Rhodium Supported on Silica-Stabilized Alumina for Catalytic Decomposition of N₂O

Abstract

Silica-stabilized alumina calcined at 1200 °C has been used as a support for rhodium catalysts, and tested in catalytic decomposition of N₂O propellant. Significant enhancement of catalytic performance was obtained on the silica-stabilized catalyst owing to the thermally stable structure favoring the stabilization of Rh⁰ species and desorption of oxygen during the decomposition process.

Graphical Abstract

Silica-stabilized alumina calcined at 1200 °C has been used as a support for rhodium catalysts, and tested in catalytic decomposition of N₂O propellant. Significant enhancement of catalytic performance was obtained on the silica-stabilized catalyst owing to the thermally stable structure favoring the stabilization of Rh⁰ species and desorption of oxygen during the decomposition process.

1 Introduction

Nitrous oxide (N₂O) has received growing interest as a promising green propellant for small satellite propulsion due to its low toxicity, self pressurizing, and multi-mode propulsion [1, 2]. The earlier successful experiences of using N₂O as a propellant within a resisto-jet propulsion system on the UoSAT-12 mini-satellite and as the oxidizer for the hybrid engine powering “Spaceship One” have shown the great potential and feasibility of N₂O as a green propellant for modern advanced propulsions. However, thermal decomposition of gaseous N₂O requires high power input due to the high activation energy of ca. 250 kJ/mol, and it leads to a temperature rise over 1000 °C because of its highly exothermic nature. Therefore, there is an urgent need to develop novel catalyst materials which possess high catalytic activities at low temperatures and thermal stabilities at high temperatures for N₂O propellant applications.

Alumina has been widely used as a catalyst or catalyst supports due to its large surface area, tunable surface property, and suitable interaction with the active metal components. There have been some reports about the alumina-supported metal catalysts for the decomposition of N₂O propellant. For example, Wallbank et al. [3] reported that Shell 405 (36 wt% Ir/Al₂O₃) exhibited a high activity in the decomposition of N₂O propellant. Zakirov et al. [4] investigated more than 50 different catalysts for the decomposition of pure N₂O for small spacecraft propulsion, and found that Rh/Al₂O₃ had the highest turnover frequency.

On the other hand, it is known that γ-Al₂O₃ is unstable in high-temperature processes; phase transformation from γ-Al₂O₃ to α-Al₂O₃ often occurs at above 1000 °C accompanied with a dramatic decrease in surface area, which results in severe deactivation of the catalysts. Therefore, it is of vital importance to improve the thermal stability of alumina support for the catalytic decomposition of N₂O propellant. Addition of promoters to alumina is an effective method for suppressing the surface area loss and avoiding transition to α-Al₂O₃ induced by high-temperature reactions [5–13]. Beguin et al. [9] reported that numerous cations, such as La³⁺, Zr³⁺, Ca²⁺, and Th⁴⁺, were extremely effective for retarding the sintering and phase transformation of alumina at high temperatures. Courthéoux et al. [10] prepared silica-doped alumina aerogels by a sol–gel method and demonstrated a marked increase in the thermal stability after calcination at 1200 °C, owing to silicon atoms stabilizing the alumina defect structure. Osaki et al. [13] also found that silica-doped alumina cryogels exhibited a high thermal stability by virtue of the synergetic effects of the very low bulk density and the dopant. These previous studies have shown that doping alumina with silicon atoms could improve the thermal stability of the materials. However, it is unknown if such doping process is helpful in promoting the activity at low reaction temperatures because the doping process also causes changes of surface properties which will influence the chemical states of the adsorbed metal particles.

Previously, we reported that mullite, a type of ceramic material with the chemical composition of 3Al₂O₃·2SiO₂, was a very good support of Rh catalysts in the decomposition of N₂O propellant [14]. More interestingly, the low-temperature activity was also enhanced together with the remarkably improved high-temperature stability as a result of the formation of mullite. In the present work, we prepared a series of silica-stabilized alumina materials with different silica concentrations, and investigated their catalytic performances for the catalytic decomposition of N₂O propellant. The promotional effect of silicon-doping was also studied in detail by using a variety of characterization techniques.

2 Experimental

2.1 Catalyst Preparation

The silica-stabilized alumina samples were prepared by co-precipitation method, using Al(NO₃)₃·9H₂O and tetraethylorthosilicate (TEOS) as starting reactants and (NH₄)₂CO₃ as a precipitant. In a typical synthesis, 2.1 g of TEOS was dissolved in 10 mL ethanol, and this ethanol solution was mixed with 50 mL of an aqueous solution of aluminum nitrate (1 mol/L). To the above mixed solution, 20 mL of an aqueous solution of (NH₄)₂CO₃ (5 mol/L) was added and followed by vigorous stirring at 60 °C for 3 h. The precipitate was then filtered, dried at 120 °C for 64 h, and calcined in air at 1200 °C for 4 h. The concentration of SiO₂ in this sample was 19 wt%, and denoted as Si_0.19Al. For comparison, samples with concentrations of SiO₂ of 5 and 32 wt% were also prepared with the same procedure, and denoted as Si_0.05Al and Si_0.32Al, respectively. In some cases, the Si_0.19Al sample was calcined at different temperatures (T), and denoted as Si_0.19Al-T.

To investigate the effect of different dopants, La₂O₃-, CeO₂-, and ZrO₂-doped alumina samples were also prepared by co-precipitation with La(NO₃)₃·6H₂O, Ce(NO₃)₃·6H₂O, and Zr(NO₃)₃·4H₂O as the precursors of dopants, and denoted as La_0.19Al, Ce_0.19Al, and Zr_0.19Al, respectively.

The Rh-supported catalysts were prepared by incipient wetness impregnation of the above-prepared supports with an aqueous solution of RhCl₃ to give a calculated Rh loading of 4.4 wt%. The catalysts were calcined at 500 °C for 4 h. For comparison, Rh/γ-Al₂O₃ was also prepared with the same procedure.

Finally, in order to investigate the thermal stability of the catalysts, the prepared catalysts were further calcined at 1200 °C for 4 h, yielding the catalysts named as 1200-Rh/Si_xAl and 1200-Rh/Al₂O₃.

2.2 Catalyst Characterization

BET surface areas of the catalysts were measured by N₂ adsorption at −196 °C using a Micromeritics ASAP 2010 apparatus. The X-ray diffraction (XRD) patterns were recorded with a PANalytical X’Pert-Pro powder X-ray diffractometer, using Cu Kα monochromatized radiation (λ = 0.1541 nm) at a scan speed of 5°/min.

Rh dispersion was derived from the measurement of H₂ chemisorption uptake at room temperature using a Micromeritics AutoChem 2090 apparatus. Before each experiment, catalysts were reduced with H₂ at 500 °C for 2 h. Temperature programmed reduction (TPR) experiments were carried out on the same apparatus. 100 mg of a catalyst was loaded into a U-shape quartz reactor and pretreated in Ar at 500 °C for 2 h. The samples were cooled to room temperature and then the flowing gas was switched to a 10 vol% H₂/Ar. The catalysts were heated to 500 °C at a ramping rate of 10 °C/min.

Oxygen adsorption experiments were performed with a BT2.15 heat-flux calorimeter. Prior to the measurement, catalyst was preheated in a special treatment quartz cell in H₂ from room temperature to 500 °C and held at that temperature for 2 h. Then, the sample was outgassed in situ in high vacuo (3 × 10⁻⁴ Pa) at 500 °C for 0.5 h. The microcalorimetric data were collected by sequentially introducing small doses (1–10 μmol) of O₂ onto the sample until it became saturated (5–6 Torr).

Transmission infrared spectra were obtained with a Bruker Equinox 55 spectrometer at a resolution of 4 cm⁻¹ using 64 scans. The self-supporting wafers (14 mg) were prepared and placed in an IR cell with NaCl windows and a temperature controller. Each sample was pretreated with H₂ at 500 °C for 45 min and then evacuated at that temperature for 15 min. The CO adsorption was carried out at room temperature under 3 Torr CO. The spectra were recorded ca. 10 min after CO adsorption.

2.3 Catalytic Activity Tests

Catalytic decomposition of N₂O over supported rhodium catalysts was carried out in a fix-bed flow reactor system at atmospheric pressure. 0.1 g of a catalyst (20–40 mesh in particle size) was mixed with 0.40 g of quartz to prevent temperature gradient, and loaded in a silica glass tube reactor of 6 mm i.d. with silica wool. The reaction temperature was measured by thermocouples located at the inlet of the catalyst bed. Prior to each run, the catalysts were in situ reduced with H₂ at 500 °C for 2 h. The reacting gas containing N₂O (30 vol%) in Ar was then introduced in the reactor at a flow rate of 50 mL/min. Reactant and product concentrations were measured by on-line gas chromatography (Agilent GC-6890N) with a Porapak Q column (for N₂O) and a 13 X column (for N₂ and O₂). N₂ and O₂ were identified as the only products (Fig. S1). N₂O conversion was determined based on the difference between its inlet and outlet concentrations.

3 Results and Discussion

3.1 Metal Oxide-Stabilized Alumina as Support for Rhodium Catalysts

Figure 1 compares the N₂O conversions as a function of reaction temperature over metal oxide-stabilized alumina supported rhodium catalysts. The referenced sample, Rh/γ-Al₂O₃, exhibited a reasonable activity compared to what have ever been reported under the same reaction conditions [14–16]. Decomposition started at 300 °C and complete decomposition was attained at 500 °C. After metal oxides of Si, La, Zr, and Ce were introduced, the resulting catalysts exhibited different catalytic activities in N₂O decomposition. The Rh/Si_0.19Al showed the highest activity, shifting the N₂O conversion curve toward a lower temperature of 100 °C. The Rh/La_0.19Al catalyst also showed a slight increase in activity, especially in the low temperature range. In contrast, the Rh/Zr_0.19Al and Rh/Ce_0.19Al samples demonstrated even lower activities than the Rh/γ-Al₂O₃. Obviously, introducing metal oxides to the alumina support has a significant influence on its characteristics as well as the subsequent catalyst. Therefore, further studies into the understanding of the doping effect are of great importance. With a view to the most obvious promotion effect in the decomposition, SiO₂ was selected as the doping material for the following research.

Fig. 1

N₂O conversion as a function of temperature over the Rh/M_xAl catalysts: Rh/Si_0.19Al (filled square), Rh/La_0.19Al (filled inverted triangle), Rh/Zr_0.19Al (filled triangle), Rh/Ce_0.19Al (filled circle), and Rh/γ-Al₂O₃ (filled diamond) (the M_xAl supports were calcined at 1200 °C, while γ-Al₂O₃ was calcined at 500 °C)

3.2 Effect of the Si Content

Figure 2 illustrates the N₂O conversion of the Rh/Si_xAl catalysts with different silica content. It can be seen that even a small amount of silica had a significant promotion effect on the catalytic activity. For example, the Rh/Si_0.05Al catalyst showed a much higher activity than the Rh/γ-Al₂O₃, starting the decomposition at 200 °C and reaching 53% N₂O conversion at 350 °C. The silica content affected the N₂O conversion in the following sequence: Rh/γ-Al₂O₃ < Rh/Si_0.32Al < Rh/Si_0.05Al < Rh/Si_0.19Al. Clearly, it appeared that there was an optimum amount of silica for the Rh/Si_xAl catalysts in the decomposition of N₂O. In the present work, the Rh/Si_0.19Al was assumed to possess the best chemical composition for the limited catalysts studied.

Fig. 2

N₂O conversion as a function of temperature over the Rh/Si_xAl catalysts with different silica content: Rh/Si_0.19Al (filled square), Rh/Si_0.05Al (filled triangle), Rh/Si_0.32Al (filled circle), Rh/γ-Al₂O₃ (filled asterisk)

The BET surface areas of the supports, fresh catalysts, and used catalysts are summarized in Table 1. A slight decrease in BET surface areas of the supports after depositing rhodium was observed on all the samples, which may be caused by the blockage of some pores by rhodium. Rh dispersions increased with increasing the BET surface areas. The γ-Al₂O₃ calcined at 500 °C had the highest specific surface area of 265 m²/g, and the highest Rh dispersion of 83% in accordance. After calcination at 1200 °C, α-Al₂O₃ was formed accompanied with a marked surface area loss to about 7 m²/g. In contrast, the silica-stabilized samples possessed rather high surface area (above 50 m²/g) after calcination at the same temperature of 1200 °C. We also examined the BET surface areas and Rh dispersions of the catalysts after activity test. No significant modification of the support and active phase occurred during the course of the experiment (Figs. S2, S3). In correlation with the catalytic performances above, we could propose that the BET surface area and the active component dispersion are not primary parameters for the activity enhancement. The modification of the support properties by doping silica probably played a key role on the decomposition of N₂O.

Table 1 Physicochemical properties of the Rh/Si_xAl and Rh/γ-Al₂O₃ catalysts

XRD patterns of the Si_xAl supports appear in Fig. 3. For the Si_0.05Al and Si_0.32Al samples, the typical diffraction peaks of θ-Al₂O₃ were detected along with traces of α-Al₂O₃. While in the case of Si_0.19Al, it appeared only the peaks of θ-Al₂O₃. According to the previous work, the γ-Al₂O₃ was completely transformed into α-Al₂O₃ during the thermal treatment at 1200 °C [14]. Clearly, the addition of 19 wt% silica strongly suppressed the phase transformation of alumina from θ to α. To be noted, the phase transformation of transition alumina is a process of dehydroxylation, resulting in the generation of oxygen vacancies [17, 18]. To be noted, among the different types of alumina (γ, δ, θ, α), oxygen vacancies are thought to be most abundant in the θ-Al₂O₃ [17]. It has been generally accepted that oxygen desorption is the rate determining step during N₂O decomposition [19]. Therefore, the oxygen vacancies on the support are believed favorable for the mobility of oxygen derived from N₂O decomposition, which led to the superior activity of the Rh/Si_0.19Al catalyst, in agreement with a previous report [20]. In order to verify the above proposition, we also tested the effect of calcination temperature of alumina support on the performance of Rh catalysts supported thereof. As shown in Fig. 4, a gradual improvement of the catalytic performance was also observed when the calcination temperature was raised from 500 to 1100 °C. XRD patterns of the Al₂O₃-1100 sample indicated that there existed a significant amount of θ-Al₂O₃ after calcination at 1100 °C, as illustrated in supplementary Fig. S4. This result further confirmed that the formation of θ-Al₂O₃ contributes significantly to the enhanced activity, as discussed above.

Fig. 3

XRD patterns of the Si_xAl supports calcined at 1200 °C: Si_0.05Al (a), Si_0.19Al (b), Si_0.32Al (c); α-Al₂O₃ (filled inverted triangle), θ-Al₂O₃ (filled circle)

Fig. 4

N₂O conversion as a function of temperature over the Rh/Al₂O₃-T catalysts: Rh/Al₂O₃-500 (filled square), Rh/Al₂O₃-1000 (filled circle), Rh/Al₂O₃-1100 (filled triangle), Rh/Al₂O₃-1200 (filled inverted triangle)

Another possibility connecting the high performance on the Rh/Si_0.19Al may involve the reducibility of the rhodium. H₂-TPR technique is known as a useful tool for gauging this property. Each TPR profile in Fig. 5 showed only a single reduction peak, which could be assigned to the reduction of Rh [21]. The rhodium on the SiO₂ stabilized supports, with the reduction peaks below 56 °C, were more easily to be reduced compared with that on the alumina (95 °C). As has been mentioned before [14], the easy reduction of Rh species may closely correlate with the appearance of oxygen vacancies. According to literatures, it is generally believed that N₂O decomposition takes place via a redox cycle [19, 20]. Therefore, the Rh⁰ species involve a redox cycle according to the following steps:

$$ {\text{N}}_{2} {\text{O}} + {\text{Rh}} \to {\text{RhO}}_{x} + {\text{N}}_{2} $$

(1)

$$ {\text{RhO}}_{x} \to {\text{Rh}} + {\text{O}}_{2} $$

(2)

Fig. 5

H₂-TPR profiles of the Rh/Si_xAl catalysts with different silica content: Rh/Si_0.19Al (a), Rh/Si_0.32Al (b), Rh/Si_0.05Al (c), Rh/γ-Al₂O₃ (d) (the M_xAl supports were calcined at 1200 °C, while γ-Al₂O₃ was calcined at 500 °C)

The easily reduced rhodium species is in favor of the stability of active center in N₂O decomposition, which leads to a great enhancement of catalytic activity.

3.3 Effect of Calcination Temperature of the Si_0.19Al Support

In general, the crystal phase of alumina changes with increasing temperature in the following order: γ → δ→θ → α. To detect the effect of alumina phase structure on the catalytic activity of N₂O decomposition, the Si_0.19Al precursor was calcined at different temperatures. Figure 6 shows the effect of calcination temperature of supports on the catalytic activity of the catalysts. A gradual improvement of the catalytic performance was observed with increase of the calcination temperature. The Rh/Si_0.19Al-500 has almost the same activity as the Rh/γ-Al₂O₃, with the temperature for complete N₂O decomposition as high as 500 °C. The Rh/Si_0.19Al-1200 catalyst was most active, indicating the onset of decomposition at about 200 °C and 93% of N₂O conversion at 350 °C. The behaviors of the Rh/Si_0.19Al-800 and Rh/Si_0.19Al-1000 catalysts are intermediate, and the Rh/Si_0.19Al-800 is less active than the Rh/Si_0.19Al-1000. Clearly, the thermal treatment before metal deposition affected remarkably the catalytic activity of the rhodium catalysts.

Fig. 6

N₂O conversion as a function of temperature over the Rh/Si_0.19Al-T catalysts: Rh/Si_0.19Al-500 (filled inverted triangle), Rh/Si_0.19Al-800 (filled triangle), Rh/Si_0.19Al-1000 (filled circle), Rh/Si_0.19Al-1200 (filled square)

Table 2 illustrates the BET surface areas of the Si_0.19Al precursor calcined at different temperatures, the fresh catalysts, and the used catalysts, as well as Rh dispersion data. The Si_0.19Al-500 had the largest surface area of 444 m²/g. A further increase of calcination temperature of the support to 1200 °C induced a decrease in surface area to a minimum value of 55 m²/g. It is also observed that the Rh dispersion decreased monotonously with the elevation of the calcination temperatures, which is in contrary to the trend of the catalytic activity. Apparently, the N₂O decomposition activity of the Rh/Si_xAl catalyst does not depend on the BET surface area of the support and the Rh dispersion, in agreement with the above results.

Table 2 Physicochemical properties of Rh/Si_0.19Al-T catalysts

Influence of calcination temperature of the support on the reduction behaviors of Rh species were also investigated by H₂-TPR, as shown in Fig. 7. Rh/Si_0.19Al catalysts gave one reduction peak which corresponds to the reduction of oxidized Rh to metallic Rh, the same as the previous results. It should be noted that the increase of calcination temperature of the support caused an increase in the reducibility of rhodium, with lower temperatures and higher amount of H₂ consumption (Table 2), which may be related to the change of support structure during phase transformation. After the calcination at 1200 °C, the Si_0.19Al displayed the θ-Al₂O₃ structure, which contains the most abundant oxygen vacancies. Thus, the Rh/Si_0.19Al-1200 sample showed better reducibility than the Rh/Si_0.19Al-500 one. By comparing the TPR results of the catalysts with the catalytic activity, it is worthy of noting that the easily reduced rhodium oxide should be responsible for the enhanced activity of the Rh/Si_0.19Al-T catalysts, again exemplified the assumption above.

Fig. 7

H₂-TPR profiles of the Rh/Si_0.19Al catalysts: Rh/Si_0.19Al-500 (a), Rh/Si_0.19Al-800 (b), Rh/Si_0.19Al-1000 (c), Rh/Si_0.19Al-1200 (d)

3.4 IR and O₂ Adsorption Microcalorimetry Characterization of the Rh/Si_0.19Al Catalyst

The promotion effect of silica doping on the N₂O decomposition was studied in more detail by IR and O₂ adsorption microcalorimetry characterization on the Rh/Si_0.19Al catalyst. The Rh species existed in the oxidation state on the as-prepared Rh/Si_0.19Al and Rh/Al₂O₃ catalysts. The IR spectra of CO adsorption at 298 K on the Rh/Si_0.19Al catalyst reduced by H₂, and pre-oxidized by N₂O are shown in Fig. 8A. For the reduced sample, there were four bands, located at 2097, 2069, 2028, and 1881 cm⁻¹, clearly observed on the Rh/Si_0.19Al catalyst. The twin bands at 2097 and 2028 cm⁻¹ are attributed to the symmetric and anti-symmetric stretching mode, respectively, of Rh^δ+(CO)₂ species, as already evidenced by previous reports [22–24]. The band at 2069 cm⁻¹, ascribed to linear carbonyl species, and at 1881 cm⁻¹, due to bridging carbonyls, both adsorbed on metallic rhodium particles. Clearly, the linear Rh⁰-CO species prevailed on the reduced sample. Similar CO adsorption bands were observed after exposure to N₂O. The linear Rh⁰-CO species continued to be the major adsorption state of CO, with some weak adsorption of gem-dicarbonyl Rh^δ+ and bridged carbonyl species as well. The general trend observed with increasing adsorption temperature of N₂O was that linear carbonyls species remained unchanged. One possible reason is that Rh⁰ species exhibited a good stability under reaction conditions. Besides that, the RhO_x species formed during the N₂O adsorption were unstable and easily reduced in the vacuum pre-treatment. However, as shown in Fig. 8B, the Rh/γ-Al₂O₃ mainly comprised of Rh^δ+(CO)₂ species. This information allows us to conclude that the silica-stabilized alumina favors the stabilization of Rh⁰ species and further suggests that Rh⁰ species are the active sites in the decomposition of N₂O.

Fig. 8

FT-IR spectra of the Rh/Si_0.19Al (A) and the Rh/γ-Al₂O₃ (B) samples arising from CO adsorption at room temperature submitted to the following pretreatments: fresh catalysts (a); reduced at 500 °C in hydrogen (b); and outgassed at 25 °C after treatment with N₂O: at 100 °C (c), at 200 °C (d), at 300 °C (e), at 400 °C (f)

In order to obtain a clear understanding of the effect of silica doping on the oxygen mobility, adsorption microcalorimetry of O₂ was accomplished on the samples. Figure 9 shows the oxygen adsorption volumetric isotherms on the silica-stabilized samples. Prior to O₂ adsorption, the samples were reduced with H₂ and then saturated with N₂O. The initial differential heats of O₂ adsorption on the four samples were quite similar, around 264 kJ/mol. However, the capacity of oxygen adsorption of the Rh/Si_0.05Al, Rh/Si_0.19Al, Rh/Si_0.32Al, and Rh/γ-Al₂O₃ catalysts were 0.58, 0.79, 0.44, and 0.25 μmol/m², respectively, well in accordance with the activity sequence of the catalysts. These results showed that there were more O₂ adsorption sites on the Rh/Si_0.19Al catalyst than those of the other Rh catalysts. It should be noted that some of the adsorption sites on the catalyst have been occupied by the oxygen generated in the N₂O decomposition during the N₂O pre-adsorption process. Thus, the O₂ adsorption in this case could be regarded as the reversible oxygen adsorption, which directly reflected the migrating ability of oxygen on the catalyst.

Fig. 9

Differential heat of O₂ adsorption as a function of O₂ coverage on Rh catalysts: Rh/Si_0.19Al (filled square), Rh/Si_0.05Al (filled asterisk), Rh/Si_0.32Al (filled circle), Rh/γ-Al₂O₃ (open triangle). The samples were reduced with H₂ and then saturated with N₂O prior to O₂ adsorption

Owing to a process of dehydroxylation in the phase transformation, oxygen vacancies are the most abundant in Si_0.19Al, which greatly facilitates the migration of oxygen species from the active sites to the support favoring its final desorption from the catalyst [25]. Therefore, the support structure may be responsible for the great capacity of oxygen adsorption of the Rh/Si_0.19Al. It has been mentioned previously that oxygen desorbing from the catalyst surface is the rate determining step for N₂O decomposition. Therefore, the oxygen migrating ability of the Rh/Si_0.19Al was thus assumed to be one main reason for the activity enhancement of N₂O decomposition.

3.5 Thermal Stability of the Rh/Si_xAl Catalysts

The thermal stabilization of a catalyst is another key parameter for evaluating its feasibility in the decomposition of N₂O as a propellant. Figure 10 presents the N₂O conversions of the Rh supported catalysts calcined at 1200 °C. The activity of the alumina supported catalyst was decreased dramatically after calcination at 1200 °C, with the N₂O conversion of 20% at 600 °C, which could be correlated with the phase transformation of alumina and the sintering and/or vaporization of active components. In contrast, the Rh/Si_xAl catalysts were only slightly affected by this treatment. Even after calcination at 1200 °C, the Rh/Si_0.19Al catalyst still initiated the N₂O decomposition at 350 °C and attained 95% conversion at 500 °C. It appears that the doping of silica not only strongly enhanced the catalytic activity of the Rh/Al₂O₃ catalyst, but also significantly improved the thermal stability of the catalyst.

Fig. 10

N₂O conversion as a function of temperature over the Rh/Si_xAl catalysts calcined at 1200 °C: 1200-Rh/Si_0.19Al (filled square), 1200-Rh/Si_0.05Al (filled triangle), 1200-Rh/Si_0.32Al (filled circle), 1200-Rh/Al₂O₃ (open diamond)

4 Conclusion

Alumina modified with silica showed a great enhancement in the activity of N₂O decomposition over the Rh catalyst supported thereof. The silica doping favors the stabilization of Rh⁰ species and the desorption of oxygen species under reaction conditions. It can be concluded that the doping of alumina with silica is an effective way to improve the catalytic activity of N₂O decomposition for propellant applications: an excellent catalytic activity at low temperatures associated with a high-temperature stability.