Abstract

Cobalt(II,III) oxide, Co3O4, is a promising catalyst for nitrous oxide direct decomposition. In this paper we report effect of doping with some rare earth (RE) elements (Pr, Sm, and Tb) on the morphology and crystallite size of Co3O4 nanorods. The various precursors (RE/Co oxalates) were prepared via the microwave assisted method and subsequent calcination. The decomposition pathway of these precursors was followed using thermogravimetric analysis (TGA). Based on thermal analysis results, Pr-, Sm-, and Tb-doped Co3O4 samples were obtained via the calcination in static air at 500°C for their oxalate precursors. X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FT-IR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and hydrogen temperature programmed reduction (H2-TPR) were used to characterize the RE-doped cobalt oxide catalysts. The activity of the prepared catalysts was investigated for N2O direct decomposition and compared with that of the undoped Co3O4 catalyst. It was shown that the promoted Co3O4 catalysts revealed higher activity compared to the unpromoted one. The dependence of the activity on both the catalysts particle size and the reduction behaviour was discussed.

1. Introduction

Due to the unawareness of its harmful effects nitrous oxide (N2O) has suffered, for a long time, from the lack of interest from the environmental scientists [1, 2]. Only during the last two decades a growing interest can be noticed with N2O abetment studies. This is especially after its citation as a second, non-CO2 greenhouse gas. Moreover, it was recognized that it causes the destruction of ozone in the stratosphere and it causes the acid rains. Due to its long lifetime, about 150 years, N2O has a Global Warming Potential (GWP) of 310 and 21 times that of CO2 and CH4, respectively [1, 2]. The industrial processes that involve N2O emission include adipic acid, nitric acid, caprolactam, acrylonitrile, and glyoxal manufactures. Moreover, N2O is emitted as a byproduct of some processes like the nonselective catalytic reduction of NO (NSCR) with cyanuric acid or urea and in three-way catalysis (TWC) to remove NO, CO, and hydrocarbons [1, 2].

Accordingly, N2O was cited as the second non-CO2 greenhouse gas Kyoto Protocol of the United Nations Convention on Climate Change (December 1997). Based on that protocol the industrial countries should reduce their collective greenhouse gas (GHG) emissions by values slightly higher than 5% below their 1990 values [1, 2]. On the other hand, the developed countries have to reduce their GHG emissions by values of around 5%. Such targets will be increased in the next protocol of the United Nations that will replace the Kyoto protocol. The catalytic routes represent the most important and applied solutions for this environmental problem. Among them the direct decomposition of nitrous oxide to its elements, that is, nitrogen and oxygen, is considered to be a simple and effective way to minimize the N2O emissions. Many catalysts categories have been developed for N2O decomposition.

Cobalt(II,III) oxide, Co3O4, is a black antiferromagnetic solid, p-type semiconductor and has direct optical band gap at 2.18–3.52 eV [3]. Co3O4 adopts the normal spinel structure, represented as AB2O4 in which A ions, Co2+(3d7), occupying tetrahedral sites and B ions, Co3+(3d6), are located in octahedral sites. Co3O4 has been investigated extensively as promising catalysts for N2O direct decomposition [414].

Russo et al. [4] have checked the activity of three series of metal (Mg, Ca, Mn, Co, Ni, Cu, Cr, Fe, and Z) spinels (cobaltites, chromites, and ferrites) prepared by the solution combustion method. Their results demonstrated that among the categories the cobalt containing spinels exhibit the best activity. Moreover, MgCo2O4 showed the best performance among the different metal cobaltites in the absence and presence of oxygen. This was attributed to its higher concentration of suprafacial-weakly chemisorbed oxygen species. Over a series of (M = Mg, Ni, Zn, ) spinel catalysts, prepared by the coprecipitation method, Yan et al. [5, 6] found that the N2O decomposition activity depends on the degree of Co2+ substitution by Mg2+ or Ni2+ or Zn2+. The highest activities were observed over Mg0.54Co0.46Co2O4, Ni0.74Co0.26Co2O4, and Zn0.36Co0.64Co2O4 catalysts. Concurrently, for and () spinel-oxide catalysts, prepared via the coprecipitation method, it was found that the highest activity was exhibited by the catalysts 0.75 and 0.50, respectively [7, 8]. Moreover, increasing the calcination temperature till 900°C for both series was accompanied by a sharp activity decrease. The activity of () catalysts, prepared by the impregnation method, was found to increase the -value [9].

With respect to the influence of promoters, Xue et al. reported a beneficial effect on doping Co3O4 spinel with ceria [10]. On varying the Ce/Co molar ratio, a complete N2O conversion below 400°C was obtained for the catalyst having a ratio of 0.05. The promotion effect was attributed to the role of CeO2 in promoting the reduction of Co3+ to Co2+, thus facilitating the desorption of adsorbed oxygen species, which is deemed to be the rate-determining step for N2O decomposition. The activity of cobalt containing spinels was also markedly enhanced by doping with alkali cations. In this regard, the presence of residual K ions has led to an activity enhancement for Co3O4-CeO2 (Ce/Co = 0.05) catalyst [11]. Potassium and cesium ions were found to be superior to lithium and sodium ones [9]. Recently, an activity enhancement was reported by the addition of K ions to Zn0.4Co2.6O4 [12], MgCo2O4 [9], and Co3O4 [13] catalysts. Doping with alkali earth cations was also reported to improve the activity of spinel oxide catalysts. In this context, N2O decomposition over a series of MCO3-Co3O4 (M = Ca, Sr, Ba) catalysts having M/Co ratios of 0.1–0.4 revealed that the catalysts containing Sr and Ba ions exhibit the highest activity [14]. The crystallite size and the calcination temperature are other factors that influence the N2O decomposition of the oxide spinel catalysts; lower values of both factors increase the catalytic activity [79, 14].

The previous literature data reveals that the activity of Co3O4 spinel oxide-based catalysts is sensitive to the preparation method, the degree of Co2+ substitution, the calcination temperature, crystallite size, and presence of dopants. The role of rare earth oxides as promoters during N2O direct decomposition was reported for some catalytic systems. For instance, Bueno-López et al. [15] demonstrated that the addition of La, Pr dopants increases the activity of Rh/CeO2 catalysts markedly. This was attributed to the role of the added RE cation in stabilizing the catalysts active phase (Rh2O3). Synergic catalytic performance compared to the single oxides was reported for CuO/CeO2 catalysts [16]. This effect was related to the enhanced catalysts reducibility as well as the facile Ce4+/Ce3+ and Cu2+/Cu+ redox cycles. Concurrently, Konsolakis et al. [17] reported a dramatic enhancement of de-N2O performance of Pt/Al2O3 by the incorporation of 20 wt% CeO2-La2O3 mixed oxide on the Al2O3 support. To the best of our knowledge, there are limited reports concerning the influence of the rare earth oxides dopants on the N2O decomposition activity of Co3O4 spinel oxide catalyst. Therefore, based on the expected promotion effect of the RE dopants as it was reported for other catalysts, in this paper we report the enhancement effect of doping Co3O4 nanorods with some rare earth (RE) elements (Pr, Sm, and Tb) during N2O direct decomposition. The examined catalysts were prepared via the microwave assisted method of cobalt oxalate and subsequent calcination. Special focus will be given to the catalysts characterization in order to check the influence of the added RE elements on the morphology and crystallite size of Co3O4 nanorods.

2. Materials and Methods

2.1. Catalysts Preparation

Cobalt nitrate (Co(NO3)2·6H2O), praseodymium nitrate (Pr(NO3)3·6H2O), samarium nitrate (Sm(NO3)3·6H2O), terbium nitrate (Tb(NO3)3·6H2O), oxalic acid (H2C2O4), and cetyltrimethylammonium bromide (CTAB) are of analytical grade and were used as received without further purification. Co3O4 was prepared by microwave assisted precipitation method (in the form of cobalt oxalate) using a microwave power of 560 W and subsequent calcination at 500°C. Detailed information about the preparation procedure is given elsewhere [18]. The rare earth-promoted Co3O4 catalysts were prepared employing the coprecipitation method. In this method the amount of a rare earth cation (Pr3+, Sm3+, and Tb3+), required to achieve rare earth (RE)/Co ratio of 0.05, was added during the precipitation of cobalt oxalate using the same working procedure. After drying the coprecipitated oxalates, they were calcined for 1 h at 500°C in air to yield the corresponding earth-promoted Co3O4 catalysts. Scheme 1 shows an illustration and mechanism of the Co3O4 and RE/Co3O4 catalyst synthesis.

2.2. Characterization Techniques

TGA and DTG curves were recorded with the aid of TA instrument apparatus (model TGA-Q500) using a heating rate of 10°C min−1 and an average sample weight of 5 mg in nitrogen flow (40 mL min−1). X-ray powder diffraction (XRD) patterns were performed for the prepared RE/Co3O4 catalysts using a computerized Philips diffractometer (type PW 103/00) operated with Cu Kα radiation (λ = 1.5405 Å) at 35 kV and 20 mA. The step scans were taken over the 2θ range of 4–80° in a step of 0.06 min−1. FT-IR spectra were performed over the wavenumber range of 4000–400 cm−1 using Attenuated Total Reflectance (ATR) sampling accessory on the Nicolet iS50 FT-IR spectrometer. Catalysts morphologies were characterized by scanning electron microscopy (JEOL-JSM-5400 LV). Prior to the SEM observations the catalysts were sputter with gold. Transmission electron microscopy (TEM) images were performed using JEOL (model JEM-1011) microscopy. Hydrogen temperature-programmed reduction (H2-TPR) experiments were performed on Quantachrome CHEMBET-3000 instrument.

2.3. Activity Measurements

The catalytic tests were carried out in a fixed-bed quartz-glass flow reactor, at atmospheric pressure, containing approximately 0.5 g of the catalyst. The experiments were carried out at the temperature range of 150–500°C and a W/F value of 0.15 g s cm−3. The N2O gas was introduced onto the catalyst bed from the bottom. The reactor was heated using a temperature-controlled furnace. The temperature in the reactor was measured using a thermocouple on the catalyst bed. Prior to the reaction all the catalysts were pretreated for 1 h in helium at 500°C. The reactant gas, 500 ppm N2O, was introduced to the catalyst bed with the aid of Bronkhorst thermal mass flow controllers using helium as a balance gas. The N2O concentrations in the inlet and outlet streams were measured using nondispersive infrared analyser (Hartmann and Braun, Uras 10E).

3. Results and Discussion

3.1. Catalysts Characterization

As it was mentioned in the introduction that the calcination temperature has a crucial effect on the catalytic behaviour of the Co3O4-based catalysts, therefore, in order to identify the appropriate calcination temperature of the parent cobalt oxalate and its RE-containing mixtures, the thermal stability of these parents has been checked by TGA during the heating up to 700°C. Our more recent work demonstrated that the preparation procedure followed has led to the formation of cobalt oxalate with the formula CoC2O4·2H2O [18]. The obtained TGA-DTG curves are shown in Figure 1. Two weight loss (WL) events can be observed for the undoped cobalt oxalate (Figure 1(a)). These events are accompanied by WL values of 19.58 and 46.11% and are maximized at 162 and 366°C. The first WL agrees well with that (19.69%) expected for the dehydration of the parent CoC2O4·2H2O. The obtained second WL% is higher than that expected for the formation of Co3O4 (36.44%) or CoO (39.36%) and lower than that anticipated for the formation of metallic Co (48.10%). This suggests that the residue formed in the second step is composed mainly of metallic cobalt. In agreement, Deng et al. [19] reported the formation of β-Co as a result of the thermal decomposition of its β-CoC2O4·2H2O precursor in N2 flow at 395°C. In this context, it is to be mentioned that conducting the experiment air flow has led to the formation of Co3O4 as a second decomposition step product [1820]. The TGA thermograms obtained for the RE/cobalt oxalate mixtures (Figures 1(b)1(d)) reveal the emergence of an early WL step (maximized at the temperature range of 54–86°C). This step could be related to the dehydration of the RE oxalates. Meanwhile, two other peaks can be observed in the three thermograms, which are maximized at the temperature ranges of 162–170°C and 366–372°C. No other WL steps can be observed in the three thermograms. Therefore, it is plausible to assign the former set of peaks to the dehydration of cobalt oxalate, whereas the second set of peaks is assigned to the decomposition of the anhydrous RE/cobalt oxalates. TGA curve of the Tb/Co mixture indicates the presence of a mild-gradual weight gain at temperatures higher than 425°C. This weight gain could be related to the oxidation of the obtained metallic cobalt by the traced amount of oxygen in the flow. According to the TGA results the various parents were calcined at 500°C for 1 h in order to obtain the relevant nanocrystalline Co3O4-based catalysts.

Figure 2 shows the XRD patterns of the calcined cobalt oxalate and its RE-containing mixtures. XRD pattern for the undoped Co3O4 catalyst (Figure 2(a)) reveals that the presence of reflections matches well with those of the cubic Co3O4 spinel oxide (space group Fd3m; JCPDS card number 42-1467) [18, 2026]. No other peaks assignable to Co2O3, CoO, and Co were detected indicating that the decomposition of cobalt oxalate in air proceeds with the formation of Co3O4 spinel oxide as a pure phase. The XRD pattern of the Pr containing catalyst (Figure 2(b)) reveals the presence of reflections due to the Co3O4 spinel oxide together with very small intensity ones at 2θ = 28.38°, 33.07°, and 47.02°. These reflections could be ascribed to the presence of Pr6O11 phase (JCPDS card number 42-1121). This indicates that this catalyst is composed of Co3O4 as a major phase together with trace amount of Pr6O11. The diffractogram of the Sm/Co catalyst (Figure 2(c)) shows the existence of high-intensity peaks characterizing the Co3O4 spinel together with two weak reflections at 2θ = 28.38° and 47.12°. These two reflections could be assigned to the Sm2O3 phase (JCPDS card number 76-0153). This, in turn, suggests the coexistence of Co3O4 (major) and Sm2O3 (trace) for this sample. Finally, the XRD pattern of the Tb/Co catalyst (Figure 2(d)) reveals the presence of all the peaks due the Co3O4 spinel as a major phase. In addition, two small reflections emerged at 2θ = 42.44° and 61.52°, which can be assigned to CoO phase (JCPDS card number 78-0431). Moreover, two other weak-intensity peaks appear at 2θ = 29.06° and 48.69°, which indicates the presence of trace amount of Tb4O7 (JCPDS card number 13-0387).

The FT-IR spectra of the calcined cobalt oxalate and its RE-containing mixtures are shown in Figure 3. For all samples, FTIR spectra revealed the presence of two absorptions located at 560–567 and 656–670 cm−1. These two absorptions can be assigned, respectively, to ν1 and ν2 stretching vibrations of the metal-oxygen bond in Co3O4 [3, 14, 18, 2527]. The RE oxide containing catalysts reveals the presence of other weak absorptions at the wavenumber range of 1700–1200 cm−1, which can be assigned to the presence of surface carbonate species [7, 8, 14]. This indicates the ability of the added RE oxides to form weak surface carbonates upon contact with atmospheric air.

Figure 4 shows the SEM images of the neat Co3O4 as well as its Pr, Sm, and Tb containing catalysts prepared by the coprecipitation method and calcined at 500°C. It is evident that the neat Co3O4 (Figure 4(a)) consists of sets of rods. Adding the RE oxides to Co3O4 resulted in dramatic morphological changes. The Pr6O11/Co3O4 catalyst (Figure 4(b)) consists of network polygons within the frame of spheres agglomerations. The Sm2O3/Co3O4 (Figure 4(c)) consists of sets of welded rods which have smaller dimensions than those of the neat Co3O4. Finally, the Tb4O7/Co3O4 catalyst (Figure 4(d)) shows agglomeration of polydisperse rectangles.

The TEM images of the neat Co3O4 as well as its Pr, Sm, and Tb containing catalysts prepared by the coprecipitation method and calcined at 500°C are depicted in Figure 5. One can clearly show that rod morphology of the unpromoted Co3O4 (Figure 5(a)) is composed of sets of capsules having diameters in the range of 35–65 nm and lengths of 50–180 nm. These rods are welded to form larger ones. The Pr6O11/Co3O4 catalyst (Figure 5(b)) shows irregular shape particles having diameters in the range of 15–50 nm. Similar morphology can be seen for the Sm2O3/Co3O4 catalyst (Figure 5(c)) with particles diameters in the range of 20–60 nm. The Tb4O7/Co3O4 catalyst (Figure 5(d)) shows irregular shaped agglomerations of particles with dimensions in the range of 20–100 nm. Thus, the electron microscopic examination reveals that the rod-like morphology of Co3O4 is very sensitive to the presence of these RE oxides. The presence of such dopants even in small concentration is sufficient to destroy this rod-like morphology.

3.2. Catalytic Activity Measurements

The activity of the three RE-promoted Co3O4 catalysts was tested for N2O decomposition and compared with that of the unpromoted Co3O4. Figure 6 shows the dependence of N2O conversion percentage on the reactor temperature for these series of catalysts being prepared by the coprecipitation method. One can easily spot the fact that the activity increases continuously with the reactor temperature over all the tested catalysts. Moreover, all the RE containing catalysts showed higher N2O decomposition activity compared to the bare Co3O4 catalyst. In the open literature, there are many papers reporting the promotion effect of various promoters during N2O decomposition over Co3O4-based catalysts. For instance, partial substitution of Co2+ ions in Co3O4 spinel oxide with various metal ions (Mg2+, Ni2+, Cu2+, and Zn2+) has led to a noticeable increasing of its activity [59]. Activity enhancement was observed also on doping Co3O4-based catalysts with some promoters like ceria [10], zirconia [28], alkali [1113, 29, 30], and alkali-earth [14] cations. In some of these papers the enhanced N2O decomposition activity was correlated with the Co3O4 crystallite size [79, 14], where high activity was observed over small Co3O4 crystallite size. Concurrently, an inverse relationship between the Ag0 crystallites size and the N2O conversion % was reported for catalysts [31]. Similarly, the higher N2O decomposition activity of the RE/Co3O4 catalysts compared to that of the bare Co3O4 one could be related to the induced Co3O4 crystallite size reduction by the added RE ions as it was observed during the TEM analysis (Figure 5).

The reported reaction mechanism of the N2O decomposition over Co3O4-based catalysts involves electron donation from an active center at the catalyst surface, which is a metal ion in a low oxidation state (probably Co2+), to the N2O molecule as follows [9, 14, 28]:This step is accompanied by Co2+ oxidation to Co3+ together with the formation of one N2 molecule and a chemisorbed oxygen atom. Thus, the recoverability of Co2+ is crucial for attaining high N2O decomposition activity as shown in (3). One hasTherefore, H2-TPR measurements were performed in order to examine the effect of the addition of RE ions on the reduction behaviour of Co3O4 spinel oxide. The obtained TPR curves of these catalysts are depicted in Figure 7. All the TPR profiles in Figure 7 reveal the existence of one broad peak, which extends over the temperature range of 280–520°C. This finding indicates that this peak could describe several overlapping reduction processes. In agreement with other literature data, such peak could be assigned to the reduction of Co3+ to Co2+ and Co2+ to Co0 according to [10, 32]:

The obtained TPR profiles indicate that the reduction behaviour of Co3O4 is greatly influenced by the added RE oxides. of the reduction peak is shifted from 448°C for Co3O4 to 441, 435, and 431°C for Pr-, Sm-, and Tb-promoted Co3O4, respectively. Moreover, the onset temperature of the reduction is shifted towards lower temperatures for the RE-promoted Co3O4 compared to the unpromoted one. These findings, in turn, suggest that the reduction behaviour of Co3O4 is enhanced by the addition of the RE oxides.

More recently, it was demonstrated that the fine tuning of Cu facilitates the reduction of Co3+ and Co2+ in spinel [33]. This promotion of the catalysts reducibility was correlated with the increase in the N2O direct decomposition over this series of catalysts. For a series of Ce-doped Co3O4 catalysts, Xue et al. [10, 11] showed that the reduction behaviour of these catalysts is influenced by the Ce/Co ratio as well as the preparation method. They concluded that the higher the Co3+ reducibility of the Co3O4-CeO2 catalyst, the higher the N2O decomposition activity. Chromčáková et al. [32] reported a direct dependence of the reduction ability and the catalytic behaviour during N2O direct decomposition of a series of Co3O4 catalysts prepared via different routes. The results of N2O decomposition over bare Pr6O11, Sm2O3, and Tb4O7 (not shown) revealed that these oxides exhibit very low activities (less than 5% conversion at 500°C). This finding suggests that the active catalysts centers could be Co2+-Co3+ redox pair. And based on the H2-TPR results, we can confidently state that the improved reducibility of the Co3+ to Co2+, induced by the incorporation of RE oxides into Co3O4, could be an additional factor for their higher N2O decomposition activity compared to that of the unpromoted Co3O4. Accordingly, the reaction mechanism can be understood as follows: as the reaction proceeds N2O is adsorbed on an active catalyst center (Co2+) as indicated by (1). This step is followed by the liberation of one nitrogen molecule leaving back and adsorbed oxygen atom over a catalyst center with higher oxidation state, that is, Co3+. The addition of the RE oxides to improve the reducibility of the catalyst, that is, enhances the recoverability of the catalyst active center (Co2+) as indicated by (3).

4. Conclusions

Series of RE-doped Co3O4 precursors were prepared by the microwave assisted method employing oxalic acid as precipitating agent. The RE-doped Co3O4 catalysts were obtained by the calcination, in air at 500°C, of these precursors. The obtained catalysts were characterized using XRD, FT-IR, SEM, TEM, and H2-TPR techniques and tested for N2O direct decomposition. It was found that the addition of the RE oxides to the Co3O4 spinel enhances its N2O decomposition activity. The superior activity of the RE-promoted Co3O4 catalysts to that of the unpromoted Co3O4 one was correlated with the observed decrease in the particle size of Co3O4 as well as the facilitated Co3+ → Co2+ reduction.

Conflict of Interests

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

Acknowledgments

This project was supported by the National Science, Technology and Innovation Plan (MAARIFAH) strategic technologies programs, no. 12-ENV2756-03, of the Kingdom of Saudi Arabia. The authors thankfully acknowledge Science and Technology Unit, Deanship of Scientific Research at King Abdulaziz University, for their technical support.