Gallium-Promoted Ni Catalyst Supported on MCM-41 for Dry Reforming of Methane

Abstract

The stability and catalytic activity of mesoporous Ni/MCM-41 promoted with a Ga loading of (0.0, 1.0, 1.5, 2.0, 2.5, and 3.0 wt %) as an innovative catalyst was examined for syngas production via CO₂ reforming of CH₄. The objective of present work was to develop a potential catalyst for CO₂ reforming of methane. For this purpose different loadings of gallium were used to promote 5% nickel catalyst supported on MCM-41. An incipient wetness impregnation method was used for preparing the catalysts and investigated at 800 °C. Physicochemical characterization techniques—including BET, XRD, TPD, TPR, TEM, and TGA—were used to characterize the catalysts. The addition of small amounts of Ga resulted in higher surface areas with a maximum surface area of 1036 m²/g for 2.5% Ga. The incorporation of Ga to the catalyst decreased the medium and strong basic sites and reduced the amount of carbon deposited. There was no weight loss for 3%Ga+5%Ni/MCM-41. The 2% Ga loading showed the highest CH₄ conversion of 88.2% and optimum stability, with an activity loss of only 1.58%. The Ga promoter raised the H₂/CO ratio from 0.9 to unity.

1. Introduction

Carbon dioxide is a greenhouse gas that contributes to more than 90% of the global greenhouse gas emissions. Increased greenhouse gas emissions increase global temperatures and change wind patterns [1]. Decreasing the amount of CO₂ in the atmosphere is the biggest challenge that needs new ideas and technologies. Chemical transformation of CO₂ into useful products is gaining attention. The high abundance and low cost of CO₂ are the main advantages of using it as a feedstock in organic syntheses [2,3]. Amongst the chemical transformation methods, dry reforming of methane (DRM), where methane and CO₂ are used as the feedstock to produce synthesis gas, is a very promising technology for environmental protection and energy production [2]. It produces H₂/CO, which is one of the main building blocks of chemical and petrochemical products, in a ratio of unity.

$${\mathrm{CO}}_2+{\mathrm{CH}}_4\to 2{\mathrm{H}}_2+2\mathrm{CO}\to \Delta {\mathrm{H}}_{298\mathrm{K}}^{°}=260.5\mathrm{kJ}/\mathrm{mol}$$

(1)

Reaction in Equation (1) transforms CH₄ and CO₂ greenhouse gases into valuable syngas (H₂ and CO). The CO₂-reforming of CH₄ has been examined with noble metals (Pt and Pd); although these metals showed considerable activity and stability, their high price and inadequate availability restrict their use. Transition metals such as Ni-based catalysts are cost-effective and exhibit similar activity to noble metals for (DRM) [4]. However, the disadvantage of Ni-based catalysts for DRM comes from the severe deactivation resulting from carbon deposition and loss of the active metal surface area by sintering [5,6,7]. Extensive research has been performed to design Ni-based catalysts that resist coking. The type of support for Ni-based catalysts shows an important role in the catalytic activity and distribution of the active metal sites [8]. There has been extensive research using different catalyst supports such as CeO₂-ZrO₂ [9,10], Al₂O₃-ZrO₂ [11], and Al₂O₃-MgO [12].

Silicate-structured mesoporous materials like MCM-41 and SBA-15 have played a significant role as supports in catalysis research [13,14]. Chen et al. [15], showed the advantages of mesoporous silica supported on a Ni-La₂O₃ catalyst. The catalyst revealed good activity, stability, and efficiency for DRM. Excellent catalytic performance using MCM-41 support for the steam reforming of methanol has been reported in the literature [16,17]. Similarly, catalysts supported on MCM-41 displayed an excellent activity and stability in the DRM [18,19,20,21]. Ni particle dispersion into the structure of the porous support is another way to suppress carbon accumulation. Xu et al. [22] designed a mesoporous alumina supported catalyst that accommodated the dispersion of Ni. They reduced the coke formation by adding basic promoters into the structure of the alumina. Ordered mesoporous silica is appropriate for accommodating the nanoparticles of metals in its mesoporous structure. Specifically, the high surface areas of SBA-15 and MCM-41 promote the dispersion of Ni. The fixed Ni particles in the silica framework were stabilized by the channels in the silica framework [23]. The enhanced stability of Ni-based catalysts on mesoporous silica supports has been reported in DRM [4,19,20].

Further, a broad research has been carried out to alter the catalyst supports via promoters to slow the deactivation of Ni-based catalysts [24,25,26]. It was found that Rh added to the Ni/MCM-41 catalyst by the one-pot procedure improved both the activity and time-on-stream stability of the catalyst [24]. An optimum amount of gallium oxide was required to avoid extreme acidity and enhance butadiene selectivity when magnesium silicate catalysts promoted with Ga were used for the conversion of ethanol into butadiene [27]. The promotion with Ga of a SiO₂-supported Cu catalyst in the hydrogenation of CO₂ to methanol was studied [28]. The results indicate that Ga promotes Cu and increases methanol selectivity, probably by forming new active sites during the methanol formation without altering the Cu oxidation state which, under reaction conditions, stays mostly metallic.

Ga is known to modify the acidic properties of the catalyst [27,29,30], while low Ga additions also increase the surface area [27]. The objective of present work was to develop a potential catalyst for CO₂ reforming of methane. For this purpose, different loadings of gallium were used to promote 5% nickel catalyst supported on MCM-41. To our knowledge, there is no reported work on using Ga as a promoter for DRM. This study will explore Ga as a promoter in Ni/MCM-41 catalysts and investigate the effect of different loadings (0.0, 1.0, 1.5, 2.0, 2.5, and 3.0 wt %) of Ga on Ni/MCM-41. The catalysts were tested using TPD, BET, XRD, TPR, TEM, and TGA.

2. Results and Discussion

Figure 1, depicts the N₂ adsorption-desorption isotherms of 5%Ni+x%Ga/MCM-41 (x = 0, 1, 1.5, 2, 2.5) catalysts, while

Table 1. Characterization of the catalysts used for methane reforming

Catalyst	BET m2 g−1	P.V cm3 g−1	P.D nm
MCM-41	1132.72	0.664	2.57
5%Ni/MCM-41	897.85	0.485	2.48
1%Ga+5%Ni/MCM-41	917.38	0.664	2.74
1.5%Ga+5%Ni/MCM-41	978.56	0.660	2.74
2%Ga+5%Ni/MCM-41	983.33	0.664	2.74
2.5%Ga+5%Ni/MCM-41	1036.02	0.593	2.57
3%Ga+5%Ni/MCM-41	979.91	0.560	2.57

BET Surface Area; P.V BJH Adsorption Cumulative Volume of Pores Between 1700 nm and 300nm diameter; P.D BJH Adsorption average port diameter (4 v/A).

presents the pore diameter, surface area, and pore volume for the modified and unmodified Ni-based catalysts. The modification by small amounts of Ga resulted in higher surface areas with a maximum surface area of 1036 m²/g for 2.5% Ga; conversely, as the loading increased to 3%, the surface area slightly decreased. A plausible explanation for the effect of the promoter Ga relates to the fact that small amounts Ga increases the dispersion of the active metals of the catalyst. However, when the amount is further increased, it is possible that it overlaps the Ni sites promoting multilayer deposition—hence reducing the available surface area. This factor was observed before [31]. The MCM-41 was prepared from different precursors and verified the tested one in this work and the surface area determination was repeated. The result was placed in the revised manuscript. The BET value for bare MCM-41 tends to be 1132.72 m²/g.

The variation of pore diameter, surface area, and the pore volume was not significant for the modified catalysts. As a consequence, there is no connection to be made between the catalytic activity and the surface area of the samples.

X-ray diffraction patterns for fresh and spent 5%Ni+x%Ga/MCM-41 (x = 0.0, 1.5, 2.0, 2.5, and 3.0) catalysts are shown in Figure 2. It can be observed that MCM-41 has no diffraction peaks due to the amorphous nature of mesoporous silica. Upon the XRD measurement for the fresh calcined Ni loaded sample (Figure 2a), NiO characteristic peaks were observed which is consistent with JCPDS no. 01-089-5881. The addition of various amounts of Ga did not lead to the appearance of any new phases, but only increased the intensity of the NiO peaks, suggesting that Ga enhanced the crystallization and growth of the Ni particles. However, after performing the H₂ reduction on the spent samples (Figure 2b), the X-ray diffraction patterns were significantly changed due to the formation of a Ni phase with its characteristic peak (111) at 2θ = 45° and (200) peak at 2θ = 53°. However, another strong diffraction peak, observed at 2θ = 28°, could be attributed to crystallized silica because of the high temperature and reduction conditions. The peak at 2θ = 30° could be attributed to the Ga metal formed from the reduction reaction. The main peaks at 2θ = 35, 45, 47, 50, and 53° could be due to the formation of Ni-Ga intermetallic compounds [32,33]. The amount of Ga loading, however, did not affect the Ni-Ga intermetallic phase.

Figure 3 shows the H₂-TPR profiles of the 5%Ni/MCM-41, 1.0%Ga5%Ni/MCM-41, 1.5%Ga5%Ni/MCM-41, 2.0%Ga5%Ni/MCM-41, 2.5%Ga5%Ni/MCM-41, and 3.0%Ga5%Ni/MCM-41 samples. All catalysts showed two main reduction peaks except the 3.0%Ga5%Ni/MCM-41. It can be viewed that the intensity of the first peak decreased and slightly shifted as the Ga loading increased and eventually disappeared for the 3.0% Ga loading. Both 2.5% Ga and 3.0% Ga presented peaks at high temperatures indicating the relatively stronger metal support interactions compared to other % loadings of Ga. It can also be seen that there was no reduction peak before 360 °C for all catalysts; this implies the disappearance of NiO on the silica pore wall surface [20]. The second peak was significantly shifted from 520 °C to 700 °C for 0.0% and 3.0% Ga, respectively, and the peak intensity broadened. This can be interpreted as the reduction of Ni ions situated at different layers in the pore wall [20], indicating an intense interaction between the Ni particles and the MCM-41 support. The shift of TPR peaks might be reasonably attributed to the strong interaction between the metal and the support and by a reduction in the particle size of the bulk material to ultrafine particles [34].

Owing to the introduction of the second metal, some variations in the reduction configurations are expected for the bimetallic MCM-41 catalysts [4]. The major peak is tentatively attributed to the reduction of Ni²⁺ in the silica structure, although the reduction of the small amount of doped Ga species cannot be excluded [4]. It is clear that the amount of incorporated Ga had a significant influence on the reducibility of the catalysts. It is known that a stronger interaction between the metal and support requires a higher temperature for hydrogen reduction [35].

CO₂-TPD was used to measure the basic sites existing in the catalysts. The temperature programed desorption technique permits the determination of the strength of acidic/basic sites present on the catalyst surface, together with total basicity. Usually in TPD the strengths of the basic sites are reported in terms of temperature range where the chemisorbed CO₂ on the basic sites is desorbed; adsorbate on weaker sites desorbs at lower temperature and that adsorbed on stronger sites desorbs at higher temperature. A low desorption temperature represents CO₂ adsorption on weak basic sites, whereas a high desorption temperature represents strong basic sites [22]. The CO₂-TPD for 5%Ni/MCM-41, 1%Ga5%Ni/MCM-41, and 2%Ga5%Ni/MCM-41 catalysts are compared in Figure 4. All catalysts showed identical CO₂-TPD profiles with three desorption peaks. As shown in

Table 2. Amount of CO₂ desorbed with respect to the peaks.

Catalyst	50–170 °C μmol g−1	170–500 °C μmol g−1	500–800 °C μmol g−1	Total μmol g−1
5%Ni/MCM-41	8.79	43.32	182.35	234.45
1.0%Ga5%Ni/MCM-41	20.26	33.87	110.60	164.73
2.0%Ga5%Ni/MCM-41	22.74	26.30	93.65	142.69
2.5%Ga5%Ni/MCM-41	23.15	42.90	79.13	145.18

and Figure 4, each desorption temperature reflects the strength of the basic sites, whereas the area under the curve of each peak represents the amount of desorbed CO₂ that correlates with the number of basic sites [36]. For example, the desorption peaks for 5%Ni/MCM-41 catalyst centered at 78, 271, and 678 °C had a total basic amount of 234.45 μmol g⁻¹, as listed in Table 2. It can be seen that with the incorporation of Ga into the Ni/MCM-41 catalyst, the desorption peaks shifted to higher temperatures. However, while the intensity of weak basic sites increased, the intensity of the strong and medium basic sites decreased. The decrease in the number of medium and strong basic sites could be ascribed to the formation of extra-framework acid sites [30]. There was a remarkable decrease in the total basic site density as the Ga loading increased e.g., from 234.45 to 142.69 μmol g⁻¹ for 5%Ni/MCM-41 and 2%Ga5%Ni/MCM-41 catalysts, respectively.

Thermogravimetric analyses (TGA) of both fresh and spent catalysts were used to quantitatively measure the carbon formation during the reaction. The TGA analysis was carried out under air atmosphere. The life of the spent catalyst was about a day. For the fresh catalyst, the TGA analysis showed no weight loss and hence no carbon deposition. The typical TGA results from spent catalysts are shown in Figure 5. A significant weight loss was observed between 600 and 750 °C. This can be explained by the combustion of the carbon deposited on the catalyst [24]. Increasing the Ga content in the catalyst decreased the amount of carbon. For example, for 5% Ni/MCM-41, the weight loss was about 16%; however, there was no weight loss for 3.0%Ga5%Ni/MCM-41. This change in weight loss cannot be correlated to the catalytic activity. It is possible that the deactivation of this catalyst is not due to carbon deposition on the external surface, but perhaps due to the deposition of carbon within the pores of the MCM-41 support which would be an irreversible deactivation. It could also be due to the destruction of the MCM-41 structure due to high Ga content.

Figure 6a,b display the TEM images of the 5%Ni/MCM-41 and 2%Ga5%Ni/MCM-41, respectively. The images show that the uniform ordered hexagonal pores of MCM-41 are still observed after metal loading [17], i.e., the long ordered arrangement of the channels in the matrix are preserved [37]. This demonstrates that the Ni particles were homogeneously dispersed in the form of round nanoparticles. The TEM images presented in Figure 6 reveal an average particle size that ranges from 25–40 nm for 5%Ni/MCM-41 and 16–33 nm for 2%Ga5%Ni/MCM-41. The uniform distribution of the Ni nanoparticles in the MCM-41 framework is apparent from the TEM image in Figure 6b.

The locations with darker contrast in the images in Figure 6a are ascribed to the Ni particles [36]. The minor dark spots are attributed to Ni particles located in the channels of the MCM-41 support. The larger dark areas over the pores correspond to Ni oxide agglomerates on the external surface, which resulted from the migration of Ni species out of the supports during the calcination process [34,37]. It can be seen that the Ni nanoparticles (dark points) were embedded within the mesoporous structure when Ga was introduced [38], which reduced the aggregation of NiO nanoparticles during the calcination step [39]. This observation further supports the bigger and larger reduction of the peak of Ga-Ni nanoparticles (TPR, Figure 2) in the chemisorption experiment. TEM images demonstrated a nearly regular hexagonal array of mesoporous channels of the catalyst [40]. The CH₄ and CO₂ conversions are presented in Figure 7. In this investigative stage, we have used TOS to reflect the stability of the Ga-promoted Ni/MCM-41. In this investigation we have not quantified the total number of active sites to determine the turn over frequency. As shown, both CH₄ and CO₂ conversions did not change for the catalysts with Ga loadings from 1–2.5% during the reaction period of 400 min. However, the 3% Ga loading showed lower activity and lesser stability than the 0% Ga-loaded catalyst. This is either because the catalyst is not easily reduced as shown in Figure 2, or due to the destruction of the framework of the MCM-41 support. Ga-incorporated Ni catalysts showed higher catalytic activity than that of the unmodified Ni-MCM-41 (0% Ga) catalyst, except for 3% Ga loading. This indicates that there should be an optimum Ga loading to enhance the catalytic activity and stability. As shown in Figure 8, 2% Ga loading showed a higher conversion of CH₄ and lower activity loss of 88.2% and 1.58%, respectively. The main disadvantage of using catalysts for this process is due to the deactivation. This phenomenon cannot be avoided, but often reduced to an acceptable level. For instance, 2% Ga loading catalyst (2%Ga5%Ni/MCM-41) showed a higher conversion of CH₄ (88.2%) and lower activity loss of 1.58%, for 400 min of TOS.

The most obvious disadvantage of Ni-based catalysts is the poor stability and a H₂/CO ratio of less than unity, which is partly caused by the reverse water gas shift (RWGS) reaction. The H₂/CO ratios for the different Ga loadings are shown in Figure 9. It is clear that the Ga promoter had a significant effect on the H₂/CO ratio as the ratio increased from 0.9 to unity with the addition of Ga. This improvement could possibly be related to the Ga-suppression of the RWGS reaction and/or the enhancement of the DRM rate (H₂ formation rate) relative to the RWGS reaction [28].

Figure 10 shows the H₂/CO ratios and the conversions of CO₂ and CH₄ versus time-on-stream and H₂ selectivity for the 2.0%Ga5%Ni/MCM-41 catalyst. The CH₄ conversion profile indicates the stability and performance. The activity reduction was ~8% during the 30 h time-on-stream. The carbon formation on the catalyst is the cause of the observed deactivation. Likewise, the CO₂ profile presents a deactivation of approximately 5.3%. The conversions of CO₂ are higher than the corresponding CH₄ conversion since CO₂ reacts with the produced H₂ to form water and additional CO that lowers the ratio of H₂/CO. The generated H₂O in this step helped to remove the carbon deposits and hence contributed to the efficiency of the catalyst. Alternatively, the H₂/CO profile gave a near-unity value, indicating good stability for the 2.0%Ga5%Ni/MCM-41 catalyst. The H₂ selectivity curve provided a steady and stable profile for all 30 h time on stream demonstrating the superiority of 2.0% Ga loading.

3. Materials and Methods

3.1. Catalyst Preparation

MCM-41 was synthesized at room temperature (25 °C) using n-hexadecyltrimethylammonium chloride (HTMACl) as the template and tetraethyl orthosilicate (TEOS) as the silica precursor. In a typical synthesis, 392 mL of distilled water, 33 mL of ammonia (28–30% aqueous solution, Sigma-Aldrich, Saint Louis, MO, USA), and 34.05 mL of HTMACl (25% aqueous solution, Sigma-Aldrich) were mixed with vigorous stirring for 10 min, following which 36.45 mL of the silica source, TEOS (98%, Sigma-Aldrich) was added under stirring. Vigorous stirring was continued for 1 h at room temperature. After that, the resulting solid was filtrated and washed several times with distilled water. The washed solid was then dried at 80 °C overnight in a vacuum furnace. Subsequently, the synthesized material was heated up 600 °C to remove the surfactant [41].

Supported Ni catalysts were obtained by the incipient wetness impregnation method. The desired weight of [Ni(NO₃)₃·6H₂O] was placed in double-distilled water, following which the MCM-41 support was impregnated with the active metal salt. The active metal impregnation of the support was performed under stirring at 80 °C for 3 h. then the impregnated catalysts were dried at 120 °C for 12 h, followed by 3 h calcination at 600 °C. The prepared catalysts were then reduced at 800 °C for 1 h under H₂/Ar flow mixture (20/80). The conventional method of catalyst reduction at temperatures below and equal to calcination temperatures is not anymore maintained, partly because it has been observed in TPR that some peaks appear at temperatures higher than that of calcination temperature. Han et al. calcined their catalyst at 550 °C and performed the reduction and reaction at 800 °C [41]. The Ni amount was fixed at 5 wt %; in a previous publication, the optimum amount of Ni was about 5% [42]. The 5 wt % Ni/MCM-41-promoted catalysts with various loadings of Ga (0.0, 1.0, 1.5, 2.0, 2.5, and 3.0 wt %) were formed from the co-impregnation of Ni (NO₃)₃·6H₂O and Ga(NO₃)₃·H₂O using an MCM-41 support as described above. An elemental analysis of fresh 5%Ni+2.0%Ga/MCM-41 catalyst was performed to verify the desired composition. Figure 11 and the embedded table display the spectrum and the elemental compositions from qualitative and quantitative points of view. The embedded table denotes the masses of the catalyst containing about 5% Ni and 2% Ga. Thus the EDX result confirmed that the desired catalyst was achieved.

3.2. Catalytic Activity Test

DRM was carried out at 1 bar in a tubular micro-reactor (ID = 9 mm). The setup of Figure 12 was supplied by PID Eng. and Tech. Prior to the DRM reaction, 0.10 g sample was activated in situ under H₂ flow (20 mL/min) at 800 °C for 60 min. The reactor was then flushed with N₂ for 20 min. The catalyst was then kept at the reaction temperature (800 °C) under N₂ flow, while the reactants CH₄, CO₂, and N₂ were simultaneously introduced at flow rates of 30, 30, and 5 mL/min respectively. The temperature, pressure, and reaction variables were monitored through the reactor panel and a GC (GC-2014 SHIMADZU) unit. Duplicate runs were performed; the variation was within ±5%. The conversions of the reactants were determined using the following equations, respectively.

$${\%\mathrm{CH}}_4\mathrm{conversion}=\frac{{\mathrm{CH}}_4\left(\mathrm{in}\right)-{\mathrm{CH}}_4\left(\mathrm{out}\right)}{{\mathrm{CH}}_4\mathrm{i}}\mathrm{X}100$$

(2)

$${\%\mathrm{CO}}_2\mathrm{conversion}=\frac{{\mathrm{CO}}_2\left(\mathrm{in}\right)-{\mathrm{CO}}_2\left(\mathrm{out}\right)}{{\mathrm{CO}}_2\left(\mathrm{in}\right)}\mathrm{X}100$$

(3)

The same units and conditions stated in our previous paper [7] were used for the BET, TPD, TPR, XRD, TEM, and TGA analyses in the characterization of catalysts before/after reactions.

In the present work, blank runs were performed using similar operating conditions. The results obtained at 800 °C reaction temperature displayed nearly no conversion activity. In addition, using the same experimental conditions we did the blank test using the MCM-41 support alone without active metal of the catalyst. Again, the activity attained was negligible.

4. Conclusions

The incipient wetness impregnation method was employed to form Ni/MCM-41 mesoporous catalysts. Syngas production via CO₂ DRM was investigated using the catalysts. The influence of different Ga loadings (0.0, 1.0, 1.5, 2.0, 2.5, and 3.0 wt %) on the catalytic performance was evaluated after testing and characterizing the catalysts. Ga doping positively affected the surface area. Moreover, the incorporation of Ga to the catalyst decreased the number of both strong and medium basic sites. Carbon deposition over the catalyst also decreased with the addition of Ga. A CH₄ conversion of 88.2% was achieved with 2% Ga loading, which also gave the highest stability with only a 1.58% activity loss. Ga as a promoter has a significant effect on the H₂/CO ratio, as the ratio increased from 0.9 to unity with the addition of Ga.