Removal of volatile organic compounds from air using supported ionic liquid membrane containing ultraviolet-visible light-driven Nd-TiO2 nanoparticles

Article abstract of DOI:10.1016/j.molstruc.2021.130023

Volatile organic compounds (VOCs) with toxicity properties discharged from industrial emissions and combustion engines threaten human health and cause environmental problems. Therefore, it is important to develop technologies to remove VOCs from air. In this study, the performance of Nd (neodymium) -TiO₂ nanoparticles embedded in a supported ionic liquid membrane (SILM) for removal of VOCs from air was investigated. Aiming at functionalizing the ionic liquid membrane with photocatalytic capability, we developed a facile and effective approach for the removal of VOCs by coupling highly efficient photocatalysts with the SILM. Nd-TiO₂ nanoparticles, which are ultraviolet-visible (UV–Vis) light-driven photocatalysts, were prepared by sol-gel and immersed in the SILM for use as a photocatalytic membrane-based reactor. The gaseous VOCs tested were toluene, acetone, chloroform, benzene, and xylene. We confirmed that at a loading of 50 wt% Nd-TiO₂ nanoparticles (1 wt% Nd content) and 1-hexyl-3-methylimidazolium hexafluorophosphate ([HMIm]PF₆) as SILM ([HMIm]PF₆/Nd-TiO₂ SILM), the as-prepared photocatalytic membrane-based reactor exhibited higher removal efficiency of VOCs from air (60~80% removal after 10 h) under UV–Vis light illumination than individual SILM or photocatalytic systems. Effluent gas analysis revealed that 25–55% of VOCs was decomposed and mineralized in the photocatalytic membrane reactor, which exhibited higher VOCs removal efficiency than membrane separation. These results indicate that the photocatalytic membrane-based reactor containing Nd-TiO₂ nanoparticles and SILM designed in this study is highly active and stable. This study on [HMIm]PF₆/Nd-TiO₂ SILM was initiated to couple membrane separation with photocatalytic degradation of gaseous VOCs and can serve as a promising way for eliminating harmful VOCs with low concentration from air.

Full text of DOI:10.1016/j.molstruc.2021.130023

Journal of Molecular Structure 1231 (2021) 130023
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstr
Removal of volatile organic compounds from air using supported ionic
liquid membrane containing ultraviolet-visible light-driven Nd-TiO2
nanoparticles
Jinlong Lia,b,∗, Boxin Lia, Guozhe Suia,b,∗, Lijuan Dua, Yan Zhuanga, Yulin Zhanga,
Yuanfang Zoua
a College of Chemistry and Chemical Engineering, Qiqihar University, Qiqihar 161006, P. R. China
b Heilongjiang Provincial Key Laboratory of Catalytic Synthesis for Fine Chemicals, Qiqihar University, Qiqihar 161006, P. R.China
a r t i c l e i n f o
a b s t r a c t
Article history:
Volatile organic compounds (VOCs) with toxicity properties discharged from industrial emissions and
combustion engines threaten human health and cause environmental problems. Therefore, it is impor-
tant to develop technologies to remove VOCs from air. In this study, the performance of Nd (neodymium)
-TiO2 nanoparticles embedded in a supported ionic liquid membrane (SILM) for removal of VOCs from air
was investigated. Aiming at functionalizing the ionic liquid membrane with photocatalytic capability, we
developed a facile and effective approach for the removal of VOCs by coupling highly efficient photocat-
alysts with the SILM. Nd-TiO2 nanoparticles, which are ultraviolet-visible (UV–Vis) light-driven photocat-
alysts, were prepared by sol-gel and immersed in the SILM for use as a photocatalytic membrane-based
reactor. The gaseous VOCs tested were toluene, acetone, chloroform, benzene, and xylene. We confirmed
that at a loading of 50 wt% Nd-TiO2 nanoparticles (1 wt% Nd content) and 1-hexyl-3-methylimidazolium
hexafluorophosphate ([HMIm]PF6) as SILM ([HMIm]PF6/Nd-TiO2 SILM), the as-prepared photocatalytic
membrane-based reactor exhibited higher removal efficiency of VOCs from air (60~80% removal after
10 h) under UV–Vis light illumination than individual SILM or photocatalytic systems. Effluent gas analy-
sis revealed that 25–55% of VOCs was decomposed and mineralized in the photocatalytic membrane re-
actor, which exhibited higher VOCs removal efficiency than membrane separation. These results indicate
that the photocatalytic membrane-based reactor containing Nd-TiO2 nanoparticles and SILM designed in
this study is highly active and stable. This study on [HMIm]PF6/Nd-TiO2 SILM was initiated to couple
membrane separation with photocatalytic degradation of gaseous VOCs and can serve as a promising
way for eliminating harmful VOCs with low concentration from air.
Received 27 December 2020
Revised 20 January 2021
Accepted 24 January 2021
Available online 29 January 2021
Keywords:
Photocatalytic membrane-based reactor
Volatile organic compounds
Supported ionic liquid membrane
Photocatalytic reaction
© 2021 Elsevier B.V. All rights reserved.
1. Introduction
environmental issue [3]. Therefore, to control the concentration of
VOC pollutants, the technology of VOC abatement has become a
Volatile organic compounds, also known as VOCs, refer to the
organic chemical compounds with significant saturated vapor pres-
sure (>70 Pa at normal temperature) and boiling points within the
range of 50–260 °C at normal pressure [1], mainly include hydro-
carbons, benzenes, organic chlorides, aldehydes, ketones, and fatty
acids [2]. These VOCs, principally discharged from industrial emis-
sions and combustion engines, generally have high-risk biologi-
cal toxicity properties, which threaten human health. In addition,
these VOCs will directly or indirectly cause environmental prob-
lems such as haze and photochemical smog, which pose a global
research topic of interest in recent years. However, conventional
techniques are less efficient at handling low concentrations of VOC
[4]; therefore, new techniques are needed to degrade VOCs present
at low concentration.
Photocatalytic technology is the optimal strategy for VOC abate-
ment, because it can mineralize VOCs to fewer polluting sub-
stances by efficient utilization of light energy [5]. Furthermore, it
is one of the most effective treatments to remove gaseous VOCs
with low concentration due to its efficiency and energy-saving
advantage at normal temperature [6]. Therefore, many catalysts
with high photoactive performance are being developed to degrade
gaseous VOCs from air. Among them, TiO2 has been widely used
owing to its good stability and low toxicity in controlling pollution
[7]. However, TiO2 has two obvious drawbacks: (1) the efficiency of
∗
Corresponding authors.
E-mail addresses: jinlong@qqhru.edu.cn (J. Li), gzhsui@qqhru.edu.cn (G. Sui).
https://doi.org/10.1016/j.molstruc.2021.130023
0022-2860/© 2021 Elsevier B.V. All rights reserved.

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
natural sunlight utilization is low because it can only absorb near-
ultraviolet light caused by wide forbidden band width (~3.2 eV)
[8]; (2) the photocatalytic efficiency is reduced due to rapid re-
combination of photogenerated electron-hole pairs [9]. Therefore,
much research attention has been devoted to the improvement of
the photocatalytic activity of TiO2 by modifications, such as doping
of metal or non-metal, and morphological modification [10].
In recent publications, rare earth metals have been proved to
be one of the most efficient dopants for effectively preventing the
recombination of photogenerated electron-hole pairs and extend-
ing the range of response spectrum [11]. As a conventional doping
metal, Nd exhibited higher photocatalytic performance and utiliza-
tion of light energy due to the 4f electron transition in rare earth
metals [12]. Alam et al. [13] synthesized Nd-V-ZnO with spindle-
like shape by using an ultrasonic-assisted sol-gel method, and the
doping of Nd and V metals enhanced the degradation efficiency of
aqueous organic pollutants under visible light. Wang et al. [14] de-
posited Nd and TiO2 on SBA-15 substrate with mesoporous struc-
ture, and the obtained Nd-TiO2-SBA-15 showed high photoactive
and good acid-base adaptability for methyl orange in the pH range
of 2–10. Nd-doped TiO2 can increase the adsorption capacity of
the catalyst surface, thus improving the photocatalytic activity [15].
However, research about gaseous VOCs abatement using Nd-doped
oxide semiconductor has been less reported, especially for remov-
ing gaseous VOCs with low concentration. Therefore, the technol-
ogy of photocatalysis is usually coupled with other technologies to
improve the efficiency of gaseous VOC removal.
(CHCl3) single-gas and multiple-gas removal performances of
[HMIm]PF6/Nd-TiO2 SILM were investigated under visible light il-
lumination, and the mechanism underlying the removal of gaseous
VOCs was discussed. We also investigated the effects of operation
conditions on the removal performance, and the experimental re-
sults were correlated to the variation of the operation conditions.
2. Experimental
2.1. Materials
Tetrabutyl titanate (C16 H36O4Ti) was selected as the Ti source to
prepare TiO2, glacial acetic acid (CH3COOH) was used as a hydrol-
ysis inhibitor of tetrabutyl titanate, nitric acid (HNO3) was used
to adjust the solution pH, neodymium nitrate (Nd(NO3)3•6H2O)
was the source of metal Nd, and absolute ethanol (CH3CH2OH)
was used as solvent. Toluene (C7H8), acetone (C3H6O), benzene
(C6H6), xylene (C8H10), and chloroform (CHCl3) were chosen as
gaseous VOC samples for the removal experiment. The ionic liq-
uid used was 1-hexyl-3-methylimidazolium hexafluorophosphate
([HMIm]PF6). The chemicals were purchased from Aladdin Chem-
istry Co. Ltd. and were of analytical grade and were used with-
out purification. Polyvinylidene fluoride film (PVDF) with 95 μm
of thickness, 0.1 μm of mean pore size, and 70% porosity was pro-
vided by Millipore Co. Ltd. The deionized water used in this exper-
iment was produced from a UPE-60 water purification equipment.
Membrane separation technology has been widely used in the
separation of VOCs because of its high recovery and high effi-
ciency [16]. Comparing with conventional polymeric membranes,
supported liquid membranes exhibit higher VOC permeability and
selectivity due to the properties of the liquid solvent. According to
the facilitated transport theory, ionic liquids are promising, offer-
ing higher VOC permeability and selectivity due to their thermal
stability and low volatility [17, 18]. Wang et al. [19] reviewed the
recent applications of supported ionic liquid membranes (SILMs)
for VOCs separation and stated the transport mechanisms of or-
ganic compounds. Uragami et al. [20] reported the separation
performance of the poly(styrene)-b-poly(dimethylsiloxane) mem-
branes containing 1-allyl-3-butylimidazilium bis (trifluoromethane
sulfonyl) imide ([ABIM]TFSI) for removing chloroform, benzene,
and toluene from aqueous solutions. The authors revealed that
[ABIM]TFSI plays an important role as an absorbent to selectively
separate VOCs from the aqueous solution. Dahi et al. [21] prepared
SILMs by immobilizing [C(4)C(1)im][BF4] in a porous Matrimid
membrane. The prepared SILMs showed better performance of sep-
aration of VOCs, such as ethanol and cyclohexane, from the so-
lution due to the best sorption capacities and the good sorp-
tion selectivity of [C(4)C(1)im][BF4] for VOCs. However, most stud-
ies only focus on VOCs separation in the liquid phase, and the
investigations on the separation of gaseous VOCs (such as ben-
zene, toluene, and formaldehyde) with low concentration from air
are much scarce. Furthermore, the gaseous VOCs separated using
membrane technology cannot be degraded into nontoxic CO2 and
H2O through its physical process. It is therefore necessary to de-
velop and study a typical membrane process coupled with other
treatment technology to improve the elimination performance of
harmful VOCs from air.
2.2. Preparation of Nd-TiO2 nanoparticles
The Nd-TiO2 nanoparticles were synthesized via
a sol-gel
method: 2 mL of C16 H36O4Ti was dissolved in 9.8 mL of
CH3CH2OH, to which 2 mL of CH3COOH was further added dur-
ing stirring. The pH of the mixture was adjusted to ~3 by adding
HNO3, and the mixture was stirred vigorously for 1 h to form so-
lution A. Next, an appropriate amount of Nd(NO3)3•6H2O, which
was varied according to the intended mass ratio of Nd:TiO2 (0,
0.5, 0.8, 1.0, and 1.5% (wt%)), was added in the deionized water
and stirred continuously until dissolved, and the obtained solution
was labeled as solution B. Solution B was slowly added to solution
A at the dropping rate of 1 mL min−1, and the resulting solution
was kept under continuous stirring for 24 h at 40 °C. The gel was
formed after aging in air for 48 h and then dried at 60 °C for 10 h
and washed with deionized water several times. The final Nd-TiO2
nanoparticles were obtained after calcination in an air furnace at
550 °C for 2 h at a heating rate of 5 °C min−1
.
2.3. Preparation of SILM
A flat membrane module with membrane area of 0.065 m2
was used for the VOCs removal measurement. The PVDF mem-
brane was on the porous polypropylene plate, which was used
as permeated gas transport channel. The detailed construction of
the flat membrane module is described in our previous report
[22]. [HMIm]PF6 (2.5 g) and Nd-TiO2 nanoparticles (0.2 g) were
mixed completely, and were then uniformly distributed onto the
hydrophilic surface of the PVDF supported membrane treated with
corona discharge. Some [HMIm]PF6 was spread on this top hy-
drophilic surface with a wetting tension of 70 mN/m and then
soaked into the pores of PVDF membrane [22]. The other sur-
face of PVDF hydrophobic bottom layer will support the top
[HMIm]PF6 liquid layer, due to the hydrophilic-hydrophobic inter-
action between the liquid and the surface of PVDF membrane. A
[HMIm]PF6/Nd-TiO2 SILM was obtained and was durable under a
transmembrane pressure up to 2–3 atm. The membrane thickness
of the obtained [HMIm]PF6/Nd-TiO2 SILM was ~20 μm, as esti-
mated from the quality of the used [HMIm]PF6.
In this study, we present the detailed study of the VOC re-
moval performance of
a coupled process of SILM and photo-
catalysis based on 1-hexyl-3-methylimidazolium hexafluorophos-
phate ([HMIm]PF6) and Nd-TiO2 nanoparticles ([HMIm]PF6/Nd-TiO2
SILM). Particularly, this study focused on the synergistic effect
of [HMIm]PF6/Nd-TiO2 SILM, i.e., the concentration of [HMIm]PF6
SILM and photocatalytic action of Nd-TiO2. Toluene (C7H8), ace-
tone (C3H6O), benzene (C6H6), xylene (C8H10), and chloroform
2

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Fig. 1. Membrane catalytic diagram for removing VOCs.
2.4. Characterization of Nd-TiO2 nanoparticles
X-ray diffraction (XRD) analysis was performed over the 2θ
range of 10–80 ° by using the D8 X-ray diffractometer (Bruker-AXS,
Germany). Fourier transform infrared (FT-IR) spectra were acquired
using the Nexus 670 infrared spectrometer (Nicolet, USA). UV-
Visible diffuse-reflectance spectra (UV–Vis DRS) were measured on
Lamda 35 UV–Vis spectrophotometer (PE, USA), using BaSO4 as a
reflectance standard. X-ray photoelectron spectroscopy (XPS) anal-
ysis was performed on ESCALAB250Xi X-ray photoelectron spec-
trometer (Thermo, USA). Scanning electron microscopy (SEM) im-
ages were recorded on the S-3400 scanning electron microscope
(Hitachi, Japan), and transmission electron microscopy (TEM) im-
ages were acquired with the H-7650 transmission electron micro-
scope (Hitachi, Japan). The specific surface area was analyzed us-
ing the Nova 2000e N2 adsorption-desorption apparatus (Quan-
tachrome, USA) in the relative pressure ranging of 0.05–0.35.
Fig. 2. XRD patterns of (a) TiO2, (b) 0.5 wt% Nd-TiO2, (c) 0.8 wt% Nd-TiO2, (d) 1.0
wt% Nd-TiO2, and (e) 1.5 wt% Nd-TiO2.
2.5. VOCs removal measurement
Gaseous VOCs removal from air was conducted, in a 0.5 m3
black chamber with Teflon film coating to avoid VOC adsorption
(Fig. 1). The air containing VOCs in the chamber was mixed com-
pletely by horizontal and vertical fans. The experimental apparatus
comprised a SILM membrane module and a 500 W xenon lamp
with a light length cut-off filter (<420 nm or >420 nm) placed in
the chamber, diaphragm vacuum pumps (DAP-6D, ULVAC, Japan),
a hot plate, a gas chromatography (GC) system (SP-6890, Lunan-
RUIHONG, China), pressure sensors, and a 500 W xenon lamp. A
certain amount of liquid VOCs, calculated through the determined
room temperature and atmospheric pressure, was injected onto
the hot plate in the chamber and then vaporized to gaseous VOCs
to achieve the set concentration of 1 000 ppm. When the liquid
VOCs were vaporized completely, the concentration in the chamber
was determined by the GC with a hydrogen flame detector (FID).
The membrane system was operated using a diaphragm vacuum
pump and a needle valve. The pressure at the permeate side of
the membrane module was maintained at 10 kPa and the sweep
air flow rate was controlled at 0.5 L min−1, they were detected us-
ing a pressure sensor and a flowmeter, respectively. After operating
for 30 min in the dark, the adsorption-desorption equilibrium of
gaseous VOCs on the [HMIm]PF6/Nd-TiO2 SILM was achieved and
then a 500 W xenon lamp, with a light length cut-off filter, was
turned on to the membrane module. The gaseous VOCs was con-
centrated on the surface of [HMIm]PF6/Nd-TiO2 SILM under the
difference of VOCs partial pressure, and most of which was de-
graded to CO2 and H2O by Nd-TiO2 under visible light illumination.
Meanwhile, a little gaseous VOCs and the degradation products
permeated the [HMIm]PF6/Nd-TiO2 SILM into the permeate side of
membrane module according to the solution-diffusion mechanism.
The composition and concentration of VOCs in the chamber and at
the outlet of the membrane module were measured by GC with a
hydrogen flame detector (FID) and a thermal conductivity detector
(TCD), respectively. In addition, the photocatalytic degradation ef-
ficiency of VOCs was calculated by measuring the concentration of
CO2 in the membrane permeate stream. To prevent negative pres-
sure in the chamber, fresh air could be supplied through the inlet
settled at the wall of the chamber.
3. Results and discussion
3.1. Characterization
The XRD patterns of TiO2 and Nd-TiO2 nanoparticles with dif-
ferent contents of Nd are shown in Fig. 2. All Nd-TiO2 nanoparti-
cles exhibited XRD profiles similar to that of TiO2, which shows
diffraction peaks at 2θ of 25.3°, 37.9°, 48.1°, 54.0°, 55.1°, 62.7°,
68.8°, 70.5°, and 75.4° that correspond to the anatase (101), (004),
(200), (105), (211), (204), (116), (220), and (215) lattice planes
3

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Fig. 3. TEM images of (a) pure TiO2, (b) 1 wt% Nd-TiO2, SEM image of (c) 1 wt% Nd-TiO2, elemental maps of (d) Ti atoms, (e) O atoms, and (f) Nd atoms, and (g) EDS spectra
of 1 wt% Nd-TiO2.
(JCPDS card no. 21–1272), respectively. This indicated that other
crystal forms of TiO2 were not formed during preparing Nd-TiO2
nanoparticles. The ionic radius of Nd3+ (0.104 nm) is larger than
Ti4+ (0.088 nm) [23], therefore, Nd3+cannot replace Ti4+ in the lat-
tice, but in the state of Nd2O3, it may be adsorbed on the sur-
face of TiO2. However, there was no characteristic diffraction peak
of Nd2O3 at 2θ = 32.5° [24] because the doping amount of Nd
was low. It can be seen from the figure that the full width half-
maximum (FWHM) of the catalysts becomes larger after doping
Nd. The average crystallite size (D) of Nd-TiO2 nanoparticles, which
was calculated from the Scherrer equation based on the FWHM of
(101) XRD peak, decreased as shown in Table 1, indicating that the
specific surface area of Nd-TiO2 nanoparticles may be increased by
Nd doping [25].
TEM and SEM images of pure TiO2 and Nd-TiO2 nanoparticles,
and the elemental maps of Ti, O and Nd in 1.0 wt% Nd-TiO2 are
shown in Fig. 3. All the prepared catalysts showed a spherical or
globular particle shape and a slight agglomeration. The average
particle size of pure TiO2 was ca. 40 nm, whereas the average
particle size of Nd-TiO2 nanoparticles was reduced to 15–20 nm.
4

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Table 1
Pore size, specific surface area and average anatase crystallite size parameters of
TiO2 and Nd-TiO2.
Pore Diameter
(nm)
Surface Area
Anatase Crystallite
Size (nm)
−
1
Sample
(m2
g
)
TiO2
7.23
6.56
6.57
6.59
6.57
45.34
21.53
13.72
12.24
12.22
11.57
Nd-TiO2(0.5 wt%)
Nd-TiO2(0.8 wt%)
Nd-TiO2(1.0 wt%)
Nd-TiO2(1.5 wt%)
105.82
130.62
150.27
129.83
Fig. 5. UV–Vis diffuse reflectance spectra of (a) TiO2, (b) 0.5 wt% Nd-TiO2, (c) 0.8
wt% Nd-TiO2, (d) 1.0 wt% Nd-TiO2, and (e) 1.5 wt% Nd-TiO2.
and the bridging mode of Ti-O-Ti bond, which are characteris-
tic peaks of TiO2 [26]. Notably, the Ti-O or Ti-O-Ti bonds in Nd-
TiO2 nanoparticles were significantly red-shifted, which might have
been caused by the doping of Nd. The absorption peak at 1623
cm−1 and the broad absorption peak at 3425 cm−1 were attributed
–
to the bending and stretching vibrations of O H, which are related
to water adsorbing on the surface of the catalyst [27]. Compared
with pure TiO2, the absorption peak at 3425 cm−1 became stronger
with increasing of Nd doping quantity. This indicated that the Nd
dopant enhances water adsorption resulting in more hydroxyl rad-
icals, increasing the photocatalytic performance of Nd-TiO2 [28].
DRS spectra of TiO2 and Nd-TiO2 nanoparticles are shown in
Fig. 5. Compared with pure TiO2, the absorption spectra of Nd-
TiO2 nanoparticles showed an obvious red shift extended to the
visible light region. This was due to the charge transfer transition
between the dopant electrons of Nd and the TiO2 conduction band
[29], indicating that the light absorption capacity of the catalyst
was greatly improved [30]. Furthermore, three absorption bands
appeared in the spectra of Nd-TiO2 nanoparticles as the reference
of Nd(NO3)3 DRS spectra, which was attributed to the 4f transition
of Nd(III) ions. By varying the content of Nd, the absorption capac-
ity of Nd-TiO2 nanoparticles under visible light illumination was
different, and Nd-TiO2 nanoparticles with 1.0 wt% of Nd exhibited
strong visible light absorption.
Fig. 4. FT-IR spectra of (a) TiO2, (b) 0.5 wt% Nd-TiO2, (c) 0.8 wt% Nd-TiO2, (d) 1.0
wt% Nd-TiO2, and (e) 1.5 wt% Nd-TiO2.
This indicated that the growth of the catalyst grains can be effec-
tively suppressed by the doping of Nd. This reduction of the parti-
cle size can increase the specific surface area of Nd-TiO2 nanopar-
ticles, which caused a further increase of the photocatalytic activ-
ity. Fig. 3(d-f) show that the elements of Ti, O, and Nd are dis-
persed uniformly in 1.0 wt% Nd-TiO2 nanoparticles. Fig. 3(g) shows
that the actual contents of Ti, Nd, and O correspond to 61.26 wt%,
0.95 wt%, and 37.79 wt%, respectively, which are consistent with
the theoretical contents in Nd-TiO2 nanoparticles. The difference
between the actual content and the theoretical content can be at-
tributed to instrument error.
FT-IR spectra of TiO2 and Nd-TiO2 nanoparticles are shown in
Fig. 4. The wide absorption bands observed over the range of 450–
800 cm−1 were assigned to the stretching mode of Ti-O bond
N2 adsorption-desorption isotherms and the corresponding
pore size distribution curves of pure TiO2 and Nd-TiO2 (1.0 wt%)
Fig. 6. (a) N2 adsorption-desorption isotherms and (b) pore size distribution curves of TiO2 and Nd-TiO2 nanoparticles.
5

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Fig. 7. (a) Overview of the XPS spectra and individual high-resolution XPS spectra for (b) Ti 2p, (c) O 1 s, (d) C 1 s, and (e) Nd 3d of Nd-TiO2 (1.0 wt%) naoparticles.
nanoparticles are illustrated in Fig. 6. Table 1 shows the pore size
and specific surface area parameters of the various component
catalysts. According to the IUPAC classification, the adsorption-
desorption isotherms of samples with an H1 hysteresis loop are all
type IV curves [31], which indicate the presence of a mesoporous
structure in the samples. The specific surface area and average pore
size of all samples are summarized in Table 1. The specific surface
a maximum when the amount of Nd was 1.0 wt%. Large specific
surface area of the catalyst can provide more reactive centers, in-
creasing the photocatalytic activity of the catalyst [32].
XPS spectra were measured to analyze the chemical states of
Nd-TiO2 (1.0 wt%) nanoparticles as shown in Fig. 7. Fig. 7(a) sum-
marizes the XPS spectra of Nd-TiO2 (1.0 wt%) nanoparticles, which
contained four elements of Ti, O, C and Nd, with the binding en-
ergies of 284.1 eV, 458.1 eV, 529.1 eV, and 974.1 eV correspond-
ing to C 1 s, Ti 2p, O 1 s and Nd 3d, respectively. As shown in
Fig. 7(b), two main peaks were observed at 458.3 eV and 464.5 eV,
which corresponded to Ti2p 3/2 and Ti2p 1/2 split peaks. This in-
dicates that the Ti element exists as Ti4+ in the samples [33, 34].
−
1, and
area was 45.34, 105.82, 130.62, 150.27, and 129.83 m2
g
the average pore size was 7.23, 6.56, 6.57, 6.57, 6.59, and 6.57 nm
for pure TiO2, 0.5 wt% Nd-TiO2, 0.8 wt% Nd-TiO2, 1.0 wt% Nd-TiO2,
and 1.5 wt% Nd-TiO2, respectively. The specific surface area of Nd-
TiO2 nanoparticles increased with increasing Nd content, reaching
6

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Fig. 10. Degradation rate and removal rate of VOCs by [HMIm]PF6/Nd-TiO2 SILM
Fig. 8. Removal performance of toluene under different methods.
under UV light and visible light illumination for 10 h.
However, these peaks shifted slightly to a higher region of bind-
ing energy because the variation in the chemical state was caused
by adding Nd in TiO2. The O1s spectrum in Fig. 7(c) shows two
peaks at 529.8 eV and 532.0 eV corresponding to the lattice oxy-
gen and the oxygen from the surface-adsorbed O–H, respectively.
In Fig. 7(d), the peak at 284.8 eV was attributed to surface impu-
rity indefinite carbon (C = C bonds), whereas the peak at 288.7 eV
was assigned to C–O–Ti, disclosing that some carbon atoms sub-
stituted titanium atoms in the TiO2 lattice [35]. In Fig. 7(e), the
peaks of binding energy at 974.8 eV and 994.8 eV were attributed
to Nd3d 5/2 and Nd3d 3/2, respectively, indicating that Nd exists
as the form of Nd3+ in the TiO2 lattice.
and transported through the [HMIm]PF6 membrane according to
the solution-diffusion mechanism. Therefore, the membrane sep-
aration process exhibited the worst performance of removal of
toluene compared with other separation methods. However, adding
Nd-TiO2 nanoparticles showed enhanced removal performance: the
half reduction time under UV light and visible light was 3.0 h and
1.5 h shorter than membrane separation. This was because toluene
was enriched on the membrane surface through membrane sepa-
ration and was degraded into CO2 and H2O by Nd-TiO2 nanoparti-
cles under UV and visible light illumination. This synergistic effect
between membrane separation and photocatalysis accelerated the
rate of toluene removal through [HMIm]PF6/Nd-TiO2 SILM under
light illumination.
3.2. VOC removal performance
The process of [HMIm]PF6/Nd-TiO2 SILM under UV and visi-
ble light illumination was further tested to evaluate the removal
of other gaseous VOCs, such as acetone, chloroform, benzene, and
xylene, and the experimental results are shown in Fig. 9. The
variation of VOC concentration in the chamber was like that of
toluene, and the photocatalytic membrane reaction process con-
taining [HMIm]PF6/Nd-TiO2 SILM exhibited high performance for
the removal of many types of VOCs. This indicated that the pho-
tocatalytic membrane reaction process can remove VOCs with low
concentration. In addition, the removal efficiency under UV light
was higher than that under visible light: for toluene, acetone, chlo-
roform, benzene, and xylene, the half reduction time was 4.2 h,
4.0 h, 3.8 h, 3.9 h, and 5.0 h under UV light, respectively, whereas
the half reduction time was 5.7 h, 6.0 h, 5.7 h, 5.2 h, and 6.5 h un-
Performing removal of VOCs were measured using only mem-
brane ([HMIm]PF6 SILM) separation, photocatalysis (Nd-TiO2) un-
der visible light, and membrane ([HMIm]PF6/Nd-TiO2 SILM) cat-
alytic separation under UV light and under visible light. The ex-
perimental results of toluene are shown in Fig. 8. All removal
methods showed similar exponential reduction curves of toluene
concentration in the chamber. The half reduction time of mem-
brane separation, photocatalysis under visible light, and membrane
catalytic separation under UV light and under visible light was
7.2 h, 6.5 h, 5.7 h, and 4.2 h, respectively. For only membrane
([HMIm]PF6 SILM) separation processes, toluene was concentrated
on the membrane surface, dissolved in [HMIm]PF6 ionic liquid,
Fig. 9. Removal performance of VOCs by [HMIm]PF6/Nd-TiO2 SILM under (a) UV light and (b) visible light illumination.
7

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
Fig. 11. Catalytic membrane-based mechanism diagram of VOCs removal through [HMIm]PF6/Nd-TiO2 SILM under UV light and visible light illumination.
der visible light, respectively. This was because the absorbance of
Nd-TiO2 in the UV light region was apparently higher than that in
the visible light region, according to the DRS analytics result of Nd-
TiO2 nanoparticles. However, performing removal of different VOCs
was different. This may be caused by two reasons: 1) the compat-
ibility between VOCs and [HMIm]PF6 ionic liquid might have fa-
cilitated the enrichment of VOCs on the surface of SILM and the
dissolution of VOCs in [HMIm]PF6 ionic liquid [36]; 2) the decom-
position of VOC molecules depends on the complexity of electronic
and spatial structure of VOC molecules.
4. Conclusion
A
SILM with photocatalytic function containing 1-hexyl-3-
methylimidazolium hexafluorophosphate ionic liquid and synthe-
sized Nd-TiO2 nanoparticles was devised. The as-prepared pho-
tocatalytic SILM under UV and visible light illumination was ef-
fective for VOC removal to improve air quality. The VOC concen-
tration in the chamber (1000 ppm of toluene, acetone, chloro-
form, benzene, and xylene) could be reduced to a lower level of
less than 200~400 ppm over 10 h of operation time. Perform-
ing VOC removal was illustrated in terms of the synergistic ef-
fect of membrane separation and photocatalytic reaction, and the
VOC removal mechanism of the catalytic membrane-based reac-
tor was presented. We conclude this SILM containing photocat-
alytic nanoparticles under UV–Vis light illumination can provide
a highly efficient and environment friendly method for removing
VOCs. mmc1.docx
Fig. 10 shows the degradation rate and rate of removal of
toluene, acetone, chloroform, benzene, and xylene by [HMIm]PF6/
Nd-TiO2 SILM under UV light and visible light illumination for
10 h. Owing to the synergy between the membrane separation
process and the photocatalysis, the VOC removal rate was higher
than 70.0% and 60.0% under UV light and visible light, respec-
tively, in which only a part of the VOCs could be degraded to
nontoxic CO2 and H2O. The concentration of VOCs was measured
at the outlet of the membrane module; the degradation rate of
toluene, acetone, chloroform, benzene, and xylene was 43.94%,
52.76%, 51.72%, 51.67%, and 39.32% under UV light illumination, and
28.28%, 39.43%, 39.82%, 39.07%, and 28.77% under visible light illu-
mination, respectively.
Declaration of Competing Interest
There is no conflict of interest.
CRediT authorship contribution statement
Whether the SILMs can be practically applied depends on its
stability. In this study, the amount of permeated air and the per-
meate side pressure were determined during the operation, which
was evaluated as the lifetime of the SILM. The flow rate of perme-
ated air and permeate side pressure were usually 1.8–5.0 cm3/min
and 1.5–2.2 kPa, respectively when the flow rate of sweep air was
0. When the operation time of [HMIm]PF6/Nd-TiO2 SILM was over
1000 h, the stability of SILM did not vary.
Boxin Li: Writing - original draft. Guozhe Sui: Conceptualiza-
tion, Data curation, Resources. Lijuan Du: Writing - review & edit-
ing, Data curation. Yan Zhuang: Data curation, Resources. Yulin
Zhang: Data curation, Investigation. Yuanfang Zou: Writing - re-
view & editing.
Acknowledgments
3.3. Removal mechanism
This work was supported by the Research Project of Edu-
cation Ministry of Heilongjiang Province of China (135209217,
135409101). We would like to thank Editage (www.editage.cn) for
English language editing.
The catalytic membrane-based mechanism diagram of VOC re-
moval through [HMIm]PF6/Nd-TiO2 SILM under UV light and vis-
ible light illumination is shown in Fig. 11. Gaseous VOCs will
be concentrated on the membrane surface under the effect of
the difference of partial pressure. Subsequently, a portion of the
VOCs dissolves in the [HMIm]PF6 ionic liquid according to the
similarity-compatible principle and then diffuses through the ionic
liquid membrane, and the ionic liquid can be dissociated and
reused. Meanwhile, most of concentrated VOCs can be degraded
to CO2 and H2O by the photocatalytic effect of Nd-TiO2 nanopar-
ticles under UV–Vis light, and the degradation products permeate
though the ionic liquid membrane. With the photocatalyst Nd-TiO2
nanoparticles, Nd acts as an electron by supplying ions electrons
to TiO2, and the photogenerated electrons react with O2 to form
O2−radicals, whereas the photogenerated holes react with water to
produce •OH radicals, which also can degrade VOCs. This synergis-
tic effect can facilitate VOC removal.
References
[1] W.X. Zou, B. Gao, Y.S. Ok, L. Dong, Integrated adsorption and photocat-
alytic degradation of volatile organic compounds (VOCs) using carbon-based
nanocomposites: a critical review, Chemosphere 218 (2019) 845–859, doi:10.
1016/j.chemosphere.2018.11.175.
[2] X.Y. Zhang, B. Gao, A.E. Creamer, C.C. Cao, Y.C. Li, Adsorption of VOCs onto
engineered carbon materials: a review, J. Hazard. Mater. 338 (2017) 102–123,
doi:10.1016/j.jhazmat.2017.05.013.
[3] Y. Naciri, A. Hsini, Z. Ajmal, J.A. Navio, B. Bakiz, A. Albourine, M. Ezahri,
A. Benlhachemi, Recent progress on the enhancement of photocatalytic prop-
erties of BiPO4 using pi-conjugated materials, Adv Colloid Interface Sci 280
(2020) 102160, doi:10.1016/j.cis.2020.102160.
[4] H. Nigar, N. Navascues, O. de la Iglesia, R. Mallada, J. Santamaria, Removal of
VOCs at trace concentration levels from humid air by microwave swing ad-
sorption, kinetics and proper sorbent selection, Sep Purif Technol 151 (2015)
193–200, doi:10.1016/j.seppur.2015.07.019.
8

J. Li, B. Li, G. Sui et al.
Journal of Molecular Structure 1231 (2021) 130023
[5] G.Z. Sui, J.L. Li, L.J. Du, Y. Zhuang, Y.L. Zhang, Y.F. Zou, B.X. Li, Preparation and
characterization of g-C3N4/Ag–TiO2 ternary hollowsphere nanoheterojunction
catalyst with high visible light photocatalytic performance, J Alloys Compd 283
(2020) 153851, doi:10.1016/j.jallcom.2020.153851.
[21] A. Dahi, K. Fatyeyeva, D. Langevin, C. Chappey, S.P. Rogalsky, O.P. Tarasyuk,
A. Benamor, S. Marais, Supported ionic liquid membranes for water and
volatile organic compounds separation: sorption and permeation properties, J
Memb Sci 458 (2014) 164–178, doi:10.1016/j.memsci.2014.01.031.
[22] J.L. Li, T.T. Zhao, G.Z. Sui, S.Q. Jia, CO2 separation from air using microporous
polyvinylidene fluoride-supported triethylene glycol/alkanolamine liquid mem-
branes, Mater Express 6 (2016) 183–190, doi:10.1166/mex.2016.1291.
[23] T.L.R. Hewer, E.C.C. Souza, T.S. Martins, E.N.S. Muccillo, R.S. Freire, Influence of
neodymium ions on photocatalytic activity of TiO2 synthesized by sol–gel and
precipitation methods, J Mol Catal A–Chem 336 (2011) 58–63, doi:10.1016/j.
molcata.2010.12.010.
[6] O. Debono, V. Hequet, L.Le Coq, N. Locoge, F. Thevenet, VOC ternary mix-
ture effect on ppb level photocatalytic oxidation: removal kinetic, reaction
intermediates and mineralization, Appl Catal B–Environ 218 (2017) 359–369,
doi:10.1016/j.apcatb.2017.06.070.
[7] L.Y. Wang, H. Xu, J.K. Gao, J.M. Yao, Q.C. Zhang, Recent progress in metal-
organic frameworks-based hydrogels and aerogels and their applications, Co-
ord Chem Rev 398 (2019) 213016, doi:10.1016/j.ccr.2019.213016.
[8] C. Wang, Y.H. Ao, P.F. Wang, J. Hou, J. Qian, Preparation of cerium and nitro-
gen co-doped titania hollow spheres with enhanced visible light photocatalytic
performance, Powder Technol 210 (2011) 203–207, doi:10.1016/j.powtec.2011.
03.015.
[24] F. Ruggieri, A.A. D’Archivio, M. Fanelli, S. Santucci, Photocatalytic degradation
of linuron in aqueous suspensions of TiO2, RSC Adv 1 (2011) 611–618, doi:10.
1039/c1ra00133g.
[25] S. Ida, P. Wilson, B. Neppolian, M. Sathish, P. Karthik, P. Ravi, Ultrasoni-
cally aided Selective Stabilization of Pyrrolic type nitrogen by one pot nitro-
gen doped and hydrothermally reduced Graphene oxide/Titania nanocompos-
ite (N-TiO2/N-RGO) for H-2 production, Ultrason Sonochem 57 (2019) 62–72,
doi:10.1016/j.ultsonch.2019.04.041.
[9] Y. Wang, Y. Wang, Y.L. Meng, H.M. Ding, Y.K. Shan, X. Zhao, X.Z. Tang, A highly
efficient visible-light-activated photocatalyst based on bismuth- and sulfur-
codoped TiO2, J Phys Chem C 112 (2008) 6620–6626, doi:10.1021/jp7110007.
[10] C. Lei, J.K. Gao, W.J. Ren, Y.B. Xie, S.Y.H. Abdalkarim, S.L. Wang, Q.Q. Ni, J.M. Yao,
Fabrication of metal-organic frameworks@cellulose aerogels composite mate-
rials for removal of heavy metal ions in water, Carbohydr Polym 205 (2019)
35–41, doi:10.1016/j.carbpol.2018.10.029.
[26] A. Bokare, M. Pai, A.A. Athawale, Surface modified Nd doped TiO2 nanoparti-
cles as photocatalysts in UV and solar light irradiation, Sol Energy 91 (2013)
111–119, doi:10.1016/j.solener.2013.02.005.
[11] F. Soumya, N. Yassine, H. Abdelghani, B. Hanane, B. Asmae, A. Abdallah, N. Ab-
derrazak, G. Taoufiq, Evolution of the physicochemical and photocatalytic prop-
erties of BaO embedded in bismuth phosphovanadates glasses, Chem Phys Lett
763 (2021) 138173, doi:10.1016/j.cplett.2020.138173.
[12] J.L. Li, S.Q. Jia, G.Z. Sui, L.J. Du, B.X. Li, Preparation of hollow Nd/TiO2 sub-
microspheres with enhanced visible-light photocatalytic activity, RSC Adv 7
(2017) 34857–34865, doi:10.1039/c7ra05228f.
[27] D.F. Sun, K. Wang, Z.J. Xu, R.X. Li, Synthesis and photocatalytic activity of sul-
fate modified Nd-doped TiO2 under visible light irradiation, J Rare Earth 33
(2015) 491–497, doi:10.1016/S1002-0721(14)60446-4.
[28] A. Ray, A. Roy, P. Sadhukhan, S.R. Chowdhury, P. Maji, S.K. Bhattacharya, S. Das,
Electrochemical properties of TiO2-V2O5, nanocomposites as a high perfor-
mance supercapacitors electrode material, Appl Surf Sci 433 (2018) 581–591,
doi:10.1016/j.apsusc.2018.02.277.
[13] U. Alam, T.A. Shah, A. Khan, M. Muneer, One-pot ultrasonic assisted sol-
gel synthesis of spindle-like Nd and V codoped ZnO for efficient photocat-
alytic degradation of organic pollutants, Sep Purif Technol 212 (2019) 427–437,
doi:10.1016/j.seppur.2018.11.048.
[29] N.R. Khalid, E. Ahmed, Z.L. Hong, Y.W. Zhang, M. Ullah, M. Ahmed, Graphene
modified Nd/TiO2 photocatalyst for methyl orange degradation under visible
light irradiation, Ceram Int 39 (2013) 3569–3575, doi:10.1016/j.ceramint.2012.
10.183.
[14] S. Wang, Z.Y. Wang, Y. Wang, C.H. Xia, E. Hong, L.M. Bai, T. Li, B.Q. Wang, Study
on the controlled synthesis and photocatalytic performance of rare earth Nd
deposited on mesoporous TiO2 photocatalysts, Sci Total Environ 652 (2019)
85–92, doi:10.1016/j.scitotenv.2018.10.154.
[30] J. Du, X. Gu, Q. Wu, J. Liu, H.Z. Guo, J.G. Zou, Hydrophilic and photocatalytic
activities of Nd-doped titanium dioxide thin films, Trans Nonferrous Met Soc
China 25 (2015) 2601–2607, doi:10.1016/S1003-6326(15)63881-X.
[31] X.L. Wang, H.Y. Yin, Q.L. Nie, W.W. Wu, Y. Zhang, L.Y. Qiu, Hierarchical Ag/AgCl-
TiO2 hollow spheres with enhanced visible-light photocatalytic activity, Mater
Chem Phys 185 (2017) 143–151, doi:10.1016/j.matchemphys.2016.10.017.
[32] Y. Chen, Q. Wu, C. Zhou, Q.T. Jin, Enhanced photocatalytic activity of La and
N co-doped TiO2 /diatomite composite, Powder Technol 322 (2017) 296–300,
doi:10.1016/j.powtec.2017.09.026.
[33] Z.S. Wei, L. Rosa, K.L. Wang, M. Endo, S. Juodkazis, B. Ohtani, E. Kowal-
ska, Size-controlled gold nanoparticles on octahedral anatase particles as ef-
ficient plasmonic photocatalyst, Appl Catal B–Environ 206 (2017) 393–405,
doi:10.1016/j.apcatb.2017.01.043.
[15] R. Kralchevska, M. Milanova, T. Tisler, A. Pintar, G. Tyuliev, D. Todorovsky, Pho-
tocatalytic degradation of the herbicide iodosulfuron by neodymium or ni-
trogen doped TiO2, Mater Chem Phys 133 (2012) 1116–1126, doi:10.1016/j.
matchemphys.2012.02.025.
[16] C. Castel, E. Favre, Membrane separations and energy efficiency, J Memb Sci
548 (2018) 345–357, doi:10.1016/j.memsci.2017.11.035.
[17] M. Diaz, A. Ortiz, I. Ortiz, Progress in the use of ionic liquids as electrolyte
membranes in fuel cells,
memsci.2014.06.033.
J Memb Sci 469 (2014) 379–396, doi:10.1016/j.
[18] A. Faghihi-Zarandi, H. Shirkhanloo, C. Jamshidzadeh, A new method for re-
moval of hazardous toluene vapor from air based on ionic liquid-phase adsor-
bent, Int J Environ Sci Te 6 (2018) 2797–2808, doi:10.1007/s13762-018-1975-5.
[19] J.F. Wang, J.Q. Luo, S.C. Feng, H.R. Li, Y.H. Wan, X.P. Zhang, Recent development
of ionic liquid membranes, Green Energy Environ 1 (2016) 47–65, doi:10.1016/
j.gee.2016.05.002.
[34] S.Q. Jia, J.L. Li, G.Z. Sui, L.J. Du, Y.L. Zhang, Y. Zhuang, B.X. Li, Synthesis of
3D flower-like structured Gd/TiO2@rGO nanocomposites via a hydrothermal
method with enhanced visible-light photocatalytic activity, RSC Adv 9 (2019)
31177–31185, doi:10.1039/c9ra06045f.
[35] Y.M. Wu, J.L. Zhang, L. Xiao, F. Chen, Properties of carbon and iron modified
TiO2 photocatalyst synthesized at low temperature and photodegradation of
acid orange 7 under visible light, Appl Surf Sci 256 (2010) 4260–4268, doi:10.
1016/j.apsusc.2010.02.012.
[36] G.Z. Sui, J.L. Li, A. Ito, Separation of VOC vapor from air by a surface-soaked
liquid membrane module using triethylene glycol, Sep Purif Technol 68 (2009)
283–287, doi:10.1016/j.seppur.2009.05.006.
[20] T. Uragami, Y. Matsuoka, T. Miyata, Permeation and separation charac-
teristics in removal of dilute volatile organic compounds from aqueous
solutions through copolymer membranes consisted of poly(styrene) and
poly(dimethylsiloxane) containing a hydrophobic ionic liquid by pervaporation,
J Memb Sci 506 (2016) 109–118, doi:10.1016/j.memsci.2016.01.031.
9