The role of zeolite FeZSM-5 porous structure for heterogeneous Fenton catalyst actuvuty and stability

Microporous and Mesoporous Materials 189 (2014) 181–188

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials
journal homepage: www.elsevier.com/locate/micromeso

The role of zeolite Fe-ZSM-5 porous structure for heterogeneous Fenton
catalyst activity and stability
K.A. Sashkina a, E.V. Parkhomchuk a,b,⇑, N.A. Rudina b, V.N. Parmon a,b
a
b

Novosibirsk State University, 2 Pirogova st., Novosibirsk 630090, Russia
Boreskov Institute of Catalysis SB RAS, 5 Lavrentieva st., Novosibirsk 630090, Russia

a r t i c l e

i n f o

Article history:
Available online 27 November 2013
Dedicated to Dr. Michael Stöcker on the
occasion of his retirement as Editor-in-Chief
of Microporous and Mesoporous Materials.
Keywords:
Fe-ZSM-5
Hierarchical zeolite
Nanozeolite
Heterogeneous Fenton catalyst
Catalyst stability

a b s t r a c t
Four types of iron containing materials have been synthesized: conventional zeolite Fe-ZSM-5 (conv),
hierarchical zeolite Fe-ZSM-5 (hier), small crystals (d = 330 nm) of zeolite Fe-ZSM-5 (nano) and ferric
oxide species supported on the amorphous silica Fe/SiO2. Samples were prepared by hydrothermal treatment, polystyrene spheres were used as a template for Fe-ZSM-5 (hier) and Fe/SiO2. The materials were
characterized by different techniques. Nature of iron-containing particles in the samples and stability of
iron species in the reaction media were suggested by using thermodynamic considerations. All solidphase Fe-containing samples as well as dissolved Fe(NO3)3 were tested in H2O2 decomposition reactions
in absence or presence of iron-complexing agent Na2EDTA, which has been used to test the catalyst stability. Catalytic activity of ferric species in hydrogen peroxide decomposition for small 330-nm crystals of
Fe-ZSM-5 was 1.4 times higher than for large zeolite crystals, and significant decrease of the activity was
observed for samples containing amorphous silica phase. Experimental results showed that ferric sites in
zeolite were stable due to the limited diffusion of Na2EDTA in zeolite phase. Wet hydrogen peroxide
oxidation of organic complexing agents by H2O2 using Fe-containing zeolites has a good potential for
purification of nuclear waste water.
Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction
Fe-ZSM-5 has been shown to be a promising heterogeneous solid-phase catalyst in total oxidation of a series of organic substrates
with low molecular weight (MW) by hydrogen peroxide [1–4].
Mineralization degree of phenol, 1,1-dimethylhydrazine and ethanol, as well as extent of H2O2 utilization is higher in such a heterogeneous system compared with the homogeneous Fenton system
due to effective adsorption of organic substrate on zeolite surface
[5]. On the other hand conventional zeolitic material is ineffective
in oxidation of high MW organics because of specific porous structure with pore size of 0.55 nm. Mineralization degree of lignin is
significantly lower in the Fe-ZSM-5/H2O2 system compared with
homogeneous ones, such as Fe(NO3)3/H2O2 and H2O2/UV [6]. This
is due to excessive distance from catalytic sites where the hydroxyl
radicals are formed inside of zeolite crystal to organic molecule adsorbed on the external surface of crystal particle. These diffusion
limitations result in prevalence of oxygen release reaction over
organics oxidation processes in case of high MW substrates.
To expand zeolite use for wet peroxide oxidation of hard convertible macromolecules accessibility of catalytic sites for high
MW substrates should be significantly increased. In order to pre⇑ Corresponding author at: Boreskov Institute of Catalysis SB RAS, 5 Lavrentieva

st., Novosibirsk 630090, Russia. Tel./fax: +7 (383)333 16 17.
E-mail address: (E.V. Parkhomchuk).
1387-1811/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
/>
pare zeolites with additional meso or macroporosity a range of
techniques may be used [7–11]. Here, hierarchically ordered zeolitic material Fe-ZSM-5 (hier) has been prepared with the use of
a template consisting of polystyrene (PS) spheres according to
the method described by Stein in [12]. Earlier hierarchical FeZSM-5 was tested in total oxidation of large organic molecules,
Na2EDTA and lignin, used as large model compounds [13]. As a result of a great increase in catalytic site accessibility performance of
Fe-ZSM-5 (hier) in oxidation of large Na2EDTA molecule and high
MW lignin appeared to be really improved compared with conventional zeolite [13]. However the question on catalyst stability during oxidation of high MW organics when a range of different
organic acids may be formed is still open. This question is obviously related to the question on the nature of iron-containing sites
in zeolites and amorphous silica phase. While there are numerous
available literature data on the nature of sites in zeolites, which are
active in selective oxidation of benzene to phenol or methane to
methanol by N2O [14–16], selective reduction of nitrogen oxides
by hydrocarbons [17,18], and decomposition of N2O [19,20], the
nature of active sites for heterogeneous Fenton reactions remains
a disputable problem. It is worth noting that hierarchical zeolite
h-Fe-ZSM-5 represents a mixture of two phases: ZSM-5 nanocrystals and amorphous silica globules with a wide size distribution
[21]. Catalytic activity as well as stability of iron containing sites
located in these two phases may be expected to be different.


182

K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

To clarify separate roles of these two iron containing phases
(zeolitic and silica) for wet peroxide oxidation catalysis we synthesized pure crystalline sample, consisting of small Fe-ZSM-5 crystals

with the size of 330 nm, and pure amorphous sample, containing
ferric species supported on the silica phase (labeled as Fe/SiO2)
and compared them with conventional and hierarchical Fe-ZSM5. On the whole four types of iron containing materials were
studied: conventional zeolite Fe-ZSM-5 (conv), hierarchical zeolite
Fe-ZSM-5 (hier), small crystals of zeolite Fe-ZSM-5 (nano) and Fe/
SiO2. In order to reveal phase compositions and textural differences
a range of techniques was used, including X-ray diffraction, low
temperature N2 adsorption, scanning and high resolution transmission electron microscopies, UV–vis diffuse reflectance spectroscopy. Samples were tested in hydrogen peroxide decomposition
reaction to determine catalytic activity in formation of hydroxyl
radical from H2O2 – the main source of oxidative activity of Fenton
reagent in organics oxidation in acidic and neutral media. Since a
wide range of organic acids may be formed during Fenton reactions
stability of catalytically active sites requires to be carefully studied.
The effect of iron-complexing agent Na2EDTA, having high stability
constants with iron ions, on catalytic activity and stability of the
samples was studied. Unlike ferric oxide species supported on
the amorphous SiO2 ferric sites in Fe-ZSM-5 appeared to be stable
under the action of organic acids.
2. Experimental
2.1. Chemicals
Styrene monomer, inhibited with 1% hydroquinone, was purchased from Ltd. ‘‘Angara-reactive’’. It was washed 4 times in separatory funnel with an equal volume of 1 M aqueous solution of
sodium hydroxide, followed by 4 times distilled water to remove
the inhibitor before polymerization. Tetraethylorthosilicate (TEOS)
was purchased from Ltd. ‘‘Angara-reactive’’, hydrogen peroxide
(30% aqueous solution) – from company ‘‘Baza No1 Khimreactivov’’,
sulfuric acid (97%) – from Moscow Chemical company ‘‘Laverna’’,
sodium hydroxide – from Ltd. ‘‘Tellura’’, iron (III) nitrate nonahydrate were produced in Boreskov Institute of Catalysis, 95% ethanol
EtOH of technical grade were obtained from Ltd. ‘‘Pharmaceya’’.
Silica fumed powder (99.8%), tetrapropylammonium hydroxide
(TPAOH, 25% solution in water), tetrapropylammonium bromide

(98%), potassium persulfate (99%) were purchased from Sigma
Aldrich, Germany.

dried and calcined at 500 °C for 5 h in air. The sample is labeled
Fe-ZSM-5 (conv-as). For catalyst pretreatment the powder of FeZSM-5 (conv-as) was suspended to the 1 M aqueous solution of
oxalic acid in concentration of 100 g L1 and stirred for 1 h at
50 °C. The catalyst was rinsed with distilled water to pH 7.0, dried
in air and calcined at 500 °C for 3 h. The pretreated sample is labeled Fe-ZSM-5 (conv).
2.2.3. Synthesis of hierarchical zeolite
Hierarchical zeolite Fe-ZSM-5 (hier) was synthesized using a PS
template as a macropore generating agent and TPAOH as a SDA.
The sample Fe-ZSM-5 (hier) was produced using ferric nitrate with
the SiO2:Fe2O3:TPAOH:H2O molar ratio of 1:0.015:0.7:17.5. Thereafter, PS template was impregnated with the gel with weight ratio
1 SiO2:1 PS. The mixture was subjected to hydrothermal synthesis
at 110 °C for 40 h. The product was washed with abundant amount
of water, then dried at an ambient temperature overnight and finally calcined at 500 °C for 8 h in air. The control sample of Fe-containing hierarchical zeolite was pretreated by the oxalic acid as
stated above, but this procedure did not change the catalytic activity, thus the sample described in the paper was not exposed to this
procedure.
2.2.4. Synthesis of small zeolite crystals
Zeolite Fe-ZSM-5 (nano) with small crystal size was synthesized
in hydrothermal conditions from the precursor gel containing
TEOS, ferric nitrate, EtOH and TPAOH with the TEOS:Fe2O3:TPAOH:EtOH:H2O molar ratio of 1:0.01:0.275:4.8:12.3. Mixture
of precursors were held in the autoclave at 115 °C for 24 h and then
at 150 °C for 24 h. The resulting suspension was characterized by
laser diffraction to measure the size of particles. Then the suspension was centrifuged, the precipitate was washed with abundant
amount of water, dried at an ambient temperature overnight and
finally calcined at 500 °C for 5 h in air. The sample is labeled FeZSM-5 (nano-as). The powder of Fe-ZSM-5 (nano-as)was suspended to the 1 M aqueous solution of oxalic acid in concentration
of 100 g L1 and stirred for 10 min at 50 °C. The pretreated catalyst
was rinsed with distilled water to pH 7.0, dried in air and calcined
at 500 °C for 3 h. The sample is labeled Fe-ZSM-5 (nano).

2.2.5. Synthesis of Fe-containing amorphous silica
Fe-containing amorphous silica Fe/SiO2 was synthesized by the
same way as h-Fe-ZSM-5 described above but without sustaining
PS templates over the boiling water for 1 h before hydrothermal
synthesis.

2.2. Catalyst preparation

2.3. Catalyst characterization

2.2.1. PS template preparation
PS were synthesized using emulsifier-free emulsion polymerization technique as described elsewhere [22,23]. Emulsion
polymerization temperature was 90 °C. PS spheres were packed
by centrifugation at relative acceleration of 390g. Obtained PS template was washed by ethanol and dried in air. Before hydrothermal
synthesis PS templates were put on the grid and sustained over
boiling water for 1 h.

The X-ray diffraction analysis was performed by a diffractometer HZG-4 with a Cu-Ka radiation in the angle range 2h from 5° to
40°. The Fe content of the catalysts and iron concentration in the
solutions after reactions were determined by the inductive coupled
plasma optical emission spectroscopy (ICP–OES). Scanning electron microscopy (SEM) images were acquired using JSM-6460LV
microscope at 15–20 kV accelerating voltage, high-resolution
transmission microscopy (HRTEM) images of the samples were
made on JEM-2010 microscope at 0.14 nm resolution and 200 kV
accelerating voltage. Particle size of sample n-Fe-ZSM-5 was measured using suspension dilution with ethanol by laser diffraction
on the Mastersizer-2000. UV-vis diffuse reflection (DR) spectra
were acquired at ambient temperature using a Shimadzu UV2501 PC at interval 11,000–54,000 sm1. Low-temperature
nitrogen adsorption isotherms were measured at 196 °C on
ASAP-2400. Prior to the measurements the samples were
outgassed at 250 °C for 8 h. The specific surface area (SBET) was

determined by applying Brunauer–Emmet–Teller (BET) equation

2.2.2. Synthesis of conventional zeolites
The synthetic Fe-containing conventional zeolite Fe-ZSM-5
(conv) was produced hydrothermally from precursor gel containing silica powder, sodium hydroxide, ferric nitrate and TPABr as a
structure-directing agent (SDA). The molar ratio in the
SiO2:NaOH:TPABr:Fe2O3:H2O
mixture
was
chosen
as
1:0.2:0.11:0.028:25, respectively. The mixture was placed to a
Teflon-coated stainless steel autoclave and kept at 150 °C for
72 h. Zeolite crystals were filtered, rinsed with distilled water,


K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

from adsorption branches in the relative pressure range of 0.05–
0.3. The external surface area (SExt) and micropore volume (Vmic)
were calculated by as-method. The value of Vtot was single point
total pore volume at P/P0 = 0.98. Hierarchy factor (HF) was calculated as SExt/SBET Vmic/Vtotal.
2.4. Catalytic activity tests
Iron-containing samples were tested in H2O2 decomposition
reactions in absence and presence of 1 g L1 sodium ethylenediaminetetraacetate (Na2EDTA). Hydrogen peroxide decomposition
was carried out in magnetically stirred glass batch reactor with
50 mL of aqueous phase and 2956 mL of gaseous phase, both
thermostated at 25 °C. 5 mM Fe(NO3)3 and 20 g L1 zeolites were
used as catalysts for homogeneous and heterogeneous reactions,
respectively. The H2O2 decomposition rate W O2 at [H2O2]0 = 1.1 M

was determined as the oxygen release rate measured barometrically in Pa s1.
3. Results and discussion
Morphologies of Fe-containing materials synthesized in given
work are shown in Figs. 1 and 2. Conventional zeolite Fe-ZSM-5
(conv) was found to have large polycrystals of 2–5 lm in diameter
(Fig. 1a). Zeolite Fe-ZSM-5 (nano) contains uniform small crystals
(Fig. 1b). According to laser diffraction analysis Fe-ZSM-5 (nano)
crystals have the mean size of 330 nm and narrow size distribution
(Fig. 1, inset). It can be also seen in the HRTEM images of the sample (Fig. 2a). The hierarchical zeolite Fe-ZSM-5 (hier) has interconnected macroporous system, obtained macropores being PS
template replica (Fig. 1c). In our previous work walls of
macropores were shown to contain small zeolite ZSM-5 crystals
stuck together with amorphous silica, both zeolite crystals and

SiO2 globules having wide particle size distribution [21]. The last
sample Fe/SiO2 consists of amorphous silica globules with wide
particle size distribution (Figs. 1d and 2b).
XRD patterns of materials obtained are given in Fig. 3. XRD reflexes for Fe-ZSM-5 (conv), Fe-ZSM-5 (nano) and Fe-ZSM-5 (hier)
samples correspond to MFI structure [24], Fe/SiO2 sample is
amorphous. It should be emphasized that Fe-ZSM-5 (conv) and
Fe-ZSM-5 (nano) have high crystallinity (Table 1). The crystallinity
of Fe-ZSM-5 (hier) is only 58% due to the presence of amorphous
phase.
Textural properties of materials are shown in Table 1, low temperature adsorption isotherms are shown in Fig. 4. One can see that
all samples have high values of total surface area (416–838 m2/g)
measured by BET method. Hierarchical zeolites and amorphous
samples are characterized also by high external surface area and
total pore volume resulting from the microporosity of particles
(Fig. 5), which in turn have a wide size distribution with a large
proportion of fine particles particularly in case of Fe/SiO2
(Fig. 2b). Hierarchy factor [HF = (Vmic/Vtotal)  (SExt/SBET)] was calculated for all obtained samples, the amorphous sample having

the highest value of HF = 0.11 (Table 1). In our previous work we
have found that the presence of amorphous phase resulted to
reduction of the micropore volume and therefore decreasing of
the HF value, but this is not always the case. Amorphous sample
Fe/SiO2 has both high micropore volume and external surface area
due to presence of TPAOH and PS template during the synthesis.
The pore size distribution confirms the micropores generation in
the amorphous sample, micropores are likely to be formed due
to the presence of TPAOH during the hydrothermal synthesis. N2
adsorption isotherms for hierarchical zeolite and amorphous sample have a large hysteresis loop indicating the presence of mesopores and wide pore distribution (Fig. 4). A horizontal hysteresis
loop is observed for highly crystallized samples Fe-ZSM-5 (conv)
and Fe-ZSM-5 (nano), indicating inkbottle-type mesopores. A

a

b
Volume, %

20

Mean size, 330 nm

15
10
5
0

0,1

1

10

100

Particle size, 10-6 m

c

183

1000

d

Fig. 1. SEM images of Fe-containing samples: (a) Fe-ZSM-5 (conv), (b) Fe-ZSM-5 (nano), (c) Fe-ZSM-5 (hier) and (d) Fe/SiO2. The particle size distribution of the sample FeZSM-5(nano), determined by laser diffraction, is shown in the inset.


184

K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

Fig. 2. HRTEM images of (a) Fe-ZSM-5 (nano) and (b) Fe/SiO2.

sloped hysteresis loop in case of samples containing amorphous
silica phase indicates a presence of cylindrical mesopores. As we
will see later in this work the type of mesopores will play a key role
for iron containing catalytic site stability in Fenton reactions.
The state of iron species in the obtained samples were studied
by HRTEM analyses and UV–vis DR spectroscopy. Fig. 6 shows

the UV–vis DR spectra of materials. For all samples two strong
bands at 46,500 b 41,500 cm1 can be ascribed to t1 ? t2 and
t1 ? e transitions due to the metal–oxygen charge transfer. The
spectrum for Fe/SiO2 with the band at 20,000 cm1 indicates that
iron presents here in large oxide aggregates [25]. For nonactivated
white zeolite samples Fe-ZSM-5 (conv-as) and Fe-ZSM-5 (nano-as)
absorption band edge at the 37,500 cm1 was observed and it was
typical for ZSM-5 zeolites. Extremely weak bands, referred to forbidden d–d transitions of the zeolitic Fe3+ in tetrahedral oxygen
coordination, shown in the Fig. 6b, may be clearly distinguished
in the Fe-ZSM-5 (conv-as) and barely seen in the Fe-ZSM-5
(nano-as). According to Tanabe–Sugano diagram the band at
22,700 cm1 is referred to transition 6 A1  4 T1 ðGÞ, at 24,600 cm1
– to 6 A1  4 T2 ðGÞ and the band at 26,800 cm1 corresponds to
sum of transitions 6 A1  4 A2 and 6 A1  4 EðGÞ [26]. UV–vis DR spec-

800
400

Fe-ZSM-5 (conv)
0

Intensity, r. u.

800
400

Fe-ZSM-5 (nano)
0
800
400

h-FeZSM-5
0
800
400

Fe/SiO2

0
5

10

15

20

25

30

35

40

2Θ (degree)
Fig. 3. XRD patterns of the Fe-containing samples.

Table 1
Iron content, crystallinity and textural properties of iron-containing samples.

Sample

Fe (wt.%)

Crystallinity (%)

SBET (m2/g)

SExt (m2/g)

Vtotal (cm3/g)

Vmic (cm3/g)

Hierarchy factor

Fe-ZSM-5 (conv)
Fe-ZSM-5 (nano)
Fe-ZSM-5 (hier)
Fe/SiO2

2.74
1.73
1.29
1.89

97
100
58
0

416
543
454
838

27
77
397
475

0.22
0.54
0.70
0.79

0.19
0.19
0.07
0.16

0.06
0.05
0.09
0.11

a

b
0.10

dV/dW, cm3/g/nm

Volume, cm3/g, STP

700
600
500
400

Fe/SiO2

200

ier)
5 (h
SMZ
e
F
Fe-ZSM-5 (nano)

100

FeZSM-5 (conv)

300

0.0

0.2

0.4

0.6

P/P0

0.8

0.08
0.06
Fe-ZSM-5 (conv)

0.04

Fe-ZSM-5 (nano)
Fe/SiO2

0.02

FeZSM-5 (hier)

1.0

0.00
5

10

15

20

Pore width, nm

Fig. 4. N2 adsorption (solid symbols) and desorption (open symbols) isotherms at 77 K (a) and pore size distribution according to DFT method (b) for different Fe-containing
samples.


K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

185

Fig. 5. HRTEM images of the (a) Fe-ZSM-5 (hier) and (b) Fe/SiO2.

Fig. 6. UV–vis DR spectra of the Fe-containing samples: (1) Fe/SiO2, (2) Fe-ZSM-5 (conv-as), (3) Fe-ZSM-5 (conv), (4) Fe-ZSM-5 (hier), (5) Fe-ZSM-5 (nano), (6) Fe-ZSM-5
(nano-as).

troscopy data and white color of nonactivated samples Fe-ZSM-5
(conv-as) and Fe-ZSM-5 (nano-as) indicated that isolated Fe3+ ions
mainly occupy the tetrahedral framework positions. Zeolite samples Fe-ZSM-5 (conv-as) and Fe-ZSM-5 (nano-as) were activated
by oxalic acid treatment followed by drying and calcination. Less
than 3 wt.% of Fe was leached during zeolite activation according
ICP–OES. Activated samples became tan indicating the formation
of small iron oxide or hydroxide clusters. This fact was also confirmed by the UV–vis DR spectra – intensive band at
35,000 cm1 for treated samples can be seen (Fig. 6). This band
may be ascribed to the metal–oxygen charge transfer in the clusters of iron in octahedral oxygen coordination. The sample FeZSM-5 (nano-as) also contains such clusters but in the fewer
quantity than in the treated Fe-ZSM-5 (nano). Crystallization of
zeolite Fe-ZSM-5 (hier) in the presence of PS templates resulted
in significant modifications of UV–vis DR spectra, which can be

described by the superposition of absorption of ferric ions in
small oxide clusters and iron in large oxide aggregates on the
surface of amorphous silica [27]. According to HRTEM analyses
all active samples contain uniformly distributed iron oxide clusters of 2–3 nm (Fig. 7). Elemental EDX analyses showed that
35 atomic% of iron in Fe-ZSM-5 (hier) was located in zeolite
crystals; the rest one was in amorphous silica phase.
Experimental determination of the phase composition of ironcontaining particles is complicated by the fact that their relative
content in zeolite is small and they have a very small size. The
nature of iron-containing particles in the samples can be assumed
theoretically from thermodynamic probability of the existence of
known iron containing oxide phases, they may be hematite aFe2O3, magnetite Fe3O4, crystal FeOOH or amorphous Fe(OH)3. In
presence of 1 M hydrogen peroxide equilibrium existence of
magnetite is not probable: free standard Gibbs energy of the
reaction

2Fe3 O4solid þ H2 O2solute () 3a  Fe2 O3solid þ H2 Oliquid

is 312.73 kJ mol1. Because of negative standard entropy change
of the reaction (69.85 J mol1 K1) phase transformation of hematite to magnetite is less probable during thermal treatment of the
samples. Due to large negative standard free energy of transformation of amorphous Fe(OH)3 to hematite, presence of the first form
may be taken out of the consideration. Among crystal forms of FeOOH there are goethite a-FeOOH, akaganit b-FeOOH and lepidocrocite c-FeOOH. According to literature data c-FeOOH, being
produced at low concentrations of ferric aqua ions from ferric
hydroxide particles with a small molecular weight, and akaganit
b-FeOOH, being formed in the presence of chloride ions, easily
transform to goethite or hematite [28]. During the aging of ferric
hydroxide polymers in alkaline medium goethite a-FeOOH preferentially is produced, while hematite is formed in acidic medium.
Thus the samples just after the hydrothermal treatment most
probably contain iron in the form of goethite. However as a result
of subsequent heat treatment of the sample, phase transformation
of goethite to hematite may occur. The known temperature of

goethite to hematite transformation is 136 °C, but if particle size
is several nanometers it may reach 700 °C. Thus a stabilization of
iron oxide species in the form of goethite in all samples is most
probable, but the existence of hematite cannot be denied.
Assuming that the most iron containing particles is goethite or
hematite, it is possible to estimate the possible amount of iron particles with diameter of 3 nm in the zeolite particle of 3 microns.
The ideal cell of calcined ZSM-5 (silicalite-1) is described by the
formula Si96xFexO192 [29]. At the iron content in the zeolite of
2 wt.%, two atoms are located in a single cell: x = 2. If the mean size
of a zeolite particle is 3 lm then the volume of this particle is
1.4  1011 cm3 (assuming a spherical form of the particle),
whereas the volume of a cell is 5.4  1021 cm3 [29]. Then zeolite
particle contains 2.6  109 cells and 5  109 iron atoms. If we assume that the volume of an iron containing particle with the size
of 3 nm is 1.4  1020 cm3, and its density is 4–5 g cm3, then each
iron particle contains 300–400 iron atoms. Thus there may be


186

K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

Fig. 7. HRTEM images of (a) Fe-ZSM-5 (conv), (b) Fe-ZSM-5 (nano), (c) Fe-ZSM-5 (hier) and (d) Fe/SiO2.

1.4  107 of nanosized iron-containing particles in primary zeolite
particle with a size of 3 lm (if all iron in zeolite is in oxidic clusters,
but really it is ca. 30 at.%, the rest iron is in the form of isolated ferric ions in tetrahedral oxygen coordination). Note that iron containing particles in samples occupy from 0.1% to 4% of total pore
volume of amorphous Fe/SiO2 and Fe-ZSM-5 (conv), respectively.
Samples obtained were tested in hydrogen peroxide decomposition as since this reaction is the main source of oxidative activity
of Fenton reagent in organic substrates oxidation. Comparative catalytic experiments were carried out in presence of EDTA anions,
which form strong complexes with iron ions, to test catalytic

stability of iron species supported on zeolite and amorphous samples. One can see H2O2 decomposition kinetic curves and values of
initial hydrogen peroxide decomposition rates in Fig. 8 and Table 2,
respectively. Catalytic activity of zeolites Fe-ZSM-5 (conv) and
Fe-ZSM-5 (nano) is lower than that of homogeneous Fenton
system, the fact was also observed in [3]. For samples, containing

Fig. 8. H2O2 decomposition kinetics at [H2O2]0 = 1.1 M in absence (open symbols)
and presence (solid symbols) of Na2EDTA in Fe-containing systems. The catalyst
concentration was 5 mM and 20 g L1 for homogeneous and heterogeneous
systems, respectively.

amorphous silica phase, Fe/SiO2 and hierarchical Fe-ZSM-5 (hier),
catalytic activity is significantly lower compared with highly crystallized zeolites. This fact seems to be a result of wide size distribution of iron containing particles unlike that one’s with specified
size of 2–4 nm in the zeolite. However specific structure of these
particles inside zeolite mesopores also should be taken into
account and requires an additional study. Values of catalytic activity in decomposition of 1.1 M H2O2 at 25 °C for Fe(NO3)3, Fe-ZSM-5
(conv), Fe-ZSM-5 (nano), Fe-ZSM-5 (hier) and Fe/SiO2 are 17, 6.4,
9.0, 0.6 and 0.2 mmol H2O2 min1 g (Fe)1. It is worth noting that
pure zeolites show also the remarkable stability in presence of
EDTA anions as no induction period indicating complex formation
is observed for Fe-ZSM-5 (conv) and Fe-ZSM-5 (nano) samples
despite of high EDTA tendency to form complexes with iron: logarithm of the stability constant is 14.3 and 25.1 for 1:1 complex of
EDTA with Fe(II) and Fe(III), respectively. Initial H2O2 decomposition rate for Fe/SiO2 in presence of EDTA is observed to be higher
than without it. After 75 min the reaction accelerates and H2O2
decomposition rate becomes close to the one in the homogeneous
system. This means that iron species supported on the amorphous
silica are not stable and form complexes with EDTA which are subsequently leached. When EDTA is oxidized by H2O2 iron in form of
hydroxides adsorbs on the surface of the catalyst since pH = 7. This
may explain why iron concentration in solutions after the reactions
was less than 5  105 M according to ICP–OES. The kinetic curve

behavior for hierarchical zeolite is similar to the amorphous sample, however longer induction period is observed, perhaps due to
presence of different iron species supported on the amorphous silica and enclosed in zeolite pores. Such low catalyst stability can be
related with the location of iron species on the external surface of
amorphous particles or in cylindrical mesopores which do not limit
the diffusion of EDTA. On the contrary iron species included in the
zeolite are likely to be mainly distributed inside crystals in the inkbottle mesopores forming the ‘‘ship in a bottle’’ catalytic sites. This
type of a structure results in spatial inaccessibility of catalytic sites
for EDTA anions. Thus the zeolite microporous structure encourages high activity and stability of catalytic sites during Fenton
reaction.


187

K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188
Table 2
Initial hydrogen peroxide decomposition rate, Pa s1in absence and presence of iron complexing agent for Fenton-type systems at 25 °C, [H2O2]0 = 1.1 M.

W O2 , [Na2EDTA] = 0
W O2 , [Na2EDTA] = 1 g/L

Fe(NO3)3

Fe-ZSM-5
(conv-as)

Fe-ZSM-5
(conv)

Fe-ZSM-5
(nano-as)

Fe-ZSM-5
(nano)

Fe-ZSM-5
(hier)

Fe/SiO2

14.4
0.16 (11.3 after 90 min)

0.63
–

10.77
9.69

6.11
5.75

9.39
6.66

0.46
0.34 (1.7 after 115 min)

0.26
0.41 (15.3 after 75 min)

Again theoretical thermodynamics estimation may support this
version. Let us consider the dissolution reaction of iron containing
particles without any supporting silicate matrix in pure water and
in presence of hydrogen peroxide. Dissolution reactions for goethite and hematite are as follows:

FeOOHsolid ỵ 3Hỵaqua () Fe3ỵ
aqua þ 2H2 Oliquid ;
Fe2 O3solid þ 6Hþaqua () 2Fe3þ
aqua þ 3H2 Oliquid :
The standard Gibbs energy change for these reactions with massive
solid phases is 0.2 and 0.4 kJ mol1, respectively [30]. In case of
nanometer spherical particle dissolving DG0 decreases on the value

 solid
2rV
;
r
 – is a molar
where r – is a surface tension, r – is a particle radius, V
volume of the solid phase. If assuming that r  0.6 J m2 and
r = 1.5 nm, than this reduction is 16.7 and 24.4 kJ mol1 for goethite
and hematite particles, respectively. Thus DG0 is 16.5 and
24 kJ mol1 for dissolution of dispersed goethite and hematite
phases, respectively. From the equation of isotherm of the
reactions:

a 3ỵ
Dr G 0
ỵ ln Fe3 6 0;
RT

aHỵ
2

a 3ỵ
Dr G 0
ỵ ln Fe6 6 0:
RT
aHỵ
follows that at 25 C the equilibrium activity of aqua iron ions
depends on the acidity as follows:

lgaFe3ỵ 6 2:9  3pH;
lgaFe3ỵ 6 2:1  3pH:
This means that in pure water only at pH < 1 the iron leaching
from the particles to the solution should be expected.
But the situation changes dramatically when the dissolution
occurs in presence of hydrogen peroxide. Let us assume that
iron-containing particles exist in the form of Fe2O3 and consider
the reaction:

Fe2 O3solid ỵ 2H2 O2soluted ỵ 4Hỵqaua

() 2Fe2ỵ
aqua

ỵ 3H2 Oliquid ỵ 2HO2aqua :

The standard Gibbs energy change for this reaction is
39.5 kJ mol1. In the case of nanosized particle dissolving the
Gibbs energy of the reaction will be less on 24.4 kJ mol1 as

described above and it is 63.9 kJ mol1. If one takes that
aH2 O2 = 1 M and aHO2 = 108 M then

lgaFe2ỵ ¼ 13:6  2pH:
Thus, the estimation means that at pH < 8 the total reductive
dissolution of nanosized iron containing particles must occur at
experimental conditions used in the work. Undoubtedly the iron
ion hydrolysis reaction should be taken into account together with
the reductive dissolution reaction, but if there are iron complexing
agents in the solution, the iron leaching from the solid particles
should be expected to prevail. Nevertheless no any significant iron

leaching and no any consecutive deactivation are observed for zeolite samples unlike Fe/SiO2 and Fe-ZSM-5 (hier), containing amorphous silica phase. This fact may indicate a key role of microporous
structure of zeolite on the high activity and stability of the heterogeneous Fenton catalyst. Catalytic site protection by zeolitic matrix
could be potentially explored for development of technology for
purification of waste water from nuclear power plants. The problem is in large amount of wastes containing radionuclides, for
example 60Co, which are enclosed in soluble complexes including
EDTA ones that pass through filters and cannot be separated from
the solution without a special pretreatment. Wet hydrogen peroxide oxidation of complexing agents by H2O2 using Fe-containing
zeolites has a good potential for this application.
4. Conclusions
Four types of iron containing materials have been synthesized
and studied: conventional zeolite Fe-ZSM-5 (conv), hierarchical
zeolite Fe-ZSM-5 (hier), small crystals (d = 330 nm) of zeolite FeZSM-5 (nano) and ferric oxide species supported on the amorphous
silica. Catalytic activity of ferric species in hydrogen peroxide
decomposition for small 330-nm crystals of Fe-ZSM-5 was
1.4 times higher than for large zeolite crystals, and significant
decrease of the activity was observed for samples containing amorphous silica phase. The experimental results show that ferric sites
composed of zeolite are stable due to the limited diffusion of Na2EDTA in microporous zeolitic phase. Ferric species supported on
the amorphous silica are extremely unstable during wet peroxide

oxidation reactions and undergo leaching to the solution.
Acknowledgements
The authors thank A.B. Ayupov, S.V. Bogdanov, E.Yu. Gerasimov
for their help in the catalyst characterization. Financial supports by
Department of Science and Education (projects Nos.
14.512.12.0005 and 8440), Russian Federation President Grant for
the Leading Scientific Schools #NSh 524.2012.3, integration
projects Nos. 35 and 24 are gratefully acknowledged.
References
[1] M.B. Kasiri, H. Aleboyeh, A. Aleboyeh, Appl. Catal. B Environ. 84 (1–2) (2008) 9–
15.
[2] A. Zecchina, M. Rivallan, G. Berlier, C. Lamberti, G. Ricchiardi, Phys. Chem.
Chem. Phys. 9 (27) (2007) 3483–3499.
[3] E.V. Kuznetsova, E.N. Savinov, L.A. Vostrikova, V.N. Parmon, Appl. Catal. B
Environ. 51 (3) (2004) 165–170.
[4] L.F. Liotta, M. Gruttadauria, G.Di. Carlo, G. Perrini, V. Librando, J. Hazard. Mater.
162 (2009) 588–606.
[5] O.A. Makhotkina, E.V. Kuznetsova, L.G. Matvienko, V.N. Parmon, Catal. Ind. 4
(2006) 30–37.
[6] O.A. Makhotkina, S.V. Preis, E.V. Parkhomchuk, Appl. Catal. B Environ. 84
(2008) 821–826.
[7] G. Soler-Illi, C. Sanchez, B. Lebeau, J. Patarin, Chem. Rev. 102 (2002) 4093–
4138.
[8] A. Corma, Chem. Rev. 97 (1997) 2373–2419.
[9] Y. Tao, H. Kanoh, L. Abrams, K. Kaneko, Chem. Rev. 106 (2006) 896–910.
[10] C.S. Cundy, P.A. Cox, Microporous Mesoporous Mater. 82 (2005) 1–78.
[11] J. Perez-Ramırez, C.H. Christensen, K. Egeblad, Ch.H. Christensen, Chem. Soc.
Rev. 37 (2008) 2530–2542.
[12] A. Stein, Microporous Mesoporous Mater. 44–45 (2001) 227–239.

188

K.A. Sashkina et al. / Microporous and Mesoporous Materials 189 (2014) 181–188

[13] K.A. Sashkina, V.S. Labko, N.A. Rudina, V.N. Parmon, E.V. Parkhomchuk, J. Catal.
299 (2013) 44–52.
[14] G.I. Panov, A.K. Uriarte, M.A. Rodkin, V.I. Sobolev, Catal. Today 41 (1998) 365–
385.
[15] A.J.J. Koekkoek, H. Xin, Q. Yang, C. Li, E.J.M. Hensen, Microporous Mesoporous
Mater. 145 (1–3) (2011) 172–181.
[16] El.-M. El-Malki, R.A. Santen, W.M.H. Sachtler, J. Catal. 196 (2000) 212–223.
[17] P. Sazama, N.K. Sathu, E. Tabor, B. Wichterlová, Š. Sklenák, Z. Sobalík, J. Catal.
299 (2013) 188–203.
[18] A. Ates, Ch. Hardacre, A. Goguet, Appl. Catal. A 441–442 (2012) 30–41.
[19] A.R. Overweg, M.W.J. Crajé, A.M. Kraan, I.W.C.E. Arends, A. Ribera, R.A. Sheldon,
J. Catal. 223 (2004) 262–270.
[20] J. Pérez-Ramírez, F. Kapteijn, G. Mul, J.A. Moulijn, Catal. Commun. 3 (2002) 19–
23.
[21] E.V. Parkhomchuk, K.A. Sashkina, N.A. Rudina, N.A. Kulikovskaya, V.N. Parmon,
Catal. Ind. 4 (2012) 23–32.

[22] D. Zou, S. Ma, R. Guan, M. Park, L. Sun, J.J. Aklonis, R.J. Salovey, Polym. Sci. A 30
(1992) 137–144.
[23] B.T. Holland, C.F. Blanford, A. Stein, Science 281 (1998) 538–540.
[24] M.M.J. Treacy, J.B. Higgins, R. Ballmoos, Collection of Simulated XRD Powder
Patterns for Zeolites, third ed., The Commission of the International Zeolite
Association, 1996.
[25] P. Ratnasami, R. Kumar, Catal. Today 9 (4) (1991) 357.
[26] E.J.M. Hensen, Q. Zhu, M.M.R.M. Hendrix, A.R. Overweg, P.J. Kooyman, M.V.

Sychev, R.A. Santen, J. Catal. 221 (2004) 560.
[27] A. Ribera, I.W.C.E. Arends, J. Catal. 195 (2000) 287.
[28] Ch. Flynn, Chem. Rev. 84 (1984) 31–41.
[29] Ch. Baerlocher, W.M. Meier, D.H. Olson, Atlas of Zeolite Framework Types,
Elsevier, 2001.
[30] I. Barin, Thermochemical Data of Pure Substances, third ed., VCH
Verlagsgesellschaft mbH, D-6945 1 Weinheim (Federal Republic of
Germany), 1995.