Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated by Temperature-Regulated Chemical Vapour Deposition as Catalysts for Acetaldehyde Oxidation

Article abstract of DOI:10.1007/s10562-017-2225-z

Abstract: Temperature-regulated chemical vapour desorption was used for deposition of Fe₂O₃ nanoparticles on the surface of mesoporous Al₂O₃ and SiO₂. The entire internal structure of mesoporous substrates, with a mean particle diameter of several hundred micrometres, was coated by Fe₂O₃ nanoparticles < 2–3?nm in lateral size. Although Fe₂O₃/Al₂O₃ had a smaller mean Fe₂O₃ particle size with superior dispersion of Fe₂O₃ nanoparticles, Fe₂O₃/SiO₂ showed a higher CO₂ evolution rate than Fe₂O₃/Al₂O₃. In combination with the results of acetaldehyde adsorption and desorption experiments, we suggest that acetaldehyde interacts more strongly with Al₂O₃ than SiO₂. This can reduce the collision frequency of acetaldehyde with catalytically active Fe₂O₃ nanoparticles deposited on Al₂O₃, thereby reducing the total oxidation rate of acetaldehyde. We demonstrate that temperature regulated-chemical vapour deposition is a promising method for preparation of mesoporous substrate-based catalysts for efficient oxidation of volatile organic compounds with different mesoporous materials. Moreover, we show that interaction between supporting material surfaces and reactant molecules is a critical factor for determining activity in heterogeneous catalysis. Graphical Abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s10562-017-2225-z

Catal Lett
DOI 10.1007/s10562-017-2225-z
Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/
SiO2 Fabricated by Temperature-Regulated Chemical Vapour
Deposition as Catalysts for Acetaldehyde Oxidation
Il Hee Kim1 · Chan Heum Park1 · Tae Gyun Woo1 · Jae Hwan Jeong1 ·
Chan Seok Jeon1 · Young Dok Kim1
Received: 19 July 2017 / Accepted: 10 October 2017
© Springer Science+Business Media, LLC 2017
Abstract Temperature-regulated chemical vapour desorp-
tion was used for deposition of Fe2O3 nanoparticles on the
surface of mesoporous Al2O3 and SiO2. The entire inter-
nal structure of mesoporous substrates, with a mean par-
ticle diameter of several hundred micrometres, was coated
by Fe2O3 nanoparticles <2–3 nm in lateral size. Although
Fe2O3/Al2O3 had a smaller mean Fe2O3 particle size with
superior dispersion of Fe2O3 nanoparticles, Fe2O3/SiO2
showed a higher CO2 evolution rate than Fe2O3/Al2O3. In
combination with the results of acetaldehyde adsorption
and desorption experiments, we suggest that acetaldehyde
interacts more strongly with Al2O3 than SiO2. This can
reduce the collision frequency of acetaldehyde with catalyti-
cally active Fe2O3 nanoparticles deposited on Al2O3, thereby
reducing the total oxidation rate of acetaldehyde. We dem-
onstrate that temperature regulated-chemical vapour deposi-
tion is a promising method for preparation of mesoporous
substrate-based catalysts for efficient oxidation of volatile
organic compounds with different mesoporous materials.
Moreover, we show that interaction between supporting
material surfaces and reactant molecules is a critical factor
for determining activity in heterogeneous catalysis.
Graphical Abstract
Il Hee Kim and Chan Heum Park have contributed equally to this
work.
Keywords Temperature regulated-chemical vapour
deposition (TR-CVD) · Heterogeneous catalysts ·
Acetaldehyde oxidation · Volatile organic compounds
(VOCs)
Electronic supplementary material The online version of this
article (doi:10.1007/s10562-017-2225-z) contains supplementary
material, which is available to authorized users.
* Young Dok Kim
ydkim91@skku.edu
1
Department of Chemistry, Sungkyunkwan University,
Suwon 16419, Republic of Korea
Vol.:(0123456789)
1 3

I. H. Kim et al.
the synthesis of mesoporous frameworks [32, 44–47]. The
post-insertion of catalytically active nanoparticles into pre-
formed mesoporous templates is advantageous, because
commercially available and cost-effective mesoporous
materials can be utilized. However, one can find several
drawbacks here; diffusion of catalytically active nanopar-
ticles is usually limited to less than several µm scale [39,
48], the nanoparticles often block the mesopores, lowering
of the surface area of the substrate [49–52], and mass pro-
duction of the catalysts in this way is challenging [53–55].
There are limited methods, such as atomic layer deposition
(ALD), which allow insertion of catalytic nanoparticles into
mesoporous frameworks without lowering the surface area
of the substrate; however, ALD uses a high vacuum and
complex control system, limiting its practical applications
for the mass-production of catalysts.
1 Introduction
The elimination of volatile organic compounds (VOCs)
from outdoor and indoor environments has been receiving
more attention recently because long-term inhalation of the
vapours of these compounds can be detrimental [1–5]. As
long as fossil fuels are used for energy, emission of these
VOCs due to incomplete combustion is unavoidable [6–9].
In indoor environments, the use of organic solvents for
interior construction can result in VOC emission, causing
sick-building syndrome. Particular attention has been paid
to acetaldehyde, which is regarded as the main cause of the
aforementioned sick-building syndrome [10–13].
Several different strategies have been considered for
removal of VOCs. The simplest and most efficient method
for VOC removal is the use of high-surface-area adsor-
bent, such as activated carbon. Recently, other more effi-
cient adsorbents such as metal–organic framework have
been developed [14–17]. However, adsorbents are regularly
regenerated, and VOCs can be re-emitted during this pro-
cess, resulting in a secondary pollution problem. Alterna-
tively, the use of catalysts for conversion of VOCs to CO2
and H2O can be considered. Photocatalysis and dark cataly-
sis with thermal energy have been widely studied [18–20].
Although the use of solar light for photocatalytic decomposi-
tion of VOCs is an attractive strategy, its low efficiency lim-
its practical applications [21, 22]. Conversely, dark catalysts
are already widely used in a variety of applications [23–25].
Among various materials, Pt-group metal nanoparticles
supported by high-surface area oxides are the most promis-
ing structures for catalytic conversion of VOCs in terms of
catalytic activity and stability [26–28]. However, the low
natural abundance and high price of Pt-group metals have
triggered the development of non-Pt-group-based catalysts.
Catalysts based on less expensive materials such as Au,
Cu, Ni, and Fe as well as bimetallic structures have been
tested for VOC oxidation [29–36]. Those non-Pt-group
metal or metal-oxide nanoparticles are catalytically active,
however, these nanoparticles generally show low stability,
and therefore easily lose their catalytic activity due to the
agglomeration of the particles upon catalytic operation at
high temperatures for long time [37–40]. This is a hurdle
of real application of these non-Pt-group catalysts, and a
strategy for increasing stability of nanoparticles of non-Pt-
group metal and metal oxide, which are catalytically active,
are needed.
In the previous studies [42, 43, 56], temperature-regulated
chemical vapour deposition (TR-CVD), which is much sim-
pler than ALD in terms of equipment structure and system
control was introduced. With this method, we could insert
catalytically active Ni-oxide and Fe-oxide nanoparticles into
the core part of mesoporous Al2O3 substrate with size of
~1 mm, and the nanoparticles were evenly distributed. And,
in the present work, we incorporated Fe-oxide nanoparticles
into mesoporous Al2O3 and SiO2, and examined further pos-
sibility of TR-CVD for applications in various substrates.
Also, the resulting structures were used for acetaldehyde
oxidation reaction. We demonstrate high efficiency of TR-
CVD for evenly depositing catalytically active nanoparticles
within the entire mesoporous network used as supporting
media of nanoparticles. We compare the structure and cata-
lytic activity of Fe2O3/Al2O3 and Fe2O3/SiO2 and highlight
that the use of different substrates can cause different activ-
ity not only due to various metal-support interactions, but
also to dissimilar interactions between various substrate sur-
faces and reactant molecules.
2 Experimental
2.1 Sample Preparations
The Fe2O3/Al2O3 and Fe2O3/SiO2 samples were prepared by
the TR-CVD method [42, 43, 56]. Commercially available
mesoporous Al2O3 (original bead size: 1 mm, mean pore
size: 11.6 nm, Sasol) and SiO2 [particle size: 250–500 µm
(35–60 mesh), mean pore size: 15 nm, Aldrich] were used
as substrates. The Al2O3 bead was mechanically fractured
and sieved so that fractured Al2O3 had similar lateral sizes
to those of mesoporous SiO2. Sieved mesoporous Al2O3
particles were used as supporting materials of Fe2O3 cata-
lysts. SiO2 was used as purchased without any additional
mechanical treatment. Bis(cyclopentadienyl) iron [Fe(Cp)2,
One promising method for increasing catalytic activ-
ity and stability of transition metal nanoparticles is use of
mesoporous materials as substrates. In this case, incorpora-
tion of catalytically active species into mesoporous struc-
tures is a challenging issue. The catalytically active species
either can be incorporated into pre-formed mesoporous
supporting materials [41–43], or can be introduced during
1 3

Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated…
Aldrich] was used as a metal precursor, and air or water
vapour originally existing in the reactor chamber was used
as an oxidizing agent. A total of 2.5 g of Fe(Cp)2 in a quartz
boat (internal size of 70 × 20 × 8 mm3) was located at the
bottom of the stainless-steel chamber, and mesoporous sub-
strates (10 g of Al2O3 or 5.3 g of SiO2) were also put inside
the chamber. The amount of substrates was adjusted so that
each substrate showed almost identical surface area. The
chamber was closed with a lid and sealed with polyimide
(PI) tape. In the first step, the temperature of the chamber
increased to 60 °C, and this temperature was maintained for
2 h; this process is required for evaporation and diffusion of
metal precursors into the mesoporous substrate. After main-
taining the temperature for 2 h, the chamber temperature was
further heated to 200 °C. This temperature was maintained
for 12 h to oxidize the Fe species adsorbed on the internal
surface of mesopores of Al2O3 or SiO2.
transfer system without exposure to the atmosphere. This
prevents possible contamination of the sample during trans-
fer from the glove box to the UHV system. XPS analysis was
carried out under a base pressure of 3×10−10 torr at UHV
conditions [39, 41–43, 56]. The photon source for XPS was
Mg Kα X-ray line (1253.6 eV), and the XPS spectra were
collected at a pass energy of 30 eV by a concentric hemi-
sphere analyser (CHA, PHOIBOS-HAS 3500, SPECS). The
binding energy of XPS spectra of each samples were cali-
brated with the respective C 1s peak centered at 284.5 eV
[57]. The intensities of the Fe 2p core-level XPS peaks of
Fe2O3/Al2O3 were normalized by the respective Al 2p peak
area, and those of the Fe 2p core-level XPS peaks of Fe2O3/
SiO2 were normalized by the respective Si 2p peak area.
2.3 Catalytic Activity Test
The catalytic activity of Fe2O3/Al2O3 and Fe2O3/SiO2 for
acetaldehyde oxidation was evaluated in a gas flow type sys-
tem comprised of a quartz tube (internal diameter of 21 mm,
length of 300 mm) equipped with a temperature control sys-
tem and mass flow controller. The concentration of acetal-
dehyde gas was controlled after addition of 200 mol ppm
acetaldehyde gas diluted with N2 (AA/N2) and O2 gas; AA/
N2 and O2 were fed into the reactor at a flow rate of 16 and
4 ml/min, respectively. As a result, 160 mol ppm of acetal-
dehyde diluted with dry air was injected into the reactor at
a flow rate of 20 ml/min. A total of 2.0 g of Fe2O3/Al2O3
or 1.2 g of Fe2O3/SiO2 were used as catalysts; the amount
of each catalyst was adjusted carefully based on the surface
area and Fe loading of each sample, i.e., different weights
of both samples were used in the reactivity experiments,
resulting in similar surface areas and Fe loadings of both
samples. The sample was placed in a quartz boat (internal
size of 70×20×8 mm3) located in the middle of the quartz
reactor. Mixed gas that had passed through the reactor was
analysed by on-line gas chromatography (GC, Hewlett Pack-
ard, HP 6890) using a capillary column (Agilent Technolo-
gies, HP-PLOT/Q, 30 m×0.53 μm), methanizer, and flame
ionization detector.
2.2 Sample Characterization
After deposition of Fe2O3 on mesoporous substrate by TR-
CVD, the elemental distribution of cross-sectional planes
of Fe2O3/Al2O3 and Fe2O3/SiO2 was analysed by scanning
electron microscopy (SEM, JEOL, JSM-7100F) equipped
with an energy dispersive spectrometer (EDS). To deter-
mine the Fe loadings in Fe2O3/Al2O3 and Fe2O3/SiO2,
inductively coupled plasma-optical emission spectroscopy
(ICP-OES) was used. After Fe2O3/Al2O3 and Fe2O3/SiO2
were annealed at 450 °C, the average pore diameter and
surface area were determined by Barret–Joyner–Halenda
(BJH) and Brunauer–Emmett–Teller (BET) methods based
on N2 isotherms (3Flex, Micromerities). The results were
compared with those of bare Al2O3 and SiO2 substrates. The
X-ray diffraction (XRD) spectra of Fe2O3/SiO2 and Fe2O3/
Al2O3 annealed at 450 °C were obtained, and compared
with that of bare SiO2 and Al2O3 [43]. Sample annealing at
450 °C was conducted under dry air flow of 30 sccm, and
XRD spectra were obtained with an X-ray diffractometer
(Rigaku, Ultima IV) using Cu Kα radiation (40 kV, 30 mA,
λ=1.54 Å) with a scanning rate of 4°/min. To obtain infor-
mation about the structure of Fe2O3 nanoparticles deposited
on the substrate, Fe2O3/Al2O3 and Fe2O3/SiO2 annealed at
450 °C were ground and analysed by high-angle annular dark
field (HAADF) scanning transmission electron microscopy
(STEM, JEOL, JEM ARM 200F) or high-resolution trans-
mission electron microscopy (HR-TEM, JEOL, JEM ARM
200F), respectively. The chemical states of the elements of
Fe2O3/Al2O3 and Fe2O3/SiO2 were analysed by X-ray photo-
electron spectroscopy (XPS). Fe2O3/Al2O3 and Fe2O3/SiO2
were annealed at 450 °C under dry air flow of 30 sccm in a
furnace located inside a glove box filled with Ar. The sam-
ples were then transferred to an ultra-high vacuum (UHV)
system by high-vacuum sealing and portable magnetic
Before the acetaldehyde oxidation experiments, Fe2O3/
Al2O3 and Fe2O3/SiO2 were pre-annealed at 450 °C for 2 h
under dry air flow of 30 ml/min. After the pre-annealing
process, the catalytic activity of catalysts as a function of
reaction temperature was evaluated. After annealing at
450 °C for 2 h, the reaction temperature was maintained at
450 °C, and 160 mol ppm gas was injected into the reac-
tor. After acetaldehyde flow was stabilized, the reactor was
cooled from 450 °C to 50 °C at a cooling rate of 1 °C/min.
After evaluating the catalytic activity as a function of reac-
tion temperature, the catalysts were again annealed at 450 °C
for 2 h to regenerate catalytic activity. The catalytic activ-
ity of each catalyst was estimated at 250 °C for 24 h. The
1 3

I. H. Kim et al.
24 h catalytic activity test of each catalyst was followed by
purging of the remaining acetaldehyde in the reactor with
30 ml/min of N2 flow for 1 h. Temperature programed oxida-
tion (TPO) was then performed to determine the amount of
carbon species remaining on catalysts during the catalytic
activity test. For TPO, the reactor temperature was increased
from 250 to 450 °C at a rate of 1 °C/min under dry air flow
of 30 ml/min.
in Fe2O3/Al2O3 and Fe2O3/SiO2 samples, respectively, and
Fe species were also distributed on the entire cutting plane
of each sample. Fe was evenly deposited not only on the
surface of the mesoporous substrate but also in the core part
of substrates with a lateral dimension of 250–500 μm. The
diffusion of Fe precursor (ferrocene) in gas phase into the
deeper part of SiO2 or Al2O3 particles via the mesoporous
network is facile, and Fe can be deposited evenly over the
entire surface of mesoporous substrates.
In addition, the catalytic activity of Fe2O3/SiO2 and
Fe2O3/Al2O3 for toluene oxidation as a function of reac-
tion time was also evaluated. The catalyst amount and pre-
annealing conditions were all identical to those of the afore-
mentioned acetaldehyde oxidation experiments. After the
pre-annealing step, reactor temperature was maintained at
450 °C, and 62 mol ppm toluene gas diluted with dry air was
injected into the reactor at a flow rate of 30 ml/min. After
the toluene flow was stabilized, the reactor temperature was
cooled from 450 to 50 °C at a cooling rate of 1 °C/min.
To determine the structures of Fe species deposited on
substrate, Fe2O3/Al2O3 was analysed by HAADF-STEM
(Fig. 1e), and Fe2O3/SiO2 was analysed by HR-TEM (Fig. 1f)
after each sample was annealed at 450 °C under dry air flow
for 2 h. Mean Fe-oxide particle size deposited on Al2O3 was
1 nm, whereas that deposited on SiO2 was ~ 2 nm. Fe2O3
particles on SiO2 were larger than those on Al2O3, possibly
due to a stronger metal-support interaction of Al2O3 as sub-
strate than SiO2. Fe2O3 nanoparticles on Al2O3 were hardly
identified by HR-TEM, most likely due to the overlap of
Al2O3 and Fe2O3 lattices. Identification of Fe2O3 crystalline
nanoparticles on amorphous SiO2 substrate is easier.
2.4 Temperature Programed Desorption (TPD)
To elucidate the effects of the surface properties of
mesoporous Al2O3 and SiO2 on catalytic activity of acet-
aldehyde oxidation by Fe-loaded catalysts, temperature
programed desorption (TPD) experiments of acetaldehyde
from bare Al2O3 and SiO2 were performed. To the reactor,
2.0 g of Al2O3 and 1.2 g of SiO2 were added and annealed
at 450 °C for 2 h under dry air flow of 30 ml/min in order
to remove surface impurities. After this pre-annealing step,
the reactor temperature was cooled to 30 °C, and 200 mol
ppm of AA/N2 gas with a flow rate of 10 ml/min for 60 h
was applied to allow acetaldehyde adsorption on the surface.
Subsequently, acetaldehyde in the reactor was purged by N2
flow. Then, reactor temperature was increased from 30 to
450 °C under N2 flow of 30 sccm at a heating rate of 1 °C/
min. The gas mixture that had passed through the reactor
during TPD experiments was analysed by an identical GC
system as was used for the aforementioned catalytic activity
experiments.
N2 isotherms of bare SiO2 and Fe2O3/SiO2 annealed at
450 °C were obtained (Fig. S1), and those of bare Al2O3
and Fe2O3/Al2O3 annealed at 450 °C were studied as well
[43]. Also, by using BET and BJH analyses of N2 isotherms,
the specific surface area and mean pore size of each sample
were determined (Table 1). For Al2O3 substrate, surface area
and pore diameter, and pore size distribution were barely
changed after Fe deposition and subsequent annealing at
450 °C. However, the surface area of SiO2, which is about
double that of Al2O3, and the porosity of SiO2 were slightly
decreased upon Fe deposition and subsequent annealing,
with a slight increase in the mean pore size. This result
might be related to partial rupture of the porous structure of
SiO2 upon annealing; as the mesoporous SiO2 network was
exposed to high temperature as 450 °C, the structure started
to partially collapse, resulting in decrease in the surface area,
porosity, and increase in the mean pore size. The specific
surface area of SiO2 is much higher than that of Al2O3 due
to the higher density of SiO2 compared to Al2O3 [58, 59].
For a given apparent volume of SiO2 and Al2O3 particles,
internal pore-volumes are estimated to be almost identical.
By ICP-OES, Fe loading on Fe2O3/SiO2 (8.32 wt%) was
more abundant than that of Fe2O3/Al2O3 (5.39 wt%). When
considering the different surface areas of various substrates,
Fe loading per unit area of Fe2O3/Al2O3 (0.35 mg/m2) and
that of Fe2O3/SiO2 (0.32 mg/m2) were almost identical.
XRD analysis of Fe2O3/SiO2 sample was conducted, and
the pattern was compared with that of bare SiO2 (Fig. S2).
Also, XRD pattern of Fe2O3/Al2O3 was compared with that
of bare Al2O3 in the previous study [43]. No iron-oxide
related peaks were found in both XRD spectra of Fe2O3/
SiO2 and Fe2O3/Al2O3. These results can be attributed to a
3 Results and Discussion
3.1 Sample Characterizations
The elemental distribution inside Fe2O3/Al2O3 and Fe2O3/
SiO2 particles was analysed by measuring SEM-EDS images
of the cut planes of mechanically fractured particles. The
green trace in Fig. 1a represents Al species in the Fe2O3/
Al2O3 sample, and the brown trace in Fig. 1c represents Si
species in the Fe2O3/SiO2 sample. These are elements of the
mesoporous substrates that were found on the entire sample
surface. Red traces in Fig. 1b, d correspond to Fe-species
1 3

Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated…
Fig. 1 Cross-sectional EDS
mapping images of a Al and
b Fe species of mechanically
cut Fe2O3/Al2O3 particles and
those of c Si and d Fe species of
Fe2O3/SiO2. e HADDF-STEM
image of Fe2O3/Al2O3 and f
HR-TEM image of Fe2O3/SiO2
annealed at 450 °C
Table 1 Surface area and average pore diameter of bare Al2O3,
XPS analyses were carried out to identify the surface
Fe2O3/Al2O3, bare SiO2, and Fe2O3/SiO2
compositions and chemical states of elements in each
catalyst. As shown in Fig. 2, each Fe 2p spectrum was
de-convoluted using six different components of a Gauss-
ian–Lorentzian function, and the normalized peak areas of
each component were listed (Fig. 2). The Fe 2p core level
spectrum of Fe(III) species shows complicated features due
to multiplet splitting and other final state contributions [60,
61]. In addition, the lowest-binding energy peak corresponds
to Fe ions with a lower oxidation state than the Fe(III) spe-
cies. The highest-binding energy peak, centred at ~714 eV,
is attributed to the surface Fe(III) species of bulk Fe2O3. The
intensity of the highest-binding energy component of Fe2O3/
SiO2 is higher than that of Fe2O3/Al2O3. This result is in line
Surface area
(m2/g)
Pore
diameter
(nm)
Bare Al2O3
450 °C-annealed Fe2O3/Al2O3
Bare SiO2
158.0
151.8
302.60
259.02
12.1
11.6
13.06
14.51
450 °C-annealed Fe2O3/SiO2
low crystallinity of Fe2O3 nanoparticles, or the small size of
Fe2O3 nanoparticles, which is in agreement with the TEM
observation.
1 3

I. H. Kim et al.
Fig. 2 Fe 2p3/2 XPS spectra of Fe2O3/Al2O3 and Fe2O3/SiO2
annealed at 450 °C. Fitting results are also displayed
with the aforementioned TEM results; for Fe2O3/Al2O3, the
smaller particle size of Fe2O3 and stronger metal-support
interaction reduce the number of surface Fe species in bulk
Fe2O3, which originate from Fe2O3 surfaces with almost no
electronic perturbation of Al2O3 to Fe2O3. Larger particles of
Fe2O3 on SiO2 are electronically considered more bulk-like
and therefore show a pronounced peak corresponding to the
surface Fe species of bulk Fe2O3.
Fig. 3 a Acetaldehyde consumption rate and b CO2 evolution rate of
450 °C-annealed Fe2O3/Al2O3 and Fe2O3/SiO2 as a function of reac-
tion temperature
temperature. In the entire temperature range of the catalyst,
Fe2O3/SiO2 generally showed a higher CO2 evolution rate
than Fe2O3/Al2O3 (Fig. 3b).
3.2 Catalytic Activity Test for Acetaldehyde Oxidation
The results in Fig. 3 suggest that, at reaction temperatures
above 300 °C, acetaldehyde is mostly removed by total oxi-
dation into CO2. Below 200 °C, partial oxidation of acetal-
dehyde followed by chemisorption of the partial oxidation
products of acetaldehyde is responsible for acetaldehyde
consumption. Total oxidation of acetaldehyde into CO2 is
more facile on Fe2O3/SiO2 than Fe2O3/Al2O3. Because Fe2O3
nanoparticles are more highly dispersed on Al2O3 than SiO2,
Fe2O3/SiO2 is not expected to show a higher activity for total
oxidation of acetaldehyde to CO2. This result is rationalized
by assuming that diffusion of acetaldehyde molecules on the
surface of SiO2 is more facile than on Al2O3. Therefore, the
collision frequency of acetaldehyde with catalytically active
Fe2O3 is higher on SiO2 than on Al2O3. Further studies using
TPD also support this idea, which is explained more in detail
Catalytic oxidation of acetaldehyde using 2.0 g of Fe2O3/
Al2O3 or 1.2 g of Fe2O3/SiO2 annealed at 450 °C was stud-
ied with decreasing reaction temperature from 450 to 50 °C
at a cooling rate of 1 °C/min. The amount of the catalysts
was determined carefully, so that each catalyst has the same
level of the surface area of the substrate, and the Fe loading
(Table S1). The starting temperature was 450 °C in these
light-off experiments because 100% conversion of acetal-
dehyde to CO2 was achieved at this temperature for both
catalysts. As shown in Fig. 3a, the acetaldehyde consump-
tion rates of both catalysts were similar. Almost all acetal-
dehyde in the feed gas was removed over the entire reaction
temperature range. On the other hand, the CO2 evolution
rate of both catalysts decreased with decreasing reaction
1 3

Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated…
in the forthcoming section. The results of light-off curves of
Fe2O3/Al2O3 or Fe2O3/SiO2 were compared with those from
the previous studies in Table S2 [62–67]. Note that most of
transition metal oxide catalysts showed ~90% of conversion
to CO2 at ~300 °C, which is comparable catalytic activity
with those of Fe2O3/Al2O3 or Fe2O3/SiO2 catalysts in the
present work. It should be emphasized that our catalysts are
comparable with other previously reported catalysts in terms
of initial reactivity, yet long-term stability of the catalytic
activity we tested in this work (which will be show next) has
rarely been reported in the literature. Moreover, Fe is known
to be one of the most cost-effective transition metals.
evolution at the initial stage of both reactivity experiments,
and CO2 evolution rapidly increased with increasing reaction
time (Fig. 4b). Analogous to the results of Fig. 3, the CO2
evolution rate of Fe2O3/SiO2 was generally higher than that
of Fe2O3/Al2O3. The amount of each sample was chosen to
have the almost identical amount of Fe loading and surface
area of both samples. According to TEM analysis, the lat-
eral size of Fe2O3 nanoparticles on Fe2O3/SiO2 is ~2 times
larger than that on Fe2O3/Al2O3, therefore, the total number
of Fe2O3 nanoparticles on Fe2O3/Al2O3 is ~8 times larger
than that on Fe2O3/SiO2 to have almost identical total vol-
ume of Fe2O3 nanoparticles, while surface area of an indi-
vidual Fe2O3 nanoparticle on Fe2O3/SiO2 is ~4 times larger
than that on Fe2O3/Al2O3. As a result, the total surface area
of Fe2O3 nanoparticles on Fe2O3/SiO2 samples should be ~2
times smaller than that on Fe2O3/Al2O3. Assuming that the
CO2 evolution rate is dependent on the number of active site
of iron oxide, CO2 evolution rate of Fe2O3/SiO2 should be
smaller than that of Fe2O3/Al2O3. However, CO2 evolution
rate of Fe2O3/SiO2 was higher than that of Fe2O3/Al2O3,
while acetaldehyde consumption rate was at the same level
for 24 h. Therefore, it can be concluded that CO2 evolution
rate was mostly determined by the diffusion rate of acetal-
dehyde or its partial oxidation intermediates on substrate
surface, not the number of catalytically active site, or the
state of iron oxide particle. This point will be more clarified
in the forthcoming discussion about TPD results. It is also
worth mentioning that both substrates showed quite similar
porosity, and therefore porosity does not seem to be a critical
factor for determining the CO2 evolution rate.
Catalytic oxidation of acetaldehyde at 250 °C using these
two catalysts was studied for 24 h to test the long-term sta-
bility of the catalytic activity (Fig. 4). There was almost no
difference between the two catalysts regarding acetaldehyde
consumption rate, and both catalysts showed nearly 100%
consumption of acetaldehyde for 24 h without a decrease
in catalytic activity (Fig. 4a). There was almost no CO2
After each experiment shown in Fig. 4, a TPO experi-
ment was subsequently carried out to determine the residual
amount of carbon species remaining on the surfaces of both
catalysts after catalytic reaction for 24 h (Fig. 5). It is notable
Fig. 4 a Acetaldehyde consumption rate and b CO2 evolution rate of
Fig. 5 CO2 evolution rate during TPO of Fe2O3/Al2O3 and Fe2O3/
Fe2O3/Al2O3 and Fe2O3/SiO2 at 250 °C as a function of reaction time
SiO2 after acetaldehyde oxidation for 24 h
1 3

I. H. Kim et al.
that most of carbon species accumulated on the surface
were removed during TPO. With increasing temperature,
CO2 desorption rate first increased and showed a maximum
value at 330 °C. Above this temperature, the CO2 desorp-
tion rate decreased, and no CO2 desorption was detected at
temperatures exceeding 450 °C (Fig. 5). The CO2 evolu-
tion rate of Fe2O3/Al2O3 was generally higher than that of
Fe2O3/SiO2, and the amount of carbon species remaining
on Fe2O3/Al2O3 after the acetaldehyde oxidation experi-
ment was much higher than that remaining on Fe2O3/SiO2.
During acetaldehyde oxidation, acetaldehyde adsorbed on
the Fe2O3/SiO2 surface was more easily removed into CO2,
leaving only a small amount of carbon-containing species
(reaction products of partial oxidation of acetaldehyde). Sta-
bilization of the partial oxidation product of acetaldehyde is
more pronounced on the Fe2O3/Al2O3 surface, resulting in a
larger amount of residual carbon-containing species on the
surface of catalysts during acetaldehyde reaction at 250 °C
for 24 h.
through samples and detected by GC increases [17, 68, 69].
When SiO2 was used as adsorbent, the acetaldehyde evolu-
tion rate slowly increased with time. In contrast, almost no
acetaldehyde was observed in the reactor outlet when Al2O3
was placed in the reactor. This indicates that all acetaldehyde
molecules in the feed gas were adsorbed onto the surface
of Al2O3, and Al2O3 shows a much higher affinity towards
acetaldehyde adsorption than SiO2 (Fig. 6).
Subsequent to the breakthrough experiments shown
in Fig. 6, TPD was performed as the reactor temperature
increased from 30 to 450 °C. High purity N2 gas passed
through the substrate during TPD experiments. During
TPD, desorption of various chemical species was identi-
fied, as shown in Fig. 7. From Al2O3, almost no acetal-
dehyde desorption was observed, whereas two acetalde-
hyde desorption states were identified from SiO2 at ~ 120
and 350 °C, which are tentatively attributed to multilayer
adsorption of acetaldehyde in SiO2 pores (physisorp-
tion) and a monolayer of acetaldehyde (chemisorption),
3.3 Temperature Programed Desorption (TPD)
To investigate the interaction between acetaldehyde and
mesoporous substrates (Al2O3 and SiO2), TPD of acetal-
dehyde from bare SiO2 and Al2O3 surfaces without Fe2O3
was performed. Before the TPD experiment, each sample
was exposed to acetaldehyde flow (10 ml/min) with a con-
centration of 200 ppm (balanced N2) for 60 h. During these
experiments, the amount of acetaldehyde passed through the
sample stage in the reactor was detected using GC (Fig. 6).
Acetaldehyde contained in the flow gas can adsorb onto the
surface of each sample. As the coverage of acetaldehyde on
the surface increases, the amount of acetaldehyde passing
Fig. 6 Changes in acetaldehyde flow rates with time during adsorp-
tion step of acetaldehyde prior to TPD were measured from gas
passed through the sample stage using Al2O3 or SiO2 substrate for
60 h
Fig. 7 TPD results of 450 °C-annealed SiO2 and 450 °C-annealed
Al2O3. a Acetaldehyde evolution rate and b CO2 evolution rate as a
function of temperature for both adsorbents
1 3

Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated…
respectively. Acetaldehyde adsorbed on the surface of
SiO2 was converted into CO2 only when the tempera-
ture exceeded 350 °C, whereas acetaldehyde molecules
adsorbed on Al2O3 were converted into CO2 at 200 °C,
much lower than the onset temperature of CO2 evolution
from SiO2 (Fig. 7b). When the oxide surface is reducible,
organic molecules can be oxidized by reaction between the
molecules and the lattice oxygen of the oxide [70–72]. In
case of SiO2, the acetaldehyde can be oxidized by the lat-
tice oxygen only when the reaction temperature was higher
than 300 °C, and therefore the contribution of the substrate
for CO2 evolution can be neglected in the results of Fig. 4.
In case of Al2O3, however, CO2 is detected from 200 °C in
the TPD, meaning that lattice oxygen of Al2O3 can totally
oxidize acetaldehyde during the reactivity experiments in
Fig. 4. However, the amount of CO2 detected from 200 to
250 °C in the TPD is very is small compared to that above
250 °C and therefore, the contribution of the lattice oxygen
of Al2O3 substrate in the acetaldehyde oxidation in Fig. 4
can be regarded to be a minor factor, even though this
process cannot be fully excluded.
3.4 Toluene Oxidation Using Fe2O3/Al2O3 or Fe2O3/
SiO2
In order to test the potential of the Fe2O3/Al2O3 of Fe2O3/
SiO2 as the catalysts for oxidation of VOCs other than
acetaldehyde, such as aromatic compounds represented by
toluene, catalytic activity for toluene oxidation using those
catalysts annealed at 450 °C was evaluated as a function of
reaction temperature from 450 to 50 °C at a cooling rate of
1 °C/min (Fig. S3). Toluene gas (62 mol ppm) was injected
into the reactor at a flow rate of 30 ml/min. A reaction tem-
perature of 450 °C was chosen as the starting temperature
for light-off experiments because 100% oxidation of toluene
to CO2 was achieved at this temperature for both catalysts.
The CO2 evolution rate of both catalysts decreased as reac-
tion temperature decreased (Fig. S3b). Fe2O3/SiO2 showed a
higher CO2 evolution rate than Fe2O3/Al2O3 within the entire
temperature range, and these results are analogous to those
of acetaldehyde. It can be said that the toluene-SiO2 interac-
tion is weaker than that of toluene-Al2O3, as acetaldehyde-
SiO2 interaction is weaker than that of acelaldehyde-Al2O3.
Collision between adsorbed toluene and Fe2O3 nanoparticles
is more facile on SiO2 than on Al2O3, a reaction that is criti-
cal for total oxidation to CO2. Most toluene molecules were
removed by total oxidation into CO2 at 450 °C (Fig. S3a);
however, with decreased reaction temperature to 250 °C,
toluene removal rate remained at the maximum rate even
though the CO2 evolution rate drastically decreased. Tolu-
ene could not only be removed by total oxidation, but also
by partial oxidation and chemisorption of the partial oxida-
tion product on the surface of catalysts. As reactor tempera-
ture became lower than 200 °C, the toluene removal rate
first decreased and then increased. The increase in toluene
removal rate as reaction temperature decreased from 150 to
50 °C is attributed to molecular adsorption of toluene on the
surface. Overall, Fe2O3/SiO2 is more efficient for total oxida-
tion of toluene, whereas Fe2O3/Al2O3 is superior to Fe2O3/
SiO2 for molecular chemisorption and activation (or partial
oxidation) of toluene, in line with the results of acetaldehyde
oxidation.
GC identified other species resulting from chemical
reactions between acetaldehyde molecules on the surface
and the substrate, and the formation of these chemical spe-
cies was much more pronounced from Al2O3 than SiO2.
This result indicates that acetaldehyde was adsorbed more
strongly and more activated on the surface of Al2O3 than
SiO2. The possible candidates for each species were specu-
lated by its retention time of GC, and listed in Table S3
[73, 74]. According to the observation, polymerization of
acetaldehyde was occurred, and C3, C4, and C6 species
were produced either on SiO2 or Al2O3. It is notable that
the decomposition of acetaldehyde into CO and CH4 was
only occurred on Al2O3.
Summarizing the results of Figs. 6 and 7, Al2O3 shows
higher acetaldehyde uptake than SiO2 at 30 °C. During
TPD, acetaldehyde molecular desorption from SiO2 pre-
vailed, whereas conversion of acetaldehyde molecules
into CO2 and other chemical species was more dominant
from Al2O3. These results all imply that Al2O3 shows
greater interaction with acetaldehyde than SiO2. A higher
Al2O3-acetaldehyde interaction hinders diffusion of acetal-
dehyde molecules on the surface. This reduces the colli-
sion frequency of Fe2O3 and adsorbed acetaldehyde species.
Greater Al2O3-acetaldehyde interaction is suggested to be
why partial oxidation of acetaldehyde is more dominant
on the surface of Fe2O3/Al2O3 than total oxidation to CO2.
Although SiO2 shows a weaker metal-support interaction
and lower dispersion of catalytically active Fe2O3 species, a
weaker interaction between acetaldehyde and SiO2 induces
increased collision frequency between Fe2O3 and acetalde-
hyde adsorbed on the surface. This increases activity towards
the total oxidation rate of acetaldehyde to CO2.
4 Conclusion
Fe2O3 nanoparticles were deposited using the TR-CVD
method and subsequent annealing. This allows deposition
of catalytically active Fe2O3 nanoparticles both on the out-
ermost surface layers and also the core part of mesoporous
substrates. On Al2O3, the mean particle size of Fe2O3 was
only ~1 nm, whereas that on SiO2 was ~2 nm. Fe2O3 was
more highly dispersed on the surface of Al2O3 due to the
higher metal-support interaction of Al2O3 than SiO2. Both
Fe2O3/Al2O3 and Fe2O3/SiO2 showed high catalytic activity
1 3

I. H. Kim et al.
17. Kim K-D, Park EJ, Seo HO, Jeong M-G, Kim YD, Lim DC
(2012) Chem Eng J 200–202:133
for acetaldehyde removal, with almost 100% acetaldehyde
removal rate at 250 °C for 24 h. Regarding CO2 evolution
rate, however, Fe2O3/SiO2 was superior to Fe2O3/Al2O3.
Together with subsequent TPO results, this result implies
that partial oxidation of acetaldehyde and adsorption of reac-
tion intermediates for total oxidation significantly contrib-
ute to acetaldehyde removal by Fe2O3/Al2O3. Conversely,
total oxidation of acetaldehyde to CO2 is more dominant
for Fe2O3/SiO2. In combination with acetaldehyde TPD
results, we suggest that interaction of SiO2 and acetalde-
hyde is weaker than interaction of Al2O3 and acetaldehyde.
Weaker interaction between substrate surface and reactant
facilitates diffusion of reactant molecules to the catalytically
active species, increasing total oxidation rate. For toluene
oxidation experiments using these two different catalysts,
total oxidation of toluene into CO2 was more efficient when
SiO2 was used as a supporting material of Fe2O3 catalyst
than when using Al2O3. This result is in line with that of
acetaldehyde oxidation. Our results show that changes in
substrate structure not only influence metal-support interac-
tion, but also support-reactant interaction, a critical factor
for determining catalytic activity and selectivity.
18. Qian X, Ren M, Yue D, Zhu Y, Han Y, Bian Z, Zhao Y (2017)
Appl Catal B 212:1
19. Selishchev DS, Kolobov NS, Pershin AA, Kozlov DV (2017)
Appl Catal B 200:503
20. Shiraishi F, Maruoka D, Tanoue Y (2017) Sep Purif Technol
175:185
21. Bae S, Kim S, Lee S, Choi W (2014) Catal Today 224:21
22. Dong H, Zeng G, Tang L, Fan C, Zhang C, He X, He Y (2015)
Water Res 79:128
23. Liu W-J, Qian T-T, Jiang H (2014) Chem Eng J 236:448
24. Rahim Pouran S, Abdul Raman AA, Wan Daud WMA (2014) J
Clean Prod 64:24
25. Yashnik SA, Denisov SP, Danchenko NM, Ismagilov ZR (2016)
Appl Catal B 185:322
26. Ordóñez S, Bello L, Sastre H, Rosal R, Díez FV (2002) Appl
Catal B 38:139
27. Liotta LF (2010) Appl Catal B 100:403
28. Chen C, Chen F, Zhang L, Pan S, Bian C, Zheng X, Meng X,
Xiao F-S (2015) Chem Commun 51:5936
29. Liotta LF, Ousmane M, Di Carlo G, Pantaleo G, Deganello G,
Boreave A, Giroir-Fendler A (2009) Catal Lett 127:270
30. Knudsen J, Merte LR, Peng G, Vang RT, Resta A, Lægsgaard E,
Andersen JN, Mavrikakis M, Besenbacher F (2010) ACS Nano
4:4380
31. Saqer SM, Kondarides DI, Verykios XE (2011) Appl Catal B
103:275
32. Yen H, Seo Y, Kaliaguine S, Kleitz F (2012) Angew Chem Int
Ed 51:12032
Acknowledgements This research was supported by the Basic Sci-
ence Research Program through the National Research Foundation of
Korea (NRF), funded by the Ministry of Science, ICT, and Future
Planning (2015R1A2A2A01003866). This work was supported by the
research grant of the Korea Basic Science Institute (D37613).
33. Liu Y, Deng J, Xie S, Wang Z, Dai H (2016) Chin J Catal
37:1193
34. Piumetti M, Bensaid S, Fino D, Russo N (2016) Appl Catal B
197:35
35. Si W, Wang Y, Zhao S, Hu F, Li J (2016) Environ Sci Technol
50:4572
36. Leung E, Shimizu A, Barmak K, Farrauto R (2017) Chem Com-
mun 97:42
References
37. Tang H, Wei J, Liu F, Qiao B, Pan X, Li L, Liu J, Wang J, Zhang
T (2016) J Am Chem Soc 138:56
1. Anton A, Ben de Lacy C, Wolfram M, Jochen S, Bogusław B,
Joachim P, Norman R, Terence R (2014) J Breath Res 8:034001
2. Buljubasic F, Buchbauer G (2015) Flavour Fragr J 30:5
3. Huang Y, Ho SS, Lu Y, Niu R, Xu L, Cao J, Lee S (2016) Mol-
ecules 21:56
38. Yang F, Chen MS, Goodman DW (2009) J Phys Chem C 113:254
39. Han SW, Kim DH, Jeong M-G, Park KJ, Kim YD (2016) Chem
Eng J 283:992
40. Jeong MG, Kim IH, Han SW, Kim DH, Kim YD (2016) J Mol
Catal A 414:87
4. Kamal MS, Razzak SA, Hossain MM (2016) Atmos Environ
140:117
41. Kim DH, Kim SY, Han SW, Cho YK, Jeong M-G, Park EJ, Kim
YD (2015) Appl Catal A 495:184
5. Paciência I, Madureira J, Rufo J, Moreira A, Fernandes EdO
(2016) J Toxicol Environ Health 19:47
42. Kim IH, Park EJ, Park CH, Han SW, Seo HO, Kim YD (2017)
Catal Today 295:56
6. Garcia-Hurtado E, Pey J, Borrás E, Sánchez P, Vera T, Carratalá
A, Alastuey A, Querol X, Vallejo VR (2014) Atmos Environ 89:85
7. Mellouki A, Wallington TJ, Chen J (2015) Chem Rev 115:3984
8. Ogunkoya D, Roberts WL, Fang T, Thapaliya N (2015) Fuel
140:541
43. Kim IH, Seo HO, Park EJ, Han SW, Kim YD (2017) Sci Rep
7:40497
44. Solsona B, Garcia T, Sanchis R, Soriano MD, Moreno M, Rodri-
guez-Castellon E, Agouram S, Dejoz A, Nieto JML (2016) Chem
Eng J 290:273
9. Tutak W, Lukács K, Szwaja S, Bereczky Á (2015) Fuel 154:196
10. LoPachin RM, Gavin T (2014) Chem Res Toxicol 27:1081
11. Nikawa T, Naya S-i, Kimura T, Tada H (2015) J Catal 326:9
12. Sun Z, Kong L, Ding X, Du C, Zhao X, Chen J, Fu H, Yang X,
Cheng T (2016) Phys Chem Chem Phys 18:9367
45. Zhu JK, Gao QM (2009) Microporous Mesoporous Mater 124:144
46. Wan HJ, Wu BS, Zhang CH, Xiang HW, Li YW, Xu BF, Yi F
(2007) Chem Commun 8:1538
47. Sun ZK, Sun B, Qiao MH, Wei J, Yue Q, Wang C, Deng YH,
Kaliaguine S, Zhao DY (2012) J Am Chem Soc 134:17653
48. King JS, Wittstock A, Biener J, Kucheyev SO, Wang YM, Bau-
mann TF, Giri SK, Hamza AV, Baeumer M, Bent SF (2008) Nano
Lett 8:2405
13. Tang X, Misztal PK, Nazaroff WW, Goldstein AH (2016) Environ
Sci Technol 50:12686
14. Qu F, Zhu L, Yang K (2009) J Hazard Mater 170:7
15. Pires J, Carvalho A, de Carvalho MB (2001) Microporous
Mesoporous Mater 43:277
49. Bang Y, Park S, Han SJ, Yoo J, Song JH, Choi JH, Kang KH, Song
IK (2016) Appl Catal B 180:179
16. Lillo-Ródenas MA, Fletcher AJ, Thomas KM, Cazorla-Amorós
D, Linares-Solano A (2006) Carbon 44:1455
50. Jiang T, Du S, Jafari T, Zhong W, Sun Y, Song W, Luo Z, Hines
WA, Suib SL (2015) Appl Catal A 502:105
1 3

Comparative Studies of Mesoporous Fe2O3/Al2O3 and Fe2O3/SiO2 Fabricated…
51. Alegre C, Galvez ME, Moliner R, Baglio V, Arico AS, Lazaro MJ
(2014) Appl Catal B 147:947
63. McCabe RW, Mitchell PJ (1983) Ind Eng Chem Res 22:212
64. McCabe RW, Mitchell PJ (1984) Ind Eng Chem Res 23:196
65. Mitsui T, Tsutsui K, Matsui T, Kikuchi R, Eguchi K (2008) Appl
Catal B 81:56
52. Coelho JV, Guedes MS, Prado RG, Tronto J, Ardisson JD, Pereira
MC, Oliyeira LCA (2014) Appl Catal B 144:792
53. Rioux RM, Song H, Hoefelmeyer JD, Yang P, Somorjai GA
(2005) J Phys Chem B 109:2192
66. Wang L, Tran TP, Vo DV, Sakurai M, Kameyama H (2008) Appl
Catal A 350:150
54. Tang Q, Hu S, Chen Y, Guo Z, Hu Y, Chen Y, Yang Y (2010)
Microporous Mesoporous Mater 132:501
67. Watanabe N, Yamashita H, Miyadera H, Tominaga S (1996) Appl
Catal B 8:405
55. Shylesh S, Singh AP (2004) J Catal 228:333
56. Han SW, Kim IH, Kim DH, Park KJ, Park EJ, Jeong M-G, Kim
YD (2016) Appl Surf Sci 385:597
68. Seo HO, Kim DH, Kim K-D, Park EJ, Sim CW, Kim YD (2013)
Adsorption 19:1181
69. Park EJ, Kim HJ, Han SW, Jeong JH, Kim IH, Seo HO, Kim YD
(2017) Chem Eng J 325:433
57. Moulder JF, Stickle WF, Sobol PE, Bomben KD (1995) Handbook
of X-ray photoelectron spectroscopy. Physical Electronics Inc.,
Chanhassen
70. Ordonez S, Bello L, Sastre H, Rosal R, Diez FV (2002) Appl Catal
B 38:139
58. Upadhyaya A, Upadhyaya GS (1995) Mater Des 16:41
59. Aly M, Hashmi MSJ, Olabi AG, Messeiry M, Abadir EF, Hussain
AI (2012) Mater Des 33:127
71. Konietzni F, Zanthoff H, Maier W (1999) J Catal 188:154
72. Garetto T, Apesteguıa C (2000) Catal Today 62:189
73. Fleischmann G, Jira R, Bolt HM, Golka K (2006) Ullmann’s
Encyclopedia of industrial chemistry. Wiley, Weinheim
74. Sivaramakrishnan R, Michael JV, Klippenstein SJ (2010) J Phys
Chem A 114:755
60. Mycroft JR, Nesbitt HW, Pratt AR (1995) Geochim Cosmochim
Acta 59:721
61. Grosvenor AP, Kobe BA, Biesinger MC, McIntyre NS (2004) Surf
Interface Anal 36:1564
62. Gallegos MV, Peluso MA, Finocchio E, Thomas HJ, Busca G,
Sambeth JE (2017) Chem Eng J 313:1099
1 3