Journal of Molecular Structure 1232 (2021) 130071

Contents lists available at ScienceDirect

Journal of Molecular Structure

journal homepage: www.elsevier.com/locate/molstr

Synthesis and characterization of magnetic ZnCl2 -activated carbon

produced from coconut shell for the adsorption of methylene blue
Hatice Karaer Yağmur a,b,∗, İsmet Kaya a
a
Polymer Synthesis and Analysis Laboratory, Department of Chemistry, Çanakkale Onsekiz Mart University, Çanakkale 17020, Turkey
b
Faculty of Science, Department of Chemistry, Dicle University, Diyarbakır 21400, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: In this research, a magnetic-activated carbon composite was synthesized to investigate its adsorption to
Received 8 January 2021 methylene blue dye. Active carbon (ACz) was produced by ZnCl2 activation of coconut shell (char: ZnCl2 ,
Revised 25 January 2021
2:1). Magnetic activated carbon (MACz) was prepared via an easy method by chemical co-precipitation of
Accepted 1 February 2021
Fe3+ /Fe2+ . The pore diameter, pore-volume, and surface area of MACz were determined from N2 adsorp-
Available online 5 February 2021
tion isotherms. The ACz and MACz were characterized by several techniques such as BET, SEM/EDX, FT-IR,
Keywords: XRF, VSM, TGA, and pHpzc analysis. The average particle size, particle size distribution, and zeta potential
Coconut shell of MACz were determined by Zetasizer. The zero point charge of the MACz was determined as approxi-
Methylene blue mately 4.9 pH. Moreover, the adsorption by MACz was investigated as kinetically and thermodynamically
Adsorption for removal of methylene blue (MB) from aqueous solution. The kinetics and thermodynamic parameters
RMSE were also calculated. According to the results, the adsorption process was determined as endothermic.
SSE
The Langmuir and Freundlich isotherms models were used and it was found that adsorption of MB onto
MACz best fitted to Freundlich model according to RMSE, SSE, and R2 values. The maximum adsorption
capacity of the composite was calculated as 156.25 mg/g. Thermodynamic parameters (Ho , Go , and
So ) were also calculated, which revealed that the adsorption process is spontaneous.
© 2021 Elsevier B.V. All rights reserved.

1. Introduction expensive and readily available. Activated carbon is the most pop-
ular adsorbent but it is expensive. So, many researchers have stud-
Water pollution is an extremely important problem and has in- ied the feasibility of using low-cost substances such as plum ker-
creased in the past decade. Industrial effluents from the metal, nels, chitin, chitosan, perlite, and natural clay as an adsorbent for
plastic, food, and pharmaceutical industries are important causes the removal of dyestuffs from wastewater [2,3].
of water pollution because most industrial wastewaters contain One of the renewable resources on the earth is biomaterials
several hazardous pollutants, such as dyes, which are extremely because they are inexpensive and very available. Since they con-
harmful to living things. Approximately 15% of dye wastes after sist of lignin, cellulose and low inorganic impurities, high inter-
dying are being released into the environment [1]. In the recent nal porosity, surface area, and are the especially promising natural
past, the introduction of a large number of synthetic dyes that matter for obtaining active carbons (ACs) [4,5]. ACs are obtained by
are toxic, mutagenic, and carcinogenic made the situation worst by carbonizing and activating agents. These biomass supplies include
disturbing normal aquatic life [2]. Dyes are stable chemicals with rice husk, coconut shells, fungi, walnut, bamboo, animal bones,
highly resistant photolysis properties [1]. To overcome dye pollu- firewood, tea leaves, coffee beans, banana peel, animal feathers
tion in the water, various methods such as flocculation, coagula- and sugarcane bagasse [6]. ACs are obtained mainly by chemical
tion, precipitation, membrane filtration, fungal decolorization, pho- and physical activation. The process of chemical activation may be
tocatalytic decolorization, and adsorption have been tested. Among achieved in one step by thermal decomposition of biomass with
them, one of the simplest, cheap, and effective physical processes activating agents such as potassium carbonate (K2 CO3 ), phospho-
is adsorption that has also simple design requirements for the ric acid (H3 PO4 ), sodium carbonate (Na2 CO3 ), potassium hydroxide
treatment of wastewater. Also, the adsorption technique provides (KOH), zinc chloride (ZnCl2 ) [7]. In the process of physical activa-
an attractive alternative treatment, especially if the adsorbent is in- tion, the sample is treated in an inert atmosphere and the activa-
tion of the solid material (char) is carried out at 80 0–10 0 0 °C, in a
∗
slightly reactive atmosphere using CO2 or steam [8]. ACs are used
Corresponding author.
E-mail address: profhatice23@hotmail.com (H.K. Yağmur).
as adsorbent materials for the water purification and removal of

https://doi.org/10.1016/j.molstruc.2021.130071
0022-2860/© 2021 Elsevier B.V. All rights reserved.
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

dyes, heavy metal treatment, and contaminated sediments reme- 2.3. Preparation of magnetic activated carbon (MACz)
diation [9]. Dyes can be seen even at very low concentrations in
aqueous media and have toxic effects such as dizziness, cyanosis, Production of MACz was followed according to a general pro-
mental health in the humans body [10]. Methylene blue (MB) is a cedure similar to the literature [9]. In general, 10 g ACz mixed
cationic dyestuff and usually used for dyeing silk, cotton, textiles with 7.32 g FeSO4 •7H2 O and 13.32 g FeCl3 •6H2 O were added in
and may cause various diseases in the human and animal body 400 mL of pure water in a beaker. The mixture was heated to 60
such as mental health, allergic effects, diarrhea, and nausea, tachy- °C with continuous stirring to make sure that Fe+2 and Fe+3 could
cardia, dyspnea irritation to the skin owing to its toxical and aro- spread into AC and cooled to 40 °C. Then NaOH (5 M) solution
matic structure [11]. Magnetic particles can use an alternative ad- was added dropwise to the mixture until pH reached in the range
sorbent for the removal of dye and heavy metal [12] and are usu- of 10–11. The suspension was stirred for 4 h and then left to settle
ally preferred owing to their magnetic property, adsorption capaci- overnight. Afterward, the solid particles were isolated using a mag-
ties, low toxicity, and biocompatibility [12]. There are many studies net and washed with pure water to be neutralized, then the impu-
on activated carbon/magnetic activated carbon and methylene blue rities were removed with acetone and ethanol. Lastly, the magnetic
adsorption in the literature. In the literature, magnetic-activated active carbon (MACz) was dried at 60 °C for 48 h (Scheme 1).
carbons were generally synthesized from almond, walnut shells,
charcoal, and chitosan [13]. However, the researches are limited on 2.4. Adsorption experiments
preparing magnetic activated carbon from coconut shell as biochar.
Therefore, magnetic activated carbon was prepared from coconut The MB was used as an adsorbate. The adsorption experiments
shells, unlike the literature. Moreover, magnetically activated car- can be classified into three sections: (a) temperature effect (b) fit-
bons have several advantages that can be added to the polymeric ting of data to adsorption isotherms (c) thermodynamic and kinet-
materials to increase the sorbent features. Also, the magnetic ad- ics studies. 10 0 0 mg L−1 MB solution was prepared as a stock so-
sorbents may be removed easily from the solution medium by a lution for experiments. The MB solutions were prepared between
magnet after adsorption experiments [14]. Therefore, a new adsor- 25- 500 mg L−1 by diluting the stock solution. To evaluate the
bent was synthesized as a magnetic active carbon from coconut effect of temperature, all kinetic experiments were performed at
shells. contact time (0–360 min), initial dye concentration of 100 mg L 1 ,
In this study, a new magnetic activated carbon(MACz) was pre- and sorbent dose of 0.5 g in a 100 mL conical flask at 298, 308,
pared from a coconut shell. The composite was characterized by and 318 K, at 120 rpm. Adsorption equilibrium studies were per-
BET, SEM/EDX, FT-IR, XRF, VSM, TGA, and pHpzc analysis. Further- formed by 0.15 g of MACz (as a sorbent) with 50 mL of dye so-
more, the zero point charge of MACz was determined. MACz was lution (25–500 mg L−1 ) in 100 mL conical flasks at 298, 308, and
used as an adsorbent for the removal of methylene blue (MB) from 318 K (S.R:120 rpm for 6 h). qe (mg g−1 ) (adsorption capacity at
an aqueous solution. The adsorption of MB onto the MACz from equilibrium) may be calculated from the following equations:
aqueous solution was studied both thermodynamically and kineti-
(Co − Ce )V
cally. qe = (1)
m
2. Materials and methods where V (L), m(g), Ce and Co (mgL−1 ) are the solution volume, the
dry adsorbent mass, the equilibrium and initial concentrations of
2.1. Materials MB, respectively.

Coconut shells (CS) were bought from a local market (Di- 2.5. Characterization
yarbakır, Turkey). After drying at room temperature coconut shells
were crushed using an electric grinder. Methylene blue (MB) The raw coconut shell and active carbons (ACs) were character-
(C16 H18 ClN3 S; λmax : 663 nm and MW : 319.85 g mol−1 ) dye, ized by using FT-IR (Fourier transform infrared spectroscopy, ATR
FeSO4 •7H2 O, NaOH, and FeCl3 • 6H2 O were supplied from Sigma- sampling accessory between 40 0 0 and 450 cm−1 ), BET (Brunauer–
Aldrich. The chemical structure of MB was given in Scheme 1. Emmett–Teller), SEM (scanning electron microscopy), and TGA
(thermogravimetric analysis). TGA analysis was carried out be-
2.2. Preparation of the activated carbon (ACz) tween 20 and 10 0 0 °C (Perkin Elmer, in N2 , rate 10 °C min−1 ).
The surface properties of activated carbons were identified via a
The chemical activation of active carbon samples from coconut SEM-EDX instrument (Zeiss Evo 40, Ametek EDX). Specific surface
shells(CS), using ZnCl2 as an activating agent, was followed accord- areas of ACs were determined by the BET surface area (SBET). The
ing to a reported procedure [7,15,16]. Coconut shells (CSs) were mesoporous volume (Vmeso ) was calculated by subtracting Vmicro
divided into small pieces before their chemical activation. The CSs from Vtotal (Vmeso =Vtotal -Vmicro ). The particle size and zeta potential
were washed with distilled water and dried at 120 °C for 72 h to measurement of the MACz were performed by using a ZetaSizer
decrease the moisture content and were impregnated with ZnCl2 Nano-ZS (Malvern Nano ZS90). The presence of major elements in
solution. The impregnation ratio (the ZnCl2 :chars ratio (w/w)) was magnetic activated carbon and activated carbon were determined
1:2. The activating agent and the coconut shells were homoge- by using an X-ray fluorescence Spectrometer (model Panalytical
neously stirred at room temperature for 24 h (900 rpm). The sam- Zetium).
ple was dried at 120 °C for 48 h to obtain the impregnated sample
which was set in a horizontal stainless-steel tubular reactor with 2.6. Determination of the point of zero charge (pHzpc)
7 cm diameter and 100 cm length. The impregnated sample was
carbonized at 500 °C for 1 h under nitrogen (N2 ) flow of 100 mL The point of zero charge was determined as reported before
min−1 with a heating rate of 10 °C min−1 . The carbonized sam- [17,18]. The procedure can be described as follows: to a series of
ple was boiled with HCl (0.1 M) solution under reflux to eliminate 50 mL conical flasks, 30 mL of 0.1 M NaCl solution was added.
impurities and decrease the amount of ash content of the active The initial solution pH values of the NaCl solutions (initial pH)
carbon. Then, the active carbon was washed with distilled water were adjusted in the range of 2–12 of using 0.1 M HCl/ NaOH
until chloride ions were not detected. Finally, the active carbon was solutions. After a constant value of initial pH had been reached,
dried at 120 °C for 48 h. 0.1 g adsorbent (MACz) was added into each conical flask and

2
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Scheme 1. Schematic representation of the synthesis of the MACz adsorbent.

capped immediately. These samples were shaken for 48 h, after O for a magnetite-active carbon may be observed in the 500–750
which pH was measured for each sample (pHf ). Variation in the cm−1 range. The FT-IR analysis confirmed the generation of Fe-O
pH (pH=pHf −pHi ) was plotted against the pHi and the point of bond from Fe3 O4 . The band at 540 cm−1 in the FT-IR spectrum
intersection of the resulting curve at which pH=0 was the pHzpc of MACz, which was not observed in ACn and ACz, corresponded
value. to the Fe-O stretching vibration in the octahedral and tetrahedral
sites of magnetite [9,21]. When the spectra of the active carbons
3. Results and discussion and CS were compared, it was observed that the stretching vibra-
tion of the O–H band at around 3300 cm−1 disappeared for active
3.1. Characterization of active carbons carbons because of the carbonization and activation together with
the removal of water [20].
Fig. 1 shows the FT-IR spectra for raw coconut shells (CS, nat- The surface morphologies of ACz and MACz were determined
ural coconut shell) and prepared activated carbons. The prepared by SEM and EDX analysis. The elements on the surface of ACz and
active carbons and magnetic active carbon were shortened as ACn MACz were determined by using EDX. The SEM images and EDX
(activated carbon without ZnCl2 ), ACz (obtained activated carbon of the active carbons were depicted in Figs. 2 and 3(A). ACn and
by ZnCl2 ), and MACz (magnetic active carbon). As given in Fig. 1, ACz have a porous and granular surface with different pore di-
the CS has different functional groups. Based on Fig. 1, the peaks ameters. After the formation of Fe3 O4 magnetic particles in the
at 3334 cm−1 , 2924 cm−1 , and 1739 cm−1 can be attributed to the MACz, it can be seen variations in the surface morphology and
hydroxyl groups (-OH), the C–H stretching, and the carbonyl groups pores of active carbon that can be owing to the generation of
(C=O), respectively. The peaks at around 1373 cm−1 and 1604 particles in the active carbon. Although the EDX analysis of the
cm−1 are due to C=C stretchings of aromatic rings. The peaks at MACz showed that the existence of peaks for C, O, Zn and Fe
1233 cm−1 and 1030 cm−1 can be attributed to the C–O stretching which were important constituents of the MACz, ACz just con-
in phenols, alcohols, or ethers. The existence of hydroxyl groups, tains carbon, oxygen, and zinc. There was no iron. The EDX analy-
ethers, carbonyl groups, phenols, and aromatic compounds con- sis confirmed the existence of iron particles in the MACz (Fig. 3(A))
firm the lignocellulosic structure of coconut shell [19]. When the [10,22].
raw material and activated carbons are compared, a decrease is ob- XRF was used to investigate the elemental compositions of ACz
served in the number of peaks, implying that there is a decrease of and MACz samples. Major elements were determined in MACz (Fe,
functional groups in the raw material. It can be seen that the peaks Si, Mn, Al, Ca, Mg, P, K, Na,) and ACz (P, Ca, Si, Mn, Al, K) by XRF
at 3300 cm−1 , 1739 cm−1 , and 1233 cm−1 (the presence of –OH, analysis. High iron content was also determined in MACz by us-
C=O, and aromatic C=C stretch) disappeared in the spectrum of ing XRF analysis. The major impurities were determined as 0.004%
activated carbons. These changes indicated that there was a break Mn3 O4 , 0.093% CaO, 0.033% K2 O, 0.023% P2 O5 , 9.872% SiO2 , 1.334%
of chemical bonds during the carbonization and activation process Al2 O3 in ACz and 43.528% Fe3 O4 , 0.02% Mn3 O4 , 0.044% CaO, 0.014%
[19]. According to the literature [15,20], the vibration band of Fe- K2 O, 0.069% P2 O5 , 0.128% SiO2 , 0.053% Al2 O3 , 0.011% MgO, 0.192

3
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 1. FT-IR spectra of CS, ACn, ACz and MACz.

Table 1
BET results of prepared active carbons.

Micro volume Meso volume Total volume (Vt) percentage of micro Average pore
Active carbons SBET(m2 g−1 ) (Vmicro ) (cm3 g−1 ) (Vmeso ) (cm3 g−1 ) (cm3 g−1 ) poresVmicro / Vt (%) diameter (nm)

ACn 3.39 0.001216 0.00039 0.001613 75.4 1.90

ACz 935.46 0.25593 0.155302 0.411232 62.2 1.76
MACz 747.71 0.175232 0.241716 0.416948 42.02 2.23

Na2 O by weight in MACz. High iron content (Fe3 O4 ) was detected ing to note that the mesoporous fraction of MACz increases with
in MACz by XRF analysis. The contents of Fe3 O4 were found as% 44 the increasing of Fe3 O4 particles. This reflects that the magnetic
for MACz. Similar results were obtained by Mohan et al. [22]. Fe3 O4 particles cause a partial pore blockage, however, an extra
Table 1 shows the porosity of the prepared AC samples. An in- mesopore structure is also being formed when the Fe3 O4 parti-
crease in the pore volume of MACz was observed. The data showed cles are added to the structure [21]. The pore size distribution in
that the SBET of the ACn increased from 3.39 m2 g−1 to 935.46 the macropore (50–300 nm), mesopore (2–50 nm), and micropore
m2 g−1 for ACz and 747.71 m2 g−1 for MACz. Based on Table 2, it (0–2 nm) diameter ranges [22] were defined by the nitrogen ad-
can see that the value of SBET , Vmicro , and Vmeso increased. Vt is sorption/desorption isotherms measured. Based on Table 1, diam-
connected to the pore development of the active carbon. The pore eter pores of ACn and ACz indicate micro but MACz is a mezzo.
development can occur from four stages: (a) opening of previously Based on the literature, the surface area of activated carbons was
inaccessible pores, (b) widening of the existing pores, (c) creation determined by BET analysis [23,24].
of new pores, and (d) merger of the existing pores because of pore The TGA curves were given in Fig. 3(B). The specific tempera-
wall breakage [11]. The surface area of MACz decreased from 935 tures of the CS, Acs, and MACz were summarized in Table 2. The
m2 g−1 to 747 m2 g−1 after iron impregnation because of partial coconut shells consist of lignin, cellulose, and hemicellulose. The
pore blockage by dispersed iron-oxide particles [4,22]. It is amaz- decomposition of coconut shells could be related to the pyrolysis

4
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 2. SEM image of ACn, ACz and MACz.

of these polymers. Similar to the literature [23], the decomposi- action leads to the aromatization process, at a low mass-loss rate
tion of lignin, cellulose and hemicellulose happened at ranges of [23]. Based on TGA, it was observed that the first mass loss started
30 0–40 0, 230–310, 180–240 °C respectively. Beyond 400 °C, the at 130 °C. The weight loss percentages were recorded as 5.3 (CS),
most significant reaction causes the aromatization, at a low mass- 3.4 (AC-n), 6.6 (AC-k), and 2.1(AC-z) wt.%, which were supposed
loss rate. According to literature, the lignin begins decomposing to be owing to low molecular weight volatile matter and moisture
at 160–170 °C and continues to decompose at a low rate until loss [25].
nearly 900 °C. Hemicellulose is the second component to begin de- The magnetic behavior of MACz composite was also investi-
composing, followed by cellulose, in a narrow temperature interval gated at 298 K (see Fig. 4(A)). The value of saturation magneti-
from about 200 to 400 °C. Beyond 400 °C, the most significant re- zation for MACz was determined as 22 emu/g.

5
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 3. EDX spectra for ACz and MACz (A), TGA curves of CS, ACn, ACz and MACz (B).

Table 2
TGA data of the active carbons.

20% Weight 50% Weight % Charat %Water

Compounds Ton a TWmax b loss loss 10 0 0 °C loss, 130 °C I.Step,°C,% II.Step,°C,% III.Step,°C,%

CS 233 312 278 339 17.82 5.3 – – –

ACn 378 491 590 – 68.25 3.4 – – –
ACz 442 289, 544, 738 433 702 84.50 2.12 130–372, 1.43 372–625, 4.68 625–1000, 9.36
MACz 254 318, 906 915 – 72.15 2.00 130–465, 2.94 465–1000, –
24.92
a
The onset temperature.
b
Maximum weight loss temperature.

3.2. Point of zero charge and size distribution of particle the pH of the solution increases above pHpzc, the surface of MACz
is negatively charged, while below the pHpzc value the surface of
The pHpzc results of the experiments were determined with MACz is positively charged because of protonation [26,27].
the MACz composite, where the pH ranged from 2 to 12. Fig. 4(B) Methylene blue (MB) is a cationic dyestuff [11]. The pKa of MB
shows that the pHpzc of the MACz was approximately 7. In other is 3.8, for values above this pH value, the cationic species were
words, at this pH, the MACz surface has zero charge [13]. So, when the preponderant MB species in the solutions. Higher solution pH

6
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 6(B) shows the removal percentage of MB by MACz at 298 K. It

can be observed that the adsorption of MB on the MACz increases
with increasing temperature. So, the adsorption process was deter-
mined as endothermic. This may be due to increasing the mobility
of the dye molecules and an increase in the number of active sites
for the adsorption with increasing temperature [33,34]. The bind-
ing of MB with the surface activated carbon due to π -π stacking
can play an important role in the binding of MB to the surface
of MACz. The main mechanism of interactions between carbona-
ceous materials and dyes are the π -π stacking; the aromatic ring
of the adsorbate (MB) interacts with the carbon surface through
the π -electron system of the ring. Similar results were also found
in other works in the literature [33,35].
The linear form of pseudo-first-order and pseudo–second-order
models [34]. were used to define the adsorption kinetics of MB on
MACz. The kinetic parameters were given in Table 3. They were
calculated for kinetic studies by Eqs. (2)–(4).

Pseudo-first order (Lagergren )log(qe − qt )= log qe−k1 /2.303t

(2)

Pseudo-second order (Ho-Mckay )t/qt = 1/k2 qe 2 +t/qe (3)

Intra particle diffusion (Weber-Morris )qt =kid t1/2 +C (4)

where qe and qt are the amounts of adsorbed in equilibrium and

at time t, respectively, k1 , k2 , and kid are the equilibrium rate
constant of pseudo-first-order adsorption (min−1 ), pseudo-second-
order adsorption (g mg−1 min−1 ), and intraparticle diffusion model
(mg g−1 min−1/2 ), and C is the intercept. The kinetic models of ad-
sorption for MB were shown in Fig. 6(C)–(E). According to Table 3,
the results indicate that the pseudo-second-order is more conve-
nient than the pseudo-first-order. According to results, adsorption
kinetic was determined as the pseudo-second-order model owing
Fig. 4. VSM for the MACz (298 K) (A), Zero point charge of MACz at 298 K (B). to its higher R2 (correlation coefficients) and a smaller difference
between qe(exp) and qe(calc) as compared to the pseudo-second-
order model. The pseudo-second-order model is based on the as-
improves the adsorption capacity of MB [28,29]. If pH of the so-
sumption that the rate-limiting step can be chemisorption involv-
lution is less than pHpzc, the adsorbent surface becomes positive
ing valence forces through sharing or exchanging electrons be-
and attracts anions from the solution. However the most-of MB
tween adsorbate and adsorbent. According to the pseudo-second-
molecules remain in molecular form and at the same time, the
order model, the adsorption rate of MB on MACz was primarily
MACz surface possesses a positive charge due to pHpzc (i.e. 7).
affected by the availability of adsorption sites (functional groups)
Conversely, if pH of the solution is greater than pHpzc, the sur-
on the sorbent [36]. ࢞qe (%) is the normalized standard deviation.
face becomes negative and attracts cations from the solution [13].
It was calculated for kinetic studies by Eq. (5).
Fig. 5(A) shows the zeta potential curves of MACz. MACz zeta po-

tential value was −7.23 mV. Therefore, MACz composite has a neg-  [(qe.exp −qe.cal )/qe.exp ]2
ative surface charge and may move towards a positively polarized qe (% )= 100 (5)
N −1
electrode under an electric field [30].
Fig. 5(B) shows that the size distribution by intensity has three where qe,exp and qe,cal (mg g−1 ), and N are the experimental
peaks, recorded as 748.6 nm for peak 1, 161 nm for peak 2, and and calculated equilibrium adsorption capacity and the number of
5150 nm for peak 3. As a whole, the z-average was recorded as data points, respectively. The applicability of the pseudo-second-
503.9 nm. The PDI displays the width of particle size distribution, order kinetic model was confirmed because of the low value of
indicating the nature of dispersion, which ranges between 0 and normalized standard deviation ࢞qe (%) [11]. Moreover, the best-
1 [31]. The value of PDI was determined as 0.611 due to the non- fitted model can determine using the average relative error (ARE)
uniformity of the particle size [31], which shows the narrow size Eq. (6) [11,37]. The equation is given as follows:
distribution [32].  
100 
n
q exp . − qcal.
ARE = (6)
3.3. Adsorption kinetics of MB n q exp .
n=1

The effects of temperature and contact time were investigated qexp. (mg/g), qcal. (mg/g) and n are the experimental and cal-
at 298, 308, and 318 K, contact time (360 min), at natural pH culated equilibrium adsorption capacity and the number of data
(for initial dye concentration of 100 mg L−1 ). Fig. 6(A) shows a points respectively. The values of qe and ARE were shown in
plot of the amount of MB, adsorbed (qe, mg g−1 ) vs. contact time Table 3. Low ARE and q (%) values indicate the accurate fitting
for different temperatures. 0–120 min, which indicates the first of the model [11,38]. According to the results (Table 3), it can be
steps of dye adsorption, suggests rapid external diffusion and sur- seen that the pseudo-second-order (PSO) model seems to be suit-
face adsorption, and 200–360 min, indicates the equilibrium state. able for modeling the adsorption of MACz.

7
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Table 3
Parameters of kinetic model.

Intra-particle
Pseudo first order Pseudo second order diffusion model

qe, exp qe, calc k1 x 103 qe, calc k2 x 104 (g kid (mg g−1
T (K) (mg g−1 ) (mg g−1 ) ࢞qe (%) (min−1 ) R2 ARE ࢞qe (%) (mg g−1 ) mg−1 min−1 ) R2 ARE min−1/2 ) R2

298 20.8 12.96 13.3 7.14 0.993 4.18 0.73 21.23 1.66 0.9898 0.23 0.79 0.9781
308 25.5 15.44 13.9 7.60 0.9521 4.38 0.84 26.11 1.49 0.9941 0.26 1.02 0.9589
318 28.9 19.31 20.4 6.68 0.9851 3.69 0.48 29.41 1.03 0.9817 0.19 1.12 0.9961

Fig. 5. Zeta potential curve obtained for water suspensions of MACz (A), Size distribution graph of MACz (B).

3.4. Adsorption isotherm tal sorption data. Fig. 7 shows the MB adsorption isotherm. The
Langmuir model supposes monolayer adsorption with a homoge-
Adsorption isotherms are generally investigated to gain insight neous distribution of adsorption sites and adsorption energy, with-
into the adsorption mechanism and adsorbate-adsorbent interac- out interactions between the adsorbed molecules [11]. In this re-
tions at a particular temperature. The surface property can be com- search, Langmuir and Freundlich isotherms Eqs. (7) and ((8)) were
mented as a multilayer or a monolayer [33]. The Freundlich and used to explain the MB and MACz interaction.
Langmuir isotherms are commonly used for modeling experimen- Ce /qe = 1/bQmax +Ce/Qmax (7)

8
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 6. Adsorption kinetic of MB on the MACz at different temperature (A) and percent dye adsorption of MB(B) at 298 K onto MACz, Fitting of pseudo-first order (C),
pseudo-second order (D) and intra-particle diffusion (E) kinetic models.

Table 4
Langmuir and Freundlich isotherm parameters.

Langmuir isotherm Freundlich isotherm

Dye T (K) −1 −1
Qmax (mg g ) b(L mg ) R 2
RMSE SSE k(mg g−1 ) 1/n R2 RMSE SSE

MB 298 144.93 0.041 0.9221 0.085 0.053 16.82 0.4065 0.9583 0.077 0.041
308 156.25 0.047 0.9157 0.095 0.040 20.99 0.3436 0.9506 0.086 0.052
318 151.52 0.063 0.8537 0.096 0.065 29.63 0.2994 0.9625 0.850 0.046

where qe (mgg−1 ), Ce (mL−1 ), Qmax (mg g−1 ), and b (L mg−1 ) 308, and 318 K are 16.82, 20.99, and 29.63 mg g−1 , respectively.
are the amount of dye adsorbed at equilibrium, the residual con- This confirms that the adsorption capacity increased as the tem-
centration of dyes in solution at equilibrium, maximum adsorp- perature was increased. The separation factor 1/n connected to
tion capacity, and Langmuir constant, respectively [33]. The val- the adsorption intensity or surface uniformity (0.4065, 0.3436 and
ues of Qmax and b of linear expression of Langmuir adsorption 0.2994) was 0 < 1/n < 1, which shows that this adsorption process
isotherm (see Table 4) were calculated from the slopes and inter- was suitable [40]. The adsorption isotherm parameters, correlation
cept of the linear plot of Ce/qe versus Ce in Fig. 7(B). According to coefficients (R2 ), values error RMSE (root mean squared error), and
Freundlich isotherm, the adsorbent surface is heterogeneous. Fre- the sum of the squares of the error (SSE) were given in Table 4.
undlich isotherm is convenient adsorption at high concentrations The values of RMSE and SSE can be calculated by the following
and inconvenient for low concentration range [22]. The linearized Eqs. (9) and ((10) respectively [37,41,42]:
Freundlich isotherm model may be shown as follows: 
1 
n
logqe = logk+(1/n )logCe (8) RMSE = (q exp . − qcal. )2 (9)
n−1
n=1
where k (mg/g) and n are the Freundlich constants. k is related to
the adsorption capacity and adsorption intensity of the dye on the 
n

different adsorbents and defines the quantity of dye adsorbed on

SSE = (qcal. − q exp . )2 (10)
n=1
adsorbent at equilibrium concentration. n represents the strength
of dye adsorption on the adsorbent. n value from 1 to 10 shows qexp. (mg g−1 ), qcal. (mg g−1 ) and n are the experimental, cal-
relatively strong adsorption [39]. k and n may be calculated from culated equilibrium adsorption capacity and the number of data
the slope and intercept of the linear plot of log qe versus log Ce points respectively. It can be seen that the values of RMSE and SSE
Fig. 7(C)). The parameters of Freundlich and Langmuir were given obtained from Freundlich were lower than those from the Lang-
in Table 4. According to Table 4, especially for the value of R2 , the muir. Low RMSE and SSE values indicate the accurate fitting of
Freundlich isotherm model (R2 =0.7973, 0.8883, and 0.8513) bet- the model [38,41]. Based on these results, the Freundlich isotherm
ter fits the equilibrium data than the Langmuir isotherm model model was the best-fitted adsorption isotherm model for the ad-
(R2 =0.3780, 0.5450, and 0.7290). Based on Table 4, for the Fre- sorption of MB on MACz. The Qmax values of MB on MACz have
undlich isotherm, k (the maximum adsorption capacities) at 298, been compared with those of other adsorbents (Table 5).

9
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

Fig. 7. Nonlinear adsorption isotherm (A), Langmuir (B) and Freundlich (C) isotherm for MB removal by MACz.

Table 5 plot of ln Kc versus 1/T. The values of G0 were defined as −4.42,
Comparison of maximum monolayer capacity for MB on the other different adsor-
−5.10, and −5.96 kJ mol−1 at 298, 308, and 318 K, respectively. The
bents.
negative Go values at 298, 308, and 318 K proposed a feasible
Adsorbents Qm (mg/g) References and spontaneous process of adsorption [35]. The So and the Ho
Magnetic wakame 479 [43] were calculated as 77 J mol−1 K, 18.57 kJ mol−1 , respectively. The
Magnetic alginate/rice husk 274 [44] adsorption process is endothermic for MB, which was confirmed
Ball-milling biochar 500 [45]
by the positive ࢞Ho values. Also, if the magnitude of H lies be-
Activated carbon from eucalyptus 977 [46]
Carbon nanotubes-based polymer nanocomposites 189 [47]
tween 80 and 200 kJ mol−1 , the process of the adsorption may
Activated carbon from coconut shell 156 This study be considered as chemical whereas physical adsorption usually is
seen in a range of 2.1–20.9 kJ mol−1 or ≤40 kJ mol−1 . In this way,
the adsorption of MB on MACz is physical (Ho was calculated as
3.5. Adsorption thermodynamics 18.57 kJ mol−1 ). The values of ࢞H lower than 20 kJ mol−1 implying
that the physisorption interactions such as Van der Waals ones are
Thermodynamic parameters such as the equilibrium constant dominated [48,49]. Positive So values show the increasing ran-
Kc (L g−1 ), the changes of enthalpy (Ho , kJ mol−1 ), standard domness in the adsorbent-solution interface during the adsorption
Gibbs free energy (Go , kJ mol−1 ), and entropy (So , J mol−1 K−1 ) process [32].
were calculated using the following Eqs. (11) and (12):

Go = −RTlnKc (11) 4. Conclusions

lnKc = S0 /R − H0 /RT (12) In this study, Firstly, the active carbon was obtained from co-
conut shells by using ZnCl2 . Magnetic activated carbon (MACz) was
where T (K), R are the absolute temperature and the gas constant prepared by using the co-precipitation method and used as an ad-
(8.314 J mol−1 K−1 ) respectively [38]. Kc (L g−1 ) was calculated by sorbent. Secondly, the active carbons (ACz and MACz) were charac-
using the Langmuir constants Qmax and b (Kc=Qmax x b) and known terized by using several techniques such as FT-IR, TGA, BET, SEM-
as the equilibrium constant. The thermodynamic parameters Ho EDX, XRF techniques. Thirdly, MACz was used as an adsorbent for
and So were calculated by using the slope and intercept of the the removal of methylene blue (MB) from aqueous solution. The

10
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

experimental data were investigated in terms of isotherm mod- [16] K. Lompe, D. Menard, B. Barbeau, K. Quek, X. Wong, Performance of biologi-
eling, kinetic and thermodynamic. The isotherm model was de- cal magnetic powdered activated carbon for drinking water purification, Water
Research 96 (2016) 42–51.
termined as a Freundlich model. The positive H° and decreased [17] E. Altintig, M. Swamy, B. Nagabhushana, R. Krishna, N. Kottam, R. Raveendra,
G° with increasing temperature are obvious indicators of the en- P. Prashanth, Fast adsorptive removal of methylene blue dye from aqueous so-
dothermic character for the MB adsorption on MACz sorbent. The lution onto a wild carrot flower activated carbon: isotherms and kinetics stud-
ies, Desalin. Water Treat. 71 (2017) 399–405.
positive S° value shows increasing irregularity within the solu- [18] M. Silva, M. Santos, M. Júnior, M. Fonseca, E. Filho, Natural palygorskite as
tion/adsorbent during the adsorption of the dye. The MACz adsor- an industrial dye remover in single and binary systems, Mater. Res. 19 (2016)
bent may be used as a low cost and effective adsorbent for MB dye 1232–1240.
[19] C. Joseph, S. Anisuzzaman, S. Abang, B. Musta, K. Quek, X. Wong, Adsorption
removal from aqueous solutions.
performance and evaluation of activated carbon from coconut shell for the re-
moval of chlorinated phenols in aqueous medium, Materials Science (Medzi-
Declaration of Competing Interest agotyra) 23 (2017) 389–397.
[20] C. Song, S. Wu, M. Cheng, P. Tao, M. Shao, G. Gao, Adsorption studies of co-
conut shell carbons prepared by koh activation for removal of lead(ii) from
The authors declare that they have no conflict of interest. aqueous solutions, Sustainability 6 (2014) 86–98.
[21] K. Raj, P. Joy, Coconut shell based activated carbon–iron oxide magnetic
CRediT authorship contribution statement nanocomposite for fast and efficient removal of oil spills, J. Environ. Chem.
Eng. 3 (2015) 2068–2075.
[22] D. Mohan, A. Sarswat, V. Singh, M. Alexandre-Franco, C. Pittman, Development
Hatice Karaer Yağmur: Conceptualization, Methodology, Inves- of magnetic activated carbon from almond shells for trinitrophenol removal
tigation, Resources, Writing - original draft, Writing - review & from water, Chem. Eng. J. 172 (2011) 1111–1125.
[23] C. Saka, BET, TG–DTG, FT-IR, SEM, iodine number analysis and preparation of
editing. İsmet Kaya: Conceptualization, Writing - review & editing.
activated carbon from acorn shell by chemical activation with ZnCl2 , J. Anal.
Appl. Pyrol. 95 (2012) 21–24.
Acknowledgment [24] C. Purnomo, E. Kesuma, I. Perdana, M. Aziz, Lithium recovery from spent
Li-on batteries using coconut shell activated carbon, Waste Manage. 79 (2018)
454–461.
This research was supported by the Scientific Research Project [25] M. Danish, T. Ahmad, R. Hashim, N. Said, M. Akhtar, J. Mohamad-Saleh, O. Su-
of Dicle University (Project Number: Fen.18. 008). laiman, Comparison of surface properties of wood biomass activated carbons
and their application against rhodamine B and methylene blue dye, Surf. In-
References terfaces 11 (2018) 1–13.
[26] L. Cuevas, E. Villacura, C. Toloza, Magellan peat (Sphagnum magallanicum) as
[1] S. Moosavi, C.W. Lai, S. Gan, G. Zamiri, G. Pivehzhani, M.R. Johan, Applica- natural adsorbent of recalcitran synthetic dyes, J. Soil Sci. Plant Nutr. 8 (2008)
tion of Efficient Magnetic Particles and Activated Carbon for Dye Removal from 31–43.
Wastewater, ACS Omega 5 (2020) 20684–20697. [27] M. Haddad, R. Mamouni, N. Saffaj, S. Lazar, Removal of a cationic dye–basic
[2] K. Krishnan, K. Ajmala, A. Faisal, T. Liji, Kinetic and isotherm modeling of red 12 from aqueous solution by adsorption onto animal bone meal, J. Assoc.
methylene blue adsorption onto kaolinite clay at the solid-liquid interface, Sep. Arab Univ. Basic Appl. Sci. 12 (2012) 48–54.
Sci. Technol. (2020), doi:10.1080/01496395.2014.965832. [28] H.R. Sousa, L.S. Silva, P.A.A. Sousa, R.R.M. Sousa, M.G. Fonseca, J.A. Osajima,
[3] K. Krishnan, S. Suresha, S. Arya, K. Sreejalekshmi, Adsorptive removal of E.C.S. Filhoa, Evaluation of methylene blue removal by plasma activated paly-
2,4-dinitrophenol using active carbon: kinetic and equilibrium modeling at gorskites, J. Mater. Res. Technol. 8 (2019) 5432–5442.
solid–liquid interface, Desalin. Water Treat. (2020), doi:10.1080/19443994. [29] Z. Ciğeroğlu, A. Haşimoğlu, O. Özdemir, Synthesis, characterization and an ap-
2014.890548. plication of graphene oxide nanopowder: methylene blue adsorption and com-
[4] K. Sun, C. Leng, J. Jiang, Q. Bu, G. Lin, X. Lu, G. Zhu, Microporous activated car- parison between experimental data and literature data, J. Dispers. Sci. Technol.
bons from coconut shells produced by self-activation using the pyrolysis gases (2020), doi:10.1080/01932691.2019.1710526.
produced from them that have an excellent electric double layer performance, [30] W. Chartarrayawadee, S. Moulton, C. Too, G. Wallace, Fabrication of graphene
New Carbon Mater. 32 (2017) 451–459. electrodes by electrophoretic deposition and their synergistic effects with PE-
[5] V. Almeida, O. Pezoti, A. Cazetta, I. Souza, K. Bedin, A. Martins, T. Silva, Ad- DOT and platinum, Chiang Mai J. Sci. 40 (2013) 750–762.
sorption studies of methylene blue onto ZnCl2 -activated carbon produced from [31] M. Singh, S. Singh, R. Rachana, Development, characterization and cytotoxicity
buriti shells (Mauritia flexuosa L.), J. Ind. Eng. Chem. 20 (2014) 4401–4407. evaluation of Gingko biloba extract (EGB761) loaded microemulsion for intra–
[6] K. Promdeen, J. Chanvidhwatanakit, S. Satitkune, C. Boonmee, T. Kawichai, nasal application, J. Appl. Pharm. Sci. 7 (2017) 24–34.
S. Jarernprasert, T. Vitidsant, Characterization of carbon materials and differ- [32] Y. Song, I. Chew, T. Choong, J. Tan, K. Tan, Isolation of nanocrystalline cellulose
ences from activated carbon particle (ACP) and coal briquettes product (CBP) from oil palm empty fruit bunch–a response surface methodology study, in:
derived from coconut shell via rotary kiln, Renew. Sustain. Energy Rev. 75 MATEC Web of Conferences, 60, 2014, p. 04009.
(2017) 1175–1186. [33] V. Kunasekaran, K. Krishnamoorthy, Multi criteria decision making to select
[7] S. Karagöz, T. Tay, S. Uçar, M. Erdem, Activated carbons from waste biomass by the best method for the preparation of solid lipid nanoparticles of rasagiline
sulfuric acid activation and their use on methylene blue adsorption, Bioresour. mesylate using analytic hierarchy process, J. Adv. Pharm. Technol. Res. 5 (2014)
Technol. 99 (2008) 6214–6222. 115–121, doi:10.4103/2231-4040.137410.
[8] R. Pereira, C. Veloso, N. Silva, L. Sousa, R. Bonomo, A. Souza, M. Souza, [34] H. Karaer, İ. Kaya, Synthesis, characterization of magnetic chitosan/active char-
R. Fontan, Preparation of activated carbons from cocoa shells and siriguela coal composite and using at the adsorption of methylene blue and reactive
seeds using H3 PO4 and ZnCl2 as activating agents for BSA and α -lactalbumin blue4, Microporous Mesoporous Mater. 232 (2016) 26–38.
adsorption, Fuel Process. Technol. 126 (2014) 476–486. [35] U. Isah, G. Abdulraheem, S. Bala, S. Muhammad, Kinetics, equilibrium and
[9] Z. Yan, Z. Hao, C. Wang, H. Jiang, H. Xu, Magnetic particles modification of co- thermodynamics studies of C.I. Reactive blue 19 dye adsorption on coconut
conut shell-derived activated carbon and biochar for effective removal of phe- shell based activated carbon, Int. Biodeterior. Biodegra. 102 (2015) 265–
nol from water, Chemosphere 211 (2018) 962–969. 273.
[10] R. Foroutan, R. Mohammadi, J. Razeghi, B. Ramavandi, Performance of algal ac- [36] K. Wong, N. Eu, S. Ibrahim, H. Kim, Y. Yoon, M. Jang, Recyclable magnetite-
tivated carbon/Fe3 O4 magnetic composite for cationic dyes removal from aque- loaded palm shell-waste based activated carbon for the 3 effective removal
ous solutions, Algal Res. 40 (2019) 101509. of methylene blue from aqueous solution, J. Clean. Prod. (2020), doi:10.1016/j.
[11] A. Cazetta, A. Vargas, E. Nogami, M. Kunita, M. Guilherme, A. Martins, T. Silva, jclepro.2015.12.063.
J. Moraes, V. Almeida, NaOH-activated carbon of high surface area produced [37] F. Guo, X. Li, X. Jiang, X. Zhao, C. Guo, Z. Rao, Characteristics and
from coconut shell: kinetics and equilibrium studies from the methylene blue toxic dye adsorption of magnetic activated carbon prepared from biomass
adsorption, Chem. Eng. J. 174 (2011) 117–125. waste by modified one-step synthesis, Colloids Surf. A 555 (2018) 43–
[12] K. Qiao, W. Tian, J. Bai, L. Wang, J. Zhao, Z. Du, X. Gong, Application of mag- 54.
netic adsorbents based on iron oxide nanoparticles for oil spill remediation: a [38] L. Fajri, M. Nawi, S. Sabar, N. Nawawi, Preparation of immobilized ac-
review, J. Taiwan Inst. Chem. Eng. 97 (2019) 227–236. tivated carbon-polyvinyl alcohol composite for the adsorptive removal
[13] D. Aht-Ong, V. Sricharoenchaikul, N. Somsesta, Adsorption removal of methy- of 2,4-dichlorophenoxyacetic acid, J. Water Process Eng. 25 (2018) 269–
lene blue onto activated carbon/cellulose biocomposite films: equilibrium and 277.
kinetic studies, Mater. Chem. Phys. 240 (2020) 122221. [39] R.L. Tseng, F.C. Wu, R.S. Juang, Characteristics and applications of the Lager-
[14] H. Zhang, X. Yao, L. Ji, J. Guo, S. Ge, W. Lu, L. Cai, Y. Wang, Magnetic activated gren’s first-order equation for adsorption kinetics, J. Taiwan Inst. Chem. Eng.
biochar nanocomposites derived from wakame and its application in methy- 41 (2010) 661–669.
lene blue adsorption, Bioresour. Technol. 302 (2020) 122842. [40] N.A. Negm, M.G. El Wahed, A.R. Hassan, M.T.H.A. Kana, Feasibility of metal ad-
[15] E. Altintig, M. Onaran, A. Sarı, H. Altundağ, M. Tüzen, G. Gao, Preparation, char- sorption using brown algae and fungi: effect of biosorbents structure on ad-
acterization and evaluation of bio-based magnetic activated carbon for effec- sorption isotherm and kinetics, J. Mol. Liq. 264 (2018) 292–305.
tive adsorption of malachite green from aqueous solution, Materials Chemistry [41] Y. Kim, J. Kim, Isotherm, kinetic and thermodynamic studies on the adsorption
and Physics 220 (2018) 313–321. of paclitaxel onto Sylopute, J. Chem. Thermodyn. 130 (2019) 104–113.

11
H.K. Yağmur and İ. Kaya Journal of Molecular Structure 1232 (2021) 130071

[42] R. Abbas, H. Hami, A. Waheb, H. Kassem, N. Mahdi, Removal of eriochrom [46] Q. Han, J. Wang, B. Goodman, J. Xie, Z. Liu, High adsorption of methylene blue
black T from aqueous solution using Al2 O3 surface: linear and non-linear by activated carbon prepared from phosphoric acid treated eucalyptus residue,
isotherm models, error analysis and thermodynamic studies, Mater. Today 20 Powder Technol. 366 (2020) 239–248.
(2020) 599–604. [47] D. Gan, J. Dou, Q. Huang, H. Huang, J. Chen, M. Liu, H. Qi, Carbon nanotubes-
[43] X. Yao, L. Ji, J. Guo, S. Ge, W. Lu, Y.Wang L.Cai, W. Song, H. Zhang, Magnetic based polymer nanocomposites: bio-mimic preparation and methylene blue
activated biochar nanocomposites derived from wakame and its application in adsorption, J. Environ. Chem. Eng. (2020), doi:10.1016/j.jece.2019.103525.
methylene blue adsorption, Bioresour. Technol. 302 (2020) 122842. [48] M. Ghaedi, S. Heidarpour, S. Kokhdan, R. Sahraie, A. Daneshfar, B. Brazesh,
[44] E. Alver, A. Metin, F. Brouers, Methylene blue adsorption on magnetic algi- Comparison of silver and palladium nanoparticles loaded on activated carbon
nate/rice husk bio-composite, Int. J. Biol. Macromol. 154 (2020) 104–113. for efficient removal of methylene blue: kinetic and isotherm study of removal
[45] Y. Li, A. Zimmerman, F. Hed, J. Chen, L. Han, H. Chen, X. Hug, B. Gao, Solven- process, Powder Technol. 228 (2012) 18–25.
t-free synthesis ofmagnetic biochar and activated carbon through ball-mill ex- [49] V. Mahida, M. Patel, Removal of some most hazardous cationic dyes using
trusion with Fe3O4 nanoparticles for enhancing adsorption of methylene blue, novel poly(NIPAAm/AA/N-allylisatin) nanohydrogel, Arabian J. Chem. 9 (2016)
Sci. Total Environ. 722 (2020) 137972. 430–442.

12