Chemosphere 164 (2016) 32e40

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Adsorption of arsenic, phosphorus and chromium by bismuth

impregnated biochar: Adsorption mechanism and depleted adsorbent
utilization
Ningyuan Zhu a, b, Tingmei Yan a, *, Jun Qiao a, Honglei Cao a, b
a
Institute of Soil Science, Chinese Academy of Sciences, No. 71 East Beijing Road, Nanjing 210008, China
b
Graduate School of the Chinese Academy of Sciences, Beijing 100039, China

h i g h l i g h t s

Bismuth activated carbons derived from wheat straw were fabricated.

Bismuth particles grown within biochar matrix.

BiBC500 showed high adsorption capacity to arsenic, phosphorus and chromium.

Adsorption mechanisms of arsenic, phosphorous and chromium were illustrated.

Phosphate depleted material could photodegrade dye pollutant.

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

Article history: Bismuth impregnated biochar were synthesized to deal with wastewater pollution. Nitrogen adsorption-
Received 20 April 2016 desorption isotherms, scanning electron microscopy (SEM), Fourier transform infrared spectroscopy
Received in revised form (FTIR), X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) were used to determine the
25 July 2016
characteristics of adsorbents and explore the main adsorption mechanism. Results showed that bismuth
Accepted 7 August 2016
Available online 28 August 2016
particle was carried successfully within the biochar matrix, making contributions to creating micropore
and boost specific surface area. The loaded bismuth, served as the adsorption site, rather than the specific
Handling Editor: X. Cao surface area played an important role in arsenic and phosphorus adsorption. Batch adsorption experi-
ments demonstrated a fit Langmuir model for arsenic (As) and phosphorus (P) and a suitable Freundlich
Keywords: model for chromium (Cr). Thermodynamic parameters depicted the endothermic nature and the
Bismuth spontaneous process for phosphate and arsenic adsorption. Besides, this contaminant-loaded carbon
Activated carbon adsorbent was further applied for the removal of methylene blue from aqueous solution.
Adsorption mechanism © 2016 Elsevier Ltd. All rights reserved.
Depleted adsorbents

1. Introduction
The continuous growth of population, increasing reinforcement
of agricultural activities and rapid development of urbanization and
industrialization worsen water quality of our water resources
which is the most essential component for life (Ali, 2012). The
excessive input of phosphorus in surface water worsened eutro-
phication and algae blooms as well as deteriorated the quality of
the groundwater severely. Heavy metals, especially the chromium
and arsenic, are more problematic and threatening to ecological
* Corresponding author.
E-mail address: tmyan@issas.ac.cn (T. Yan).
environment and human beings because of their high toxicity, non-
biodegradation and accumulation through food chain. Thus, the
World Health Organization recommended that the maximum
permissible limit for arsenic should be 10 mg L1 while the
maximum limit for chromium be 0.05 mg L1 in drinking water (Liu
et al., 2008; Tang et al., 2014; Wang et al., 2010).
Therefore, it is very imperative to deal with contaminated
wastewater prior to its discharge into soil and water environment.
Many researchers have explored various methods such as chemical
precipitation, gravity separation, solvent extraction, reverse
osmosis, ion-exchange, electrocoagulation and electrodialysis,
flotation and adsorption for removing phosphate or heavy mental
irons from sewage system (Hua et al., 2012). For instance, Afkhami
et al. synthesized DNPH (nano-alumina modified with 2,4-

http://dx.doi.org/10.1016/j.chemosphere.2016.08.036
0045-6535/© 2016 Elsevier Ltd. All rights reserved.
 N. Zhu et al. / Chemosphere 164 (2016) 32e40
dinitrophenylhydrazine) with a satisfying adsorption capacity to-
ward Pb (II), Cr (III) and Cd (II) ions in a multiple-metal solution,
indicating that DNPH is a promising adsorbent for heavy mental
treatment (Afkhami et al., 2010). Nevertheless, few researches
developed effective, efficient, economical and eco-friendly mate-
rials or methods for reduction of arsenic, phosphate and chromium
pollutant which are more difficult to be removed because of their
anion form (Alemayehu et al., 2011; Onnby et al., 2014). In addition,
the removal efficiency of As (III) is much lower than As (V) for
adsorption process was unable to act on the uncharged form of As
(III) especially at environmentally relevant pH (Sun et al., 2013).
But, arsenite is more toxic, soluble, and mobile than arsenate.
Fortunately, metal salt, metal oxides and hydrous metal oxides
such as granular ferric hydroxide (Banerjee et al., 2008), TiO2 (Liu
et al., 2008), akaganeite (Lazaridis et al., 2005), and Na2SO3/FeSO4
(Pan et al., 2014) have been well explored for arsenic, phosphorus
or chromium treatment. Adsorption is recognized as one of the
most available option because of its low cost and high efficiency.
However, handling and disposal of the waste sludge is still a
cumbersome problem limiting their actual industrial application.
Furthermore, there is still lack of economical way to transform
more agricultural residues like straw resources to valuable and
green by-product whether in production sites or in factory for their
high-fibre content and low-protein quality. After harvest season,
farmers burn crop stalk in great quantities, causing serious envi-
ronmental problems such as haze which may pose a serious threat
to public health.
Biochar, originated from biomass pyrolysis under oxygen-
limited environment can be used as soil amendment to increase
soil fertility, double or triple plant yields, fix carbon and cut
greenhouse gas emissions in a vast scale. Besides, the chemical
modified biochar also adsorb contaminants either by electrostati-
cally attractive forces or by ligand exchange mechanism without
the adverse waste sludge production (Loganathan et al., 2013;
Marris, 2006). Thus, application of the adsorption methods with
biochar based adsorbent is a promising way for arsenic, phosphate
and chromium treatment. Herein, to enhance the adsorption ability
of biochar and probe into the rational straw utilization of straw
resources, bismuth oxide and wheat straw were explored to pre-
pare a cost-effective but high efficiency material for wastewater
treatment for the first time. The specific objectives of this work
were to: (1) prepare and characterize Bi2O3 impregnated biochar
composites, (2) test its sorption capacities for As (III), P and Cr (VI),
and (3) investigate the possible mechanisms involved in the As (III),
P and Cr (VI) adsorption.
2. Materials and methods
2.1. Preparation of adsorbents
Bismuth oxide solution (Bi-solution) was prepared by adding
10 mmol Bi2O3 to 20 ml hydrochloric acid (0.12 mol) and then
diluted to 100 ml with distilled water. A stock solution of
3000 mg L1 KH2PO4, 200 mg L1 sodium arsenite and 200 mg L1
potassium dichromate were prepared in a volumetric flask and
diluted to the required concentrations (60e1800 mg L1). Bi2O3,
KH2PO4, NaAsO2, K2Cr2O7 and other reagents used in this work
were all at analytical grade.
Wheat straw (WS) was obtained from Wuxi city, Jiangsu prov-
ince, China. It was milled and sized into particle with diameters
between 0.60 and 0.80 mm. Bismuth oxide activated carbon (BiBC)
and control biochar (CBC) were produced as follow: 10 g WS was
mixed with 100 ml Bi-solution and hydrochloric acid (0.12 mol)
respectively, stirred vigorously at 80  C for 3 h, exposed to ultra-
sonic treatment to get the targeted particles as small as possible
33
and then dried at 105  C. The biochar precursors were then heated
to setting temperature (400, 500 and 600  C) and finally main-
tained for 60 min in a furnace with the heating rate of 10  C min1
under nitrogen flow. Finally, the biochar materials were washed
with 0.01 M NaHCO3 solution and distilled water for three times.
Activated carbons were labelled as BiBC400, BiBC500 and BiBC600
while the control labelled as CBC400, CBC500 and CBC600.
2.2. Characterization of adsorbents
Porosity and surface characteristics were measured by N2
(0.162 nm2) adsorption using a NOVA-2000E (Quantachrome, USA)
surface area analyzer. Brunauer-Emmertt-Teller (BET) surface area
and average pore diameter of the BiBCs and CBCs were determined
by multipoint BET analysis of adsorption data points with relative
pressures of 0.05e0.3. Surface functional groups were detected by
Fourier transform infrared spectroscopy (FTIR) (Nicolet IS10,
Thermo Electron Co, USA) at a spectral range of 4000 cm1 to
400 cm1 with a resolution of 8 cm1. X-ray diffraction (XRD)
patterns of the synthesized biochars were measured by a Siemens
D-501 diffractometer with Ni filter and graphite monochromator.
Furthermore, X-ray photoelectron spectroscopy (XPS) and scanning
electron microscopy (SEM) were used to further investigate the
adsorption mechanism and determine its surficial morphology
characteristics.
2.3. Adsorption experiments
Adsorption experiments were used to reveal the adsorption
potential of different biochar samples. Effect of pH on phosphorus,
arsenic and chromium adsorption was investigated in a conical
flask containing 0.1 g adsorbent and 50 ml of 50 mg L1 adsorbate
(phosphate, sodium arsenite and potassium dichromate). The pH
level was adjusted at the range of 2e10 with hydrochloric acid and
sodium hydroxide. To evaluate the stability of the BiBC during those
three adsorbates adsorption process, the desorbed bismuth con-
centration from aqueous solution was also analyzed.
Adsorption dosage experiments were executed by adding 0.01,
0.05, 0.1, 0.25, 0.5, and 1 g sorbent in a conical flask with 50 mL
solution under the optimum pH value at initial 50 mg L1 con-
centration of phosphate, sodium arsenite or potassium dichromate.
The mixture was agitated at 150 rpm in an orbital shaker (SHA-C) at
25  C for 2 h and then placed for 24 h to reach the equilibrium
condition.
Adsorption kinetics experiments were conducted as follows:
0.1 g bismuth oxide activated carbon was added in 50 mL phos-
phate, arsenic or chromium solution with initial concentration of
300, 10, and 20 mg L1 under the optimum pH level, respectively.
The sorption amount of phosphorus, arsenic and chromium were
investigated at different time intervals (0, 0.5, 1, 2, 5, 10, 30, 60, and
120 min). The mixture was agitated at 200 rpm in an orbital shaker
(SHA-C) at 25  C for 24 h to reach the equilibrium condition.
Adsorption isotherm experiments were conducted as follows:
0.1 g adsorbent was mixed with 50 ml adsorbate solution with
concentration ranging from 60 to 1800 mg L1 for phosphate and
concentration ranging from 5 to 200 mg L1 for potassium di-
chromate or sodium arsenite under the optimum pH value in the
centrifuge tube respectively. All the vessels above were shaken at
150 rpm in the oscillator for 2 h and then placed in the water bath
under constant temperature (15, 25 and 45  C) for 24 h to reach
equilibrium.
Besides, adsorbate saturated materials were prepared by mixing
4 g adsorbent with 2 L 500 mg L1 phosphate or 100 mg L1 sodium
arsenite/potassium dichromate solution under the optimum pH
value and then named as BiPBC500, BiAsBC500 and BiCrBC500
34 N. Zhu et al. / Chemosphere 164 (2016) 32e40
respectively.
The concentration of phosphate was determined using double
UVevis spectrophotometer (UV-2450 Shimadzu, Japan) at its
maximum wavelength of 700 nm. Inductive coupled plasma
emission spectrometry (ICP) and Inductive coupled plasma emis-
sion spectrometry-mass spectrum (ICP-MS) were used to detect the
concentration of arsenic, chromium and the leaching bismuth.
3. Results and discussion
3.1. Characteristics of biochar
3.1.1. Surface area and pore volume
Nitrogen adsorption is a standard procedure for the determi-
nation of the porosity of adsorbents. The N2 adsorption-desorption
isotherms of biochar based materials contained an almost hori-
zontal plateau at higher pressures (Fig. 1). According to the IUPAC,
both BiBCs and CBCs exhibited a type I isotherm and a type H4
hysteresis loops, indicating relatively small external surfaces and
highly microporous within the materials. Besides, the isotherm of
CBC400 and CBC600 raised rapidly near P/P0 ¼ 1 demonstrating the
presence of macropore.
Carbonization temperature had a significant impact upon the
pore development and specific surface area of the activated agent
(Table 1). With temperature increasing from 400 to 600  C, BET
surface areas, pore volumes as well as pore size distribution of all
the materials did show a disciplinary change. They increased firstly
and then reduced with the continuously increasing pyrolysis tem-
perature. It was known that the BET surface area, pore volume and
pore size distribution of activated carbons were closely related to
the final pyrolysis temperature, raw material and retention time.
With the same raw material and retention time, increasing
carbonization temperature within a certain temperature range in-
creases the evolution of volatile matters from the precursor, ben-
efits pore development and even creates new pores. This
pyrolyzation effect, when out of the range, will cause the collapse of
micropore and widen a significant amount of micropore to meso-
pore, quality just drop (Demiral and Gunduzoglu, 2010).
BiBCs had larger BET surface area and pore volume, especially
the micropore area and micropore volume, but a lower average
pore diameter than that of CBCs under the same carbonization
temperature (Table 1). This could be attributed to the fact that the
activating agent (bismuth oxide) helped in creating new micropore,
Fig. 1. N2 adsorptionedesorption isotherms of various activated carbons.
Table 1
Surface areas and pore volumes of activated carbons.
Sample SBET (m2/g) SMicro (m2/g) Dp (nm) VMicro (cm3/g) VTotal (cm3/g)
CBC400 6.34 e 7.03 e 0.011
CBC500 124.44 46.82 2.58 0.019 0.080
CBC600 38.49 e 5.54 e 0.053
BiBC400 87.42 52.15 2.49 0.023 0.054
BiBC500 190.40 124.55 2.00 0.047 0.095
BiBC600 106.70 22.26 2.95 0.010 0.079
inhibited the formation of tars which jam up the pores and, thus,
decreased micropore collapse of the samples. The average diameter
of 2.00 nm indicated that the synthesized BiBC500 was in the
micropore region according to the IUPAC classification implying
high adsorption potential. Highly developed micropore structure
means high specific surface area and high adsorption potential to
certain substances.
3.1.2. Morphology study
The morphology and particle size of the materials were
analyzed by SEM. The pore structure was observed obviously for
the CBC500 and CBC600 (Appendices Fig. 1). However, the pore
structure of CBC500 was inerratic while the pore structure of
CBC600 was unordered. The surface of CBC400 was rough but
apparently absence of porosity. The flourishing pore structure was
observed for all the bismuth activated carbon. Besides, when the
bismuth oxide impregnated biochar precursor was carbonized at
400, 500, and 600  C, globular particles were inlaid on the porous
surface of biochar, with a diameter between 0.5 and 1 mm (see
Appendices Fig. 1). The size of the pores on the surface of biochar
decreased with the increasing carbonization temperature. Besides,
few larger particle size with an uneven surface was also observed
for the BiBC600. The very small white globular particles were firmly
attached within the biochar matrix which may represent the bis-
muth serve as the adsorption site on the surface of activated
carbon.
3.2. The main adsorption mechanism
BiBC500 was used to investigate the adsorption performance of
bismuth activated carbons because of their high BET surface area
and abundant pore volume. The raw biochar (CBC500) removed
negligible phosphorus, arsenic and chromium while BiBC500
showed high adsorption ability to those kinds of anion.
3.2.1. FTIR
FTIR was used to identify the functional groups and the roles
they played during the modification and adsorption procedure
(Fig. 2). Functional groups of BiBC500 were not changed greatly
after being doped with bismuth (Fig. 2b). The new peaks at
513 cm1 and 622 cm1 were observed which demonstrated the
existence of BieO bond on BiBC500 (Fruth et al., 2006). The strong
new band centered at 1027 cm1, 593 cm1 and 536 cm1 were also
noticed (Fig. 2c). Those new peaks were linked with the stretching
vibration of PO4 groups and the bend vibration of O-P-O (Luwang
et al., 2011). The bands at 3435 cm1 which represented the hy-
droxyl groups were all existed in CBC500, BiBC500 BiAsBC500,
BiPBC500 and BiCrBC500. But this peak for BiAsBC500, BiPBC500
and BiCrBC500 was much more intense than BiBC500, implying O-
H of the water molecules were adsorbed on the activated carbon
surface with the adsorption process as well (Xue et al., 2009).
It was reported that the characteristic absorption bands of
adsorbed arsenate was 650e1050 cm1 for As-OH or As-O
stretching vibration (Hu et al., 2015). Peak at 1045 cm1
 N. Zhu et al. / Chemosphere 164 (2016) 32e40
Fig. 2. FTIR graph of a) CBC500, b) BiBC500, c) BiAsBC500, d) BiPBC500 and e)
BiCrBC500.
represented a weak AseO bond suggesting As (III) surface complex
was non protonated (Pena et al., 2006) while peak at 876 cm1
implied the As-OH vibration(Fig. 2d). The FTIR results suggested
that electrostatic attraction contributed slightly to As (III) adsorp-
tion. However, no any new peak was observed after Cr adsorption
(Fig. 2d). The chromium adsorption mechanism should be studied
by other technological means.
3.2.2. XRD study and XPS investigation
There was no obvious peaks found for the raw biochar (Fig. 3a).
Diffraction peaks were observed at 2q ¼ 27.165, 37.949, 39.618,
44.553, 46.007 and 48.689 which were corresponding to 012, 104,
110, 015, 113 and 202 planes respectively, marching bismuth
structure (Fig. 3b). Those diffraction peaks could also be observed
after arsenic, phosphorus or chromium adsorption. The new
diffraction peaks at 2q ¼ 14.605, 20.068, 25.472, 29.033, 29.504,
31.316 and 41.866 which were assigned to 100, 101, 110, 111, 200,
102 and 211, respectively (Fig. 3c) demonstrating the existence of
bismuth phosphate compound (Nithya et al., 2015). However, no
obvious new peaks were found after arsenic or chromium
adsorption, suggesting that there was no crystal generated.
XPS was used to further investigate the mechanism of arsenic,
phosphorus and chromium sorption by the bismuth modified
biochars (Fig. 4). Bismuth peaks corresponding to Bi4f were found
almost same in the surface of Bi-impregnated biochar and other
adsorbed materials. The binding energy of 159.6 and 164.9 eV
35
Fig. 3. XRD pattern of a) CBC500, b) BiBC500, c) BiAsBC500, d) BiPBC500 and e)
BiCrBC500.
represented the existence of bismuth structure. Prominent peak
corresponding to P2p and As3d were observed in the surface of
BiPBC500 and BiAsBC500 respectively demonstrating phosphorus
or arsenic was fixed successfully on the material after adsorption.
The binding energy of 134.5 eV demonstrated the PO4 group
structure (Lo et al., 1994). The binding energy was 48.75 suggesting
As (III) oxidation was not happen. XPS investigation also displayed
the major adsorbed chromium was Cr (III) (Fig. 4). This result
demonstrated that reduction of hexavalent chromium was
happened in the process of Cr adsorption.
3.2.3. The mechanism of adsorption of arsenic, phosphorous and
chromium
Bismuth, with the electronic configuration of (Xe)
4f145d106s26p3, is one of the most thoroughly investigated main
group elements, which has been known as ‘the wonder metal’
owing to the easy involvement in chemical combinations for the
electrons in its p orbital (Sun et al., 2014). Thus, the bismuth
impregnated biochar may have the great potential in adsorbing
some inorganic anions contaminant through ligand exchange or
electrostatic incorporation.
XPS results demonstrated phosphorus, arsenic and chromium
were immobilized firmly on the surface of bismuth activated car-
bon. The FTIR detected new peaks after arsenic and phosphorus
adsorption suggesting these two species were adsorbed by chem-
ical adsorption. The As (V) could be adsorbed through the elec-
trostatic force (ion-exchange on the protonated surface based on
Coulombic force) and Lewis acidebase interactions (ligand ex-
change reactions) while the As (III) adsorption was mainly involved
a Lewis acidebase (ligand exchange) reaction (Munoz et al., 2002).
XPS results also indicated the adsorbed arsenic was As (III). Thus,
the main arsenic adsorption mechanism could be attributed to the
Lewis acidebase reaction between the bismuth atom and arsenite.
Based on the results above, the main phosphorus sorption process
could be described as follow: bismuth atom was loaded on the
surface of activated carbon, served as the sorption site and then
become hydroxylated under the acidic condition. Then the hy-
droxylated bismuth attracted phosphate group and formed BiPO4
which has been reported as a high effective photocatalyst for
degradation methylene blue efficiently under the ultraviolet irra-
diation (Pan and Zhu, 2010). This is our ongoing word for the uti-
lization of phosphate exhausted adsorbent. FTIR detected no new
36 N. Zhu et al. / Chemosphere 164 (2016) 32e40
Fig. 4. High-resolution XPS spectra of Bi4f, As3d, P2p and Cr2p.
peaks after Cr adsorption which confirming Cr adsorption was
mainly a physical adsorption process. Previous studies proved that
the pristine biochar displayed low adsorption ability to Cr (VI) but
high affinity to Cr (III) (Pan et al., 2014). Therefore, the main chro-
mium adsorption mechanism could be described as follow: Cr (VI)
was adsorbed by the pronated material through electrostatic phe-
nomenon, then reduced to Cr (III), an easily immobilized form, and
finally fixed within the biochar matrix.
The surface of BiAsBC500 became roughly with intensive flake
around the white particle after arsenic adsorption while the surface
of BiCrBC500 changed mildly. This result confirmed that bismuth
played an important role in arsenic sorption but contributed
slightly to chromium adsorption. The acicular crystal structure
which related to bismuth phosphate on the surface of the globular
particle was obvious after phosphate adsorption. Herein, we
concluded that physicochemical actions other than the traditional
consideration mechanism, namely related to its specific surface
area, contributed mainly to the adsorption to arsenic and phos-
phorus by the prepared bismuth activated carbon. Still, there was
no sufficient information to prove that bismuth made a difference
on the chromium reduction process. Therefore, further researches
are still needed to understand the mechanism comprehensively
and thoroughly.
3.3. Adsorption study
Before evaluation of adsorption potential of the prepared ma-
terial, the leaching risk of bismuth was studied first. Results dis-
played that the concentration of residual bismuth ions was low in
water at various pH (Appendices Fig. 2). In low pH range, the
released Bi2þ was less than 0.2 mg/L (at pH 2) and it decreased to
0.01 mg/L (at pH 5). This result manifested that the most of bismuth
atoms were firmly loaded within the BiBC matrix. Within the pH
range from 5 to 10, the residual bismuth concentrations were near
zero. The leaching loss amounts of bismuth increased under the
presence of chromium. The application of bismuth activated carbon
for chromium adsorption should be cautious. There was fewer
bismuth ions detected in the aqueous solutions under the phos-
phorus or arsenic existence condition, confirming BiBC could be
considered as a kind of environmental-friendly material for phos-
phorus and arsenic pollutant removal.
3.3.1. Effect of pH and sorbent dosage
The effect of initial solution pH was investigated firstly. The
removal efficiency of arsenic for BiBC500 went up gradually and
then decreased with increasing the initial solution pH level
(Appendices Fig. 3). The maximum adsorption efficiency of As (III)
was obtained at pH 9.3 which was coincident with the first disso-
ciation constant of H3AsO3 (pK1 ¼ 9.23). Similar adsorption be-
haviors have been reported for As on metal based adsorbents such
as TiO2 and iron oxides or hydroxide (Liu et al., 2008; Meng et al.,
2000). Arsenous acid is a kind of weak acid (pKa1 ¼ 9.23,
pKa2 ¼ 12.10, and pKa3 ¼ 13.41). Under the acid condition, the
surface of adsorbent tends to be protonated with one H2O molecule
and one hydrion which hampered the Lewis acidebase reaction
between the active adsorption site and the uncharged species
H3AsO03 (Tripathi et al., 2015).
The adsorption capability of BiBC500 to phosphate was also
 N. Zhu et al. / Chemosphere 164 (2016) 32e40
affected by the initial pH dramatically. The maximum removal ef-
ficiency was obtained at pH 3 and BiBC500 showed considerable
adsorption ability to phosphate at pH between 2 and 6. Metal ox-
ides possessed strong ability to adsorb negatively charged com-
pounds, such as phosphate and arsenite (Yao et al., 2011). With the
change of pH, phosphate acid can dissociated to different anion
species as follows:
H3 PO4 4H2 PO4 þH
þ
pka1 ¼ 2:15
H2 PO4 4HPO4 þHþ
2 2
pka2 ¼ 7:20
þ first-order, pseudo-second-order and intra-particle-diffusion
HPO2 3
4 4PO4 þH pka3 ¼ 12:33
When the solution pH was at 2, namely pH < pKa1, the phos-
phate species in solution contained the uncharged H3PO4 which
could strongly be attached onto the sites of BiBC500 by coordina-
tion interaction between bismuth atom and phosphate group. The
high adsorption capacity then decreased in the range of pH from 3
to 6. Under this condition phosphate mainly existed as the form of
monovalent H2PO 4 which could be adsorbed by electrostatic
interaction with protonated bismuth atom. However, when pH was
higher than 7.2, most H2PO 2
4 was transformed to HPO4 because of

concentration of OH . On the other hand, the bismuth was stabi-
lized and unlikely protonated under the alkaline condition. Thus,
the binding affinity between phosphate and bismuth was almost
negligible.
Chromium adsorption was also affected dramatically by the
initial solution pH (Appendices Fig. 3). Removal efficiency
decreased with the increasing initial pH value ranging from 2 to 10.
The pH level affected the form of Cr (VI) in aqueous solution, which
arose from the hydrolysis reaction of Cr2O2 7 which was expressed
as follow (Giri et al., 2012):
The dominant form of hexavalent chromium at acidic condition
was HCrO 4 . The maximum adsorption capacity was obtained at pH
3, indicating HCrO 4 was easier to be adsorbed over the protonated
surface of adsorbent based on electrostatic phenomenon.
It was obvious that the removal efficiency was enhanced along
with the increasing sorbent dosage which affected the distribution
coefficient KD. KD was used for describing the binding ability of an
adsorbate element over adsorbent surface. KD could be calculated
by the following equation:
Ce
KD ¼ (1)
Ca
Where Ce and Ca are the element concentrations on adsorbents
(mg g1) and in aqueous solution (mg L1), respectively. With a
certain pH, KD would show two different situations: one is to keep a
37
constant with adsorbent dose increasing meaning a homogeneous
surface, the other is to rise with the sorbent dose implying a het-
erogeneous surface (Tian et al., 2011). BiBC500 had a KD with the
latter form implying a heterogeneous surface of BiBC500
(Appendices Fig. 4).
3.3.2. Adsorption kinetic
Batch aqueous sorption experiments displayed that BiBC500
possessed much better phosphorus, arsenic and chromium
adsorption capabilities. Three widely used kinetic models (pseudo-
models) were selected to study the experimental data based on
three initial solution concentrations to further depict the adsorp-
tion process clearly (Calisto et al., 2014). Those modes could be
described as follows in turn:
lnðqe  qt Þ ¼ ln qe  k1 t (2)
t 1 t
¼ þ (3)
qt k2 q2e qe
1
qt ¼ kp t 2 þ Ci (4)
Where t is the adsorption time (min), qt and qe are the amount of
adsorbate adsorbed on a given period of time t (mg g1) and at
equilibrium (mg g1), k1 (1 min1), k2 (g mg1 min1), and kp
(mg g1 min1/2) are the adsorption rate constant of the three ki-
netic model respectively and Ci is a constant in the intra-particle-
diffusion equation, corresponding to the thickness of boundary
layer.
Arsenic, phosphorus and chromium adsorption process pro-
ceeded very quickly and could attach equilibrium in 120 min
(Fig. 5). All the regression coefficients of the pseudo-second-kinetic
model were higher than that of pseudo-first-kinetic model, and the
experimental qe values were very close to that calculated from the
model (Table 2), suggesting the adsorption process was chemi-
sorption. The intra-particle-diffusion model is always used to
comprehensively illustrate the adsorbate diffusion mechanism and
determine whether intra-particle-diffusion is the rate-limiting
factor controlling the adsorption process especially in a solid-
liquid adsorption system. The determination coefficients of the
whole process were all between 0.920 and 0.941, indicating that the
intra-particle-diffusion model was not the only rate-limiting step, a
multi-step process is expected to work during the entire time range
(Table 2). The linear plot (for phosphate adsorption) did not pass
through the origin and could be divided into two stages according
to the changing phosphate adsorption rate (Fig. 5). The first sharp
region is instantaneous adsorption stage or the transport of phos-
phate molecules from the bulk solution to the adsorbent external
surface driven by the initial phosphate gradients. The second
potion is the intra-particle-diffusion process, phosphate molecules
transported from external surface into the pores of the BiBC500,
showing chemical sorption characteristic.
3.3.3. Adsorption isotherm
The Langmuir, Freundlich isotherm and Temkin models were
used to analyze arsenic phosphorus and chromium adsorption
experimental data (Song et al., 2016; Davis and Di Toro, 2015). The
Langmuir model (i.e. L model) is based on a monolayer adsorption
assumption and the equation could be expressed in the following
linear form:
Ce Ce 1
¼ þ (5)
qe Q0 bQ0
38 N. Zhu et al. / Chemosphere 164 (2016) 32e40

Fig. 5. Adsorption kinetics for arsenite, phosphate and dichromate removal.

Table 2
Parameters of adsorption kinetic models for removal of arsenic, phosphorous and chromium.

Adsorbate Pesudo-first-order kinetic model Pesudo-second-order kinetic model Intra-particle-diffusion model

k1 qe R2 k2 qe R2 kp Ci R2

As 0.015 1.91 0.912 0.047 2.23 0.996 0.20 0.22 0.920

P 0.023 106.29 0.966 0.008 131.58 0.982 11.69 2.74 0.941
Cr 0.028 2.60 0.989 0.020 3.27 0.998 0.29 0.13 0.935

Freundlich model (i.e. F model) derived from a multilayer (mg g1) and b (L mg1) were Langmuir constants. Kf (mg g1) and
adsorption condition with an inhomogeneous distribution of heat n were Freundlich constants. A (L mg1) and bT (J g1) were Temkin
of adsorption over the surface. Freundlich equation could be writ- constants. T was absolute temperature and R was universal gas
ten as follows: constant 8.314 J mol1 K1. Results showed that the L model was
mostly suitable for the equilibrium data of arsenic and phosphorus
1 adsorption (r2L > r2T > r2F) assuming a monolayer adsorption pro-
ln qe ¼ ln Kf þ ln Ce (6)
n cess for As and P respectively (Appendices Fig. 5). The maximum
Temkin isotherm (i.e. T model) assumes that the heat of adsorption capacities were 125.40 mg g1 for P and16.21 mg g1 for
adsorption decreases linearly. The equation could be expressed in As (III), which were more excellent than other commonly used
the following form: adsorbents reported in the literature (Qian et al., 2013; Cui et al.,
2013). A dimensionless equilibrium parameter (RL) can be further
RT RT analyzed by the Langmuir equation. The lower RL value means the
qe ¼ lnA þ ln Ce (7)
bT bT more favorable adsorption provided at the range between 0 and 1.
The values of RL were 0.034e0.583 for BiBC500 to arsenic and
Where Ce (mg L1) was the equilibrium concentration, and qe 0.045e0.541for BiBC500 to phosphate in the C0 range of our study,
(mg g1) was the amount of adsorption capacity at equilibrium. Q0 implying a favorable adsorption process to arsenic and phosphate.
 N. Zhu et al. / Chemosphere 164 (2016) 32e40 39

The Temkin plots also had good linearity indicating the strong   
chemical interaction between adsorbate and adsorbent. These re- d ln Cqee
sults illustrated the homogenous distribution of active sites over DHobs ¼ R   (11)
sorbent surface.
d T1
Generally, the Langmuir model is suitable for the uniformly
distributed metal salts, namely, the active sorption site to some
certain contaminants on the surface of activated carbon. The Where qe and Ce were calculated by the Langmuir parameters.
Freundlich model, however, was best fitted to the chromium DHobs can be estimated from the slope of line yields from plots of
sorption data (r2F > r2L > r2T) suggesting a multilayer sorption ln(qe/Ce) versus reciprocal reaction temperature (Liu et al., 2015).
process for chromium. The maximum Cr (VI) adsorption capacity Results showed that adsorption of As needed much more adsorp-
was detected as 12.23 mg g1 in our study. This result also indi- tion heat than phosphate adsorption (Table 3). An increase of DHobs
cated that those active adsorption sites were inoperative to with increasing qe was observed which could contribute to the
chromium. reason that a greater repulsive interaction between the adsorbed As
or P and the free adsorbate in aqueous solution needs to be over-
come at a higher qe (Appendices Fig. 6). Besides, the increasing rate
of change in the DHobs curve (both P and As adsorption) indicated
3.3.4. Thermodynamic parameters that the higher adsorption capacity means the more heat of
The Langmuir model was best fitted for experimental data of P adsorption after adsorbed certain substances.
and As (III) adsorption. The Gibbs free energy equation and the
linearized Van't Hoff equation (i.e., the van't Hoff plot) were then 3.4. Utilization of phosphate depleted adsorbents
used for estimating the thermodynamic parameters of P and As
adsorption, which were expressed as follows: As Bi2O3 and BiPO4 has been reported with an excellent pho-
todegradation ability to organic pollutant such as methylene blue

DG ¼ RTlnK (8) under the ultraviolet light irradiation (Pan and Zhu, 2010). As bio-
char has been reported as a considerable adsorption capacity for
  methylene blue. A dark experiment (without ultraviolet radiation)
DS DH
lnK ¼  (9) was first executed as follow: 0.1 g phosphate saturated material was
R RT
mixed with 50 ml methylene blue solution in a conical flask with
Where DG (kJ mol1) is the change of free energy, DS (kJ mol1) is concentration ranging from 10 to 150 mg L1. Correspondingly,
the change of entropy, T is the absolute temperature, DH (kJ mol1) 0.1 g BiPBC500 was mixed with 50 ml methylene blue solution in a
is the change of enthalpy (integral enthalpy), R is the ideal gas quartz container with concentration ranging from 10 to 150 mg L1.
constant, and K is the dimensionless equilibrium coefficient esti- All the vessels were standing at 25  C for 24 h to reach the equi-
mated from the Langmuir constant (b) as librium sufficiently and then detected the methylene blue con-
centration. Then, the quartz container was placed under ultraviolet
K ¼ bCw (10) irradiation environment at 25  C. The photodegradation methylene
blue amounts was investigated at different time intervals (0.5, 1, 2,
Where Cw is the water concentration (5.56  104 mmol L1). 3, 4 and 5 h). The MB removal efficiency which was enhanced after
Thermodynamic parameters for phosphate and arsenic (III) sufficient UV irradiation especially at a medium concentration
adsorption were summarized in Table 3. The negative values of (Appendices Fig. 7). Obviously, the reaction rates decreased with
free energy change DG demonstrate the spontaneous nature of the increasing of the initial concentration of MB solution. This could
the adsorption process. The DG was within the range of 20 be attributed to the high dye concentration effect which impeded
to 80 kJ mol1, suggesting that chemisorption and phys- the light penetration and resulted in fewer photon reaching to the
isorption coexist in the adsorption process. The positive values in catalyst surface. This results demonstrated BiBC500 could be used
the differential heat of adsorption (DH ) indicate that the for phosphate removal and further make contribution to
adsorption of P or As (III) onto the BiBC500 is an endothermic adsorption-photodegradation of dye.
process. Investigating the enthalpy changes with the amount of
adsorption capacity (qe) would be conducive to comprehend the 4. Conclusion
interaction between adsorbate (P or As) and BiBC500. The
observed molar differential enthalpy (DHobs) of arsenic or phos- Bismuth activated carbons displayed high sorption ability to
phate sorption on bismuth activated carbon were further ob- arsenite, phosphate and dichromate. Bach adsorption experiments
tained from the differential Van't Hoff equation which were reveled the spontaneous nature and endothermic process for P and
expressed as follows: As adsorption. The interaction between phosphate group and bis-
muth particles played a fatal role in phosphate adsorption and
formed bismuth phosphate which could photodegrade MB solu-
tion. Ligand exchange contributed mainly to arsenic removal while
Table 3 electrostatic phenomenon along with the chromium reduction
Parameters of Langmuir isotherms and thermodynamics of As (III) and P adsorption.
played an important role in chromium sorption. This work
Adsorbate Temperature(K) Langmuir parameters Thermodynamics demonstrated that bismuth modified biochar provides an innova-
Q0 b R2 DG DS DH tive and effective way to utilize straw resources by preparing cost-
effective adsorbents for anion pollutants treatment.
P 288 1.18 13.88 0.998 26.47 0.12 6.82
298 1.32 18.86 0.995 27.76
318 1.48 22.84 0.994 29.97 Acknowledgement
As 288 0.182 1.13 0.99 32.48 0.15 11.86
298 0.216 1.31 0.992 34.36 This study was funded by the National Natural Science Foun-
318 0.273 1.5 0.997 37.18
dation of China (41171236), the National Key Technology R&D
40 N. Zhu et al. / Chemosphere 164 (2016) 32e40
Program of China (2012BAD15B03) and the National S&T Major
Project of China (2012ZX07204-003).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.chemosphere.2016.08.036.
References
Afkhami, A., Saber-Tehrani, M., Bagheri, H., 2010. Simultaneous removal of heavy-
metal ions in wastewater samples using nano-alumina modified with 2,4-
dinitrophenylhydrazine. J. Hazard. Mater. 181, 836e844.
Alemayehu, E., Thiele-Bruhn, S., Lennartz, B., 2011. Adsorption behaviour of Cr(VI)
onto macro and micro-vesicular volcanic rocks from water. Sep. Purif. Technol.
78, 55e61.
Ali, I., 2012. New generation adsorbents for water treatment. Chem. Rev. 112,
5073e5091.
Banerjee, K., Amy, G.L., Prevost, M., Nour, S., Jekel, M., Gallagher, P.M.,
Blumenschein, C.D., 2008. Kinetic and thermodynamic aspects of adsorption of
arsenic onto granular ferric hydroxide (GFH). Water Res. 42, 3371e3378.
Calisto, V., Ferreira, C.I.A., Santos, S.M., Gil, M.V., Otero, M., Esteves, V.I., 2014. Pro-
duction of adsorbents by pyrolysis of paper mill sludge and application on the
removal of citalopram from water. Bioresour. Technol. 166, 335e344.
Cui, H., Su, Y., Li, Q., Gao, S., Shang, J.K., 2013. Exceptional arsenic (III,V) removal
performance of highly porous, nanostructured ZrO2 spheres for fixed bed re-
actors and the full-scale system modeling. Water Res. 47, 6258e6268.
Davis, C.W., Di Toro, D.M., 2015. Modeling nonlinear adsorption to carbon with a
single chemical parameter: a lognormal Langmuir isotherm. Environ. Sci.
Technol. 49, 7810e7817.
Demiral, H., Gunduzoglu, G., 2010. Removal of nitrate from aqueous solutions by
activated carbon prepared from sugar beet bagasse. Bioresour. Technol. 101,
1675e1680.
Fruth, V., Ianculescu, A., Berger, D., Preda, S., Voicu, G., Tenea, E., Popa, M., 2006.
Synthesis, structure and properties of doped Bi2O3. J. Eur. Ceram. Soc. 26,
3011e3016.
Giri, A.K., Patel, R., Mandal, S., 2012. Removal of Cr (VI) from aqueous solution by
Eichhornia crassipes root biomass-derived activated carbon. Chem. Eng. J.
185e186, 71e81.
Hu, X., Ding, Z., Zimmerman, A.R., Wang, S., Gao, B., 2015. Batch and column
sorption of arsenic onto iron-impregnated biochar synthesized through hy-
drolysis. Water Res. 68, 206e216.
Hua, M., Zhang, S.J., Pan, B.C., Zhang, W.M., Lv, L., Zhang, Q.X., 2012. Heavy metal
removal from water/wastewater by nanosized metal oxides: a review. J. Hazard.
Mater. 211, 317e331.
Lazaridis, N.K., Bakoyannakis, D.N., Deliyanni, E.A., 2005. Chromium(VI) sorptive
removal from aqueous solutions by nanocrystalline akaganeite. Chemosphere
58, 65e73.
Liu, G., Zhang, X., Talley, J.W., Neal, C.R., Wang, H., 2008. Effect of NOM on arsenic
adsorption by TiO2 in simulated As(III)-contaminated raw waters. Water Res.
42, 2309e2319.
Liu, C.-H., Chuang, Y.-H., Chen, T.-Y., Tian, Y., Li, H., Wang, M.-K., Zhang, W., 2015.
Mechanism of arsenic adsorption on magnetite nanoparticles from water:
thermodynamic and spectroscopic studies. Environ. Sci. Technol. 49,
7726e7734.
Lo, P.H., Tsai, W.T., Lee, J.T., Hung, M.P., 1994. Role of phosphorus in the
electrochemical-behavior of electroless Ni-P alloys in 3.5 wt.percent Nacl so-
lutions. Surf. Coat. Technol. 67, 27e34.
Loganathan, P., Vigneswaran, S., Kandasamy, J., 2013. Enhanced removal of nitrate
from water using surface modification of adsorbents e a review. J. Environ.
Manag. 131, 363e374.
Luwang, M.N., Ningthoujam, R.S., Srivastava, S.K., Vatsa, R.K., 2011. Disappearance
and recovery of luminescence in Bi3þ, Eu3þ codopecl YPO4 nanoparticles due
to the presence of water molecules up to 800 degrees C. J. Am. Chem. Soc. 133,
2998e3004.
Marris, E., 2006. Putting the carbon back: black is the new green. Nature 442,
624e626.
Meng, X.G., Bang, S., Korfiatis, G.P., 2000. Effects of silicate, sulfate, and carbonate on
arsenic removal by ferric chloride. Water Res. 34, 1255e1261.
Munoz, J.A., Gonzalo, A., Valiente, M., 2002. Arsenic adsorption by Fe(III)-loaded
open-celled cellulose sponge. thermodynamic and selectivity aspects. Envi-
ron. Sci. Technol. 36, 3405e3411.
Nithya, V.D., Hanitha, B., Surendran, S., Kalpana, D., Kalai Selvan, R., 2015. Effect of
pH on the sonochemical synthesis of BiPO4 nanostructures and its electro-
chemical properties for pseudocapacitors. Ultrason. Sonochem. 22, 300e310.
Onnby, L., Kumar, P.S., Sigfridsson, K.G., Wendt, O.F., Carlson, S., Kirsebom, H., 2014.
Improved arsenic(III) adsorption by Al2O3 nanoparticles and H2O2: evidence of
oxidation to arsenic(V) from X-ray absorption spectroscopy. Chemosphere 113,
151e157.
Pan, C., Zhu, Y., 2010. New type of BiPO4 oxy-acid salt photocatalyst with high
photocatalytic activity on degradation of dye. Environ. Sci. Technol. 44,
5570e5574.
Pan, J.-j., Jiang, J., Xu, R.-k., 2014. Removal of Cr(VI) from aqueous solutions by
Na2SO3/FeSO4 combined with peanut straw biochar. Chemosphere 101, 71e76.
Pena, M., Meng, X.G., Korfiatis, G.P., Jing, C.Y., 2006. Adsorption mechanism of
arsenic on nanocrystalline titanium dioxide. Environ. Sci. Technol. 40,
1257e1262.
Qian, W., Zhao, A.-z., Xu, R.-k., 2013. Sorption of As(V) by aluminum-modified crop
straw-derived biochars. Water Air Soil Pollut. 224.
Song, G., Zhu, X., Chen, R., Liao, Q., Ding, Y.D., Chen, L., 2016. An investigation of CO2
adsorption kinetics on porous magnesium oxide. Chem. Eng. J. 283, 175e183.
Sun, Y., Zhou, G., Xiong, X., Guan, X., Li, L., Bao, H., 2013. Enhanced arsenite removal
from water by Ti(SO4)2 coagulation. Water Res. 47, 4340e4348.
Sun, H.-T., Zhou, J., Qiu, J., 2014. Recent advances in bismuth activated photonic
materials. Prog. Mater. Sci. 64, 1e72.
Tang, W.-W., Zeng, G.-M., Gong, J.-L., Liang, J., Xu, P., Zhang, C., Huang, B.-B., 2014.
Impact of humic/fulvic acid on the removal of heavy metals from aqueous so-
lutions using nanomaterials: a review. Sci. Total Environ. 468e469, 1014e1027.
Tian, Y., Wu, M., Liu, R., Wang, D., Lin, X., Liu, W., Ma, L., Li, Y., Huang, Y., 2011.
Modified native cellulose fibers-A novel efficient adsorbent for both fluoride
and arsenic. J. Hazard. Mater. 185, 93e100.
Tripathi, S., Bose, R., Roy, A., Nair, S., Ravishankar, N., 2015. Synthesis of hollow
nanotubes of Zn2SiO4 or SiO2: mechanistic understanding and uranium
adsorption behavior. ACS Appl. Mater. Interfaces 7, 26430e26436.
Wang, X.S., Chen, L.F., Li, F.Y., Chen, K.L., Wan, W.Y., Tang, Y.J., 2010. Removal of Cr
(VI) with wheat-residue derived black carbon: reaction mechanism and
adsorption performance. J. Hazard. Mater. 175, 816e822.
Xue, F., Li, H., Zhu, Y., Xiong, S., Zhang, X., Wang, T., Liang, X., Qian, Y., 2009. Sol-
vothermal synthesis and photoluminescence properties of BiPO4 nano-cocoons
and nanorods with different phases. J. Solid State Chem. 182, 1396e1400.
Yao, Y., Gao, B., Inyang, M., Zimmerman, A.R., Cao, X., Pullammanappallil, P., Yang, L.,
2011. Removal of phosphate from aqueous solution by biochar derived from
anaerobically digested sugar beet tailings. J. Hazard. Mater. 190, 501e507.