Iron-Functionalized Silica Nanoparticles as a Highly Efficient Adsorbent and Catalyst for Toluene Oxidation in the Gas Phase

Article abstract of DOI:10.1002/cctc.201200562

Catalytic oxidation is one of the most important industrially applicable processes for the decomposition of volatile organic compounds (VOCs) in polluted air. The advanced VOC removal process is composed of an adsorption unit and a catalytic incinerator. Many efforts have been made to design a combined adsorption-catalytic unit with optimal activity and selectivity. We demonstrate that iron-functionalized silica nanoparticles with interparticle mesoporosity (FeKIL-2) act as highly efficient adsorbents and catalysts with optimal Fe/Si molar ratios of 0.01 in toluene oxidation as model VOCs in the gas phase. By using UV/Vis, FTIR, and Moessbauer spectroscopic techniques, we prove that the enhanced activity of the catalyst is attributed to iron incorporated into the silica matrix, which depends on the iron content. The iron content with Fe/Si≤0.01 leads to the formation of stable Fe³⁺ ions in the silica matrix, which ensures easier oxygen release from the catalyst (Fe³⁺/Fe²⁺ redox cycles). The increase in the iron content with Fe/Si>0.01 leads to the formation of oligonuclear iron complexes. The material thus introduces a promising, environmentally friendly, cost-effective, and highly efficient catalyst with combined adsorption and catalytic properties for the removal of low-concentration VOC from polluted air.

Full text of DOI:10.1002/cctc.201200562

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201200562
Iron-Functionalized Silica Nanoparticles as a Highly
Efficient Adsorbent and Catalyst for Toluene Oxidation in
the Gas Phase
[
b]
[a]
[c]
[a]
[b]
Margarita Popova, Alenka Risti c´ , Karoly Lazar, Darja Mau cˇ ec, Mihaela Vassileva, and
[b, d]
Nata sˇ a Novak Tu sˇ ar*
Catalytic oxidation is one of the most important industrially ap-
plicable processes for the decomposition of volatile organic
compounds (VOCs) in polluted air. The advanced VOC removal
process is composed of an adsorption unit and a catalytic in-
cinerator. Many efforts have been made to design a combined
adsorption–catalytic unit with optimal activity and selectivity.
We demonstrate that iron-functionalized silica nanoparticles
with interparticle mesoporosity (FeKIL-2) act as highly efficient
adsorbents and catalysts with optimal Fe/Si molar ratios of
we prove that the enhanced activity of the catalyst is attribut-
ed to iron incorporated into the silica matrix, which depends
on the iron content. The iron content with Fe/Siꢀ0.01 leads to
3
+
the formation of stable Fe ions in the silica matrix, which en-
3
+
2+
sures easier oxygen release from the catalyst (Fe /Fe redox
cycles). The increase in the iron content with Fe/Si>0.01 leads
to the formation of oligonuclear iron complexes. The material
thus introduces a promising, environmentally friendly, cost-ef-
fective, and highly efficient catalyst with combined adsorption
and catalytic properties for the removal of low-concentration
VOC from polluted air.
0.01 in toluene oxidation as model VOCs in the gas phase. By
using UV/Vis, FTIR, and Mçssbauer spectroscopic techniques,
Introduction
Volatile organic compounds (VOCs) are the main class of air
ties were described by using the Mars–van Krevelen mecha-
[
1–12]
[9]
pollutants emitted by various industrial processes.
The list
nism. The nature of the support is also an important factor
of pollutants includes more than 300 compounds, such as oxy-
genates, aromatic hydrocarbons, and halogen hydrocarbons.
The catalytic oxidation has been considered as the most ap-
propriate method for VOC removal from main industrial emis-
sions, such as the petroleum industry. In addition, many local
sources such as printing and painting processes emit low-con-
centration VOCs. In this case, the adsorption process is the
most promising control technology. The advanced VOC remov-
al process is composed of an adsorption unit and a catalytic in-
because the surface area and functionality determine the
nature and dispersion of metal oxide particles and therefore
[1,8]
their catalytic behavior.
Over the last decades, mesoporous
silica supports have been of interest as catalyst supports be-
cause of their uniform mesoporous channel structure and high
[15–26]
specific surface area.
Various metal ions (Fe, Ti, V, Cr, Cu,
etc.) have been introduced into the silica matrix to obtain
modified mesoporous materials with tunable catalytic proper-
[21–23]
ties.
The applied method of modification strongly influen-
[
13,14]
cinerator.
Many efforts have been made to design catalysts
ces the nature (location, dispersion, and oxidation state) of the
[
1–12]
[24–28]
with good activity and selectivity.
Transition-metal oxides
loaded metal species.
Their introduction in the host matrix
are an alternative to the currently used noble metal-containing
catalysts because of their resistance to halogens, low cost, and
could be realized during the silica synthesis process as well as
by various postsynthesis techniques (e.g., impregnation and
[
1–4,8]
[24–28]
high catalytic activity and selectivity.
Their catalytic proper-
grafting).
Adsorption and catalytic oxidation of aromatic
VOCs are of importance because they are emitted from diverse
sources, for example, from the petrochemical industry, printing,
pressing, automobile exhaust, and traffic-related processes.
The total oxidation of toluene as a model aromatic VOC has
been studied with different metals (Pt, Cu, Fe, V, Co, Cr) on dif-
ferent types of supports (Al O , CeO2, TiO2 activated
[
a] Dr. A. Risti c´ , D. Mau cˇ ec
National Institute of Chemistry
1
000 Ljubljana (Slovenia)
[
b] Dr. M. Popova, M. Vassileva, Prof. Dr. N. Novak Tu sˇ ar
Institute of Organic Chemistry with Centre of Phytochemistry
Bulgarian Academy of Sciences
2
3
[2,3,7,8,20,29–31]
carbon).
1113 Sofia (Bulgaria)
We show that iron-functionalized silica nanoparticles with in-
terparticle porosity (disordered mesoporous KIL-2 material
functionalized by iron) act as highly efficient adsorbents and
catalysts for toluene elimination. Spectroscopic techniques
demonstrate that the superior activity of the catalyst can be at-
[
c] Dr. K. Lazar
Centre for Energy Research
Institute of Isotopes
Hungarian Academy of Sciences
1
525 Budapest (Hungary)
[
d] Prof. Dr. N. Novak Tu sˇ ar
3
+
tributed uniquely to the formation of stable Fe ions in the
University of NovaGorica
5
000 Nova Gorica (Slovenia)
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 986

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
silica matrix, which ensures easier oxygen release from the cat-
3
+
2+
alyst (Fe /Fe redox cycles).
Results and Discussion
Physicochemical characterization of the samples
We have prepared the heterogeneous catalysts through the in-
corporation of iron into the silicate nanoparticles of the disor-
dered mesoporous KIL-2 structure with interparticle porosity,
[31–33]
which have recently been developed in our laboratory.
Heterogeneous iron catalysts with various iron concentrations
Fe/Si ratios ranging from 0.005 to 0.05) were prepared
(
through direct two-step solvothermal synthesis (see details in
Experimental Section). The XRD patterns of FeKIL-2 samples at
low angles indicate their disordered mesoporous nature (not
shown). At high angles, they do not show detectable diffrac-
tion peaks that could be assigned to iron oxide phases
Figure 2. SEM of a) 005FeKIL-2 (Fe/Si=0.005), b) 01FeKIL-2 (Fe/Si=0.01),
c) 02FeKIL-2 (Fe/Si=0.02), and d) 05FeKIL-2 (Fe/Si=0.05). Scale
bars=200 nm.
(Figure 1); thus, either such oxides are not present within
Figure 1. XRD patterns of 005FeKIL-2 (Fe/Si=0.005), 01FeKIL-2 (Fe/Si=0.01),
Figure 3. Nitrogen sorption isotherms with pore size distribution curves of
FeKIL-2 samples. PSD=pore size distribution.
0
2
2FeKIL-2 (Fe/Si=0.02), and 05FeKIL-2 (Fe/Si=0.05): a) at low angles in the
q range from 0.5 to 58 (not shown) and b) at high angles in the 2q range
from 5 to 608.
and 01FeKIL-2 samples show adsorption isotherms typical for
[
31–33]
FeKIL-2 samples or their crystallites are very small with typical
dimensions of 5 nm or less and thus are not detectable by
using XRD. High-angle XRD patterns (Figure 1) show only one
KIL-2 silica,
that is, with a relatively narrow type IV hystere-
[35]
sis loop. The 005FeKIL-2 sample shows the adsorption iso-
therm with a capillary condensation step at relative pressures
of approximately 0.88, which indicates the presence of large
o
diffraction peak at 2q=44.39 corresponding to the sample
holder and one broad peak at
o
2
qꢁ23 corresponding to glass-
like amorphous silicate nanopar-
[a]
Table 1. Physicochemical properties of the studied samples.
[
34]
ticles.
SEM images of the
Sample
Fe/Si
Fe
S
BET
2
w
[nm]
BJH
V
me
V
t
V
mi
S
me
2
S
ex
[m g ]
FeKIL-2 catalysts are shown in
Figure 2.
ꢂ1
3
ꢂ1
3
ꢂ1
3
ꢂ1
ꢂ1
2
ꢂ1
[
wt%]
[m g
]
[cm g
]
[cm g
]
[cm g
]
[m g
]
KIL-2
–
–
545
472
556
213
366
19.9
27.0
21.6
–
1.392
1.149
1.326
0.546
0.368
1.480
1.292
1.459
0.976
0.640
0.054
0.080
0.077
–
370
246
332
143
136
38
37
39
73
70
Nitrogen adsorption isotherms
for FeKIL-2 materials are shown
in Figure 3, whereas structural
parameters determined on the
basis of these isotherms are
listed in Table 1. The difference
of isotherms types can be seen
clearly in Figure 3. 005FeKIL-2
0
0
05FeKIL-2
1FeKIL-2
0.005
0.01
0.02
0.05
0.47
0.92
2.91
3.84
02FeKIL-2
05FeKIL-2
–
0.069
[a] SBET =BET surface area; Sex =external surface area, Sme =mesoporous surface area, Vme =primary mesopore
volume, and Vmi =micropore volume, each evaluated by using the a plot method; V =total pore volume eval-
uated from adsorption isotherm at the relative pressure of 0.98; and wBJH =mesopore diameters at the maxi-
mum of the BJH pore size distribution.
s
t
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 987

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
mesopores. The sample 01FeKIL-2 undergoes a capillary con-
densation step at a lower relative pressure of approximately
05FeKIL-2 sample could be explained by the higher amount of
iron slowing the toluene adsorption.
0
.82, which evidences smaller mesopores in this sample com-
All catalysts were tested in total toluene oxidation, and the
results are shown in Figure 5. CO2 is the only registered
pared to 005FeKIL-2. Thus, the increase in the iron amount re-
sults in an increase in the pore volume (from 1.292 to
carbon-containing product in all cases and H O. Of all the
2
3
ꢂ1
[36]
1.459 cm g ) and a decrease in the pore diameter. Notably,
higher amounts of the incorporated iron lead to a marked
change in the shape of the nitrogen adsorption isotherms; for
example, the nitrogen isotherms of 02FeKIL-2 and 05FeKIL-2
samples change to a type II isotherm, which also confirms the
[
37]
presence of macroporosity. A capillary condensation step of
2FeKIL-2 and 05FeKIL-2 samples occurs at a very high relative
0
pressure (p/p ꢃ0.9), which is an evidence of an interparticular
0
[
38]
porosity.
BET specific surface areas of 005FeKIL-2 and 01FeKIL-2 mate-
rials increase with a higher amount of iron, whereas the specif-
ic surface areas of 02FeKIL-2 and 05FeKIL-2 materials decrease
significantly. The pore size distribution of different mesoporous
materials has been determined by using the BJH model widely
[
39,40]
used for this type of samples.
Although this model often
underestimates pore sizes, it is appropriate for comparative
Figure 5. a) Catalytic activity versus temperature and b) catalytic activity
versus time on stream at 3808C on the iron-containing silicate nanoparticles
with Fe/Si molar ratios 0.005 (denoted as 005FeKIL-2), 0.01 (denoted as
01FeKIL-2), 0.02 (denoted as 02FeKIL-2), and 0.05 (denoted as 05FeKIL-2).
[
41]
purposes. The pore size distribution determined from ad-
sorption isotherms shows one distinguished maximum for
005FeKIL-2 and 01FeKIL-2 samples. 02FeKIL-2 and 05FeKIL-2
Operating conditions for part a: T=300–4808C, atmospheric pressure,
samples show very broad pore size distributions with no ap-
parent maximum because of the technique limits of the nitro-
ꢂ1
P
toluene =0.9 kPa, weight hourly space velocity=1.2 h . TOF=turnover
frequency.
[
38]
gen adsorption–desorption apparatus.
modified materials, 01FeKIL-2 possesses the best catalytic ac-
tivity and stability. The catalytic activity increases with the in-
crease in the iron content (up to 0.9 wt%); a further increase
in the iron content (02FeKIL-2 and 05FeKIL-2) leads to lower
catalytic activity. Of the modified materials, 005FeKIL-2 and
01FeKIL-2 possess the best catalytic stability at 3808C (Fig-
ure 5b); the conversion remains approximately constant over
3 h. The catalysts maintain their catalytic activity in three reac-
tion cycles. A well-defined trend of catalytic activity decrease is
observed for 02FeKIL-2. The stable activity of 05FeKIL-2 could
be related to the low activity of this sample.
Adsorption and catalytic performance
The toluene adsorption data of the studied materials are pre-
sented in Figure 4. All samples show close total adsorption ca-
pacity (29.3–34.9 mg toluene/100 mg sample) but different
times (t=5–25 min) to achieve the total capacity. The disor-
dered silica structure of KIL-2 materials facilitates the access to
the adsorption sites compared with ordered mesoporous sili-
[
42]
cates. The earlier breakthrough of toluene (Figure 4) for the
The formation of SiꢂOꢂFe species results in easier oxygen
release, which is an essential step in the oxidation reaction ac-
[43–45]
cording to the Mars–van Krevelen mechanism.
The mecha-
nism of oxidation assumes the adsorption of the VOC molecule
on the catalyst surface and its oxidation with lattice oxygen
following the oxidation of the reduced catalysts. The surface
properties (acidic and basic) of the catalyst affect the sorption
of organic molecules and therefore its conversion in the oxida-
tion reaction. The activity in the total oxidation of toluene is
connected with the interaction of aromatic electrons with
acidic centers of the catalyst, which increases the possibility of
the electrophilic attack of adsorbed oxygen and the combus-
tion of toluene molecules. Iron ions could act as Lewis acid
sites, which facilitates the interaction of toluene molecules
with the support and thus increases the catalytic activity. The
reduction properties of the samples can be used for the inter-
pretation of their catalytic performance as well (see the next
section).
Figure 4. Breakthrough curves for toluene adsorption on the initial silicate
nanoparticles (denoted as KIL-2) and iron-containing samples with Fe/Si
molar ratios 0.005 (denoted as 005FeKIL-2), 0.01 (denoted as 01FeKIL-2), 0.02
(denoted as 02FeKIL-2), and 0.05 (denoted as 05FeKIL-2). Operating condi-
tions: Tads =258C, atmospheric pressure, Ptoluene =0.9 kPa.
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 988

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Local environment of iron
clusters within the pores appear between 300 and 450 nm and
Fe O nanoparticles at the external surface of the crystal above
2
3
[47,48]
The information on the local environment of iron is obtained
through FTIR, UV/Vis, Mçssbauer, and temperature-pro-
grammed reduction (TPR) experiments.
450 nm.
more or less distorted isolated Fe sites; for tetrahedral Fe
ions in the framework positions of iron silicalite-1, they are ob-
Two charge transfer transitions were observed for
3
+
3+
3
+
A comparison of the FTIR spectra of parent and iron-func-
tionalized materials is shown in Figure 6. The asymmetric
stretching vibrations of SiꢂOꢂSi from the silica matrix appear
served at 214 and 240 nm, whereas for isolated Fe ions in
the octahedral symmetry, they are observed between 250 and
[49]
300 nm. Similar observations were found in disordered mes-
[19,33]
oporous materials.
The analysis of UV/Vis spectra of
0
2
05FeKIL-2 and 01FeKIL-2 samples reveals peaks at 245 and
90 nm. These bands are assigned to the charge transfer be-
tween the iron and oxygen atoms of FeꢂOꢂSi in the network,
which indicates the presence of tetrahedrally and octahedrally
3
+
[47,49]
coordinated Fe species, respectively.
The UV/Vis spectra
of 02FeKIL-2 and 05FeKIL-2 samples are different from those of
the other two spectra. The analysis of both spectra reveal
peaks at 250 and 355 nm. The first peak corresponds to isolat-
3
+
ed Fe species, whereas the latter can be assigned to the oli-
gonuclear iron complex. An additional peak at 213 nm attribut-
3
+
able to the presence of the isolated tetrahedral Fe ions is
also present in the spectrum of 02FeKIL-2. No contribution
greater than 450 nm is observed, which excludes the occur-
[48]
rence of iron oxide particles in all samples.
Figure 6. FTIR spectra of the initial and modified samples.
The TPR thermogravimetric (TG) profiles (Figure 8) show dif-
ferent reduction behaviors of the modified materials in terms
of dependence of iron content. 02FeKIL-2 and 05FeKIL-2 sam-
ples are characterized with two reduction peaks at low temper-
ꢂ1
at approximately 1090 cm . The slight shift of this band to
lower wavenumbers is observed for the iron-containing materi-
ꢂ1
als. The band at 960 cm probably arises from both SiꢂOꢂH
and SiꢂOꢂTi stretching vibrations. The simultaneous shift of
ꢂ1
the bands at 1090 and 960 cm to the lower wavenumbers is
usually ascribed to the presence of tetrahedrally coordinated
[
19]
metal ions in the silica matrix. The negligible changes in the
spectrum of 02FeKIL-2 and 05FeKIL-2 in comparison with
005FeKIL-2 and 01FeKIL-2 silicas are indicative of the lower
degree of iron incorporation into the silica matrix of these
samples.
Diffuse reflectance UV/Vis spectroscopy was used to investi-
3
+
gate the nature of Fe ions in FeKIL-2 samples (Figure 7). The
UV/Vis spectra of iron-containing zeolites are characterized by
3
+
intense Fe !O charge transfer bands, the position of which
provided information on the coordination and degree of ag-
Figure 8. TPR-TG profiles of the iron-modified samples. The samples were re-
[
46]
3+
3
ꢂ1
duced in a 50 vol% H
2
in Ar (flow rate=100 cm min ) to a temperature of
glomeration.
Thus, isolated Fe
ions give rise to bands
o
ꢂ1
600 C (heating rate=58Cmin ), which was maintained for 1 h. DTG=differ-
below 300 nm, whereas signals of small oligonuclear Fe O
x
y
ential thermal gravimetry.
atures in the intervals 250–3808C and 480–6008C, whereas
005FeKIL-2 and 01FeKIL-2 samples possess a single reduction
peak at high temperatures (480–6008C). The reduction peak at
low temperatures is attributed to the presence of oligonuclear
iron clusters, which are formed mainly in the samples with
high iron content (02FeKIL-2 and 05FeKIL-2). The reduction of
iron ions incorporated into the silica matrix of 005FeKIL-2 and
[42]
0
1FeKIL-2 is observed at high temperatures. This assumption
is also supported by the FTIR and UV/Vis data of the samples,
3
+
and the presence of tetrahedrally coordinated Fe ions can
be observed in the latter case. The TPR data of 02FeKIL-2 and
05FeKIL-2 samples show a peak at a low temperature and
Figure 7. UV/Vis spectra of the modified samples.
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 989

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
3
+
2+
ions, or if the Fe to Fe reduction proceeds, then the re-
Table 2. Redox properties of iron-modified KIL-2 materials.
2
+
duced Fe irons are expelled from the tetrahedral framework
[48]
Samples
Weight loss
Theoretical weight loss
site to the extra-framework position.
In contrast, in mesoporous materials, the reversible
3
+
2+
2+
0
3+
0
2
30–4108C 410–6008C Fe !Fe
Fe !Fe Fe !Fe
[
wt%]
[wt%]
[wt%]
[wt%]
[wt%]
3+
2+
Fe ÐFe reduction proceeds on isolated iron ions because
0
0
0
0
05FeKIL-2
0
0.07
0.13
0.25
0.33
0.07
0.13
0.42
0.55
0.13
0.26
0.83
1.10
0.20
0.39
1.25
1.65
the structure is more flexible and the crystallinity is less strict.
1FeKIL-2 0.02
2FeKIL-2 0.06
5FeKIL-2 0.18
2+
In comparison to the formation of silanolic SiꢂOH groups, Fe
3
+
[50,51]
ꢂOH groups form during the reduction of Fe ꢂO bonds.
This reversibility plays a role during the oxidation reaction by
[19]
3+
2+
transferring of oxygen atoms. Furthermore, the Fe !Fe
reduction also proceeds on the mentioned polyferrate FeꢂOꢂ
FeꢂOꢂ chains through the simple extraction of oxygen. How-
ever, this latter process is not reversible when the hydrogen
lower reduction ability; thus, the redox cycles in these samples
are more difficult compared with those in 005FeKIL-2 and
2
+
0
1FeKIL-2 samples (Table 2).
treatment is reverted to the repeated evacuation. The Fe
To obtain specific information on the local environment of
component formed upon the extraction of oxygen starts to
form oligomeric separate Fe O species, which does not partici-
iron, in situ Mçssbauer spectra were recorded for 01FeKIL-2,
2FeKIL-2, and 05FeKIL-2 samples under different conditions
after calcination and subsequent treatments at 640–650 K, i.e.,
x
3
y
+
2+
0
pate in further reversible Fe ÐFe process, that is, in trans-
ferring oxygen. The proportion of this “immobilized” iron in ag-
glomerates increases with the iron content, as demonstrated
(
evacuation and reduction in hydrogen, respectively). The spec-
tra are shown in Figure 9, and the data obtained from the de-
composition of spectra are listed in Table 3.
2
+
by the increase in the remaining Fe part after the hydrogen
treatment and the subsequent second evacuation. Thus, the
Mçssbauer studies demonstrate the concentration dependence
3
+
of the proportions of tetrahedrally coordinated single Fe
ions, and they also prove that a large proportion of tetrahe-
drally coordinated iron ions can participate in the reversible
3
+
2+
Fe ÐFe
redox process. Similar in situ spectra were ob-
tained on a related analogous mesoporous Fe-TUD-1 (Fe/Si=
[52]
0
.02) sample.
Hypothesis
The results from physicochemical studies of the local iron envi-
ronment indicate that iron is incorporated into the framework
for materials with an Fe/Si ratio up to 0.01. If the iron concen-
tration exceeds this value, iron deposits in the form of an oli-
gonuclear iron complex. The iron ions incorporated into the
Figure 9. In situ Mçssbauer spectra recorded for 01FeKIL-2, 02FeKIL-2, and
05FeKIL-2 samples a) as received, b) after evacuation at 3608C, c) after reduc-
tion in H at 3508C, and d) after 2nd evacuation at 3608C.
2
3+
2+
silica matrix can participate in Fe /Fe redox cycles more
easily, which is an essential step in the oxidation process ac-
cording to the Mars–van Krevelen mechanism. We assume
that, in the case of samples with lower iron content (005FeKIL-
2 and 01FeKIL-2), iron is incorporated predominantly as stable
3
+
The signal of the tetrahedrally coordinated isolated Fe
ions can be easily identified in the spectra. These ions demon-
3
+
3+
strate a characteristic Fe doublet with large quadrupole split-
Fe ions into the silica structure, whereas the oligonuclear
ꢂ1
ting (quadrupole splitting=1.8–2.0 mms and isomer shift=
iron complexes formed in samples with higher iron content
(02FeKIL-2 and 05FeKIL-2) are more prone to the agglomera-
tion process. The TPR-TG experiments (Figure 8) support our
idea and there is no significant change in the reduction ability
for 01FeKIL-2 during the second TPR-TG experiment
(Figure 10). The same experiment is performed with 05FeKIL-2.
The second TPR-TG experiment with 05FeKIl-2 shows a shift of
the peak at low temperature to a higher temperature of
4148C. This effect can be explained with some difficulties in
ꢂ1
[50]
0
.26–0.30 mms ) in the spectra recorded after evacuation.
Notably, the proportion of this component decreases with the
increase in the iron content; the corresponding relative intensi-
ty value is 67% for 01FeKIL-2, whereas it is only 44% in
05FeKIL-2 in the first series of evacuation. The other character-
istic observation in the analysis of spectra is that a large
3
+
2+
amount of iron participates in the reversible Fe ÐFe pro-
cess. In microporous ferrisilicates, this process is not prevailing;
3
+
2+
3+
2+
in their rigid crystalline structure, the Fe to Fe reduction is
the Fe !Fe transition related to the presence of an oligo-
3
+
not allowed for the framework-substituted tetrahedral Fe
nuclear iron complex.
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 990

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Conclusions
Table 3. Comparison of characteristic data extracted from in situ Mçssbauer spectra
for FeKIL-2 samples.
[
a]
Iron-containing silicate nanoparticles with interparti-
cle mesoporosity (pore diameter 2–50 nm) are highly
active adsorbents and catalysts for toluene removal
from polluted air. The iron content influences the
nature of the metal species in the silica matrix. The
Samples Treatment
Component IS
QS
[mms
FWHM
[mms
RI
[%]
ꢂ1
ꢂ1
ꢂ1
[mms
]
]
]
3
3
3
3
2
3
3
2
3
3
2
+
+
+
+
+
+
+
+
+
+
+
0
0
0
1FeKIL-2 as received
evacuation at 3508C
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
0.24
0.35
0.29
0.26
1.03
0.37
0.84
1.13
0.35
0.30
0.87
0.73
1.19
1.08
1.78
1.96
0.54
2.02
2.17
1.11
1.89
2.47
0.50
0.84
0.75
0.76
0.67
0.90
0.35
0.57
0.75
0.51
1.00
20
80
23
67
11
45
123
42
59
21
20
3
+
presence of Fe ions in the tetrahedral position is
observed for all modified materials. Oligonuclear iron
complexes are present in FeKIL-2 samples with an
Fe/Si molar ratio greater than 0.1. All the iron-func-
tionalized samples show high adsorption capacity.
The highest catalytic activity is observed for the
H
2
at 3608C
2nd evacuation at 3508C Fe
Fe
Fe
0
1FeKIL-2 sample with an Fe/Si molar ratio of 0.01.
3
+
The predominate incorporation of isolated Fe ions
into the silica matrix is observed for 01FeKIL-2, which
3
3
3
3
2
3
3
3
3
3
2
+
+
+
+
+
+
+
+
+
+
+
2FeKIL-2 as received
evacuation at 3508C
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
0.32
0.36
0.33
0.28
1.09
0.37
0.92
1.21
0.33
0.30
1.06
0.74
1.21
1.21
1.91
2.28
0.73
1.95
2.04
1.15
1.77
2.11
0.43
0.65
0.67
0.67
0.49
1.00
0.52
0.51
0.62
0.52
0.70
30
70
37
57
6
32
31
37
53
30
17
3+
2+
favors the fast Fe !Fe transition, essential for the
oxidation reaction. Iron incorporated into silica nano-
particles with interparticle mesoporosity (disordered
mesoporous KIL-2) was more efficient than iron incor-
porated into ordered mesoporous silicates MCM-41
H
2
at 3608C
[42]
2nd evacuation at 3508C Fe
and SBA-15.
Fe
Fe
Experimental Section
3
3
3
3
2
3
3
3
3
3
2
+
+
+
+
+
+
+
+
+
+
+
5FeKIL-2 as received
evacuation at 3508C
at 3608C
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
0.34
0.36
0.35
0.29
1.08
0.42
0.91
1.18
0.34
0.30
1.05
0.77
1.27
1.15
1.83
2.01
0.50
1.92
2.06
1.05
1.65
2.07
0.42
0.58
0.62
0.59
0.72
0.82
0.51
0.53
0.59
0.55
0.73
40
60
37
44
19
15
39
46
41
32
27
Synthesis
Iron with Fe/Si molar ratios ranging from 0.005 to 0.05
was incorporated into a new inorganic silicate matrix
from the KIL family (KIL-2) through a two-step solvother-
mal synthesis. KIL-2 is a mesoporous silicate formed
from silicate nanoparticles (10–20 nm in size) with inter-
H
2
2nd evacuation at 3508C Fe
[27]
particle porosity. We denoted the potential catalysts
Fe
Fe
005FeKIL-2 (Fe/Si=0.005, 0.270 g), 01FeKIL-2 (Fe/Si=
0
0
.01, 0551 g), 02FeKIL-2 (Fe/Si=0.02, 1.10 g), and
5FeKIL-2 (Fe/Si=0.05, 2.75 g). In the first step, iron(III)
[
a] IS=isomer shift, related to metallic iron; FWHM=full width at half maximum;
QS=quadrupole splitting; RI=relative intensity.
chloride hexahydrate (FeCl ·6H O, Aldrich; 005FeKIL-2:
3
2
0
0
4
.270 g, 01FeKIL-2: 0551 g, 02FeKIL-2: 1.10 g, and
5FeKIL-2: 2.75 g), tetraethyl orthosilicate (98%, Acros;
2.52 g), and triethanolamine (99%, Fluka; 15.07 g) were
stirred for 30 min. Finally, demineralized water was added to the
above mixture, followed by the addition of tetraethylammonium
hydroxide (20%, Acros; 14.73 g). The solution was mixed with
a magnetic stirrer to obtain a homogeneous gel. The final gel with
a molar composition of tetraethyl orthosilicate/0.5 triethanolamine/
(0.005–0.05) iron/0.1 tetraethylammonium hydroxide/11 water was
aged overnight at RT and then dried in an oven at 508C for 24 h.
In the second step, the gel was solvothermally treated in ethanol
in Teflon-lined stainless steel autoclaves at 1508C for 48 h. The
template was removed through calcination at 5008C for 10 h
ꢂ1
(
ramp rate=18Cmin ) in the flow of air. This also proved the ther-
mal stability of the product. The hydrothermal stability of the prod-
ucts was verified by keeping them in boiling water for 2 h.
Characterization
Figure 10. Second TPR-TG profiles of the iron-modified samples. After the
first reduction (Figure 8), the samples were reoxidized in a flow of air to
a temperature of 5008C (heating rate=108Cmin ), which was maintained
Basic characterization
ꢂ1
The basic structural characterization of prepared catalysts was per-
2
for 1 h, and then the samples were reduced again in a 50 vol% H in Ar
3
ꢂ1
(
flow rate=100 cm min ) to a temperature of 6008C (heating rate=58
formed by using XRD, SEM, and N sorption measurements. XRD
2
ꢂ1
Cmin ), which was maintained for 1 h.
diffractograms were recorded on a PANalytical X’Pert PRO (HTK)
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 991

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
high-resolution diffractometer with CuKa1 radiation (1.5406 ꢁ) in
the 2q range from 0.5 to 58 (100 s per step of 0.0398) and from 5
to 608 (100 s per step of 0.0168) for the samples and from 108 to
Adsorption
The system for toluene adsorption assessment consisted of a fixed-
bed flow reactor coupled with a GC system. The samples were
treated at 4008C for 1 h before the adsorption experiment that
used a N2 flow of 60 mLmin . The toluene adsorption was per-
formed at 258C in N flow (flow rate=30 mLmin ). By using this
technique, the breakthrough curves, which represented the con-
centration of toluene in the outlet gas versus time, were obtained.
From the numerical integration of these curves, an adsorption ca-
pacity of 880 ppm toluene was determined.
9
08 (100 s per step of 0.0168) for the sample holder equipped with
a fully opened X’Celerator detector. Morphology and surface prop-
erties of the samples were observed by using SEM on a Zeiss
Supra 3VP microscope. Elemental analysis of all samples was per-
formed by using the energy-dispersive X-ray method with INCA
Energy system attached to a Zeiss Supra 3VP microscope. Nitrogen
physisorption measurements were performed at 77 K with a Micro-
meritics ASAP 2020 volumetric adsorption analyzer. Before the ad-
sorption analysis, the samples were outgassed under vacuum at
ꢂ1
ꢂ1
2
4
73 K for 2 h in the port of the adsorption analyzer. The BET specif-
Catalysis
ic surface area was calculated from the adsorption data in the rela-
tive pressure range 0.06–0.165. The total pore volume was esti-
mated on the basis of the amount adsorbed at a relative pressure
of 0.98. The mesopore volume Vme and the external surface area
Sex were determined by using the a plot method from the adsorp-
tion data in the range of the standard reduced adsorption from 2.5
to 3.0.
microporosity in these materials. The micropore volume V was es-
timated from the a interval 0.75–1.00. In the a plot calculations,
[53]
Before the catalytic test, the samples were pretreated in air at
4
508C for 1 h. Toluene oxidation was studied at atmospheric pres-
[36]
sure by using a fixed-bed flow reactor, air as the carrier gas, and
the catalyst (30 mg; prepared as agglomerates with 0.2–0.8 mm
particle size) diluted with glass beads (60 mg) of the same diame-
ter previously checked to be inactive. The air stream was passed
through a saturator filled with toluene and equilibrated at 08C
s
[54,55]
The same method was applied to show the presence of
mi
[55]
s
s
(ptoluene =0.9 kPa). The activity was determined in the temperature
a
2
macroporous silica material LiChrospher Si-1000 (SBET
2.1 m g ) was used as a reference adsorbent. The pore size
=
ꢂ1
interval of 350–4508C at a weight hourly space velocity of 1.2 h .
The online analysis of the reaction products was performed by
using the GC apparatus equipped with a flame ionization detector
and a thermal conductivity detector on a 25 m PLOT Q capillary
column (25 mꢂ0.53 mmꢂ20 mm) with the polystyrene–divinylben-
zene stationary phase (Chrompack). The turnover frequency was
calculated as the number of toluene molecules converted per iron
atom per second. The samples were treated in air at 5008C for 1 h
after every catalytic cycle.
2
ꢂ1
[56]
distributions were calculated from nitrogen adsorption data by
using an algorithm based on the ideas of Barrett, Joyner, and Ha-
lenda. The mesopore diameters were determined as the maxi-
mum on the pore size distribution for given samples.
[40]
Acknowledgements
Local environment of iron
The financial support from the Bulgarian Scientific Fund (Bulga-
ria-Slovenia bilateral project HTC 02/205), the Slovenian Research
Agency (research program P1-0021, research project Z1-9144,
Slovenia-Hungary bilateral project BI-HU/11-12-010, Slovenia-Bul-
garia bilateral project BI-BG/11-12-009), and the EN-FIST Centre
of Excellence is acknowledged. The authors thank Dr. M. Dimitrov
for technical support.
The local environment of iron was investigated by using diffuse re-
flectance UV/Vis, FTIR, and Mçssbauer spectroscopies. The diffuse
reflectance UV/Vis spectra of the samples in the UV/Vis region
were recorded on a Lambda 40P (Perkin–Elmer) UV/Vis spectropho-
tometer equipped with an integrating sphere. Spectra were decon-
voluted to the Gaussian bands by using the Microcal Origin 8.1
software). The FTIR spectrum was recorded from KBr pellets
(
99 wt% of KBr) on a Brucker Vector 22. In situ Mçssbauer spectra
[49]
were recorded for samples treated in a special measuring cell.
Pellets were pressed from the 400–800 mg amounts of samples
and were evacuated at approximately 0.1 Pa and treated subse-
quently in hydrogen (both treatments at 640 K). The 57Co/Rh
Keywords: heterogeneous
mesoporosity · iron–silica · nanoparticles · oxidation
catalysis
·
interparticle
(
1 GBq) source was used for the measurements. Isomer shift values
[
[
[
1] F. Bertinchamps, C. Gregoire, E. M. Gaineaux, Appl. Catal. B 2006, 66,
0–22.
2] Ch.-H. Wang, Sh.-Sh. Lin, Ch.-L. Chen, H.-Sh. Weng, Chemosphere 2006,
4, 503–509.
3] A. A. dos Santos, K. M. Lima, R. T. Figueiredo, S. M. da Silva Egues,
A. L. D. Ramos, Catal. Lett. 2007, 114, 59–63.
are related to a-iron. The obtained spectra were decomposed to
Lorentzian-shaped components. The accuracy of the positional pa-
1
ꢂ1
rameters is approximately ꢄ0.03 mms . The TPR-TG investigations
6
were performed in a Setaram TG92 instrument. Typically, 40 mg of
the sample was placed in a microbalance crucible and heated in
3
ꢂ1
a 50 vol% H in Ar (flow rate=100 cm min ) to a temperature of
6
Before the TPR-TG experiments, the samples were treated in situ in
a flow of air to a temperature of 5008C (heating rate=108Cmin ),
which was maintained for 1 h. After the first reduction, the samples
were reoxidized in a flow of air (heating rate=108Cmin ) to
a temperature of 5008C, which was maintained for 1 h, and then
the samples were heated in a 50 vol% H in Ar (flow rate=
[4] R. S. G. Ferreira, P. G. P. de Oliveira, F. B. Noronha, Appl. Catal. B 2001, 29,
2
ꢂ1
008C (heating rate=58Cmin ), which was maintained for 1 h.
275–283.
[
5] M. Zhang, B. Zhou, K. T. Chuang, Appl. Catal. B 1997, 13, 123–130.
[6] M. A. Centeno, M. Paulis, M. Montes, J. A. Odriozola, Appl. Catal. A 2002,
34, 65–78.
ꢂ1
2
[7] V. Gaur, A. Sharma, N. Verma, Carbon 2005, 43, 3041–3053.
[8] S. Ch. Kim, W. G. Shim, J. Hazard. Mater. 2008, 154, 310–316.
[9] C. Doornkamp, V. Ponec, J. Mol. Catal. A-Chem. 2000, 162, 19–32.
ꢂ1
2
[10] J. C.-S. Wu, T.-Y. Chang, Catal. Today 1998, 44, 111 – 118.
[11] P. Papaefthimiou, Th. Ioannides, X. E. Verykios, Appl. Catal. B 1997, 13,
175–184.
3
ꢂ1
1
5
00 cm min ) to
a
temperature of 6008C (heating rate=
ꢂ1
8Cmin ), which was maintained for 1 h.
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 992

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[
12] S. K. Gangwal, M. E. Mullins, J. J. Spivey, P. R. Caffrey, B. A. Tichenor, Appl.
[36] T. A. Zepeda, J. L. G. Fierro, B. Pawelec, R. Nava, T. Klimova, G. A.
Fuentes, T. Halachev, Chem. Mater. 2005, 17, 4062–4073.
[37] M. Kruk, M. Jaroniec, Chem. Mater. 2001, 13, 3169–3183.
[38] A. Lemaire, J. C. Rooke, L.-H. Chen, B.-L. Su, Langmuir 2011, 27, 3030–
3043.
[39] W. W. Lukens, P. Schmidt-Winkel, D. Zhao, J. Feng, G. D. Stucky, Lang-
muir 1999, 15, 5403–5409.
[40] E. P. Barrett, L. G. Joyner, P. P. Halenda, J. Am. Chem. Soc. 1951, 73, 373–
380.
Catal. 1988, 36, 231.
[
[
13] S.-W. Baek, J.-R. Kim, S.-K. Ihm, Catal. Today 2004, 93–95, 575–581.
14] H. L. Tidahy, S. Siffert, J.-F. Lamonier, R. Cousin, E. A. Zhilinskaya, A.
Aboukais, B.-L. Su, X. Cannet, G. De Weireld, M. Frere, J.-M. Giraudon, G.
Leclercq, Appl. Catal. A 2007, 70, 377–383.
[
[
15] Y. Wang, Q. Zhang, T. Shishido, K. Takehira, J. Catal. 2002, 209, 186–196.
16] Ch. Wu, Y. Kong, F. Gao, Y. Wu, Y. Lu, J. Wang, L. Dong, Microporous Mes-
oporous Mater. 2008, 113, 163–170.
[
[
[
[
17] W. A. Carvalho, P. B. Varaldo, M. Wallau, U. Schuchardt, Zeolites 1997, 18,
[41] P. T. Tanev, L. T. Vlaev, J. Colloid Interface Sci. 1993, 160, 110–116.
[42] M. Popova, A. Szegedi, Z. Cherkezova-Zheleva, I. Mitov, N. Kostova, T.
Tsoncheva, J. Hazard. Mater. 2009, 168, 226–232.
[43] S. H. Taylor, C. S. Heneghan, G. J. Hutchings, I. D. Hudson, Catal. Today
2000, 59, 249–259.
408–416.
18] T. Kawabata, Y. Ohishi, S. Itsuki, N. Fujisaki, T. Shishido, K. Takaki, Q.
Zhang, Y. Wang, K. Takehira, J. Mol. Catal. 2005, 236, 99–106.
19] M. S. Hamdy, G. Mul, J. C. Jansen, A. Ebaid, Z. Shan, A. R. Overweg, Th.
Maschmeyer, Catal. Today 2005, 100, 255–260.
20] R. S. Araffljo, D. C. S. Azevedo, E. Rodrıguez-Castellon, A. Jimenez-Lopez,
C. L. Cavalcante, Jr., J. Mol. Catal. 2008, 281, 154–163.
[44] S. Minicò, S. Scire, C. Crisafulli, R. Maggiore, S. Galvagno, Appl. Catal. B
2000, 28, 245–251.
[45] E. M. Cordi, P. J. O’Neill, J. L. Falconer, Appl. Catal. B 1997, 14, 23–36.
[46] J. PØrez-Ramírez, M. S. Kumar, A. Brückner, J. Catal. 2004, 223, 13–27.
[47] S. Bordiga, R. Buzzoni, F. Geobaldo, C. Lamberti, E. Giamello, A. Zec-
china, G. Leofanti, G. Petrini, G. Tozzola, G. Vlaic, J. Catal. 1996, 158,
486–501.
[
[
[
21] A. Taguchi, F. Schüth, Microporous Mesoporous Mater. 2005, 77, 1–45.
22] C. C. Rodrigues, Catal. Commun. 2007, 8, 1227–1231.
23] I. Sobczak, M. Ziolek, M. Renn, P. Decyk, I. Nowak, M. Daturi, J.-C. Laval-
ley, Microporous Mesoporous Mater. 2004, 74, 23–36.
[
[
24] A. Szegedi, M. Heged u˝ s, J. L. Margitfalvi, I. Kiricsi, Chem. Commun. 2005,
[48] J. PØrez-Ramírez, J. Catal. 2004, 227, 512–522.
[49] A. Risti c´ , K. Lµzµr, H. Solt, V. Kau cˇ i cˇ , CrystEngComm 2011, 13, 1946–
1952.
[50] K. Lµzµr, G. Pµl-BorbØly, . Szegedi, H. K. Beyer, Stud. Surf. Sci. Catal.
2002, 142, 1347–1354.
1441–1443.
25] A. Szegedi, Y. Kónya, D. MØhn, E. Solymµr, G. Pal-BorbØly, Y. Horvµth, L.
Biró, I. Kiricsi, Appl. Catal. A 2004, 272, 257–266.
[26] M. Kang, M.-H. Lee, Appl. Catal. A 2005, 284, 215–222.
[27] M. Zhou, J. Yu, B. Chang, J. Hazard. Mater. 2006, 137, 1838–1847.
[
28] N. Novak Tu sˇ ar, S. Jank, R. Gläser, ChemCatChem 2011, 3, 254–269.
[29] A. Szegedi, M. Popova, Z. Cherkezova-Zheleva, A. Dimitrova, I. Mitov, Mi-
croporous Mesoporous Mater. 2010, 136, 106–114M.
[51] K. Lµzµr, . Szegedi, G. Pµl-BorbØly, A. N. Kotasthane, P. Fejes, Catal.
Today 2005, 110, 239–246.
[52] N. Novak Tu sˇ ar, A. Risti c´ , S. Cecowski, I. Ar cˇ on, K. Lµzµr, H. Amenitsch, V.
Kau cˇ i cˇ , Microporous Mesoporous Mater. 2007, 104, 289–295.
[53] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309–319.
[54] A. Sayari, P. Liu, M. Kruk, M. Jaroniec, Chem. Mater. 1997, 9, 2499–2506.
[55] M. Kruk, M. Jaroniec, C. H. Ko, R. Ryoo, Chem. Mater. 2000, 12, 1961–
1968.
[
[
[
[
[
[
30] M. Popova, . Szegedi, K. Lµzµr, Z. Kµroly, Microporous Mesoporous
Mater. 2012, 151, 180–187.
31] N. Novak Tu sˇ ar, D. Mau cˇ ec, M. Rangus, I. Ar cˇ on, M. Mazaj, M. Cotman, A.
Pintar, V. Kau cˇ i cˇ , Adv. Funct. Mater. 2012, 22, 820–826.
32] N. Novak Tu sˇ ar, A. Risti c´ , G. Mali, M. Mazaj, I. Ar cˇ on, D. Ar cˇ on, V. Kau cˇ i cˇ ,
N. Zabukovec Logar, Chem. Eur. J. 2010, 16, 5783–5793.
33] A. Risti c´ , D. Mau cˇ ec, S. K. Henninger, V. Kau cˇ i cˇ , Microporous Mesoporous
Mater. 2012, 164, 266–272.
[56] B. Tan, H. J. Lehmler, S. M. Vyas, B. L. Knutson, S. E. Rankin, Nanotechnol-
ogy 2005, 16, S502–S507.
34] S. Kahraman, M. Önal, Y. Sarikaya, I. Bozdogan, Anal. Chim. Acta 2005,
552, 201–206.
35] K. S. W. Sing, D. H. Everett, R. A. Haul, L. Moscou, R. A. Pirroti, J. Rouquer-
Received: August 16, 2012
ol, T. Siemieniewska, Pure Appl. Chem. 1985, 57, 603–619.
Published online on January 4, 2013
ꢀ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 986 – 993 993