Applied Clay Science 137 (2017) 213–224

Contents lists available at ScienceDirect

Applied Clay Science

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

Research paper

Effect of ball milling process on the structure of local clay and its
adsorption performance for Ni(II) removal
Shahrzad Maleki, Ayoub Karimi-Jashni ⁎
Department of Civil and Environmental Engineering, Shiraz University, Po. Box 7134851156, Shiraz, Iran

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

Article history: The effect of ball milling process on the properties of the natural local clay including structural changes and ad-
Received 19 March 2016 sorption capacity for the removal of Ni(II) ions from aqueous solutions, was investigated. The local clay was
Received in revised form 23 October 2016 ground at varying times from 5 to 20 h with a 10:1 or 20:1 weight ratio of the balls to powder, which produced
Accepted 5 December 2016
six different ball milled clays (BM-Clay1 to BM-Clay6). Ground clays were then characterized by particle size
Available online 20 December 2016
measurement, X-ray diffraction, Fourier transform infrared spectroscopy, scanning electron microscopy analysis,
Keywords:
and adsorption experiments. These analyses were performed to evaluate the changes in particle size distribution,
Ball milling morphology, crystallinity, and adsorption characteristics. According to XRD analysis, the degree of amorphization
Natural clay increased with grinding time, and the present crystalline phase in the final ball milled clay was found to be
Ni(II) removal quartz, which was the associated phase in the original unmilled clay mineral. FTIR studies indicated the destruc-
Isotherm tion of the layers of the montmorillonite upon milling. Adsorption experiments revealed that under similar con-
Kinetics ditions, BM-Clay2 has the most adsorption capacity for Ni(II) ions. Adsorption characteristics of this ball milled
clay for the removal of Ni(II) ions from aqueous solutions was examined in batch adsorption studies under dif-
ferent conditions of contact time, solution pH, adsorbent dose, and initial Ni(II) concentration. The Langmuir
maximum adsorption capacity was found to be 29.76 mg/g at pH 7 and 25 °C. The adsorption kinetics data
showed better agreement with the pseudo second-order kinetic model. Also, both surface adsorption and
intra-particle diffusion contributed to the rate limiting steps in the adsorption of Ni(II) on clay adsorbent. In con-
clusion, grinding of local clay using ball milling process can significantly increase the adsorption capacity of the
clay for Ni(II) removal from aqueous solutions.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction electrochemical treatment technologies, ion exchange, and adsorption

(Fu and Wang, 2011; Kurniawan et al., 2006). Among these methods,
Heavy metals in wastewater released from the waste discharge of in- adsorption has been found to be one of the most effective, economical,
dustrial manufacturing processes are one of the major environmental and easy to adapt process for the removal of heavy metals from solu-
concerns due to their toxicity, persistency, non-biodegradability, and tions (Qiu et al., 2015; Sis and Uysal, 2014).
accumulation in living organisms (Glatstein and Francisca, 2015; Jiang Recently, clays or clay materials have gained much attention as the
et al., 2010; Xing et al., 2011). In particular, nickel is released into aque- adsorbent. Clay minerals, which are important constituents of soil for
ous systems from the manufacturing process of stainless steel, super al- immobilization of contaminants, play this role by taking up various pol-
loys, metallic alloys, electroplating, batteries, etc. (Ijagbemi et al., 2009; lutants as water flows over the soil or penetrates into the ground. Im-
Vieira et al., 2010a). Nickel exceeding its critical level, has been reported mobilization of contaminants takes place through either ion exchange
to be toxic and carcinogenic. It can cause health problems such as lung or adsorption processes, or a combination of both. The high adsorption
and kidney problems, gastrointestinal distress, headache and skin der- properties, non-toxicity, abundant availability, high specific surface
matitis (Dawodua and Akpomie, 2014; Fu and Wang, 2011; Ijagbemi area, mechanical stability, layered structure, and high cation exchange
et al., 2009; Vieira et al., 2010a). Different treatment techniques have capacity (CEC), make clays and clay minerals to be attractive adsorbents
been used to remove heavy metals from wastewater such as chemical for the removal of different pollutants (Djomgoue et al., 2012; Gupta
precipitation, coagulation–flocculation, flotation, membrane filtration, and Bhattacharyya, 2008; Hao et al., 2014). A number of studies have
been reported on the adsorption of Ni(II) by natural clays without any
⁎ Corresponding author.
treatment including natural kaolinite and montmorillonite (Dawodua
E-mail addresses: [email protected] (S. Maleki), [email protected], and Akpomie, 2014; Donat et al., 2005; Gupta and Bhattacharyya,
[email protected] (A. Karimi-Jashni). 2008; Ijagbemi et al., 2009; Jiang et al., 2010; Shirvani et al., 2015;

http://dx.doi.org/10.1016/j.clay.2016.12.008
0169-1317/© 2016 Elsevier B.V. All rights reserved.
214 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

Vieira et al., 2010b). In addition, different modification techniques Table 1

have been used for the improvement of clay quality and its characteris- Characteristics of various BM-Clays.

tics, including intercalation and pillaring, and acid activation Clay name Ball to powder weight ratio Grinding time (h)
(Bhattacharyya and Gupta, 2008a). In recent years, various forms of BM-Clay1 10:1 5
modified clays have been used by researchers to remove Ni(II) ions BM-Clay2 10:1 10
from aqueous solutions (Bhattacharyya and Gupta, 2008b, 2008c; BM-Clay3 10:1 15
Ijagbemi et al., 2010). BM-Clay4 10:1 20
BM-Clay5 20:1 10
It is well known that during dry grinding of mineral particles, me-
BM-Clay6 20:1 20
chanical effects cause surface, structural, and morphological changes
(Djukić et al., 2013; Li et al., 2014). There are some reports on the effects
of dry grinding on calcite (Li et al., 2014), palygorskite (Boudriche et al.,
2014; Liu et al., 2012), and pyrophyllite (Zhang et al., 2015). In addition,
grinding of clay minerals was the subject of several studies, namely the 20 h. Samples of Ball Milled Clay (BM-Clay) are denoted as Table 1
grinding of montmorillonite (Hrachová et al., 2007; Ramadan et al., through the whole text.
2010; Vdović et al., 2010) and kaolinite (Gamelas et al., 2014;
Hamzaoui et al., 2015; Shahverdi-Shahraki et al., 2015; Valášková et 2.3. Material characterization
al., 2011; Vdović et al., 2010; Zbik and Smart, 2005). These studies
have focused on investigating the effects of dry grinding on the mor- The cation exchange capacity (CEC) of the purified clay was
phology, microstructure, aggregation, and changes of surface properties determined using the method described by Meier and Kahr (1999)
of particles, especially specific surface area and cation exchange capaci- by complex of copper ion with triethylenetetramine (Cu(II) sulfate:
ty. Structural deformation caused by grinding such as distortion, frag- assay ≥ 99%, Merck, Germany; triethylenetetramine: assay ≥ 95%,
mentation, and reduction of particle size, followed by an increase in Merck, Germany). The point of zero charge (pHpzc) was determined
specific surface area, exfoliation of particles and amorphization, can en- using technique described by Lazarevića et al. (2007) and Ijagbemi et
hance the adsorption capacity of the clays (Dukić et al., 2015b). al. (2009).
Although numerous studies report the effect of mechanical grinding The X-ray powder diffraction (XRD) patterns of the samples were
on the structure of clay minerals, only a limited number of them have measured using a Bruker D8 advance diffractometer with Cu-Kα radia-
investigated the role of particle size changes caused by ball milling of tion (λ = 1.54 Å) from 5° to 90° (2θ) in 0.05° steps. Chemical composi-
clay minerals on their ability for the removal of heavy metals from solu- tion of the clay was determined by using X-ray fluorescence (XRF)
tions (Djukić et al., 2013; Dukić et al., 2015a; Vhahangwele and Mugera, spectroscopy (Philips PW1480).
2015). The particle size distribution of samples were measured on
This study aims to investigate the effects of ball milling process on dispersed particles in an appropriate dispersant using a dynamic light
the properties of the local clay minerals, including structural changes scattering particle size analyzer (Horiba LB-550), which covers ranges
and adsorption capacity for the removal of Ni(II) ions from aqueous so- from 1 to 6000 nm. The method is based on the measurement of the
lutions. For that, the influence of different milling times and intensities laser light scattered from a sample cell that contains suspension. Ultra-
on adsorption behavior of the clay was investigated. Moreover, the ef- sonic disrupter UD-201 was used for dispersion of particles in the
fects of contact time, solution pH, adsorbent dose, and initial metal solu- dispersant.
tion concentration were investigated for the Ball Milled Clay (BM-Clay) Morphologies and phase distribution in powder samples were ana-
with the highest adsorption capacity. Various adsorption isotherm and lyzed by scanning electron microscope (SEM) TESCAN VEGA3 equipped
kinetic models were applied to describe the main mechanisms involved with the energy dispersive X-ray spectroscopy (EDX) detector. The Fou-
in the removal process. rier transform infrared spectra (FTIR) of the samples in the region 4000–
400 cm−1 were obtained in transmission by a PerkinElmer spectrum RX
I using the KBr pellets technique.
2. Materials and methods
2.4. Adsorption experiments
2.1. Preparation of clay sample
Standard solution of Ni(II) (1000 mg per liter) was prepared by dis-
Natural local clay was obtained from Jahrom Mine, Fars province,
solving certain amount of nickel sulfate (NiSO4 · 6H2O) in distilled
south of Iran, and was purified by sedimentation technique as follows.
water, which was further diluted to the desired concentrations required
The purified clay fraction was obtained by dispersing 10 g of clay
for the adsorption experiments. Before the experiments, the initial pH of
lumps in 1 L distilled water, allowed to swell for 24 h, and stirred for
the Ni(II) solutions was adjusted using diluted HCl or NaOH. All
30 min. Then, the supernatant dispersion of particles was separated
chemicals used in this research were of analytical reagent grade
after about 2 min sedimentation and particles b2 μm were collected
(Merck, Germany).
after 20 h. Finally, the collected clay was dried at 90 °C in oven and
Batch adsorption experiments were performed in 250 mL capped
was ground to pass through a sieve with a pore size of 74 μm
bottles by addition of various amounts of adsorbent into 100 mL of Nick-
(200 mesh). Afterward, a dry grinding was performed as explained in
el solutions and agitating in a temperature-controlled rotating shaker
the next section.
for a desired time interval. After the shaking procedure, the suspensions
were centrifuged and concentrations of Ni(II) remaining in the superna-
2.2. Ball milling of clay tant liquid were measured with an air acetylene flame atomic adsorp-
tion spectrometer (Shimadzu A-A 680).
Mechanical milling of clay minerals was carried out using a laborato- The removal of Ni(II) ions was studied as a function of pH (3.0–10.0),
ry planetary ball mill consisting of two containers (7.0 cm inner diame- initial Ni(II) concentration (10–90 mg/L), contact time (0–1440 min),
ter), using stainless balls with various diameters from 5 to 20 mm as and clay amount (0.2–5.0 g/L). For the determination of adsorption
milling media. In order to investigate the relationship between the isotherms, solutions of Ni(II) ions with different concentrations of
grinding intensity and the rate of change of clay structure, various ratios 10 to 90 mg/L at pH 7 were agitated with 2 g/L of the adsorbent until
of the weight of balls to powder including 10:1 and 20:1 were used. The equilibrium was achieved. The kinetics of adsorption was studied by in-
milling speed was set at 500 rpm, and the grinding time varied from 5 to vestigating the Ni(II) ions uptake from the solution of 50 mg/L
 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224 215

concentration at different time intervals by using 2 g/L of clay. Temper- external film diffusion, intraparticle diffusion, and adsorption to active
ature and pH were kept constant at 25 °C and 7.0, respectively. Experi- sites (Ijagbemi et al., 2010; Malamis and Katsou, 2013).
ments were conducted in duplicate and the results were given as Assuming pseudo first-order kinetic, the rate of adsorption can be
averages. The relative standard deviation was lower than 2%. determined by using the Lagergren equation (Dawodua and Akpomie,
The amount of metal ion adsorbed was calculated using the follow- 2014; Sdiri et al., 2011):
ing equation:
k1 t

 logðqe −qt Þ ¼ logqe − ð7Þ
C i −C f V 2:303
qt ¼ ð1Þ
m where qe and qt are the adsorbed quantity per unit mass (mg/g) at equi-
where, qt is the adsorbed mass of metal ion per gram of adsorbent librium and at any time t, respectively, and k1 (min−1) is the adsorption
(mg/g), Ci and Cf are the initial and final concentrations of Ni(II) rate constant.
(mg/L) in the solution, respectively, V is the volume of solution (L), The pseudo second-order kinetic rate equation is given as
and m is the mass of adsorbent (g). (Bhattacharyya and Gupta, 2008a; Gupta and Bhattacharyya, 2006;
Sdiri et al., 2011):
2.5. Data analysis
t 1 t
¼ þ ð8Þ
qt k2 q2e qe
2.5.1. Isotherm models
The Langmuir adsorption isotherm is based on the assumption of
where k2 (g/mg min) is the pseudo second-order rate constant.
monolayer adsorption that takes place at specific homogeneous sites
When there is a possibility of diffusion of adsorbate species into the
within adsorbent. The linear form of this model is described as
interior of the pores of the adsorbent, the intra-particle diffusion model
(Ijagbemi et al., 2010; Sdiri et al., 2011):
can be used as (Bhattacharyya and Gupta, 2008a; Vassileva et al., 2013):
Ce Ce 1
¼ þ ð2Þ qt ¼ kid t 0:5 ð9Þ
qe qm K L qm
where kid is the intra-particle diffusion rate constant. When intra-
where C e is the equilibrium concentration of the adsorbate in the
particle diffusion plays a significant role in controlling the kinetics of
liquid phase (mg/L), q e the amount of the adsorbate in the solid
the adsorption process, the plots of qt vs. t0.5 yield straight line passing
phase per gram of adsorbent (mg/g), and q m and K L are Langmuir
through the origin.
constant.
The liquid film diffusion model may be applied when the flow of the
The Freundlich adsorption isotherm is applicable to the adsorption
adsorbate from the bulk solution to the surface of the solid phase plays
on heterogeneous surfaces or surfaces supporting sites of varied affini-
the most significant role in adsorption (Bhattacharyya and Gupta,
ties. The linear form of Freundlich model is given as (Dawodua and
2008a):
Akpomie, 2014; Ijagbemi et al., 2010; Sdiri et al., 2011):
ln ð1−F Þ ¼ −kfd t ð10Þ
1
logqe ¼ logK F þ log C e ð3Þ
nF
where F = qt/qe is the fractional attainment of equilibrium and kfd is the
where KF and nF are Freundlich coefficients. adsorption rate constant. A linear plot of −ln (1 − F) versus t with zero
The Temkin model unlike two previous models takes into account intercept suggests that the kinetics of the adsorption process is con-
the adsorbing species and adsorbent interactions (Dawodua and trolled by diffusion through the liquid film surrounding the solid
Akpomie, 2014; Ijagbemi et al., 2010): adsorbent.

RT RT 3. Results and discussion

qe ¼ lnAT þ ln C e ð4Þ
bT bT
3.1. Characteristics of local clay
where T is the temperature (K), R is the ideal gas constant (8.314 J/mol K),
and AT and bT are constants. In this section, at first the characteristics of the initial local clay pass-
Dubinin–Radushkevich (D–R) isotherm describes the amount ing through mesh 200 are investigated. The CEC value for the clay was
adsorbed at equilibrium as a function of Polanyi potential, which can 89.2 cmol(+)/kg. In comparison to the CEC of various clays, this sample
be applied to the adsorption data in order to identify physical or chem- is mainly composed of montmorillonite (Bergaya et al., 2006). The X-ray
ical adsorption (Dawodua and Akpomie, 2014; Sari et al., 2007; Sdiri et diffraction pattern of the clay before ball milling is displayed in Fig. 1.
al., 2011): According to XRD pattern, the sample mainly contained montmorillon-
ite and quartz. As shown in Fig. 1, the reflections at 2θ = 6.0, 19.8, 34.9,
lnqe ¼ ln qm −βε 2 ð5Þ 61.8 and 2θ = 26.7 are the characteristic for montmorillonite and
quartz, respectively.
where qm is the monolayer adsorption capacity, β is the activity coeffi- The chemical composition of this local clay by XRF is shown in Table
cient related to the mean sorption energy, and ε is the Polanyi potential 2. The major oxides in the clay were SiO2 and Al2O3. The ratio of SiO2/
expressed as: Al2O3 is equal to 3.32, suggesting that the studied clay sample was rela-
  tively enriched in silica.
1 The pH at which the surface charges of a suspended solid takes a zero
ε ¼ RT ln 1 þ ð6Þ
Ce value is defined as point of zero charge (pHpzc). This value gives an in-
sight into the type of chemical species that are more likely to be re-
moved from aqueous solution. When the pH of the solution is greater
2.5.2. Kinetic models than pHpzc, sorbent surface is negatively charged and could adsorb pos-
Adsorption kinetics can be controlled by several independent mech- itive metal species. The pHpzc of the clay was found to be about 4.0. The
anisms acting in parallel or in series, which may include bulk diffusion, obtained value indicates that the purified local clay could be used to
216 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

Table 3
Parameters of the Langmuir and Freundlich isotherm models for natural clay.

Montmorillonite Equilibrium models Parameters

Quartz
Langmuir qm (mg/g) 17.33
KL (L/mg) 0.09
R2 0.99
Freundlich KF ((mg/g)(L/mg)1/n) 2.85
nF 2.36
R2 0.98

Unmilled 3.2. Comparison of BM-Clays

3.2.1. X-ray diffraction

The XRD patterns of the local clay before and after grinding with ball
BM-Clay1 milling device are shown in Fig. 1. In BM-Clay1, milling resulted in a
rapid decrease in the intensity of the reflection at 2θ = 6.0 and a lower de-
Intensity

crease of other reflections of the montmorillonite. By increasing the grind-

BM-Clay2 ing time and intensity for BM-Clay2 to BM-Clay6, profound structural
changes were observed, which was evident by the progressive reduction
of the reflection intensities. Loss of the crystalline phase of the montmo-
BM-Clay3 rillonite has occurred, which resulted in increasing amorphization.
According to XRD patterns, reflection of quartz was not altered through
ball milling process. Therefore, the main crystalline phase presented in
BM-Clay4 the final ground clay was quartz, which was the associated phase in the
original unmilled clay sample. The harder quartz grains present in the
sample can work as additional mass of grinding bodies and accelerate
the mechanically amorphization of the montmorillonite (Mako et al.,
BM-Clay5
2001). Since the quartz content of the ball milled samples were the
same in this research, the observed XRD patterns is due to ball milling
time and intensity and is not related to the quartz content.
BM-Clay6

3.2.2. FTIR spectroscopy

5 15 25 35 45 55 65 75 85
Fig. 2 shows the FTIR spectra of the initial and ground samples. In the
2theta (degree) FTIR spectra of the unmilled clay, the absorption bands at 3630 and
3422 cm−1, reflect the stretching modes of \\OH which correspond to
Fig. 1. XRD patterns of the local clay before and after grinding.
the interlayer groups and water molecule adsorbed on the clay surface,
respectively (Dukić et al., 2015b). The band at 1643 cm−1 may be attrib-
uted to the\\OH bending vibration of water (Dukić et al., 2015b). The ab-
adsorb Ni(II) ions efficiently at pH values above 4.0, because the adsor-
sorption band at 1030 cm−1 can be assigned to Si\\O stretching mode
bent surface becomes negative at these pH values.
(Hrachová et al., 2007). The bands at 521 and 469 cm−1 correspond to
Adsorption characteristics of the purified natural clay for the remov-
the Al\\O\\Si and Si\\O\\Si vibrations, respectively (Hrachová et al.,
al of Ni(II) ions were studied by varying the initial metal ion concentra-
2007).
tion from 10 to 90 mg/L. Experimental results were analyzed using the
The peak at 3630 cm−1 decreased rapidly during the first 5 h of
Langmuir and Freundlich isotherm models which are reported in
grinding for the BM-Clay1 and was absent from the FTIR spectrum of
Table 3. Results show that correlation coefficient of Langmuir model is
the remaining ground samples, indicating breakdown of the \\OH
higher than Freundlich model, thus Langmuir isotherm model is an ap-
bonds, and destruction of the clay sheets (Hrachová et al., 2007;
propriate model for the simulation of sorption data, suggesting that the
Ramadan et al., 2010). However, the peaks at 3422 and 1643 cm−1 cor-
adsorption occurs through the formation of a monolayer coverage of
responding to the adsorbed surface water and bending vibration of
Ni(II) ions on the surface of the adsorbent clay. Langmuir monolayer ad-
water were still present even for BM-Clay6. Meanwhile, redistribution
sorption capacity of Ni(II) ions on the natural clay was estimated to be
of the intensity in the region from 800 to 500 cm−1 occurred upon mill-
17.33 mg/g.
ing. The absorption band at 521 cm−1 which corresponds to Al\\O\\Si
vibration, gradually decreased and disappeared in BM-Clay4, while an-
Table 2
other absorption band at 721 cm− 1 was observed in the spectra of
Chemical composition of the local clay.
BM-Clay1 to BM-Clay6, which is attributed to the Al\\O stretching
Chemical composition Weight content (%) vibration. These changes indicate the destruction of the bonds
SiO2 54.96 between tetrahedral and octahedral sheets of the montmorillonite
Al2O3 16.55 and decomposition of the layers (Hrachová et al., 2007), which may in-
Fe2O3 2.08
fluence the adsorption properties of the samples which are discussed
CaO 1.52
K2O 1.46
later.
MgO 5.42
Na2O 0.02
MnO 0.006 3.2.3. Particle size analysis
TiO2 0.167 The particle size distribution of samples were measured by Dynamic
P2O5 0.052 Light Scattering (DLS). DLS is a technique that can be used to determine
SO3 0.052 the size distribution of small particles in suspension based on the ran-
LOI 17.52
dom changes measurement of the intensity of scattered light from
 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224 217

Fig. 2. FTIR spectra of the unmilled and ground samples.

particles. The most used parameters obtained from particle size analysis to longer grinding times were fragmented into smaller particles, how-
(PSA) are the median particle size D50, the volume mean diameter D4,3, ever the resulting particles tended to agglomerate. The tendency of
and the span value or distribution width ((D90 − D10) / D50). As shown the particle agglomeration of the samples increased for BM-Clay3 and
in Table 4, D50 and D4,3 decreased after ball milling for BM-Clay2 and BM-Clay4 (Fig. 4e and f) in comparison to BM-Clay2. In the cases of
BM-Clay6. Higher span values are indication of spreading of the size dis- BM-Clay5 and BM-Clay6, particles were in the form of agglomerated
tribution. The size distributions of clay particles for samples of BM- clusters of smaller particles, therefore an increase of particle size was
Clay1, BM-Clay2 and BM-Clay6 are shown in Fig. 3. After 5 h of grinding observed (Fig. 4g and h). These results are not in full agreement with
for BM-Clay1, particles display a broad bimodal size distribution in the the PSA data, where the reduction of the particles size was observed
range between 670 and 6000 nm with mean value of 2388 nm. By in- for BM-Clay6. This inconsistency could be the consequence of the sam-
creasing the time of milling to 10 h for BM-Clay2, a monomodal narrow ple preparation for PSA, as the sample particles were dispersed in the
distribution was achieved. The average particle size was reduced to dispersant by sufficient sonication, which could lead to the separation
370 nm and the distribution was from 225 to 670 nm. Finally, increasing of the agglomerates that are observed in the SEM images of Fig. 4.
the ball to clay weight ratio from 10:1 to 20:1 and increasing the milling In conclusion, the mechanical milling of clay samples occurred in
time from 10 to 20 h, improved the reduction of the particle size which two consecutive steps. At the beginning of the grinding, particle size
resulted in particle size distribution between 15 and 226 nm with the
mean size of 35 nm. 35

3.2.4. SEM imaging 30 BM-Clay1

The morphological properties of raw and ground samples were ob- BM-Clay2
25 BM-Clay6
Frequency (%)

served by SEM (Fig. 4). The unground clay displayed leafy structure
(Fig. 4a and b). Milling of clay for 5 h (BM-Clay1) decreased the particle 20
size, and irregularly shaped particles with varying sizes were formed
(Fig. 4c). In BM-Clay2, particles were smaller in comparison with BM- 15
Clay1 as a result of the duration of milling (Fig. 4d). Samples subjected
10

Table 4 5
Particle size data for various BM-Clays.
0
Samples D50 (nm) D4,3 (nm) Span 1 10 100 1000 10000
BM-Clay1 2090 2388 1.55 Diameter (nm)
BM-Clay2 365 370 0.48
BM-Clay6 26 35 1.19
Fig. 3. Particle size distribution of local clay at different grinding times.
218 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

linear form of Langmuir equation are depicted in Fig. 6. In addition, re-

sults of the fitted isotherm models are summarized in Table 5.
According to Table 5, the correlation coefficients (R2) of the Lang-
muir model are higher than Freundlich model for all of the milled
clays, which suggest the monolayer adsorption on a homogeneous sur-
face. These results are the same as the unmilled clay, independently of
the microstructural changes induced by milling process. The maximum
adsorption capacity for Ni(II) on the analyzed clays and therefore their
removal efficiency, follows the order BM-Clay2 N BM-Clay1 N BM-
Clay3 N BM-Clay4 N BM-Clay5 N BM-Clay6.
The Langmuir adsorption capacities of Ni(II), expressed in mg/g,
were compared with adsorption capacity of the main unmilled clay,
which was equal to 17.33 mg/g. This shows that the maximum adsorp-
tion capacity achieved for the BM-Clay1 (26.04 mg/g) is much higher
than the capacity of the unmilled local clay. For BM-Clay2, the maxi-
mum adsorption capacity is 29.76 mg/g, which is the highest value
and has increased about 71.7% compared to the initial unmilled clay.
By increasing the grinding time for BM-Clay3 and BM-Clay4, the maxi-
mum adsorption capacities are found to be higher than the unmilled
clay, but lower than BM-Clay2. The improvement of the adsorption ca-
pacities of the BM-Clay1 to BM-Clay4 compared to the initial unmilled
clay can be attributed to the decrease in the clay particle size and exfo-
liation of them, which leads to the increasing surface available for ad-
sorption. Increasing grinding time initially raised the Ni(II) adsorption
capacity to some extent, but then lowered it with longer grinding
times. This is consistent with the result of SEM, that at the beginning
of the grinding, particle size was reduced but prolonged milling pro-
duced aggregates. In other words, samples subjected to longer grinding
times are fragmented into smaller particles, however resulted particles
tend to agglomerate in BM-Clay3 and BM-Clay4. A similar trend could
be deduced for specific surface area, such that fragmentation of particles
with increasing milling time in BM-Clay1 and BM-Clay2, increases the
specific surface area of the samples, but prolonged milling decreases it
for BM-Clay3 and BM-Clay4 as a result of agglomeration of smaller par-
ticles. Increasing the weight ratio of the balls to powder from 10:1 to
20:1 for BM-Clay5 and BM-Clay6 decreased the maximum adsorption
capacity to even less than that of the initial unmilled clay. According
to PSA results, decrease in the clay particle size was seen for BM-Clay6
which could improve the adsorption capacity of the milled clay. Howev-
er, decreasing the adsorption capacity for BM-Clay5 and BM-Clay6 may
be related to the decrease in specific surface area, which is the result of
amorphization and agglomeration of finer particles due to interactions
between particles as a result of cohesion and adhesion. This was evident
from the SEM results, which were shown in Fig. 4g and h where clusters
of aggregates have been produced. Moreover, destruction of the bonds
between tetrahedral and octahedral sheets was observed in the FTIR
spectra of BM-Clay4 to BM-Clay6. Decomposition of the layers was in-
creased by increasing the weight ratio of the balls to powder, which is
Fig. 4. SEM images of (a and b) unground sample, (c) BM-Clay1, (d) BM-Clay2, (e) BM- evident from the presence of a new band with increased intensity
Clay3, (f) BM-Clay4, (g) BM-Clay5 and (h) BM-Clay6. near 721 cm−1 in the spectra of BM-Clay5 and BM-Clay6. These miner-
alogical changes may have affected the adsorption mechanism and re-
was reduced (Fig. 4c–f) and prolonged milling produced aggregates as sulted in decreasing the adsorption capacity of the clays. Therefore,
clearly shown in Fig. 4g and h. BM-Clay2 was chosen as the most efficient clay for Nickel removal in
EDX spectrum of the unmilled clay for the region shown in Fig. 4a is this study, and the effect of different parameters on its adsorption is
presented in Fig. 5a. The composition of the clay was dominated by Si, Al discussed in the next section.
and O which were attributed to montmorillonite. Fig. 5b and c show the
EDX analyses of BM-Clay2 and BM-Clay6. Again, distinct signals from Si,
Al and O were detected. The spectrum was nearly the same for the two 3.3. Adsorption study
ball milled clays.
3.3.1. Effect of contact time
The effect of contact time on Nickel removal by BM-Clay2 was stud-
3.2.5. Adsorption of Ni(II) ied at different time intervals in order to establish the equilibrium time
The evaluation of the Ni(II) adsorption capacity of the milled clays for maximum uptake, by using 2 g/L of clay and solution of 50 mg/L con-
were performed using Langmuir and Freundlich isotherm models. For centration. Temperature and pH were kept constant at 25 °C and 7.0, re-
each ground clay, solutions of Ni(II) ions with varying concentrations spectively. The variation in the percentage removal of Ni(II) ions with
of 10 to 90 mg/L were investigated. Equilibrium data in terms of the the adsorption time is presented in Fig. 7.
 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224 219

(a)

(b) (c)

Fig. 5. The EDX analysis of the (a) unmilled clay; (b) BM-Clay2 and (c) BM-Clay6.

The adsorption rate was quite rapid at the beginning, with 50% of ad- The pseudo second-order kinetic model was also used to describe
sorption occurred during the first 30 min of contact time, because at the the adsorption process. Both kinetic model constants, k2 and qe, were
beginning the active sites on the adsorbent surface were more available. determined from the intercept and slope of the linear plot of t/qt against
Afterwards, the removal increased, but with a much lower rate until the t, respectively. According to Table 6, the R2 value was close to unity and
equilibrium was reached. The equilibrium was reached around 780 min qe obtained from the plot was 19.96 mg/g which agreed well with the
and increased slightly up to 1440 min. The slow uptake in the later stage experimental value (19.38 mg/g), indicating that the process of Ni(II)
is probably due to the decrease of adsorption sites on the clay, which is removal better fits to the pseudo second-order than the pseudo-first
gradually filled up with metal ions. Also kinetics become more depen- order kinetic model. Therefore, the adsorption capacity is proportional
dent on the rate at which the metal ions were transported from the sur- to the number of active sites on adsorbent and the rate of removal is
face to the internal adsorption sites (Bhattacharyya and Gupta, 2008b; controlled by chemical reaction (Qiu et al., 2015).
Dawodua and Akpomie, 2014; Olgun and Atar, 2012). To apply intra-particle diffusion model to the data, the plot of qt ver-
Meanwhile, adsorption mechanism of Ni(II) onto BM-Clay2 was in- sus t0.5 is provided in Fig. 8a, which shows two separate regions with
vestigated using several kinetic models, including the pseudo first- different slopes. The initial portion of the plot indicates the boundary
order, pseudo second-order, intra-particle diffusion, and liquid film dif- layer effect representing the diffusion of Ni(II) ions through the solution
fusion models. to the external surface of the BM-Clay2, while the second linear portion
The experimental and calculated parameters of the pseudo first- indicates the intra-particle diffusion of metal ions reaching the active
order and pseudo second-order models are presented in Table 6. The adsorption sites on the internal clay surfaces. By comparing the calculat-
pseudo first-order equation provides a good fit to the experimental ed intra-particle diffusion constants for the first phase (kid,1) and second
data. This is indicated by the high value of the coefficient of determina- phase (kid,2), the lower value is kid,2 indicating that the rate limiting step
tion (R2) which is equal to 0.98. However, the calculated qe value was for Ni(II) adsorption process is intra-particle diffusion. However, the
not in a good agreement with the experimental value, thus the pseudo intra-particle diffusion is not the only rate limiting mechanism, since
first-order model is not suitable for modeling the adsorption of Ni(II) the line did not pass through the origin (Sarkar and Majumdar, 2011;
onto BM-Clay2. Shirvani et al., 2015).
220 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

3 (a) 90
80
2.5 BM-Clay1
BM-Clay2 70
2 BM-Clay3 60

% Removal
Ce/qe (g/L)

BM-Clay4 50
1.5
40
1 30
20
0.5
10
0 0
0 10 20 30 40 50 0 200 400 600 800 1000 1200 1400 1600
Ce (mg/L) Time (min)

Fig. 7. Effect of contact time on Ni(II) removal by BM-Clay2.

14 (b)
12 BM-Clay5 range of 3.0 to 10.0 for the concentration of 50 mg/L and constant clay
10 BM-Clay6 amount of 2 g/L.
The influence of pH on the removal of Ni(II) ions with clay sample as
Ce/qe (g/L)

8 a sorbent is given in Fig. 9. The total percentage removal of Ni(II) with

6 adsorbent increases from 56.4% to 99.3% with increasing pH of the solu-
tion from 3.0 to 10.0. Lowest removal was obtained at pH of 3.0, while
4 an increase in the removal was observed from pH 4.0 to 8.0. A sharp in-
crease in the removal was recorded at pH 9.0. This can be explained by
2
surface properties of clay particles such that at low pH values, large
0 number of H+ ions compete with Ni(II) ions for active sites, which con-
0 10 20 30 40 50 60 70 80 sequently lead to a low total percentage removal. However, as pH in-
Ce (mg/L) creases, concentration of H+ ions decreases and Ni(II) ions are
adsorbed on free binding sites of the clay (Bhattacharyya and Gupta,
Fig. 6. Langmuir isotherm plot for the removal of Ni(II) by (a) BM-Clay1 to BM-Clay4, and 2008a; Olgun and Atar, 2012; Sis and Uysal, 2014).
(b) BM-Clay5 and BM-Clay6. In order to evaluate the attribution of bulk precipitation to the remov-
al of Ni(II), the change of solution concentrations in the absence of clay
Kinetic data was also applied to the liquid film diffusion model by sample was also examined. According to Fig. 9, no Ni(II) removal is pos-
plotting − ln(1 − F) versus t (Fig. 8b). As explained before, diffusion sible without clay sample at low pH values, but chemical precipitation as
from the bulk liquid phase to the surface of an adsorbent might also hydroxide is significant at pH N 8.0. Therefore, an increase in the total re-
play an important role in determining the rate processes. According to moval of Ni(II) ions is seen and N 99% removal was achieved at pH 10.0.
the Fig. 8b, the non-zero intercept again shows that despite giving linear The net removal of Ni(II), which is the difference between the total
plot, the liquid film diffusion might not be the only rate determining removal percentage and the percentage removed by precipitation is
step in adsorption of Ni(II) on clay (Gupta and Bhattacharyya, 2006). shown in Fig. 9. At pH value of 7.0, this difference is the greatest and
As a result, Ni(II) adsorption to BM-Clay2 is considered as a complex the maximum adsorption takes place. Thus, an optimized pH of 7.0
process, in which both intra-particle diffusion and surface sorption con- was selected for further experiments.
tribute to the rate limiting steps.

3.3.3. Effect of adsorbent dose

3.3.2. Effect of pH Adsorbent dose is an important parameter, because it determines
The pH of the aqueous solution is one of the most important param- the capacity of an adsorbent for a given initial concentration of the ad-
eters for the adsorption of metal ions on the clay minerals, because the sorbate at the operating conditions. The influence of adsorbent dosage
surface charge of the adsorbents and degree of ionization depend on the on the adsorption of Ni(II) was studied by varying the mass of BM-
initial pH of the solution (Gupta and Bhattacharyya, 2006; Jiang et al., Clay2 between 0.2 and 5 g/L with a constant initial Ni(II) concentration
2010; Sari et al., 2007). To investigate the influence of the initial solution of 50 mg/L. Fig. 10 shows the equilibrium adsorption capacity, qe, of the
pH on Ni(II) removal, a set of experiments were conducted over the pH clay mineral and the percentage of metal ions adsorbed as a function of
mass of clay. Evidently, there was an increase in the percentage removal

Table 5
Equilibrium isotherm model parameters for Ni(II) sorption on BM-Clays.
Table 6
BM-Clay
Equilibrium models Kinetic parameters for the pseudo first-order and pseudo second-order models.
Parameters 1 2 3 4 5 6
Kinetic models Parameters
Langmuir
qm (mg/g) 26.04 29.76 22.47 19.68 9.21 7.07 Pseudo first-order k1 (min−1) 0.003
KL (L/mg) 0.30 0.31 0.27 0.19 0.16 0.07 qe Cal. (mg/g) 8.00
R2 0.99 0.99 0.99 0.99 0.99 0.99 qe Exp. (mg/g) 19.38
R2 0.98
Freundlich Pseudo second-order k2 (g/(mg min)) 0.0009
KF ((mg/g)(L/mg)1/n) 7.51 8.04 6.11 4.15 2.75 1.19 qe Cal. (mg/g) 19.96
nF 2.67 2.46 2.70 2.33 3.47 2.53 qe Exp. (mg/g) 19.38
R2 0.97 0.93 0.96 0.92 0.97 0.98 R2 0.99
 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224 221

25 (a) 60 100

20 50 80
qt (mg/g)

15 40

% Removal
qe (mg/g)

qe (mg/g)
60
10 30 % Removal
40
5 20

0 10 20
0 10 20 30 40
0 0
t0.5
0 1 2 3 4 5 6
6 (b) Clay (g/L)
5
Fig. 10. Effect of BM-Clay2 dose on the amount of Ni(II) adsorbed per unit mass (qe) and
4 percentage removal.
-ln(1-F)

3
2 3.3.4. Effect of the initial Ni (II) concentration
1 The changes in initial concentration of the metal ions in the
0 liquid phase can affect the adsorption process. The effect of the initial
concentration on Ni(II) removal was investigated by varying initial con-
0 300 600 900 1200 1500
centration in the range of 10–90 mg/L with a constant adsorbent dosage
Time (min) of 2 g/L. As shown in Fig. 11, the amount of Ni(II) adsorbed per unit mass
(qe) increased with the increase in the initial concentration of Ni(II)
Fig. 8. Plot of (a) intra-particle diffusion model and (b) liquid film diffusion model for
Ni(II) adsorption by BM-Clay2.
ions, whereas a decrease in the percentage removal was observed.
With increasing the initial metal ion concentration, the concentration
gradient increases which acts like a driving force to overcome the resis-
tance to mass transfer of the metal ions between the adsorbate and ad-
of nickel with an increase in the amount of clay. This is possibly due sorbent species, thus the amount of metal ion adsorbed per unit mass of
to increased sorptive surface area and the availability of more active the adsorbent increases (Dawodua and Akpomie, 2014; Ijagbemi et al.,
binding sites on the surface of the adsorbent with the increase in adsor- 2010; Jiang et al., 2010; Malamis and Katsou, 2013). On the other
bent dose (Dawodua and Akpomie, 2014; Vieira et al., 2010b). The in- hand, the decrease in adsorption percentages is due to the fact that all
crease in percentage removal of nickel is significant for the mass of adsorbents have a fixed number of active sites and at higher concentra-
clay up to 3 g/L, after which the rate of increase slows down. This can tions the active sites become saturated, thus preventing further metal
be explained by the fact that concentration of metal ions in the solution ion binding (Ozdes et al., 2011).
has reached a low value and the adsorption driving force is therefore
minimum. 3.3.5. Adsorption isotherms
On the other hand, the equilibrium adsorption capacity, qe, of clay The adsorption capacity at equilibrium is an important parameter
decreased with increase in adsorbent dose. When the adsorbent dose because it determines how much adsorbent is required quantitatively
is small, the metal ions compete for adsorption at limited adsorption for the removal of an adsorbate from solution (Zhu et al., 2013). The
sites and as a result sorbent sites become saturated, and therefore qe is equilibrium data obtained from the adsorption of Ni(II) ions onto the
high. With increasing adsorbent content, smaller amount of metal ions BM-Clay2 was fitted to Langmuir isotherm model by plotting Ce/qe ver-
per unit mass of clay is available and as a result, sorbent sites remain un- sus Ce (Fig. 12a). Maximum adsorption capacity (qm) of the BM-Clay2
saturated. Similar results have been reported for the adsorption of for nickel was 29.76 mg/g, and the Langmuir equilibrium constant KL,
heavy metals in various studies (Dawodua and Akpomie, 2014; Gupta had a value of 0.31 L/mg. The linear Langmuir equation gives a good
and Bhattacharyya, 2006; Vieira et al., 2010b). fit for the adsorption isotherm at 25 °C (R2 = 0.99). Based on further

100 30 100
90
80 25 80
70 20
% Removal
% Removal

qe (mg/g)

60 60
50 15
40 with adsorbent 40
10
30 without adsorbent qe (mg/g)
5 20
20 net removal % Removal
10 0 0
0 0 20 40 60 80 100
0 2 4 6 8 10 12 C0 (mg/L)
Initial pH
Fig. 11. Effect of initial concentration on the amount of Ni(II) adsorbed per unit mass (qe)
Fig. 9. Effect of the initial pH and precipitation on Ni(II) removal by BM-Clay2. and percentage removal.
222 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

1.2 (a) where C0 (mg/L) is the initial amount of adsorbate. Depending on the RL
value, there are four probabilities for adsorption; (i) favorable adsorp-
1 tion if 0 b RL b 1, (ii) unfavorable adsorption when RL N 1, (iii) linear ad-
sorption for RL = 1, and (iv) irreversible adsorption for RL = 0. By
Ce/qe (g/L)

0.8
computing adsorption intensity, the RL varies from 0.26 to 0.04 for the
0.6 initial solution concentration of 10 to 90 mg/L. Based on the RL values,
0.4 adsorption is favorable. Moreover, the values tended toward zero,
representing the ideal irreversible adsorption, thus the BM-Clay2 is a
0.2 suitable adsorbent for the adsorption of Ni(II) from aqueous solution.
0 The adsorption isotherm data was also analyzed on the basis of the
Freundlich equation by plotting log qe versus log Ce (Fig. 12b). The coef-
0 5 10 15 20 25 30
ficient of determination (R2) was 0.93 which is less than R2 for Langmuir
Ce (mg/L) plot. The constants KF and nF were derived from the intercept and slope
of the linear plot to be 8.04 (mg/g)(L/mg)1/n and 2.46, respectively. The
(b) exponent 1/nF is an index of the diversity of free energies associated
2 with the adsorption of the solute by adsorbent, which gives an idea
for the favorability of sorption process. For 1/nF b 1, adsorption is favor-
log qe (mg/g)

1.5 able for the entire range of adsorbate concentrations and when 1/nF N 1,
adsorption is favorable only at higher concentrations of adsorbate
1 (Ijagbemi et al., 2010; Sari et al., 2007; Sarkar and Majumdar, 2011).
In the present case, the value of 1/nF was found to be less than unity, im-
0.5 plying that the adsorption of Ni(II) onto BM-Clay2 was favorable under
study conditions.
0 Temkin isotherm was also chosen for adsorption evaluation. Accord-
-0.5 0 0.5 1 1.5 ing to the plot of qe versus ln Ce (Fig. 12c), R2 was 0.99. The constants AT
and bT were calculated from the plot, showing that the adsorption po-
log Ce (mg/L) tential, AT, of clay for Ni(II) was 5.84 L/g and Temkin constant, bT, related
to heat of adsorption was 0.47 kJ/mol.
(c) Finally, Dubinin–Radushkevich isotherm based on the heteroge-
30 neous surface of the adsorbate was applied to the adsorption data
25 in order to distinguish the predominant adsorption type as physical
or chemical sorption. Fig. 12d shows the D-R isotherm plot for the
qe (mg/g)

20 adsorption of Ni(II) ions onto BM-Clay2. The values of monolayer sorp-

15 tion capacity (qm) and β constant were calculated from the plot as
102.32 mg/g and 3 × 10−9 mol2/J2, respectively. The mean energy of ad-
10 sorption E (kJ/mol) can be calculated from the equation (Ijagbemi et al.,
5 2010; Sdiri et al., 2011):

0
1
-1 0 1 2 3 4 E ¼ pffiffiffiffiffiffiffiffiffiffiffi ð12Þ
−2β
ln Ce (mg/L)

The removal mechanism occurs basically via physical interaction

ε2×109 (d) when E b 8 kJ/mol, and via ion-exchange if 8 kJ/mol ≤ E ≤ 16 kJ/mol.
0 0.2 0.4 0.6 0.8 1 The value of E was found to be 12.91 kJ/mol, which is between 8 and
16 kJ/mol, indicating the adsorption process may be carried out via
-7
ion-exchange mechanism.
-7.5 The adsorption constants and correlation coefficients calculated
from the isotherms are presented in Table 7. By comparing the R2 values
-8 obtained from Langmuir, Freundlich, Temkin and D-R isotherm models,
ln qe

-8.5
Table 7
-9 Parameters of the adsorption isotherm models.

-9.5 Equilibrium models Parameters

-10 Langmuir qm (mg/g) 29.76

KL (L/mg) 0.31
R2 0.99
Fig. 12. Plot of (a) Langmuir; (b) Freundlich; (c) Temkin and (d) Dubinin-Radushkevich
Freundlich KF ((mg/g)(L/mg)1/n) 8.04
isotherms for the removal of Ni(II) by BM-Clay2.
nF 2.46
R2 0.93
Temkin AT (L/g) 5.84
bT (kJ/mol) 0.47
analysis of Langmuir equation, the adsorption intensity (RL) is defined R2 0.99
as (Sari et al., 2007; Sdiri et al., 2011): Dubinin–Radushkevich qm (mg/g) 102.32
β (mol2/J2) 3 × 10−9
1 R2 0.96
RL ¼ ð11Þ E (kJ/mol) 12.91
1 þ K L C0
 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224 223

both Langmuir and Temkin models are more suitable for Ni(II) ions with TiO 2 on the heavy metal adsorption characteristics. Ceram. Int. 41,
5129–5137.
adsorption. Fu, F., Wang, Q., 2011. Removal of heavy metal ions from wastewaters: a review.
J. Environ. Manag. 92, 407–418.
Gamelas, J.A.F., Ferraz, E., Rocha, F., 2014. An insight into the surface properties of calcined
4. Conclusions kaolinitic clays: the grinding effect. Colloids Surf. A Physicochem. Eng. Asp. 455,
49–57.
In this study, microstructure and adsorption properties of ball milled Glatstein, D.A., Francisca, F.M., 2015. Influence of pH and ionic strength on Cd, Cu and Pb
removal from water by adsorption in Na-bentonite. Appl. Clay Sci. 118, 61–67.
clays prepared from the local clay, which was mainly composed of Gupta, S.S., Bhattacharyya, K.G., 2006. Adsorption of Ni(II) on clays. J. Colloid Interface Sci.
montmorillonite and quartz, using planetary ball mill were investigated. 295, 21–32.
To study the effect of grinding, properties of the purified clay was ob- Gupta, S.S., Bhattacharyya, K.G., 2008. Immobilization of Pb(II), Cd(II) and Ni(II) ions on
kaolinite and montmorillonite surfaces from aqueous medium. J. Environ. Manag.
served after 5 to 20 h of ball milling at 500 rpm, with 10:1 or 20:1 of 87, 46–58.
weight ratio of balls to powder. XRD and FTIR results indicated that Hamzaoui, R., Muslim, F., Guessasma, S., Bennabi, A., Guillin, J., 2015. Structural and ther-
the layer structure of montmorillonite has been broken by increasing mal behavior of proclay kaolinite using high energy ball milling process. Powder
Technol. 271, 228–237.
the time and intensity of milling, and the only existing crystalline
Hao, J., Xiao, L., Liu, H., Shi, L., Xu, X., Lian, B., Liu, C., 2014. Adsorption of hexavalent chro-
phase in the final ground product (BM-Clay6) was quartz. According mium onto organic bentonite modified by the use of iron(III) chloride. Water Sci.
to the PSA, the particle size of the local clay was reduced by ball milling. Technol. 70, 664–670.
Hrachová, J., Komadel, P., Fajnor, V.Š., 2007. The effect of mechanical treatment on the
Moreover, according to SEM studies, the milling of clay samples oc-
structure of montmorillonite. Mater. Lett. 61, 3361–3365.
curred in two steps, which included particle size reduction at the begin- Ijagbemi, C.O., Baek, M.-H., Kim, D.-S., 2009. Montmorillonite surface properties and sorp-
ning of the grinding and aggregation of particles for prolonged milling tion characteristics for heavy metal removal from aqueous solutions. J. Hazard. Mater.
that led to increasing particle size. Also, the adsorption properties of 166, 538–546.
Ijagbemi, C.O., Baek, M.-H., Kim, D.-S., 2010. Adsorptive performance of un-calcined
milled clays for the removal of Ni(II) ions from aqueous solutions sodium exchanged and acid modified montmorillonite for Ni2+ removal: equilibri-
changed during the milling process. The results revealed that under um, kinetics, thermodynamics and regeneration studies. J. Hazard. Mater. 174,
similar conditions, BM-Clay2 has the most sorption capacity for Ni(II) 746–755.
Jiang, M.-Q., Jin, X.-Y., Lu, X.-Q., Chen, Z.-l., 2010. Adsorption of Pb(II), Cd(II), Ni(II) and
ions. Apparently, the different responses of the milling process on the Cu(II) onto natural kaolinite clay. Desalination 252, 33–39.
clays are a consequence of structural changes such as decrease in parti- Kurniawan, T.A., Chan, G.Y.S., Lo, W.-H., Babel, S., 2006. Physico–chemical treatment tech-
cle sizes, exfoliation of clay minerals and aggregation of fine particles. niques for wastewater laden with heavy metals. Chem. Eng. J. 118, 83–98.
Lazarevića, S., Janković-Častvan, I., Jovanović, D., Milonjić, S., Janaćković, D., Petrović, R.,
Adsorption characteristics of Ni(II) ions onto BM-Clay2 were studied 2007. Adsorption of Pb2+, Cd2+ and Sr2+ ions onto natural and acid-activated sepi-
under different conditions of solution pH, contact time, initial metal ions olites. Appl. Clay Sci. 37, 47–57.
concentration, and clay dose. Various isotherm models were used to Li, T., Sui, F., Li, F., Cai, Y., Jin, Z., 2014. Effects of dry grinding on the structure and granu-
larity of calcite and its polymorphic transformation into aragonite. Powder Technol.
represent experimental data. The monolayer adsorption capacity was
254, 338–343.
found to be 29.76 mg/g by using Langmuir model. By applying the kinet- Liu, Y., Wang, W., Wang, A., 2012. Effect of dry grinding on the microstructure of
ic models to the experimental data, it was found that the adsorption of palygorskite and adsorption efficiency for methylene blue. Powder Technol. 225,
124–129.
Ni(II) onto BM-Clay2 followed the pseudo second-order rate kinetic.
Mako, E., Frost, R.L., Kristof, J., Horvath, E., 2001. The effect of quartz content on the mech-
Moreover, the liquid film diffusion as well as intra-particle diffusion anochemical activation of kaolinite. J. Colloid Interface Sci. 244, 359–364.
mechanisms had significant role in the adsorption of Ni(II) onto BM- Malamis, S., Katsou, E., 2013. A review on zinc and nickel adsorption on natural and mod-
Clay2. Results demonstrated that ball milling of the local clay for 10 h ified zeolite, bentonite and vermiculite: examination of process parameters, kinetics
and isotherms. J. Hazard. Mater. 252–253, 428–461.
with 10:1 weight ratio of ball to powder enhanced the adsorption ca- Meier, L.P., Kahr, G., 1999. Determination of the cation exchange capacity (CEC) of clay
pacity of Ni(II) ions by 71.7% compared to the initial unmilled clay. minerals using the complexes of copper(II) ion with triethylenetetramine and
tetraethylenepentamine. Clay Clay Miner. 47, 386–388.
Olgun, A., Atar, N., 2012. Equilibrium, thermodynamic and kinetic studies for the adsorp-
References tion of lead (II) and nickel (II) onto clay mixture containing boron impurity. J. Ind.
Eng. Chem. 18, 1751–1757.
Bergaya, F., Theng, B.K.G., Lagaly, G., 2006. Handbook of clay science. first ed. Develop- Ozdes, D., Duran, C., Senturk, H.B., 2011. Adsorptive removal of Cd(II) and Pb(II) ions from
ments in Clay Science Vol. 1. Elsevier. aqueous solutions by using Turkish illitic clay. J. Environ. Manag. 92, 3082–3090.
Bhattacharyya, K.G., Gupta, S.S., 2008a. Adsorption of a few heavy metals on natural and Qiu, G., Xie, Q., Liu, H., Chen, T., Xie, J., Li, H., 2015. Removal of Cu(II) from aqueous solu-
modified kaolinite and montmorillonite: a review. Adv. Colloid Interf. Sci. 140, tions using dolomite–palygorskite clay: performance and mechanisms. Appl. Clay Sci.
114–131. 118, 107–115.
Bhattacharyya, K.G., Gupta, S.S., 2008b. Influence of acid activation on adsorption of Ni(II) Ramadan, A.R., Esawi, A.M.K., Gawad, A.A., 2010. Effect of ball milling on the structure of
and Cu(II) on kaolinite and montmorillonite: kinetic and thermodynamic study. Na+-montmorillonite and organo-montmorillonite (Cloisite 30B). Appl. Clay Sci. 47,
Chem. Eng. J. 136, 1–13. 196–202.
Bhattacharyya, K.G., Gupta, S.S., 2008c. Adsorption of Fe(III), Co(II) and Ni(II) on ZrO–ka- Sari, A., Tuzen, M., Citak, D., Soylak, M., 2007. Equilibrium, kinetic and thermodynamic
olinite and ZrO–montmorillonite surfaces in aqueous medium. Colloids Surf. A studies of adsorption of Pb(II) from aqueous solution onto Turkish kaolinite clay.
Physicochem. Eng. Asp. 317, 71–79. J. Hazard. Mater. 149, 283–291.
Boudriche, L., Chamayou, A., Calvet, R., Hamdi, B., Balard, H., 2014. Influence of different Sarkar, M., Majumdar, P., 2011. Application of response surface methodology for optimi-
dry milling processes on the properties of an attapulgite clay, contribution of inverse zation of heavy metal biosorption using surfactant modified chitosan bead. Chem.
gas chromatography. Powder Technol. 254, 352–363. Eng. J. 175, 376–387.
Dawodua, F.A., Akpomie, K.G., 2014. Simultaneous adsorption of Ni(II) and Mn(II) ions Sdiri, A., Higashi, T., Hatta, T., Jamoussi, F., Tase, N., 2011. Evaluating the adsorptive capac-
from aqueous solution unto a Nigerian kaolinite clay. J. Mater. Res. Technol. 3, ity of montmorillonitic and calcareous clays on the removal of several heavy metals
129–141. in aqueous systems. Chem. Eng. J. 172, 37–46.
Djomgoue, P., Siewe, M., Djoufac, E., Kenfack, P., Njopwouo, D., 2012. Surface modification Shahverdi-Shahraki, K., Ghosh, T., Mahajan, K., Ajji, A., Carreau, P.J., 2015. Effect of dry
of Cameroonian magnetite rich clay with Eriochrome Black T. Application for adsorp- grinding on chemically modified kaolin. Appl. Clay Sci. 105–106, 100–106.
tion of nickel in aqueous solution. Appl. Surf. Sci. 258, 7470–7479. Shirvani, M., Rafiei, H.R., Bakhtiary, S., Azimzadeh, B., Amani, S., 2015. Equilibrium, kinetic,
Djukić, A., Jovanović, U., Tuvić, T., Andrić, V., Novaković, J.G., Ivanović, N., Matović, L., 2013. and thermodynamic studies on nickel removal from aqueous solutions using Ca-ben-
The potential of ball-milled Serbian natural clay for removal of heavy metal contam- tonite. Desalin. Water Treat. 54, 464–472.
inants from wastewaters: simultaneous sorption of Ni, Cr, Cd and Pb ions. Ceram. Int. Sis, H., Uysal, T., 2014. Removal of heavy metal ions from aqueous medium using
39, 7173–7178. Kuluncak (Malatya) vermiculites and effect of precipitation on removal. Appl. Clay
Donat, R., Akdogan, A., Erdem, E., Cetisli, H., 2005. Thermodynamics of Pb2+ and Ni2+ ad- Sci. 95, 1–8.
sorption onto natural bentonite from aqueous solutions. J. Colloid Interface Sci. 286, Valášková, M., Barabaszová, K., Hundáková, M., Ritz, M., Plevová, E., 2011. Effects of brief
43–52. milling and acid treatment on two ordered and disordered kaolinite structures. Appl.
Dukić, A.B., Kumrić, K.R., Vukelić, N.S., Dimitrijević, M.S., Baščarević, Z.D., Kurko, S.V., Clay Sci. 54, 70–76.
Matović, L.L., 2015a. Simultaneous removal of Pb2+, Cu2+, Zn2+ and Cd2+ from high- Vassileva, P., Detcheva, A., Uzunov, I., Uzunova, S., 2013. Removal of metal ions from aque-
ly acidic solutions using mechanochemically synthesized montmorillonite–kaolinite/ ous solutions using pyrolyzed rice husks: adsorption kinetics and equilibria. Chem.
TiO2 composite. Appl. Clay Sci. 103, 20–27. Eng. Commun. 200, 1578–1599.
Dukić, A.B., Kumrić, K.R., Vukelić, N.S., Stojanović, Z.S., Stojmenović, M.D., Milošević, Vdović, N., Jurina, I., Škapin, S.D., Sondi, I., 2010. The surface properties of clay minerals
S.S., Matović, L.L., 2015b. Influence of ageing of milled clay and its composite modified by intensive dry milling – revisited. Appl. Clay Sci. 48, 575–580.
224 S. Maleki, A. Karimi-Jashni / Applied Clay Science 137 (2017) 213–224

Vhahangwele, M., Mugera, G.W., 2015. The potential of ball-milled South African benton- Zbik, M., Smart, R.S.C., 2005. Influence of dry grinding on talc and kaolinite morphology:
ite clay for attenuation of heavy metals from acidic wastewaters: simultaneous sorp- inhibition of nano-bubble formation and improved dispersion. Miner. Eng. 18,
tion of Co2+, Cu2+, Ni2+, Pb2+, and Zn2+ ions. J. Environ. Chem. Eng. 3, 2416–2425. 969–976.
Vieira, M.G.A., Almeida Neto, A.F., Gimenes, M.L., da Silva, M.G.C., 2010a. Removal of nickel Zhang, J., Yan, J., Sheng, J., 2015. Dry grinding effect on pyrophyllite–quartz natural
on Bofe bentonite calcined clay in porous bed. J. Hazard. Mater. 176, 109–118. mixture and its influence on the structural alternation of pyrophyllite. Micron
Vieira, M.G.A., Almeida Neto, A.F., Gimenes, M.L., da Silva, M.G.C., 2010b. Sorption kinetics 71, 1–6.
and equilibrium for the removal of nickel ions from aqueous phase on calcined Bofe Zhu, Z., Gao, C., Wu, Y., Sun, L., Huang, X., Ran, W., Shen, Q., 2013. Removal of heavy metals
bentonite clay. J. Hazard. Mater. 177, 362–371. from aqueous solution by lipopeptides and lipopeptides modified Na-montmorillon-
Xing, S., Zhao, M., Ma, Z., 2011. Removal of heavy metal ions from aqueous solution using ite. Bioresour. Technol. 147, 378–386.
red loess as an adsorbent. J. Environ. Sci. 23, 1497–1502.