Sie sind auf Seite 1von 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/222259086

Monopropellant decomposition catalysts: V. Thermal decomposition


and reduction of permanganates as models for the preparation of
supported MnOx catalysts

Article  in  Applied Catalysis A General · August 2002


DOI: 10.1016/S0926-860X(02)00220-X

CITATIONS READS

43 215

7 authors, including:

Charles Kappenstein Laurence Pirault-Roy


Université de Poitiers Université de Poitiers
132 PUBLICATIONS   2,189 CITATIONS    63 PUBLICATIONS   501 CITATIONS   

SEE PROFILE SEE PROFILE

Tarek Wahdan Asma Ali


Suez Canal University Higher Colleges of Technology
12 PUBLICATIONS   203 CITATIONS    11 PUBLICATIONS   77 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Surface Acid-Base Properties of Metal Oxides View project

Generation of nanoparticles using microemulsion View project

All content following this page was uploaded by Mohamed I Zaki on 15 October 2017.

The user has requested enhancement of the downloaded file.


Applied Catalysis A: General 234 (2002) 145–153

Monopropellant decomposition catalysts


V. Thermal decomposition and reduction of permanganates as
models for the preparation of supported MnOx catalysts夽
Charles Kappenstein a,∗ , Laurence Pirault-Roy a , Maurice Guérin a , Tarek Wahdan b ,
Asma A. Ali c , Fakhreia A. Al-Sagheer c , Mohamed I. Zaki c,1
a Laboratoire de Catalyse en Chimie Organique, Faculté des Sciences, UMR-CNRS 6503, Catalyse par les Métaux,
40 Avenue du Recteur Pineau, 86022 Poitiers, France
b Institute of Natural Resources and Environmental Research, Riyadh, Saudi Arabia
c Chemistry Department, Faculty of Science, Kuwait University, P.O. Box 5969, Safat, Kuwait

Received 29 November 2001; received in revised form 18 March 2002; accepted 18 March 2002

Abstract
The need for highly concentrated hydrogen peroxide solutions to fuel spacecraft propulsion systems necessitates the
development of new catalytic formulations mainly based on supported manganese oxides. In order to improve the preparation
of such catalysts, the thermal decomposition and the hydrothermal reduction of different permanganate precursors were studied,
as well as the effect of washing on the products. The test samples were characterized by X-ray power diffraction (XRD) and
surface area determination, and compared for their catalytic activity towards H2 O2 decomposition using a constant pressure
reactor (for 1.71% H2 O2 ) and a constant volume reactor (for 50% H2 O2 ). The catalysts prepared by thermal decomposition
of KMnO4 or NaMnO4 ·H2 O have shown the intermediate formation of potassium (or sodium) manganate, and its partial
dissolution during washing with water; higher catalytic activity and larger surface area were observed for the decomposition
products of KMnO4 . The hydrothermal reduction yielded products exhibiting larger surface area and higher specific catalytic
activity when carried out under stoichiometric conditions. Comparable kinetic parameters were determined using both catalytic
reactors. Unsupported and alumina supported catalysts displayed comparable activity per gram of manganese.
© 2002 Elsevier Science B.V. All rights reserved.
Keywords: Hydrogen peroxide decomposition; Manganese oxide catalysts; Potassium permanganate; Sodium permanganate; Permanganate
reduction; Thermal decomposition

1. Introduction
夽 These results were originally presented at the 3rd Interna-

tional Hydrogen Peroxide Propulsion Conference, Gulfport, MS, The use of more concentrated hydrogen peroxide
November 2000 (Conference proceedings CD). solutions (i.e. up to 98%) leads to a strong increase of
∗ Corresponding author. Tel.: +33-5-49-453860;
the adiabatic decomposition temperature (632 ◦ C for
fax: +33-5-49-454020. 85%; 755 ◦ C for 90%; 953 ◦ C for 98%), which is in-
E-mail address: charles.kappenstein@univ-poitiers.fr
(C. Kappenstein).
adequate for long-term application of conventional sil-
1 Present address: Chemistry Department, Faculty of Science, ver or silver plated screen catalysts. The melting point
Minia University, Egypt. of silver (962 ◦ C) is near the adiabatic temperature

0926-860X/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 6 - 8 6 0 X ( 0 2 ) 0 0 2 2 0 - X
146 C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153

and the formation of silver oxides will preclude the hydrothermal reduction of different permanganate
long-term use of such type of catalysts. Other cat- precursors in bulk state (unsupported samples), as a
alytic formulations have been proposed to replace the model for the precursor–precursor interactions, which
silver catalysts [1]; most of them are based on man- may occur inside the support pores of supported cat-
ganese oxides supported on alumina (MnOx /Al2 O3 ), alysts, (ii) to follow the effect of washing with water
silica (MnOx /SiO2 ) or cordierite monolith honeycomb on the active phases, and (iii) to compare the cat-
(MnOx /2MgO·2Al2 O3 ·5SiO2 ). alytic activity of the different manganese oxide sam-
The impregnation procedure is important for a good ples, thus, produced towards H2 O2 decomposition in
anchorage of the active phase onto the support, and the solution.
most efficient catalyst precursor compounds are the
permanganates of potassium, sodium, calcium or am-
monium [2,3], or even permanganic acid HMnO4 [4]. 2. Experimental
Formation of the manganese oxides can be effected
either by thermal decomposition of the precursor in 2.1. Preparation of test catalysts
the solid state, or chemical reduction of the MnO4 −
anions in aqueous or non-aqueous solutions. Decomposition of potassium permanganate (KM-
The preparation course of a supported manganese nO4 , analytical grade, Labosi, France; 2 g in a flat
oxide catalyst often involves three main processes: covered porcelain crucible) or sodium permanganate
(i) wet impregnation of the support with an aque- (NaMnO4 ·H2 O, 98% pure, Strem Chemicals, USA)
ous solution containing manganese precursor species was performed in static air at 400 ◦ C for 1 or 9 h at
at a given concentration; (ii) drying at 80–120 ◦ C; a heating rate of 23 ◦ C min−1 . After calcination and
(iii) calcination at 400–700 ◦ C, in oxidative (air), in- cooling to ambient temperature, the solid residues
ert (N2 , Ar) or reductive gas atmosphere (H2 , CO). were washed with 250 ml distilled water, dried at
Process (ii) can be described as a combination of 110 ◦ C for 5 h, and calcined further at 600 ◦ C for 5 h.
the following two competitive solid–solid interac- For simplicity, the catalysts thus obtained are denoted
tions: (a) a precursor–support surface interaction, below by the chemical symbols of the alkali metal
leading to strong anchorage of the precursor species and manganese atoms, followed by the successive
onto the support; and (b) a precursor–precursor bulk designations: an Arabic numeral indicating the time
interaction within the support pores, leading to crys- duration (in hours) of the calcinations at 400 ◦ C, w
tallization of the precursor followed by its thermal if washed with water, and 600 if calcined further at
decomposition throughout the calcination process 600 ◦ C for 5 h. Thus, KMn9h indicates the 400 ◦ C cal-
(iii). Stronger precursor–precursor interactions than cination product of KMnO4 for 9 h, whereas KMn9hw
precursor–support interactions lead to weakly bound signifies the same product following washing with
precursor species, which could be dissolved in the water and drying at 110 ◦ C for 5 h.
H2 O2 solution and washed away during the first Reduction of the permanganate precursors was
pulses. This may obviously be promoted by the strong performed in an autoclave implementing the Guyard
solvation capacity of the hydrogen peroxide, which is (redox) reaction [7] with manganous ions in aque-
close to that of water. ous solution (at pH 1 or 7) containing foreign ion
Another way to prevent the partial dissolution of the additives, under controlled hydrothermal conditions
calcined precursor is to perform a direct redox reac- (100 ◦ C, 120 kPa, 16 h). The amount of the reac-
tion between permanganate and a mild reductant under tants and details of the reaction variables have been
controlled hydrothermal conditions, in the presence of reported elsewhere [5]. Following cooling to ambient
appropriate foreign ion additives. This procedure has temperature, the resulting precipitates were filtered,
been studied in the bulk state with manganous ions as washed with distilled water, dried at 120 ◦ C for
a reductant and led to the formation of a variety of 16 h, and ground to pass through a standard sieve of
tunnel-structured MnO2 [5,6]. 250 mesh. Designations of the different test catalysts
Therefore, the main objectives of the present work thus obtained are given corresponding to a preparation
were: (i) to study the thermal decomposition or briefing in Table 1.
C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153 147

Table 1
Designation, preparation conditions and characteristics of the different test samples
Samples Preparation XRD phasea SBET (m2 g−1 )

Thermal decomposition
KMn1h Calcination of KMnO4 , 400 ◦ C, 1 h K2 MnO4 + K2 Mn4 O8 (?) 7.6
KMn9h Calcination of KMnO4 , 400 ◦ C, 9 h K2 MnO4 + K2 Mn4 O8 (?) 6.5
KMn1hw KMn1h washed, water, dried 110 ◦ C K2 Mn4 O8 (?) 24
KMn9hw KMn9h washed, water, dried 110 ◦ C K2 Mn4 O8 (?) 22
KMn9hw600 KMn9hw calcined, 600 ◦ C, 5 h K2−x Mn8 O16 12
NaMn9h NaMnO4 ·H2 O calcined, 400 ◦ C, 9 h Na0.7 MnO2 2.8
NaMn9hw600 NaMn9h washed, calcined 600 ◦ C, 5 h Na0.7 MnO2 + Mn7 O13 ·5H2 O 2.1
Hydrothermal reductionb
S1 MnSO4 + excess KMnO4 , pH 1 ␣-MnO2 (fibrils) 71
S2 Excess MnSO4 + KMnO4 , pH 1 ␣-MnO2 (fibrils) 85
S3 MnSO4 + NaMnO4 , stoichiometric, pH 1 ␥-MnO2 + ␤-MnO2 (minor) 127
S4 MnSO4 + KMnO4 , stoichiometric, pH 7 Amorphous-MnO2 165
S5 Excess MnSO4 + NaMnO4 + Ca(NO3 )2 , pH 1 t-MnO2 62

2−x Mn8 O16 (␣-MnO2 ) 44–1386; Na0.7 MnO2 27–751; Mn7 O13 ·5H2 O 23–1239;
a PDF number: K MnO 12–264; K Mn O 16–205; K
2 4 2 4 8
␥-MnO2 17–510; ␤-MnO2 24–735; t-MnO2 (todorokite) 38–475.
b Precise amounts of the reactants and additives can be found in [5].

2.2. Catalyst characterization 2.3. Catalytic activity

The test catalysts were characterized by X-ray pow- Two reactors were used to determine the activity
der diffractometry (XRD) and nitrogen sorptometry of the test catalysts towards the decomposition of
for BET surface area determination. The XRD was H2 O2 solutions. The first reactor was designed to
carried out at ambient temperature, using a D5000 work at constant pressure and temperature and has
Bruker diffractometer equipped with a source of back- been already described in a previous paper [9]. It can
monochromatized Cu K␣ radiation (λ = 0.15406 nm). be used to measure the volume of the evolved oxygen
The diffractometer was operated at 40 kV and 30 mA, up to 60 ml. Diluted H2 O2 (10 ml, 1.71 wt.%) was
and the data were acquired step-wise in the 2θ range placed in the reactor at 20 ◦ C and, then, the catalyst
5–80◦ with a step size of 0.02 or 0.04◦ , a time step (10–200 mg) was added to initiate the decomposition
of 1–10 s depending of the sample, and a divergence reaction. The second reactor is a constant volume
slit of 1◦ , using an online microcomputer. An auto- batch reactor (170 ml) with an operating pressure
matic PDF library search and match was conducted, between vacuum and 2 bar, the design of which has
using the standard SEARCH and DIFFRAC soft- been described in earlier communications [10,11]. A
ware (Bruker), for crystalline phase identification 200 mg portion of the test catalyst was placed inside
purposes. the reactor at ambient temperature, and the H2 O2
Nitrogen sorptometry was carried out at liquid nitro- solution (50 wt.%, Prolabo, France; 100 ␮l) was in-
gen temperature (−195 ◦ C), using an automatic ASAP jected using a microsyringe.
2010 Micromeritics sorptometer with outgassing plat-
form and online data acquisition and handling system 3. Results and discussion
operating BET [8] analytical software for the adsorp-
tion data. The nitrogen gas was a 99.999% pure prod- 3.1. Thermal decomposition of KMnO4
uct of KOAC (Kuwait). A 300 mg portion of the test
catalyst was outgassed at 100 ◦ C and 10−5 Torr for The thermal decomposition of KMnO4 was fol-
12 h, prior to exposure to the nitrogen atmosphere. The lowed by XRD, the diffractograms obtained are dis-
reproducibility of the measurements was better than played in Fig. 1. The calcination at 400 ◦ C for 1 h is
98%. shown to result in the formation of a solid residue
148 C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153

Increasing the calcination period led to an obvious de-


crease of the manganate phase in the product KMn9h.
Following washing with distilled water and subsequent
drying, the potassium manganate phase disappeared
completely in the product KMn9hw, most probably
as a result of dissolution in water. XRD identification
of the remaining ill-crystallized material was difficult
to identify (Fig. 1) and different phases could be pro-
posed on the basis of the PDF data file: K2 Mn4 O8
(16–205); K0.5 Mn2 O4 ·1.5H2 O (42–1317); KMnO2
(18–1035, 44–1025); and MnO2 (44–141). Despite
of an apparent agreement with the standard pattern
of K2 Mn4 O8 , none of these phases is in complete
agreement with the position or intensity of the XRD
peaks observed. The further calcination of KMn9hw
at 600 ◦ C for 5 h is shown (Fig. 1) to yield a material
(KMn9hw600) assuming the crystalline structure
of cryptomelane phase (K2−x Mn8 O16 , PDF 44–
1386).
The intermediate formation of potassium man-
ganate is in agreement with data compiled elsewhere
[12]. The washing of KMn9h gave a green solution
containing manganate MnO4 2− ions, which turned
violet rapidly due most likely to the following man-
ganese disproportionation reaction:
Fig. 1. XRD powder diffractograms for the indicated calcination
products of KMnO4 .
3MnO4 2− (aq) + 2H2 O

→ 2MnO4 − (aq) + MnO2 (s) + 4OH− (aq)


(KMn1h) containing a mixture of well-crystallized
potassium manganate (K2 MnO4 , PDF 12–264) and a After washing, the surface area of the solid residues,
poorly crystallized phase, as well as in the develop- whether KMn1hw or KMn9hw, is shown to increase
ment of a large strong diffraction peak near 2θ = 13◦ . by a factor of three (Table 1) at the expense of the

Table 2
Thermal decomposition of KMnO4 . Mass loss of the different steps observed, major XRD phase, crystallite size and oxidation number
of the Mn atoms [13]
Samples Mass lossa (%) Major crystalline phase Crystallite sizeb (nm) Mn oxidation number

Observedc Calculatedd

KMn1h 11.0 K2 MnO4 12 5.0 4.8


KMn9h 11.5 K2 MnO4 11 5.1 4.7
KMn1hw 46.8 K2 Mn4 O8 (?) 12 5.3
KMn9hw K2 Mn4 O8 (?) 11 5.0 5.0
KMn9hw600 3.5 K2−x Mn8 O16 14 4.3 4.5
a Determined by means of TG analysis.
b Calculated, adopting the XRD line broadening technique.
c Determined by TPR analysis.
d Based on chemical analysis results.
C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153 149

Table 3
Suggested reaction pathways for the thermal decomposition of the permanganate compounds, and derived mass losses
Thermal decomposition reaction Mass loss % (calculated)

2KMnO4 (s) → K2 MnO4 (s) + MnO2 (s) + O2 (g) 10.1


6KMnO4 (s) → 2K2 MnO4 (s) + K2 Mn4 O8 (s) + 4O2 (g) 13.5
2KMnO4 (s) → K2 MnO4 (s) + MnO1.84 (s) + 1.08O2 (g) 11.0
2NaMnO4 ·H2 O(s) → Na2 MnO4 (s) + MnO2 (s) + 2H2 O(g) + O2 (g) 21.3
3NaMnO4 ·H2 O(s) → Na2 MnO4 (s) + NaMn2 O4 (s) + 3H2 O(g) + 2O2 (g) 24.6a
a Observed total mass loss for NaMnO4 ·H2 O is 24.0%.

loss of a part of the active phase (i.e. the water soluble the PDF database has not contained the corresponding
K2 MnO4 ). file.
The mass loss conceded by the different decom- Nevertheless, two crystalline phases in the cal-
position products of KMnO4 treatments, as well as cination products of NaMnO4 are in fair agree-
their crystallite size, and calculated and observed man- ment with standard diffraction data of Na0.7 MnO2.05
ganese oxidation numbers, are given in Table 2. It (27–751) and Mn7 O13 ·5H2 O (23–1239). The ef-
is obvious that the mean oxidation number of Mn in fected mass loss (24%) throughout the calcinations
the 400 ◦ C calcination products is about 5.0, which at 400 ◦ C is in fair agreement with the formation of
is compatible with a mixture of manganate (6.0) and a mixture consisting of Na2 MnO4 , and NaMn2 O4
MnO2 (4.0) or K2 Mn4 O8 (3.5). However, mass losses (Table 3; calculated loss = 24.6%). Following wash-
calculated on the basis of decomposition reactions ing, about one-fourth of the manganese is retrieved
suggested in Table 3 do not agree with the observed as MnO4 − (aq) ions. The calcination products are
ones (Table 2). A reasonable compromise would be shown (Table 1) to assume poor surface areas;
the formation at 400 ◦ C of a non-stoichiometric oxide
(MnO1.84 ). The calcination at 600 ◦ C led to a decrease
in the surface area from 22 to 12 m2 g−1 (Table 1) in
relation to the formation of crystalline cryptomelane.
Meanwhile, a slight mass loss of 3.5% and a signifi-
cant decrease in the oxidation number from 5.0 to 4.3
are shown (Table 2) to ensue.

3.2. Thermal decomposition of NaMnO4 ·H2 O

In contrast to the well-crystallized potassium per-


manganate, the sodium permanganate monohydrate
(as well as the decomposition products) is a very
hygroscopic powder. The XRD results are exhibited
in Fig. 2, but the identification of the crystalline phase
composition of the decomposition products is far
from being complete. This is essentially attributable
to the lack of sufficient reference data. For instance,
while the intermediate formation of sodium man-
ganate (Na2 MnO4 ) following calcination at 400 ◦ C
of the permanganate could be accounted for on basis
of the development of the characteristic intense green
color in the filtrate of water washing of the 400 ◦ C Fig. 2. XRD powder diffractograms for the indicated calcination
calcination product (turns violet in a few seconds), products of NaMnO4 ·H2 O.
150 C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153

namely, 2.8 m2 g−1 for NaMn9h and 2.1 m2 g−1


for NaMn9hw600. Therefore, it is concluded that
sodium permanganate is not an adequate precursor
for Mn-oxide based catalysts.

3.3. Hydrothermal reduction of permanganates


in solution

The Guyard-like redox reaction [7] carried out is


given by the following stoichiometric equation:

2MnO4 − (aq) + 3Mn2+ (aq) + 2H2 O


→ 5MnO2 (s) + 4H+ (aq)
The products were poorly crystalline powders with
high surface area, as compared with the thermal de-
composition products (Table 1). The highest surface
areas are those assumed by the samples S3 (␥-MnO2 Fig. 3. Catalytic activity of KMn9hw600 and NaMn9hw600
+ ␤-MnO2 ) and S4 (amorphous MnO2 ), which towards 10 ml 1.71 wt.% H2 O2 .
were obtained via a reaction between stoichiometric
amounts of the oxidant and reductant [5]. The struc-
ture of the manganese oxides obtained has been shown profiles revealing an important induction period for
to depend intimately on the reaction conditions [5]. the reaction onset. The linear portion of the curve is
Moreover, the product oxides were found to accommo- characteristic of a zero-order kinetics for the reaction
date alkali and alkaline earth ions both in the surface [9]. Rate constant values derived from the slopes are
and the bulk structure. Calcination at 250 ◦ C neither given in Table 4. These data are of the same order
changed the surface area, nor the bulk structure. of magnitude as the results obtained [9] for supported
Ir/Al2 O3 grain catalyst (k0 = 35 × 10−8 mol s−1 (mg
3.4. Catalytic activity catalyst)−1 ). Catalysts derived from KMnO4 exhibited
a higher activity and clearer first-order reaction kinet-
3.4.1. Constant pressure reactor ics than the NaMnO4 -derived catalysts (Fig. 3); the
Results of the catalytic decomposition of H2 O2 ob- rate constant values are also shown in Table 4.
tained with the constant pressure reactor are compared Fig. 4 shows the volume–time curves con-
for KMn9hw600 and NaMn9hw600 samples in Fig. 3. structed for the catalysts prepared by hydrother-
For both catalysts, the activity increases expectedly mal reduction. All the catalyst samples exhibit
with the catalyst mass. The catalysts derived from clearly first-order kinetics; the whole set of re-
the sodium permanganate precursor display S-shaped sults is displayed in Table 5. The highest spe-

Table 4
Rate constant values for the catalytic decomposition of H2 O2 . Zero-order kinetics for NaMnO4 based catalysts; first-order kinetics for
KMnO4 based catalysts
Sample Mass (mg) k0 (mol s−1 ) k0 (mol s−1 (g catalyst)−1 )

NaMn9h 13 6.6 × 10−6 51 × 10−5


NaMn9h 24 12 × 10−6 48 × 10−5
NaMn9h 29 13 × 10−6 46 × 10−5
Sample Mass (mg) k1 (ml s−1 ) k1 (ml s−1 (g catalyst)−1 )

KMn9h 10 0.09 9
KMn9h 15 0.16 11
C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153 151

Fig. 4. Catalytic activity of the oxides obtained by hydrothermal


reduction towards 10 ml 1.71 wt.% H2 O2 .

cific rate constant k1 (i.e. per gram of catalyst)


is that determined for the sample S3 (␥-MnO2 +
Fig. 5. Decomposition of 50% H2 O2 on the hydrothermal reduction
␤-MnO2 ), though, in general, no clear relationship product S3. Variation of pressure, and catalyst and gas phase
can be established between activity and surface area. temperatures vs. time for three successive injections of 100 ␮l
The intrinsic rate constant (k1 , per square meter of from 1 bar argon pressure.
surface area) is highest for sample S5 (t-MnO2 ).
This obvious disparity may be related to the fact that
specific and intrinsic rate constants assume different specific activity is one order of magnitude lower, most
dependencies; the specific rate constant must take into likely due to the low manganese content; but the spe-
account the influences of both textural (surface area, cific activity (per gram of manganese, last column in
porosity) and non-textural parameters, whereas the Table 5) ranks this supported catalyst atop of the list.
intrinsic rate constant rules out the textural ones. On
the other hand, the sample KMn9hw600 prepared by 3.4.2. Constant volume reactor
decomposition of KMnO4 displays a specific activity Two catalyst samples (S1 and S3) were tested using
of the same order of magnitude but a much higher the constant volume (batch) reactor for three suc-
intrinsic rate constant due mainly to its low surface cessive 50% H2 O2 injections. Variations of the pres-
area. For the supported sample MnOx /Al2 O3 , the sure, the catalyst temperature and the gas phase

Table 5
First-order rate constant values for H2 O2 decomposition on hydrothermal reduction products
Sample Mn Mass k1 k1 SBET k1 k1 
(wt.%) (mg) (ml s−1 ) (ml s−1 g−1 ) (m2 g−1 ) (ml s−1 m−2 ) (ml s−1 (g Mn)−1 )
S1 47.3 13 0.13 10 71 0.14 21
S2 52.3 13 0.26 20 85 0.23 38
S3 57.6 8 0.22 28 127 0.22 49
S4 49.8 9 0.18 20 165 0.12 40
S5 46.5 12 0.21 18 62 0.29 39
KMn9hw600 51.4 15 0.16 11 12 0.90 21
MnOx /Al2 O3 1.6 177 0.17 1.0 – – 60
152 C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153

Table 6
Decomposition of 50% H2 O2 . Successive pulses of 100 ␮l H2 O2 on catalyst samples S1 and S3 (200 mg). Pressure increase and pressure
slope, initial and maximum temperature of the catalyst, temperature slope, maximum temperature of the gas phase
Sample Injection P increase P slope Tinitial Tmax T slope Tmax gas
number (mbar) (mbar s−1 ) (◦ C) catalyst (◦ C) (◦ C s−1 ) (◦ C)
S1 1 130 225 20 84 53 37
2 111 233 20 75 45 38
3 108 250 20 71 45 30
S3 1 133 255 20 95 47 52
2 109 271 20 98 80 59
3 105 317 20 97 85 35

temperature, are graphically presented for the sample oxide dose. For the first pulse, the pressure step is
S3 in Fig. 5. The pressure peaks are shown to be very highest, due to the liquid–vapour water equilibrium
sharp, indicating strong activity. A zoom at the results (23 mbar at 20 ◦ C). The pressure slope, the tempera-
obtained throughout the second injection is presented ture maxima and the temperature slopes are definitely
in Fig. 6; the three records display parallel variations higher for sample S3 than S1, thus indicating a higher
versus time, and the slope of each curve can be deter- activity for the former catalyst. The activity of the
mined. The results are reported in Table 6. The pres- S3 catalyst is, hence, comparable to the activity of
sure increase sequence after each injection is shown supported MnOx /Al2 O3 catalyst, but slightly lower
to be the same for both test catalysts, and corresponds than the activity of powdered Ag/Al2 O3 catalyst
to a complete decomposition of the hydrogen per- [9].

4. Conclusions

The results presented and discussed for the catalyst


samples prepared by thermal decomposition of solid
permanganates may facilitate drawing the following
conclusions:
1. The intermediate formation of solid potassium or
sodium manganate in the thermal decomposition
course of the corresponding permanganate precur-
sor.
2. The dissolution in water of the manganate propor-
tion of the decomposition products.
3. A higher activity and larger surface area exhibited
by catalysts derived from KMnO4 .
On the other hand, the catalysts prepared by hy-
drothermal reduction of permanganates in solution
are distinct by: (i) a larger surface area, (ii) a stronger
specific activity for the samples prepared under sto-
ichiometric conditions, (iii) comparable results for
Fig. 6. Zoom at the second pulse for the decomposition of H2 O2
both catalytic reactors, and (iv) comparable activity
on sample S3. The data frequency is 10 Hz for each gauge. Note per gram of manganese for unsupported and alumina
the full scale (3 s) on the x-axis. supported catalyst samples.
C. Kappenstein et al. / Applied Catalysis A: General 234 (2002) 145–153 153

Acknowledgements [4] C. Kappenstein, T. Wahdan, D. Duprez, M.I. Zaki, D.


Brandes, E. Poels, A. Bliek, Stud. Surf. Sci. Catal. 91 (1995)
We thank the French Space Agency, Centre Na- 699.
[5] A.A. Ali, M.Sc. Thesis, Kuwait University, 1999.
tional d’Etudes Spatiales (CNES), Toulouse for fund-
[6] A.A. Ali, F.A. Al-Sagheer, M.I. Zaki, Intern. J. Inorg. Mater.
ing of, and interest in, this study, Mr. Ahmed Nohman 3 (2001) 427.
(Minia University, Egypt) for the design and realiza- [7] A. Guyard, Bull. Soc. Chim. Fr. 1 (1964) 89.
tion of the constant pressure reactor, and the Faculty [8] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and
of Post-graduate studies, SAF and EMU of the Fac- Porosity, Academic Press, New York, 1967, pp. 36–41
[9] C. Kappenstein, L. Pirault-Roy, M. Guerin, R. Eloirdi, N.
ulty of Science for supporting the work carried out in
Pillet, in: Proceedings of the 3rd International Hyrogen
Kuwait University. Peroxide Conference, Gulfport, MS, November 2000.
[10] R. Eloirdi, S. Rossignol, C. Kappenstein, D. Duprez, N. Pillet,
AIAA Paper 00-3553.
References
[11] R. Eloirdi, S. Rossignol, C. Kappenstein, D. Duprez, N.
Pillet, in: Proceedings of the 3rd International Conference on
[1] J.J. Rusek, J. Propul. Power 12 (1996) 574. Spacecraft Propulsion, Cannes, France, October 2000.
[2] F.M. Radwan, A.M. Abdel-Hameed, Z. Phys. Chem. (Leipzig) [12] Gmelin Handbuch der Anorganischen Chemie, System
271 (1990) 1169. Number 56, Manganese, Part C2, Springer, Berlin,
[3] A.K.H. Nohman, D. Duprez, C. Kappenstein, S.A.A. 1975.
Mansour, M.I. Zaki, Stud. Surf. Sci. Catal. 63 (1991) 617. [13] T. Wahdan, Ph.D. Thesis, Mansoura University, Egypt, 1994.

View publication stats

Das könnte Ihnen auch gefallen