152.30 24 6.35 - -
Lack of Fit 143.18 19 7.54 4.13 0.0612
Pure Error 9.12 51.82 - -
Cor Total 2927.51 29 - - -
Fit Statistics
R20.9480 SD 2.52
Adjusted R20.9371 CV 3.16
Predicted R20.9140 AP 34.40
68
the antagonistic effect. The effect of the initial
concentration (20-100 mg L-1) was investigated.
The removal efciency for ACT is highly
dependent on the initial ACT concentration. Figure.
4 illustrated that removal efciency decreases from
92.3 to 72.6%, increasing ACT concentration
from 20 to 100 mg L-1. The effect of initial ACT
concentration depends on the immediate relation
between the concentration of the ACT and
the available sites on an adsorbent surface. In
general, the removal efciency decreases with an
increase in the initial ACT concentration due to
the saturation of adsorption sites on the adsorbent
surface. On the other hand, the increase in initial
ACT concentration will cause an increase in the
capacity of the adsorbent, which may be due to the
high driving force for mass transfer at a high initial
ACT concentration.
Figure 4 shows the effect of solution pH ranging
from 4 to 10 on the adsorption of ACT onto MCS@
MWCNTs. Figure 4 implies that the adsorption of
ACT onto MCS@MWCNTs at pH solution between
4 and 7.0 is a pH-dependent phenomenon, so the
adsorption performance was around 85% in all pH
ranges. The increase of solution pH to 7.0 caused
improvement of the adsorption. A decreasing trend
in adsorption of ACT was observed for the solution
pH over 7.0. Accordingly, the optimum solution
pH at which the maximum adsorption of the
ACT under the selected experimental conditions
was obtained was found to be 6.5. The effect of
solution pH on the adsorption of ACT onto MCS@
MWCNTs can be justied considering the pHzpc
of MCS@MWCNTs (6.8) and pKa of ACT (9.4).
According to Figure 4, the adsorption of ACT
onto MCS@MWCNTs is almost independent of
solution pH over the solution pHs between 4.0
and 7.0. At this pH range, the ACT molecules
remain mostly neutral and nonionic and thus
unfavorable for electrostatic and π-π interactions
with the functional groups on the surface of
MCS@MWCNTs. Therefore, the hydrophobic
Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
X1: ACT Concentration (mg L
-1
)
20 40 60 80 100
10.0
8.0
6.0
4.0
X2: pH
Fig. 4. Contour plot of ACT removal showing the effect of variables of ACT concentration
and pH (MCS@MWCNTs dosage of 300 mg L-1 and reaction time of 17.5 min).
69
interactions might be the primary mechanism
anticipated in the ACT adsorption under these
conditions. Considering that the pKa of ACT is
9.4, the main fraction of ACT molecules was in
anionic form at solution pHs above this value. In
contrast, the surface of MCS@MWCNTs (pHzpc
= 6.8) is negatively charged at alkaline solution
pH. Therefore, the reduction of ACT adsorption at
the pH over 7.0 can be related to the electrostatic
repulsion that occurred between anionic ACT
molecules and the negatively charged functional
groups on the surface of MCS@MWCNTs. This
effect tends to enhance the adsorption of ACT
onto MCS@MWCNTs. The greater the solution
pH over 7.0, the more negative the surface of
MCS@MWCNTs and the more ionized the ACT
molecules. Therefore, the greater the repulsive
force leads to reduced adsorption. The MCS@
MWCNTs dosage as adsorbent was one of the
dominant parameters controlling the adsorption of
ACT. The relation of MCS@MWCNTs dosage and
pH on removal efciency of ACT is illustrated in
Figure 5. Increasing MCS@MWCNTs dosage from
100 to 500 mg L-1 leads to an improvement in ACT
removal efciency from 71.4% to 94.2% at ACT
concentration of 60 mg L-1, pH of 7, and reaction
time 17.5 min. The increase in removal efciency
with increasing adsorbent dose is probably due to
the greater adsorbent surface area and pore volume
available at higher adsorbent dose providing more
functional groups and active adsorption sites that
result in higher removal efciency.
The dependence of the removal efciency of ACT
on reaction time is shown in Figure 6. This Figure
shows that removal efciency increases with
time, and adsorption reaches equilibrium in about
25 minutes. It indicates that the rapid increase in
removal efciency is achieved during the rst 20
min. The fast adsorption at the initial stage may
be due to the higher driving force making an
immediate transfer of adsorbate ions to the surface
of MCS@MWCNTs particles and the availability
of the uncovered surface area and the remaining
active sites on the adsorbent. According to the
results, an equilibrium time was set to 25 min for
adsorption of ACT onto MCS@MWCNTs.
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
Removal Efficiency (%)
Fig. 5. Response surface plot of ACT removal showing the effect of variables of MCS@MWCNTs dosage
and pH (ACT concentration of 60 mg L-1 and reaction time of 17.5 min).
70
Removal Efficiency (%)
3.3. Optimization and validation test
The optimization of operating conditions was
conducted to determine the optimum values
of these parameters required to achieve the
highest ACT removal efficiency. Optimization
was performed by numerical technique built
in the Design-Expert software. The desired
goal for the variables was chosen as “in
range”, while removal efficiency (response)
was chosen as “maximize”. According to the
output results, the removal efficiency of ACT
could reach a maximum value of 98.1% with
the ACT concentration of 45 mg L-1, pH of
6.5, MCS@MWCNTs dosage of 400 mg L-1,
and the reaction time of 23 min. An additional
experiment was conducted to validate the model
prediction (the optimal conditions) in this study.
It was found that the results were greatly agreed
with the predicted value through the quadratic
model (Table 3). The experimental data were
close to predicted data, indicating the accurate
prediction ability of the model.
3.4. Adsorption kinetics and isotherms studies
The calculated kinetic parameters for pseudo-rst-
order and pseudo-second-order models are listed
in Table 4. The pseudo-second-order model’s
correlative coefcient (R2) was better than that
of the pseudo-rst-order model. These results
indicated that the kinetic data tted well with
pseudo-second-order models. The pseudo-second-
order model assumes that the rate-limiting step
might be chemical adsorption in the adsorption
process.
The adsorption isotherm is critically important
in designing an adsorption system. The sorption
data were tted by the Langmuir and Freundlich
equations, and the calculated parameters are
Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
Fig. 6. Response surface plot of ACT removal showing the variables effect of reaction time
and ACT concentration (MCS@MWCNTs dosage of 300 mg L-1, and pH of 7).
71
summarized in Table 5. It was found that the
Langmuir model provided a better t to the observed
data for the ACT, with high correlation coefcients
(R2=0.996). The maximum ACT sorption capacity
(qm) was 256.4 mg g-1. The values of n and RL were
obtained at 3.52 and 0.026, respectively, suggesting
that the adsorbent was favorable for removing ACT
from the aqueous solution.
3.5. Regeneration studies
Desorption studies are necessary to complete the
investigation of the mechanism involved in the
adsorption of an adsorbate by an adsorbent and to
regenerate the adsorbent for economic success. In
the present study, desorption was explored, varying
the pH from 4.0 to 10.0 and keeping the adsorbent
dosage constant at 300 mg L-1. An increase in pH
favored ACT desorption from MCS@MWCNTs
because of electrostatic repulsion between
negatively charged sites on the adsorbent surface
and ACT molecules. The feasibility of using MCS@
MWCNTs in successive adsorption-desorption
cycles was examined by contacting 45 mg L-1 ACT
solution with 400 mg L-1 recycled adsorbent at pH
10.0. Under these conditions, ACT removal by
MCS@MWCNTs and recycled MCS@MWCNTs
was 98.1% and 86.7%, respectively. Such a marked
loss of sorption capacity suggests that the reuse
of desorbed MCS@MWCNTs would need some
regeneration before recycling.
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
Table 3. Optimization and validation tests for ACT removal efciency.
NO ACT concentration
(mg L-1)pH MCS@MWCNTs
dosage (mg L-1)
Reaction
time (min)
Experimental removal
efciency (%)
Predicted removal
efciency (%)
I 45 6.5 400 23 98.7 98.1
II 60 7.0 300 18 84.5 85.3
III 40 8.5 400 12 79.4 80.0
IV 80 8.5 200 25 71.0 69.4
Table 4. Parameters of kinetic equations for the adsorption of ACT.
Pseudo-First-Order model Pseudo-Second-Order model
qe(mg g-1) k1(min-1)R2qe(mg g-1) k2 (g mg-1 min-1)R2
121.47 0.0689 0.9452 217.4 0.0008 0.9972
Table 5. Langmuir and Freundlich constants for the adsorption of ACT.
Langmuir Freundlich
qm (mg g-1) b (L mg-1)RLR2Kf (L g-1) n R2
256.4 0.658 0.026 0.9961 103.1 3.52 0.9542
72 Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
4. Conclusions
The process was optimized using central
composite design (CCD), a statistical tool used to
optimize response surface methodology (RSM).
A second-order polynomial model adequately
t the experimental data with an adjusted R2 of
0.9270, showing that the model could efciently
predict the ACT removal. It was found that all
selected variables signicantly affect ACT removal
efciency. Under these conditions, the maximum
adsorption capacity for MCS@MWCNTs was
found to be 256.4 mg g-1. The results showed that
the Langmuir and the pseudo-second-order kinetic
models presented better ttings for the adsorption
equilibrium and kinetics data. This study showed
that MCS@MWCNTs are a useful adsorbent for
the removal of ACT from aqueous solutions.
5. Acknowledgements
The authors would like to express their appreciation
to the student research committee of Kerman
University of Medical Sciences [Grant number
400000753] for supporting the current work.
Funding: This work received a grant from the
Kerman University of Medical Sciences [Grant
number 400000753].
Conict of interest: The authors declare that they
have no conict of interest regarding the publication
of the current paper.
Ethical approval: The Ethics Committee of
Kerman University of Medical Sciences approved
the study (IR.KMU.REC.1400.503).
6. References
[1] B. Zaied, M. Rashid, M. Nasrullah,
A. Zularisam, D. Pant, L. Singh, A
comprehensive review on contaminants
removal from pharmaceutical wastewater
by electrocoagulation process, Sc. Total
Environ., 726 (2020) 138095. https://doi.
org/10.1016/j.scitotenv.2020.138095.
[2] A. Kumar, M. Chandel, A. Sharma, M.
Thakur, A. Kumar, D. Pathania, L. Singh,
Robust visible light active PANI/LaFeO3/
CoFe2O4 ternary heterojunction for the
photo-degradation and mineralization of
pharmaceutical efuent: Clozapine, J.
Environ. Chem. Eng., 9 (2021) 106159.
https://doi.org/10.1016/j.jece.2021.106159.
[3] C.-C. Su, C.A. Cada Jr, M.L.P. Dalida, M.-
C. Lu, Effect of UV light on acetaminophen
degradation in the electro-Fenton process,
Sep. Purif. Technol., 120 (2013) 43-51.
https://doi.org/10.1016/j.seppur.2013.09.034.
[4] E. Bailey, H. Worthington, P. Coulthard,
Ibuprofen and/or paracetamol
(acetaminophen) for pain relief after surgical
removal of lower wisdom teeth, a Cochrane
systematic review, British Dent. j., 216 (2014)
451. https://doi.org/10.1038/sj.bdj.2014.330.
[5] M. Qutob, M. Rafatullah, M. Qamar, H.S.
Alor, A.N. Al-Romaizan, M.A. Hussein,
A review on heterogeneous oxidation
of acetaminophen based on micro and
nanoparticles catalyzed by different activators,
Nanotechnol. Rev., 11 (2022) 497-525.
https://doi.org/10.1515/ntrev-2022-0030.
[6] S. Wang, J. Wu, X. Lu, W. Xu, Q. Gong, J. Ding,
B. Dan, P. Xie, Removal of acetaminophen
in the Fe2+/persulfate system: Kinetic model
and degradation pathways, Chem. Eng. J., 358
(2019) 1091-1100. https://doi.org/10.1016/j.
cej.2018.09.145.
[7] H.N.P. Vo, G.K. Le, T.M.H. Nguyen, X.-T.
Bui, K.H. Nguyen, E.R. Rene, T.D.H. Vo,
N.-D.T. Cao, R. Mohan, Acetaminophen
micropollutant: Historical and current
occurrences, toxicity, removal strategies
and transformation pathways in different
environments, Chemosphere, 236 (2019)
124391. https://doi.org/10.1016/j.
chemosphere.2019.124391.
[8] Z.B. Khalid, M. Nasrullah, A. Nayeem, Z.A.
Wahid, L. Singh, S. Krishnan, Application
of 2D graphene-based nanomaterials for
pollutant removal from advanced water
and wastewater treatment processes, in:
Adapting 2D Nanomaterials for Advanced
Applications, ACS Publications, ACS
Symposium Series, 1353 (2020) 191-217.
73
Acetaminophen removal from aqueous environment by MCS@MWCNTs Ebrahim Nabatian et al
https://doi.org/10.1021/bk-2020-1353.ch009.
[9] I. Mustafa, Methylene blue removal from
water using H2SO4 crosslinked magnetic
chitosan nanocomposite beads, Microchem.
J., 144 (2019) 397-402. https://doi.
org/10.1016/j.microc.2018.09.032.
[10] H. Zeng, L. Wang, D. Zhang, P. Yan, J. Nie,
V.K. Sharma, C. Wang, Highly efcient
and selective removal of mercury ions
using hyperbranched polyethylenimine
functionalized carboxymethyl chitosan
composite adsorbent, Chem. Eng. J., 358
(2019) 253-263. https://doi.org/10.1016/j.
cej.2018.10.001.
[11] S. Wang, Y.-Y. Zhai, Q. Gao, W.-J. Luo, H.
Xia, C.-G. Zhou, Highly efcient removal
of acid red 18 from aqueous solution by
magnetically retrievable chitosan/carbon
nanotube: batch study, isotherms, kinetics,
and thermodynamics, J. Chem. Eng. Data,
59 (2014) 39-51. https://doi.org/10.1021/
JE400700C.
[12] M. Abbas, M. Arshad, M. Raque, A.
Altalhi, D. Saleh, M. Ayub, S. Sharif, M.
Riaz, S. Alshawwa, N. Masood, Chitosan-
polyvinyl alcohol membranes with improved
antibacterial properties contained Calotropis
procera extract as a robust wound healing
agent, Arab. J. Chem., (2022) 103766. https://
doi.org/10.1016/j.arabjc.2022.103766.
[13] M. Jawaid, A. Ahmad, N. Ismail, M.
Rafatullah, Environmental remediation
through carbon based nano composites,
Springer, 2021. https://www.springer.com/
series/8059
[14] K. Li, Q. Gao, G. Yadavalli, X. Shen, H. Lei, B.
Han, K. Xia, C. Zhou, Selective adsorption of
Gd3+ on a magnetically retrievable imprinted
chitosan/carbon nanotube composite with
high capacity, ACS Appl. Mater. Interfaces, 7
(2015) 21047-21055. https://doi.org/10.1021/
acsami.5b07560.
[15] F. Jamali‐Behnam, A.A. Najafpoor, M.
Davoudi, T. Rohani‐Bastami, H. Alidadi, H.
Esmaily, M. Dolatabadi, Adsorptive removal
of arsenic from aqueous solutions using
magnetite nanoparticles and silica‐coated
magnetite nanoparticles, Environ. Prog.
Sustain. Energy, 37 (2018) 951-960. https://
doi.org/10.1002/ep.12751.
[16] K. Xu, Y. Wang, H. Zhang, Q. Yang, X. Wei,
P. Xu, Y. Zhou, Solid-phase extraction of
DNA by using a composite prepared from
multiwalled carbon nanotubes, chitosan,
Fe3O4 and a poly (ethylene glycol)-based
deep eutectic solvent, Microchim. Acta, 184
(2017) 4133-4140. https://doi.org/10.1007/
s00604-017-2444-4.
[17] H. Alidadi, M. Dolatabadi, M. Davoudi,
F. Barjasteh-Askari, F. Jamali-Behnam,
A. Hosseinzadeh, Enhanced removal
of tetracycline using modied sawdust:
optimization, isotherm, kinetics, and
regeneration studies, Process Saf. Environ.
Prot., 117 (2018) 51-60. https://doi.
org/10.1016/j.psep.2018.04.007.
[18] A. Fegousse, A. El Gaidoumi, Y. Miyah,
R. El Mountassir, A. Lahrichi, Pineapple
bark performance in dyes adsorption:
optimization by the central composite design,
J. Chem., 2019 (2019) 3017163. https://doi.
org/10.1155/2019/3017163.
[19] R.A. Khera, M. Iqbal, A. Ahmad, S.M.
Hassan, A. Nazir, A. Kausar, H.S. Kusuma,
J. Niasr, N. Masood, U. Younas, Kinetics
and equilibrium studies of copper, zinc,
and nickel ions adsorptive removal on to
Archontophoenix alexandrae: conditions
optimization by RSM, Desalin. Water Treat.,
201 (2020) 289-300. https://doi.org/10.5004/
dwt.2020.25937.
[20] V.U. Siddiqui, A. Ansari, M.T. Ansari, M.
Akram, W.A. Siddiqi, A.M. Alosaimi, M.A.
Hussein, M. Rafatullah, Optimization of facile
synthesized ZnO/CuO nanophotocatalyst
for organic dye degradation by visible
light irradiation using response surface
methodology, Catalysts, 11 (2021) 1509.
https://doi.org/10.3390/catal11121509.
[21] S. Noreen, S. Ismail, S.M. Ibrahim, H.S.
74 Anal. Methods Environ. Chem. J. 5 (1) (2022) 61-74
Kusuma, A. Nazir, M. Yaseen, M.I. Khan, M.
Iqbal, ZnO, CuO and Fe2O3 green synthesis
for the adsorptive removal of direct golden
yellow dye adsorption: kinetics, equilibrium
and thermodynamics studies, Z. fur Phys.
Chem., 235 (2021) 1055-1075. https://doi.
org/10.1515/zpch-2019-1599.
[22] A. Kausar, K. Naeem, M. Iqbal, H.N. Bhatti,
A. Ashraf, A. Nazir, H.S. Kusuma, M.I. Khan,
Kinetics, equilibrium and thermodynamics
of dyes adsorption onto modied chitosan: a
review, Z. fur Phys. Chem., (2021). https://
doi.org/10.1515/zpc-2019-1586.
[23] P. Mishra, Z.A. Wahid, R.M. Zaid, S. Rana,
S. Tabassum, A. Karim, L. Singh, M.A.
Islam, X. Jaing, M. Sakinah, Kinetics and
statistical optimization study of bio-hydrogen
production using the immobilized photo-
bacterium, Biomass Convers. Bioren.,
(2020) 1-12. https://doi.org/10.1007/s13399-
020-00835-6.
