Catalytic decomposition of hydrazine on tungsten carbide: The influence of adsorbed oxygen

Article abstract of DOI:10.1006/jcat.2002.3634

Hydrazine is one of the most frequently utilized liquid fuels in microthrusters. Passivated and nonpassivated tungsten carbide (WC) powders were prepared by a temperature-programmed reaction and utilized as catalysts in the hydrazine decomposition reactions. All catalysts were active at 313 K. The only reaction products were NH₃ and nitrogen, suggesting that the reaction route and the rate-determining step were not modified by adsorbed oxygen. Surface oxygen atoms reacted with hydrazine, leading to a larger decrease in the reaction rate for the passivated solids. Excess carbon on the surface of the carbide inhibited hydrazine decomposition and led to a decrease of ～ 48% in the specific surface area of the tungsten carbide. The excess carbon might be removed in flowing hydrogen at 1000 K. When compared to commercial Ir catalyst (Shell 405), WC was less active in the hydrazine decomposition reaction.

Full text of DOI:10.1006/jcat.2002.3634

Journal of Catalysis 210, 1–6 (2002)
doi:10.1006/jcat.2002.3634
Catalytic Decomposition of Hydrazine on Tungsten Carbide:
The Influence of Adsorbed Oxygen
∗
∗,1
J. B. O. Santos, G. P. Valen c¸ a, and J. A. J. Rodrigues†
∗
Department of Chemical Processes, School of Chemical Engineering, State University of Campinas, UNICAMP, C.P. 6066, 13081-970 Campinas, SP,
Brazil; and †National Institute Space Researche, INPE, Rod. Presidente Dutra, Km 40, 12630-000 Cachoeira Paulista, SP, Brazil
Received June 18, 2001; revised March 15, 2002; accepted April 11, 2002
catalysts for space applications, new materials have been
Passivated and nonpassivated tungsten carbide (WC) powders studied as prospective catalysts for the decomposition of
were prepared by a temperature-programmed reaction and used N H .
2
4
as catalysts in the hydrazine decomposition reaction. All catalysts
were active at 313 K. The only products of reaction were nitro-
gen and ammonia, suggesting that the reaction route and the rate-
determining step were not modified by adsorbed oxygen. Surface
oxygen atoms reacted with hydrazine, resulting in a larger decrease
in the rate of reaction for the passivated solids. Excess carbon on
the surface of the carbide inhibited the decomposition of hydrazine
and resulted in a decrease of ca. 48% in the specific surface area of
the tungsten carbide. The excess carbon may be removed in flow-
ing hydrogen at 1000 K. When compared to commercial Ir cata-
lyst (Shell 405), WC is less active in the hydrazine decomposition
reaction. ꢀc 2002 Elsevier Science (USA)
Several transition metals, such as Ir (2–4), Rh (2, 5), Ni
2), Co (2), Pt (2, 6), Ru (2), Pd (2), Ag (2), Cu (2), Fe (7),
and W (8), are active catalysts for the decomposition of
N2H4. The decomposition of N2H4 may occur by the fol-
lowing reactions:
(
N2H4 → N2 + 2H2
3N2H4 → N2 + 4NH3.
[1]
[2]
Reactions 1 and 2 were observed on Rh (2), Pt (2), and
Pd (2) between 333 and 433 K, with an activation energy of
−
1
−1
Key Words: hydrazine decomposition; tungsten carbide; 46 kJ mol for Rh and 67 kJ mol for Pt and Pd, while only
temperature-programmed reaction; WC catalyst; passivated WC or mainly reaction 2 was observed on Ir, Ni, Co, Ru, Ag,
powders; nonpassivated WC powders; surface oxygen.
and Cu, with activation energies varying between 46 and
−
1
7
1 kJ mol . For a 4% Ni/γ -Al2O3 catalyst, at 292 K and a
N2H4 partial pressure of 5.87 mbar, Valen c¸ a (9) observed
only reaction 1 while on Ni (111), the formation of diimide
1
. INTRODUCTION
(
10, 11) was also observed.
Rodrigues and co-workers (12–13) showed that either
Hydrazine, N2H4, has several applications and is one of
the most frequently used liquid fuels in microthrusters (1).
The use of N2H4 in satellite attitude control is due to the low
activationenergyofitsdecompositionandthehighexother-
micity of this reaction. These features combine two proper-
tieswhichareimportantforspaceapplications:theinjection
of liquid reactants at room temperature (RT) and the for-
mation of gaseous products at high temperature (>1000 K)
in a short period of time (<1 s). For a solid to be a suitable
catalyst for space applications, the decomposition of N2H4
should occur at RT. The commercial catalyst, SHELL 405
tungsten or molybdenum carbides or nitrides may be used
as catalysts in the decomposition of N2H4. They observed
that all catalysts were active in this reaction at RT. These
authors also observed that at 333 K Mo2C was the most
active solid catalyst. On the other hand, during tests per-
formed on a 2-N thruster rocket engine under conditions
that simulate the actual conditions found in space, W2C was
the most efficient catalyst, and in some cases it was more
active than the commercial catalyst (13).
Tungsten carbides or nitrides are interstitial alloys used
in metallurgical applications (14). Passivated tungsten and
molybdenum carbides were first used as catalysts for the
isomerization of hydrocarbons, a reaction typically cata-
lyzed by noble metals (15). The study of these materials as
catalysts became feasible in the 1980s, when Oyama (16)
and Volpe and Boudart (17, 18) developed a methodol-
ogy to prepare high surface area tungsten or molybdenum
carbides or nitrides. In spite of their similarities to noble
metals, carbides and nitrides have unique features. For
(
∼36% w/w Ir/Al2O3), has physical and chemical proper-
ties which are unique for space applications, such as ther-
mal stability, bimodal distribution of pore diameters, and
uniform size distribution of Ir particles (d < 2 nm) on the
surface of the Al2O3. However, due to the high price of Ir
and the high loading of metal necessary to prepare suitable
1
To whom correspondence should be addressed. E-mail: gustavo@
feq.unicamp.br.
1
0
021-9517/02 $35.00
c 2002 Elsevier Science (USA)
All rights reserved.
ꢀ

2
SANTOS, VALEN C¸ A, AND RODRIGUES
example, the catalytic properties of transition metal car- Some WC-f samples were treated with H2 at 998 K for 1.5 h
bides may be affected by adsorbed oxygen. Iglesia et al. and these solids are called freshly reduced carbides, WC-fr.
(
19, 20) found that oxygen adsorbed on WC introduces Finally, some WC-fr samples were passivated in a flow of
WOx sites that catalyze the isomerization of n-heptane with 1.0% O2/He mixture for 1 h. These samples are called
high selectivity. Ribeiro et al. (21, 22) found that oxygen ad- passivated-reduced carbides, WC-rp. Each sample (ca. 0.5 g)
sorbed on WC or on β-W2C inhibits the rate of hydrogenol- was prepared at least twice.
ysis and catalyzes the isomerization of neopentane and
The gases that evolved during the preparation of the WC
n-hexane, suggesting a bifunctional mechanism. Blekkan samples were analyzed by an on-line gas chromatograph
et al. (23) observed the bifunctional character of molyb- (CG 90), using a thermal conductivity detector and a Chro-
denum carbide exposed to oxygen in the isomerization of mosorb 102 column maintained at 373 K.
n-heptane. In addition, excess carbon on the surface of car-
2.2. Catalyst Characterization
bides affects their catalytic properties.
Although the catalytic decomposition of N2H4 has been
extensively studied, little is known about its sequence of
steps. The mechanisms proposed in the literature involve
the breaking of an N–N bond followed by recombination
of adsorbed N atoms or a step-by-step dehydrogenation of
the hydrazine molecule (24–26). For example, Maurel and
Menezo (27) concluded that on metals between 333 and
The passivated samples were characterized by X-ray
diffraction (XRD), N2 adsorption, and chemical analysis.
The bulk structure of the passivated samples was deter-
mined by the powder X-ray diffraction technique using a
Philips model PW 1830 diffractometer, with Cu Kα radia-
tion. The surface area was determined by the BET method
(
29). For nonpassivated samples, the analyses were per-
5
73 K, N2 is always formed from one hydrazine molecule
formed in a glass adsorption system, while the passivated
samples were analyzed in a Micromeritics-ASAP 2010C. To
measure the nonpassivated carbides BET specific surface
area, the cell was isolated and transferred to a glass volu-
metric adsorption system. The samples were then degassed
without cleavage of the N–N bond. Alberas et al. (28), how-
ever, concluded that cleavage of both the N–N bond and the
N–H bond could occur depending on the metal used. Thus,
an understanding of the decomposition kinetics of N2H4 is
necessary for the design of more efficient catalysts for space
applications.
The objective of this work is to prepare and study passi-
vatedandnonpassivatedWCinthedecompositionofN2H4.
As the tungsten and molybdenum are active catalysts for
the decomposition of N2H4 (12, 13), an understanding of
the kinetics of this reaction on these solids is of fundamen-
tal importance for the design of catalysts feasible for use
in space applications. In addition, there is very little in-
formation in the literature on the decomposition of N2H4
on tungsten carbides. Moreover, the influence of adsorbed
oxygen on the surface of tungsten carbides used for the
decomposition of N2H4 has not been reported.
−5
at 1.33 × 10 Pa for 0.5 h at RT, while passivated samples
were degassed at 673 K for 2 h prior to BET surface area
measurements.
The amounts of carbon and oxygen in the samples were
determined by chemical analysis. The amount of tungsten
was determined as the difference between the amount of
carbide and that of carbon and oxygen.
2
.3. Kinetic Measurements
The rate of decomposition of N2H4 was measured in the
system used to prepare the solids. The nonpassivated car-
bides were prepared and reduced in situ while the passi-
vated carbides were prepared in advance and stored in a
dissector. The amount of sample used in each experiment
was ca. 10 mg. All reactions were carried out at atmospheric
pressure. The catalysts were spread as a thin layer in the re-
actor that was attached to the reaction rig. The system was
2
. EXPERIMENTAL METHODS
2
.1. Catalyst Preparation
Tungsten carbide, WC, was prepared by temperature- then purged with flowing He for at least 1 h and the catalyst
2
3
−1
programmed reaction of WO3 powder (Carlo Erba, 3 m
g
was reduced in flowing H2 (1.64 cm s ) at 623 K for 1 h.
−
1
, 99.5%) in a flowing mixture of 20% v/v CH4 (White The flowofH2 wasthenswitchedtoHeandthetemperature
Martins, 99.97%) and H2 (White Martins, 99.999%) at an was decreased to RT. The system remained at RT for 1 h
−
1
overall flow rate of 71 µmol s at atmospheric pressure. until no H2 was observed by gas chromatography. Helium
3
−1
Typically, ca. 0.5 g of WO3 was spread over a quartz disk (0.93 cm s ) was then flowed first through a saturator con-
in a quartz cell. The sample was heated from RT to 1123 K taining N2H4 and then through a condenser, maintained at
−1
at a rate of 5 K min and then maintained at this temper- 303 K and 293 K, respectively. At 293 K, the partial pressure
ature for 1 h. The sample was then cooled in He (White of N2H4 is 1.24 kPa (30). After 0.5 h, the N2H4/He mixture
Martins, 99.999%) to RT. Samples thus prepared are called was flowed through the reactor and the decomposition of
fresh carbides, WC-f. Some WC-f carbide samples were pas- N2H4 was studied for at least 1 h. The gases that evolved
sivated in a flow of 1.0% O2/He (White Martins) mixture during the decomposition of N2H4 went through a H2SO4
for 1 h and these solids are called passivated carbides, WC-p. trap that removes NH3 and unreacted N2H4. The remaining

CATALYTIC DECOMPOSITION OF HYDRAZINE ON WC
3
gases, H2 or N2, were analyzed by gas chromatography
CG 90), using a thermal conductivity detector and a Car-
3
5000
(
boxen 1000 column maintained at 323 K.
30000
The reaction temperature for the decomposition of N2H4
on WC was varied between 303 and 333 K and the partial
pressure of hydrazine was fixed at 1.24 kPa for the calcu-
lation of the apparent activation energy. To determine the
reaction order with respect to N2H4, the partial pressure of
hydrazine was varied between 0.63 and 1.24 kPa by varying
the saturator temperature.
2
2
1
1
5000
0000
5000
0000
3
. RESULTS AND DISCUSSION
5
000
0
Both WC-p and WC-rp have the same hexagonal crystal
structure as that of WC reported in the literature. No other
phases, such as W, W2C, WO3, or WO2, were observed for
these solids (Fig. 1). In addition, the XRD patterns sug-
gest that flowing H2 at 1000 K for 1.5 h did not change
the bulk structure of the carbide and that the carbon re-
moved was only excess carbon on the surface of the sam-
0
20
40
60
80
100
Time / min
FIG. 2. Removal of the excess carbon from the WC-f surface.
ple. This result is in good agreement with a previous result
12).
The specific surface areas of the WC-f and WC-p were
about 13 m g , indicating that passivation had little in-
fluence on the surface structure of the sample containing
excess carbon. The formula of the WC-p sample calculated
from chemical analysis was WC1.14O0.07, suggesting that lit-
tle O is added to the sample surface after passivation. Thus,
excess carbon decreases the amount of adsorbed oxygen
and it can be used as a passivation layer (31).
To remove the excess carbon from the surface of the car-
bide, the standard procedure is to expose the solid to a flow
of H2 at high temperatures (21, 32). However, carbon from
the bulk may also be removed if the carbide is exposed to
flowing H2 at high temperatures for long periods of time
(32). Thus, in this work the excess carbon on the surface of
the WC-f was removed as CH4 by treating the carbide with
H2 at 1000 K for 1.5 h. The CH4 that evolved was measured
by gas chromatography (Fig. 2) using a thermal conductivity
detector.
(
a
2
−1
8
6
4
2
000
000
000
000
0
2
0
30
40
50
60
70
80
Diffraction Angle / 2θ
b
1
0000
Figure 2 shows the areas of the chromatographic peaks
corresponding to methane partial pressure versus reduc-
tion time. The methane partial pressure increases quickly
during the first 20 min, reaching a maximum after about
8
6
4
2
000
000
000
000
0
4
0 min. Then the amount of methane decreases slowly un-
til it reaches a constant pressure at about 80 min. Leclercq
et al. (32) found that a longer treatment of carbides with H2
can change their bulk structure. In this work, the carbide
samples were exposed to a flow of H2 for 90 min and no
change in the bulk structure was observed by XRD.
The specific surface area of the WC-rp sample was about
5 m g , a 90% increase from that of the WC-f sample.
2
−1
2
2
0
30
40
50
60
70
80
This fact suggests that the carbon removed was located on
the surface of the carbide, in agreement with the results
of XRD.
Diffraction Angle / 2θ
FIG. 1. X-ray diffraction patterns. (a) WC-p; (b) WC-rp.

4
SANTOS, VALEN C¸ A, AND RODRIGUES
No products were observed when N2H4 was flowed was similar to that of the solid prior to reaction, suggesting
through WC-f or WC-p at 313 K, suggesting that excess that the formation of nitrides did not occur or was restricted
carbon on the carbide surface prevented the decomposi- to the surface.
tion of N2H4. After excess carbon was removed from the
carbide surface (WC-fr and WC-rp samples), N2 and NH3 served for the passivated carbide could be due to a de-
reaction 2) were observed as products for the decomposi- crease in the area of the solid exposed to N2H4. Thus, al-
However, the larger initial decrease in reation rate ob-
(
tion of N2H4 at 313 K. This is in good agreement with the though the passivated carbide was reduced in flowing H2
results reported by Rodrigues (12), who also observed that at 623 K for 1 h prior to reaction, only a small fraction of
excess carbon totally prevents the decomposition of N2H4. the adsorbed O can be removed by H2 at this temperature.
In this work the reaction rate is defined as moles of N2H4 For example, Ribeiro et al. (33) found that only 10% of
decomposed per gram of catalyst per second. The initial the surface oxygen can be removed by H2 below 700 K.
value of the reaction rate is the same within the experimen- Thus, since there were O atoms on the surface of the pas-
tal error for passivated and nonpassivated tungsten carbide sivated carbides, H2O could be formed upon reaction with
samples. In addition, the reaction rate for the decomposi- N2H4. In that case, H2O would fill the micropores of the
tion of N2H4 on all carbide samples decreases with time carbide. In a similar experiment one of the authors ob-
(
Fig. 3). On passivated carbides, however, the decrease in served that after reaction at 313 K the surface area de-
the reaction rate is larger during the first 20 min of reaction, creased ca. 50%. However, when the sample was heated
while on nonpassivated samples (Fig. 3) the decrease in the in vacuo above 600 K the surface area had the same value
reaction rate is continuous over time. For example, after as that of the carbide prior to reaction (12), suggesting that
6
0 min the reaction rate on WC-rp is 30% lower than the H2O was formed during the reaction between N2H4 and ad-
initial value while on the WC-fr the reaction rate is 10% sorbed O. The blocking of micropores of α-WC with H2O
lower than the initial value. was also observed after titration of adsorbed O with H2 by
1
29
Since the rate has units of moles per gram per second it the decrease in intensity of the Xe NMR peak (34) and
is important to know whether the observed decrease is due in the reaction between H2 and O2 (33). Thus, although
to a change in the surface structure of the carbide or to a the formation of surface nitrides cannot be ruled out, none
decrease in the number of sites available for reaction. Al- of the techniques used in this work suggest this and a
thoughtheXRDpatternsofthecarbidesamplesusedinthis more sensitive surface technique would be necessary to ob-
work were not taken after reaction, in a similar experiment serve it. However, the decrease observed in the reaction
one of the authors observed that the XRD of passivated ratecanbeexplainedbyfillingthemicroporesofthecarbide
tungsten carbides did not change after reaction between with H2O formed during the reaction between N2H4 and
3
03 and 313 K compared to those of the solid prior to reac- surface O.
tion. However, when the reaction was done under the same The reaction order with respect to N2H4 was zero for
conditions as those used in satellite microthrusters (1100 K) the WC-rp catalyst, in good agreement with that reported
13), the XRD of the solid after reaction had patterns of by Rodrigues (12). A zero-order reaction suggests that the
(
tungsten nitrides in addition to the carbide patterns (12). carbide surface is covered with a strongly adsorbed species.
In addition, the chemical analysis after reaction at 313 K From the measured value for the initial rate of reaction at
−
7
−1 −1
2
−1
3
03 K (8.7 × 10 mol g
s ), the surface area (25 m g ),
and the value of surface density obtained by Ribeiro et al.
1
5
−2
(
33) (0.4 × 10 cm ), the estimated turnover rate for the
−
3
−1
1
1
1
1
1
0
0
.8
.6
.4
.2
.0
.8
.6
decomposition of N2H4 at 303 K is 5 × 10
s . From the
WC-rp
WC-fr
2 4
values for partial pressure of N H , activation energy of
adsorption (vide infra), and preexponential factor for ad-
sorption (37) it is possible to estimate the fraction of empty
−7
sites on the surface of the carbide as 7.7 × 10 , suggesting
an almost covered surface during reaction.
The reaction temperature was varied between 303 and
3
13 K for calculation of the apparent activation energy
−
7
(
Fig. 4). The initial rates varied between 8.7 × 10 and
−
6
−1 −1
2.5 × 10 mol g
s
and the activation energy for WC-rp
−
1
was about 50 kJ mol . Ribeiro et al. (33) observed that
NH3 adsorbs strongly on the surface of WC and that it des-
orbed from the surface as N2with a temperature maximum
of 1200 K. Thus, a rough estimation of the enthalpy of ad-
1
0
20
30
40
50
60
70
Time / minutes
FIG. 3. Rate of decomposition of hydrazine on WC-fr and WC-rp at
13 K.
−
1
−1
3
sorption of N atoms, QN, is 300 kJ mol
K .

CATALYTIC DECOMPOSITION OF HYDRAZINE ON WC
5
∗
∗
∗
∗
1
1
0
0
.5
.0
.5
.0
NH + NH → NH3 + NH +
[9]
[10]
[11]
2
2
∗
∗
∗
NH + H → NH3 + 2
NH + H → NH +
N + N → N + 2
2
∗
∗
∗
∗
2
∗
∗
∗
2
[12]
[13]
[14]
∗
∗
∗
∗
∗
N + H → NH +
∗
∗
∗
N + NH → N2 + H + .
-
-
0.5
The activation energies for steps 6 through 11 are each
−
1
equal to zero. In addition, for QN less than 450 kJ mol the
activation energy for the combination of two adsorbed N
atoms is also zero. Thus, as adsorbed NHx fragments form
on the carbide surface they react to form either NH3 or ad-
sorbed N. Once NH3 is formed, it is desorbed from the car-
bide surface while adsorbed N accumulates on the surface.
1.0
3
.05
3.10
3.15
3.20
3.25
3.30
3.35
-
1
-3
-1
T
/ 1.0 x 10 K
FIG. 4. Arrhenius plot for catalyst WC-rp.
Using this value we estimated the activation energy for In addition, no H2 was observed as a product and the esti-
some elementary steps of the catalytic decomposition of mated activation energy for the bimolecular desorption of
−
1
N2H4 (35) based on the bond order conservation method H2 is ca. 100 kJ mol . Thus, as reaction proceeds, the prob-
of Shustorovich (36). The first step in the decomposition of ability of finding two neighboring NHx species or adsorbed
N2H4 is a dissociative adsorption on a bare surface. N2H4 H decreases and the only possible steps on the surface are
may be adsorbed according to
those involving adsorbed N. These possible steps include
the reaction between adsorbed N and adsorbed H (step 13)
or adsorbed NH (step 14). The activation energy for step
∗
∗
∗
N2H4 + 2 → N2H + H
[3]
[4]
3
−
1
−1
1
3 is 56 kJ mol and for step 14 it is 44 kJ mol , values
or
−
1
∗
∗
∗
close to the 50 kJ mol value we measured in this work. In
addition, the activation energy for the reverse reaction for
step 14 is 223 kJ mol , a value that suggests that this step
is irreversible. Thus, it seems that the rate-determining step
involves adsorbed N but is not the bimolecular desorption
of adsorbed N.
N2H4 + 2 → NH + NH
2
2
−
1
where ∗ represents an empty surface site.
Of these two steps the more favorable is the breaking of
one N–H bond on the surface to form adsorbed N2H3 and
−1
H (Ea = 67 kJ mol ). Once adsorbed N2H3 is formed it
may react with adsorbed H to form NH3 and NH or it may
break one N–N bond or one N–H bond. The activation
4. CONCLUSION
−
1
energy for reaction with adsorbed H is 0 kJ mol while
the activation energies for the breaking of the N–N bond is
WC was prepared by a temperature-programmed reac-
tion of WO3 with a 20%v/v CH4/H2 mixture. We were able
to prepare ca. 0.5 g of tungsten carbide reproducibly, as
confirmed by XRD, BET surface area measurements, and
chemical analysis. This reproducibility is an advance in the
preparation of tungsten carbides as it is difficult to control
the solid state reaction (WO3 → WC). Indeed, there is very
little work in the literature on the preparation of the tung-
sten carbide samples of more than 0.4 g.
Excess carbon on the carbide surface reduced the specific
surface area by 40%. Samples of WC with excess carbon on
the surface showed no catalytic activity during the decom-
position of N2H4 at 313 K, suggesting that excess carbon
totally inhibits the reaction.
−1
−1
1
28 kJ mol and that for the N–H bond is 67 kJ mol .
Thus, even on a bare surface the most favorable step for
decomposition of adsorbed N2H3 is the reaction with H to
form NH3 and adsorbed NH. Details about the calculation
of the activation energy for different steps in the catalytic
decomposition of N2H4 can be found elsewhere (35).
Adsorbed NH may react with adsorbed H, NH, or NH2 to
form adsorbed NH2, adsorbed N and NH2, or adsorbed N
−
1
and NH3, respectively. All these steps have Ea = 0 kJ mol
.
Adsorbed NH2 may react with adsorbed H or NH2 to form
NH3 and NH3 and adsorbed NH. The activation energy for
these steps is also equal to zero. The possible steps are as
follows:
∗
∗
∗
RemovalofexcesscarboninflowingH2 at1000Kfor1.5h
did not change the hexagonal structure of the carbide and
N2H4 + 2 → N2H + H
[5]
[6]
[7]
[8]
3
∗
∗
∗
∗
2
−1
N2H + H → NH3 + NH +
3
resulted in solids with a specific surface area of 25 m g
.
∗
∗
∗
∗
The only products of decomposition of N H on WC-fr and
2
4
NH + NH → NH + H
2
WC-rp were N2 and NH3, suggesting that this reaction takes
place on the same sites on the surface of both passivated
∗
∗
∗
∗
NH + NH → NH3 + N +
2

6
SANTOS, VALEN C¸ A, AND RODRIGUES
and fresh carbides. Although the initial rate of reaction was 13. Rodrigues, J. A. J., Cruz, G. M., Bugli, G., Boudart, M., and Dj e´ ga-
Mariadassou, G., Catal. Lett. 45, 1 (1997).
the same for both carbides, passivated catalysts deactivated
1
4. Toth, L. E., in “Transition Metal Carbides and Nitrides” (John L.
Margrave, Series Ed.), p. 29. Academic Press, New York, 1971.
5. Boudart, M., and Ptak, L. D., J. Catal. 16, 90 (1970).
more rapidly due to formation of H2O by the reaction of ad-
sorbed O and N2H4. Analysis of kinetic data and estimated
activation energy of elementary steps suggest a surface
1
1
6. Oyama, S. T., Ph.D. dissertation, Stanford University, 1981.
covered with N during reaction and a rate-determining step 17. Volpe, L., and Boudart, M., J. Solid. State Chem. 59, 332 (1985).
1
1
8. Volpe, L., and Boudart, M., J. Solid. State Chem. 59, 348 (1985).
9. Iglesia, E., Baumgartner, J. E., Ribeiro, F. H., and Boudart, M., J. Catal.
involving these species.
Finally, the decomposition of N2H4 on the tungsten car-
bides prepared in this work has values of activation energy
and order of reaction similar to those found for commercial
131, 523 (1991).
2
0. Iglesia, E., Ribeiro, F. H., Boudart, M., and Baumgartner, J. E., Catal.
Today 15, 455 (1992).
catalysts, indicating that carbides are promising catalysts for 21. Ribeiro, F. H., Dalla Betta, R. A., Boudart, M., Baumgartner, J. E.,
and Iglesia, E., J. Catal. 130, 86 (1991).
space applications.
2
2
2
2. Ribeiro, F. H., Boudart, M., Dalla Betta, R. A., and Iglesia, E., J. Catal.
30, 498 (1991).
1
ACKNOWLEDGMENTS
3. Blekkan, E. A., Cuong, P. H., Ledoux, M. J., and Guille, J., Ind. Eng.
Chem. Res. 33, 1657 (1994).
4. Khomenko, A. A., and Apel’Baum, L. O., Kinet. Katal., 17(3), 691
J.B.O.S. acknowledges the financial support received from FAPESP
(Grant N. 96/2108-6). We are also grateful to INPE (LCP), where this
(
1976).
work was conducted.
2
2
5. Contour, J. P., and Pannetie, G., J. Catal. 24, 434 (1972).
6. Szware, M., Proc. R. Soc. London 198, 267 (1949).
REFERENCES
. Schmidt, E. W., in “Hydrazine and Its Derivatives: Preparation, Prop-
27. Maurel, R., and Menezo, J. C., J. Catal. 51, 293 (1978).
28. Alberas, D. J., Kiss, J., Liu, Z. M., and White, J. M., Surf. Sci. 278, 51
1
(1992).
erties, Applications” (E. W. Schmidt, Ed.), Chap. 6, p. 714. Wiley-
Interscience, New York, 1984.
29. Brunauer, S., Emmet, P. H., and Teller, E. J., J. Am. Chem. Soc. 60, 309
(1938).
2
. Maurel, R., Menezo, J. C., and Barrault, J., J. Chim. Phys. 70, 1221
30. Lange, N. A., in “Handbook of Chemistry” (John A. Dean, Ed.), 13th
ed., pp. 10–32. McGraw Hill, New York, 1985.
(1973).
3
4
5
6
. Sawin, H. H., and Merrill, R. P., J. Chem. Phys. 73(2), 996 (1980).
. Wood, B. J., and Wise, H., J. Catal. 39, 471 (1975).
. Prasad, J., and Gland, J. L., Langmuir 7, 722 (1991).
. Alberas, D. J., Kiss, J., Liu, Z. M., and White, J. M., Surf. Sci. 278, 51
31. Keller, V., Wehrer, P., Garin, F., Ducros, R., and Maire, G., J. Catal.
153, 9 (1995).
32. Leclercq, G., Kamal, M., Giraudon, J. M., Devassine, P., Feigenbaum,
L., Leclercq, L., Frennet, A., Bastin, J. M., Lofberg, A., Decker, S.,
and Dufour, M., J. Catal. 158, 142 (1996).
(1992).
7
8
. Grunze, M., Surf. Sci. 81, 603 (1979).
. Cosser, R. C., and Tompkins, F. C., Trans. Faraday Soc. 67, 526
33. Ribeiro, F. H., Dalla-Betta, R. A., Guskey, G. J., and Boudart, M.,
Chem. Mater. 3, 805 (1991).
(
1971).
34. Valen c¸ a, G. P., Ph.D. dissertation, Stanford University, 1990.
35. Medeiros, J. E., and Valen c¸ a, G. P., Brazilian J. Chem. Eng. 15, 126
(1998).
◦
9
. Valen c¸ a, G. P., in “Anais do 7 Semin a´ rio Brasileiro de Cat a´ lise,” Vol. 1,
p. 375. IBP, Intituto Brasibin de Peh o´ leo, Rio de Janeiro, 1993.
1
1
0. Huang, S. X., Rufael, T. S., and Gland, J. L., Surf. Sci. 290, L673 (1993).
1. Gland, J. L., Fisher, G. B., and Mitchell, G. E., Chem. Phys. Lett. 119(1),
36. Shustorovich, E., Adv. Catal. 37, 101 (1990).
37. Boudart, M., and Dj e´ ga-Mariadassou, G., in “Kinetics of Hetero-
geneous Catalytic Reactions” (John M. Prausnitz and Leo Brewer,
Series Eds.), p. 74. Princeton University Press, Princeton, NJ,
1984.
8
9 (1985).
1
2. Rodrigues, J. A. J., Th e` se de Doctorat, Universit e´ Paris VI Pierre et
Marie Curie, Paris, 1996.