As an important energy resource for replacing petroleum, coal and biomass have attracted great interest. Gasification is an important process as regards to the utilization of them in energy recycling, namely, converting into a producer gas mixture consisting of carbon monoxide (CO), hydrogen (H₂), carbon dioxide (CO₂) and other trace species (Huber et al. 2006; Zhu et al. 2008; Yung et al. 2009). As the gas mixture has a low energy density, its utilization value is very low. To increase the energy density or calorific value, the producer gas is often reformed to produce useable H₂ (Antal et al. 2000; Lee et al. 2002; Hao et al. 2003, 2005; Sınag et al. 2003; Lu et al. 2006; Byrd et al. 2007), CO (Ashcroft et al. 1991; Fox 1993; Rostrup-Nielsen and Bak-Hansen 1993) and synthetic natural gas (SNG) (Seemann et al. 2006, 2010; Kopyscinski et al. 2009, 2010). Because CO₂ is the most difficult producer gas to convert, its conversion remains an important topic.

The conversion of CO₂ to valuable chemical materials has been proposed as a possible way of utilizing non-combustible CO₂ (Jessop et al. 2004; Sakakura et al. 2007; Centi and Perathoner 2009; Federsel et al. 2010; Mikkelsen et al. 2010; Bourrez et al. 2011). In terms of CO₂ conversion, since CO is a valuable material in many chemical processes, the conversion of CO₂ to CO through the reverse water gas shift (RWGS) reaction by catalytic hydrogenation is recognized as the most promising process. In addition, CO₂ methanation is also an important process, and has been investigated for the production of energy carriers (Weatherbee and Bartholomew 1981; Peebles et al. 1983; Yamasaki et al. 2006; Ocampo et al. 2009), since this reaction is considerably faster than other reactions that form hydrocarbons or alcohols (Inu and Takeguchi 1991).

Nickel(Ni)-based catalysts remain the most extensively studied well-dispersed materials, and Ni-based catalysts have been investigated for the RWGS reaction and methanation (Vance and Bartholomew 1983; Wang et al. 2008). The discovery of ordered mesostructured silica SBA-15 with an adjustable pore size and a high specific surface area (Zhao et al. 1998, 2014) has provided the opportunity to produce well-dispersed Ni catalysts. NiO/SBA-15 has been realized with a post synthesis method (Vradman et al. 2005; Cheng et al. 2009; Lu and Kawamoto 2012), and methanation has been investigated (Lu and Kawamoto 2013). In addition, highly loaded and well-dispersed NiO/SBA-15 has been obtained by the direct synthesis method for the first time (Lu and Kawamoto 2012). However, the differences in location and dispersion of NiO particles caused by synthesis method in SBA-15 are not yet clear.

Therefore, we synthesized NiO/SBA-15 with 10 wt% NiO using direct and post synthesis methods, respectively. Their TEM observations were carried out to investigate the dispersion of NiO particles in SBA-15. H₂-TPR was used to study the location and dispersion of NiO particles in SBA-15. The RWGS reaction and methanation were carried out using NiO/SBA-15 obtained with different synthesis methods and NiO loadings.

According to the previous report (Lu and Kawamoto 2012), the direct synthesis method was used to prepare NiO/SBA-15. The block copolymer Pluronic P123 was mixed with water, H₂SO₄, Ni(NO₃)₂·6H₂O, and tetraethyl orthosilicate at room temperature. The mixture was aged at 60 °C until a white precipitate appeared and then immediately evaporated at 100 °C overnight. The solid product was dried at 150 °C for 5 h and then calcined at 500 °C for 10 h then 800 °C for 2 h.

With post synthesis method, SBA-15 was firstly synthesized at 60 °C according to the method (Chen et al. 2003). Then as described in the literature (Lu and Kawamoto 2013), the SBA-15 was added to EtOH containing dissolved Ni(NO₃)₂·6H₂O and dispersed for 6 h under supersonic. The mixture was then dried at room temperature and calcined at 400 °C for 5 h. The solid products were characterized by TEM observation. The catalyst reducibility was also studied by temperature programmed reduction (TPR) in BET-CAT catalyst analyzer (BEL Japan Inc.) using H₂.

The RWGS reaction and methanation were performed as follows. The catalysts were pre-reduced before the reaction. The reaction was performed at atmospheric pressure in a fixed-bed quartz reactor. A 30-mm long catalyst (2 g) between two layers of quartz wool was loaded into the reactor. The reactor was heated in a furnace. All reactant gases were monitored by mass flow meter and controller. The flow of the product was measured with a film flow meter and analyzed by GC-TCD after the reaction had become stable.

As shown in Fig. 1a, the small-angle XRD patterns of NiO/SBA-15 with NiO loading of 10 wt% had three well-resolved peaks, indicative of the typical two-dimensional hexagonally ordered mesostructure of SBA-15, regardless of the synthesis method. However, the angles of all of the diffraction peaks of NiO/SBA-15 obtained with direct synthesis method were smaller than those obtained with post synthesis method. This result suggests that the pore size of SBA-15 differed with synthesis method—in other words, the synthesis method affected the pore size of SBA-15. With post synthesis method, the NiO particles were inserted into the SBA-15 pores, resulting in the decrease of pore size. The wide-angle XRD peaks could be indexed to a face-centered cubic crystalline NiO structure, irrespective of synthesis method (Fig. 1b), indicating that NiO particles were also dispersed on the outer surface besides on pore surface of SBA-15. All wide-angle peak intensities of NiO/SBA-15 obtained with direct synthesis method were greater than those with post synthesis method. This result suggests that, with direct synthesis method, the NiO amount located on outer surface of NiO/SBA-15 was high. In addition, a broad peak at around 2θ = 23° caused by the amorphous SiO₂ structure of SBA-15 was clearly observed in the wide-angle XRD pattern, indicating that the ordered mesoporous SBA-15 structure was not disturbed by the included NiO.

XRD patterns of 10 wt% NiO/SBA-15

We examined the nitrogen adsorption–desorption isotherms of NiO/SBA-15 with NiO loading of 10 wt% obtained by using two different synthesis methods (not shown). The isotherms were similar to the type IV IUPAC classification, clearly indicating that these materials possessed mesoporous structures, and a type-H1 hysteresis loop was observed, regardless of synthesis method. The isotherms were typical of SBA-15, indicating the formation of a p6mm structure with open cylindrical mesopores. It was very clear that the decrease of pore size was caused by NiO particles included through post synthesis method.

The TEM images of NiO/SBA-15 with NiO loading of 10 wt% showed an ordered mesoporous structure typical of SBA-15, irrespective of synthesis method (Fig. 2a, b). With direct synthesis method (Fig. 2a), NiO particles were highly dispersed (monodispersed) into the SiO₂ structure of SBA-15. With post-synthesis method, aggregates of NiO particles were observed clearly (Fig. 2b).

TEM images of NiO/SBA-15 obtained by direct synthesis and post-synthesis methods

TPR investigation is generally useful as a fingerprint of a metal species’ interaction with the support material. As shown in Fig. 3a, for NiO/SBA-15 obtained with direct synthesis method, H₂-TPR profiles exhibited three sharp peaks. The former two peaks were ascribed to NiO particles on the outer surface of SBA-15, while the last peak was assigned to those on the pore surface of SBA-15, indicating that NiO particles were highly and well-dispersed into SBA-15, and the located boundary of NiO particles on between pore and outer surface could be observed. However, for NiO/SBA-15 obtained with post synthesis method (Fig. 3b), H₂-TPR profiles exhibited only a broad peak centered at 378 °C, indicating that NiO particles located either in pore or on outer surface could be observed, but their location boundary could not be distinguished clearly. Therefore, with post synthesis method, NiO particles were more poorly dispersed (aggregated) in SBA-15 than that with direct synthesis method.

H₂-TPR profiles of NiO/SBA-15 obtained by direct synthesis and post synthesis methods

We used 10 wt% NiO/SBA-15 obtained by direct synthesis and post-synthesis methods to investigate the RWGS reaction (Fig. 4). When NiO/SBA-15 obtained by direct synthesis method was used, CO₂ conversion increased with increasing temperature (Fig. 4a). Using NiO/SBA-15 obtained by post synthesis method, CO₂ conversion increased with increasing temperature at temperatures of 600 °C or above; however, the CO₂ conversion efficiency was lower than that obtained using NiO/SBA-15 synthesized by direct synthesis method. At low temperatures (lower than 500 °C), because CO₂ methanation occurred, the CO₂ conversion percentage was higher than that obtained using NiO/SBA-15 synthesized by direct synthesis method. Therefore, NiO/SBA-15 obtained by direct synthesis method favoured the RWGS reaction, but that obtained by post synthesis method catalyzed both RWGS reaction (by-reaction) and methanation (main reaction), which suited for methanation. Using NiO/SBA-15 obtained by direct synthesis method, the CO selectivity was 100 %, regardless of the temperature (Fig. 4b). Using NiO/SBA-15 obtained by post synthesis method, the CO selectivity was 100 % at temperatures of 700 °C and above. But at lower temperatures it was not 100 %, owing to the formation of CH₄. Therefore, CO selectivity seems to be affected by synthesis method. However, based on the above TEM observations, we found that the dispersion of NiO particles, instead of synthesis method, affected the selectivity to CO.

CO₂ conversion and CO selectivity with different 10 wt% NiO/SBA-15 obtained by the direct synthesis and post-synthesis methods. Notes CO₂:H₂ = 60:60 mL/min

We varied the NiO loading to investigate the RWGS reaction using NiO/SBA-15 obtained by direct synthesis method. CO₂ conversion increased with increasing temperature, regardless of the NiO loading, which is far higher than conventional Cu catalysts (Chen et al. 2003, 2004). When the NiO loading was less than 15 wt%, the CO selectivity was close to 100 %, regardless of temperature. This indicates that NiO particles were monodispersed when the NiO loading exceeded 20 wt%, and the CO selectivity was 100 % at temperatures of 600 °C or above, regardless of the NiO loading (Fig. 5a). However, at temperatures below 600 °C the CO selectivity was lower than 100 %, owing to the formation of CH₄, indicating NiO particles aggregated. Therefore, CO selectivity was influenced by NiO loading at low temperatures. To obtain CO with a selectivity of 100 %, the RWGS reaction should therefore be performed at an appropriate temperature and using NiO/SBA-15 with an appropriate NiO loading (NiO dispersion) as a catalyst.

CO selectivity using NiO/SBA-15 obtained by the direct synthesis method with different NiO loadings and CO₂ conversion using NiO/SBA-15 obtained by the post-synthesis method with different NiO loadings. Notes CO₂:H₂ = 30:210 mL/min

Finally, methanation was performed with NiO/SBA-15 obtained by post synthesis method and with different NiO loadings. CO₂ conversion increased with increasing NiO loading (Fig. 5b). When 70 wt% NiO/SBA-15 was employed, CO₂ conversion of 99.2 % was achieved. These values are the highest yet reported comparison with the previous report (Perkas et al. 2009). The value was still 89.4 % at an H₂ to CO₂ ratio of 4:1. These values are the highest yet reported. The CH₄ selectivity was 100 % when the NiO amount exceeded 50 wt%, regardless of the NiO/SBA-15 preparation method; in other words, release of CO as a by-product was completely suppressed.

Novel NiO/SBA-15 catalysts were obtained with different synthesis methods for producer gas conversion. The synthesis method affected the location and dispersion of NiO particles in SBA/15. High temperature facilitated RWGS reaction, regardless of synthesis method. At low temperature, CO₂ conversion reaction was dependent on synthesis method. With direct synthesis method, NiO loading affected CO selectivity only at low temperatures owing to CH₄ formation. With post synthesis method, CO₂ conversion to CH₄ through methanation increased with increasing NiO loading.