Abstract

Different compositions of Co3O4 catalysts on CeO2-ZrO2 solid solution (Ce0.9Zr0.1O2 and Ce0.8Zr0.2O2) have been studied for the oxidation of propylene. The optional amount of Co3O4 active phase on CeO2-ZrO2 support of 30 wt% was found. The mixed Co3O4-CeO2-ZrO2 with the same composition of the optimal supported ones showed approximately the same activity, which was not higher than the activity of the mixed Co3O4-CeO2 catalyst. Catalytic activities of mixed Co3O4-CeO2 with different loading contents supported on high surface area supports (Al2O3, SiO2) were then measured. The optimal composition of active phase was still 30 wt% but the minimum temperature of the highest activity increased to above 300°C due to the inert nature influence of the support.

1. Introduction

Cobalt-based oxides are one of the most active catalysts in catalytic oxidation of hydrocarbons. Co3O4 is only slightly weaker than the noble metal catalysts for low temperature catalytic combustion [1, 2]. Furthermore, cobalt spinel in both pure and bimetallic forms seems promising for its low cost and good activity in volatility organic compound combustion [35], diesel soot oxidation [6], and particularly total oxidation of light hydrocarbons [7, 8]. The high catalytic activity in oxygen involving reactions of the Co3O4-based catalysts is most likely related to the high bulk oxygen mobility [9] and facile formation of highly active electrophilic oxygen (O or O−2) species for hydrocarbon oxidation [10].

Ceria has been the most frequently used catalyst due to its beneficial properties like the ability to enhance the water-gas shift reaction, to improve catalytic activity at the interfacial metal-support sites, to thermally stabilize the alumina support, to promote noble metal dispersion, and to exhibit its own catalytic activity especially when it is used in nanoparticles [1115]. However ceria’s most prominent effect is its ability to undergo rapid redox cycles, 2CeO2 → Ce2O3 + (1/2)O2, thus acting as an oxygen buffer by storing/releasing O2 due to the Ce4+/Ce3+ redox couple. Addition of zirconium is considered the most effective way for ceria stabilization against its sintering at high temperature reactions. Ceria-zirconia solid solution also yields an improvement in ceria’s oxygen storage capacity (OSC), redox properties, thermal resistance, and catalytic activity at low temperatures [1618]. Therefore, system based on CeO2 is widely used as support or active phase for oxidation reaction of hydrocarbon.

Although both cobalt oxide and ceria/ceria-zirconia catalysts have been widely applied for the oxidation of hydrocarbons, their mixtures are still rare investigated. A series of Co3O4 on mesoporous (, 0.85, 0.95, 1) catalysts synthesized by surfactant-assisted coprecipitation with subsequent incipient wetness impregnation method were studied. Results showed that the appropriate Ce/(Ce + Zr) ratio in ceria-zirconia support and cobalt oxide loading is required, and the 20 wt.% Co3O4/meso-Ce0.85Zr0.15O2 catalyst exhibits outstanding catalytic performance [19]. Another investigation by Xu et al. has shown that -Co3O4 catalysts exhibit a better catalytic activity when compared with pure Co3O4. The catalyst with the Ce/Co atomic ratio 1 : 16 exhibits the best activity [20]. Nanometric CeO2-supported cobalt oxide catalysts (molar ratio of 100 Co/Ce = 0.1, 1, 4, 10, 20, 50, 100) were prepared by the method of ultrasonic-assisted incipient wetness impregnation. The best catalytic activity was obtained over the catalyst containing 100 Co/Ce molar ratio of 20 [21]. Studies by Iwaneka et al. have shown that -CeO2 catalysts are in general more active than the undoped cobalt catalyst. It has been found that catalysts with a Co/(Co + Ce) ratio not higher than 0.64 sinter less than those with a higher cobalt content. The catalyst containing 86% cobalt exhibits the highest activity among the studied systems [22].

For the treatment of unburned hydrocarbons, which are toxic components in automobile exhaust gases, mixtures of Co3O4 (30 wt %) with CeO2 and CeO2-ZrO2 (Ce/Zr = 6/4) prepared by precipitation method have also been tested for the combustion of CH4 under exceeds oxygen conditions. The results showed that the methane oxidation activity of Co3O4 (30 wt.%)-CeO2/(CeO2-ZrO2) oxides is comparable with pure Co3O4 for fresh samples, while the composite systems exhibit higher activity and thermal stability than single Co3O4 after ageing at 750°C. The catalysts were able to convert 100% CH4 at rather high temperature (500°C). The role of ceria and ceria-zirconia is assumed to maintain a good combustion activity of the cobalt composite oxides by dispersing the active phase Co3O4 [23].

Meanwhile, the study on the mixtures of Co3O4 with CeO2 or CeO2-ZrO2 for the oxidation of hydrocarbons is still limited with only one ratio of Ce/Zr (Ce/Zr = 6/4) while we found from our previous study that the optimal Ce/Zr ratios for the highest conversion of hydrocarbons are 8/2 and 9/1 [24]. Moreover, the catalytic ability of the reported mixed Co3O4 with CeO2 or CeO2-ZrO2 catalysts was only high at high temperatures (500°C) [24]. Therefore, it is still worth studying more logically the different mixtures of Co3O4 with CeO2 or CeO2-ZrO2 or Co3O4 catalysts supported on CeO2 or CeO2-ZrO2 under different operation oxidation conditions of hydrocarbons (rich and stoichiometric conditions). It may also be interesting to compare the use of active CeO2-ZrO2 support with other popular nonactive but high surface area and cheap supports as Al2O3, SiO2. That will be the focus of this present paper.

2. Experimental

2.1. Catalyst Synthesis

The work uses several commercial supports such as Al2O3 (purity 100%, Merck) and SiO2 (purity 100%, Merck). To prepare active phase and mixtures of different components, a sol-gel method was used as described in more detail. It was previously shown that sol-gel method leads to the formation of very pure and homogeneous catalyst powders exhibiting high surface area [25].

13.5697 g Ce(NO3)3·6H2O (98.5%, Merck) was dissolved in 250 mL distilled water under stirring condition (solution A).

10.0703 g ZrOCl2·8H2O (99.0%, Merck) was dissolved in 250 mL distilled water under stirring condition (solution B).

27.7778 g citric acid monohydrate, C6H8O7·H2O (99.5%, Merck), was dissolved in 250 mL distilled water under stirring condition (solution C).

9.0937 g Co(NO3)2·6H2O (99.0%, Merck) was dissolved in 250 mL distilled water under stirring condition (solution D).

To prepare mixed oxide CeO2-ZrO2 supports (Ce0.9Zr0.1O2 and Ce0.8Zr0.2O2), solution B was dropped into a suitable volume of solution A corresponding to different CeO2/ZrO2 molar percentages. If precipitation occurred, a suitable amount of concentrated HNO3 solution was added. A suitable amount of solution C was dropped into the obtained solution in order to obtain molar ratio of citric acid to metals of 2.6.

Co3O4 supported on Ce0.9Zr0.1O2 or Ce0.8Zr0.2O2 and CeO2-Co3O4 supported on Al2O3, SiO2 were synthesized by impregnation method by mixing solution D and solution C, or solution A, solution D, and solution C, with different ratios and suitable amount of active phases. Concentrated HNO3 was then added to obtain solution in order to control pH of solution below 1. The obtained solutions were stabilized for 30 minutes. Then, a suitable amount of Ce0.9Zr0.1O2 (or Ce0.8Zr0.2O2, Al2O3, SiO2) support was added to solution to obtain suspensions. After stirring vigorously for 1 hour, the suspension was vaporized at 60–80°C until the slurry was dried. The obtained solid was dried at 120°C for 2 hours. Then, these samples were calcined at 550°C for 3 hours.

Co3O4 supported on CeO2-ZrO2 was labeled as A Co/Ce0.9Zr0.1O2 or Ce0.8Zr0.2O2, in which A is the weight percentage of the active phase.

CeO2-Co3O4 supported on Al2O3 (SiO2) was identified as E CeCo P-Q/Al (Si), in which E is the weight percentage of the active phase, P-Q is the ratio of CeO2/Co3O4, Ce is the symbol of CeO2, Co is the symbol of Co3O4, Al is the symbol of Al2O3, and Si is the symbol of SiO2.

2.2. Characterization

The catalysts were characterized by using X-ray, SEM, and BET techniques. The X-ray powder diffraction (XRD) patterns were recorded with a D8 Bruker Advanced diffractometer (Germany).

The specific surface areas of the samples were measured at 77 K by the BET method using N2 adsorption on an ASAP 2010 and a Gemini VII Micromeritics apparatus.

The morphology of these catalysts was examined on a Hitachi X4800 (Japan) scanning electron microscope.

2.3. Measurement of Catalytic Activities

Catalytic activities (the propylene conversion, which was determined as the ratio of consumed propylene to the amount of propylene introduced into the reactor, and the CO2 selectivity, which was determined as the amount of the obtained CO2 per the total amount of all products) were measured in a micro (or differential) reactor set up with an internal diameter of 0.4 cm and length of 60 cm. 0.1 g of catalyst (particle sizes are within the range 250–300 μm, which were prepared by pressing catalyst powders in a hydraulic presser, grinding the obtained pellets, and sieving into desired particle sizes) was used with a total gas flow of 80 mL/min at a pressure of 1 atm. The catalyst bed length is 5 mm; thus the gas hourly space velocities (GHSV), which were determined from the standard flow rate of the gas divided by the catalyst bed volume, were 76,000 h−1 in the catalyst reactor. The volume composition of gas flow was C3H6/O2/N2 = 2.5/2.5/95 (%) and the reaction temperatures range from 200°C to 500°C. Under the stoichiometric condition (oxygen sufficient), the composition of the gas flow was C3H6/O2/N2 = 2.5/11.25/86.25 (%). The reaction temperature was measured using a thermocouple attached at the position of the catalyst bed inside an electric furnace. The temperature of the catalytic bed was directly controlled by the proportional integral derivative (PID) controller of the furnace. Analysis of propylene, oxygen, CO2, CO, and oxygenated products was performed using an online Focus-Thermo scientific gas chromatograph with a thermal conductivity detector (TCD). C3H6 and oxygenated products were detected with a column of 80/100 Chromosorb and a column of Carbowax 20 M in series while CO2, CO, and O2 were detected with a column of 60/80 Carboxen and a column of 80/100 Porapak in series.

3. Results

3.1. Characterization and Catalytic Activity of Co3O4 Catalysts on CeO2-ZrO2 Supports

From our previous experiment, it was found that mixtures of Co3O4 and CeO2 exhibited high activity for the oxidation of propylene but were unstable at high temperatures. The situation may be improved if Co3O4 is impregnated on a stable phase containing CeO2. CeO2-ZrO2 mixtures have been determined to be active for the oxidation of propylene as in our previous publication [24]. Amongst CeO2-ZrO2 mixtures, ones with Ce/Zr ratios = 8/2 and 9/1 showed highest activity and possessed high ability of CO2 formation. Therefore, CeO2-ZrO2 mixtures with Ce/Zr ratios = 8/2 and 9/1 were chosen as supports for Co3O4 catalysts. The Co3O4 supported on CeO2-ZrO2 samples was characterized and tested for the oxidation of propylene under rich condition.

BET surface area of Co3O4 impregnated on different CeO2-ZrO2 supports (Ce0.9Zr0.1O2 and Ce0.8Zr0.2O2) is shown in Table 1. The results shows that Ce0.8Zr0.2O2 support possesses a little higher surface area than Ce0.9Zr0.1O2 support. When Co3O4 was impregnated on the supports, surface area tends to decrease maybe due to lower surface area of Co3O4 catalyst (11.44 m2/g) and the influence of the calcination after the impregnation. However, samples with 5% of Co3O4 on CeO2-ZrO2 supports showed increased surface area, which is not clearly understood since both SEM and XRD of these samples showed no evidence of strange phases, particle sizes, or morphology.

All samples show structures similar to those of CeO2 and CeO2-ZrO2 mixed oxides as seen from XRD patterns (Figure 1). The evidence of Co3O4 was only seen clearly when the content of Co3O4 is over 30%. For the samples with lower Co3O4 contents, the presence of Co3O4 only made the base line rougher, which indicates the more amorphous nature of Co3O4 (which was seen from XRD pattern of its pure sample).

SEM images of Co3O4 catalyst on Ce0.8Zr0.2O2 support in Figure 2 show a fine dispersion of Co3O4 on the honeycomb-like structure of the support. The particle size of Co3O4 is only few nm (circle marked in Figures 4(b) and 4(c)) and became larger when Co3O4 content increased (sample 30% Co3O4). On the sample 30% Co3O4/Ce0.8Zr0.2O2, there are aggregations of Co3O4 particles at some places (Figure 4(d)) but the support surface was not impregnated completely yet at some other places (Figure 4(c)).

The comparison of the activity of catalysts with different Co3O4 loading contents (5–50%) on CeO2-ZrO2 supports is shown in Figures 3-4. The catalysts were tested in rich condition (oxygen deficient) since it has been found that this condition is close to the real high speed operating conditions of motorcycles, the field of application of these catalysts [24]. It was also expected that if a catalyst exhibits good activity for complete oxidation under rich condition, it might exhibit even better activity under stoichiometric or lean conditions.

It can be seen that CeO2-ZrO2 supports could only reach to high propylene conversions (refer propylene conversions higher than 22.22%, the maximum conversion of propylene for the complete oxidation at the C3H6/O2 ratio of 1) at high temperatures (above 300°C). Meanwhile, pure Co3O4 and Co3O4 catalysts on CeO2-ZrO2 supports exhibited high conversion from 250°C. After reaching to those high conversions at this temperature, the propylene conversion was rather stable. The sample 30% Co3O4/Ce0.9Zr0.1O2 exhibited a sudden increased conversion at 500°C but together with decreased CO2 selectivity, which is the same as pure Co3O4 catalyst. However, pure Co3O4 even exhibited much lower CO2 selectivity at high temperatures. Thus, the deposition of Co3O4 catalyst on CeO2-ZrO2 supports decreased the temperature to obtain high activity of the catalysts compared with that of the support (250°C for catalysts on support instead of 350°C for the support) even at low Co3O4 loading contents (3–5%). The deposition of Co3O4 catalyst on CeO2-ZrO2 supports also simultaneously increased CO2 selectivity compared with that of the pure Co3O4 catalyst at high temperatures (400–500°C).

To investigate the role of active site and support, samples prepared by chemical mixing using sol-gel synthesis with the same compositions as 30% Co3O4/Ce0.9Zr0.1O2 and 30% Co3O4/Ce0.9Zr0.1O2 samples were also tested. The samples prepared by chemical mixing showed similar activities as supported sample prepared by impregnation method. These samples also possess high surface area as those of CeO2-ZrO2 support. This observation shows that CeO2-ZrO2 components in 30% Co3O4/Ce0.9Zr0.1O2 and 30% Co3O4/Ce0.9Zr0.1O2 catalysts play roles as active sites too. Therefore, it is possible to simply prepare Co3O4-CeO2-ZrO2 mixed catalyst by sol-gel method instead of more complex impregnation method while still maintaining surface area and catalytic activity of the catalysts.

3.2. Characterization and Catalytic Activity of Co3O4-CeO2 Supported on Al2O3 and SiO2

In our previous study, it was found that the sample containing 50 and 20 mol% of CeO2 (CeCo 50-50 and CeCo 20-80, resp.) exhibited excellent catalytic activities, which were even higher than those of Co3O4 catalysts on CeO2-ZrO2 supports described in Section 3.1. Co3O4-CeO2 catalysts were able to convert a large amount of propylene from 200°C already [24]. The catalysts also possessed the same high surface area as those of CeO2-ZrO2 supports (around 45 m2/g). However, these catalysts (the same as 30–50% Co3O4/CeO2-ZrO2 catalysts) were unstable at high temperatures; the catalyst particles were observed broken at high temperatures (400–500°C), resulting in blocking of the reactor. Therefore, it may be helpful to impregnate Co3O4-CeO2 catalysts on other high surface area and thermal stable supports (γ-Al2O3 and SiO2) to increase surface area of the catalysts as well as make them more stable at high temperatures. Thus, we decided to study Co3O4-CeO2 catalysts on support. CeO2-ZrO2 would not be chosen as a support for Co3O4-CeO2 catalysts since the composition of the catalyst already contains CeO2; moreover, CeO2-ZrO2 supports possess much less surface areas. Although high surface area is not the most important factor in the reaction where a Mars Van Krevelen mechanism is usually invocated as in hydrocarbons oxidation [26], high surface area still helps in the adsorption of reactants.

The use of high surface area Al2O3 and SiO2 supports resulted in much high surface area catalysts than in the case of using CeO2-ZrO2 supports. The surface area of the CeO2-Co3O4 catalysts on Al2O3 and SiO2 supports was only slightly lower than that of pure supports even if the catalyst loading contents were high (30–50 w%) (Table 2). Morphology of CeO2-Co3O4 catalysts on Al2O3 and SiO2 supports is a little different. Figure 5 shows SEM images of CeO2-Co3O4 catalyst with Ce/Co molar ratio of 5/5 on Al2O3 and SiO2 supports. The deposition of CeO2-Co3O4 catalysts on Al2O3 resulted in more amorphous morphology and less particle sizes. The images also show that the catalyst loading content of 30 w% made the surface of the supports covered almost completely by CeO2-Co3O4 catalyst particles. This may be the reason for the fact that the samples with catalyst loading contents above 30 w% were the samples which exhibited highest propylene conversion (Figure 6).

Propylene conversions and CO2 selectivity (under rich condition) of CeO2-Co3O4 catalyst on Al2O3 and SiO2 supports were presented in Figures 6-7. The catalytic loading content was investigated in a wide range (5–50 w%) with the aim of optimizing it and finding out a catalyst with as low catalytic loading content as possible but as high activity as possible. It can be seen that the pure CeO2-Co3O4 active phases were able to reach to the maximum activity at lower temperatures than when they were deposited on the supports. The samples with high CeO2-Co3O4 loading contents (30–50 w%) on Al2O3 support could convert a maximum amount of propylene at 300°C while the samples with lower CeO2-Co3O4 loading contents (5–10 w%) could only obtain the maximum conversion at 400–450°C (Figure 6). Thus, when the catalyst was impregnated on the support, high loading contents of the active phase (above 30%) are required to lower the minimum temperature of the maximum conversion. High CeO2-Co3O4 loading contents also help to maintain CO2 selectivity at high temperature close to that of the pure active phase. The sample with low CeO2-Co3O4 loading contents (5–10 w%) showed decreased CO selectivity at high temperatures, which is similar to that of the pure support. The results also show that catalytic activities of the samples with Ce/Co molar ratios of 5/5 and 2/8 are quite similar. The sample containing 50 w% CeO2-Co3O4 (Ce/Co = 2/8) on SiO2 support exhibited a little better activity than those of other samples since it could convert a maximum amount of propylene sooner (from 250°C). CO2 selectivity on catalysts using SiO2 support is also a little higher than those of the catalysts using Al2O3 support. This is probably due to the higher acidity of Al2O3 support, which made some exposed places of the support become sites for coke formation. The catalysts on Al2O3 support after the reaction are usually observed becoming black. Catalytic activity of the best CeO2-Co3O4 catalyst on Al2O3 and SiO2 supports was not higher than those of the best of Co3O4 catalyst on CeO2-ZrO2 supports although the activity of the unsupported CeO2-Co3O4 catalyst was higher than that of the unsupported Co3O4 catalyst, showing that active CeO2-ZrO2 supports made certain meanings compared with inert Al2O3 and SiO2 supports.

3.3. Activity of CeO2-Co3O4 Catalysts for the Treatment of Propylene under Stoichiometric Condition

To understand details about the ability of the catalysts in different reaction conditions, CeO2-Co3O4 catalysts on supports were also tested for the oxidation of propylene under stoichiometric condition (enough oxygen), in which the ratio of O2 to propylene was maintained at 4.5 according to the stoichiometric ratio of the complete oxidation reaction. The best Co3O4 catalyst on Ce0.9Zr0.1O2 support (30% Co3O4/Ce0.9Zr0.1O2) was also tested under stoichiometric condition for comparison. The results show that all catalysts were able to convert 100% propylene into CO2 (Table 3). However, only 30% Co3O4/Ce0.9Zr0.1O2 catalyst was able to convert almost 100% propylene from low temperature (250°C); the catalysts on Al2O3 and SiO2 supports were only able to convert propylene completely at high temperatures (from 400°C). Although Al2O3 and SiO2 supports possess much higher surface area than that of Ce0.9Zr0.1O2 support, they are almost inert for the oxidation of propylene while Ce0.9Zr0.1O2 support, itself, is a good catalyst for the oxidation of propylene as seen in Section 3.1; therefore, the catalyst on Ce0.9Zr0.1O2 support exhibited an excellent activity. Between Al2O3 and SiO2 supports, catalyst on Al2O3 showed worse activity than that of SiO2, which may be due to higher acidity of Al2O3, making exposed Al2O3 particles become sites for coke formation even under stoichiometric condition.

Different from the rich condition, under stoichiometric condition, the only product is CO2 at all temperatures. Thus, the catalysts catalyze well the complete oxidation of propylene but under rich condition, CO and oxygenated products were still formed since there was not enough oxygen for the reaction of propylene.

When the complete oxidation was performed under lean condition (excess oxygen), the catalytic activities of these catalysts were even better, resulting in the decrease of the minimum temperature to convert 100% propylene of all catalysts.

4. Conclusions

Two series of catalysts with suitable loading contents (above 30 w%) investigated in this work, Co3O4 on CeO2-ZrO2 supports and CeO2-Co3O4 catalysts on Al2O3 and SiO2 supports, exhibited good activities for the treatment of propylene under all rich, stoichiometric, and lean conditions. Under stoichiometric condition, 30% Co3O4/Ce0.9Zr0.1O2 catalyst was able to oxidize 100% propylene to pure CO2 from 250°C; 30% CeO2-Co3O4 catalysts on Al2O3 and SiO2 supports were able to do the same at higher temperatures (400°C). Thus, Ce0.9Zr0.1O2 was proved to be the more active support than inert Al2O3 and SiO2 supports for the catalysts for the complete oxidation of propylene since Ce0.9Zr0.1O2, itself, was also a good active phase for the reaction. However, catalysts on Ce0.9Zr0.1O2 exhibited much lower surface areas and much less thermal stability. Further work will be performed to explore more details on the characteristics of the catalysts and the supports.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgment

This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under Grant no. 104.03-2011.16.