Journal of Colloid and Interface Science 539 (2019) 553–562

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

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

Regular Article

Adsorption behavior and mechanism of Fe-Mn binary oxide

nanoparticles: Adsorption of methylene blue
Kun Lu a, Tingting Wang a, Li Zhai a, Wei Wu b, Shipeng Dong a, Shixiang Gao a, Liang Mao a,⇑
a
State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, Nanjing 210093, China
b
Dragonfly Agri (Jiangsu) Research Corp. LTD, Nanjing 210000, China

g r a p h i c a l a b s t r a c t
mechanism

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

Article history: Wastewater containing organic dyes has caused worldwide concern. It is thus imperative to develop
Received 16 October 2018 materials to remove organic dyes from wastewater. In this study, a nano-structured Fe-Mn binary oxide
Revised 18 December 2018 (nFMBO) was synthesized via a facile coprecipitation approach and used for methylene blue (MB)
Accepted 26 December 2018
removal from aqueous solution. Characteristic results indicated that the as-prepared nFMBO had a typical
Available online 27 December 2018
wrinkled structure. The adsorption performance of the nFMBO was then investigated by batch experi-
ments. The adsorption kinetics was well fitted to a pseudo-second-order kinetics model, and the adsorp-
Keywords:
tion isotherms agreed well with the Langmuir model with a maximum adsorption capacity of 72.32 mg/g
Nano-structure
Fe-Mn binary oxide
at 25 °C. Solution pH was a key factor for adsorption and the absorbent exhibited better removal effi-
Adsorption ciency for MB in solution with high pH. In addition, it was found that the investigated coexisting anions
Methylene blue (CO2– 2
3

3 , SO4 , PO3 ) did not have a significant influence on MB removal. More importantly, the nFMBO
Reusability could be easily separated from the water and regenerated by acid elution, and the adsorption efficiency
of the nFMBO only decreased to 85.1% of the initial capacity after five adsorption-regeneration cycles.
These results indicate that the nFMBO can become an alternative adsorbent for the removal of MB from
wastewater.
Ó 2018 Elsevier Inc. All rights reserved.

1. Introduction result in serious water pollution problems [1–3]. These dyes are
known as organic materials composed of aromatic compounds
Dyes and pigments released from various industries, such as the and contained phenyl, azo and amino groups, which make them
textile, printing, and leather industries, into the environment will highly toxic and difficult to degrade [4–6]. Methylene blue (MB)
is thought to be one of the most common aromatic-cationic dyes,
which is widely used in textile industry [7]. Previous studies have
⇑ Corresponding author.
demonstrated that exposure to MB could trigger skin damage and
E-mail address: lmao@nju.edu.cn (L. Mao).

https://doi.org/10.1016/j.jcis.2018.12.094
0021-9797/Ó 2018 Elsevier Inc. All rights reserved.
554 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562
burning sensation in the eyes, while its direct ingestion caused
increased heart rate, digestive disease, and methemoglobinemia
[8,9]. Therefore, it is urgently necessary to remove MB from
wastewater for reducing their threats to humans and ecosystems.
Various techniques, including photo-catalysis, adsorption, coagula-
tion and membrane treatment, have been adopted to solve the dye
pollution problems [10–16]. Among these treatment approaches,
adsorption is regarded as a promising and effective technique to
remove the dyes from water due to its advantages: a simple treat-
ment process, low cost, high efficiency and extensive applicability
[10,11,17,18]. However, conventional adsorbents may have a slow
adsorption rate, low adsorption capacity, and poor reusability
[19,20]. Therefore, it is necessary to develop novel high-
performance adsorbents to overcome the above challenges.
Among the currently available adsorbents, iron oxides are a
group of emergent nanomaterials that have been studied in depth
as adsorbent for the removal of environmental pollutants from
aqueous solutions because of their outstanding adsorption perfor-
mance, geological abundance and environmentally friendly prop-
erties [21–28]. Previous studies have demonstrated that,
compared to single iron oxides compounds, binary metal oxides
composites usually had superior properties, such as improved
adsorption capacity [22,29–31]. For example, it had been reported
that the introduction of high valence Mn into Fe-containing adsor-
bents enhanced its adsorption capacity [30]. Zhong et al. synthe-
sized a urchin-like Fe-Mn binary oxides with higher adsorption
capacity for Cd(II) [25]. Most of these studies focused on the
removal of heavy metals, with few studies on the removal of MB
from the aquatic environments [7,32,33]. It had been demon-
strated that the adsorption of heavy metal on the metal oxide
nanoparticles was mainly attributed to electrostatic interaction
[22,25]. As methylene blue (MB) is a typical cationic dye [7], we
speculated that the metal oxides composites could also be used
as an efficient adsorbent for the removal of MB.
The objective of this study was to explore the adsorption behav-
ior of MB on a nano-structured Fe-Mn binary oxide (nFMBO) adsor-
bent through a systemic investigation using batch experiments.
The influence of solution pH, ionic strength, co-anions and natural
organic matter (NOM) on MB adsorption process was also
explored. Furthermore, the reusability of the adsorbent was also
investigated to evaluate its applicability in real practice. Finally,
Fourier transform infrared spectroscopy (FT-IR) analysis was
performed to further elucidate the adsorption mechanism. This
study provides valuable insights into the development of mixed
metal oxide nanoparticles-based adsorbents for environmental
remediation.
2. Materials and methods
2.1. Materials
All chemical reagents used in this work were analytical grade or
better. Potassium permanganate (KMnO4, >99.5%) and ferrous sul-
fate heptahydrate (FeSO47H2O, >99.7%) were purchased from Nan-
jing Chemical Reagent Co. Ltd., China. Methylene blue (MB) was
purchased from Sigma-Aldrich. Suwannee River natural organic
matter (NOM) was obtained from the International Humic
Substances Society and its elemental composition is provided in
Table S1.
2.2. Synthesis of the nFMBO
The adsorbent was synthesized according to a modified litera-
ture method [25]. In a typical procedure, KMnO4 dissolved in
100.0 mL deionized water (0.075 mol/L) were heated to 373 K
and simultaneously stirred at 400 rpm using a heater/magnetic
stirrer, and then 100.0 mL of a 0.225 mol/L FeSO47H2O solution
was slowly added into the KMnO4 solution. Fifteen mL of a
5 mol/L NaOH solution was thereafter added drop-wise into the
boiling mixture. After that, the mixture was stirred continuously
for 6 h without heating, and then the mixture was kept for another
6 h without stirring. The precipitate was obtained by centrifuga-
tion, washed several times using de-ionized water to remove the
impurities, and then freeze-dried until a constant weight was
obtained.
2.3. Characterization of the nFMBO
The crystal phases of the nFMBO were analyzed by powder
X-ray diffraction (XRD) using a Bruker D8 Advance (Bruker AXS,
Germany) with Cu Ka radiation at a setting of 40 kV and 40 mA
and angular variation of 10–80. Scanning electron microscopy
(SEM) and energy dispersive spectroscopy (EDS) were carried out
with a QUANTA FEG 250 scanning electron microscope to analyze
the surface morphology and element composition of the nFMBO.
Transmission electron microscopy (TEM) was carried out on a
JEM-200CX. X-ray photoelectron spectroscopy (XPS) was per-
formed on a PHI 5000 VersaProbe with a monochromatic Al Ka
X-ray source. The specific surface area of the obtained nFMBO
was measured using an ASAP 2020 instrument (Micromeritics
Instrument Co., Norcross, GA, USA) at 77 K, and then calculated
by the Brunauer-Emmett-Teller (BET) method. The surface f poten-
tial of the nFMBO was determined by the immersion technique
using a Malvern Nano ZS instrument [34]. FT-IR spectroscopy anal-
ysis of the nFMBO, MB and MB-adsorbed nFMBO were performed
by a Vertex 70v spectrometer (Bruker, Germany) with a resolution
of 4 cm
1.
2.4. Evaluation of adsorption performance
Sorption experiments were carried out using a batch adsorption
approach as reported in our earlier studies [35,36]. Generally, all
MB experimental samples were added at pre-determined concen-
trations in 250 mL glass bottles by dilution of the stock solution
using a background electrolyte (NaCl at 0.01 mol/L), and then
adjusting the pH by the addition of 0.1 mol/L NaOH or HCl in water.
The adsorbent concentration used was 1.0 g/L. The mixture was
placed into a rotary shaker with a speed of 170 rpm at room tem-
perature. After sorption experiments were completed, samples
were removed, centrifuged at 3000 rpm for 5 min, and then the
residual MB concentration in the supernatant was determined
using a UV/Vis spectrophotometer at 664 nm [33].
Adsorption kinetic experiments were performed with an initial
MB concentration of 100 mg/L at 25 °C and pH 7.0 to explore the
sorption behavior of MB on nFMBO (1.0 g/L) across time. At pre-
specified intervals (0–12 h), the concentration of MB was mea-
sured. The quantity of MB adsorbed (qt (mg/g)) at time t was calcu-
lated according to the concentration difference between initial
solution concentration (Ci) and final solution concentration (Cf),
as shown in the following equation:
ðC i
C f ÞV
qt ¼ ð1Þ
m
in which V (L) is the volume of the solution and m (g) is the mass of
the absorbent.
Isothermal adsorption experiments of MB on the nFMBO were
conducted by adding a different MB concentration ranging from
10 to 100 mg/L, qe (mg/g) was calculated by the following
equation:
 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562
ðC i
C e ÞV
qe ¼
m glass conical bottles, and then 100 mg nFMBO was added into
in which Ce is the equilibrium concentration of MB (mg/L).
Solution pH, ionic strength, co-ions and natural organic matter
(NOM) were varied to evaluate the adsorption capacity of the
nFMBO in different aquatic quality conditions. The effect of solu-
tion pH on the adsorption of MB by the nFMBO was performed
with an initial MB concentration of 50 mg/L and the solution pH
ranged from 3.0 to 11.0 at room temperature. The influence of
the ionic strength on the adsorption of MB by the nFMBO was car-
ried out by adding NaCl or CaCl2 with concentration ranging from
0.1 to 1.0 mol/L into 50 mg/L MB solutions at room temperature
with a solution pH of 7.0. Moreover, three common co-anions,
namely CO2– 2
3

3 , SO4 and PO4 were used to investigate the effect
of co-ions on the adsorption performance. The presence of NOM
on the MB removal was also studied with various concentrations
of NOM (0–10 mg/L, TOC). The removal efficiency of MB was calcu-
lated according to the following equation:
ðC 0
C t Þ
g% ¼  100%
C0
where g was the removal efficiency of MB, C0 was the initial
MB concentration (mg/L) and Ct was the final MB concentration
(mg/L) after a certain period of time.
2.5. Reusability of the nFMBO
The reusability of the adsorbent is one of the most important
factors to assess its potential use [37]. Therefore, regeneration
experiments were conducted to evaluate the reusability of the
Fig. 1. Characterization of the nFMBO. (a) XRD patterns of sample, (b) SEM image of sample, (c) Histogram of the size distribution for Fe-Mn binary oxides (n = 200, n is the
number of particles) (d) EDS pattern of the Fe-Mn binary oxides.
555
nFMBO. 100 mL MB solution (50 mg/L) was prepared in 250 mL
ð2Þ
the solution. The mixture was placed into a rotary shaker at
170 rpm for 12 h. After that, the mixture was filtered using vacuum
filtration to separate the adsorbent from the solution. The MB-
loaded nFMBO was regenerated by shaking in 0.1 mol/L HNO3 for
30 min, and then by washing for three times using de-ionized
water. Finally, the regenerated nFMBO was used again in the suc-
ceeding cycles. Moreover, the leaching of Fe and Mn from the
nFMBO were measured using ICP-MS (NexION 300, Perking Elmer,
USA).
3. Results and discussions
3.1. Characterization of the nFMBO
The phase purities and crystallographic structure of the as-
prepared nFMBO were investigated by XRD. As shown in Fig. 1a
the diffraction peaks for the nFMBO in the 2h range from 10 to
ð3Þ 70° were indexed to the pure orthorhombic phase of FeOOH (JCPDS
29-713) [38]. However, no obvious diffraction crystalline peaks of
manganese oxides were detected, which was mainly attributed
to the fact that most of manganese oxides existed in amorphous
form in the as-prepared nFMBO [29].
Additionally, SEM analysis and EDS analysis were performed to
further identify the successful synthesis of the composite nFMBO.
Fig. 1b showed the SEM micrograph of the as-prepared nFMBO,
where it was observed that the nFMBO was composed of many
uniform three-dimension sphere and the surface of nFMBO was
wrinkle and hierarchical, which increased its specific surface area
556 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562

(Fig. S1). The surface of the nFMBO was further observed by TEM,
and it was found that the sphere was surrounded by many irregu-
lar branching structures (Fig. S2). In order to figure out the size dis-
tribution of the nFBMO, we randomly selected about two hundred
spheres from the SEM images and analyzed their diameters using a
Gaussian distribution. The result was presented in Fig. 1c. It was
found that more that 80% of the total count was in the range of
450–550 nm (average diameter: 496.7 ± 63.2 nm), which demon-
strated that the size of the prepared nFMBO was relatively uniform.
Furthermore, the EDS line scan spectra and elemental mappings
can characterize the surface element composition and the distribu-
tion of element of the sample. From Fig. 1d, the EDS spectrum
depicted that the prepared nFMBO contained Fe, Mn and O, and
the molar ration of Fe/Mn on surface is about 3:1, which matched
well with the amount of precursors used in the synthesis process.
EDS elemental mappings (Fig. 2a–d) further confirmed that the
heterogeneous structure of the nFMBO which is composed of three
elements: Fe, Mn and O.
X-ray photoelectron spectroscopy was used to examined the
oxidation states of the iron and manganese in the nFMBO. In the
high-resolution Fe2p XPS spectrum, as shown in Fig. 3a, the peaks
of Fe2p1/2 and Fe2p3/2 located at 724.8 and 711.1 eV were the char-

Fig. 2. (a) SEM-EDS elemental mapping of the prepared nFBMO, (b)–(d) the EDS elemental mappings of O, Fe and Mn, respectively.

4500 1300
(a) (b)
4000 Mn2p3/2
Fe2p3/2
Fe2p1/2 1250
Intensity (a.u.)

3500
Intensity (a.u.)

Mn2p1/2
3000 1200

2500 1150
2000
1100
1500

1000 1050
740 735 730 725 720 715 710 705 700 660 655 650 645 640 635
Binding Energy (ev) Binding Energy (ev)
Fig. 3. The high-resolution XPS spectra of Fe2p (a) and Mn2p (b) in the nFMBO.
 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562 557

(a) (b)
120 20

Quantity Adsorbed (cm /g STP)

0.010

dV/dD (cm3/g ⋅nm)

0.008 15
100
0.006 10

Zeta Potential (mV)

3

80 0.004
5
0.002
60 0
0.000
0 10 20 30 40 50
Pore Diameter (nm) -5
40
-10
20 Adsorption -15
Desorption
0 -20
0.0 0.2 0.4 0.6 0.8 1.0 3 4 5 6 7 8 9
Relative Pressure (P/Po) pH
Fig. 4. (a) Nitrogen adsorption and desorption isotherms of the nFMBO with the corresponding pore size distribution (inset) calculated by the BJH method from the
desorption branch. (b) Variation of zeta potential with solution pH for the nFMBO at 25 °C.

acteristic positions of a-FeOOH, indicating the existence of a- in which qe (mg/g) and qt (mg/g) were the amount of MB adsorbed
FeOOH in the nFMBO [29]. The binding energies at 653.3 and on the nFMBO at equilibrium and at the sampling time (t), and k1
641.6 ev were assigned to Mn2p1/2 and Mn2p3/2 (Fig. 3b), which was the pseudo-first-order-rate constant.
matched well with the characteristic peaks of MnO2 [29]. These The pseudo-second-order model can be expressed as the fol-
results demonstrated that the iron and manganese in the nFMBO lowing form:
were in the oxidation states +III and +IV, respectively.
t 1 1
The specific surface area of the nFMBO was analyzed by Nitro- ¼ þ t ð5Þ
gen adsorption and desorption. As shown in Fig. 4a, the obtained qt k2 qe 2 qe
nitrogen adsorption isotherm demonstrated a typical type IV iso- in which qe (mg/g) and qt (mg/g) were the amount of MB adsorbed
therm with H3 hysteresis loop, indicating a characteristic distribu- on the nFMBO at equilibrium and at the sampling time (t), and k2 is
tion of slit-shaped mesopores. According to the BET model, the the pseudo-second-rate constant.
obtained nFMBO had a specific surface area of 67.5 m2/g. More- The kinetic parameters and correlation coefficients for the
over, based on the BJH method, the obtained nFMBO had a pore adsorption of MB on the nFMBO were presented in Table 1. The
volume of 0.15 m3/g and a pore diameter of 12.65 nm. The high value of the correlation coefficient (R2) for the pseudo-second-
specific surface area and pore volume suggested that the nFMBO order model was 0.99, which was much higher than the one
may be a promising candidate for MB adsorption. Furthermore, obtained from the pseudo-first-order model. Moreover, the calcu-
we determined the variation of its zeta potential with the change lated adsorption capacity (qe,cal) obtained from the pseudo-
of solution pH, and found that the estimated point of zero charge second-order model was also closer to the experimental adsorp-
(PZC) for the nFMBO was around 6.2, as presented in Fig. 4b.

3.2. MB adsorption kinetics 100

3.2.1. Effect of contact time

The adsorption kinetics of MB on the nFBMO across time was 80
studied with an initial MB concentration of 100 mg/L (Fig. 5). It
was found that the adsorption rate of MB on the nFMBO increased
quickly during the first 1 h, and that the equilibrium were achieved 60
after 2 h. It was noteworthy that the nFMBO exhibited a high
qt (mg/g)

adsorption capacity of 80.13 mg/g for the initial MB concentration

of 100 mg/L. The adsorbent prepared in this study possesses a
40
wrinkled structure, which is expected to become a promising
adsorbent for rapid and effective treatment of the MB-containing
wastewater due to its features of fast adsorption and high capacity.
20
3.2.2. Kinetics models
In this study, two common kinetic models, namely pseudo-first-
0
order and pseudo-second-order, were used to further explore the
0 2 4 6 8 10 12
mechanism of the adsorption process of MB on the nFBMO [39].
Time (h)
The pseudo-first-order model can be expressed as the following
form: Fig. 5. Adsorption kinetic of methylene blue on the nFMBO. Experimental condi-
tions: [nFMBO]0 = 1.0 g/L, [MB]0 = 100 mg/L], T = 25 °C, pH = 7.0. The dotted line is
k1
log ðqe
qt Þ ¼ logqe
t ð4Þ non-linear curve fitting by the pseudo-first-order model; the solid line is non-linear
2:303 curve fitting by pseudo-second-order model.
558 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562

Table 1
Parameters of adsorption kinetics equation.

Pseudo-first-order Pseudo-second-order
2
qe,exp (mg/g) k1 (1/h) qe,cal (mg/g) R k2 (1/h) qe,cal (mg/g) R2
80.13 10.54 68.99 0.59 12.63 78.76 0.99

tion capacity (qe,exp). These results demonstrated that the kinetics
experimental data were better fitted to the pseudo-second-order
model, indicating that the adsorption process was likely to be a
chemisorption process.
3.3. MB adsorption isotherms
The adsorption isotherm experiments were carried out with the
initial MB concentration ranging from 10 to 100 mg/L to under-
stand the interaction between the MB and the adsorption sites of
the nFMBO. The adsorption time was 12 h, which was an adequate
time for equilibrium to occur (the equilibrium were achieved after
2 h). As shown in Fig. 6, it was obvious that the adsorption capacity
of the nFMBO increased with the increasing equilibrium concen-
tration of MB and reached saturation progressively.
Two equilibrium isotherm models, namely the Langmuir iso-
therm and the Freundlich isotherm [40], were used to fit the exper-
imental data. The Langmuir isotherm model and the Freundlich
isotherm model were presented in Eqs. (6) and (7), respectively.
K L qm C e
qe ¼
1 þ K LCe
1
log qe ¼ log Ce þ log KF
n
where qe (mg/g) was the adsorption capacity of MB at equilibrium.
Ce (mg/L) was the concentration of MB when the sorption process
reached equilibrium. qm (mg/g) represented the maximum adsorp-
tion capacity of MB and KL was a constant according to the Lang-
muir isotherm model. KF and n were respectively the adsorption
constant and linearity index according to the Freundlich isotherm
model.
The parameters from two adsorption isotherm models were
listed in Table 2. The value of correlation coefficient (R2) was used
to assess the validity of isotherm models. From Table 2, it was
observed that the adsorption of MB was better fitted to the Lang-
muir model than Freundlich model due to the higher correlation
coefficient value (R2 = 0.96), which indicated that the adsorption
of MB on the surface of the nFMBO took place in a monolayer
adsorption manner. In addition, the maximum sorption capacity
calculated according to the Langmuir model was 72.32 mg/g,
which was close to the experimental data.
Table 3 showed the comparison between the nFMBO prepared
in this study and other Fe or Mn oxide nanoparticles previously
used for the adsorption of MB from aqueous solutions. It was
ð6Þ clearly observed that the adsorption capacity of MB on the nFMBO
was higher than many other previously reported adsorbents,
which indicated that the nFMBO prepared in this study had great
ð7Þ application potential for the removal of MB in practical wastewater
treatment.

3.4. Influence of solution pH on MB sorption

100
Solution pH is one of the important parameters that greatly
affect the feasibility of environmental remediation involving to
80 adsorption treatment because it influences the surface charge
and binding sites of the adsorbents as well as the ionization of
the adsorbates [48]. Therefore, the effect of solution pH on the
60 removal of MB by the nFMBO was studied. From Fig. 7a, it was
observed that the removal of MB adsorbed by the nFMBO increased
qe (mg/g)

with the solution pH increasing from 3.0 to 7.0 and plateaued in

40 the pH range of 7.0–11.0. For example, only 17.6% of MB was
adsorbed by the nFMBO when the solution pH was 3.0, while at
solution pH 7.0–11.0, the MB adsorption rate reached nearly 92%.
20 The estimated point of zero charge (PZC) for the nFMBO was
around 6.2, as presented in Fig. 4b. Therefore, at lower solution
pH (pH < PZC), the surface charge of the nFMBO was positive, so
0 stronger electrostatic repulsion would exist between the positively
0 3 6 9 12 15 charged the nFMBO and MB (a kind of cationic dye), which inhib-
Ce(mg/L) ited the adsorption of MB on the nFMBO. Additionally, in acidic pH,
there existed competitive adsorption between the excess H+ and
Fig. 6. Adsorption isotherm of MB on the nFMBO. Experimental conditions: MB. With the solution pH increasing, however, the surface charge
[nFMBO]0 = 1.0 mg/L, T = 25 °C, pH = 7.0, and adsorption time = 12 h. The solid line of the nFMBO was negative and the electrostatic repulsive force
represents non-linear curve fitting by Langmuir model; the dotted line represents
became weaker, enhancing the adsorption of MB on the surface
non-linear curve fitting by Freundlich model.
of the nFMBO. Based on the above results, it was concluded that

Table 2
Parameters of adsorption isotherm model.

Langmuir Isotherm Freundlich Isotherm

qe,exp (mg/g) KL R2 1/n KF R2
72.32 3.57 0.96 0.23 45.73 0.94
 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562 559

Table 3
Comparison of maximum MB adsorption capacities of Fe or Mn oxide nanoparticles.

Adsorbents Experimental Conditions qmaxa (mg/g) Cib (mg/L) Reference

Dosage (g/L) T (°C) pH
Fe3O4@SiO2-Congo red 0.3 25 11.0 31.44 27 [27]
Fe3O4@MIL-100(Fe) 1.0 25 7.0 49.41 60 [41]
chitosan/Fe3O4/graphene 0.4 25 7.0 30.10 5.0 [26]
Fe3O4@MWCNTs 0.4 25 7.0 48.06 30 [42]
Fe3O4@Graphene 0.4 25 7.0 45.27 – [23]
Fe3O4@Graphene-MWCNTs 0.2 25 7.0 65.79 35 [43]
Fe3O4/PDA/Si-Ca-Mg 0.4 25 7.0 100.23 50 [44]
MnOx@cotton fibers 2.0 30 7.0 46.30 20 [45]
Fe-Mn oxides@carbon aerogels 0.6 30 7.0 9.37 30 [46]
MnFe2O4 0.15 25 7.0 44.90 – [47]
nFMBO 1.0 25 7.0 72.32 100 This test

–: the initial concentration of MB was not given in the reference.

a
qmax (mg/g): the maximum adsorption capacity of MB.
b
Ci (mg/L): the initial concentration of MB.

(a) (b)
120 100
+
Na
2+
100 Ca
80

80
60
η%
η%

60
40
40

20
20

0 0
2 3 4 5 6 7 8 9 10 11 12 0.01 0.1 0.2 0.4 0.8 1
pH Concentration (mol/L)
(c) (d)
120 110
0 mM 1 mM NOM
10 mM 50 mM
100 100

80 90
η%
η%

60 80

40 70

20 60

0 50
2- 2- 3- 0 2 4 6 8 10
CO3 SO4 PO4 Concentration (mg/L)
Fig. 7. (a) Effect of solution pH on the MB adsorption capacity of the nFMBO; (b) Effect of ionic strength on the adsorption of MB on the nFMBO; (c) Effect of co-anions (CO2–
3 ,
SO2
3

4 and PO4 ) with different concentrations (0, 1, 10 and 50 mM) on the adsorption of MB on the nFMBO; (d) Effect of NOM with different concentrations (0–5 mg/L) on the
adsorption of MB on the nFMBO.

adjusting the pH value of wastewater is a desirable method to
improve the MB adsorbed on the nFMBO in practical application.
3.5. Influence of ionic strength, co-anions and NOM on MB sorption
Inorganic salts are always used as additive in the dyes produc-
tion. For example, in the treatment of printing and dyeing, inor-
ganic salts are usually added as dye-promoting agents and
levelling agents [49]. As electrostatic interaction played a crucial
role in the adsorption of MB on the nFMBO, these inorganic salts
in wastewater may affect the adsorption process by competing
for the charged sites on its surface. Therefore, it is necessary to
study the influence of inorganic salts on the adsorption perfor-
mance of the nFMBO. The influence of Na+ and Ca2+ on the adsorp-
tion of MB on the nFMBO was presented in Fig. 7b. It was found
that, at a background electrolyte of 0.01 mol/L Na+, 96.5% of MB
560 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562
was adsorbed by the nFMBO, while as the concentration of Na+
increased, the adsorbed MB decreased from 96.5% to 61.3%. For
Ca2+, the adsorbed MB decreased from 96.5% to 38.4%. These
results demonstrated that the presence of Na+ and Ca2+ inhibited
the adsorption of MB, and the inhibiting effect became stronger
with the ionic strength increased. The main reason was that there
was a competitive adsorption between salt ions and cationic MB on
the surface of the nFMBO, that is, some salt ions occupied a certain
amount of adsorption sites, leading to a decrease in the adsorption
sites of MB. Similar observations were reported for the adsorption
of MB using Zirconium-based metal organic frameworks loaded on
polyurethane foam membrane [16]. In addition, compared the Na+
with Ca2+, it was found that the inhibiting effect of Ca2+ was stron-
ger than that of Na+ at the same concentration, which was attribu-
ted to the fact that Ca2+ had a higher positive charge than Na+. Once
Ca2+ is adsorbed on the surface of the nFMBO, it could produce
stronger electrostatic repulsion to MB than Na+.
Additionally, the influence of co-anions, such as CO2– 2

3 , SO4 ,
PO3

3 and NOM on the adsorption of MB by the nFMBO was studied.
As shown in Fig. 7c. it was found that the presence of CO2– 2

3 , SO4 ,
PO3

3 did not significantly alter the adsorption capacity of the
nFMBO. The presence of NOM had little influence on the adsorp-
tion of MB by the nFMBO, as shown in Fig. 7d. For example, when
the concentration of NOM was at 4.0 mg/L, the removal of MB
could reach 98.1%, which is higher than that without NOM
(94.6%). The decreased MB removal may be attributed to the fact
that part of NOM bind to the surface of the nFMBO. Although the
removal of MB decreased when the concentration of NOM was at
10.0 mg/L, the removal rate remained above 90%. These results
demonstrated that the nFMBO could be an excellent adsorbent
for the treatment of cationic dye wastewater in natural conditions.
3.6. Reusability of nFMBO
The reusability of adsorbent is an important criterion to evalu-
ate its application in practical wastewater treatment. Recycling
usage of adsorbent can not only save the production cost but also
reduce secondary pollution. As nFMBO exhibited a poor adsorption
capacity at lower pH values, acid treatment may be considered as a
suitable approach for the regeneration of the nFMBO. Therefore,
HNO3 was selected as the eluent for the regeneration of the
nFMBO. Fig. 8a showed the stability of the nFMBO. It was clearly
observed that the adsorption capacity of the regenerated nFMBO
decreased by only about 10% after five cycles, indicating the excel-
(a)
100
80
60
η%
40
20
0 0.0
0 1 2 3 4 5
Cycle Times
Fig. 8. (a) MB removal by the nFMBO over five successive adsorption-desorption cycles. (b) The percentage of Fe and Mn leached from the nFMBO during adsorption-
desorption cycles.
lent regeneration performance of nFMBO. Moreover, the percent-
age of Fe and Mn leached from the nFMBO was measured, as
shown in Fig. 8b. It was found that the leaching of Fe and Mn from
the nFMBO was less than 2.0%, which indicated that the obtained
nFMBO had a satisfactory stability. In addition, in the experimental
process, we found the nFMBO could be easily separated from the
aqueous phase by simple sedimentation. Therefore, it is concluded
that the nFMBO could be easily regenerated, which guarantees its
long term usage in the treatment of actual dye wastewater.
3.7. Possible adsorption mechanisms
The above results showed that the adsorption of MB on the
nFMBO is pH-dependent and ionic strength-dependent. For pH,
the adsorption capacity of MB on nFMBO increased at pH > PZC,
on the contrary, the adsorption capacity decreased at pH < PZC,
which suggested that the intermolecular electrostatic attraction
should be considered during the adsorption process. Moreover,
the salt effect study showed that the adsorption capacity decreased
with the ionic strength increasing, which further confirmed the
existence of electrostatic interaction between MB and nFMBO.
To further gain insight into the interaction between the adsorp-
tion sites of nFMBO and MB, the FT-IR spectra of MB, nFMBO, and
nFMBO with adsorbed MB were analyzed. According to Fig. 9a, it
could be inferred that chemical bonding between nFMBO and MB
may take place. Typically, the characteristic peak at 3238 cm
1,
assignable to the hydroxyl group stretching vibration, shifted to
3234 cm
1 after adsorption, which indicated that the electrostatic
attraction between the cationic MB dye and the hydroxyl group of
the nFMBO [44]. Moreover, the new stretching vibration adsorp-
tion bands at 1595 cm
1, 1392 cm
1 and 879 cm
1, attributed to
the CAC stretching vibration of aromatic cycle and the aromatic
skeletal group from MB, were found on the spectra of nFMBO after
MB adsorption [50]. These results reflected the evidence for the
strong interaction between MB and nFMBO.
Based on the above discussion, a reasonable MB adsorption
mechanism of the nFMBO towards MB was proposed, as shown
in Fig. 9b. At lower pH (pH < PZC) values in which case, the surface
charge of the nFMBO and MB were both negative, the major
adsorption process was controlled by hydrogen bonding, formed
by the interaction between the nitrogen atoms from MB and
hydroxyl groups from the surface of the nFMBO. On the other hand,
for the higher pH (pH > PZC), the surface charge of the nFMBO was
(b)
2.0
Fe
Leaching of Fe and Mn (%)
Mn
1.6
1.2
0.8
0.4
6 1 2 3 4 5
Cycle Times
 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562 561

results from the study demonstrated that the nFMBO as an easy-

(a)
prepared adsorbent would be a promising material for efficient
(i) treatment of wastewater containing MB.
3238

Notes
% Transmittance

(ii)
3346
The authors declare no competing financial interests.

1595
1392
Acknowledgements

879
(iii)
We acknowledge the financial support from the National Natu-
3234
ral Science Foundation of China (21677074 and 21806076) and the
Foundation Research Funds for the Central Universities
(021114380082). We thank Elijah J. Petersen for reviewing the
manuscript.
4000 3500 3000 2500 2000 1500 1000 500
Wavenumbers (cm-1) Appendix A. Supplementary material
(b)
pH < PZC pH > PZC Supplementary data to this article can be found online at
H3C CH3 https://doi.org/10.1016/j.jcis.2018.12.094.
N N
H 3C CH3
References
N N
Fe-Mn Fe-Mn
S S
[1] E. Forgacs, T. Cserhati, G. Oros, Environ. Int. 30 (2004) 953–971.
[2] S. Vahidhabanu, A.A. Idowu, D. Karuppasamy, B.R. Babu, M. Vineetha, ACS
Sustain Chem. Eng. 5 (2017) 10361–10370.
N N [3] C. Zhao, H. Zheng, Y. Sun, S. Zhang, J. Liang, Y. Liu, Y. An, Sci. Total. Environ. 640
H3C CH3 H3C CH3
(2018) 243–254.
[4] N.M. Mahmoodi, M. Arami, N.Y. Limaee, N.S. Tabrizi, Chem. Eng. J. 112 (2005)
Hydrogen bond Electrostatic attraction 191–196.
H3C CH3 [5] R.O. Cristovao, A.P.M. Tavares, A.S. Ribeiro, J.M. Loureiro, R.A.R. Boaventura, E.A.
N
H 3C
N Macedo, Bioresource Technol. 99 (2008) 4768–4774.
CH3
[6] J. Khatri, P.V. Nidheesh, S.A. Singh, M.S. Kumar, Chem. Eng. J. 348 (2018) 67–73.
N N [7] U. Pal, A. Sandoval, S.I.U. Madrid, G. Corro, V. Sharma, P. Mohanty,
Fe-Mn Fe-Mn
S S Chemosphere 163 (2016) 142–152.
[8] M. Albert, M.S. Lessin, B.F. Gilchrist, J. Pediatr. Surg. 38 (2003) 1244–1245.
-

N (2018) 21605–21615.
H 3C CH3
Fig. 9. (a) FTIR spectra of (i) nFMBO, (ii) MB and (iii) MB-loaded nFMBO; (b)
Schematic illustration of the governing mechanisms of MB adsorption onto the
nFMBO.
negative, so the cationic MB could be strongly adsorbed on the
negative nFMBO through electrostatic attraction.
4. Conclusions
Based on the previously reported preparation approaches
[29,48], the present study has demonstrated a simple and facile
coprecipitation approach to prepare the nFMBO. The characteriza-
tion results indicated that the as-prepared nFMBO exhibited wrin-
kled structure and excellent surface performance, which facilitated
the removal of MB. Compared with other previously reported
adsorbents [46,47], the nFMBO showed higher adsorption capacity
for MB (72.32 mg/g). The sorption kinetic process was better fitted
by a pseudo-second-order kinetic model. The Langmuir isotherm
model was confirmed to be more suitable for describing the
adsorption behavior of the nFMBO than the Freundlich model.
Additionally, the nFMBO exhibited better removal efficiency for
MB in solution with high pH, and the co-anions (CO2– 2

3 , SO4 ,
PO3

3 ) did not significantly alter the adsorption capacity of the
nFMBO. Zeta potential and FT-IR analysis showed that the possible
mechanism for MB removal involved the electrostatic attractions
and H-binding interactions. Moreover, the nFMBO had a good
reusability and the adsorption capacity still kept 85.1% after five
cycles of the adsorption/desorption process, which was better than
the adsorbents used in previous studies [43,44]. Overall, these
[9] S. Sharma, A. Hasan, N. Kumar, L.M. Pandey, Environ. Sci. Pollut. Res. Int. 25
N
H3C CH3 [10] G. Crini, Bioresource Technol. 97 (2006) 1061–1085.
[11] M. Rafatullah, O. Sulaiman, R. Hashim, A. Ahmad, J. Hazard. Mater. 177 (2010)
70–80.
[12] R. Li, B. Gao, K. Guo, Q. Yue, H. Zheng, Y. Wang, Bioresource Technol. 240
(2017) 59–67.
[13] S. Zhao, Z. Wang, J. Membrane. Sci. 524 (2017) 214–224.
[14] J.T. Adeleke, T. Theivasanthi, M. Thiruppathi, M. Swaminathan, T. Akomolafe, A.
B. Alabi, Appl. Surf. Sci. 455 (2018) 195–200.
[15] N.P. de Moraes, M.L. Caetano Pinto da Silva, L.A. Rodrigues, Mater. Lett. 228
(2018) 486–489.
[16] J. Li, J.L. Gong, G.M. Zeng, P. Zhang, B. Song, W.C. Cao, H.Y. Liu, S.Y. Huan, J.
Colloid. Interf. Sci. 527 (2018) 267–279.
[17] P.C.C. Faria, J.J.M. Orfao, M.F.R. Pereira, Water Res. 38 (2004) 2043–2052.
[18] S.B. Wang, Y. Boyjoo, A. Choueib, Z.H. Zhu, Water Res. 39 (2005) 129–138.
[19] X. Yuan, S.P. Zhuo, W. Xing, H.Y. Cui, X.D. Dai, X.M. Liu, Z.F. Yan, J. Colloid.
Interf. Sci. 310 (2007) 83–89.
[20] Y. Song, Y. Duan, L. Zhou, J. Colloid. Interf. Sci. 529 (2018) 139–149.
[21] J. Hu, G.H. Chen, I.M.C. Lo, Water Res. 39 (2005) 4528–4536.
[22] M. Hua, S. Zhang, B. Pan, W. Zhang, L. Lv, Q. Zhang, J. Hazard. Mater. 211 (2012)
317–331.
[23] Y. Yao, S. Miao, S. Liu, L.P. Ma, H. Sun, S. Chem, Eng. J. 184 (2012) 326–332.
[24] Q. Liu, L.B. Zhong, Q.B. Zhao, C. Frear, Y.M. Zheng, ACS Appl. Mater. Inter. 7
(2015) 14573–14583.
[25] L.B. Zhong, J. Yin, S.G. Liu, Q. Liu, Y.S. Yang, Y.M. Zheng, RSC Adv. 6 (2016)
103438–103445.
[26] H.V. Tran, L.T. Bui, T.T. Dinh, D.H. Le, C.D. Huynh, A.X. Trinh, Mater. Res. Exp. 4
(2017) 035701.
[27] Y. Dai, J. Zou, D. Liu, L. Niu, L. Zhou, Y. Zhou, X. Zhang, Colloid. Surface. A. 550
(2018) 90–98.
[28] N. Kataria, V.K. Garg, Chemosphere 208 (2018) 818–828.
[29] G. Zhang, J. Qu, H. Liu, R. Liu, R. Wu, Water Res. 41 (2007) 1921–1928.
[30] G. Zhang, X. Xu, Q. Ji, R. Liu, H. Liu, J. Qiu, J. Li, ACS Appl. Mater. Inter. 9 (2017)
14868–14877.
[31] M. Ghobadi, M. Gharabaghi, H. Abdollahi, Z. Boroumand, M. Moradian, J.
Hazard. Mater. 351 (2018) 308–316.
[32] L.M. Ma, Z.G. Ding, T.Y. Gao, R.F. Zhou, W.Y. Xu, J. Liu, Chemosphere 55 (2004)
1207–1212.
[33] M.A. Salam, Chem. Eng. J. 270 (2015) 50–57.
[34] K. Bourikas, J. Vakros, C. Kordulis, A. Lycourghiotis, J. Phys. Chem. B. 107 (2003)
9441–9451.
[35] K. Lu, J. Hu, S. Gao, L. Mao, Environ. Pollut. 231 (2017) 863–870.
562 K. Lu et al. / Journal of Colloid and Interface Science 539 (2019) 553–562
[36] K. Lu, Q. Huang, T. Xia, X. Chang, P. Wang, S. Gao, L. Mao, Environ. Pollut. 220
(2017) 264–273.
[37] V.K. Gupta, A. Nayak, Chem. Eng. J. 180 (2012) 81–90.
[38] X. Liang, X. Wang, J. Zhuang, Y. Chen, D. Wang, Y. Li, Adv. Funct. Mater. 16
(2006) 1805–1813.
[39] S.M. Maliyekkal, A.K. Sharma, L. Philip, Water Res. 40 (2006) 3497–3506.
[40] G. Zhang, H. Liu, R. Liu, J. Qu, J. Colloid. Interf. Sci. 335 (2009) 168–174.
[41] Y. Shao, L. Zhou, C. Bao, J. Ma, M. Liu, F. Wang, Chem. Eng. J. 283 (2016) 1127–
1136.
[42] L. Ai, C. Zhang, F. Liao, Y. Wang, M. Li, L. Meng, J. Jiang, J. Hazard. Mater. 198
(2011) 282–290.
[43] P. Wang, M. Cao, C. Wang, Y. Ao, J. Hou, J. Qian, Appl. Surf. Sci. 290 (2014) 116–
124.
[44] P. Chen, Z.F. Cao, S. Wang, H. Zhong, Colloid. Surface. A. 552 (2018) 89–97.
[45] J. Wang, L. Kou, Z. Huang, L. Zhao, RSC Adv. 8 (2018) 21577–21584.
[46] S. Ye, W. Jin, Q. Huang, Y. Hu, Y. Li, B. Li, Int. J. Biol. Macromol. 92 (2016) 1169–
1174.
[47] X. Wu, Z. Ding, W. Wang, N. Song, S. Khaimanov, N. Tsidaeva, Powder Technol.
295 (2016) 59–68.
[48] H. Liu, Y. Yang, J. Kang, M. Fan, J. Qu, J. Environ. Sci. 24 (2012) 242–247.
[49] B. Yu, B. Yang, G. Li, H. Cong, J. Text. I 105 (2013) 321–326.
[50] Y. Wang, W. Wang, A. Wang, Chem. Eng. J. 228 (2013) 132–139.