Accepted Manuscript

Removal of copper and nickel from water using nanocomposite of magnetic

hydroxyapatite nanorods

Dong Nguyen Thanh, Pavel Novák, Jana Vejpravova, Hong Nguyen Vu,
Jaromír Lederer, Tasnim Munshi

PII: S0304-8853(17)33604-1
DOI: https://doi.org/10.1016/j.jmmm.2017.11.064
Reference: MAGMA 63404

To appear in: Journal of Magnetism and Magnetic Materials

Received Date: 27 November 2016

Revised Date: 9 November 2017
Accepted Date: 15 November 2017

Please cite this article as: D.N. Thanh, P. Novák, J. Vejpravova, H.N. Vu, J. Lederer, T. Munshi, Removal of copper
and nickel from water using nanocomposite of magnetic hydroxyapatite nanorods, Journal of Magnetism and
Magnetic Materials (2017), doi: https://doi.org/10.1016/j.jmmm.2017.11.064

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Removal of copper and nickel from water using nanocomposite
of magnetic hydroxyapatite nanorods

Dong Nguyen Thanha, Pavel Novákb, Jana Vejpravovac, Hong Nguyen Vub, Jaromír
Lederera,Tasnim Munshid*
a
Unipetrol Centre of Research and Education, Chempark Litvínov, Záluží 1– Litvínov, 436 70, Czech
Republic
b
University of Chemical Technology, Technická 5, 166 28 Prague 6, Czech Republic
c
Institute of Physics AS CR, v.v.i., Na Slovance 1999/2 182 21 Praha 8, Czech Republic
d
School of Chemistry, University of Lincoln, Brayford Pool, Lincoln, Lincolnshire, LN6 7TS, United
Kingdom
*
Corresponding author: E-mail: TMunshi@lincoln.ac.uk

Abstract
A nanocomposite of magnetic hydroxyapatite was synthesized and tested as an adsorbent for the
removal of copper (Cu (II)) and nickel (Ni(II)) from aqueous solution. The adsorbent was investigated
using Transmission Electron Microscopy (TEM), Scanning Electron Microscopy equipped with an
Energy Dispersive Spectrometer (SEM/EDS), X-Ray powder diffraction (XRD) and the Brunauer–
Elmet–Teller nitrogen adsorption technique (BET-N2 adsorption). Batch experiments were carried out
to determine and compare the adsorption parameters of Fe3O4 and its composite with hydroxyapatite.
It was found that the adsorbent is nanostructured and has a specific surface area of 101.2 m2g-1. The
Langmuir adsorption isotherm was found to be an appropriate model to describe the adsorption
processes, showing the adsorption capacities of Cu(II) and Ni(II) of 48.78 mg g-1 and 29.07 mg g-1,
respectively. In addition to the high adsorption capacity, the fully-adsorbed material could be easily
separated from aqueous media using an external magnetic field. These results suggested that the
utilization of new hydroxyapatite - Fe3O4 nanocomposite for the removal of Cu(II) and Ni(II) is a
promising method in water technology.

Keywords

Nanocomposite; Hydroxyapatite; Magnetite; Copper; Nickel; Adsorption mechanism

1. Introduction

Heavy metals are reported as the major pollutants in both the industrial wastewater and rainwater.
They are non- biodegradable and accumulate in the environment, causing both short and long term
harmful effects1,2. Copper and nickel are the most commonly used heavy metals in metal processing.
It has been found that nickel and copper are accumulative and toxic to most forms of life, even in
small quantities. Exposure to nickel and its compounds can have adverse effects on human health.
Acute nickel poisoning causes headaches, dizziness, nausea and vomiting, chest pain, a dry cough,
shortness of breath, rapid respiration, cyanosis and extreme weakness. Copper can cause stomach and
intestinal distress, liver and kidney damage and anaemia 3. Therefore, the removal of these metals
from polluted wastewater or soils has attracted great attention. The actual removal of heavy metals
from water can be based on a range of techniques, such as ion exchange, adsorption, oxidation–
precipitation, coagulation/electrocoagulation or membrane filtration 4. Identifying an effective
technique for the removal of heavy metals is a challenge and it depends on various parameters, such
as cost and eco-toxicity of secondary by-products. Adsorption is one of the most efficient physico-
5, 6
chemical treatment processes that can replace other more expensive water treatment methods .
Various adsorbents, such as laterite iron concretion, clays, activated carbon, biomasses and zeolites,
6-9
are being used for the treatment of heavy metal polluted water .
Apatite-group minerals form naturally and are stable across a wide range of geologic conditions for
hundreds of millions of years 10. They have a great scientific and economic significance due to their
wide availability, low cost, ease of use and self regeneration. Apatite can also be produced
synthetically by calcium and phosphate precipitation reactions 11. Because of their special chemical
composition and crystal structure, apatite-group minerals are the most promising materials for the
treatment of wastewater and soil containing heavy metals 10, 12, 13. As a member of the apatite mineral
family, hydroxyapatite (Ca10(PO4)6(OH)2 - A) is an ideal material for the long-term containment of
contaminants because of its low cost, high stability and high sorption capacity for heavy metals. In
addition, the formed metal phosphates have a low solubility in water 13.
With the progress of nanotechnology, the focus is gradually shifting towards nanomaterials as either
alternatives to conventional adsorbents or their enhancements. Nanomaterials can exhibit a plenty of
useful properties making them attractive for the adsorption studies especially in water treatment
applications 5. Therefore, there has been significant research on the synthesis and use of
nanostructured hydroxyapatite for the removal of cadmium14, nickel 15, lead and arsenic 16, 17. It has
been shown that nanostructured hydroxyapatite is an effective adsorbent material as indicated by its
sorption capacity.
However, similarly as for other nanoadsorbents, due to its small size it is difficult to separate
nanostructured hydroxyapatite entirely from the treated solution after adsorption as the hydroxapatite
is often suspended in solution and therefore magnetic separtion is a more cleaner, effiecient way of
removing the nanostructured hydroxyapatite . Magnetic separation has been widely used as a fast,
efficient and economic separation method, particularly in water treatment technology 18-20. It implies
that the combination of hydroxyapatite, having a high sorption capacity, with magnetite, allowing
magnetic separation, would be a promising way to make a novel magnetic nanoadosorbent. It leads to
the development of hydroxyapatite – magnetite composites. Recently, Yuan et al. 21 and Lijing et al.22
reported that this type of composite had high sorption capacity for Cd2+, Zn2+ and Pb2+. They found
that the maximum sorption capacities were 220.77, 140.65 and 598.8 mg g−1 for Cd2+, Zn2+ and Pb2+,
respectively.

In this study, we propose a new method for the synthesis of this type of composite which consists of
Fe3O4 nanoparticles incorporated with hydroxyapatite nanorods (MIA), and investigate its adsorption
behaviours and mechanisms towards Cu(II) and Ni(II) in comparison with Fe3O4 (M). As the authors
know, the use of MIA as an adsorbent in the removal of these heavy metals from aqueous solutions
has yet to be reported.
2. Material and methods

2.1. Materials
Ni(NO3)2.5H2O, CuSO4.5H2O, FeCl2.4H2O, FeCl3.6H2O, NH4OH (25%), CaCl2.2H2O, NH4H2PO4
and ethanol (analytical grade) were purchased from Merck and Sigma-Aldrich. Cu(II) and Ni(II) stock
solutions were prepared from their nitrate and sulphate salts.

2.2. Preparation of nanostructured magnetic adsorbents

In the first step, the preparation of magnetite Fe3O4 nanoparticles was performed according to the
literature 23. The magnetite synthesis is based on the co-precipitation of Fe2+ and Fe3+ ions in an
aqueous ammonia solution, as shown in the equation:

2FeCl3 + FeCl2 + 8NH3 + 4H2O → Fe3O4 + 8NH4Cl (1)

Briefly, 23.3 g FeCl3⋅6H2O and 8.6 g FeCl2⋅4H2O were dissolved in 400 ml of distilled water. The
solution was stirred under a constant nitrogen gas flow until a yellow-red colour was observed.
Subsequently, the mixture was heated up to 80°C and 30 ml of NH 4OH 25% were added to the
solution which immediately became dark brown indicating the formation of the magnetite particles.
The solution was stirred continuously for a further 30 minutes at a temperature of 80°C. After cooling
down to room temperature, the magnetic Fe3O4 nanoparticles were washed repeatedly with distilled
water.

In the second step, hydroxyapatite nanorods were synthesized with the addition of Fe 3O4 nanoparticles
using an in-situ method 5, 23, 24 to obtain the desired composite. 14.7 g CaCl2, 6.9 g NH4H2PO4 and
11.6 ml C2H5OH were added to the solution of the synthesized Fe3O4 in the first step. The molar ratio
of Fe3O4 to hydroxyapatite was 1:4. After stirring for 30 minutes, NH4OH 10% was added to the
mixture until the pH reached the value of 11. The suspension was kept at room temperature for 1 h,
the obtained hydroxyapatite - magnetite composite was separated by a magnet, washed repeatedly
with distilled water, dried at 500C and ground by mortar before use. In order to compare the
adsorption behaviour of the MIA with that of the unmodified material, magnetite Fe3O4 nanoparticles
(M) were also synthesized.

2.3 Adsorbent characterization

The size and morphology of the synthesized M and MIA particles were characterized by a
Transmission Electron Microscope (JEOL JEM-1010) operating at 80 kV. The structure of the
adsorbents was analyzed by X-ray powder diffraction (XRD). XRD patterns were recorded at the
room temperature by a PANalytical X’pert PRO X-ray diffractometer with Cu Kα radiation
(λ=1.5406 Å). The average specific surface area, pore size, and pore volume were determined using a
Micromeritics ASAP 2020 - Accelerated Surface Area and Porosimetry analyzer. The pH was
determined using a WTW InoLab pH 720 pH-meter. The pH point of zero charge (pHPZC) was
determined by the potential titration method 25. The Cu(II) and Ni(II) concentrations were determined
using a flameless graphite furnace atomic adsorption spectrometry (Varian Spectr AAS 880). The
content of the Cu(II) and Ni(II) ions in the MIA after adsorption tests was determined by the EDS
method using a TESCAN VEGA 3 LMU scanning electron microscope equipped with Oxford
Instruments INCA 350 EDS analyser.

2.4 Adsorption experiments

All the adsorption experiments of Cu(II) and Ni(II) were conducted by the batch method, when the
adsorbent materials were placed in a bath of ionic metal solution and the amount of adsorbed metal
was calculated by the measurement of bulk concentration changes in the batches. The effect of pH on
the adsorption of Cu(II) or Ni(II) was studied by mixing 0.1 g adsorbent with 50 ml of 10 mg l -1
Cu(II) or Ni(II) solutions. It was expected that the precipitation of metal hydroxide occurred at
pH>6.3 for Cu(OH)2(s) and pH>8.3 for Ni(OH)2(s) 26. On the other hand, by using NaOH titration, the
result of precipitation titration experiments showed that the 100 mg l-1 solution of Cu(II) and Ni(II)
started to form a suspension of hydroxide precipitation at a pH of 5 and 7, respectively. Therefore, to
ensure that the precipitation process did not occur, the effect of pH on the Cu(II) and Ni(II) adsorption
was studied by adjusting the initial pH of the solution within the range of 3–5 and 3-7 for Cu(II) and
Ni(II), respectively. The pH of the solutions was adjusted and maintained at selected values for 24
hours to ensure equilibrium to be reached.

The kinetics of Cu(II) and Ni(II) adsorption on M and MIA adsorbent were studied by adding the
adsorbent into the solutions of Cu(II) or Ni(II) and measuring their residual concentrations after
various time intervals. Briefly, 1000 ml of Cu(II) or Ni(II) 5 mg l-1 were adjusted to a pH of 5.0 or 7.0
for Cu(II) and Ni(II) by adding HCl (0.1M) or NaOH (0.1M). Then, 0.5 g of adsorbent was added and
stirred to obtain a 0.5 g l-1 suspension. Approximately 10 ml samples containing both solution and
suspension were taken from the reaction vessel in following time intervals: 0.08, 0.16, 0.33, 0.50,
0.75, 1, 2, 3 and 4h of reaction time.
Experiments aiming to determine the equilibrium adsorption isotherm were performed by mixing 0.1
g of adsorbent with 50 ml of solution with initial Cu(II) or Ni(II) concentration ranging from 5 to 50
mg l-1 and 10 -100 mg l-1 for the adsorbent M and MIA, respectively. The pH was monitored in 30
minute intervals and re-adjusted when necessary by the addition of the acid (0.1 mol.l -1 HCl) or the
base (0.1 mol.l-1 NaOH) as required. Generally, the pH did not change significantly after the first
adjustment. All the samples from the above described experiments were filtered through a 0.45 μm
membrane filter and the filtered solutions of Cu(II) or Ni(II) were analyzed to determine the residual
concentration.

The Cu(II) or Ni(II) uptake q (mg g-1) was calculated as

(C0  Ct ).V
q (2)
m

where Co and Ct are the initial and the final concentrations of Cu(II) or Ni(II) (mg L -1), V is the
volume of the solution (L) and m is the mass of the adsorbent (g).

The moles of exchanged calcium can be calculated by the following formular:

CM Ca 2 R .V C M Ca 2 B .V
R  nCa B  
2 2
Ca EXC.  nCa (3)
M Ca M Ca

where nCa2+R and nCa2+B are the number of the moles of released Ca 2+ in the real sample (adsorption
of Cu(II) or Ni(II)) and in distilled water as well as the blank sample (mol.); CM Ca2+R and CM Ca2+B
are the concentrations of released Ca 2+ in real samples (adsorption of Cu(II) or Ni(II)) in distilled
water, and in the blank sample (mg.L-1.); V is the volume of the solution (L); MCa is the atomic weight
of calcium (g mol-1.).

The contribution of Ca2+ exchange mechanism to the total adsorption capacity was calculated as

Ca . M Cu / Ni
 EXC
.100% (4)
qeCu / Ni

where MCu/Ni is the atomic weight of copper or nickel (g mol-1.); qeCu/Ni is Cu(II) or Ni(II) is the
maximum adsorption capacity of MIA.

2.5.3 Adsorption kinetics

To describe the kinetics of the adsorption of Cu(II) and Ni(II) onto the adsorbents, the pseudo first-
order (Eq.3) and the pseudo second-order (Eq.4) kinetic models 5, 48 have been used:
 logqe  qt   log qe 
k1
t (3)
2.303

t 1 t
 2

qt k 2 qe qe (6)

where qt and qe are the amount of adsorbed Cu(II) and Ni(II) per unit weight of adsorbent (mg g−1) at
equilibrium and at time t, respectively; k1 is the rate constant of first-order sorption (min−1) and the
value was calculated from the slope of the linear plot of log (qe–qt) versus t; k2 is the rate constant of
pseudo second-order adsorption and the value was calculated from the slope of the linear plot
of t versus t/qt. In addition, the initial sorption rate, h, can be obtainedfrom the following equation 5:

h = k2qe2 (7)

The slope and intercept values were obtained by least-squares regression to give the value of the
pseudo-second order rate constant. A substantial knowledge of sorption kinetics is very important for
the design and sizing (scale-up) of both batch and continuous adsorbers especially for industrial or
environmental applications.

2.5.4 Adsorption isotherms

The adsorption equilibrium data were analyzed using the Langmuir (Eq. 8) and Freundlich (Eq. 9)
adsorption isotherm models as shown below 49:

Ce 1 C
  e (8)
qe qm K L qm

1
log qe  log K F  log Ce (9)
n

where, qe (mg g-1) is the amount of adsorbed Cu(II) or Ni(II) per weight unit of the adsorbent in
equilibrium, Ce (mg L-1) is equilibrium concentration of Cu(II) or Ni(II) in solution, qm (mg g-1) is the
maximum monolayer adsorption capacity, KF and KL are the Freundlich and Langmuir adsorption
coefficients and n is the Freundlich exponent.

3. Results and discussion

3.1. Properties of prepared adsorbents

The XRD pattern of Fe3O4 nanoparticles exhibits the diffraction lines of Fe3O4 at 18.46°, 30.22°,
35.60°, 43.26°, 53.70°, 57.18°, 62.73° and 74.35° (Fig.1). The TEM image in Fig.2 shows Fe3O4
particles with an average size of 15 nm. However, the particles are not uniform in size, having the
diameter between 6 and 20 nm. The XRD pattern (Fig.1) of the MIA composite adsorbents shows its
hybrid nature as it contains the peaks of the both magnetite nanoparticles (JCPDS reference code: 01-
076-7168) and hydroxyapatite (JCPDS reference code: 04-014-8416). There are five characteristic
diffraction peaks of hydroxyapatite (25.93◦, 32.13◦, 39.80◦, 46.79◦ and 49.52◦) and eight ones
belonging to Fe3O4 (18.21°, 30.23°, 35.57°, 43.24°, 53.51°, 57.20°, 62.84° and 74.28°).

From the TEM images in Fig. 3 it can be concluded that the as-synthesized hydroxyapatite consists of
needle- shaped nanorods having a length of 80 nm and a width of 5 nm. The uniform distribution of
the magnetic Fe3O4 particles over the aggregated hydroxyapatite nanorods can be observed.

The average specific surface area for M and MIA was found to be 72.9 and 101.2 m2 g-1, respectively.
It indicates that hydroxyapatite nanorods contribute to the increase in surface area of nanocomposite
MIA compared with unmodified magnetic Fe3O4. The potentiometric titrations determining the value
of point of zero charges (pHPZC) are shown in Figs.S1 and S2. This parameter is an indicator of the net
surface charge of the adsorbents. Therefore, it may have a strong influence on their adsorption
behaviour in aqueous media at a different pH. Its values for M and MIA are 6.9 and 8.3, respectively.

Fig. 4 displays the EDS spectra of MIA adsorbents before and after being loaded with Cu(II) and
Ni(II), respectively. After the Cu(II) and Ni(II) loading of MIA, copper or nickel spectral lines were
observed. It implies that Cu(II) and Ni(II) had been adsorbed on the surface of MIA successfully.
Moreover, after loading with heavy metal, a increase of calcium concentration could be found. This
phenomenon can be explained by the interchange in the Cu(II) and Ni(II) on the surface.

Hysteresis curves of M and MIA nanoparticles are presented in Fig. 5. No remanence or coercivity
was observed for either material. The Magnetic moment per unit mass of M nanoparticles (80
Am2/kg) decreased to 31 Am2/kg for MIA nanoparticles because of the reduction of ratio Fe3O4 from
100% to 40.4% in M and MIA, respectively. Nevertheless, the latter can still be easily separated from
aqueous solution by use of an external magnetic field in a few minutes.

3.2. Effect of pH and adsorption mechanism

The pH of the aqueous solution plays an important role in adsorption since it influences not only the
distribution of metal species in solution but also the surface properties of the adsorbents. The main
species of Cu(II) in the pH range of 3.0-5.0 include Cu2+, Cu2(OH)22+, Cu(OH)+, Cu2(OH)3+ and
Cu3(OH)42+ 20. At pH <7, the predominant nickel species is Ni2+ 28. As seen in the Fig. 5a, the
percentage of adsorbed Cu(II) and Ni(II) on the adsorbents M and MIA almost increased with the
increase in solution pH, indicating that solution pH has a strong influence on the adsorption. This
result is similar to that obtained in previous studies of the adsorption behaviour of Ni(II), Cu(II) on
kaolinite and montmorillonite 29, Co, Ni, Cu, and Zn on hydrous manganese dioxide 30, Cd on
crystallite hydroxyapatite 14 and Co on hydroxyapatite 31. The adsorption efficiency of MIA and M for
Ni(II) removal shows significant difference comparing to that for Cu(II). The maximum adsorption
capacity of both adsorbents for Cu(II) and Ni(II) was obtained at pH 5.0 and 7.0, respectively. Thus,
the optimum pH for the removal of Cu (II) and Ni(II) by adsorption were fixed at 5.0 and 7.0 for the
next steps of the adsorption experiments. The pH value is similar to the one which was previously
used in the adsorption of Cu(II) and Ni(II) onto iron oxide-coated sand from aqueous solution 32.

The sorption mechanisms of the heavy metals are diverse and mainly include: ion exchange,
formation of surface complexs, dissolution/precipitation, electrostatic and Columbic intereaction 13, 33,
34
. It is often not easy to distinguish among mechanisms and impossible to explain the adsorption
process by a single mechanism 35, 36. In this case, there were two types of adsorbents, M (Fe3O4
nanoparticles) and MIA in which Fe3O4 was bonded with hydroxyapatite nanorods forming
heterogeneous surface of the composite. To explain the accumulation of cationic heavy metals on
adsorbents based on magnetite and hydroxyapatite, the four possible mechanisms, including surface
complexation, ion exchange, dissolution,precipitation and electrostatic have been proposed 37, 38.

3.2.1 Iner-sphere surface complexation

The first mechanism involves the formation of inner-sphere surface complexes between Cu(II) and
Ni(II) and deprotonated surface hydroxyls of Fe3O4 and hydroxyapatite in M and MIA. These types of
surface complex formation equilibria are summarized schematically as follows 35, 36:

S − OH + M2+ → S − O − M+ + H+ (10)

S − OH + M2+ → (S − O)2 − M + 2H+ (11)

39
where S stands for surface of Fe3O4 and hydroxyapatite 38, M2+ stands for Cu(II) or Ni(II) ions.
According to the Le Chatelier's principle and schematic reactions (3,4), the effect of pH can be
described clearly. At high pH values, the decrease in H+ concentration caused the cation exchange
between H+ and Cu(II) and Ni(II) cations to form surface complexes. Similarly, this mechanism was
also modelled as the main uptake mechanisms on the surface adsorption of Co 2+ ions on Fe3O4
nanoparticles 39 and hydroxyapatite 38. The results of SEM-EDS and FTIR on removal Cu(II) and
Ni(II) from aqueous solution by synthetic hydroxyapatite 40, 41 the literature also supports the
contribution of surface complexation to the metal retention by hydroxyapatite. This mechanism has
been determined as the initial rapid step while the subsequent slower step has been related to ionic
exchange with calcium ions 40.

3.2.2 Calcium exchange mechanism

Generally, hydroxyapatite selectivity towards divalent metal cations is the result of an ion-exchange
process with Ca2+cations in the hydroxyapatite lattice 14, 34. Hydroxypatite surfaces provide the
nucleation sites for the newly crystallizing phase 33. The ionic radius of copper (0.72 Å) and nickel
(0.69 Å) 42 differ from that of Ca2+ (0.99 Å) only slightly, and therefore they can substitute Ca2+ in the
hydroxyapatite crystal lattice. The reaction mechanism corresponds to equimolar exchange of Cu(II)
or Ni(II) and calcium yielding Ca 10 − xCux(PO4)6(OH)2 or Ca10 − xNix(PO4)6(OH)2 where x can vary from
0 to 10 depending on the reaction time and experimental conditions. In general, Cu(II) or Ni(II)
cations are firstly adsorbed onto the nano-hydroxyapatite surface and then substitute the Ca 2+
according to the following equation:

Ca10(PO4)6(OH)2 + xM2+ → Ca10−xMx(PO4)6(OH)2 + xCa2+ (12)

The EDS spectra show the lowering of the Ca 2+ peak intensity of MIA – Cu(II), and MIA – Ni(II) in
comparison with that in MIA. In addition, copper peaks and nickel peaks could be observed. The
chemical analysis by EDS (Fig.4) showed that the amount calcium decreased from 16.51 % (MIA) to
14.09 % (MIA-Cu(II)) and 15.64% (MIA-Ni(II)), while the concentration of copper and nickel were
determined to be 5.35 % and 1.71 % in the samples after adsorbing. This phenomenon indicates that
copper and nickel had been adsorbed on the surface of MIA successfully. The adsorption occurred
with the participation of cationic exchange mechanism between Cu(II) or Ni(II) and Ca2+on the
surface of MIA adsorbents. In order to confirm this supposition, some additional experiments were
carried out in order to determine and to compare the concentration of Ca2+ released in the solution
after the completion of adsorption process (0.3 g MIA in 200 ml of the solution containing 300 mg l -1
Cu(II) or Ni(II)) and in blank sample (MIA in distilled water). The results showed that the
concentration of Ca2+ in the solution increased remarkably after heavy metal adsorption compared
with the blank sample (0.43, 27.0 and 15.5 mg l-1 for blank sample, MIA-Cu (II) and MIA-Ni(II),
respectively). Therefore, it can be concluded that there was a cation exchange mechanism in the
adsorption process. Similarly, according to the spectroscopy result (XPS, XRD) in 15, 43, Cu(II) or
Ni(II) substitution Ca2+ in the hydroxyapatite crystal lattice has been demonstrated.

In addition, the data in Fig.6 and Table 1 show that the number of moles (n) of released Ca 2+ is
smaller than the number of moles of adsorbed Cu(II) and Ni(II). They are equal to 51.1% and 61.9 %
of total Cu(II) and Ni(II) removed. Thus there were other mechanisms which also contributed to the
removal of Cu(II) and Ni(II).

3.2.3 Dissolution and precipitation mechanism

Beside the above mechanisms, the third mechanism involves hydroxyapatite dissolution, followed by
phosphate reaction with dissolved Cu(II) or Ni(II) and precipitation. Hydroxypatite acts as a
phosphate provider in solutions where the present metals can precipitate 33. In the similar reported
studies on the removal mechanism of copper by hydroxyapatite, phosphate rocks, and synthetic
hydroxyapatite 34, 44, XRD analysis performed on synthetic hydroxyapatite samples after Cu 2+ and Ni2+
uptake showed a reduction of hydroxyapatite peak intensity and the appearance of peaks attributed
to Cu3(PO4)2, libethenite, Cu3(PO4)(OH) and Ni3(PO4)2. In the present study, the new phosphate
substances did not occur in XRD patterns of the samples of MIA after the adsorption of Cu(II) and
Ni(II). It can be suggested that the third mechanism may exist resulting in the formation of low
concentration of new amorphous phosphate substances 45 that do not show diffraction lines. However,
the third mechanism does not play an important role in the overall adsorption process.

3.2.4 The electrostatic mechanism

Magnetite and synthetic hydroxyapatite are amphoteric solids. Therefore, their reactive surface sites
exhibit buffering properties in solution 46, 47:

≡ FeOH + H+ ↔ ≡FeOH2+ (13)

≡ FeOH ↔ ≡FeO- + H+ (14)

≡ PO- + H+ ↔ ≡POH0 (15)

≡ CaOH2+ ↔ ≡CaOH0 + H+ (16)

The value of pHPZC of the M and MIA were determined to be 6.9 and 8.3. At pH lower than 5 and 7,
the positively charged FeOH 2+, ≡ CaOH2+ and neutral ≡POH0 sites prevail on M and MIA surfaces,
making their surface charges in this pH region positive which is unfavourable for the adsorption of
Cu(II) and Ni(II). Therefore, we suggest that if an electrostatic mechanism existed, it did not play
important role in adsorption process and contributed in unremarkable percentage to the total of
adsorption capacity.

3.3. Adsorption Kinetic

The adsorption kinetic parameters obtained from the evaluation of experimental data are summarized
in Table 2. The plots of log (qe–qt) and t/qt versus t for the first- and second-order model are shown in
Figures 7, 8. Based on correlation (R2) given in Table 2 as well as on the visual assessment of the
plots in Figs. 7 and 8, it can be concluded that the pseudo second-order kinetic model is appropriate to
describe the adsorption of Cu(II) and Ni(II) on two adsorbents. Moreover, it can be seen from Table 1
that the experimental adsorption capacities, qe,exp are in good agreement with the theoretical
adsorption capacities (qe,cal) calculated by pseudo second-order equation (11).
The initial adsorption rates (h) of Cu(II) and Ni(II) species onto the MIA composite adsorbent are
2.115 mg g−1 min−1 and 1.479 mg g−1 min−1, which are significantly higher than that of M (0.919 and
0.506 mg g−1 min−1). Therefore the hydroxyapatite nanorods in the MIA composite have a higher
adsorption rate for Cu(II) and Ni(II) than that on iron oxides nanoparticles M.

3.4 Adsorption isotherm

The adsorption isotherms for Cu(II) and Ni(II) onto M and MIA and their Langmuir and Freundlich
adsorption isotherm plots are shown in Fig.9 and Fig.10.

The corresponding values of the constants appearing in the Langmuir and the Freundlich equations
are summarized in Table 3. The correlation coefficients (R2) obtained from the Langmuir and
Freundlich model are 0.998, 0.969 for Cu (II) adsorption and 0.997, 0.935 for Ni(II) adsorption on M,
respectively. These results show that the Langmuir model gives a better fit of the experimental data
than the Freundlich model for the adsorbent M. It would indicate that the surface of the Fe3O4
nanoparticles (M) is energetically homogeneous, the adsorbed species are independent of each other
(no associative adsorption) and they form a monolayer.

The correlation coefficients (R2) obtained from the Langmuir and Freundlich model are 0.992, 0.995
for Cu (II) adsorption and 0.989, 0.994 for Ni(II) adsorption on MIA, respectively.
This result imply that the Cu (II) and Ni(II) adsorption on MIA can be described reasonably well by
either the Langmuir or Freundlich isotherm models. However, the experimental values fit slightly
better to the Freundlich model than to the Langmuir one in comparison with that in M adsorbent. The
adsorption behaviour may result from the adsorption mechanism on the heterogeneous surface of the
composite material MIA which consists of both hydroxyapatite nanorods and Fe3O4 nanoparticles.
The Langmuir monolayer adsorption capacity (qm, mg g−1) value toward Cu(II) and Ni(II) obtained by
the linear method of analysis was 6.19, 48.78 mg g−1 and 5.86, 29.07 mg g−1 for M and MIA,
respectively.

The higher adsorption capacity of MIA compared to M can be explained by the enhancement in
specific surface area and metallic affinity due to small size and chemical structure, although its
pHPZC is higher than that of M which is unfavourable for adsorption process. It can be concluded that
the hydroxyapatite nanorods play an important role in enhancement of Cu(II) and Ni(II) removal
efficiency of the carrier material - magnetic Fe3O4 nanoparticles. In addition, from Fig S3 and Fig. 11
it can be seen that the existence of a magnetic core facilitates the application of MIA in a magnetic
separation system that is becoming widely used in water and wastewater treatment.
As the qm value of the adsorbent varies with the experimental conditions, direct comparison with
literature values is not always possible. Nevertheless, from the adsorption data of other adsorbents in
Table 4, we can conclude that MIA is the promising adsorbent for Cu(II) and Ni(II) removal from
aqueous solution.

4. Conclusion
A novel composite adsorbent has been prepared by a simple in-situ method, and its adsorption
behaviour and mechanisms toward Cu (II) and Ni(II) have been investigated. The composite
adsorbent, which consists of hydroxyapatite nanorods, had a high surface area of 101.2 m2 g-1 and
adsorptive affinity contributing to the high removal capacity. Ca2+ exchange and surface complexation
were proved to be the dominant mechanisms for the adsorption of the toxic divalent heavy metals.
Although the composite material precipitates out of the reaction mixture, due to the nature and size of
the material, separtion is difficult and not effiecient, the magnetic separtion properties and the ease of
separtion using a magnetic field, makes it a fast and effiecient process that can be easily used. The
high uptake capability and magnetic-separation property of the new composite adsorbent makes it
potentially attractive for the removal of Cu (II) and Ni(II) in wastewater treatment, especially for the
treatment of toxic effluent from metal finishing process.

Acknowledgments

The publication is a result of the project Development of the UniCRE Centre (project code
LO1606) – “Green synthesis”, which was financially supported by the Ministry of Education,
Youth and Sports of the Czech Republic under the National Programme for Sustainability.

References

1. Abe M, Usuda K, Hayashi S, Ogawa I, Furukawa S, Igarashi M and Nakae D, Carcinogenic

risk of copper gluconate evaluated by a rat medium-term liver carcinogenicity bioassay
protocol. Archives of Toxicology 82: 563-571 (2008).

2. Genç-Fuhrman H, Wu P, Zhou Y and Ledin A, Removal of As, Cd, Cr, Cu, Ni and Zn from
polluted water using an iron based sorbent. Desalination 226: 357-370 (2008).

3. Zeledón-Toruño Z, Lao-Luque C and Solé-Sardans M, Nickel and copper removal from

aqueous solution by an immature coal (leonardite): effect of pH, contact time and water
hardness. Journal of Chemical Technology & Biotechnology 80: 649-656 (2005).

4. Fu F and Wang Q, Removal of heavy metal ions from wastewaters: A review. Journal of
Environmental Management 92: 407-418 (2011).
5. Thanh DN, Singh M, Ulbrich P, Štěpánek F and Strnadová N, As(V) removal from aqueous
media using α-MnO2 nanorods-impregnated laterite composite adsorbents. Materials
Research Bulletin 47: 42-50 (2012).

6. Mohan D and Pittman JCU, Arsenic removal from water/wastewater using adsorbents--A
critical review. Journal of Hazardous Materials 142: 1-53 (2007).

7. Vieira MGA, Neto AFA, Gimenes ML and da Silva MGC, Removal of nickel on Bofe
bentonite calcined clay in porous bed. Journal of Hazardous Materials 176: 109-118 (2010).

8. Garg UK, Kaur MP, Garg VK and Sud D, Removal of Nickel(II) from aqueous solution by
adsorption on agricultural waste biomass using a response surface methodological approach.
Bioresource Technology 99: 1325-1331 (2008).

9. Dang SV, Kawasaki J, Abella LC, Auresenia J, Habaki H, Gaspillo P-aD and Kosuge H,
Removal of arsenic from synthetic groundwater by adsorption using the combination of
laterite and iron-modified activated carbon. Journal of Water and Environment Technology 6:
43-54 (2008).

10. Conca JLW, Judith, An Apatite II permeable reactive barrier to remediate groundwater
containing Zn, Pb and Cd. Applied Geochemistry 21: 2188-1300 (2006).

11. Conca JL and Wright J, An Apatite II permeable reactive barrier to remediate groundwater
containing Zn, Pb and Cd. Applied Geochemistry 21: 1288-1300 (2006).

12. Admassu W and Breese T, Feasibility of using natural fishbone apatite as a substitute for
hydroxyapatite in remediating aqueous heavy metals. Journal of Hazardous Materials 69:
187-196 (1999).

13. Krestou A, Xenidis A and Panias D, Mechanism of aqueous uranium(VI) uptake by

hydroxyapatite. Minerals Engineering 17: 373-381 (2004).

14. Mobasherpour I, Salahi E and Pazouki M, Removal of divalent cadmium cations by means of
synthetic nano crystallite hydroxyapatite. Desalination 266: 142-148 (2011).

15. Mobasherpour I, Salahi E and Pazouki M, Removal of nickel (II) from aqueous solutions by
using nano-crystalline calcium hydroxyapatite. Journal of Saudi Chemical Society 15: 105-
112 (2011).

16. Yadanaparthi SKR, Graybill D and von Wandruszka R, Adsorbents for the removal of
arsenic, cadmium, and lead from contaminated waters. Journal of Hazardous Materials 171:
1-15 (2009).
17. Liu G, Talley JW, Na C, Larson SL and Wolfe LG, Copper Doping Improves Hydroxyapatite
Sorption for Arsenate in Simulated Groundwaters. Environmental Science & Technology 44:
1366-1372 (2010).

18. Tuutijärvi T, Lu J, Sillanpää M and Chen G, As(V) adsorption on maghemite nanoparticles.

Journal of Hazardous Materials 166: 1415-1420 (2009).

19. Tran HV, Tran LD and Nguyen TN, Preparation of chitosan/magnetite composite beads and
their application for removal of Pb(II) and Ni(II) from aqueous solution. Materials Science
and Engineering: C 30: 304-310 (2010).

20. Wang XS, Zhu L and Lu HJ, Surface chemical properties and adsorption of Cu (II) on
nanoscale magnetite in aqueous solutions. Desalination 276: 154-160 (2011).

21. Feng Y, Gong J-L, Zeng G-M, Niu Q-Y, Zhang H-Y, Niu C-G, Deng J-H and Yan M,
Adsorption of Cd (II) and Zn (II) from aqueous solutions using magnetic hydroxyapatite
nanoparticles as adsorbents. Chemical Engineering Journal 162: 487-494 (2010).

22. Dong L, Zhu Z, Qiu Y and Zhao J, Removal of lead from aqueous solution by
hydroxyapatite/magnetite composite adsorbent. Chemical Engineering Journal 165: 827-834
(2010).

23. Panella B, Vargas A and Baiker A, Magnetically separable Pt catalyst for asymmetric
hydrogenation. Journal of Catalysis 261: 88-93 (2009).

24. Mobasherpour I, Heshajin MS, Kazemzadeh A and Zakeri M, Synthesis of nanocrystalline

hydroxyapatite by using precipitation method. Journal of Alloys and Compounds 430: 330-
333 (2007).

25. Marczewski AW, Adsorptive and Colloidal Properties of Soil Fractions, in 15 th International
Congress of Chemical and Process Engineering CHISA 2002, Prague, Czech Republic, pp.
25-29 (2002).

26. Baes CF and Mesmer RE, The hydrolysis of cations. Krieger, Malabar (Florida) (1986).

27. Chang Y-C and Chen D-H, Preparation and adsorption properties of monodisperse chitosan-
bound Fe3O4 magnetic nanoparticles for removal of Cu(II) ions. Journal of Colloid and
Interface Science 283: 446-451 (2005).

28. Ren Y, Yan N, Wen Q, Fan Z, Wei T, Zhang M and Ma J, Graphene/δ-MnO2 composite as
adsorbent for the removal of nickel ions from wastewater. Chemical Engineering Journal
175: 1-7 (2011).
29. Bhattacharyya KG and Gupta SS, Influence of acid activation on adsorption of Ni(II) and
Cu(II) on kaolinite and montmorillonite: Kinetic and thermodynamic study. Chemical
Engineering Journal 136: 1-13 (2008).

30. Kanungo SB, Tripathy SS and Rajeev, Adsorption of Co, Ni, Cu, and Zn on hydrous
manganese dioxide from complex electrolyte solutions resembling sea water in major ion
content. Journal of Colloid and Interface Science 269: 1-10 (2004).

31. Huang Y, Chen L and Wang H, Removal of Co(II) from aqueous solution by using
hydroxyapatite. Journal of Radioanalytical and Nuclear Chemistry 291: 777-785 (2012).

32. Boujelben N, Bouzid J and Elouear Z, Adsorption of nickel and copper onto natural iron
oxide-coated sand from aqueous solutions: Study in single and binary systems. Journal of
Hazardous Materials 163: 376-382 (2009).

33. Valsami-Jones E, Ragnarsdottir KV, Putnis A, Bosbach D, Kemp AJ and Cressey G, The
dissolution of apatite in the presence of aqueous metal cations at pH 2–7. Chemical Geology
151: 215-233 (1998).

34. Fanny M-R and Michel F, Sorption of Inorganic Species on Apatites from Aqueous Solutions,
in Encyclopedia of Surface and Colloid Science, Second Edition. Taylor & Francis, pp. 5700-
5725 (2007).

35. Stumm W, Aquatic surface chemistry: chemical processes at the particle-water interface.
Wiley (1987).

36. Pitter P, Hydrochemistry. Institute of chemical technology Prague (VŠCHT Praha), pp. 579
(2009).

37. Miretzky P and Fernandez-Cirelli A, Phosphates for Pb immobilization in soils: a review.

Environmental Chemistry Letters 6: 121-133 (2008).

38. Smičiklas I, Dimović S, Plećaš I and Mitrić M, Removal of Co2+ from aqueous solutions by
hydroxyapatite. Water Research 40: 2267-2274 (2006).

39. Uheida A, Salazar-Alvarez G, Björkman E, Yu Z and Muhammed M, Fe3O4 and γ-Fe2O3

nanoparticles for the adsorption of Co2+ from aqueous solution. Journal of Colloid and
Interface Science 298: 501-507 (2006).

40. Fernane F, Mecherri MO, Sharrock P, Fiallo M and Sipos R, Hydroxyapatite interactions with
copper complexes. Materials Science and Engineering: C 30: 1060-1064 (2010).
41. Corami A, D’Acapito F, Mignardi S and Ferrini V, Removal of Cu from aqueous solutions by
synthetic hydroxyapatite: EXAFS investigation. Materials Science and Engineering: B 149:
209-213 (2008).

42. Patil DR and Chougule BK, Effect of copper substitution on electrical and magnetic
properties of NiFe2O4 ferrite. Materials Chemistry and Physics 117: 35-40 (2009).

43. Choudary BM, Sridhar C, Kantam ML and Sreedhar B, Hydroxyapatite supported copper
catalyst for effective three-component coupling. Tetrahedron Letters 45: 7319-7321 (2004).

44. Mavropoulos E, Rocha NCCd, Moreira JC, Bertolino LC and Rossi AM, Pb 2+, Cu 2+and Cd
2+ ions uptake by Brazilian phosphate rocks. Journal of the Brazilian Chemical Society 16:
62-68 (2005).

45. Oliva J, De Pablo J, Cortina J-L, Cama J and Ayora C, Removal of cadmium, copper, nickel,
cobalt and mercury from water by Apatite II™: Column experiments. Journal of Hazardous
Materials 194: 312-323 (2011).

46. Wu L, Forsling W and Schindler PW, Surface complexation of calcium minerals in aqueous
solution: 1. Surface protonation at fluorapatite—water interfaces. Journal of Colloid and
Interface Science 147: 178-185 (1991).

47. Hajdú A, Tombácz E, Illés E, Bica D and Vékás L, Magnetite Nanoparticles Stabilized Under
Physiological Conditions for Biomedical Application Colloids for Nano- and Biotechnology,
ed by Hórvölgyi Z and Kiss É. Springer Berlin / Heidelberg, pp. 29-37 (2008).

48. Ho YS, Ng JCY and McKay G, Removal of Lead (II) from effluents by sorption on peat using
second-order kinetics. Separation Science and Technology 36: 241 - 261 (2001).

49. Nguyen Thanh D, Singh M, Ulbrich P, Strnadova N and Štěpánek F, Perlite incorporating γ-
Fe2O3 and α-MnO2 nanomaterials: Preparation and evaluation of a new adsorbent for As(V)
removal. Separation and Purification Technology 82: 93-101 (2011).

Figure captions

Fig.1. The X-ray diffraction patterns of M and MIA

Fig. 2. TEM image of the magnetic Fe3O4 nanoparticles (M)

Fig. 3. TEM images of the magnetite-hydroxyapatite composite (MIA): a) overview of the particles’
cluster, b) detail showing the morphology of the components

Fig. 4. EDS spectra of the samples: (a) MIA, (b) MIA –Cu(II): after adsorbing Cu(II), (c) and MIA –
Ni(II)

Fig. 5. Effect of the solution pH on Cu(II) (a) and Ni(II) (b) adsorption onto M and MIA.

Fig. 6. The plot of the relation between adsorbed Cu(II) and Ni(II) and exchanged Ca 2+
(n - number of moles)

Fig. 7. Kinetic experiment (a) and second – order kinetic model plot for adsorption of Cu(II) on M
and MIA

Fig. 8. Kinetic experiment (a) and second – order (b) kinetic model plot for adsorption of Ni(II) on M
and MIA

Fig. 9. Langmuir (a) and Freundlich (b) adsorption isotherm plot for adsorption of Cu(II) on M and
MIA

Fig. 10. Langmuir (a) and Freundlich (b) adsorption isotherm plot for adsorption of Ni(II) on M and
MIA

Fig. 11. Schematic explanation of magnetic separation process in wastewater treatment

Table captions

Table 1

The relation between the adsorbed Cu(II) and Ni(II) and the exchanged Ca 2+ cations

Table 2

Comparison of the pseudo first- and second-order adsorption rate constants for Cu(II) and Ni(II)

Table 3

Langmuir and Freundlich adsorption isotherm constants for Cu(II) and Ni(II) adsorption on M and
MIA.

Table 4
Comparison of adsorption capacities for Cu(II) and Ni(II) by various adsorbents
 100
a)

80

% Cu(II) removed
60

40

20
M
MIA

0
2 3 4 5 6
pH

100
b)

80
% Ni(II) removed

60

40

20
M
MIA

0
1 2 3 4 5 6 7 8 9 10
pH
 0.30
nCa2+ in distiled water
0.258 n adsorbed Cu2+
0.25
Number of moles (mol.)

nCa2+ exchanged for Cu2+

n adsorbed Ni2+
nCa2+ exchanged for Ni2+
0.20

0.15 0.132
0.121

0.10
0.075

0.05
0.01

0.00
 50
11 b)
a)
2
10 R = 0.998
40
9
30
8
qt(mg. g )

t/qt
2
-1

R = 0.999
7
20
6

5 10
M
M
MIA
4 MIA
0
0 50 100 150 200 250 300 0 50 100 150 200 250

time (min.) time (min.)

 8 60
a) b)
50
6 2
R = 0.995
40

t/qt
qt(mg. g )
-1

4 30

2
20 R = 0.999
2
M 10
M
MIA
MIA
0 0
0 50 100 150 200 250 300 0 50 100 150 200 250
time (min.)
time (min.)
 7 2.0
a)
b)
6 M 2
MIA 2
R = 0.804
R =0.998 1.5

Log qe
5
Ce/qe (g L )

4
-1

1.0 2
R = 0.969
3

2 0.5
2
R =0.990
1 MIA
M
0 0.0
0 10 20 30 40 50 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
-1
Ce (mg L ) Log Ce
 9 1.5
a) b)
8 2
R = 0.997 2
R = 0.992
7
6 1.0
Ce/qe

Log qe
2
R = 0.918
4 2
R = 0.989
3 0.5
2
M MIA
1 MIA M
0 0.0
0 10 20 30 40 50 60 70 80 0.0 0.5 1.0 1.5 2.0
-1
Ce (mg g ) Log Ce
 Table 1 The relation between the removed Cu(II) and Ni(II) and the exchanged Ca2+

Parameter Blank sample Cu(II) adsorption Ni(II) adsorption

Residual metal concentration (mg L -1) - 218.0 264.5

Concentration of the released Ca2+ (mg L-1) 0.43 27.0 15.5

Number of moles of adsorbed metal (mol.) - 0.258 0.121

Number of moles exchanged Ca2+ (mol.) 0.011 0.132 0.075

Contribution of Ca2+ exchange mech. (%) - 51.1 61.9

Table 2 Comparison of the pseudo first- and second-order adsorption rate constants of Cu(II)

and Ni(II) removal.

Metal Adsorbent P.first-order kinetic model P.second-order kinetic model

qe,exp qe,cal k1 qe,cal h

2
R R2
(mg g -1) (mg g-1) (min-1) (mg g-1) (mg g-1 min-1)

M 5.75 46.21 0.00021 0.821 5.80 0.919 0.998

Cu(II)
MIA 10.31 42.80 0.00041 0.538 10.58 2.115 0.999

M 4.46 47.78 0.00020 0.565 4.66 0.506 0.995

Ni(II)
MIA 7.14 44.76 0.00004 0.729 7.35 1.479 0.999
Table 3 Langmuir and Freundlich adsorption isotherm constants for Cu(II) and Ni(II)

adsorption on M and MIA.

Metal Adsorbent Langmuir constants Freundlich constants

qm KL KF
2
R n R2
mg g-1 (L mg-1) (mg. g-1)(L. mg-1)1/n

M 6.19 0.829 0.998 5.476 3.349 0.969

Cu(II)
MIA 48.78 0.156 0.990 7.794 3.4657 0.804

M 5.86 0.211 0.997 3.050 1.696 0.918

Ni(II)
MIA 29.07 0.111 0.989 3.177 6.878 0.992
Table 4 Comparison of adsorption capacities for Cu(II) and Ni(II) by various adsorbents

Adsorption capacity

Adsorbent (mg g-1) Reference

Cu(II) Ni(II)

Iron-coated sand 2.04 - [31]

Nanohydroxyapatite/chitosan composite 26.11 - [49]

Nano-hydroxyapatite/chitin composite 21.45 - [49]

Gum arabic modified magnetic nano-adsorbent 38.5 - [50]

Amino-functionalized magnetic nanosorbent 25.77 - [51]

Chitosan-bound Fe3O4 21.5 - [26]

Fe3O4 incorporated with hydroxyapatite (MIA) 48.78 - Present work

Iron-coated sand - 1.0 [31]

ZrO–kaolinite - 8.8 [52]

ZrO-montmorillonite - 22.0 [52]

Graphene/ δ-MnO2 - 46.6 [53]

Oxidized multi-walled carbon nanotubes - 7.61 [54]

Fe3O4 incorporated with hydroxyapatite (MIA) - 29.07 Present work