Role of promoters over yttria-zirconia supported Ni catalyst for dry reforming of methane

Abstract

Dry reforming of methane (DRM) bears great hope for the catalytic community as well as environmentalists for its potential to convert two greenhouse gases, CH₄ and CO₂, together into synthetic feedstock “syngas”. The stable tetragonal zirconium yttrium oxide phase over the “Yttria-zirconia supported Ni” catalyst (Ni/YZr) brings >70% CH₄ conversion against 50% CH₄ conversion over zirconia supported Ni catalyst) in 7 h time-on-stream (TOS). The use of the second metal oxide (MO_x; M = Ho, Ga, Gd, Ba, Cs) in a small amount (4 wt%) over Ni/YZ catalyst is found to promote the catalytic activity further. Herein, we have prepared such metal-promoted yttria-zirconia supported Ni catalyst, employed them for DRM and characterized them with surface area porosity, X-ray diffraction, spectroscopic techniques, temperature programmed techniques and transmission electron microscopy. A fine correlation of characterization results with catalytic activity brings out various useful information that would be useful for establishing yttria-zirconia supported Ni catalyst for DRM. Ni stabilized over cubic zirconium holmium oxide phase in 5Ni4Ho/YZr catalyst, cubic zirconium gadolinium oxide phase in 5Ni4Gd/YZr catalyst and cubic zirconium barium oxide phase in 5Ni4Ba/YZr catalyst perform excellent toward DRM. Catalytically, 5Ni4Ho/YZr catalyst achieves CH₄ conversion as high as ~85% whereas 5Ni4Ba/YZr and 5Ni4Gd/YZr show CH₄ conversions of about ~80%. Even in 30 h TOS study, 5Ni4Ho/YZr catalyst showed >81% CH₄ conversion with retaining highest H₂/CO (0.97).

INTRODUCTION

Global warming gases (CH₄ and CO₂) are converted into syngas (H₂ and CO) over the appropriate catalyst at high-temperature conditions. The process has synthetic utility on the expense of depleting the concentration of global warming gases. The reaction is popularly known as dry reforming of methane (DRM). Methane is dissociated over Pt, Pd, and Ni active metals where Ni has minimal methane dissociation energy.¹ In a detailed DFT calculation, Niu et al. showed that the step/edge site of Ni particle has less energy barrier for C–H dissociation than a flat surface.² Further, the dissociated carbon species is oxidized into CO by CO₂. Additionally, the low cost of Ni has drawn huge interest from the DRM community. However, Ni undergoes serious sintering at high-temperature endothermic DRM reactions which leads to severe coke deposition and finally deactivation. In the DFT study, the growth process of subsequent carbon species into graphene or carbon nanotube-like structures was discussed by Zhang et al.³ SiO₂, Al₂O₃, and ZrO₂ supported Ni catalysts had enhanced metal-support interaction which resist the sintering at high-temperature DRM reaction.^4–6 Silica is just a neutral support providing a large surface area (from micropores to mesopores range). Alumina support is acidic and acid-basic property of alumina was also optimized by the addition of lanthana or MgO.^7–10 Addition of <5 wt.% Zr along with 6.14 wt.% Ni over MgAl₂O₄ support had gained coke-resistant property as there was no plugging. The CH₄ conversion was retained >50% up to 25 h TOS over this catalyst system.¹¹ The current research trend has shifted toward zirconia-supported catalysis due to oxygen-endowing capacity and subsequent vacancy formation in zirconia rather than silica and alumina. Among different ZrO₂ phases, the energy barrier of CO₂ dissociation on the defective surfaces of cubic-ZrO₂, tetragonal-ZrO₂, and monoclinic-ZrO₂ was found to reduce 1/2, 1/4, and 1/5 of the perfect surface.¹² A series of metal oxide-promoted “zirconia” or “metal oxide-zirconia” supported Ni-based catalysts are used for DRM (Figure 1).

Figure 1. A presentation of promoted “zirconia” or “metal oxide-zirconia” supported Ni catalysts tested for dry reforming of methane (DRM) reaction.

The addition of alkali and alkaline earth metals was expected to increase the basicity of the catalyst which induces the CO₂ interaction with the catalyst surface. Apart from basicity, it may also induce surface parameters such as Ni dispersion, metal-support interaction, NiO reducibility, coke resistance, and support stability. 0.6Wt. % Na promotional addition caused strong metal-support interaction due to the formation of NiO_xH_y species.¹³ Barium addition induces self-coking of carbon deposits by –OH and –O species and strong sinter resistance by NiO_x particles.¹⁴ It showed >65% conversion, >0.8 H₂/CO up to 33 h. 0.4Wt. %Mg presence over zirconia supported Ni catalyst which brought higher surface area, stabilized the t-ZrO₂ phase, and coke resistance.¹⁵ Using 2:1 ratio of Ni (0.04 mol) and Ca over 0.1 mol ZrO₂ enriched the surface parameters (surface area, pore volume and pore diameter), strengthened metal-support interaction, and induced a low degree of graphitization. It resulted into >70 conversion % 0.7 H₂/CO ratio and 3 × 10³ H₂ molar production in 350 min. Upon addition of a high amount of Mg (7Wt. %) over ZrO₂ supported Ni catalyst, a reducible dispersed NiO-MgO solid solution was formed which resulted into 85% CH₄ conversion and 92% CO₂ conversion 0.94 H₂/CO in 7 h TOS.¹⁶ Further, the promotional addition of 0.5Wt. % K over a MgO-ZrO₂ (5:2) supported Ni catalyst caused an increase in surface area and high Ni/Mg ratio. The catalyst retained about 90% CH₄ conversion at 750°C up to 14 h TOS.¹⁷

Promotional addition of p/d/f-block elements in ZrO₂ supported Ni catalysts was found to stabilize zirconia phases, optimize Ni crystallite size, induce the formation of adsorbed oxygen species, and tune the reducibility-basicity profile. It was reported that Si promotional addition in ZrO₂ support optimized Ni size up to 9–12 nm.^18,19 “Ni covered with porous silica” and then sequentially coated with a porous ZrO²⁻ shell had extra optimized Ni (6 nm). The catalyst caused higher adsorption energy and a lower dissociation barrier of CO₂ than the catalyst without ZrO₂ coating.¹⁹ So, it showed more about 60% CH₄ conversion, and a 0.7H₂/CO ratio up to 150 h at 700°C. 5Wt. %Al-5Wt. %Mn modified ZrO₂ supported Ni catalyst had small Ni crystallite, dense basic sites and adsorbed oxygen species. This resulted in >85% conversion up to 5 h TOS.²⁰ 15Wt. % Ni incorporated “80%CeO₂ and 20% ZrO₂” catalyst gave more than 97% CH₄ at 800°C up to 100 h.²¹ 5Wt. % Pr, Y, and La promoted ceria-zirconia supported Ni catalysts had similar basicity and similar size of metallic Ni (after reaction) which gave very similar catalytic activity (~90%) with >0.9 H₂/CO ratio.²² An equal proportion of Mn and Ni over ceria-zirconia (50:50) support was found to facilitate CH₄ decomposition and CO₂ activation through –CH₂OH species at low temperatures.²³ It gave a 27% conversion with a 0.75 H₂/CO ratio at 550°C. All ceria-zirconia catalyst systems were found excellent, but the high weight percentage of Ni is needed to decrease from an economic point of view.

10 mol%Pr₂O₃-zirconia, 10 mol%Nd₂O₃-zirconia, 10 mol%La₂O₃-zirconia, 10 mol%ceria-zirconia, and 10 mol%Y₂O₃-zirconia supported systems having 1Wt. % Ni dispersion was applied for DRM under an electric field (up to 5 W) and external heat of 150°C.²⁴ The reaction was presumed to initiate the activation of CH₄ and formate species. 10 mol%La₂O₃-zirconia supported Ni catalyst showed 33% CH₄ conversion, 43% CO₂ conversion and 0.7 H₂/CO ratio without coke decomposition. Lanthana-zirconia, Samaria-ZrO₂ supported, and yttria-zirconia supported systems were able to optimize Ni particle size to 13.5 nm and to attain conversion of more than >50% with a high H₂/CO (>0.8) ratio.²⁵ La₂O₃ promoter over ZrO₂ supported Ni catalyst had increased mesoporosity, increased CO₂ interaction (by La⁺³ and CO₂ coordination), and increased metal-support interaction (than unsupported one).^26,27 3Wt. % La addition over ZrO₂ supported Ni catalyst resulted in 70% conversions and >0.9 H₂/CO at 700°C in 50 h TOS. Ca-promoted lanthana-zirconia supported Ni catalyst brought additional basicity, more CO₂ interaction, efficient carbon deposit oxidation, and minimum carbon deposition.²⁸ So, it resulted into ~78% CH₄ conversion, 83% CO₂ conversion and a 0.9 H₂/CO ratio. Ce promotional addition over this support induced prominent CH₄ decomposition and enhanced mobility of lattice oxygen which resulted into >80% CH₄ conversion.^29,30 Ce promotional addition over “tungsten-zirconia supported Ni catalyst” had the additional redox capability and basic sites.^31,32 It showed conversion close to 80%. Cr or Gd-promoted catalyst systems had an excellent regeneration capability of reducible NiO whereas Ga-promoted catalyst had increased metal-support interaction. Cr-promoted catalyst had a low energy gap and enhanced reducibility whereas Gd-promoted catalyst had a stable zirconium-gadolinium phase and strong 3d-4f electron exchange and spin-orbit interaction between Gd and Ni.^28,33,34 Cr, Gd, and Ga are quite competitive and shows >80% conversion and >0.9 H₂/CO ratios. Yttria addition brought additional surface adsorbed oxygen species causing coke gasification. It showed 50-60% conversion and ~64% Hydrogen yield with >0.8 H₂/CO ratio. The addition of yttria in zirconia support prevented the phase transition of zirconia at high reaction temperature.^35–37 Cs and Sr promoters were applied to a 15Wt.%Y85Wt%ZrO₂ supported Ni system where earlier one showed 56% H₂ yield (and 2.6% weight loss) whereas later one showed 63.4% H₂ yield (and 9.7% weight loss) against 55.3% H₂-yield over a nonpromoted catalyst (and 25.6% weight loss).³⁸

The yttria-zirconia-supported Ni catalysts had all the essential qualities like temperature sustainability, lattice oxygen endowing capability, Ni size optimization, and adequate basic sites in favor of DRM. The promotional addition of metal oxide over yttria-zirconia supported Ni catalyst may further enhance the activity. Herein, we have prepared metal oxides (like Ho, Ba, Ga, Gd, and Cs) promoted yttria-zirconia-supported Ni catalysts. These catalysts were investigated for DRM reaction. Surface area and porosity, X-ray diffraction, RAMAN, UV-vis spectroscopy, thermogravimetry, temperature programmed study and transmission electron microscopy were used to characterize these catalysts properly. The fine-tuning characterization results and catalytic activities would set a concluding remark for DRM over metal oxide-promoted “yttria-zirconia supported Ni catalysts.”

EXPERIMENTAL Material

Ni(NO₃)₂.6H₂O (98%, Alfa Aesar), zirconia (gifted by Kagaku Daiichi Kogyo Co. Ltd), yttria (obtained from China), CsNO₃, Ga(NO₃)₃._XH₂O, Ba(NO₃)₂, Ho(NO₃)₃.5H₂O, Gd(NO₃)₂.6H₂O and deionized water.

Catalyst preparation

Yttria-zirconia support is prepared by mechanically mixing the respective metal oxides (8 wt. % yttrium oxide and 92 wt. % zirconium oxide). Nickel nitrate aqueous solution (equivalent to 5 wt% Ni) and 4 wt.% promoter-based metal nitrate aqueous solution (Metal = Cs, Ga, Ba, Ho, Gd) are added to the Yttria-zirconia support by impregnation method. The prepared mixture is stirred at 80°C, and dried at 120°C overnight in an oven. Subsequently, the dried product is calcined at 600°C for 3 h. Zirconia-supported nickel catalyst, yttria-zirconia-supported Ni catalyst, metal oxide-promoted yttria-zirconia supported nickel catalysts are abbreviated as 5Ni/Zr, 5Ni/YZr and 5Ni4M/YZr (M = Cs, Ga, Ba, Ho, Gd) respectively.

Catalyst characterization

Detailed information on characterization technique, equipment, specification and methods are mentioned in Table 1.

Table 1 Detailed information on characterization technique, equipment, specification and methods.

Techniques	Equipment	Methods/specification	Other information
Surface area and porosity	Micromeritics Tristar II 3020 using dewar and liquid nitrogen	Brunauer–Emmett–Teller (BET) method for surface area and pore size Barrett–Joyner–Halenda (BJH) method for pore size distribution	The N2 adsorption-desorption isotherm is plotted up to 1 relative pressure (P/Po)
X-ray diffraction (XRD)	Rigaku Ultima 4 diffractometer operated at 40 kV and 40 mA.	Radiation source: CuKα (λ = 1.54056 Å) Step size 0.01 Scanning range 5–100	Phases were documented using a standard JCPDS database
Raman analysis	Laser Raman (NMR-4500) Spectrometer (JASCO, Japan) equipped with Spectra Manager Ver.2	Beam wavelength for excitation: 532 nm Range: 150–1250 cm−1	The Laser intensity was adjusted to 1.6 mW for 10 s of exposure time at three accumulations.
Ultraviolet-visible spectroscopy	V-570, JASCO, USA	Resolution: 1 nm Range: 200–1000 nm Scanning speed 200 nm/min	-
H2-temperature programmed reduction (TPR)	Micromeritics Auto Chem II 2920, USA equipped with a thermal conductivity detector	10% H2-Ar flow rate = 40 mL/min Heating ramp: 10°C/min	70 mg Sample was heated up to 1000°C under 10% H2-Ar mixture
CO2-temperature programmed desorption (TPD)	Micromeritics Auto Chem II 2920, USA equipped with a thermal conductivity detector	10% CO2/He flow rate: 30 mL/min Heating ramp: 10°C/min	70 mg Sample was fed by 10% CO2/He at 50°C for 30 min. CO2 desorption is monitored upon raising the temperature up to 1000°C.
Thermal gravimetric analysis (TGA)	Shimadzu TGA-51	Sample weight: 10 mg Heating ramp: 10°C/min Heating Ranges up to 1000°C	-

Catalyst activity test

0.1 g 5Ni4M/YZr catalyst is packed in a tubular stainless-steel reactor (Diameter = 9.1 mm and height = 300 mm) supplied by PID Eng. and Tech Micro activity Reference Company. To monitor the temperature of the catalyst bed, a K-type thermocouple is injected axially at the catalyst bed. First, the reductive pretreatment is carried out over a catalyst under H₂ (flow rate 30 mL/min) for 1 h at 700°C. Further, the reactor is flushed with N₂ for 15 min to remove the physiosorbed H₂. DRM reaction is performed at 800°C by passing CO₂, CH₄, and N₂ at 30, 30, and 10 mL/min (to generate 42,000 mL/h·g_cat of gas hourly space velocity) respectively. The effluent is examined by an online GC equipped with molecular sieve 5 A and Porapak Q columns and a TCD detector using Ar carrier gas. CH₄ conversions, CO₂ conversions and H₂/CO ratio are calculated by the following expressions: [Image Omitted. See PDF] [Image Omitted. See PDF] [Image Omitted. See PDF]

RESULTS Catalytic activity results

The catalytic activity of 5Ni/Zr, 5Ni/YZr, and 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, Ga) catalysts in terms of CH₄ conversions, CO₂ conversions and H₂/CO ratio at 800°C reaction temperature is shown in Figure 2. The error bars of catalytic activity results are also shown in Supporting Information: Figure S1. It was noted that the error of catalytic activity is 3%–4% for all the experiments. Toward DRM, zirconia supported Ni system (5Ni/Zr) is found efficient to carry out more than 50% CH₄ conversion and ~61% CO₂ conversion whereas yttria-zirconia supported Ni catalyst system (5Ni/YZr) is capable of improving the CH₄ conversion up to >70% and CO₂ conversion up to >80% respectively. The catalytic activity remains constant over 5Ni/YZr catalyst over the tested time (420 min). Upon the addition of 4 Wt.% Ga over yttria-zirconia supported Ni catalyst, the catalytic activity (CH₄ conversion) drops down to 56.8%. 4 wt% Cs incorporated catalyst (5Ni4Cs/YZr) has least catalytic activity (53.7% CH₄ conversion) among the rest promoted catalyst system. In the alkaline earth metal group, Ba location is next to caesium. On adding 4Wt. % Ba over yttria-zirconia supported Ni catalyst (5Ni4Ba/YZr), the CH₄ and CO₂ conversions is raised to 80% and ~87%, respectively up to 420 min time on stream. 5Ni4Gd/YZr is quite competent with 5Ni4Ba/YZr in the mean of catalytic activity during 420 min of test time. 5Ni4Ho/YZr is found high performed catalyst. They outperformed with ~85% CH₄ conversion and ~91% CO₂ conversion during the testing time of 420 min. The long-time catalyst study (30 h) over 5Ni4Ho/YZr and 5Ni4Cs/YZr is also carried out (Supporting Information: Figure S2). Over 5Ni4Ho/YZr catalyst, the CH₄ conversion and CO₂ conversion remain above 81% and 86% respectively up to 30 h time-on-stream study. The catalytic activity remains constant (53% CH₄ conversion) over 5Ni4Cs/YZr catalyst up to 30 h time on stream.

Figure 2. Catalytic activity versus time on stream study of different catalysts at 800°C reaction temperature (A) CH4 versus time on stream (B) CO2 conversions versus time on stream (C) H2/CO ratio versus time on stream.

As per the DRM reaction (CO₂ + CH₄ → 2H₂ + 2CO), the H₂/CO ratio should be equal to 1. However, over each catalyst, H₂/CO ratio is found less than one. It indicates the presence of H₂ consuming reaction like reverse water gas shift reaction (CO₂ + H₂ → CO + H₂O).³¹ At the end of 420 min on stream, the H₂/CO ratio over different catalysts is found in the following order; 5Ni4Ho/YZr (H₂/CO = 0.97) > 5Ni4Ba/YZr (H₂/CO = 0.94) > 5Ni4Gd/YZr (H₂/CO = 0.90) > 5Ni/YZr (H₂/CO = 0.89) > 5Ni4Ga/YZr (H₂/CO = 0.78) ~ 5Ni4Cs/YZr > 5Ni/YZr (H₂/CO = 0.71). H₂/CO ratio closer to 1 indicates a strong inhibition of RWGS reaction. In a long-time study of up to 30 h, the H₂/CO ratio of 5Ni4Ho/YZr and 5Ni4Cs/YZr catalyst remained constant (as like as 7 h TOS).

Characterization results

Surface area and porosity results show that 5Ni/Zr, 5Ni/YZr, and 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, Ga) catalysts are mesopores and bimodal (Figure 3 and Supporting Information: Figure S3). Yttria-zirconia-supported Ni has half of the surface area than zirconia-supported Ni catalyst. The promotional addition of Ga, Gd, Cs, and Ba does not bring about major changes in the surface area and porosity patterns. However, the average diameter of different catalysts is variable. The average pore size of catalyst is found in following order; 5Ni 4Ga/YZr (20 nm) < 5Ni 4Gd/YZr (22.3 nm) < 5Ni4Ho/YZr (23.6 nm) < 5Ni 4Ba/YZr (24.3 nm) < 5Ni 4Cs/YZr (28.4 nm).

Figure 3. N2-adsorption isotherm and porosity distribution profile of (A) 5Ni/Zr and 5Ni/YZr (B) 5Ni4M/YZr (M = Gd, Ho) catalyst.

The X-ray diffraction pattern of the 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, Ga) catalysts are shown in Figure 4. Patel et al. previously showed that the addition of yttria to zirconia support brought a phase transition from unstable monoclinic zirconium oxide to stable tetragonal zirconium yttrium oxide.³⁶ In all, yttria-zirconia supported Ni catalysts (promoted and unpromoted), stable tetragonal zirconium yttrium oxide phases at 2θ = 30.10°, 34.90°, 50.1°, 59.56°, 62.16° (JCPDS reference number 01-082-1242 and 01-082-1243) are found frequently. However, a 4Wt. % Gd or Cs or Ga promoted yttria-zirconia supported Ni catalyst has diffuse peaks for unstable monoclinic zirconium oxide also at Brag angle 2θ = 28.18°, 31.4°, 34.9° and 62.6° (JCPDS reference number 01-078-0047). The 5Ni4Gd/YZr sample has additional mixed oxide phases as cubic zirconium gadolinium oxide at Brag angles 2θ = 28.2°, 62.6°, and 74.11° (JCPSD reference number 01-080-0469). In the same way, the mixed oxide phase as cubic zirconium barium oxide is also detected in 5Ni4Ba/YZr catalyst at Brag angle 2θ = 30.12°, 43.18°, 53.32° and 62.7° (JCPDS reference number 00-006-0399). 4wt. % Ba and 4wt.% Ho promoted YZr supported catalyst systems has less crystallinity than 4wt. % Gd, 4Wt. % Cs, and 4wt.% Ga promoted catalyst system. Most of the diffraction peaks of 5Ni4Ba/YZr and 5Ni4Ho/YZr catalysts are found in the same 2θ range (Figure 4F–G). Cubic zirconium holmium oxide phases are merged with the cubic zirconia oxide phases (JCPDS reference number 01-078-1306). Ga or Cs promoted YZr supported Ni catalyst shows crystalline cubic NiO diffraction peak at Brag angle 2θ = 37.28° and 43.29° and 62.91° (JCPDS reference number 00-004-0835). Higher crystalline peaks of NiO in 5Ni4Ga/YZr and 5Ni4Cs/YZr catalysts indicate a lower dispersion of NiO than the rest catalyst system.

Figure 4. X-ray diffraction (XRD) of (A–E) 5Ni4M/YZr (M = Ba, Ho, Gd, Cs, Ga) catalyst (F–G) XRD of 5Ni4Ba/YZr and 5Ni4Ho/YZr catalyst.

The Raman spectra of 5Ni/Zr, 5Ni/YZr and 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, Ga) catalysts are shown in Figure 5. Raman spectra also concludes the information of ZrO₂ phases. Zirconia supported Nickel catalyst system shows intense peaks for the monoclinic ZrO₂ phase at 178, 334, 380, 476, and 610 cm⁻¹ where the peak intensity of 476 cm⁻¹ is found higher than 610 cm⁻¹.^39,40 It is reported that the Raman bands 476 and 610 cm⁻¹ are present in both monoclinic-ZrO₂ phase and tetragonal-ZrO₂ phase but the relative intensity of these two bands are different for the two phases.⁴⁰ For monoclinic phase, the intensities of earlier bands are much stronger than the later band. On incorporation of 8 wt% Yttria along with ZrO₂ support, the 5Ni/YZr catalyst has diffused/depleted peaks for monoclinic ZrO₂ and new peaks for cubic ZrO₂ at 258 cm⁻¹.⁴¹ Interestingly, Gallium-promoted catalyst (5Ni4Ga/YZr) has the highest intensity of the cubic ZrO₂ phase.

Figure 5. Raman spectra of 5Ni/Zr, 5Ni/YZr and 5Ni4M/YZr (M = Gd, Ba, Cs, Ga) catalyst in 0–800 cm−1 wavenumber range.

The UV-vis spectra of the catalyst samples show a charge transfer band and d-d transition band (Figure 6). It is interesting to note that on the introduction of yttria along with zirconia as support, the intensity of charge transfer bands at 230 nm for O²⁻ → Zr⁴⁺,³⁰ 257 nm for O²⁻ → Ni²⁺⁴ and 290 nm for O²⁻ → Ni²⁺/Zr⁴⁺) are suppressed largely (Figure 6A) as well as the bandgap between the valance band and conduction band is decreased to 2.23 eV (against 3.15 eV in 5Ni/Zr) (Supporting Information: Figure S4). On promotional addition of Ba, Gd and Ho, bandgap has increased and remains about 3.13 to 3.17 eV. In the case of the promotional addition of 4 wt% Ba, again the charge transfer band for O²⁻ → Ni²⁺/Zr⁴⁺ is restored. In the previous study, it was also found that with the increase of Ba²⁺ concentration, the intensity of the bands in the charge transfer region also increased⁴² due to charge transfer between Ba²⁺–O²⁻. So, it can be expected that due to the combined charge transfer of O²⁻ → Ni²⁺/Zr⁴⁺/Ba²⁺, the charge transfer band at 295 nm is restored in 5Ni4Ba/YZr. That means the incorporation of Ba⁺² does not inhibit the charge transfer of O²⁻ than Ho and Gd. Cs-promoted yttria-zirconia Ni catalyst has additional UV peaks in the charge transfer regions (Figure 6B).

Figure 6. (A) UV-vis spectra of 5Ni/Zr, 5Ni/YZr and 5Ni4M/YZr (M = Ho, Gd, Ba) (B) UV-vis spectra of 5Ni4M/YZr (M = Cs, Ga).

The H₂-TPR profile of 5Ni/Zr, 5Ni/YZr, and 5Ni4M/YZr (M = Ho, Gd, Ba,) catalyst is mostly centered about 425°C (Figure 7A). The reduction peak below 425°C is attributed to the reduction of “weakly interacted NiO-species” whereas the reduction peak > 425°C is attributed to the reduction of “moderately interacted NiO-species.”⁴³ In zirconia-supported Ni catalyst, the intensity of the reduction peak for “moderately interacted NiO species” is much higher than for “weakly interacted NiO species” (Figure 7A). On incorporating yttria in support along with zirconia, the intensity of reduction peak for “moderately interacted NiO species” is greatly suppressed and the 5Ni/YZr catalyst has a comparable amount of both types of reducible NiO species (“weakly interacted NiO species” and “moderately interacted NiO species”). Upon incorporation of 4Wt. %Ho promoter as well as 4Wt. % Ba promoter, a relatively lower temperature reduction peak is marginally larger than a higher temperature reduction peak. 4Wt. % Gd promotional addition in yttria-zirconia supported Ni catalyst, lower temperature reduction peak became more prominent than the higher temperature reduction peaks. It indicates that “weakly interacted reducible NiO species” are grown relatively more than “moderately interacted reducible NiO species” over 5Ni4Gd/YZr, 5Ni4Ba/YZr and 5Ni4Ho/YZr catalysts. In the case of 4Wt. % Ga incorporation over yttria-zirconia supported Ni catalyst was unique (Supporting Information: Figure S5A). The reduction peak profile is shifted to a higher temperature such that the lower temperature peak completely vanishes and a new peak centered about 600°C appears which is attributed to the reduction of “strongly interacted NiO-species” over the catalyst surface. 4Wt. % Cs promoted yttria zirconia catalyst had merged reduction peaks for Cs₂O (about 354°C and 374°C) along with reduction peaks for interacted NiO species⁴⁴ (Supporting Information: Figure S5A). It is also noticeable that, upon Cs addition, the reduction peak position for “interacted NiO-species” was extended to a relatively higher temperature (than Gd, Ba, Ho promoted catalyst) indicating enhanced metal-support interaction upon Cs addition.

Figure 7. (A) H2-TPR of 5Ni/Zr, 5Ni/YZr, 4Ni4M/YZr (M = Ba, Gd. Ho) catalysts (B) CO2-TPD of 5Ni/Zr, 5Ni/YZr, 4Ni4M/YZr (M = Ba, Gd. Ho, Cs, Ga) catalyst.

The CO₂-TPD profiles of 5Ni/Zr, 5Ni/YZr, and 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, Ga) catalysts can be broadly classified into three regions (Figure 7B). The peaks below 175°C for weak basic site regions (attributed to basicity due to surface hydroxyl), from 175°C to 450°C for intermediate strength basic site regions (attributed to basicity due to isolated O^2- species) and above 450°C for strong basic sites (attributed to basicity due to unidentate carbonates)^33,45 Most of the catalysts had CO₂-TPD peaks in the low and intermediate temperature regions claiming weak and intermediate strength basic sites. Zirconia-supported Ni catalyst has the highest density of weak basic sites than the rest catalyst system. 4Wt. % Gd promoted yttria-zirconia supported Ni catalyst is characterized by the highest density of intermediate strength basic sites among all catalysts. Cs promoted as well as Ba promoted yttria-zirconia supported Ni catalyst had strong basic sites. The thermogravimetric profile of 5Ni/Zr, 5Ni/YZr, and 5Ni4M/YZr (M = Ho, Gd, Ba, Cs, and Ga) catalyst is shown in Supporting Information: Figure S5B. Zirconia supported Ni and yttria zirconia supported Ni catalysts had 18%–19% weight loss. On incorporation of promoters like Ho into Yttria-supported Ni catalyst, 14%–15% weight loss is observed. Further, 5Ni 4Ga/YZr, 5Ni 4Ba/YZr, and 5N4Gd/YZr showed 7.8%, 6.8%, and 5.8% weight loss. Cs promoted yttria-zirconia supported Ni catalyst has no weight loss.

The TEM image fresh and spent 5Ni/Zr, 5Ni/YZr and 5Ni4M/YZr (M = Cs, Ga, Ba, Gd, Ho) is shown in Figure 8. The carbon deposit in the form of nanotubes is found in all spent catalyst systems. The particle size distribution plots and particle size in fresh and spent catalyst samples are calculated by TEM image on a 200 nm scale (Supporting Information: Figures S6 and S7, Table S1). The particle size of Ni species is found smallest in the case of Ho-promoted catalyst. It is interesting to note that thermogravimetry analysis of 5Ni4Cs/YZr shows no weight loss but the HRTEM image of the spent-5Ni4Cs/YZr catalyst (Figure 7D) evidence the presence of carbon nanotube. In the thermogravimetry analysis, analysis is performed under oxygen which may oxidize the carbon nanotube as well as the metal oxide. It is reported that under a particular oxygen environment, various types of caesium-related oxides like caesium oxide, caesium peroxide, caesium suboxide, caesium superoxide as well as 80 different types of Caesium oxide clusters are nurturing.^46,47 So, it can be concluded that net zero weight loss in TGA of 5Ni4Cs/YZr does not mean that there is no carbon deposit or there is no oxidation of coke over catalyst but it may be due to equivalent weight gain by caesium compounds under oxygen.

Figure 8. High Resolution Transmission Electron Microscope image of different fresh and spent catalyst.

DISCUSSION

The DRM reaction is known as the potential conversion of equimolar CH₄ and CO₂ into syngas. But here, over each catalyst, CO₂ conversion was found to be higher than CH₄ conversion. This indicates that additional CO₂ is engaged in some other reactions than DRM. Metallic Ni is also known for H₂ dissociation.⁴⁸ Thereby, supported Ni catalyst may also felicitate the reverse water gas shift reaction (CO₂ + H₂ → CO + H₂O) as a major side reaction during DRM⁴ which caused higher CO₂ conversion than CH₄ conversion. The %weight loss over different catalysts during DRM reaction is found in the following order 5Ni4Cs/YZr (0% weight loss) < 5Ni4Gd/YZr (5.8% weight loss) < 5Ni4Ba/YZr (6.8% weight loss) < 5Ni4Ga/YZr (7.8% weight loss) < 5Ni4Ho/YZr (14%–15% weight loss) <5Ni/YZr (18%–19% weight loss) ~5Ni/Zr. However, DRM activity results over promoted catalysts are found in the given order 5Ni4Ho/YZr (~85% CH₄ conversion) > 5Ni4Ba/YZr (~80% CH₄ conversion) ~5Ni4Gd/YZr (~80% CH₄ conversion) > 5Ni/YZr (70% CH₄ conversion) > 5Ni4Ga/YZr (56.8% CH₄ conversion) > 5Ni4Cs/YZr (53.7% CH₄ conversion). Even at high coke deposition, the catalyst active site of 5Ni4Ho/YZr remains exposed and the catalysts perform excellently. It seems that coke deposition is not a crucial factor in affecting activity. It is possible if the rate of carbon formation (at the catalytic active sites) properly matches the rate of carbon diffusion (away from the catalytic active sites).⁴⁹ Other factors like support stability, reducibility, and basicity also need to be considered in correlating the catalytic activity.

“Bare zirconia” supported Nickel has a substantial amount of basic sites (from weak to moderate strength), a higher intensity of “moderately interacted NiO-species” than “weakly interacted NiO-species” and 3.15 eV bandgap (between valance band and conduction band). However, unstable monoclinic ZrO₂ phases against high-temperature DRM reaction limit its activity (CH₄ conversion) below 50% in 420 min time-on-stream. On incorporation of 8Wt.% Y₂O₃ along with ZrO₂, a mixed oxide support named “tetragonal zirconium yttrium oxide” is formed which can withstand the high-temperature DRM reaction. The intensity of the charge transfer band from O^2-→Mⁿ⁺ (Mⁿ⁺ = Zr⁴⁺, Ni²⁺) is suppressed and the bandgap between the valance band and conduction band is decreased to 2.23 eV. However, the intensity of “moderately interacted NiO-species” is decreased, the concentration of all types of basic sites are also decreased and coke deposition doesn't decrease. This indicates that despite high coke deposition, the catalytic active sites (interacted Ni species) over stable support remain exposed for the DRM reaction and performed constantly. Overall, it can be said that upon the incorporation of 8 wt% Y₂O₃ along with zirconia, the 5Ni/YZr catalyst has mixed oxide (zirconium yttrium oxide) thermally stable phases and decreased bandgap (between valance band and conduction band) which results into higher and more constant catalytic performance (>70% CH₄ conversion and >80% CO₂ conversion) than 5Ni/Zr. The search for optimum catalyst performance is carried out by the use of different promoters especially Ba, Gd and Ho. Ho, Gd, and Ba promoted yttria-zirconia supported Ni catalysts have additional mixed oxide phases along with a zirconium yttrium oxide phase. 5Ni4Ho/YZr, 5Ni4Gd/YZr, and 5Ni4Ba/YZr catalysts have cubic holmium zirconium oxide phase, cubic zirconium gadolinium oxide phase and cubic zirconium barium oxide phase respectively. The bandgap of these catalysts also lies between 3.13 and 3.17 eV. The reducibility profile of the top three catalysts (5Ni4Ho/YZr, 5Ni4Gd/YZr, and 5Ni4Ba/YZr) is not found much different. All three catalysts have a relatively lower concentration of “moderately interacted NiO-species” than “weakly interacted NiO-species.” In terms of basic sites distribution, 5Ni4Ho/YZr has the presence of weak basic sites and intermediate strength basic sites, 5Ni4Gd/YZr acquires excessive intermediate strength basic sites and 5Ni4Ba/YZr may be characterized by the additional presence of strong basic sites. The basic profiles of 5Ni4Gd/YZr and 5Ni4Ba/YZr are quite different but the catalytic performance is very similar. This indicates that optimum catalytic activity requires an adequate amount of basic sites and a proper amount of stable metallic Ni sites. Excess amounts for one type of basic site are not needed to drive higher activity. Overall, it can be said that all three catalysts (5Ni4Ho/YZr, 5Ni4Gd/YZr, and 5Ni4Ba/YZr) had comparable reducibility profiles and adequate amounts of basic sites over the surface. The presence of different supports that stabilize the metallic Ni had key roles in deciding activity. 5Ni4Ho/YZr had stable tetragonal zirconium yttrium oxide as well as cubic zirconium holmium oxide phases to stabilize catalytic active Ni. The particle size over 5Ni4Ho/YZr catalyst is minimum (3.67 nm) than the rest promoted catalyst and the size remains limited to a minimum (4.68 nm) during the entire DRM reaction. Previously, it was observed that Ni particle size < 9 nm undergoes stronger metal-support interaction which restricts the thermal sintering of nickel.⁵⁰ 5Ni4Ho/YZr performs excellent toward DRM with ~85% CH₄ conversion at the end of 420 time on stream and more than 81% CH₄ conversion up to 30 h time-on-stream. As H₂/CO ratio over 5Ni4Ho/YZr catalyst is maximum (0.97) and so it has the strongest inhibition for reverse water gas shift reaction than the rest catalysts. 5Ni4M/YZr (M = Gd, Ba) has both stable tetragonal zirconium yttrium oxide and Cubic zirconium-gadolinium oxide or Cubic-zirconium barium oxide phase to stabilize catalytically active Ni. Both 5Ni4Gd/YZr and 5Ni4Ba/YZr are about equally efficient with 80% CH₄ conversion. As H₂/CO ratio of 5Ni4Ba/YZr is higher than 5Ni4Gd/YZr and 5Ni4Ba/YZr catalyst has stronger inhibition for RWGS reaction 5Ni4Gd/YZr catalyst.

4 wt% Ga or 4 wt% Cs addition over yttria-zirconia supported Ni catalyst is not found effective in terms of catalytic activity toward DRM and catalytic activity of 5Ni4Ga/YZr and 5Ni4Cs/YZr are less than even nonpromoted catalyst (5Ni/YZr). XRD results show less dispersion of NiO over the catalyst surface. Clearly, less-dispersed-NiO generates fewer active sites over the catalyst surface upon reduction resulting in inferior activity. The comparative catalytic activity (in terms of CH₄ conversion and CO₂ conversion) of the metal-promoted “metal oxide-zirconia supported Ni” based catalyst system is shown in Figure 9. The catalytic activity results with detailed reaction conditions are also mentioned in Supporting Information: Table S2. Metal-promoted “yttria-zirconia supported Ni” is found superior to metal-promoted “tungsten-zirconia supported Ni” catalyst system and quite competitive to metal-promoted “lanthana-zirconia supported Ni” catalyst system^{25,26,28,46,51,52} (Figure 9). The promotional effect over 5Ni/YZr catalyst was studied extensively studied where Ho-promoted catalyst were found be the best whereas in rest case (5Ni/LaZr, 5Ni/WZr) ceria promoted catalyst showed excellency. Alumina and silica support are also thermally stable. The promotional effect of different metal oxides over alumina-supported Ni was investigated intensively for DRM.^4,53–59 Over Ni/Al₂O₃, 79-81% CH₄ conversion was quite noticeable over W (in 7.3 h TOS), Co (in 24-h TOS) and Yb (in 20 TOS) promotors. Various groups have investigated Metal oxide promoted “silica supported Ni” and “ordered mesoporous silica (MCM-41, SBA-15)-supported Ni” catalysts for the DRM reaction.^60–67 However, preparation of ordered mesoporous silica again needs skill and is more costly. So, these catalysts gain less consideration. Over Ni/SiO₂ catalyst, promotors like Lathana were quite impressive. It showed 80% CH₄ conversion up to 24 h. In the current study, 5Ni4Ho/YZr catalyst is prepared by simple mechanical mixing and it outperforms others. It shows 85% CH₄ conversion in 7 h TOS and 81% CH₄ conversion in 30 h TOS with 0.97 H₂/CO.

Figure 9. The CH4 conversion (CCH4 ${{\rm{C}}}_{{\mathrm{CH}}_{4}}$) and CO2 conversion (CCO2 ${{\rm{C}}}_{{\mathrm{CO}}_{2}}$) of over different nonpromoted and metal-promoted “metal oxide-zirconia supported Ni” catalysts. In the case of the promoted catalyst, the atomic symbol of the promoted metal is shown in parentheses just below the activity value.

CONCLUSION

The catalytic activity of ZrO₂-supported Ni is limited by an unstable ZrO₂ phase under drastic DRM conditions. Upon incorporation of Y₂O₃ along with ZrO₂ support, a thermostable support of zirconium yttrium oxide is formed and catalytic activity shoots to 70% CH₄ conversion. Along with adequate basic sites, 5Ni4Ho/YZr has zirconium yttrium oxide phase along with cubic zirconium holmium oxide phase to stabilize the optimum size of Ni (3.67 nm) which is minimum among other promoted catalysts. 5Ni4Ho/YZr catalyst had highest catalytic activity with strongest inhibition for RWGS. It shows about 85% CH₄ conversion in 7 h time-on-stream and >81% CH₄ conversion in 30 h time-on-stream with 0.97 H₂/CO. Including an adequate amount of basic sites, the 5Ni4Gd/YZr and 5Ni4Ba/YZr catalysts have promoter-incorporated support phases (as cubic zirconium gadolinium oxide and cubic zirconium barium oxide respectively) to stabilize Ni. Both catalysts are equally efficient ( ~ 80% CH₄ conversion) toward DRM. However, in terms of RWGS inhibition, 5Ni4Ba/YZr catalysts are found to be better than 5Ni4Gd/YZr catalysts. The low dispersion of catalytic active Ni sites over 5Ni4Cs/YZr and 5Ni4Ga/YZr results in inferior catalytic activity than other catalyst systems.

ACKNOWLEDGMENTS

The authors would like to extend their sincere appreciation to Researchers Supporting Project number (RSP2023R368), King Saud University. Rawesh Kumar and Kenit Acharya acknowledge Indus University for supporting research.