Journal of Environmental Chemical Engineering 6 (2018) 5990–5998

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

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

Optimization of carboxyl-functionalized mesoporous silica for the selective T

adsorption of dysprosium
⁎
Takamasa Kanekoa,b, Fukue Nagatab, Shinichi Kugimiyaa, Katsuya Katob,
a
Materials Chemistry Course, Graduate School of Engineering, Aichi Institute of Technology, 1247 Yachigusa, Yakusa-cho, Toyota, 470-0392, Japan
b
National Institute of Advance Industrial Science and Technology, 2266-78, Anagahora, Moriyamaku, Nagoya, Aichi, 463-8560, Japan

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

Keywords: Dysprosium (Dy), one of the rare-earth elements, was adsorbed on the four types of mesoporous silica, having
Mesoporous pore size 3–22 nm and functionalized with amino and carboxyl groups on the surface. The influence of the pore
Silica size on the selective Dy ion adsorption capacity was examined using a metal ion mixture. The synthesized
Carboxyl materials was characterized by field emission scanning electron microscopy, transmission electron microscopy,
Dysprosium
Brunauer-Emmett-Teller analyzer. In addition, the functionalized materials were characterized by thermo-
Selectivity
gravimetric analyzer, Fourier transform infrared, and zeta-potential. The adsorption amount of Dy ions on the
functionalized particles was evaluated by inductively coupled plasma optical emission spectrometry. The ma-
terial having a large pore size (22 nm) as well as amino and carboxyl chains manifested an excellent adsorption
capacity for Dy ions (32.9 μg/mg). The recycling efficiency of adsorption-desorption process retained approxi-
mately 100% of the capability at least five cycles. These results suggested that the functionalized material with
pore size 22 nm can be used for an adsorbent for the selective and efficient removal of Dy ions from mixtures.

1. Introduction
Dysprosium (Dy), one of the rare-earth elements as neodymium
magnets, is useful in the automobile industry. Although the increasing
of Dy consumption in recent years is attributed to the increasing de-
mand for its factor, however, it is difficult to assure a constant supply
because of low production and high cost. In 2013, approximately 90%
of the total world production of Dy element occurred in China.
However, the significant problem related to this element is generated
due to the amount of supply and price stability since the most part of
resources was dependent on other countries. Therefore, usual and
simple recycling systems of Dy are necessary.
To date, various materials were reported as adsorbents for metal
ions, e.g., ceramic materials and polymers [1–4]. It is well known that
various adsorbates were adsorbed on the surface of these inorganic
materials via physisorption; however, they manifested low selective
adsorption [5,6]. The removal of a particular ion from the mixture of
various metal ions is an important industrial approach [7,8]. In pre-
vious study, Dy ions were adsorbed from a metal ion mixture on non-
porous the silica gel modified with the amino and carboxyl groups. It
was reported that these silica materials selectively adsorbed Dy on their
surface based on a complex formation between the organic groups and
the outermost shell of the Dy ion, with an adsorption capacity of
0.146 mmol/g [9,10]. Several studies have also reported on Dy ion
adsorption using the functionalized amino and carboxyl groups on the
solid surface. In terms of efficiency, silica samples with high surface
area such as porous particles are promising materials since they have
high adsorption capacity [11–13]. Therefore we assumed that meso-
porous silica (MPS) was considerably attractive for developing highly
efficient materials for Dy adsorption.
Ordered MPS has uniform nanosize pores that are formed with the
aid of surfactant micelles [14–20]. MPS is an inorganic material with
carefully controlled pore sizes (2–50 nm), large surface area, and high
pore volume [20–22]. We previously reported that enzyme im-
mobilization on MPS showed enhanced enzyme activity and stability
[23]. For immobilization inside the pores, the available free space must
be considered to provide sufficient enzyme mobility for retaining their
catalytic activity. Additionally, the free space ensures substrate diffu-
sion to the catalytic sites. Herein, we demonstrate that the porous
materials with adequate pore structure allow for the adsorption of
metal ions and focus on the effects of Dy adsorption amounts on the
functionalized MPS with different pore sizes. MPS with pore size and
pore distribution, which were easily controlled, was synthesized via a
templating process using a surfactant as the organic template and tet-
raethoxysilane (TEOS) as silica sources. The formation of the solid
supports is based on the hydrolysis and polycondensation processes of

⁎
Corresponding author.
E-mail addresses: takamasa-kaneko@aist.go.jp (T. Kaneko), katsuya-kato@aist.go.jp (K. Kato).

https://doi.org/10.1016/j.jece.2018.09.018
Received 20 July 2018; Received in revised form 31 August 2018; Accepted 14 September 2018
Available online 18 September 2018
2213-3437/ © 2018 Elsevier Ltd. All rights reserved.
T. Kaneko et al. Journal of Environmental Chemical Engineering 6 (2018) 5990–5998

Scheme 1. Schematics of amino-functionalized mesoporous silica (MPS) materials employing (a) APTES and −NH2 and (b) N6-APTES and −2NH2 and the following
carboxyl-functionalized MPS materials employing diglycolic anhydride: (c) −NH-COOH and (d) −2NH-2COOH, respectively.

Fig. 1. FE-SEM images of (a) mesoporous silica (MPS) (3 nm), (b) MPS (5 nm), (c) MPS (12 nm), and (d) MPS (22 nm).

silica precursors. The process is that allows designing characteristics the
synthesized materials. MPS has SiOH groups on the surface, allowing
for functionalization with various organic groups. Herein, we synthe-
sized MPS with various pore sizes (3, 5, 12, and 22 nm) using a sol–gel
method, while the surface was functionalized with the amino and car-
boxyl groups using commonly employed coupling reagents. The amount
of metal ion adsorbed on MPS was systematically monitored.
In addition, the effect of the pore size on the Dy ion adsorption
capacity of the resulting materials was studied. The recycling property
of the adsorbents was also investigated.
2. Experimental procedure
2.1. Materials
1-Hexadecytrimethylammoniumchloride (CTAC) and Pluronic P123
(P123) were purchased from Sigma-Aldrich Co., St. Louis., MO. TEOS
and 3-aminopropyltrimethoxysilane (APTES) were commercially ob-
tained from Shin-Etsu Chemical Co., Tokyo, Japan. Diglycolic hydride
was purchased from Tokyo Chemical Industry Co., Tokyo, Japan. 1,3,5-
Trimethylbenzene (TMB), acetic acid, toluene, n-decane, aqueous am-
monia, ethanol, and hydrochloric acid were obtained from Wako Pure
Chemical Industries, Osaka, Japan. N-(6-aminohexyl)aminopropyl-
trimethoxysilane (N6-APTES) was obtained from AZmax Co., Chiba,
Japan.

5991
T. Kaneko et al.
Fig. 2. TEM images of (a) MPS (3 nm), (b) MPS (5 nm), (c) MPS (12 nm), and (d) MPS (22 nm).
2.2. Characterization
The morphology of the synthesized MPS materials was investigated
using field emission scanning electron microscopy (FE-SEM; S-4300,
Shimadzu Co., Kyoto, Japan) at an accelerating voltage of 10.0 kV, via
transmission electron microscopy (TEM; JEM 2010, JELO Ltd., Tokyo,
Japan) at an accelerating voltage of 200 kV. Scanning TEM (STEM;
JEOL JEM-2100 Plus, JOEL Ltd., Tokyo, Japan) images were operated
at 200 kV, and a copper grid with coating carbon was used for metal
substrates. The nitrogen (N2) adsorption and desorption measurements
were performed using a Shimadzu TriStar 3000 system after preheating
at 105℃, and the specific surface area was calculated according to the
Brunauer–Emmett–Teller (BET) method, and the pore size distribution
was obtained by the Barrett–Joyner–Halenda (BJH) method. The
amounts of the amino and carboxyl functional groups on MPS were
established in heating temperature range from room temperature to
1000℃ (10℃/min) using a thermogravimetric analyzer (TG-DTA; TGA-
502, Shimadzu Co., Kyoto, Japan). The organic functional groups on
the surface were also analyzed via Fourier transform infrared (FT-IR)
spectroscopy in the wavenumber range of 400–4000 cm−1 using an
MFT-2000 spectrophotometer (JASCO Co., Tokyo, Japan). The spectra
were obtained in the transparent mode using a pellet made of 99% KBr
and 1% sample. Additionally, to prove the existence of functional
groups on MPS, the surface potential of MPS was measured using a zeta-
potential analyzer (ELSZ-1000, Otsuka Electronics Co., Tokyo, Japan).
Before the measurements, the particles were dispersed in distilled water
with sonication for 5 min. The amount of metal ions adsorbed on MPS
was determined using inductively coupled plasma optical emission
spectrometry (ICP-OES; IRIS Advantage, Thermo Fisher Scientific, MA).
5992
Journal of Environmental Chemical Engineering 6 (2018) 5990–5998
2.3. Synthesis of MPS
2.3.1. Synthesis of MPS (3 and 5 nm) particles
MPS particles (3 and 5 nm) were synthesized according to a pre-
viously reported synthesis procedure and using CTAC as an organic
template [21]. CTAC (2.6 g) was added to 30 mL of aqueous solution
(pH 0.5), and the mixture was stirred for 2 h at room temperature. To
obtain mesopores of 5 nm, TMB (2.6 g) was added to the solution
containing CTAC. Subsequently, TEOS (3.5 g) was added dropwise and
the solution was stirring at 4 °C. After 2 h stirring, NH3 (3.0 g, 28 wt.%,
14.7 M) was added to the solution and the reaction mixture was stirred
for 24 h. Finally, the product was centrifuged at 6000 rpm for 20 min.
washed and calcined at 500 °C for 5 h in air (heating rate of 1℃/min) to
remove the organic template.
2.3.2. Synthesis of MPS (12 and 22 nm) particles
The procedure for preparing MPS (12 nm) is described as follows:
first, triblock copolymer P123 (1 g), used as the template, was com-
pletely dissolved in HCl (35 mL, 1.07 M). Then, n-decane (7.6 g) was
added to the mixture. After 2 h stirring, ammonium fluoride (NH4F;
11.5 mg) was added to the solution. Next, TEOS (2.13 g) was added to
the mixture and the mixture was stirred for 1 h. Subsequently, the re-
action mixture was incubated at 40 °C for 20 h in a temperature-con-
trolled bath, after which the suspension was hydrothermally treated for
24 h at 110 °C. The MPS (22 nm) particle was also prepared using TMB
(1 g) as a pore expander. First, P123 was dissolved in distilled water
(26 mL). After gentle stirring, TMB (1 g) was added to the mixture and
the mixture was stirred overnight. Next, NH4F (11.5 mg) and TEOS
(2.1 g) were added to the solution and the reaction mixture was stirred
for 10 min at room temperature. Then, a concentrated HCl (4 mL) so-
lution was gradually added and the solution was stirred for 30 min at
T. Kaneko et al. Journal of Environmental Chemical Engineering 6 (2018) 5990–5998

Fig. 3. N2 adsorption–desorption (A–D) isotherms and pore size distribution curves (a–d) of the MPS materials. (A, a) MPS (3 nm), (B, b) MPS (5 nm), (C, c) MPS
(12 nm), and (D, d) MPS (22 nm). These samples were measured after the samples were preheated at 105℃.

Table 1 room temperature. Subsequently, the resulting mixture was incubated

The amount of amino and carboxyl groups on functionalized MPS materials. at 35 °C for 20 h, after which it was transferred to an autoclave and
Sample -NH2a -NH-COOHa −2NH2a −2NH-2COOHa treated at 110 °C for 24 h. To remove the template, MPS was calcined at
(μg/m2)b (μg/m²)b (μg/m²)b (μg/m²)b 500 °C for 5 h. The heating temperature were ranged from room tem-
perature to 500℃ (heating rate 1 ℃/min).
MPS (3 nm) 20 137 26 106
MPS (5 nm) 234 412 349 494
MPS (12 nm) 377 637 292 709
2.4. Surface functionalization of MPS
MPS (22 nm) 293 629 379 797

a
Amino-functionalized MPS materials using APTES; −NH2 or N6-APTES; To adsorb metal ions on the MPS surface, amino- and carboxyl-
−2NH2, followed by functionalization with carboxyl groups; −NH-COOH and functionalized MPS materials were prepared using a method similar to
−2NH-2COOH, respectively. that described in previous research [24,25]. Amino-functionalized MPS
b
The amount of functional groups on MPS materials was calculated from the was prepared by dispersing MPS (100 mg) in toluene (10 mL), and 1 mL
weight losses in TG curves. of 3-aminopropyltriethoxysilane (APTES; −NH2) or N-(6-aminohexyl)
aminopropyltrimethoxysilane (N6-APTES; −2NH2) were added to the
solution for 10 min. Next, the mixture was refluxed for 6 h. The product
was centrifuged at 6000 rpm for 20 min, and the precipitate was

5993
T. Kaneko et al.
Fig. 4. FTIR spectra of MPS materials. a) red line denotes MPS (12 nm). b) The
MPS (12 nm)-NH2 and c) MPS (12 nm)-2NH2 samples are denoted by yellow
and blue lines, respectively. d) The MPS (12 nm)-NH-COOH and e) MPS
(12 nm)-2NH-2COOH samples are denoted by green and purple lines, respec-
tively. These sample were also analyzed in the wavenumber range of
400–4000 cm−1.
Fig. 5. TEM image of MPS (22 nm) −2NH-2COOH.
washed with ethanol and acetone, and dried at room temperature.
Carboxyl-functionalized MPS was prepared as follows: 150 mg of MPS-
NH2 and MPS-2NH2 were redispersed into toluene (15 mL), and the
solution was stirred at 50 °C for 10 min. Subsequently, diglycolic an-
hydride (261 mg) was added to the solution and the mixture was in-
cubated at 50 °C. After 1 h, the reaction mixture was stirred at room
temperature. Finally, the products were washed with ethanol and
acetone for dryness (Scheme 1).
2.5. Adsorption of metal ions on functionalized MPS materials
The concentration of metal ions (copper, Cu; dysprosium, Dy; iron,
Fe; neodymium, Nd; and zinc, Zn) was 10 ppm for each element. The
solutions were prepared in distilled water. The adsorption tests were
conducted in a solution at pH 1.75, adjusted by titration with 1 M
HNO3. Metal ion (Cu, Dy, Fe, Nd, and Zn) mixture were used in this
experiment, and the adsorption experiments were carried out at the
room temperature (20℃). The adsorption process was carried out as
follows: first, carboxyl-functionalized MPS (2 mg) was added to 10 mL
of the metal ion solution (Cu, Dy, Fe, Nd, and Zn; concentration was
10 ppm for each element) and the mixture was stirred for 2 h at room
temperature. The supernatant was separated by centrifugation at
5994
Journal of Environmental Chemical Engineering 6 (2018) 5990–5998
6000 rpm for 20 min. The supernatant was filtered using a 0.45 μm
filter, and the amount of metal ions in supernatant was measured via
ICP-OES. The amount of metal ions adsorbed on carboxyl-functiona-
lized MPS was calculated from the change in the concentration of the
Dy ions into the solution using Eq. (1). Where Qb and Qa is the initial
Dy ions concentration and after adsorption, respectively, and I0 is the
initial Dy ions content.
Eq = (Qb – Qa) I0 (1)
2.6. Adsorption isotherms
Adsorption isotherm experiments were conducted at room tem-
perature using MPS (12 nm)-2NH-2COOH and MPS (22 nm)-2NH-
2COOH. These samples having large amount of Dy ions adsorption were
added to Dy ions solutions with various concentrations (2–10 ppm), and
adsorption process was conducted for 2 h at the room temperature.
After centrifugation, the Dy ions concentration of the supernatant was
measured by ICP-OES, and adsorption amount of Dy ions on functio-
nalized MPSs was calculated. To establish the metal ion adsorption
mechanism, the data were further processed by Langmuir and
Freundlich equations. Detailed definitions of the variables can be found
in the literature [26–29]. The Langmuir adsorption isotherm model is
given by Eq. (2):
Qe = (QmKLC)/ (1 + KLC) (2)
where Qe (μg/mg) is the amount of metal ions on MPS adsorbed at
equilibrium, C (mol/L) is the equilibrium concentration in the super-
natant, and KL is the Langmuir constant. Qm denotes the maximum
monolayer surface coverage adsorption capacity on MPS.
The Freundlich isotherm model indicates multilayer adsorption, and
the equation of this model is expressed as follows:
Qe = KFCe1/n (3)
where KF (mg/g) is the Freundlich constant that denotes the adsorption
capacity, the 1/n value stands for the favorable adsorption condition.
2.7. Cycling tests
Cycling tests was conducted using carboxyl-functionalized MPS
(22 nm)-2NH-2COOH. The adsorption and desorption experiments
were performed at room temperature only for Dy ions. The adsorption
process was carried out as follows: first, carboxyl-functionalized MPS
materials (5 mg) were added to 25 mL of Dy ions solution (10 ppm) and
the mixture was stirred for 2 h. The supernatant was separated by
centrifugation at 6000 rpm for 20 min, and the solution was filtered
using a 0.45 μm filter. To desorb Dy ions from the adsorbent, the pre-
cipitate was added to 12.5 mL of 1 M HNO3 solution and the mixture
was stirred for 2 h. The supernatant was separated by centrifugation at
6000 rpm for 20 min, and the solution was filtered using a 0.45 μm
filter. Subsequently, the adsorbent was washed twice with distilled
water. The adsorption experiment was repeated five times. After each
filtration, the amount of Dy ions in the supernatant was measured via
ICP-OES. The amount of Dy ions adsorbed on carboxyl-functionalized
MPS was calculated from the change in the concentrations of Dy ions in
the solution before adsorption; Qc and after desorption; Qd. To evaluate
the cycling property of MPS, the adsorption capacity QR was calculated
based on the concentration of Dy ions in the supernatant using Eq. (4).
The value of 100% capacity illustrated the adsorption Dy ions desorbed
from the MPS completely by desorption process.
QR (%) = (Qd/Qc) × 100 (4)
T. Kaneko et al.
Fig. 7. Adsorption of the five metal ions (Cu, Dy, Fe, Nd, and Zn) on functionalized MPS materials, with initial concentration of 10 ppm, pH: 1.75, room temperature
and adsorption time of 2 h.
3. Results and discussion
3.1. Properties of synthesized MPS particles
The as-synthesized MPS materials with various pore sizes were
characterized using FE-SEM, TEM, and N2 physisorption. The mor-
phology of the synthesized MPS materials was characterized using FE-
SEM and TEM. The FE-SEM and TEM images of the synthesized MPS
particles (3, 5, 12, and 22 nm) are shown in Fig. 1(a)–(d) and Fig. 2,
respectively. According to the FE-SEM images in Fig. 1(a)–(d), MPS
(3 nm) and MPS (5 nm) was composed of small particles (50–500 nm),
while the MPS (12 nm) and MPS (22 nm) show agglomerated
morphologies with polyhedral particles with mainly in the range of
1–10 μm. As observed in TEM images, MPS particles was confirmed the
hexagonal structure and exhibited the presence of mesoporous channels
with each diameter and appearance in Fig. 2(a)–(d).
N2 physisorption was performed for MPS particles, and the resulting
adsorption–desorption isotherms [(A)–(D)] and pore size distribution
curves [(a)–(d)] are shown in Fig. 3. MPS with pore size (3, 5, 12,
22 nm) were exhibited the isotherms of type IV, furthermore, MPS
having pore size 12 nm and 22 nm are indicative of H1 hysteresis loops,
respectively. The textural properties calculated based on the obtained
5995
Journal of Environmental Chemical Engineering 6 (2018) 5990–5998
Fig. 6. EDX profiles and EDX mapping images
of MPS (22 nm) −2NH-2COOH. STEM ele-
mental mapping images of MPS
(22 nm)−2NH-2COOH. Yellow, violet, and
green indicate silicon, nitrogen, and oxygen,
respectively. STEM images were operated at
200 kV, and a copper grid with coating carbon
was used for metal substrates.
isotherms are summarized, and the values for the specific surface area
of MPS with different pore sizes (3, 5, 12, and 22 nm) were 998, 865,
608, and 546 m2/g, respectively, while those for the pore volumes were
1.27, 0.73, 1.44, and 1.95 cm3/g, respectively. The actual size of the
pores was controlled at approximately 2.6, 5.2, 12.5, and 22.5 nm.
Based on these values, the samples were denoted as MPS (3, 5, 12, and
22 nm), respectively. From these results, it could be concluded that the
synthesized MPS materials have a mesoporous structure and optional
pore diameter.
3.2. Characterization of functionalized MPS materials
MPS particles were functionalized with APTES and N6-APTES, fol-
lowed by functionalization with carboxyl groups using diglycolic an-
hydride, as shown in Scheme 1, and named as (a) −NH2, (b) −2NH2,
(c) −NH-COOH, and (d) −2NH-2COOH, respectively.
To confirm the functionalization of the MPS surface, FTIR, TG-DTA,
N2 physisorption, and ζ-potential analyses were performed. To de-
termine the amount of the amino and carboxyl groups on the MPS
surface, thermal gravimetric analysis was performed. The corre-
sponding weight losses were attributed to the decomposition of organic
groups on the surface. From the exothermic peak attributed to the
T. Kaneko et al. Journal of Environmental Chemical Engineering 6 (2018) 5990–5998

797 μg/m2, respectively. According to these results, the weight losses

were higher, with each successive functionalization indicating that the
amount of the amino and carboxyl groups on MPS progressively in-
creased with the pore size. In particular, as the amount of organic
groups on MPS-2NH-2COOH increased in comparison with that on
MPS-NH-COOH, it was possible to observe the highly increasing
amount of functionalized carboxyl groups on MPS (12 and 22 nm)-2NH-
2COOH at 709 and 797 μg/m2, respectively. The amount of organic
groups on MPS (12 and 22 nm)-2NH-2COOH greatly increased in
comparison to that on MPS (3 and 5 nm)-2NH-2COOH. It can be as-
certained that in the last two cases, the pores of MPS were obstructed by
the organic groups, while for MPS (12 and 22 nm), the pores were
sufficiently large to avoid obstruction during the second step of func-
tionalization.
ζ-Potential measurements were performed to analyze the surface
charge for each functionalized MPS. It was observed that the obtained
values were mainly affected by the functional groups attached on the
surface. The ζ-potentials MPS (22 nm), MPS (22 nm)-NH2, and MPS
(22 nm)-2NH-2COOH were 8.83, 18.1, and −37.4 (mV), respectively.
From these results, the presence of the amino and carboxyl functional
groups on MPS-2NH2 and MPS-2NH-2COOH was confirmed. The FTIR
Fig. 8. Illustration of pore size and surface functionalization on MPS with small spectra of MPS samples are shown in Fig. 4. The spectra of all MPS
(3 nm) and large (12 nm) pore sizes. (A) Functionalization of MPS with amino
samples contain absorption bands characteristic to the silanol groups
groups, (B) subsequent functionalization of MPS (pore size = 3 nm) with car-
O–Si–O and Si–O–Si at 943 and 1089 cm−1, respectively. For the
boxyl groups, and (C) subsequent functionalization of MPS (pore size = 12 nm)
amino-functionalized MPS samples, including MPS (12 nm)-NH2 and
with carboxyl groups.
MPS (12 nm)-2NH2, the absorption bands at 2865, 2980 and
1540 cm−1 attributed to −CH2, −CH2 and NeH bonds, respectively,
Table 2 were noticed, providing evidence for the presence of organic groups on
Langmuir and Freundlich constants for Dy ion adsorption. the silica surface [30–32]. Thus, according to the registered spectra for
Langmuir plot Freundlich plot MPS (12 nm)-NH2, MPS (12 nm)-2NH2, MPS (12 nm)-NH-COOH, and
MPS (12 nm)-2NH-2COOH, the amino groups derived from APTES
Sample Qma R2
1/nb R2
(−NH2) or N6-APTES (−2NH2) were bonded to the surface. Finally, for
MPS (12 nm) 39.1 0.99 0.54 0.95 the carboxyl-functionalized MPS samples, including MPS (12 nm)-NH-
−2NH-2COOH COOH and MPS (12 nm)-2NH-2COOH, the absorption peak at
MPS (22 nm) 44.8 0.99 0.59 0.94 1680 cm−1 corresponding to the eC]O group was observed for both
−2NH-2COOH samples [32]. Therefore, the FTIR spectra of the MPS (22 nm)-NH2 and
a MPS (22 nm)-2NH2 samples showed the presence of amino groups,
Qm denotes the maximum monolayer surface coverage adsorption capacity
on MPS materials. while those of the MPS (22 nm)-NH-COOH and MPS (22 nm)-2NH-
b
1/n value stands for the favorable adsorption condition. 2COOH samples proved the presence of both the amino and carboxyl
groups. In addition, the TEM images of the structure and morphology of
MPS materials before and after functionalization were shown in Fig. 5.
As shown in the TEM images, it was indicated that the porous structure
was retained like small domains of cylindrical pores. The results dis-
played no difference in the ordered mesoporous structure between the
samples before and after functionalization, thereby showing that the
porosity of the functionalized MPS materials was preserved. To confirm
the presence of functionalized groups in nanoparticles, the elemental
distribution of MPS (22 nm)-2NH-2COOH was investigated by STEM-
EDX analysis. EDX spectrum of the functionalized MPS (22 nm)-2NH-
2COOH is depicted in Fig. 6, which indicates the presence of silicon,
oxygen, and nitrogen. The nitrogen is based on the functionalized
groups containing amino groups. The elemental mapping confirms a
uniform distribution of the functionalized groups around the nano-pore
structure of silica nanoparticles.
Fig. 9. Cycling test of MPS (22 nm)-2NH-2COOH performed five times for Dy
ions adsorption, with initial concentration of 10 ppm, pH: 1.75, room tem-
3.3. Metal ion adsorption on MPS samples
perature and adsorption time of 2 h. The desorption process was performed by
adding to the 1 M HNO3 solution and the mixture was stirred for 2 h.
Subsequently, the absorbent was washed twice with distilled water. The ad- Different metal ions (i.e., Cu, Dy, Fe, Nd, and Zn) were adsorbed on
sorption-desorption experiment was repeated five times. the amino- and carboxyl-functionalized MPS materials via physical
adsorption. The amount of ions adsorbed at equilibrium on these
samples is shown in Fig. 7. As shown in the figure, the adsorption of Cu,
decomposition of the amino and carboxyl groups, the amount of or-
Fe, Nd, and Zn ions on the functionalized MPS samples did not occur,
ganic groups on the MPS samples were calculated (Table 1). Therefore,
whereas the Dy ions were found to have adsorbed. According to pre-
the amount of organic groups on MPS (3, 5, 12, and 22 nm)-NH-COOH
vious research, the adsorption of Dy ions has been explained via a
was 137, 412, 637, and 629 μg/m2, respectively, while on MPS (3, 5,
complexation reaction occurring between the Dy ions and the carbonyl
12, and 22 nm)-2NH-2COOH, the amount was 106, 494, 709, and
oxygen on the functional groups [9–11]. Therefore, the amount of Dy

5996
T. Kaneko et al.
ions adsorbed at equilibrium on the MPS (3, 5, 12, and 22 nm) samples
functionalized with N6-APTES and diglycolic acids (−2NH-2COOH)
was 5.7, 8.7, 26.6, and 32.9 μg/mg silica, respectively. From these re-
sults, it can be observed that Dy adsorption increased as the pore size
increased. In contrast, for the MPS (3, 5, 12, and 22 nm) samples
functionalized with APTES and diglycolic acids (−NH-COOH), the
amount of Dy adsorbed was 2.6, 4.8, 6.4, and 7.0 μg/mg silica, re-
spectively. It is expected that for MPS samples having higher pore size
and higher amount of organic functional groups (i.e., MPS (12 nm)-
2NH-2COOH and MPS (22 nm)-2NH-2COOH), the amount of adsorbed
Dy greatly increased up to 26 and 32 μg/mg, respectively. According to
a previous report, materials such as silica gel have a Dy metal ion ad-
sorption capacity of 0.146 mmoL/g (24 μg/mg) [10]. Our results illu-
strated that MPS (22 nm)-2NH-2COOH, the adsorption capacity in-
creased more than 130% in comparison with that reported for silica gel.
The MPS-2NH-2COOH sample with smaller pore sizes (3 and 5 nm)
exhibited the lowest adsorption capacity (5.7 and 8.7 μg/mg, respec-
tively). On the contrary, MPS functionalized with the amino and car-
boxyl groups (−2NH-2COOH) with pore sizes of 12 and 22 nm had
higher Dy adsorption capacity. These results indicate that the amount of
the functional groups on MPS is considerably important for ensuring the
high adsorption capacity of these materials toward Dy ions. Based on
our experimental findings, a schematic of the relation between Dy ad-
sorption and surface functionalization of MPS samples is shown in
Fig. 8. It can be seen that the pores with a diameter of 3 nm were
collapsed by the organic groups attached on their internal surface, and
there is no space in the pores. Therefore, it was difficult to retain the
functionalization groups. In conclusion, it can be affirmed that the pore
size and organic functional groups on MPS greatly affect the adsorption
of Dy ions.
3.4. Adsorption isotherm of Dy ions on MPS samples
The results for the Dy adsorption capacity of the functionalized MPS
samples (12 and 22 nm) are summarized in Table 2. The adsorption
models were evaluated from the Langmuir and Freundlich plots. The
Langmuir model showed better fit than the Freundlich model with re-
gard to the R2 value. As summarized in Table 2, the R2 values for the
Langmuir plots corresponding to MPS (12 nm)-2NH-2COOH and MPS
(22 nm)-2NH-2COOH were the same, i.e., 0.99, showing a good fit. The
R2 values for Freundlich plots corresponding MPS (12 nm) -2NH-
2COOH and MPS (22 nm)-2NH-2COOH were lower (0.95 and 0.94,
respectively). From these results, it can be stated that Dy ions are ad-
sorbed on these functionalized MPS materials as a monolayer. The ex-
perimental maximum Dy adsorption capacity for MPS (12 nm)-2NH-
2COOH and MPS (22 nm)-2NH-2COOH (Qm) calculated using the
Langmuir model was 39.1 and 44.8 (μg/mg), respectively. Thus, these
results also confirm that higher pore sizes are more favorable for Dy
adsorption capacity.
3.5. Cycling test of adsorption on MPS
Cycling efficiency is one of the most important factors when col-
lection materials are used as resource adsorbents. In the cycling test, an
experiment was performed using MPS (22 nm)-2NH-2COOH. The ad-
sorption process was performed according to the procedure described in
Section 2.7. The desorption process was performed using 1 M HNO3.
The cycling test was performed five times, the results of which are
shown in Fig. 9. It can be observed that the adsorption capacity did not
change after five cycles.
4. Conclusion
Herein, we reported that the rare-earth element Dy can be suc-
cessfully adsorbed on the amino- and carboxyl-functionalized MPS
materials with different pore sizes. These materials were synthesized
5997
Journal of Environmental Chemical Engineering 6 (2018) 5990–5998
via soft templating method using organic surfactants, and the pore
diameter was controlled by an organic reagent. The surface of MPS was
functionalized using a silane coupling reagent and diglycolic acid. To
determine the mesoporous character of the materials, FE-SEM, TEM,
and N2 physisorption were performed. The presence of the amino and
carboxyl groups on the surface were confirmed via TG-DTA, FTIR, and
zeta-potential. Among two types of carboxyl functionalization of MPS
with pore size (3, 5, 12 and 22 nm), the highly Dy ion adsorption was
shown for MPS (12 and 22 nm)-2NH-2COOH with pore sizes of 12 and
22 nm. This was due to the effect of pore size on MPS, as shown by N2
physisorption results in Fig. 3. The pore diameters of MPS (12 and
22 nm) provide sufficient space for Dy adsorption after functionaliza-
tion with carboxyl groups; however, smaller pore sizes of 3 and 5 nm
were filled with the amino and carboxyl functional groups that hin-
dered the adsorption of metal ions. These results indicated that the
investigation of the pore size along with a high degree of functionali-
zation could improve the adsorption of Dy ions.
Acknowledgement
This work is partly supported by Grant-in-Aid for Scientific Research
(C) No. 15K06474 from the Japan Society for the Promotion of Science
(JSPS).
References
[1] J. Aguado, J.M. Arsuaga, A. Arencibia, M. Lindo, V. Gascon, Aqueous heavy metals
removal by adsorption on amine-functionalized mesoporous silica, J. Hazard.
Mater. 163 (2009) 213–221.
[2] A.M. Liu, K. Hidajat, S. Kawi, Y. Zhao, A new class of hybrid mesoporous materials
with functionalized organic monolayers for selective adsorption of heavy metal
ions, Chem. Commun. (2000) 1145–1146.
[3] A. Sayari, S. Hamoudi, Y. Yang, Applications of pore-expanded mesoporous silica. 1.
Removal of heavy metals cations and organic pollutants from wastewater, Chem.
Mater. 17 (2005) 212–216.
[4] P.K. Jal, S. Patel, B.K. Mishra, Chemical modification of silica surface by im-
mobilization of functional groups for extractive concentration of metal ions, Talanta
62 (2004) 1005–1028.
[5] T.A. Saleh, A. Sar, M. Tuzen, Optimization of parameters with experimental design
for the adsorption of mercury using polyethylenimine modified-activated carbon, J.
Environ. Chem. Eng. 5 (2017) 1079–1088.
[6] T.A. Saleh, M. Tuzen, A. Sarı, Magnetic activated carbon loaded with tungsten oxide
nanoparticles for aluminum removal from waters, J. Environ. Chem. Eng. 5 (2017)
2853–2860.
[7] T.A. Saleh, Simultaneous adsorptive desulfurization of diesel fuel over bimetallic
nanoparticles loaded on activated carbon, J. Clean. Prod. 172 (2018) 2123–2132.
[8] T.A. Saleh, Mercury sorption by silica/carbon nanotube and silica/activated carbon:
acomparison study, J. Water Supply Res. Technol. 64 (2015) 892–903.
[9] N. Zhang, B. Hu, C. Huang, A new ion-imprinted silica gel sorbent for on-line se-
lective solid-phase extraction of dysprosium (Ⅲ) with detection by inductively
coupled plasma-atomic emission spectrometry, Anal. Chim. Acta 597 (2007) 12–18.
[10] T. Ogata, H. Narita, M. Tanaka, Adsorption behavior of rare earth elements on silica
gel modified with diglycol amic acid, Hydrometallurgy 152 (2015) 179–182.
[11] X. Zheng, E. Liu, F. Zhang, Y. Yan, J. Pan, Efficient adsorption and separation of
dysprosium from NdFeB magnets in an acidic system by ion imprinted mesoporous
silica sealed in a dialysis bag, Green Chem. 18 (2016) 5031–5040.
[12] M. Anbia, K. Kargosha, S. Khoshbooei, Heavy metal ions removal from aqueous
media by modified magnetic mesoporous silica MCM-48, Chem. Eng. Res. Des. 93
(2015) 779–788.
[13] C. Liu, Y. Guo, Q. Hong, C. Rao, H. Zhang, Y. Dong, L. Huang, X. Lu, N. Bao, Bovine
serum albumin adsorption in mesoporous titanium dioxide: pore size and pore
chemistry effect, Langmuir 32 (2016) 3995–4003.
[14] J.Y. Ying, C.P. Mehnert, M.S. Wong, Synthesis and applications of supramolecular-
templated mesoporous materials, Angew. Chem. Int. Ed. 38 (1999) 56–77.
[15] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Triblock copolymer syntheses of mesoporous silica with periodic 50 to 300 ang-
strom pores, Science 279 (1998) 548–552.
[16] F. Gao, C. Lian, L. Zhou, H. Liu, J. Hu, Phase separation of mixed micelles and
synthesis of hierarchical porous materials, Langmuir 30 (2014) 11284–11291.
[17] B.G. Trewyn, I. Slowing, S. Giri, H.T. Chen, V.S.Y. Lin, Synthesis and functionali-
zation of a mesoporous silica nanoparticle based on the sol-gel process and appli-
cations in controlled release, Acc. Chem. Res. 40 (2007) 846–853.
[18] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartull, J.S. Beck, Ordered meso-
porous molecular sieves synthesized by a liquid-crystal template mechanism,
Nature 359 (1992) 710–712.
[19] H. Choi, J.J. Kim, Y.H. Mo, B.M. Reddy, S.E. Park, Novelty of dynamic process in
the synthesis of biocompatible silica nanotube by biomimetic glycyldodecylamide
as a soft template, Langmuir 33 (2017) 10707–10714.
T. Kaneko et al.
[20] K. Nakanishi, M. Tomita, K. Kato, Synthesis of amino-functionalized mesoporous
silica sheets and their application for metal ion capture, J. Asian Ceram. Soc. 3
(2015) 70–76.
[21] T. Orita, M. Tomita, M. Harada, K. Kato, Binding activity of avidin to the biotin
within mesoporous silica materials for bioanalytical applications, Anal. Biochem.
425 (2012) 1–9.
[22] J.L. Paris, M. Colilla, L.L. Barba, M. Manzano, M.V. Regi, Tuning mesoporous silica
dissolution in physiological environments: a review, J. Mater. Sci. 52 (2017)
8761–8771.
[23] K. Nakanishi, M. Tomita, K. Kato, Improvement in the catalytic activity of cyto-
chrome c by immobilization on a novel mesoporous silica sheet, RSC Adv. 4 (2014)
4732–4735.
[24] Y. An, M. Chen, Q. Xue, W. Liu, Preparation and self-assembly of carboxylic acid-
functionalized silica, J. Colloid Interface Sci. 311 (2007) 507–513.
[25] R. Hikosaka, F. Nagata, M. Tomita, K. Kato, Adsorption and desorption character-
istics of DNA onto the surface of amino functional mesoporous silica with various
particle morphologies, Colloids Surf. B Biointerfaces 140 (2016) 262–268.
[26] S.J. Allen, G. Mckay, J.F. Porter, Adsorption isotherm models for basic dye
5998
Journal of Environmental Chemical Engineering 6 (2018) 5990–5998
adsorption by peat in single and binary component system, J. Colloid Int. Sci. 280
(2004) 322–333.
[27] A. Kapoor, J.A. Ritter, R.T. Yang, An extended Langmuir model for adsorption of
gas mixtures on heterogeneous surfaces, Langmuir 6 (1990) 660–664.
[28] M.D. Levan, T. Vermeulen, Binary Langmuir and Freundlich isotherms for ideal
adsorbed solutions, J. Phys. Chem. 85 (1981) 3247–3250.
[29] S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular layers, J.
Am. Chem. Soc. 60 (1938) 309–319.
[30] A. Suwalski, H. Dabboue, A. Delalande, S.F. Bensamoun, F. Canon, P. Midoux,
G. Saillant, D. Klatzmann, J. PSalvetat, C. Pichon, Accelerated achilles tendon
healing by PDGF gene delivery with mesoporous silica nanoparticles, Biomaterials
31 (2010) 5237–5245.
[31] A.S.M. Chong, X.S. Zhao, Functionalization of SBA-15 with APTES and character-
ization of functionalized materials, J. Phys. Chem. B 107 (2003) 12650–12657.
[32] T.A. Saleh, Isotherm, kinetic, and thermodynamic studies on Hg(II) adsorption from
aqueous solution by silica- multiwall carbon nanotubes, J. Water Supply Res.
Technol. 21 (2015) 16721–16731.