Pd+Al₂O₃-Supported Ni-Co Bimetallic Catalyst for H₂ Production through Dry Reforming of Methane: Effect of Carbon Deposition over Active Sites

Abstract

Dry reforming of methane (DRM) is gaining global attention due to its capacity to convert two greenhouse gases together. It proceeds through CH₄ decomposition over active sites (into CH_4−x) followed by CH_4−x oxidation by CO₂ (into syngas). Furthermore, CH_4−x oligomerization into coke cannot be neglected. Herein, xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are prepared, investigated for DRM, and characterized with X-ray diffraction, UV-Vis, transmission electron microscopy, temperature-programmed reduction/desorption techniques, and thermogravimetry. Fine-tuning among stable active sites, graphitic carbon deposits, and catalytic activity is noticed. The total reducibility and basicity are found to decrease upon increasing the Co proportion up to 2.5 wt% in the Ni-Co bimetallic Pd+Al₂O₃-supported catalyst. The active sites derived from strong metal–support interaction species (NiAl₂O_x or dispersed CoO_x) are found to be promising in higher levels of activity. However, activity is, again, limited by graphitic carbon which is increased with an increasing Co proportion in the Ni-Co bimetallic Pd+Al₂O₃-supported catalyst. The incorporation of 1.25 wt% Co along with 3.75 wt% Ni over Pd+Al₂O₃ results in the generation of fewer such active sites, extensive oxidizable carbon deposits, and inferior catalytic activity compared to 5Ni/Pd+Al₂O₃. The 2.5Ni2.5Co/Pd+Al₂O₃ catalyst has lower crystallinity, a relatively lower coke deposit (than the 3.75Ni1.25Co/Pd+Al₂O₃ catalyst), and a higher number of stable active sites. It attains a 54–51% H₂ yield in 430 min TOS and 0.87 H₂/CO (similar to 5Ni/Pd+Al₂O₃)

1. Introduction

Global warming is now not only limited to the melting of glaciers and the rise of sea levels but it has become the major cause of tremendous disturbances in seasonal cycles across the globe. The main cause of this issue is attributed to greenhouse gases like CO₂ and CH₄. The catalytic conversion of both CH₄ and CO₂ into syngas (CO+H₂), popularly known as dry reforming of methane (DRM), has brought major attention to the scientific community. More importantly, the product syngas has synthetic utility in industry and may be a hydrogen source to be exploited in order to achieve future clean energy goals. The catalytic conversion of CH₄ and CO₂ into syngas is summarized into two sequential steps. The first step is the decomposition of CH₄ into CH_4−x and (x/2)H₂ [1]. The second step comprises two routes related to CH_4−x oligomerization (into coke) and CH_4−x oxidation by CO₂ (into H₂ and CO). Coke formation over the catalyst surface seriously affects the H₂ and CO yield. Additionally, the deposition of inert coke-like graphite over catalytic active sites may deactivate the catalyst permanently.

Various noble metals, Co, and Ni dispersed over a proper thermally sustainable support are found to be catalytically active for DRM reactions [2]. The methane dissociation energy over Ni was, again, less than that over Pt and Pd [3]. Comparing the activation barrier for CO₂ and CH₄ dissociation, the catalytic activity of Ni was found to be better than that of noble metals [4]; the interaction energy of CH₄ over Ni was found to be 18 kcal/mol whereas over Co, it was 0.7 kcal/mol [5]. After interaction with CH₄, the electronic configuration of metallic Co is not changed, but the electronic configuration of Ni is changed to s^0.54d^9.42 (concerning d⁸s² electronic configuration of metallic Ni) [6]. Cobalt addition increased the active oxygen amount/high oxygen affinity and coke resistance [6,7,8]. The Co⁺²/Co³⁺ ratio over the catalyst’s surface also nurtures oxygen vacancy. The vacancy induces abundant unsaturated coordination sites and high-energy dangling bonds. These vacancies are prominent sites for CO₂ adsorption and dissociation [9,10]. Huang et al. identified adsorption structures of species involved in DRM over a specified Co plane [11]. The CH₄ decomposition capacity of cobalt nanoparticles was also recognized [12]. Jana et al. showed 33 mol H₂ production per mol of Co from CH₄ at 600 °C. Cobalt–aluminum mixed oxides like Co₂AlO₄ were reduced under a CH₄ stream and can decompose CH₄ further [13]. Using DFT and micro-kinetic modeling, it was found that CH₄ dissociation was a rate-determining step in DRM, and the Cu (111) face also induced surface carbon coupling which tended to form surface carbon clusters and ultimately prompted catalyst deactivation [14]. Gonzalez-Delacruz et al. found that adjacent nickel atoms prevented carbon deposition over cobalt sites, and cobalt sites remained protected from deactivation towards DRM [15]. The Ni and Co bimetallic combination has strong synergy [16]. It nurtures the Ni-Co alloy phase and NiCo₂O₄ species over the surfaces of catalysts [17,18]. Ni-Co alloy had excellent CO₂ dissociation ability and weak chemisorption of H₂ and NiCo₂O₄ had higher reducibility [17,19]. Ni-Co synergy also increased Ni dispersion over alumina support [20]. Pd could transfer electrons to CO₂ and bring about strong structural changes [21]. Pd–bidentate format species (from a CO₂ carbon source) as well as Pd–CO (from a CH₄ carbon source) [22] justified the strong interaction of Pd with CO₂ as well as CH₄. The spillover of H₂ over Pd was found to increase the reducibility of NiO also [23]. However, the Ni–Pd synergy was found to be greatly dependent on types of support [24]. Ni–Pd-based catalysts are known for a higher coke tolerance compared to monometallic catalysts [25]. Ni–Pd over alumina induced more exposure of Ni° sites with better dispersion of Ni [23].

The idea of dispersion of bimetallic N–Co over an Al₂O₃ support seems more efficient than alumina-supported Ni and alumina-supported Co catalysts. The incorporation of a small amount of Pd along with support may induce CO₂ interaction, NiO reducibility, and CH₄ dissociation over a Ni–Co bimetallic Al₂O_3-supported catalyst. Herein, we have prepared a Pd+Al₂O₃-supported 5wt% Ni–Co catalyst, employed it for DRM, and characterized it with X-ray diffraction, “surface area & porosity”, ultra-violet spectroscopy, transmission electron microscopy, temperature-programmed reduction/desorption techniques, and thermogravimetry techniques. The novelty of this research lies in explaining the synergic interaction between different ratios of Ni and Co over a Pd+Al₂O₃ support in terms of surface reducibility, crystallinity, and basicity, which are the regulating factors in DRM activity and coke precipitation. The fine correlation between characterization results and catalytic activity may help to develop a practical DRM catalyst soon.

2. Results

2.1. Characterization Results and Discussion

The X-ray diffraction pattern of fresh and spent xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalyst systems are shown in Figure 1. The Pd+Al₂O_3-supported 5 wt% Ni catalyst has tetragonal PdO phase (at Bragg’s angle 2θ = 34°, 55°, 71.24°; JCPDS reference number 01-075-0584) and cubic NiAl₂O_x phase (at Bragg’s angle 2θ = 37.6°, 39.6°, 45.8°, 61°, 67°; JCPDS reference number 00-020-0776) (Figure 1A). It is interesting to note that after the reaction over the spent-5Ni/Pd+Al₂O₃ catalyst system, the tetragonal PdO phase mostly disappears, and the intensity of the cubic NiAl₂O_x phase is depleted but the cubic alumina phase intensifies. Most of the cubic phase of Al₂O₃ overlaps with the cubic NiAl₂O_x phase excluding the diffraction pattern at 42.8° (JCPDS reference number 00-004-0880). Upon dispersing 3.75 wt% Ni and 1.25 wt% Co active sites together over the Pd+Al₂O₃ support, the same diffraction patterns are observed (Figure 1B). However, the spent 3.75Ni1.25Co/Pd+Al₂O₃ catalyst showed a higher-intensity alumina peak of about 42.8° (than the spent-5Ni/Pd+Al₂O₃) and a new diffraction peak for graphitic carbon (at Bragg’s angle 2θ = 26.43°; JCPDS reference number 00-008-0415) [26]. Upon incorporating an equal proportion of Co along with Ni over the Pd+Al₂O₃ support (2.5Ni2.5Co/Pd+Al₂O₃), the intensity of diffraction patterns declines (Figure 1C, Figure S1). Upon a further increase in the proportion of Co in the bimetallic supported system, the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst showed an additional diffraction peak for cubic Co₃O₄ (at Bragg’s angle 2θ = 31.22°, 36.81°, 59.2°, 65.11°; JCPDS reference number 01-080-1533) (Figure 1D). The 5Co/Pd+Al₂O₃ catalyst has a lower-intensity peak than the 5Ni/Pd+Al₂O₃ catalyst; the former has cubic CoAl₂O₄ mixed oxide phase (at Bragg’s angle 2θ = 31.36°, 36.5°, 45.6°, 59.63°, 65.23°; JCPDS reference number 00-003-0896), but the latter has mixed cubic NiAlO_x phase (Figure 1F). The intense diffraction pattern for cubic Al₂O₃ phase (at Bragg’s angle 2θ = 37.4°, 42.82°, 45.79°, 67.31°; JCPDS reference number 00-004-0880) is also present over the 5Co/Pd+Al₂O₃ catalyst.

The information regarding graphitic carbon peaks over the spent xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalyst system is needed for emphasis (Figure 1G). Spent 5Ni/Pd+Al₂O₃ has no graphitic carbon peak. Among Ni-Co active sites, upon increasing the Co proportion over Pd+Al₂O₃, the intensity of the crystalline carbon peak increases over the spent catalyst. For spent 1.25Ni3.75Co/Pd+Al₂O₃, the graphitic carbon peak is shifted towards a lower angle, indicating an expansion of lattice parameters in graphitic crystallites. However, the spent 5Co/Pd+Al₂O₃ catalyst shows a diffuse intensity of the graphitic carbon peak. This indicates that graphitic carbon synthesis over the catalyst’s surface excels when active sites are composed of two metals.

The N₂ adsorption isotherm and porosity distribution of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are shown in Figure 2A,B. The surface area, pore volume, and pore diameter of each catalyst are shown in Figure 2C. All catalysts have a type IV isotherm having an H1 hysteresis loop. This indicates the presence of cylindrical mesopores. The pore size distribution plot (dV/dlogW vs. W) shows a mono-modal pore size distribution of about 8 nm and the average pore size falls between 8.4–8.6 nm. The pore size is not affected much by different loadings of Ni and Co over the Pd/Al₂O₃ support. The pore size distribution and average pore size of catalysts are not affected much upon Ni and Co loading over the Pd/Al₂O₃ support.

Pd+Al₂O₃-supported Ni has enhanced surface parameters (surface area, pore volume, and pore diameter) compared to the Pd+Al₂O_3-supported Co catalyst. This indicates the extent of interactions of two different metal oxides over the Pd+Al₂O₃ surface. Cobalt oxide crystallites are more readily deposited inside the pores of the Pd+Al₂O₃ support than nickel oxide. 5Ni/Pd+Al₂O₃ has a 135 m²/g surface area, a 0.37 cm³/g pore volume, and a 8.4 nm pore diameter. If 5 wt% metal loading over the Pd+Al₂O₃ support is maintained by “3.75 wt% Ni and 1.25 wt% Co”, the surface area is decreased by 25% whereas pore volume is diminished by 27% compared to the 5Ni/Pd+Al₂O₃ catalyst. This indicates that the dispersion of bimetallic oxide (Ni and Co) is not as similar as the dispersion of monometallic nickel oxide over the Pd+Al₂O₃ support. Some of the metal oxides may also diffuse into the pores and diminish the surface area and pore volume over the 3.75Ni1.25Co/Pd+Al₂O₃ catalyst more so than the 5Ni/Pd+Al₂O₃ catalyst.

However, if an equal proportion of Ni and Co is dispersed over the Pd+Al₂O₃ support, surface area, and pore volume are improved in comparison to the rest of the supported bimetallic systems. This indicates that equal proportions of both Ni and Co have a better interaction with the Pd+Al₂O₃ catalyst than the rest of the supported bimetallic system.

H₂ temperature-programmed reduction (H₂-TPR) profiles of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are shown in Figure 3A. The 5Ni/Pd+Al₂O₃ catalyst has a single broad reduction peak in the region of 500–825 °C which is attributed to “strongly interacted reducible NiO” or the reduction of NiAl₂O_x species [27]. Also, in XRD, cubic NiAl₂O_x phase is detected. Previously, metallic Ni exsolved from the reduction of NiAl₂O_x (upon reduction) was claimed to be the cause of the generation of stable catalytic active sites for dry reforming of methane [27]. On decreasing the Ni content to 3.75 wt% and incorporating 1.25 wt% Co over the Pd+Al₂O₃ support, the reduction peak intensity for “strongly interacted reducible NiO” is decreased and two new peaks in the low-temperature range of about 100 °C and 375 °C are noticed. However, the total uptake of H₂ is marginally decreased. The previous reduction peak (about 100 °C) represents the reduction of highly dispersed PdO species and the second one (375 °C) represents for the reduction of Co₃O₄ [28,29,30]. Upon equal proportions of Ni and Co in the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst, lower reduction temperature peaks (about 100 °C and 375 °C) disappear whereas the peak intensity of about 500–825 °C increases. Here, a fine-tuning between Ni and Co is seen. The enhanced peak intensity of about 500–825 °C may be attributed to the enhanced reduction of strongly interacted NiO species or the reduction of a “strongly interacted highly dispersed CoO_x cluster” [28]. In XRD, the crystalline peak intensity for cubic NiAl₂O_x phase is decreased more so for the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst than for the 3.75Ni1.25Co/Pd+Al₂O₃ catalyst. So, it can be said that improving the reduction profile in the temperature range of 500–800 °C is a combined contribution of the reduction of “strongly interacted NiO” and the reduction of “highly dispersed CoO_x species”. It is noticeable that, upon increasing the Co content (or decreasing the Ni content) up to 2.5 wt%, the total amount of H₂ consumption decreases or the total concentration of reducible species decreases. In Pd+Al₂O₃-supported Ni-Co bimetallic catalysts, if the wt% of Co is increased more so than Ni, the total H₂ consumption of the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst is greater than that of 2.5Ni2.5Co/Pd+Al₂O₃. However, the reduction profile of the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst is altered. The reduction peak for strongly interacted NiO species was suppressed extensively as well as the reduction peaks for PdO species and Co₃O₄ is intensified over the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst. This indicates that the fine-tuning between Co and Ni is disturbed completely over the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst. However, it can be said that the presence of a higher wt% of Co may induce a higher population of PdO species to reduce and vice versa. It is noticeable that, over the 5Co/Pd+Al₂O₃ catalyst, both reduction peak of PdO and Co₃O₄ is shifted to a relatively higher temperature, indicating higher levels of metal–support interaction. Except for when Ni/Co = 1, higher concentrations of reducible Co₃O₄ and PdO species are observed with a decreasing ratio of Ni/Co.

CO₂ temperature-programmed desorption (CO₂-TPD) profiles of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are shown in Figure 3B. It is categorized into three basic regions. The CO₂ desorption peak is about 100 °C for weak basic sites, about 300 °C for moderate-strength basic sites, and about 700 °C for strong basic sites [31]. The basicity of the surface is contributed to by surface hydroxyl and surface oxygen [32]. Surface hydroxyl constitutes weak basicity. The surface oxygen which generates carbonate/formate intermediate upon interaction with CO₂ is related to medium-strength basic sites. The surface oxygen which generates more stable carbonates upon reaction with CO₂ constitutes strong basic sites [31,33]. From the CO₂-TPD, a general observation can be pointed out. The total amount of CO₂-desorbed gases is decreased with increasing Co amounts up to 2.5 wt% (or decreasing Ni amounts to 2.5 wt%). Upon further increasing the Co loading, the CO₂-desorbed amount is increased over the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst. The CO₂-TPD profile representing a higher loading of Co than Ni over the 5Co/Pd+Al₂O₃ support has a different look than the rest of the catalyst. It has an additional peak of about 850 °C. The peak intensity of about 850 °C is increased upon further loading of Co. This indicates that the CO₂ desorption peak is related to the interaction of CO₂ over Co. It can also be said that CO₂ strongly interacts with Co, more so than Ni.

The cyclic H₂TPR–CO₂TPD–H₂TPR experiments for xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 2.5, 0) catalysts are shown in Figure 3C–E. After sequential H₂-TPR (reduction), CO₂-TPD (oxidation), and H₂-TPR (reduction) treatments of the catalyst, a reducible peak of about 750–900 °C is observed. The reduction peak in such a high-temperature range was previously attributed to the reduction of NiO which was supported over M–O–M′ (M ≠ M′; M = W, Zr, Al and M′ = W, Zr, Al) species [34,35]. Here also, the reduction of NiO or Co₃O₄ over a Pd–O–Al support can be expected at such a high reduction temperature. The cyclic H₂TPR–CO₂TPD–H₂TPR experiments indicate that under high-temperature reactions (800 °C), some of the active species, such as Ni and Co, are regenerated into NiO and Co₃O₄ under the CO₂ stream (which is reduced by the last reductive treatment). This indicates that under the oxidizing environment (in the CO₂ stream) during DRM, some active sites are oxidized and become inactive for further catalyzing DRM reactions. It can be noticed that the intensity of such a peak is maximal for the 5Co/Pd+Al₂O₃ catalyst (Figure 3F).

The bandgap of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts and thermogravimetry analysis of spent xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are shown in Figure 4. Without Co, the bandgap is 1.58 eV over 5Ni/Pd+Al₂O₃. It is noticeable that the band gap between the valence band and conduction band is continuously decreasing as the proportion of Co is increased in xNi(5−x)Co/Pd+Al₂O₃ (x = 0,1.25, 2.5, 3.75, and 5) catalysts. The band gap is minimal (0.34 eV) over the 5Co/Pd+Al₂O₃ catalyst. Thermogravimetric analyses of fresh and spent xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts are carried out and shown in Figure 4B. The fresh catalyst samples have minimal loss compared to the spent catalyst samples. The TGA results indicate that Pd+Al₂O₃-supported Ni catalysts have minimal coke decomposition (~4% weight loss) whereas, upon introduction of just 1.25 wt% Co (along with 3.75 wt% Ni) over the Pd+Al₂O₃ support, the coke deposition is exceptionally high (46.87%). Upon increasing the proportion of Co further, the coke deposition is decreased. The 2.5Ni2.5Co/Pd+Al₂O₃ catalyst shows ~21% weight loss whereas ~11–13% weight loss is evident over 1.25Ni3.75Co/Pd+Al₂O₃ and 5Co/Pd+Al₂O₃ catalysts. To confirm the carbon type, Raman analysis of spent catalyst samples is undertaken (Figure 4C). All catalysts have three Raman bands at 1070 cm⁻¹, 1340 cm⁻¹, and 1570 cm⁻¹ for vibration of C–C sp³ (T band), defect carbon (D band), and ordered carbon (G band), respectively [36], [37] Interestingly, D bands and G bands are most intense for the spent 3.75Ni1.25Co/Pd+Al₂O₃ catalyst. Thus, it can now be concluded that over the spent 3.75Ni1.25Co/Pd+Al₂O₃ catalyst had a maximal coke deposit and the coke deposit was mostly sp² hybridized carbon with defective and ordered carbon structures. The relatively higher intense C–C sp³ (T band) is noticed over the spent 5Co/Pd+Al₂O₃ catalyst (more so than other catalysts).

Transmission electron microscopy of catalysts and particle size distribution are depicted in Figure 5. In spent catalyst samples, the larger particle size is clearly evident. In the case of the spent 5Ni/Pd+Al₂O₃ catalyst, no carbon formation is noticed but over the spent 2.5Ni2.5Co/Pd+Al₂O₃ catalyst, carbon nanotubes are visualized easily. In the XRD, the diffraction pattern for graphitic carbon was absent. This indicates that 5Ni/Pd+Al₂O₃ is coke-resistant.

2.2. Catalytic Activity Results and Discussion

Catalytic activity results of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts in terms of H₂ yield (%) and CO yield (%) are shown in Figure 6A,B. The activity experiments were reproduced twice for all the catalysts. It was noted that the error of catalytic activity was 2–4% for all the experiments. The catalytic activity results and TOF results (at the end of 430 min TOS) are also shown in tabular form in Figure 6C. H₂ yield vs. the H₂/CO ratio of catalysts is presented in Figure S1. After the decomposition of CH₄ over active sites (into CH_4−x; x = 1–4) over different catalyst systems, two phenomena ran parallel. One is the oxidation of CH_4−x by CO₂, leading to DRM reaction, and another is the polymerization of CH_4−x species, leading to coke deposit. Further, the coverage of inert carbon species over active sites may seriously affect DRM activity. Apart from the coke deposit, if only DRM reactions take place over a catalyst, the H₂ yield and CO yields should be similar (CO₂ + CH₄ → 2CO + 2H₂). But here, the CO yield is always found to be above the H₂ yield over each catalyst system. This indicates the presence of a reverse water gas shift reaction which consumes the H₂ (CO₂ + H₂ → CO + H₂O) and so the CO yield surpasses the H₂ yield.

The H₂-TPR and CO₂-TPD profiles of the catalyst systems reflect the information regarding the catalysts’ reducibility and basicity (extent of CO₂ interaction). A general trend of reducibility and basicity over xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts is observed. Upon decreasing the Ni wt% from 5 to 2.5 (or increasing the Co wt% from 0–2.5), the total H₂ consumption and CO₂ desorption of the catalyst are decreased from 2.7 to 1.7 cm³/g and from 2.1 to 0.4 cm³/g, respectively. This means that total reducibility and total basicity (CO₂ interaction) are decreased upon increasing the Co proportion up to 2.5 wt% over xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalyst systems. However, the catalyst activity results of xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts do not follow this trend. The 3.75Ni1.25Co/Pd+Al₂O₃ catalyst has minimal catalytic activity among bimetallic catalysts. This indicates that only total reducibility and total basicity do not monitor the reaction. Other factors like types of reducible species, types of surface basic species, total carbon deposits, and types of carbon in deposit should also be considered.

Ni supported over Pd+Al₂O₃ has a stable Ni site derived from the reduction of cubic NiAl₂O_x phases. The TGA result of the spent 5Ni/Pd+Al₂O₃ catalyst shows the presence of a minimum carbon deposit (4% weight loss) and the Raman result shows the presence of both diamond and graphitic carbon bands over the spent 5Ni/Pd+Al₂O₃ catalyst. Regarding the carbon deposit, due to the lesser amount of graphitic carbon over the spent 5Ni/Pd+Al₂O₃ catalyst, a graphic carbon crystalline peak is not observed in XRD analysis. The 5Ni/Pd+Al₂O₃ catalyst does not show crystalline carbon peaks after the DRM reaction. It seems that the catalytic active Ni site derived from NiAl₂O_x is highly stable; it dissociates the CH₄ (into CH_4−x; x = 1–4) and it also resists the accumulation of CH_4−x species for oligomerization. Cobalt supported over Pd+Al₂O₃ also has active Co sites derived from the cubic Co₃O₄ phase. The stability of the 5Ni/Pd+Al₂O₃ catalyst’s active sites is higher than those derived from the 5Co/Pd+Al₂O₃ catalyst. The instability of Co active sites is also evident in the cyclic H₂TPR–CO₂TPD–H₂TPR experiment. The cyclic experiment shows the highest-intensity reduction peak at about 800 °C (in the last reductive treatment) over 5Co/Pd+Al₂O₃, indicating the oxidation of active Co species (or inactiveness of active sites) by CO₂ during DRM. Further, the spent 5Co/Pd+Al₂O₃ catalyst also has a remarkable peak intensity of graphitic carbon whereas, over the spent 5Ni/Pd+Al₂O₃ catalyst, the graphitic carbon peak is absent. TGA results are similar. The spent 5Ni/Pd+Al₂O₃ catalyst has minimal weight loss (4.13%) whereas the spent 5Co/Pd+Al₂O₃ catalyst has three times more weight loss. The spent 5Ni/Pd+Al₂O₃ catalyst has no graphitic carbon but a small amount of oxidizable carbon deposit whereas the spent 5Co/Pd+Al₂O₃ catalyst has marked growth of both amorphous and graphitic carbon deposits. Altogether, the 5Ni/Pd+Al₂O₃ catalyst has higher and constant activity and a 57–56% H₂ yield after 430 min on stream. The catalyst activity of 5Co/Pd+Al₂O₃ is lowest (43% H₂ yield) and it declines very fast (34% H₂ yield) within the 430-minute timeframe on stream. The H₂/CO ratio of the 5Ni/Pd+Al₂O₃ catalyst is a maximum (0.87) whereas it is a minimum (0.78) over 5Co/Pd+Al₂O₃ (Figure S1). Low H₂ production over 5Co/Pd+Al₂O₃ results in a lower H₂/CO ratio of the catalyst. The major cause of inferior activity over 5Co/Pd+Al₂O₃ seems to be due to less stable active sites vis-à-vis relatively higher coke deposition during DRM reactions.

Upon incorporation of 1.25 wt% Co along with 3.75 wt% Ni over a Pd+Al₂O₃ support, the catalyst experiences in a sharp drop in surface area and pore volume. The surface of 3.75Ni1.25Co/Pd+Al₂O₃ is populated by relatively lesser amounts of catalytic active Ni species (derived from NiAl₂O₄) or catalytic active Co (derived from dispersed CoO_x species) and Co species derived from Co₃O₄ species. The XRD patterns show a higher intensity of graphitic carbon (than 5Ni/Pd+Al₂O₃) and the TGA result shows extensive weight loss over the spent 3.75Ni1.25Co/Pd+Al₂O₃ catalyst. This indicates, upon incorporating Co, that the distribution of reducible surface species is modified greatly and brings DRM and carbon deposition. The most intense Raman band over the spent 3.75Ni1.25Co/Pd+Al₂O₃ catalyst again confirms the high-level coke deposit over the catalyst. Raman analysis of catalysts showed that these carbon deposits belong to sp² hybridized defective diamond carbon (D band) and ordered graphitic carbon (G) structures. In total, the catalytic activity of 3.75Ni1.25Co/Pd+Al₂O₃ is decreased (54–51% H₂ yield during 430 min of TOS) but weight loss is increased extensively (~47%) with respect to the 5Ni/Pd+Al₂O₃ catalyst.

Dispersing an equal proportion of Ni and Co over the Pd+Al₂O₃ catalyst shows a decrease in total crystallinity and an adequate population of catalytic active Co or Ni species, which is derived by the reduction of CoO_x or NiAl₂O_x species (strongly interacted reducible species). Interestingly, frequently observed reducible species like PdO and Co₃O₄ are completely absent over the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst. The surface area and pore volume of the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst also progress more than those of the 3.75Ni1.5Co/Pd+Al₂O₃ catalyst. The TOF of 2.5Ni2.5Co/Pd+Al₂O₃ catalyst is found to be 165 h⁻¹ which is greater than that of other catalyst systems in this study (Figure 6C). The H₂/CO ratio of the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst, again, reaches the maximum, 0.87 (equal to the 5Ni/Pd+Al₂O₃ catalyst). The total carbon deposit over the spent 2.5Ni2.5Co/Pd+Al₂O₃ catalyst decreases impressively but graphitic carbon is still intensified along with Co loading. Despite the strong tuning of Co and Ni in favor of DRM reactions, the catalytic activity is improved with respect to the 3.75Ni1.5Co/Pd+Al₂O₃ catalyst but is not better than that if the 5Ni/Pd+Al₂O₃ catalyst. Overall, the evolution of stable active sites over the more expanded 2.5Ni2.5Co/Pd+Al₂O₃ catalyst surface results in an enhancement in catalytic activity (55–53% H₂ yield after 430 min) with respect to the 3.75Ni1.5Co/Pd+Al₂O₃ catalyst.

The incorporation of a greater amount of Co than Ni over Pd+Al₂O₃ causes a depletion in the reduction peak at high temperatures as well as a rise in the reduction peak at low and intermediate temperatures. This indicates the loss of strongly interacted active sites (derived from the reduction of NiAl₂O_x and dispersed CoO_x species) from the catalyst surface. However, cobalt-related stable carbonates are observed over the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst. XRD spectra of the spent catalyst also showed the highest intensity peak for graphitic carbon. Overall, it can be said that the lack of strongly interacted active sites and potential graphitic carbon deposits over the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst resulted in the lowest catalyst performance among bimetallic supported catalysts. The 1.25Ni3.75Co/Pd+Al₂O₃ catalyst showed a 43% H₂ yield with a markable decrease in H₂/CO (0.84) after 430 min on stream.

Overall, the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst showed the highest TOF and activity among the bimetallic catalysts investigated in this study. A comparative table of the different catalyst systems is shown in

Table 1. The comparative table of catalytic activity (in terms of H₂-yield) over different catalyst systems.

Sr. No.	Catalyst Name	Active Sites (wt.%)	CP	AC (g)	GHSV (L/h gcat)	RT (°C)	TOS (h)	Y (H2) %	Ref.
1	Ni/ZrO2-I	10 (Ni)	I	0.05	60	700	-	50	[38]
2	Ni/ZrO2	5 (Ni)	I	0.1	42	700	7	43	[39]
3	Ni/CeO2-ZrO275	5 (Ni)	Co-I	0.1	30	700	24	28	[40]
4	Ni1Ce/ZrO2	5 (Ni)	I	0.1	42	700	7	47	[41]
5	Ni/18wt%CeO2-82wt%ZrO2	8 (Ni)	I	0.15	40	750	50	35	[40]
6	Ni/28mol%CeO2-72mol%ZrO2	5 (Ni)	I	0.1	30	700	21	35	[40]
7	Ni/SiO2	5 (Ni)	I	0.1	24	700	23	22	[23]
8	Ni-SiO2-OA	5 (Ni)	I-OA	0.1	24	700	23	25	[23]
9	Ni-MSN	5 (Ni)	I	0.1	36	700	25	49	[42]
10	Ni/Al2O3	10 (Ni)	I	-	70	600	70	22.7	[38]
11	Ni3TiAl	5 (Ni)	MM	0.1	42	700	7	30	[27]
12	Ni3MoAl	5 (Ni)	MM	0.1	42	700	7	39	[27]
13	Ni/CeO2-Al2O3	10 (Ni)	M1	0.1	30	750	12	12.5	[43]
14	5Ni/5Y-Zr	5 (Ni)	Sg	0.1	42	700	7	45	[44]
15	5Ni/5Mg-Zr	5 (Ni)	Sg	0.1	42	700	7	23	[44]
16	2.5Ni2.5Co/Pd+Al2O3	5 (Ni & Co)	I	0.1	42	800	7	53	This Study

CP: Catalyst’s Preparation Method, AC: Amount of catalyst taken for the reaction, RT: Reaction temperature, TOS: Time on stream, Y: Yield, MSN: Mesoporous Silica Nanoparticle, I: Impregnation method, Co-I: Coprecipitation followed by impregnation, Sg: Sol-gel method, MM: Mechanical mixing, Method 1 (M1): The support is prepared by 0.5–1.2 nm diameter Al₂O₃ sphere saturated by 5 wt.% ceria solution.

. By comparing the catalytic activity results with closely related catalysts, the current catalyst system was found to be more effective than others.

3. Materials and Methods

3.1. Catalyst Preparation

The xNi(5−x) Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalysts were prepared using a wet impregnation method using nitrate precursor solutions of Co and Ni and a 5 wt% Pd+Al₂O₃ (Aldrich Chemical Co., Milwaukee, WI, USA) support. Nickel nitrate hexahydrate (Ni(NO₃)₂6H₂O; 99% purity, Fisher, Schwerte, Germany) and cobalt nitrate hexahydrate (Co(NO₃)₂6H₂O; 99% purity, Aldrich Chemical Company Inc., Milwaukee, WI, USA) were dissolved in distilled water and stirred continuously to form a uniform solution. Thereafter, the catalyst support (5 wt% Pd+Al₂O₃) was added to the precursor solution. The mixture was stirred under heating until a paste was formed. Further, the paste was dried for 12 h at 120 °C and calcined at 600 °C calcination for 3 h. All elements claimed in the catalyst synthesis were confirmed by EDX analysis (Figure S2). The catalysts were abbreviated as xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) where x is wt%.

3.2. Catalyst Performance Evaluation

The dry reforming of methane reaction was carried out over 0.1 g catalysts at 800 °C at 1 atm pressure in a packed-bed stainless steel reactor (internal diameter of 0.91 mm and a length of 30 cm, PID Eng. and Tech Micro Activity) equipped with a K-type stainless steel sheathed thermocouple. An axially positioned thermocouple close to the catalyst bed was used to monitor the reaction temperature. Before the reaction, the catalyst was reduced/activated under H₂ at 800 °C for 60 min then again after the remnant H₂ was purged by N₂. After catalyst activation, a DRM gas feed (CH₄:CO₂:N₂ in the volume ratio of 3:3:1, respectively) was passed through the catalyst bed with 42,000 mL/($g_{cat.}$h) space velocity at an 800 °C reaction temperature for 430 min on stream. The gas composition in the outlet stream was analyzed online using a gas chromatograph (Shimadzu GC-2014, Kyoto, Japan) equipped with a thermal conductivity detector. H₂ yield percent, CO yield percent, and TOF (h⁻¹) were calculated from the following formula [45]:

$${\mathrm{H}}_2 \mathrm{yield} (\%)=\frac{\mathrm{M}\mathrm{o}\mathrm{l}\mathrm{e} \mathrm{o}\mathrm{f} {\mathrm{H}}_2 \mathrm{i}\mathrm{n} \mathrm{P}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}{2\times \mathrm{M}\mathrm{o}\mathrm{l} \mathrm{o}\mathrm{f} {{\mathrm{C}\mathrm{H}}_4}_{\mathrm{i}\mathrm{n}}}\times 100$$

(1)

$$\mathrm{CO} \mathrm{yield} (\%)=\frac{\mathrm{M}\mathrm{o}\mathrm{l}\mathrm{e} \mathrm{o}\mathrm{f} \mathrm{C}\mathrm{O} \mathrm{i}\mathrm{n} \mathrm{P}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}{\mathrm{M}\mathrm{o}\mathrm{l} \mathrm{o}\mathrm{f} {{\mathrm{C}\mathrm{H}}_4}_{\mathrm{i}\mathrm{n}}+\mathrm{M}\mathrm{o}\mathrm{l} \mathrm{o}\mathrm{f} {{\mathrm{C}\mathrm{O}}_2}_{\mathrm{i}\mathrm{n}}}\times 100$$

(2)

$$\mathrm{TOF} ({\mathrm{h}}^{-1})=\frac{{\mathrm{v}}_{{\mathrm{C}\mathrm{H}}_4\times {\mathrm{X}}_{{\mathrm{C}\mathrm{H}}_4}\times \mathrm{P}}}{\mathrm{R}\times \mathrm{T}\times {\mathrm{S}}_{\mathrm{N}\mathrm{i}+\mathrm{C}\mathrm{o}}}$$

(3)

${\mathrm{v}}_{{\mathrm{C}\mathrm{H}}_4}$ = Flow rate of methane, ${\mathrm{X}}_{{\mathrm{C}\mathrm{H}}_4}$ = Conversion ratio of methane, P = Pressure in atm, R = Gas constant in L·atm·K⁻¹·mol⁻¹, T ¼= Temperature in Kelvin, ${\mathrm{S}}_{\mathrm{N}\mathrm{i}+\mathrm{C}\mathrm{o}}$ = The mole of surface Ni and Co atoms in the sample.

3.3. Catalyst Characterization

The catalysts were characterized by X-ray diffraction, surface area and porosity, ultra-violet spectroscopy, H₂ temperature-programmed reduction (H₂-TPR), CO₂ temperature-programmed desorption (CO₂-TPD), cyclic H₂TPR–CO₂TPD–H₂TPR experiments, thermogravimetry analysis, and transmission electron microscopy. Detailed instrument specifications and analysis procedures are described in the supporting information file.

4. Conclusions

To conclude, 5 wt% Ni over a Pd+Al₂O₃ catalyst has Ni as active sites that are derived primarily from strong metal–support interaction species (or NiAl₂O_x) and it is oxidized the least under a CO₂ stream at a DRM reaction temperature. It has no graphitic carbon deposit and acquires the highest 57–56% H₂ yield. Further, 5 wt% Co over the Pd+Al₂O₃ catalyst has Co as active sites that are derived from Co₃O₄ and the active sites are oxidized exclusively under a CO₂ stream during DRM reactions. The 5Co/Pd+Al₂O₃ catalyst has the least activity (43–34% H₂-yield) and a markable graphitic carbon deposit. Upon introducing a 1.25 wt% cobalt proportion in the Ni-Co bimetallic catalyst over a Pd+Al₂O₃ support, the graphitic carbon deposit is enhanced abruptly. Total reducibility and the extent of CO₂ interaction are decreased upon increasing the Co proportion up to 2.5 wt% over xNi(5−x)Co/Pd+Al₂O₃ (x = 5, 3.75, 2.5, 1.25, 0) catalyst systems. Incorporation of 1.25 wt% Co along with 3.75 wt% Ni over Pd+Al₂O₃ results in fewer active sites derived from strong metal–support interaction species (NiAl₂O_x or dispersed CoO_x species), resulting in extensive oxidizable carbon deposits, higher-level graphitic carbon deposits, and relatively less DRM activity (54–51% H₂ yield) compared to 5Ni/Pd+Al₂O₃. Upon incorporation of an equal amount of Co and Ni over Pd+Al₂O₃, the catalyst’s crystallinity decreases and, again, nurtures a higher number of active sites derived from strong metal–support interaction species. The TOF of the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst is a maximum 165 h⁻¹ and the H₂/CO ratio of the 2.5Ni2.5Co/Pd+Al₂O₃ catalyst, again, reaches the maximum 0.87 (equal to 5Ni/Pd+Al₂O₃ catalyst). Among all bimetallic supported catalysts, the catalyst performance of the 1.25Ni3.75Co/Pd+Al₂O₃ catalyst is inferior due to the deficit of strong metal–support interaction species and the highest level of graphitic carbon deposits.