Chemical Engineering Journal 350 (2018) 1000–1009

Contents lists available at ScienceDirect

Chemical Engineering Journal

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

Classical theory and electron-scale view of exceptional Cd(II) adsorption T

onto mesoporous cellulose biochar via experimental analysis coupled with
DFT calculations
⁎
Quan Chena, Jiewei Zhenga, Liuchun Zhengc, Zhi Dangb, Lijuan Zhanga,
a
School of Chemistry and Chemical Engineering, South China University of Technology, Guangzhou 510640, PR China
b
School of Environment and Energy, South China University of Technology, Guangzhou 510006, PR China
c
School of Chemistry and Environment, South China Normal University, Guangzhou 510006, PR China

H I GH L IG H T S G R A P H I C A L A B S T R A C T

• Three mesoporous cellulose biochar

(MCB) adsorbents were prepared.
Illustration of Cd(II) adsorption mechanism onto mesoporous cellulose biochar.

• The structure-performance relation-

ship in Cd(II) adsorption was eluci-
dated.
• Asorption
new strategy for understanding ad-
micro-mechanism was pro-
posed.
• Provided theoretical guidance for the
subsequent biochar preparation.

A R T I C LE I N FO A B S T R A C T

Keywords: The micro-mechanism of heavy metal cations adsorption onto biochar is critical for the renovation of heavy
Mesoporous cellulose biochar metal contamination. In this work, we prepared three mesoporous cellulose biochar (MCB) adsorbents with
Cd(II) different surface area, O and N functionalities content properties through a 300 °C carbonization-KOH activa-
Classical theory tion-700 °C carbonization-HNO3 oxidation process. The Cd(II) adsorption performance of three MCB was com-
Density functional theory
pared, the MCB-1h exhibited best Cd(II) adsorption capacity (368.8 mg/g), and even remained 89% after 5
Micro-mechanism
cycles. DFT calculations identified that MCB-1h had comparatively stronger Cd(II) binding ability, and the re-
sults were well consistent with the experimental data. The adsorption micro-mechanism was analyzed in the
view of classical theory and electron-scale. The classical theory results proposed that the adsorption process was
dominated by chemisorption, and electron sharing or exchange between Cd(II) and biochar occurred. Electron-
scale mechanism analysis found that the functional groups and aromatic could provide lone pair electrons and π
electrons for the Cd(II) adsorption, respectively. The N functionalities such as amino, pyridine and pyrrole
groups could raise the adsorption ability of the biochar adsorbent. Overall, our results not only provide new
insights into the heavy metal adsorption, but also has significant reference value for the subsequent biochar
adsorbent preparation.

⁎
Corresponding author.
E-mail address: celjzh@scut.edu.cn (L. Zhang).

https://doi.org/10.1016/j.cej.2018.06.054
Received 11 April 2018; Received in revised form 7 June 2018; Accepted 9 June 2018
1385-8947/ © 2018 Elsevier B.V. All rights reserved.
Q. Chen et al.
1. Introduction
In recent decades, heavy metal cadmium containing waste has been
directly or indirectly discharged into the environment with the rapid
development of metallurgy, batteries, mining, electrolysis, electrical
appliance manufacturing industries [1,2]. Cadmium is a human non-
essential element with high chemical activity, toxicity and non-de-
gradable, and easy to enter the plant and human body, triggering a
variety of illnesses [3]. In order to reduce the harm of cadmium to the
environment and human, many techniques have been used for cad-
mium removal, including adsorption, membrane separation, ion-ex-
change, chemical precipitation, penetration, animal and plant restora-
tion [4,5].
Among the above techniques, adsorption has been broadly con-
cerned due to its simplicity of operation, high removal efficiency, and
low cost [6,7]. It should be pointed out that its practical application
heavily depends on the adsorption efficiency of adsorbent, which needs
to be inexpensive, having strong sorbate affinity and large number of
active adsorption sites. Among the numerous adsorbents, biochar have
been most frequently mentioned due to its economic advantages, large
surface area, and good adsorption performance [8,9]. However, if a
large fraction of the pores in the biochar are micropores (< 2 nm), its
adsorption ability might be restricted by hydrothermal unstable, low
adsorption capacity, and relatively slow diffusion kinetic [10]. Meso-
porous biochar is a very promising heavy metal adsorbent with high
adsorption capacity and hydrothermal stability, while retaining the
large surface area and pore volume [11,12]. On the other hand, biochar
is usually prepared using bioresource waste as precursors. Cellulose is a
cheap biomass, and its content in wheat straw, corn straw, and rice
straw are 33–40%, about 45% and about 40%, respectively [13]. Re-
garding cellulose as raw material to prepare biocahr adsorbent could
not only remove heavy metals, but also take full advantage of biomass
resources, and turn “waste” into treasure [14,15].
Considerable efforts have devoted to the heavy metal removal using
biochar adsorbent. Huang and his co-workers prepared biochar ad-
sorbent using peanut hull as raw material. When the Cd(II) initial
concentration was 200 mg/L, the removal percentage was up to 90%,
and the adsorption isotherm theoretical maximum adsorption capacity
was 28.99 mg/g [16]. A cellulose biochar was reported by Reddy and
his co-workers, the Pb(II) adsorption process of the adsorbent was in
accordance with pseudo-second-order kinetic model, and the maximum
adsorption capacity was 316.3 mg/g [17]. However, much work has
focused on preparation of adsorbent, evaluation of adsorption proper-
ties, factors that affect the adsorption performance, and adsorption
model. Nevertheless, the structure-performance relationship in Cd(II)
adsorption onto biochar has not yet been systematically cleared, and
further efforts are required.
Moreover, heavy metals adsorbed onto biochar were controlled by a
number of mechanisms, such as physical sorption, complexation, ion
exchange, precipitation, and electrostatic interactions (chemisorption)
[18]. But rarely systematic study concerned the adsorption micro-me-
chanism, especially at the electron-scale. Whereas, the micro-me-
chanism is extremely helpful to guide the preparation of high-efficiency
heavy metal biochar adsorbent, thus reducing the practical cost of ad-
sorption method. There is thereby an urgent need to investigate the
adsorption mechanism. It’s generally established that the traditional
theoretical analysis could propose the adsorption mechanism, but it
remains challenge to understand on the basis of experimental data
alone. In recent years, computer calculations based on density func-
tional theory (DFT) have been employed to better study the adsorption
micro-mechanism, which is a quantum mechanic method for studying
the electron structure of multi-electron systems, and hold the ad-
vantages in wide applications and simple algorithm [19]. Yu et al. in-
vestigated the adsorption of organic compounds onto MOFs to ratio-
nalize the difference in adsorption performance and interpret the
adsorption micro-mechanism by DFT calculations [20]. In Liang’s work,
1001
Chemical Engineering Journal 350 (2018) 1000–1009
DFT calculations were studied to explore the surface interactions be-
tween heavy metal ions and different active sites of Mn-doped α-Fe2O3
nanocrystals, and to further understand the ion exchange mechanism in
the adsorption [21]. Unfortunately, the clearly adsorption mechanism
of Cd(II) onto modified biochar has not yet been revealed.
Taking the aforementioned aspects into consideration, we herein
show our attempt to produce three mesoporous cellulose biochar (MCB)
adsorbents with different surface area and morphology, O and N
functionalities content adsorbents, evaluate and compare their Cd(II)
adsorption performance, inspect the structure-performance relationship
and factors that affect adsorption performance. The adsorption micro-
mechanism is to be investigated by the combination of classical theory
(based on the experimental data) and electron-scale (based on DFT
calculations). In addition, our DFT calculations not only interpret the
adsorption micro-mechanism, but also rationalize the differences in
adsorption performance. Our mechanistic study provides meaningful
insights into the preparation of efficient biochar adsorbents for en-
vironmental renovation.
2. Materials and methods
2.1. Materials
Cellulose (Cellulose, Powder, Sigma-Aldrich), Cadmium chloride
hemi(pentahydrate) (CdCl2·2.5H2O, 99%, J&K), Copper(II) chloride
dihydrate, CuCl2 2H2O, 99%, J&K). Potassium hydroxide (KOH, AR,
Guangzhou Chemical Reagent Factory), ethylene diamine tetraacetic
acid (EDTA, AR, Guangzhou Chemical Reagent Factory), nitric acid
(HNO3, AR, Guangzhou Chemical Reagent Factory). Cadmium standard
solution was purchased from Tianjin Kemiou Chemical Reagent Co.,
Ltd, China. All the chemicals are of analytical grade and used without
further treatment or purification.
2.2. Preparation of mesoporous cellulose biochar
Typically, dried cellulose powder was evenly spread in the crucible
and placed in the central of tube furnace. The heating rate was 5 °C/
min, the nitrogen flow rate was 50 mL/min, the temperature was raised
from room temperature to 300 °C and maintained for 1 h, and then
cooled to room temperature to obtain cellulose biochar (designated
CB). The CB and KOH solids were mixed at a mass ratio of 1:3, and
deionized water was added and stirred (KOH concentration: 30%), then
placed in 80 °C oven to obtain a cellulose alkali (named CB-KOH). The
CB-KOH was calcined in the tube furnace. The heating rate was set as
5 °C/min and the nitrogen flow rate was 50 mL/min, the temperature
raised from room temperature to 700 °C, and kept for 1 h, 2 h, 3 h to
ensure the formation of aromatic structure, then cooled to room tem-
perature to achieve cellulose activated carbon (designated CB-1h, CB-
2h, CB-3h). Then, 30% nitric acid solution was slowly mixed with 20 g
of CB-1–3h and stirred at 70 °C for 3 h. The reaction solution was di-
luted with large amount of deionized water and filtered. And then
washed with a large amount of deionized water to neutral. The samples
were dried in a vacuum oven to obtain the targeting product (named
MCB-1h, MCB-2h, MCB-3h).
2.3. Characterizations
FTIR (Vector 33-IR, Bruker) was employed to analyze functional
groups in the samples obtained by the KBr pellet technique. The spec-
trums were collected with wavenumber from 400 to 4000 cm−1. N2
adsorption-desorption isotherms were obtained from an ASAP 2010
analyzer (Micromeritics) at 77 K. The surface area was determined
based on BET analyses. Pore size distribution was derived from the
desorption branches of the isotherms based on BJH model, and the total
pore volume was calculated using BJH analysis. The surface mor-
phology, microstructure and elements of the samples were observed by
Q. Chen et al.
high resolution scanning electron microscopy (SEM, Merlin, Zeiss). X-
ray diffraction spectroscopy (XRD, D8 AdvanceX, Bruker) with Cu Kα
radiation generated at 45 kV and 40 mA was used to measure the crystal
structure of the samples. XRD spectra were obtained from 5 to 80°. The
surface element and functional groups composition of the samples were
characterized by X-ray photoelectron spectroscopy (XPS, K-Alpha,
Thermo Fisher Scientific) and analyzed with XPS PEAK 4.1 software.
2.4. Batch adsorption test
The adsorption performance of Cd(II) onto MCB and CB was eval-
uated and compared by static adsorption method. Cd(II) stock solution
was prepared by dissolving CdCl2·2.5H2O in ultrapure water and then
gradient dilution. The solution pH was adjusted by 0.1 M HCl solution
or 0.1 M NaOH solution. The typical experimental procedure was as
follows: 20 mL of 200 mg/L Cd(II) solution (original pH 5.5) was placed
in a conical flask and 20 mg of adsorbent was added. At 25 °C, shaken at
150 rpm for 1 h, then filtered through a 0.22 μm filter membrane, and
the filtrate was determined by an inductively coupled plasma-atomic
emission spectrometry (ICP-AES, Prodigy, Leeman Labs). The effect of
factors on the adsorption performance was investigated by changing the
test conditions, such as the initial concentration of Cd(II) solution
(10–500 mg/L), adsorbent dose (4–200 mg), pH (2–7), temperature
(10–40 °C) and contact time (10 min–10 h). The adsorption capacity qe
and the removal percentage Pe were calculated from Eqs. (1) and (2):
(c0−ce ) V
qe =
m (1)
100(c0−ce )
Pe =
c0 (2)
where C0 and Ce are the initial and equilibrium concentration of Cd(II)
in the solution, respectively; m is the added amount of adsorbent; and V
is the volume of the solution; qe is the amount of adsorbed Cd(II) at
equilibrium.
Desorption experiments were conducted using 0.1 M EDTA as
eluent. Firstly, adsorption was performed using 100 mg MCB-1h in
200 mg/L Cd(II) solutions (100 mL) at 25 °C, shaken at 150 rpm for 1 h,
then filtered through a 0.22 μm filter membrane. The desorbed ad-
sorbent was washed thoroughly with 0.1 M EDTA and ultrapure water
and dried, then subjected to the next cycle. The solution was set to a
concentration of 1.0 mg/mL depending on the amount of the adsorbent.
The regeneration performance of MCB-1h was evaluated by the ad-
sorption capacity of Cd(II).
All the adsorption experiments were conducted in three parallel
groups and expressed as standard deviation.
2.5. Quantum chemistry computational simulations
The micro-mechanism of Cd(II) adsorption onto MCB in electron-
scale was studied by the spin-polarized density functional theory (DFT).
DFT calculations were carried out using the Dmol3 package in Materials
Studio 2017 R2. An atomic centered grid was used for the atomic basis
function. The electronic basis set was dual numerical polarization
(DNP4.4) all-electron basis set. The self-consistent field convergence
value was 1.0 × 10−6. The “DFT semi core pseudopots (DSPPs)”
method was adopted as the core treatment. The dispersion was cor-
rected by the Grimme scheme to avoid the limitations in handling weak
interactions by DFT. The conduct or like screening model (COSMO)
with a permittivity of 78.54 (water) was used to mimic structures en-
cased by the aqueous layer, and the temperature was set as 298 K. The
size of the base set is similar to that of the Gaussian function type 6-31G
(d, p) basis set, but is more accurate than that. This high-precision
numerical basis set can reduce the base group overlap effect, and the
system is accurately described. Exchange-correlation functions were
described by the generalized gradient correction (GGA) Perdew-Bruke-
1002
Chemical Engineering Journal 350 (2018) 1000–1009
Ernzerhof (PBE) function. A simple graphene structure composed of
seven aromatic rings was selected as the model compound of MCB, and
the k-points were set to 4 × 4 × 1 after convergence. Integration was
performed in the Brillouin zone and a 20 Å vacuum layer. The form of
cadmium is a positively charged Cd(II), and the adsorption binding
energies were calculated by the following Eq. (3):
Eads = E (X …Cd (II ))−E (X )−E (Cd (II )) (3)
where E(X) is the energy of the isolated adsorbent; E(Cd(II)) is the en-
ergy of the isolated adsorbate Cd II) cluster; E(X…Cd(II)) is the energy
of the steady adsorption state on different adsorbents.
In order to clarify the electron exchange or sharing behavior in the
Cd(II) adsorption process, the charge density distribution of the ad-
sorbent and its overlap with the electron cloud of Cd(II) were analyzed.
3. Results and discussion
3.1. Physical/chemical properties of mesoporous cellulose biochar
Cellulose was carbonized at 300 °C to evaporate the free and crystal
water, then activated with strong alkali KOH to promote the formation
of pore structure and O-containing functional groups. After that, the
intermediate was carbonized at 700 °C to form the aromatic structure.
Lastly, strong acid HNO3 was added to further oxidize the biochar by
the reaction of strong acid and alkali, thereby introducing N-containing
functional groups and mesoporous structure. The technological condi-
tions are of great importance for the characteristics of cellulose biochar,
in which the carbonization time at 700 °C affects the surface structure
most. Herein, we prepared three cellulose biochar with 1 h, 2 h and 3 h
carbonization time, respectively, and their microstructure and Cd(II)
adsorption performance will be further discussed below.
The surface microscopic morphology of the adsorbents was ob-
served by SEM. Fig. 1(A–H) displayed the typical SEM images of cel-
lulose biochar (CB) and mesoporous cellulose biochar (MCB). The
surface of CB was relatively flat, only some wrinkles were observed due
to the high-temperature dehydration. For MCB, a clear internal layer,
large number of nanoscale pores and channels were observed, which
could increase the contact probability of adsorbent and adsorbate,
thereby conducing to the heavy metal ion adsorption. In addition, the
surface structure of MCB gradually collapsed with the increase of car-
bonization time.
The content of C, N, and O was detected in the MCB through SEM-
EDX, and the accurate content (excluding H) of the elements was fur-
ther confirmed by XPS, the results were listed in Table 1. Compared to
CB, N was newly introduced into MCB, and the O content increased at
the same time, owing to the oxidation of nitric acid and parts hydroxyl
groups were oxidized to carboxyl groups [22]. The O content in the
MCB decreased with the increase of carbonization, while the N content
slightly increased.
The nitrogen adsorption-desorption isotherms and pore size dis-
tributions of MCB and CB were shown in Figs. 1(I and J) and S1, and the
related data were listed in Table 1. As can be seen from Fig. 1(I), there
were obvious hysteresis loops in the relative pressure range of
0.45–0.95, were type IV isotherms, and the average pore diameters
were from 3.13 to 3.23 nm, confirming that MCB was mesoporous
structure. While the nitrogen adsorption-desorption isotherm of CB
(Fig. S1A) is the type III isotherm, indicating weak interactions with
nitrogen and adsorption affinity of CB [23]. According to the data in
Table 1, the specific surface area and pore volume of CB were 8.1 m2/g
and 0.03 cm3/g. However, for those of MCB, both increased sig-
nificantly, were about 550–600 m2/g and 0.35 cm3/g, respectively. It
can be seen from Figs. S1(B) and 1(J) that the pore size of CB was
widely distributed, but the pore size of MCB was narrowly distributed
with average of about 3.2 nm. In addition, the results had a certain
difference with the SEM results, owing to the different principles of the
two test methods and the series treatments (KOH activation and HNO3
Q. Chen et al. Chemical Engineering Journal 350 (2018) 1000–1009

Fig. 1. SEM images of CB (A and E), MCB-1h (B and F), MCB-2h (C and G) and MCB-3h (D and H) with scale bar of 10 μm (A–D) and 1 μm (E–H); nitrogen adsorption-
desorption isotherms (I) and pore size distributions (J) of MCB.

oxidation) on the MCB surface. In general, a larger specific surface area
and pore volume of the adsorbent lead to the better adsorption per-
formance. Therefore, the increase of specific surface area and pore
volume were beneficial to the Cd(II) adsorption. In addition, the spe-
cific surface area and pore volume of MCB decreased with the increase
of carbonization time, due to the collapse of the surface structure.
The functional groups and crystalline structure of the adsorbents are
essential for the adsorption performance and were confirmed by XRD,
FTIR, and XPS. Fig. 2(A) displayed the XRD patterns of the CB and MCB.
The pattern for CB had two diffraction peaks at 15° and 22° that can be
well-indexed to the 1 0 1 and 0 0 2 crystal planes of natural cellulose,
respectively, indicating that the crystalline structure of cellulose has not
been destroyed this moment [24]. However, for MCB, the diffraction
peak at 15° completely disappeared, suggesting that the 1 0 1 crystal
planes were destroyed by carbonization and HNO3 oxidation. And the
diffraction peak at 22° weakened, indicating the 0 0 2 crystal planes
were damaged, and the proportion of crystalline area sharply reduced.
Moreover, the newly produced peak at 43° was the characteristic peak
of aromatic structure, indicating that aromatic structure was formed
into MCB [25].
FTIR spectra of CB and MCB (Fig. 2B) was obtained in the wave-
number range from 400 to 4000 cm−1. The characteristic peaks at
3328 cm−1, 2913 cm−1, and 1053 cm−1 in the CB pattern were at-
tributed to C–O–H, C–H, and C–O–C or C–O–H stretch vibrations,

Table 1
Surface elemental composition and BET data of CB and MCB.
Sample XPS (wt%) SEM-EDX (wt%) BET surface Pore volume Pore size

C 1s O 1s N 1s C O N Others

CB 88.1 11.9 – 83.9 16.1 – – 8.1 0.03 4.99

MCB-1h 78.56 19.66 1.78 76.4 21.56 1.20 0.84 597.7 0.36 3.23
MCB-2h 80.47 17.68 1.85 76.6 20.66 1.24 1.50 563.8 0.34 3.13
MCB-3h 82.41 15.73 1.87 77.3 19.38 1.35 1.97 557.1 0.33 3.19

Notes: The others are Si and K; BET surface (m2/g), Pore volume (cm3/g), Pore size (nm).

1003
Q. Chen et al.
Fig. 2. XRD patterns (A), FTIR spectra (B) and XPS survey spectra (C) of the CB and MCB, XPS C1s spectra (D) of CB and MCB, and XPS N1s spectra (E) of MCB.
respectively. After series treatment, parts of the hydroxyl moiety were
destroyed or oxidized to carboxyl. Therefore, the characteristic peak of
C]O appeared at 1720 cm−1 in the curves of MCB. At the same time,
the peak at 1053 cm−1 disappeared, and the peaks appeared between
1200 and 1400 cm−1 belonged to the aromatic ring, indicating that the
aromatic structure was introduced into MCB, which were in good
consistent with XRD results. Moreover, a new peak was observed at
1240 cm−1, allocated to the C–N stretch vibration, possibly due to the
introduction of nitro, amino, pyridine, and pyrrole groups [26]. In all, a
large number of active functional groups were produced after series
treatment, which could provide electrons for Cd(II) adsorption and
promote adsorption performance.
The XPS survey spectra of the CB and MCB, XPS C1s spectra of CB
and MCB, and XPS N1s spectra of MCB were shown in Fig. 2(C–E),
respectively. XPS survey spectra (Fig. 2C) revealed that nitrogen atoms
were successfully incorporated into the carbon skeleton of MCB, the
content of elements was summarized in Tables 1, while S1 listed the
detailed content of O and N functionalities, which obtained from the
C1s and N1s spectra. The C1s spectrum of CB comprised two peaks with
a binding energy of 284.6 (C1) and 286.7 eV (C2), the C1 peaks could
be allocated to the C–C, C]C, and C–H, and the C2 peaks could be
assigned to C–O–H and C–O–C, respectively. New peaks with a binding
energy of 289.0 eV (C3) were found in the C1s spectra of MCB, which
corresponded to the newly formed carboxyl groups. Due to the Cd(II)
1004
Chemical Engineering Journal 350 (2018) 1000–1009
complexation ability of carboxyl was stronger than that of hydroxyl,
thus it is beneficial to enhance the adsorption capacity by increasing
carboxyl content. The N1s spectrum of MCB showed 7 peaks with
binding energy of 398 eV (N1), 400 eV (N2), 401.3 eV (N3), 402.2 eV
(N4), 404.2 eV (N5), 406.1 eV (N6) and 408.2 eV (N7), corresponding
to pyridine N, pyrrole N, tetravalent N, amino N, nitrogen oxides N,
chemisorbed N and nitrate N, respectively. Identical conclusion was
obtained in the study of FTIR. The binding energy of same functional
groups in the three MCB may be slightly deviated, but the overall
species are basically the same [27,28]. The pyridine N, pyrrole N,
amino N and nitrate N could complex with Cd(II), and the total content
of these groups in MCB decreased with the increase of carbonization
time.
Based on the above characterization and analysis, three MCB ad-
sorbents were successfully prepared with different surface area and
morphology, O and N functionalities content, which might provide lone
pair electrons for the Cd(II) adsorption and would benefit the Cd(II)
complexation with MCB.
3.2. Cd(II) adsorption performance
The Cd(II) adsorption performance of the adsorbent was heavily
affected by various interference factors and their own surface structure.
The effects of adsorbent dose and pH value on the adsorption capacity
Q. Chen et al. Chemical Engineering Journal 350 (2018) 1000–1009

Fig. 3. Effect of adsorbent dose (A) and pH value (B) on the adsorption capacity of CB and MCB, recycle Cd(II) adsorption performance of MCB-1h (C).

Table 2 that of CB. The adsorption ability of MCB was significantly enhanced
The relative functional groups content of the adsorbents and their Cd(II) after series treatments, which increased the specific surface area, pore
binding energies. volume, and active adsorption functional groups. The maximum Cd(II)
Matrix –OH –COOH –NO2 –NH2 –C5H5N –C4H5N adsorption capacity of MCB-1h was slightly higher than that of MCB-2h
and MCB-3h, also higher than most of reported adsorbents (listed in
CB 27.7 – – – – – Table S2). The reason might be that the specific surface area and pore
MCB-1h 21.1 5.3 4.2 20.0 5.1 19.1
volume of MCB-1h were a bit larger, which was in favor of the contact
MCB-2h 11.5 3.1 22.1 13.1 11.1 10.3
MCB-3h 17.3 3.2 16.3 7.6 13.3 12.4
of the adsorbent with Cd(II). Otherwise, the contents of carboxyl, pyr-
idine pyrrole and amino groups in MCB-1h were relatively higher, and
Binding energy −32.45 −41.04 −39.78 −61.11 −48.03 −68.49
(kcal/mol)
enhanced the electron donor ability of the adsorbent, which was ben-
eficial to the Cd(II) adsorption.
To gain more insights of structure-performance relationship in Cd
of CB and MCB were evaluated by single factor experiment, and the (II) adsorption onto MCB, we calculated the Cd(II) binding energy of O
adsorption performance of CB and MCB-1–3h was compared through and N functionalities modified MCB based on DFT, and the results were
parallel experiments, the results were displayed in Figs. 3 and S2. The listed in Table 2. The MCB adsorbents were highly aromatized mate-
CB and MCB showed the same trend of Cd(II) adsorption capacity, so rials, and their structure was similar to that of graphene, which is
we just focused on MCB in the due discussions. composed of aromatic rings. In order to simplify the simulate calcula-
The adsorbent dose is critical for evaluating the effectiveness of the tions, we employed a simple graphene structure composed of seven
adsorbent, and could directly determine the economic impact of the aromatic rings as the MCB model compound to calculate the binding
adsorption process. Fig. 3(A) showed the Cd(II) adsorption capacity of energies of different functional groups modified MCB and Cd(II)
CB and MCB with various adsorbent dose from 2 to 100 mg in 20 mL [29,30]. The Cd(II) adsorption equilibrium configuration images were
stock solution, the contact time was 1 h, temperature was 25 °C, Cd(II) shown in Fig. S3. The bigger absolute value of binding energy, the
initial concentration was 200 mg/L, original pH (5.5). For MCB, the stronger Cd(II) binding ability of the adsorbent. The Cd(II) binding
adsorption capacity decreased with the increase of adsorbent dose, ability of the adsorbents was in the following order: C4H5N-MCB >
owing to the decrease of adsorbed Cd(II) by per unit mass adsorbent. NH2-MCB > C5H5N-MCB > COOH-MCB > NO2-MCB > OH-MCB.
When the adsorbent dose was 2 mg, the adsorption capacity of CB and The N-containing functional groups modified MCB showed stronger
MCB-1–3 h was 102 mg/g, 368 mg/g, 358 mg/g and 346 mg/g, re- binding ability than those of O-containing functional groups modified
spectively. Taking the comparatively high adsorption capacity of MCB- MCB, due to the electronegative of O is stronger and less capable to
1–3h at 20 mg adsorbent dose into account, and to avoid the waste of provide electrons. The COOH-MCB had higher O content and stronger
adsorption sites and spaces, the adsorbent dose was set as 20 mg. electron-donating ability than those of OH-MCB, so its Cd(II) binding
The surface properties, protonation of functional groups, and dis- ability was stronger. The O content of MCB was higher than that of CB,
tribution of adsorbed metal ions on adsorbent were significantly af- resulting in a stronger Cd(II) binding ability and higher adsorption
fected by pH value. Adsorption properties were not tested for Cd(II) at capacity. The structure of C4H5N-MCB and C5H5N-MCB was similar, but
initial pH above 7, wherein the formation of metal hydroxide was showed a bit difference in binding ability, owing to the higher charge
supposed to precipitate. The effect of pH value on adsorption capacity density of five-membered ring six-electron structure for C4H5N-MCB
(Fig. 2B) was studied by varying pH value from 2 to 7, the solution than that of six-membered ring six-electron structure for C5H5N-MCB.
volume was 20 mL, the adsorbent dose was 20 mg, contact time was 1 h, The lone pair of electrons on NH2-MCB also was more easily captured
Cd(II) initial concentration was 200 mg/L, temperature was 25 °C. by Cd(II), so the Cd(II) binding ability was comparatively stronger.
When the pH value was in the range of 2.0–4.0, the adsorption capacity Moreover, the CB only had hydroxyl groups, which Cd(II) binding
increased obviously with the pH value. A possible explanation is that ability was the weakest, resulting in the lowest Cd(II) adsorption ca-
the H+ concentration was much higher than that of Cd(II) at pH < 4.0, pacity of CB. In the O-containing groups modified MCB, the COOH-MCB
and it competed with Cd(II) for the adsorption sites and spaces of the exhibited stronger binding ability than OH-MCB, and the MCB-1h had
adsorbents, leading to a lower adsorption capacity. With the increase of the highest relative –COOH content. Furthermore, the –NH2 and
pH, H+ quantity decreased, more adsorption sites were vacated for Cd –C4H5N groups content in MCB-1h was relatively higher, and the
(II), resulting in an increase of adsorption capacity. Considering that the binding energy of the two functional groups modified MCB was much
adsorption capacity was comparatively high at original pH, we didn’t lower than the other groups modified MCB, thus the MCB-1h had the
adjust the pH value in the following single factor tests. highest adsorption capacity of Cd(II). Most of all, the DFT calculation
In conclusion, the adsorption performance of the adsorbents was in results were well consistent with the experimental data.
the order of MCB-1h > MCB-2h > MCB-3h > CB. The maximum The recycling performance is also a key factor to evaluate an ad-
adsorption capacity for Cd(II) of MCB was about 4 times higher than sorbent, what really matters the cost of practical application. In this

1005
Q. Chen et al.
study, the strong complexing agent EDTA was selected as desorption
agent to study the influence of cycling number on Cd(II) adsorption
capacity of MCB-1h [31], and the results were shown in Fig. 3(C). The
mass concentration of the adsorbent was 1 mg/mL, contact time was
1 h, Cd(II) initial concentration was 200 mg/L, the temperature was
25 °C, original pH (5.5). The Cd(II) adsorption capacity decreased from
154.02 mg/g to 137.10 mg/g after 5 cycles. The declining activity may
arise from the consumption of washing process, changes in the physical
and chemical properties including reduction of specific surface area and
pore size, and weakening of surface functional groups [32]. However,
the Cd(II) adsorption capacity of the adsorbent after 5 cycles remained
89% to that of the fresh adsorbent, indicating that the MCB was a
suitable Cd(II) adsorbent with good regeneration ability and adsorption
performance.
3.3. Microscopic adsorption mechanism analysis
To shed light on the micro-mechanism of Cd(II) adsorption onto
MCB, the adsorption mechanism was studied and analyzed by the
combination of classical theory (adsorption thermodynamics and ki-
netics) and electron-scale (charge density and differential charge).
3.3.1. Adsorption thermodynamic
The mechanism of Cd(II) adsorption onto MCB was studied in the
view of adsorption isotherms and thermodynamic quantities. The ad-
sorption isotherm was used to describe the interactions between equi-
librium adsorption amount and concentration at a certain temperature,
and to analyze the adsorption interfacial phenomenon. The thermo-
dynamic quantities were chosen to study the adsorption process and
type. The Langmuir and Freundlich isotherm models were studied [33],
where the Langmuir model was suitable for monolayer sorption systems
[34], while the Freundlich model was appropriate for multi-layer ad-
sorption or high adsorbate concentration system [35]. The adsorption
isotherms were investigated by varying Cd(II) initial concentration
from 10 to 500 mg/L at a temperature of 25 °C (Fig. S2(A)). The fitting
parameters derived from isotherm models were displayed in Table 3,
and the fitting curves were shown in Fig. 4(A) and (B). Based on the
effects of temperature on Cd(II) adsorption performance varying from
10 to 40 °C (Fig. S2(B)), the adsorption Gibbs free energy (ΔG) were
calculated directly [36], and adsorption enthalpy (ΔH) and adsorption
entropy (ΔS) were derived from the intercept and slope of the linear
relationship between ln KD and 1/T (Fig. 4(C)), all data were given in
Table 4. All the related equations were listed in the Supporting
Information.
As reflected with correlation coefficients (R2) in Table 3, the Cd(II)
adsorption onto MCB can be better described by the Freundlich model
compared to the Langmuir model, while the Cd(II) adsorption onto CB
was more suitable for Langmuir model. The results suggesting that the
Cd(II) adsorption on CB was monolayer adsorption, and on MCB was
multi-layer adsorption. Moreover, the value of 1/n was smaller than 1
and the RL values were in the range of 0–1, indicating that Cd(II) was
favorably adsorbed onto MCB [37]. In addition, the KF values of MCB
were higher than those of CB, further confirmed the stronger adsorption
ability of MCB.
Table 3
Adsorption isotherm parameters of Cd(II) adsorption onto CB and MCB at 25 °C.
Sample Langmuir isotherm model Freundlich isotherm model
qmax b (L/mg) R2
RL range 1/n KF
(mg/g)
CB 108.8 0.005 0.988 0.29–0.95 0.751 0.905
MCB-1h 458.7 0.016 0.768 0.11–0.86 0.634 13.91
MCB-2h 429.2 0.015 0.755 0.12–0.87 0.582 15.48
MCB-3h 414.9 0.013 0.807 0.13–0.89 0.634 11.14
Chemical Engineering Journal 350 (2018) 1000–1009
In the range of 10–40 °C, the adsorption Gibbs free energies (ΔG)
were negative and decreased with the increase of temperature, in-
dicating that the thermodynamic feasibility and spontaneity of the
adsorption process for the MCB despite the increase of environmental
temperature. At the same temperature, the ΔG values of MCB were
smaller than those of CB, suggesting the more spontaneous adsorption
of Cd2+ onto CB. The lowest ΔG value of MCB-1h indicated the highest
adsorption capability for Cd(II), owing to the biggest specific surface
area and more active adsorption functional group. Moreover, the po-
sitive values of adsorption enthalpy (ΔH) suggested that the adsorption
of Cd(II) was endothermic, which was consistent with the effect of
temperature on Cd(II) adsorption capacity (Fig. S2B). This is because
that the elevated temperature was conducive for Cd(II) to overcome the
steric hindrance, and thereby accelerating the adsorption process. In
addition, the positive values of adsorption entropy (ΔS) suggested the
increased disorder in the system during the process of adsorption, and a
good Cd(II) affinity of the adsorbents. The ΔS and ΔH values of MCB
were bigger, and the ΔG values were smaller than those of CB, in-
dicating that the Cd(II) affinity of MCB was stronger than CB [38].
3.3.2. Adsorption kinetic
To achieve a better understanding of the mechanism of Cd(II) ad-
sorption onto MCB and CB. Several kinetic models such as the pseudo-
first-order kinetics [39], pseudo-second-order kinetics [40], and Elo-
vich model were employed [41], and the effects of contact time on the
Cd(II) adsorption performance (Fig. S2(C)) were fitted. The kinetic
fitting parameters were listed in Tables 5 and S2, and the fitting curves
of pseudo-second-order model, pseudo-first-order model, and Elovich
model were shown in Figs. 4(D–F), and S4. All the related equations
were listed in the Supporting Information.
The correlation coefficient (R2) of pseudo-first-order kinetic were in
the range of 0.94–0.99, and the calculated equilibrium adsorption ca-
pacity qe (cal) were not consistent with the experimental equilibrium
adsorption capacity qe (exp), suggesting that the pseudo-first-order
model was inappropriate to describe the adsorption process. In contrast,
the R2 values of the pseudo-second-order were almost all bigger than
0.990 at each C0, and the calculated qe (cal) values well agreed with qe
(exp). Thus, it can be concluded that the adsorption of Cd(II) on MCB
followed the pseudo-second-order kinetic model, and the adsorption
process was mainly donated by chemisorption. They were mainly
caused by the electron sharing or exchange between Cd(II) and the
functional group in the adsorbents [40]. The results of Elovich model
suggested that the formation of complexes in the adsorption process,
mainly contributed by the carboxyl, pyridine and pyrrole groups in the
MCB, owing to the strong electron-donating ability of these groups. The
α values of MCB were larger than those of CB, and the β values were
smaller, indicating that MCB has more active adsorption sites and
stronger electron-donating ability. Compared the FTIR and XPS survey
spectra of MCB-1h (Fig. S5), the characteristic peaks of C]O
(1600–1750 cm−1) and C–O (1053 cm−1) significantly shifted to bigger
wavenumbers after Cd(II) adsorption, and a new peak was observed at
461 cm−1, allocated to the Cd–O stretch vibration. These shifts in-
dicated that the mondentate and bidentate chelation of Cd(II) and the
functional groups [41]. Moreover, a new peak with a binding energy of
405.0 eV was found in the pattern of Cd(II) adsorbed MCB-1h, also
confirming that the Cd(II) was successfully adsorbed by the MCB-1h.
The pseudo-second-order rate constant k2 of all the adsorbents de-
creased with the increase of Cd(II) initial concentration, due to the
R2
increase of mass transfer driving force and binding rate, and it is ben-
eficial to the adsorption process [42]. The k2 values of MCB were
smaller than those of CB, indicating the stronger adsorption affinity of
0.976 MCB for Cd(II). At the same, Cd(II) initial concentration, the k2 values
0.995
of MCB-1h were the smallest, which indicated the fastest binding rate of
0.990
0.998 Cd(II), owing to its biggest specific surface area and more active ad-
sorption functional groups.
1006
Q. Chen et al. Chemical Engineering Journal 350 (2018) 1000–1009

Fig. 4. Langmuir (A) and Freundlich isotherm fitting curves (B) of Cd(II) adsorption onto CB and MCB at 25 °C, linear relationship between ln KD and 1/T fitting
curves (C), pseudo-second-order kinetic fitting curves of Cd2+ adsorption onto CB and MCB at C0 of 100 mg/L (D), 200 mg/L (E), 500 mg/L (F).

Table 4 adsorption process through DFT calculations [43,44]. The charge den-
Thermodynamic parameters of Cd(II) adsorption onto CB and MCB. sity of the adsorbent and its overlap degree with the electron cloud of
Parameters T (°C) CB MCB-1h MCB-2h MCB-3h Cd(II) had a direct effect on the binding ability. The Cd(II) adsorption
charge density images were shown in Fig. 5(A–F).
ΔG (kJ/mol) 10 −12.32 −16.70 −16.62 −16.51 Among all the adsorbents, the NH2-MCB exhibited highest charge
15 −12.86 −18.08 −17.99 −17.66 density and biggest electron cloud overlapped degree, the electron
20 −13.63 −19.18 −19.04 −18.81
25 −14.13 −20.11 −19.75 −19.45
cloud of Cd(II) overlapped with both the π electron of aromatic and the
30 −14.88 −20.80 −20.69 −20.45 lone pair electron of amino groups, resulting in a lower binding energy.
35 −15.60 −21.70 −21.49 −21.44 The adsorbents containing strong withdrawing electron groups such as
40 −16.34 −22.74 −22.68 −22.51 carboxyl and nitro groups exhibited lower charge density, the electron
ΔH (kJ/mol) 25.72 37.75 37.56 38.42
of Cd(II) only overlapped with the π electron of the aromatic, and far
ΔS (kJ/mol.K) 0.134 0.194 0.192 0.195
from the functional groups, resulting in a higher binding energy. The
charge density of C4H5N-MCB and C5H5N-MCB was higher, the location
Table 5 of the Cd(II) near the middle of the adsorbent, and most electrons of Cd
Pseudo-second-order model kinetic fitting parameters of Cd(II) adsorption onto (II) overlapped with the π electron of aromatic and only parts over-
CB and MCB. lapped with the N-containing heterocycles.
Fig. 5(G-L) showed the Cd(II) adsorption differential charge images
Adsorbent C0 (mg/ qe (exp) Pseudo-second-order model
L) (mg/g) of functional groups modified MCB, which was used for the analysis of
k2 (g mg−1 min−1) qe (cal) (mg/ R2 electron contribution in the adsorption process [45]. The blue part
g) means the missing of electron, and red means acquisition of electron of
MCB. The O atom in the carboxyl and hydroxyl groups and the N atom
CB 100 26.34 3.27 × 10−3 28.27 0.988
200 46.20 1.92 × 10−3 51.20 0.990 in the amino groups could provide electron for Cd(II) adsorption. But
500 72.80 1.13 × 10−3 82.78 0.989 the N atom provided electron with higher electron density, which fa-
MCB-1h 100 80.04 1.37 × 10−3 92.25 0.994
cilitated the adsorption of Cd(II), resulting in a stronger Cd(II) ad-
200 154.0 0.71 × 10−3 177.9 0.995 sorption ability. Although the nitro groups contained N atom, but parts
500 352.8 0.22 × 10−3 400.0 0.994 of electrons were captured by the two O atom, leading to a weaker
MCB-2h 100 78.90 1.39 × 10−3 90.42 0.993 binding ability. The C5H5N-MCB increased the amount of aromatic ring
200 148.7 0.75 × 10−3 174.5 0.992 and the density of π electron. However, the p-orbit where the lone pair
500 338.6 0.29 × 10−3 400.0 0.992 electrons of N atom from the aromatic ring couldn’t provide electrons
MCB-3h 100 75.55 1.76 × 10−3 81.97 0.994 for Cd(II) adsorption, thus colored red near N atom. While the C4H5N-
200 143.9 0.82 × 10−3 163.1 0.995 MCB also increased the amount of aromatic ring and the density of π
500 317.3 0.37 × 10−3 362.3 0.995 electron, but the lone pair electrons of N atom could participate in the
formation of conjugated ring, which could provide electron for Cd(II)
adsorption, so colored blue near N atom. Therefore, the Cd(II) binding
3.3.3. Electron-scale adsorption mechanism
ability of C4H5N-MCB was stronger than that of C5H5N-MCB.
The above characterizations, thermodynamic and kinetic analysis
The functional groups and aromatic could provide long pair elec-
indicated the existence of electron sharing or electron exchange phe-
trons and π electrons for the Cd(II) adsorption, respectively. The N
nomenon. Hence, we further analyzed the electron donation in the
functionalities such as amino, pyridine and pyrrole groups could raise

1007
Q. Chen et al.
Fig. 5. The Cd(II) adsorption charge density and differential charge images of functional groups modified MCB (A and G: OH-MCB; B and H: COOH-MCB; C and I:
NO2-MCB; D and J: NH2-MCB; E and K: C5H5N-MCB; F and L: C4H5N-MCB).
the adsorption ability of the biochar adsorbents.
4. Conclusion
In summary, we have successfully prepared three mesoporous cel-
lulose biochar adsorbents (MCB) with different physicochemical prop-
erties for heavy metal Cd(II) removal. The MCB exhibited pretty good
Cd(II) adsorption performance, the adsorption capacity of MCB-1h was
up to 368.8 mg/g, better than most of the reported biochar adsorbent,
and showed good recycling performance. The structure-performance
relationship in Cd(II) adsorption onto MCB results of experimental data
are supported by the DFT calculations. Moreover, we demonstrated a
new strategy for understanding heavy metal adsorption mechanism by
the combination of classical theory and electron-scale analysis, this
combined study provided a comprehensive image of the Cd(II) ad-
sorption mechanism onto biochar. The functional groups and aromatic
could provide lone pair electrons and π electrons for the Cd(II) ad-
sorption, respectively. The N functionalities such as amino, pyridine
and pyrrole groups could raise the adsorption ability of the biochar
adsorbents. In total, the future looks bright for the application of MCB
as Cd(II) adsorbent, and our mechanistic study provides valuable per-
spectives for the preparation of efficient heavy metal biochar ad-
sorbents.
1008
Chemical Engineering Journal 350 (2018) 1000–1009
Acknowledgments
This work was financially supported by the Science and Technology
Planning Project of Guangdong Province (No. 2016B020242004; No.
2015B020237008), China.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.cej.2018.06.054.
References
[1] Z. Li, Z. Ma, T.J. van der Kuijp, Z. Yuan, L. Huang, A review of soil heavy metal
pollution from mines in China: pollution and health risk assessment, Sci. Total
Environ. 468 (2014) 843–853.
[2] F.-J. Zhao, Y. Ma, Y.-G. Zhu, Z. Tang, S.P. McGrath, Soil contamination in China:
current status and mitigation strategies, Environ. Sci. Technol. 49 (2014) 750–759.
[3] Y.S. Ok, A.R. Usman, S.S. Lee, S.A. Abd El-Azeem, B. Choi, Y. Hashimoto, J.E. Yang,
Effects of rapeseed residue on lead and cadmium availability and uptake by rice
plants in heavy metal contaminated paddy soil, Chemosphere 85 (2011) 677–682.
[4] Z.B. Bouabidi, M.H. El-Naas, D. Cortes, G. McKay, Steel-making dust as a potential
adsorbent for the removal of lead(II) from an aqueous solution, Chem. Eng. J. 334
(2018) 837–844.
[5] X. Fang, J. Li, X. Li, S. Pan, X. Zhang, X. Sun, J. Shen, W. Han, L. Wang, Internal
pore decoration with polydopamine nanoparticle on polymeric ultrafiltration
membrane for enhanced heavy metal removal, Chem. Eng. J. 314 (2017) 38–49.
[6] S. Matsui, T. Kureha, S. Hiroshige, M. Shibata, T. Uchihashi, D. Suzuki, Fast
Q. Chen et al.
adsorption of soft hydrogel microspheres on solid surfaces in aqueous solution,
Angew. Chem. Int. Ed. 56 (2017) 12146–12149.
[7] L. Ling, W.-J. Liu, S. Zhang, H. Jiang, Achieving high-efficiency and ultrafast re-
moval of Pb(II) by one-pot incorporation of a N-doped carbon hydrogel into FeMg
layered double hydroxides, J. Mater. Chem. A 4 (2016) 10336–10344.
[8] Q. Lin, X. Bu, A. Kong, C. Mao, X. Zhao, F. Bu, P. Feng, New heterometallic zir-
conium metalloporphyrin frameworks and their heteroatom-activated high-surface-
area carbon derivatives, J. Am. Chem. Soc. 137 (2015) 2235–2238.
[9] M. Zhang, L. Song, H. Jiang, S. Li, Y. Shao, J. Yang, J. Li, Biomass based hydrogel as
an adsorbent for the fast removal of heavy metal ions from aqueous solutions, J.
Mater. Chem. A 5 (2017) 3434–3446.
[10] C. Liang, Z. Li, S. Dai, Mesoporous carbon materials: synthesis and modification,
Angew. Chem. Int. Ed. 47 (2008) 3696–3717.
[11] A. Yaumi, M.A. Bakar, B. Hameed, Reusable nitrogen-doped mesoporous carbon
adsorbent for carbon dioxide adsorption in fixed-bed, Energy 138 (2017) 776–784.
[12] D. Saha, S. Barakat, S.E. Van Bramer, K.A. Nelson, D.K. Hensley, J. Chen,
Noncompetitive and competitive adsorption of heavy metals in sulfur-functiona-
lized ordered mesoporous carbon, ACS Appl. Mater. Interfaces 8 (2016)
34132–34142.
[13] D. Sud, G. Mahajan, M. Kaur, Agricultural waste material as potential adsorbent for
sequestering heavy metal ions from aqueous solutions – a review, Bioresour.
Technol. 99 (2008) 6017–6027.
[14] V. Gupta, P. Carrott, R. Singh, M. Chaudhary, S. Kushwaha, Cellulose: a review as
natural, modified and activated carbon adsorbent, Bioresour. Technol. 216 (2016)
1066–1076.
[15] S. Hokkanen, A. Bhatnagar, M. Sillanpää, A review on modification methods to
cellulose-based adsorbents to improve adsorption capacity, Water Res. 91 (2016)
156–173.
[16] Q. Cheng, Q. Huang, S. Khan, Y. Liu, Z. Liao, G. Li, Y.S. Ok, Adsorption of Cd by
peanut husks and peanut husk biochar from aqueous solutions, Ecol. Eng. 87 (2016)
240–245.
[17] M. Ilangovan, V. Guna, S. Olivera, A. Ravi, H. Muralidhara, M. Santosh, N. Reddy,
Highly porous carbon from a natural cellulose fiber as high efficiency sorbent for
lead in waste water, Bioresour. Technol. 245 (2017) 296–299.
[18] M.I. Inyang, B. Gao, Y. Yao, Y. Xue, A. Zimmerman, A. Mosa, P. Pullammanappallil,
Y.S. Ok, X. Cao, A review of biochar as a low-cost adsorbent for aqueous heavy
metal removal, Crit. Rev. Environ. Sci. Technol. 46 (2016) 406–433.
[19] J. Wei, W. Sun, W. Pan, X. Yu, G. Sun, H. Jiang, Comparing the effects of different
oxygen-containing functional groups on sulfonamides adsorption by carbon nano-
tubes: experiments and theoretical calculation, Chem. Eng. J. 312 (2017) 167–179.
[20] K. Liu, S. Zhang, X. Hu, K. Zhang, A. Roy, G. Yu, Understanding the adsorption of
PFOA on MIL-101 (Cr)-based anionic-exchange metal–organic frameworks: com-
paring DFT calculations with aqueous sorption experiments, Environ. Sci. Technol.
49 (2015) 8657–8665.
[21] Q. Yuan, P. Li, J. Liu, Y. Lin, Y. Cai, Y. Ye, C. Liang, Facet-dependent selective
adsorption of Mn-doped α-Fe2O3 nanocrystals toward heavy-metal ions, Chem.
Mater. 29 (2017) 10198–10205.
[22] J.J. Ternero-Hidalgo, J.M. Rosas, J. Palomo, M.J. Valero-Romero, J. Rodríguez-
Mirasol, T. Cordero, Functionalization of activated carbons by HNO3 treatment:
influence of phosphorus surface groups, Carbon 101 (2016) 409–419.
[23] A. Shrotri, H. Kobayashi, A. Fukuoka, Air oxidation of activated carbon to syn-
thesize a biomimetic catalyst for hydrolysis of cellulose, ChemSusChem 9 (2016)
1299–1303.
[24] H. Zhao, J.H. Kwak, Z.C. Zhang, H.M. Brown, B.W. Arey, J.E. Holladay, Studying
cellulose fiber structure by SEM, XRD, NMR and acid hydrolysis, Carbohydr. Polym.
68 (2007) 235–241.
[25] M. Yu, J. Li, L. Wang, KOH-activated carbon aerogels derived from sodium car-
boxymethyl cellulose for high-performance supercapacitors and dye adsorption,
Chem. Eng. J. 310 (2017) 300–306.
[26] X. Luo, X. Lei, N. Cai, X. Xie, Y. Xue, F. Yu, Removal of heavy metal ions from water
by magnetic cellulose-based beads with embedded chemically modified magnetite
1009
Chemical Engineering Journal 350 (2018) 1000–1009
nanoparticles and activated carbon, ACS Sustainable Chem. Eng. 4 (2016)
3960–3969.
[27] D. Guo, R. Shibuya, C. Akiba, S. Saji, T. Kondo, J. Nakamura, Active sites of ni-
trogen-doped carbon materials for oxygen reduction reaction clarified using model
catalysts, Science 351 (2016) 361–365.
[28] L. Qie, W.M. Chen, Z.H. Wang, Q.G. Shao, X. Li, L.X. Yuan, X.L. Hu, W.X. Zhang,
Y.H. Huang, Nitrogen-doped porous carbon nanofiber webs as anodes for lithium
ion batteries with a superhigh capacity and rate capability, Adv. Mater. 24 (2012)
2047–2050.
[29] Y. Cao, J. Huang, Y. Li, S. Qiu, J. Liu, A. Khasanov, M.A. Khan, D.P. Young, F. Peng,
D. Cao, One-pot melamine derived nitrogen doped magnetic carbon nanoadsorbents
with enhanced chromium removal, Carbon 109 (2016) 640–649.
[30] P.K. Tamukong, W.D. Peiris, S. Kilina, Computational insights into CdSe quantum
dots' interactions with acetate ligands, Phys. Chem. Chem. Phys. 18 (2016)
20499–20510.
[31] H. Ge, J. Wang, Ear-like poly (acrylic acid)-activated carbon nanocomposite: a
highly efficient adsorbent for removal of Cd(II) from aqueous solutions,
Chemosphere 169 (2017) 443–449.
[32] T. Wang, W. Liu, N. Xu, J. Ni, Adsorption and desorption of Cd(II) onto titanate
nanotubes and efficient regeneration of tubular structures, J. Hazard. Mater. 250
(2013) 379–386.
[33] J. Fu, Z. Chen, M. Wang, S. Liu, J. Zhang, J. Zhang, R. Han, Q. Xu, Adsorption of
methylene blue by a high-efficiency adsorbent (polydopamine microspheres): ki-
netics, isotherm, thermodynamics and mechanism analysis, Chem. Eng. J. 259
(2015) 53–61.
[34] R. Mallampati, L. Xuanjun, A. Adin, S. Valiyaveettil, Fruit peels as efficient re-
newable adsorbents for removal of dissolved heavy metals and dyes from water,
ACS Sustainable Chem. Eng. 3 (2015) 1117–1124.
[35] O. Hamdaoui, E. Naffrechoux, Modeling of adsorption isotherms of phenol and
chlorophenols onto granular activated carbon: Part I. Two-parameter models and
equations allowing determination of thermodynamic parameters, J. Hazard. Mater.
147 (2007) 381–394.
[36] T. Maneerung, J. Liew, Y. Dai, S. Kawi, C. Chong, C.-H. Wang, Activated carbon
derived from carbon residue from biomass gasification and its application for dye
adsorption: kinetics, isotherms and thermodynamic studies, Bioresour. Technol.
200 (2016) 350–359.
[37] G. Zeng, Y. Liu, L. Tang, G. Yang, Y. Pang, Y. Zhang, Y. Zhou, Z. Li, M. Li, M. Lai,
Enhancement of Cd(II) adsorption by polyacrylic acid modified magnetic meso-
porous carbon, Chem. Eng. J. 259 (2015) 153–160.
[38] P. Wang, M. Cao, C. Wang, Y. Ao, J. Hou, J. Qian, Kinetics and thermodynamics of
adsorption of methylene blue by a magnetic graphene-carbon nanotube composite,
Appl. Surf. Sci. 290 (2014) 116–124.
[39] S. Lagergren, Zur theorie der sogenannten adsorption geloster stoffe, Kungliga
svenska vetenskapsakademiens, Handlingar 24 (1898) 1–39.
[40] L. Ma, S.M. Islam, C. Xiao, J. Zhao, H. Liu, M. Yuan, G. Sun, H. Li, S. Ma,
M.G. Kanatzidis, Rapid simultaneous removal of toxic anions [HSeO3]−, [SeO3]2–,
and [SeO4]2–, and metals Hg2+, Cu2+, and Cd2+ by MoS42– intercalated layered
double hydroxide, J. Am. Chem. Soc. 139 (2017) 12745–12757.
[41] P. Ramachandran, R. Vairamuthu, S. Ponnusamy, Adsorption isotherms, kinetics,
thermodynamics and desorption studies of reactive Orange16 on activated carbon
derived from Ananas comosus (L.) carbon, J. Eng. Appl. Sci. 6 (2011) 15–26.
[42] I. Tan, J. Chan, B.H. Hameed, L.L.P. Lim, Adsorption behavior of cadmium ions onto
phosphoric acid-impregnated microwave-induced mesoporous activated carbon, J.
Water Proc. Eng. 14 (2016) 60–70.
[43] R. Fei, L. Yang, Strain-engineering the anisotropic electrical conductance of few-
layer black phosphorus, Nano Lett. 14 (2014) 2884–2889.
[44] L. Yan, J. Song, T.-S. Chan, C. Jing, Insights into antimony adsorption on 0 0 1 TiO2:
XAFS and DFT study, Environ. Sci. Technol. 51 (2017) 6335–6341.
[45] X. Guo, X. Zhang, S. Zhao, Q. Huang, J. Xue, High adsorption capacity of heavy
metals on two-dimensional MXenes: an ab initio study with molecular dynamics
simulation, Phys. Chem. Chem. Phys. 18 (2016) 228–233.