Stability of hydrogen peroxide during perhydrolysis of carboxylic acids on acidic heterogeneous catalysts

Article abstract of DOI:10.1007/s11164-010-0149-y

This paper describes a study of the stability of hydrogen peroxide in the presence of different aluminosilicate materials, in connection with an investigation of carboxylic acid perhydrolysis. During the reaction, aluminosilicate materials such as H-β zeolites, mesoporous material H-MCM-41 and alumina initiate the decomposition of hydrogen peroxide. The reason of the spontaneous decomposition of H₂O₂ is related to the partial dealumination of these zeolites. However, in the case of experiments carried out with H-ZSM-5 zeolite catalysts, a slight catalytic effect on the perhydrolysis and no spontaneous decomposition of hydrogen peroxide were noticed. The use of cation exchange resins as catalysts is more kinetically beneficial than H-ZSM-5 zeolite catalysts. Springer Science+Business Media B.V. 2010.

Full text of DOI:10.1007/s11164-010-0149-y

Res Chem Intermed (2010) 36:389–401
DOI 10.1007/s11164-010-0149-y
Stability of hydrogen peroxide during perhydrolysis
of carboxylic acids on acidic heterogeneous catalysts
S e´ bastien Leveneur
•
Narendra Kumar
•
Tapio Salmi Dmitry Yu. Murzin
•
Received: 16 August 2009 / Accepted: 7 January 2010 / Published online: 12 June 2010
Ó Springer Science+Business Media B.V. 2010
Abstract This paper describes a study of the stability of hydrogen peroxide in the
presence of different aluminosilicate materials, in connection with an investigation
of carboxylic acid perhydrolysis. During the reaction, aluminosilicate materials such
as H-b zeolites, mesoporous material H-MCM-41 and alumina initiate the decom-
position of hydrogen peroxide. The reason of the spontaneous decomposition of
H O is related to the partial dealumination of these zeolites. However, in the case
2
2
of experiments carried out with H-ZSM-5 zeolite catalysts, a slight catalytic effect
on the perhydrolysis and no spontaneous decomposition of hydrogen peroxide were
noticed. The use of cation exchange resins as catalysts is more kinetically beneficial
than H-ZSM-5 zeolite catalysts.
Keywords Perhydrolysis ꢀ Aluminosilicate materials ꢀ Decomposition ꢀ
Hydrogen peroxide
Introduction
Interest in hydrogen peroxide has increased in the last decade. For instance, the
world production of hydrogen peroxide was around 1.9 million tonnes in 1994, and
grew to 2.2 million tonnes in 2006 [1]. The applications of this compound are
versatile: it is used as an oxidizing agent for inorganic and organic compounds, a
S. Leveneur (&) ꢀ N. Kumar ꢀ T. Salmi ꢀ D. Yu. Murzin
Laboratory of Industrial Chemistry and Reaction Engineering, Process Chemistry Centre,
˚
˚
Abo Akademi University, Biskopsgatan 8, 20500 Abo/Turku, Finland
e-mail: sleveneu@abo.fi
S. Leveneur
LSPC-Laboratoire de S e´ curit e´ des Proc e´ d e´ s Chimiques, INSA Rouen, Place Emile Blondel,
BP8, 76131 Mont-Saint-Aignan Cedex, France
123

390
S. Leveneur et al.
bleaching agent for textiles and wood, as an antiseptic for therapeutic use, as a
disinfectant for wastewater and as a sterilizing agent.
The use of H O as an oxidizing agent [2] in chemistry avoids the use of
2
2
conventional heavy-metal oxidants which form toxic waste; application of nitric
acid which forms the greenhouse gas N O; and utilization of molecular oxygen,
2
which requires safety precautions. Scheme 1 illustrates the application of hydrogen
peroxide in some fine chemical synthesis reactions [3].
From an environmental and economical point of view, these reactions should be
carried out with heterogeneous catalysts, which should not decompose hydrogen
peroxide.
The aim of this paper is to treat the stability issues of H O in an acidic
2
2
environment over different heterogeneous catalysts, and especially aluminosilicate
materials. Traditionally, synthetic aluminosilicate materials, and more particularly
synthetic zeolites, are widely used as catalysts in the petrochemical industry, for
instance in fluid catalytic cracking and hydro-cracking. The use of zeolites for fine
chemicals production is widespread: acylation, alkylation, hydroxyalkylation of
aromatic and heterocyclic compounds, and rearrangement reactions [4]. Selectivity
(
including shape selectivity), catalytic effects and environmental regulations explain
the application of such materials in industry. In particular, oxidation reactions such
as epoxidation of olefins, hydroxylation of arenes, and oxidation of O-, S- and
N-functionalities with hydrogen peroxide are conducted using Ti-zeolites. The
originality of this paper is the investigation of an oxidation reaction with H O in
2
2
the presence of titanium-free aluminosilicate materials.
To illustrate the stability of H O , the peroxycarboxylic acids synthesis was
selected as a model reaction (Scheme 2).
2
2
2 2
Scheme 1 Applications of H O in fine chemical reactions
1
23

Stability of hydrogen peroxide during perhydrolysis
391
Scheme 2 Synthesis of PCA from the parent carboxylic acid and hydrogen peroxide
This reaction occurs at acidic pH [5–7], and is traditionally catalyzed by sulfuric
acid. This reaction was selected because spontaneous decomposition of hydrogen
peroxide or peroxycarboxylic acids is negligible [8–10] in traditional experimental
conditions (pH: 1–3; and temperature: 20–50 °C); and configurational diffusion can
be neglected because the size of the reactant molecules is smaller compared to the
pore size of the materials used (Table 1). Thus, by using heterogeneous catalysts for
this reaction, only decomposition due to the solid acid catalyst can occur.
In a previous paper, we have demonstrated that cation-exchange resins [11] are
suitable heterogeneous catalysts for the carboxylic acids perhydrolysis. Our
objective in the present study is to elucidate hydrogen peroxide stability towards
aluminosilicate type materials, and to compare the catalytic efficiency of these
aluminosilicate materials versus cation-exchange resins in such reactions.
Except for the communication of Palani et al. [12], in which the efficacy of
different zeolites and mesoporous materials was investigated, there are no other
scientific papers dealing with perhydrolysis of carboxylic acids over these materials.
Experimental section
Apparatus and experimental procedures
All experiments were performed in a 250-mL jacketed glass reactor vessel. The
reactor was equipped with a mechanical agitator and a temperature probe. Water
was pumped through the outer jacket of the vessel to control the temperature of the
reaction mixture. A pitched blade impeller (PTFE coated) was used to ensure a
vigorous mixing during the reaction.
A reflux condenser was attached to the top of the reactor (adjusted at 0 °C) to
avoid loss of volatile liquid-phase compounds. In case decomposition of peroxy-
propionic acid (PPA) or H O appeared, an inert carrier gas (Helium, AGA,
2
2
Table 1 Channel structure of the tested aluminosilicate materials
Aluminosilicate
materials
Channel structure
˚
3-Dimensional pore system; 12-ring channel in c direction with pores 7.6 9 6.4 A
b
plus two 12-ring channels in a direction perpendicular to c-direction with pores
7
˚
.6 9 6.4 A and 5.5 9 5.5 A
˚
˚
3-Dimensional pore system; straight 10 member-ring 5.2 9 5.7 A channels
ZSM-5
˚
connected by sinusoidal 5.3 9 5.6 A channels. Intersection cavity: 9 A
˚
MCM-41
1-Dimensional hexagonal arrangement of uniformly open channels from 15 to
˚
1
00 A
123

392
S. Leveneur et al.
99.996%) was introduced to the reactor through one of the necks to avoid the
accumulation of oxygen in the gas phase.
To prevent contamination induced by alkaline and metal components, which
initiate the catalytic decomposition of peroxypropionic acid and hydrogen peroxide,
all parts of the reactor system in contact with the reaction solution were washed with
hydrochloric acid, followed by another washing with a phosphate-free detergent
solution.
In the experiments carried out for testing the stability of hydrogen peroxide
towards zeolite, at the first stage, deionized water and zeolites were mixed together
in the reactor. As the desired reaction temperature was reached, the preheated
hydrogen peroxide solution (Merck, 30 wt.%) was added through a dropping funnel.
In fact, hydrogen peroxide was preheated at the same temperature as the reaction
temperature. At the time ‘‘zero’’, the required amount was introduced into the
reactor.
In the experiments carried out for checking the catalytic effect of zeolites on the
propionic acid perhydrolysis reaction, at the first stage, propionic acid (Acros,
99 wt.%) and the catalyst were mixed together in the reactor. Thereafter, the same
protocol as described above was followed.
According to a patent [13], during the reaction of hydrogen peroxide and
propionic acid in water in the presence of an acid catalyst, the danger of explosion is
suppressed by employing a temperature of maximum 60 °C.
From our previous study [9, 11], the kinetics of propionic acid perhydrolysis is
demonstrated to be dependant on the Brønsted acid sites in the cases of
homogeneous and heterogeneous catalysis. One should keep in mind that the goal
is to check the stability of hydrogen peroxide towards aluminosilicate materials in
an acidic environment by studying the reaction of perhydrolysis. Due to the low
Brønsted acid site amount of these catalysts and the difficulty of measuring them in
the temperature range 40–50 °C, the kinetic comparisons were based on catalyst
loading. Table 2 introduces the experimental matrix.
Analytical methods
Samples were withdrawn from the reaction mixture by a plastic syringe (to avoid
contamination of the solution by trace of metals) and they were analyzed by the
Greenspan and Mackellar method [14]. The concentration of hydrogen peroxide was
determined by titration by using a standard solution of ammonium cerium sulphate
Table 2 Experimental
parameters for the catalyst
comparison in the batch reactor
Reaction temperature (°C)
Rotating speed (rpm)
40–50
250–350
-
1
)
H
2
O
2
(mol L
5.75–6.10
5.70–580
23.85–24.10
8.26–43.40
-1
PA (mol L
)
-1
H
2
O (mol L
)
-1
Catalysts loading (g L
)
1
23

Stability of hydrogen peroxide during perhydrolysis
393
(
0.1 N). The concentrations of propionic and peroxypropionic acids were
determined by titration with an automatic titrator (Metrohm 751 GPD Titrino) by
using a standard solution of sodium hydroxide (0.2 N).
Catalyst properties
The acidic properties of the different aluminosilicate materials are summarized in
Table 3. The Brønsted and Lewis acid sites were determined by FTIR using
pyridine as probe molecule [15–18]. The specific surface area was determined by N2
adsorption, calculated by the Dubinin method for the microporous materials and by
the BET method for the mesoporous H-MCM-41 [15–18].
A comparison between a cation exchange resin, i.e., Amberlite IR-120 and
zeolites, was carried out. The characteristics of this commercial resin are
summarized in Table 4.
The pretreatment of Amberlite IR-120 consisted of a drying step at 70 °C for
8 h, since a higher drying temperature might affect the sulfonic acid sites on the
4
catalyst. The parameter capacity by dry weight represents the number of Brønsted
groups bearing (i.e., sulfonic group) by the resins, which is equivalent to the
Brønsted acid sites of zeolites displayed in Table 3.
To dealuminate zeolite, the protocol of Marques et al. [19] was applied to H-b-25
zeolite, i.e., an acid treatment at 30 °C by HCl (1 N) for 10 min was performed.
The Al concentration in the reaction mixture was measured by ICP-OES
(
PerkinElmer Optima 5300DV).
Table 3 Properties of the aluminosilicate materials
SiO /Al
mol/mol
2
2
O
3
Brønsted acid sites
at 523 k lmol/g
Lewis acid sites
at 523 k lmol/g
Specific surface
2
area m /g
H-b-25
25
75
300
23
31
–
269
147
82
416
–
162
39
30
36
–
807
664
805
443
–
H-b-75
H-b-300
H-ZSM5-23
H-ZSM5-31
2
Al O
3
7
156
40
299
1242
H-MCM-41
40
26
Table 4 Properties of
Amberlite IR-120
Supplier
Aldrich
Polymer type
Gel
8
Cross-linking (%)
Moisture content (% mass)
45
Capacity by dry weight (meq/g)
Native particle size range (mm)
4.4
0.3–1.2
123

394
S. Leveneur et al.
Results and discussion
Experiments of the perhydrolysis of propionic acid over different aluminosilicate
materials gave two antagonist observations: either the catalyst was active in
perhydrolysis, or no formation of peroxypropionic acid was observed due to the
decomposition of hydrogen peroxide over these acidic materials.
Aluminosilicate materials initiating the decomposition of hydrogen peroxide
During the reaction of perhydrolysis, decomposition of hydrogen peroxide was
observed with the following materials: H-b zeolites, H-MCM-41 and alumina. For
½
H
2
O
2ꢁ
the sake of simplicity, the ratio
, where [H O ] is the initial concentration,
2 2 0
½
H
2
O
2ꢁ
0
was plotted versus the reaction time. An experiment with only hydrogen peroxide
without any catalyst was carried out at 50 °C; no spontaneous decomposition of this
reactant was observed.
According to Sengupta et al. [20], the decomposition of hydrogen peroxide on
un-impregnated silica-alumina material is dependent of the strength of acid sites. In
our previous studies [9, 11] no decomposition of hydrogen peroxide was noticed due
to Brønsted acidity in either sulfuric acid or sulfonic groups bearing by the resins.
For that reason Lewis acid sites in such materials can be considered to be
responsible for the decomposition of hydrogen peroxide.
However, Rocha et al. [21] have studied the kinetics of peroxyacetic acid
formation from acetic acid and hydrogen peroxide in the presence of scandium (III)
triflate. They have demonstrated that the use of this strong Lewis acid enhanced the
reaction rate sixfold compared to the catalyst absence; and no decomposition of
hydrogen peroxide was reported.
Several articles treat the dealumination of H-b zeolites [22, 23], or type A and Y
zeolites [24] by acidic leaching. Marques et al. [19] have shown that treatment of
HBEA zeolite (Si/Al = 12.5) with an HCl solution at 30 °C causes a rapid
dissolution of the extraframework Al species, especially the monomeric ones and a
slow dissolution of the framework Al atoms bridging OH groups and structure
defects. These Al species released in the reaction mixture could catalyze the H O
2
2
decomposition.
In this section, the reason of hydrogen peroxide decomposition will be discussed
regarding the Lewis acid, partial dealumination and SiO /Al O ratio.
2
2 3
Figure 1 displays three different experiments carried out with H-b-25 zeolites.
The difference of pH is due to the presence of PA, which acidifies the mixture. In
experiments carried out with a mixture PA and H O , the concentration of PA
2
2
remains constant. Even if the decomposition of H O is strongly dependent on the
2
2
pH, one can conclude that the production of PPA or the presence of PA does not
influence the kinetic of decomposition. From Fig. 1, the kinetics of hydrogen
peroxide decomposition for experiments carried out with PA and H O mixture in
2
2
the presence of H-b-25 and dealuminated H-b-25 by acidic treatment is similar.
This observation implies that the decomposition of hydrogen peroxide is due to Al
species, in particular dissolution of Al atoms from the framework of the zeolite. The
difference of aluminium concentration between the end and the beginning of the
1
23

Stability of hydrogen peroxide during perhydrolysis
395
1
1
0
0
0
0
0
0
0
.02
.00
.98
.96
.94
.92
.90
.88
.86
H-β -25: Only H2O2; pH : 3.53
H-β -25: PA+H2O2; pH : 1.83
Dealuminated H-β -25 by acid treatment:
PA+H2O2; pH : 1.93
0
50
100
150
200
250
300
350
400
Time [min]
Fig. 1 Experiments carried out in the presence of H-b-25 zeolite at 50 °C, at 350g, and loading of
8.25 g/L
1
experiment carried out with H-b-25 and PA and H O mixture was measured to be
2
2
-
1
equal to 181 mg L , which can explain the decomposition of hydrogen peroxide.
From Fig. 1, one can notice the presence of an initial retarding period, which is
more significant in the case of the experiment carried out with H-b-25 and PA and
H O mixture. The presence of the initial retarding effect, occurring for a long time
2
2
period, during the decomposition of hydrogen peroxide is not the first observation in
the literature. For instance, this phenomenon of retarding phase was noticed in the
case of H O decomposition by Fe(III). De Laat et al. [25] explained it by the
2
2
.
formation of OH radicals via Fenton’s reaction, with the duration of this phase
depending on various parameters, such as the initial concentration of Fe(III). In the
present case, this retarding period could be explained by the dissolution of Al.
Indeed, in the case of experiments carried out over dealuminated H- b-25, this
retarding period seems to be shorter, because Al atoms from the extra-framework
have already been removed by acidic treatment.
To check if the decomposition of hydrogen peroxide in such conditions is due to
the partial dissolution of Al species in the liquid phase, an experiment with
aluminium oxide was carried out (Fig. 2). Alumina was chosen instead of an
aluminium salt because its structure is closer to Al species present in the zeolites.
Figure 2 shows that hydrogen peroxide decomposes over Al O , which implies
2
3
that the assumption of H O decomposition is related to the partial dissolution of Al
2
2
species is plausible. The ratio [PA]/[PA] remained constant during this experiment.
0
To check if the decomposition of hydrogen peroxide is related to the ratio SiO2/
Al O , H-b zeolites with three different SiO /Al O ratios were used as shown in
2
3
2
2 3
Fig. 3. No production of PPA was noticed, but instead, decomposition of hydrogen
peroxide was observed.
During these experiments, the concentration of PA remained constant (Fig. 3).
The kinetic curves are similar, but there is an initial retarding effect during the
experiments carried out with H-b-75 and H-b-25. No decomposition of hydrogen
123

396
S. Leveneur et al.
1
1
0
0
0
0
0
0
0
.02
.00
.98
.96
.94
.92
.90
.88
.86
pH: 2.15
Alumina
0
50
100
150
200
250
300
350
400
450
Time [min]
Fig. 2 Experiment carried out with a mixture of PA and H
loading of 18.25 g/L
2
O
2
over alumina at 50 °C, at 350g, and
1
1
0
0
0
0
0
0
0
.02
.00
.98
.96
.94
.92
.90
.88
.86
pH: 1.7-1.8
H-β -75
H-β -25
H-β -300
0
50
100
150
200
250
300
350
400
Time [min]
2 2
Fig. 3 Experiments carried out with a mixture of PA and H O over H-b zeolites at 50 °C, at 350g, and
loading of 18.25 g/L
peroxide was reported in the paper of Palani et al. [12] during the synthesis of
peroxyacetic acid using H-b with a ratio Si/Al equal to 12, catalyst loading lower
than 1 g/l and at 25 °C. From Fig. 3, it can be concluded that the decomposition is
not directly related to the SiO /Al O ratio.
2
2 3
Although the amount of Lewis acid sites is 5 times higher in the case of H-b-25
than H-b-75 or H-b-300 (Table 3), the kinetics of decomposition is similar. Thus,
the Lewis acid sites bearing on these materials are not responsible for the H O
2
2
1
23

Stability of hydrogen peroxide during perhydrolysis
397
decomposition. Furthermore, the amount of Lewis acid sites in H-ZSM-5-23, which
does not induce the decomposition of H O , is of the same order of magnitude as in
2
2
H-b-75, H-b-300 or H-MCM-41 (Table 3). Thus, the assumption that the
decomposition of hydrogen peroxide is due to the presence of Lewis acid sites in
aluminosilicate materials can be rejected.
During the experiments, the decomposition of H O appeared only at zeolite
2
2
loadings exceeding 8.26 g/L, which explains the absence of decomposition in the
study of Palani et al. [12]. Below this catalyst loading, the perhydrolysis of
propionic acid due to the protolysis of the carboxylic acids occurs [7]. The change
of the catalyst loading in the range 10.24–18.25 g/L does not influence the kinetics
of H O decomposition at 50 °C (Fig. 4). In both cases, there is a retarding effect;
2
2
however, this induction period is longer for the experiment carried out with a
loading of 18.25 g/L.
A mesoporous material, H-MCM-41, was used to study its behavior in the
mixture of PA and H O . Figure 5 displays the kinetics of H O decomposition
2
2
2 2
versus time for this mixture over H-MCM-41 and H-b-75 at 40 °C. During these
reactions, the concentration of PA remains constant. From Fig. 5, one can notice
the same trend as with alumina or H-b materials, i.e., a decomposition phase of
hydrogen peroxide. The retarding period is more pronounced at the beginning of
the reaction (until 300 min) for H-MCM-41 than for H-b-75, which can be
explained by the higher stability of the mesoporous material towards an acidic
environment. After the induction period, the kinetics of H O decomposition is
2
2
similar in both cases.
The decomposition of hydrogen peroxide is due to the presence of Al species
dissolved from the framework structure of the H-b zeolites and mesoporous material
H-MCM-41. The slow kinetics of H O is related to the fact that the dissolution of
2
2
Al species from the framework is a slow phenomenon [19]. Furthermore, the Al
1
1
0
0
0
0
0
0
.02
.00
.98
.96
.94
.92
.90
.88
pH: 1.7-1.8
H-β -75 at 18.25 g/l
H-β -75 at 10.24 g/l
0
50
100
150
200
250
300
350
400
Time [min]
Fig. 4 Experiments carried out with a mixture of PA and H
50g
2 2
O over H-b-75 zeolites at 50 °C and at
3
123

398
S. Leveneur et al.
1.03
0.98
0.93
0.88
0.83
0.78
0.73
0.68
H-β -75 at 17.54 g/l; pH : 1.9-2.0
H-MCM-41 at 23.77 g/l; pH : 1.7-1.8
0
200
400
600
800 1000 1200 1400 1600 1800 2000
Time [min]
Fig. 5 Experiments carried out with a mixture of PA and H2O2 over H-MCM-41 and H-b-75 at 40 °C
and at 350g
species released are in oxide or hydroxide form, and according to Mani et al. [26],
their catalytic activity towards hydrogen peroxide decomposition is minor.
Materials which catalyze the perhydrolysis reaction
In the case of experiments carried out with H-ZSM-5 or Amberlite IR-120, no
decomposition of hydrogen peroxide was noticed, while the catalytic effect on the
perhydrolysis of propionic acid was visible. The purpose of this section is to propose
an explanation for the selectivity of the aluminosilicate materials towards the
reaction between H O and PA, and to perform a comparative study of these
2
2
materials. This comparison was based on equal catalyst loadings, and the
experiments were carried out under similar conditions.
Figure 6 shows that H-ZSM-5 zeolite catalyzes the perhydrolysis of propionic
acid. Moreover, the activity of H-ZSM-5-23 is higher than H-ZSM-5-31, due to a
higher amount of Brønsted acid sites. In the case that no external catalyst is added,
some perhydrolysis takes place (Fig. 6). This is due to the fact that propionic acid
itself is a weak Brønsted acid having some catalytic effect.
Figure 6 demonstrates that cation exchange resins are more beneficial than
zeolites for the synthesis of peroxypropionic acid from hydrogen peroxide and
propionic acid. An estimation of the apparent second-order rate constants based on
the product [PA]*[H O ] is provided in Table 5. The apparent Brønsted acid
2
2
?
concentration [H ] is also displayed for each catalyst (except for H-ZM5-31), which
is the sum of the amount of Brønsted acid sites of the solid catalyst and the number
of hydroxonium ions released due to the protolysis of PA present per litre of
reaction mixture.
Figure 7 displays the rate constants versus the apparent Brønsted acid concen-
?
tration [H ] for each catalyst. As can be seen, the apparent rate constants are
?
linearly proportional to [H ], which confirms the fact that the reaction of
perhydrolysis is catalyzed by the Brønsted acid sites. Amberlite IR-120 catalyzes 4
1
23

Stability of hydrogen peroxide during perhydrolysis
399
5
4
4
3
3
2
2
1
1
0
0
.0
.5
.0
.5
.0
.5
.0
.5
.0
.5
.0
H-ZSM5-23
H-ZSM5-31
Amberlite IR-120
No catalyst
0
50
100
150
200
250
300
350
400
450
Time [min]
Fig. 6 Formation of PPA over different catalysts at 50 °C at 43.38 g/L of loading and 350g
Table 5 Apparent rate
?
constants and [H ]
-1
-1
s
?
Apparent [H ]
k(L mol
)
-1
(
mol L )
-
6
-2
H-ZSM-5-23
5.51 9 10
2.95 9 10
2.18 9 10
1.15 9 10
1.80 9 10
-6
-6
-6
H-ZSM-5-31
–
-1
-3
Amberlite IR-120
Without catalyst
1.91 9 10
5.64 9 10
-1
-1
k (l.mol .s )
2
2
1
1
5
.5E-05
.0E-05
.5E-05
.0E-05
.0E-06
y = 0.0001x + 2E-06
2
R
= 0.98
0.0E+00
0
0.05
0.1
0.15
0.2
0.25
+
-1
Apparent [H ] (mol.l )
?
Fig. 7 Rate constants versus apparent [H ]
123

400
S. Leveneur et al.
times faster the perhydrolysis of PA than H-ZSM-5-23, and for this reason cation
exchange resins are preferred in practice.
According to Lutz et al. [27], the dense-structured zeolites of types ZSM-5 (MFI)
are relatively stable under treatment by liquid water for 72 h up to 513 K, whereas
the open-structured zeolites of type beta (BEA) undergo strong decomposition in the
same range. This statement shows that H-b zeolites are more fragile than H-ZSM-5
zeolites regarding the dealumination. In other words, the dissolution of Al species is
easier in the case of H-b zeolites than H-ZSM-5 zeolites.
More generally, material with an MFI structure seems to be used preferentially in
the case of oxidation by hydrogen peroxide. For instance, titanium silicate zeolite
TS-1, which has an MFI structure, is a material specifically designed for industrial
oxidation reaction with hydrogen peroxide [28].
Conclusions
This paper describes a study of hydrogen peroxide stability over different
aluminosilicate materials in the perhydrolysis of propionic acid. Several experi-
ments were carried out in an isothermal batch reactor at temperatures of 40–50 °C,
equimolar concentrations of the reactants, and with hydrogen peroxide alone.
It was observed that some aluminosilicate materials (such as H-b, H-MCM-41 or
alumina oxide) decompose hydrogen peroxide, but others (such as H-ZSM-5)
catalyze the perhydrolysis reaction in a moderate way.
Experiments carried out with H-b zeolites showed that the decomposition of
hydrogen peroxide is not directly proportional to the ratio SiO /Al O or to the
2
2 3
Lewis acid site bearing by these materials. But it appears that the decomposition of
hydrogen peroxide over these catalysts is related to the partial dealumination of the
zeolites framework, releasing Al species in the midst which induce the decompo-
sition mechanism. This assumption was supported by the literature, and by the
observation that alumina oxide material decomposes hydrogen peroxide during the
propionic acid perhydrolysis reaction. A certain minimal amount of material (ca.
8
the range 10.24–18.25 g/L does not change the kinetics of H O decomposition.
.26 g/L) is needed to initiate the decomposition. The change of catalyst loading in
2
2
The retarding period preceding the H O decomposition is connected to the
2
2
partial dealumination period of the zeolites. The duration of this period depends on
several parameters, such as pH of the solution, catalyst loading, and the nature of the
zeolite structure.
H-ZSM-5 does not decompose hydrogen peroxide, and in the case of a high
number of Brønsted acid sites, a catalytic effect on the perhydrolysis of propionic
acid is observed. Amberlite IR-120 was, however, even more beneficial by 4 times
compared to H-ZSM-5-23.
˚
Acknowledgements The financial support from the Abo Akademi Forskningsinstitut and the Finnish
Graduate School in Chemical Engineering (GSCE) are gratefully acknowledged. This work is part of the
˚
activities at the Abo Akademi Process Chemistry Centre (PCC) within the Finnish Centre of Excellence
Programme (2006–2011) by the Academy of Finland.
1
23

Stability of hydrogen peroxide during perhydrolysis
401
References
1
2
3
. R. Hage, A. Lienke, Angew. Chem. Int. Ed. 45, 206 (2006)
. R. Noyori, Chem. Commun. 14, 1807 (2005)
. C.W. Jones, Application of Hydrogen Peroxide and Derivatives, 1st edn. (Royal Society of Chem-
istry, Cambridge, 1999), pp. 49–53
. M.G. Clerici, Top. Catal. 13, 373 (2000)
4
5
6
7
. L.V. Dul’neva, A.V. Moskvin, Russ. J. Gen. Chem. 75, 1125 (2005)
. X. Zhao, T. Zhang, Y. Zhou, D. Liu, J. Mol, A. Catal, Chemical 271, 246 (2007)
. S. Leveneur, T. Salmi, D.Yu. Murzin, L. Estel, J. W a¨ rn a˚ , N. Musakka, Ind. Eng. Chem. Res. 47, 656
(2008)
8
. N. Musakka, T. Salmi, J. W a¨ rn a˚ , J. Ahlkvist, M. Piironen, Chem. Eng. Sci. 61, 6918 (2006)
. S. Leveneur, T. Salmi, N. Musakka, J. W a¨ rn a˚ , Chem. Eng. Sci. 62, 5007 (2007)
9
1
0. X. Zhao, K. Cheng, J. Hao, D. Liu, J. Mol, A. Catal, Chemical 284, 58 (2008)
1. S. Leveneur, D. Yu. Murzin, T. Salmi, J.-P. Mikkola, N. Kumar, K. Er a¨ nen, L. Estel, Chem. Eng.
J. 147, 323 (2009)
1
1
1
1
1
2. A. Palani, A. Pandurangan, Catal. Commun. 7, 875 (2006)
3. G. Prescher, O. Weiberg, H. Waldmann, H. Seifert, International Patent US4088679 (1978)
4. F.P. Greenspan, D.G. MacKellar, Anal. Chem. 20, 1061 (1948)
5. D. Kubi cˇ ka, N. Kumar, P. M a¨ ki-Arvela, M. Tiitta, V. Niemi, T. Salmi, D.Yu. Murzin, J. Catal. 222,
6
5 (2004)
6. P. M a¨ ki-Arvela, N. Kumar, V. Nieminen, R. Sj o¨ holm, T. Salmi, D. Yu Murzin, J. Catal. 225, 155
2004)
1
1
(
7. E.M. Sulman, V.V. Alferov, Yu.Yu. Kosivtsov, A.I. Sidorov, O.S. Misnikov, A.E. Afanasiev,
N. Kumar, D. Kubicka, J. Agullo, T. Salmi, D.Yu. Murzin, Chem. Eng. J. 134, 162 (2007)
8. A. Aho, N. Kumar, K. Er a¨ nen, T. Salmi, M. Hupa, D.Yu. Murzin, Fuel 87, 2493 (2008)
9. J.P. Marques, I. Gener, P. Ayrault, J.C. Bordado, J.M. Lopes, F.R. Ribeiro, M. Guisnet, C. R. Chimie
1
1
8, 399 (2005)
20. P.K. Sengupta, R.N. Tiwari, S. Bhagat, Chem. Eng. World 32, 89 (1997)
21. G.O. Rocha, R.A.W. Johnstone, B.F. Hemming, P.J.C. Pires, J.P. Sankey, J. Mol, A. Catal, Chemical
1
86, 127 (2002)
2. A.E.W. Beers, J.A. van Bokhoven, K.M. de Lathouder, F. Kapteijn, J.A. Moulijn, J. Catal. 218, 239
2003)
2
(
2
2
2
2
2
2
3. D.M. Roberge, H. Hausmann, W.F. H o¨ lderich, Phys. Chem. Chem. Phys. 4, 3128 (2002)
4. R.L. Hartman, H.S. Fogler, Ind. Eng. Chem. Res. 44, 7738 (2005)
5. J. De Laat, G.T. Le, B. Legube, Chemosphere 55, 715 (2004)
6. B. Mani, Ch. Ravi Mohan, V. Sitakara Rao, React. Kinet. Catal. Lett. 13, 277 (1980)
7. W. Lutz, H. Toufar, R. Kurzhals, M. Suckow, Adsorption 11, 405 (2005)
8. M. Guisnet, F.R. Ribeiro, Les z e´ olithes, un nanomonde au service de la catalyse; EDP sciences
(2006)
123