Role of Ca, Cr, Ga and Gd promotor over lanthana‐zirconia–supported Ni catalyst towards H2‐rich syngas production through dry reforming of methane

Abstract

Hydrogen production through methane dry reforming holds the promise of lowering greenhouse gases, that is CO₂ and CH₄, concentrations. Herein, Ca‐, Cr‐, Ga‐ and Gd‐promoted lanthana‐zirconia–supported Ni catalysts are investigated and characterized by X‐ray diffraction, Raman, infrared and UV‐vis spectroscopy, CH₄‐temperature programmed surface reaction and cyclic reduction‐desorption experiment. All promoted catalyst systems had high and constant hydrogen yield (>70%) due to pronounced reoxidation capacity of reduced ‘NiO species strongly interacted with support’ through oxygen replenishment by CO₂. The presence of mixed oxide and regeneration of reduced catalyst up to optimum level through oxygen replenishment by CO₂ in Gd, as well as Cr‐promoted catalyst, outperformed (80% initially) than other promotors, Ca and Ga. In the long run (440 min to 33 h), Cr‐promoted catalyst system performed better than Gd‐promoted catalyst system as H₂ yield remained constant ~79% due to the smallest energy gap between valance and conduction band, Ni‐Cr interaction species for wide range CH₄ decomposition and chromate species for profound carbon deposit oxidation.

INTRODUCTION

Hydrogen is a clean, promising and non-polluting source of energy. It has three times the energy (143 MJ kg⁻¹) of gasoline (46.4 MJ kg⁻¹) and emits no greenhouse gases when burned with oxygen z.¹ It is produced from a variety of non-renewables and renewables feedstocks, including coal, gasoline, natural gas, biomass and water electrolysis. Among these are hydrogen production from natural gas via dry reforming of methane (DRM), steam reforming methane (SRM) and partial oxidation of methane (POM). Hydrogen production from DRM is a great scientific step that holds out the prospect of green energy in the future by reducing greenhouse gas emissions (CH₄ & CO₂). CH₄ is typically dissociated over noble and transition metals such as Pt, Ru, Rh and Ni metal. Despite serious agglomeration of the catalyst at high temperatures, cheap Ni-based catalyst has drawn much attention in DRM reaction at endothermic conditions. Albeit the dispersion of Ni over thermally stable supports such as Al₂O₃, ZrO₂ and SiO₂ and strong metal-support interaction find the way against Ni agglomeration. The DRM reaction scheme for H₂ production is summarized in Figure 1 as the dissociation of CH₄ into CH_y and adsorbed (4-y) H over Ni, dissociation of CO₂ into CO and oxygen atom, oxidation of CH_y with oxygen atom into CO and yH and finally association of adsorbed hydrogen into H₂ (Figure 1). Overall, the dry reforming reaction (Figure 1, equation 1) is highly endothermic and operated at high-temperature conditions (600–800°C). Along with DRM reaction (CH₄ + CO₂ → 2H₂ + 2CO), thermal cracking of methane (CH₄ → C + 2H₂) and CO₂ gasification of carbon reaction (C + CO₂ → 2CO) also become feasible in temperature range 650–700°C. However, H₂ consuming reactions as gasification of carbon deposit by H₂ (Figure 1, equation 2) and reverse water gas shift (RWGS) reaction (Figure 1, equation 3) also run parallel with DRM, which seriously affect the H₂ yield.

FIGURE 1. Reaction scheme over Ni-based catalyst system

Among the various supports, the partial oxidation of carbon deposit by lattice oxygen of ZrO₂ is indeed an additional advantage over Al₂O_3, and SiO₂ supports in DRM.^2,3 Therdthianwong et al. also showed suppression of coke deposition up to 22% without affecting H₂ yield (60%) over 10 wt.% ZrO₂-promoted Al₂O₃-supported Ni catalyst (against 30% carbon deposition in unpromoted catalyst).⁴ Further, the trends for using zirconia as support are started by the catalyst community. 8 wt% Ni impregnated over commercially available pelletized zirconia catalyst (8Ni/ZrO₂) was utilized for CH₄:20CO₂:50Ar gas feed (WHSV = 120 L g⁻¹ h⁻¹) conversion. It showed 22% CH₄ conversion, 41% CO₂ conversion, 19% H₂ yield and 0.62 H₂/CO ratio at 750°C up to 28 h TOS.⁵ Impregnation of 1 wt% Ni colloidal solution (mixture of Ni nitrate precursor, pyrogallol[4] arene CHCl₃ and H₂O) over ZrO₂ support had an average particle diameter of 1.1 nm, and nearly all Ni⁶ atoms are at the interface and perimeter. 50 mg of catalyst converted 1CH₄:1CO₂:2Ar feed (120 L g⁻¹ h⁻¹ WHSV) and showed 18% H₂ yield in 15 h TOS. Towards low-temperature dry reforming of methane, ‘Si-promoted ZrO₂-supported Ni catalyst’ (NiSi/ZrO₂) prepared by co-impregnation method had brought attention.⁷ It retained the size of metallic Ni in the range of 12–9 nm at 400°C up to 15-h reaction and showed 4.3% CH₄ conversion and 3.8% CO₂ conversion, 1.5% H₂ yield.

Increasing different metal oxide portions with support ZrO₂ has been reported worthily in Ni-supported system in favour of DRM either as catalytic excellence or as coke resistance. Further addition of promotor/modifier over ZrO₂-supported/ ZrO₂ + MO_x-supported system was found to modulate the acido-basic surface property, metal-support interaction and reducibility, greatly favouring DRM. 8 wt% Ni impregnated over commercially available pelletized cerium-zirconia (8Ni/CeO₂-ZrO₂) showed enhanced basicity and enhanced oxygen ion lability and less carbon deposit.⁵ 100 mg 8Ni/CeO₂-ZrO₂ catalyst was employed for 30CH₄:20CO₂:50 Ar gas feed (WHSV = 120 L g⁻¹ h⁻¹) which showed about 40% CH₄ conversion, 60% CO₂ conversion, 35% H₂ yield, 0.9 H₂/CO ratio at 750°C up to 28 h TOS. In longer TOS, resistance against thermal sintering of Ni particle was attained due to protective action of a [O^δ−, O^δ+] dipolar layer created on the surface of Ni particles via spontaneous thermally driven spill over of O²⁻ ions.⁸ The catalyst 8 wt%Ni/18 wt%CeO₂-ZrO₂ converted 55CH₄:35CO₂:10 Ar (WHSV = 40 L g⁻¹ h⁻¹) and showed 38% CH₄ conversion, 67% CO₂ conversion, 37% H₂ yield, 0.8 H₂/CO ratio at 750°C up to 50 h TOS. xCeO₂-(100-x) ZrO₂ (x = 28–100) support prepared by co-precipitation method was also used of dispersion as well as stabilization of 5 wt% Ni. Ceria-zirconia solid solution, increased lattice defects and increased reducibility of catalyst system were the main cause of good catalytic activity.⁹ The zirconia-rich catalyst system was found to attain higher activity and coke resistance than the ceria-rich catalyst system. 5 wt%Ni/28 mol%CeO₂-72 mol%ZrO₂ catalyst over 1CH₄:1CO₂ gas feed (GHSV = 30 L g⁻¹ h⁻¹) showed 47% CH₄ conversion, 53% CO₂ conversion, 34% H₂ yield, 0.9 H₂/CO ratio at 750°C up to 21 h TOS. 5 wt%Ni-impregnated ceria-zirconia (60:40 atom ratio) catalyst had higher reducibility of NiO.¹⁰ It was employed over 2CH₄:2CO₂:1N₂ gas feed (WHSV = 60 L g⁻¹ h⁻¹), resulting in 63% CH₄ conversion, 57% H₂ yield and 0.88 H₂/CO at 700°C during 6-h time on stream. 15 wt%Ni dispersed on nanocrystalline cubic Ce_0.8Zr_0.2O₂ support by co-precipitation method had dispersed-exposed nano-sized NiO_x crystallite in intimate contact with support.¹¹ Additionally, ceria induced enhance oxygen transfer. So, it interacts with 1CH₄:1.08CO₂:1N₂ feed (108 L/h g GHSV) and showed >95% CH₄ conversion, >95% CO₂ conversion and >95% H₂ yield at 800°C up to 100-h reaction.

Yttria was noticed for advancement in surface adsorbed oxygen species.¹² 15 wt.% yttria incorporation in support ZrO₂ caused size optimization of NiO up to 22 nm (against 36.8 nm in zirconia-supported Ni catalyst) and enhanced surface parameters.¹³ It bears the high intensity of strong basic site/lattice oxygen and unidentate carbonate species along with bidentate carbonate and formate species. Overall, at 700°C, 15 wt.% yttria-85 wt.% zirconia-supported Ni catalyst showed 67.5% hydrogen yield, which decreased to 63.7% at a 420-min time on stream (TOS). In Ni-based catalyst system, incorporation of 9 wt.% WO₃ in ZrO₂ support caused enhanced redox property, additional CH₄ decomposition sites and stabilization of monoclinic ZrO₂ during high-temperature DRM reaction.¹⁴ It showed 55% H₂ yield, but it decreased due to the limited re-oxidizing capability of the surface by CO₂ and shading of the catalytic active site by thermally stable carbonates. Ceria promotional addition over tungsten oxide-zirconia-supported Ni catalyst was noticed for sequential oxygen vacancy generation.¹⁵ After adding ceria as promotor, surface-enhanced CO₂ interaction and re-oxidizing capability (by CO₂) up to pristine level. 2.5 wt.% ceria promotional addition in the catalyst had constantly produced 78% H₂ yield up to 420 min. Ni-based catalyst incorporation of 9 wt.% P on ZrO₂ support provided more than one CH₄ decomposition site (as nickel metal and nickel phosphate), brought charge transfer band closer, physisorbed and chemisorbed (as carbonate and carboxylate) CO₂ species.¹⁶ It showed a 76% H₂ yield with an H₂/CO ratio of 0.95. 3 wt.% ceria promotional addition offers more CH_x species held by surface with the different extents and inputs mobile lattice oxygen into the catalytic system for possible carbon deposit oxidation. It had additional bicoordinated formate species, and the surface reoxidizible capability by CO₂ had also increased to a great extent. It showed H₂ yield up to 97% with an H₂/CO ratio of ~1.

The 8 wt.%Ni impregnated over commercially available pelletized lanthana-zirconia (8Ni/La₂O₃-ZrO₂) showed enhanced basicity and catalytic activity than ZrO₂-supported Ni catalyst.⁵ 100 mg 8Ni/La₂O₃-ZrO₂ was employed for 30CH₄:20CO₂:50 Ar gas feed (WHSV = 120 L g⁻¹ h⁻¹) which showed about 38% CH₄ conversion, 54% CO₂ conversion, 28% H₂ yield, 0.8 H₂/CO¹⁷ ratio at 750°C up to 28 h TOS. Lanthana-zirconia support prepared by urea hydrolysis method (calcined in air or argon at 250°C) was used to disperse 5 wt% Ni and tested for low-temperature DRM.¹⁷ Catalyst calcined under argon lead less mobility of NiO and the formation of small Ni particles where carbon deposits found less time to interact with each other, and it showed coke resistance. It interacted with 1CH₄:1CO₂:8N₂ feed (14.4 L g⁻¹ h⁻¹ WHSV) and showed 6.3% CH₄ conversion, 10.8% CO₂ conversion and 4.5% H₂ yield at 400°C reaction temperature up to 100-h TOS. On incorporating 9 wt.% La₂O₃ in ZrO₂ support, Ni-supported catalyst caused stabilization of ZrO₂ phase and presence of additional lanthanum coordinated oxycarbonate species (La₂O₃·CO₂), prominent decomposition of CH₄ decomposition at Ni in the vicinity of LaZr interface.¹⁸ It showed 75%–80% H₂ yield up to 460 min in TOS. Nonetheless, NiO crystallite formation during the process was induced by catalyst deactivation. The addition of ceria (2.5 wt.%) promoted the formation of a ternary redox system lanthana-ceria-zirconia, increased NiO support interaction, increased the density of La₂O₃.CO₂ and increased oxygen replenishment from CO₂ to the catalytic surface. It demonstrated a maximum H₂ yield of 87%, which remained constant throughout the 460 min in TOS.

Apart from ceria, the role of another promotor over the lanthana-zirconia system has to be investigated, where the use of Ca, Cr, Ga and Gd promoters may be effective. The addition of Ca is known for increasing the basicity of the support and thereby inducing more CO₂ interaction with the catalyst's surface and oxidation of carbon deposit.¹⁹ Previously, 0.02 mol Ca and 0.04 mol Ni containing 0.10 mol ZrO₂ support prepared by reflux-mediated co-precipitation method had shown high surface area and higher Ni dispersion than that prepared by the impregnation method,² and 50 mg of it showed >60% H₂ yield with H₂/CO ~ 0.9 up to 35 h in the TOS. In 5 wt.% Ni impregnated 8 wt% CaO-ZrO₂ support (prepared by polymerization method) catalyst, CaO made intimate contact with Ni causing reducibility of metallic Ni²⁰ 100 g catalyst had shown 74% CH₄ conversion, 82% CO₂ conversion, 0.78 H₂/CO ratio, 3 × 10³ H₂ molar production with 36 L g⁻¹ h⁻¹ WHSV in 350 min at 800°C. A deep literature report on a series of the catalyst system in terms of CH₄ conversion, CO₂ conversion, H₂/CO ratio on different reaction conditions is shown in Table S1.^{12,15,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47} Potassium (K) promotional addition was marked for increasing the mesoporosity and reducibility of the catalyst,⁴¹ Na addition for raising the basicity as well as the formation of strong interacting Ni species (NiO_xH_y),⁴² Ba addition for the formation of barium zirconate which stabilizes the NiO_x species strongly over the surface against agglomeration⁴³ and Mg addition for increasing the basicity as well as reducible NiO-MgO solid solution in favour of DRM.⁴⁸ Mg-Al (4:1) promoter brought dispersed small size NiO, strong metal-support interaction, high reducibility and high surface area,⁴⁴ Al and Mn (1:1) had optimized Ni size, high basicity and high adsorbed oxygen species.⁴⁵ Chromium was found to enhance CO₂ absorption over the surface.⁴⁹ Also, the catalyst prepared by co-impregnating Ni (5 wt%) and Cr (7.8 wt%) precursor over char had better dispersed Ni and higher activity than sequential impregnation of precursor. Chromium addition over Ni-based catalyst system was previously identified for modifying support basicity as well as forming Ni-Cr alloy, which is oxidation resistant, tough and stable for the high-temperature conditions.^6,7 Ga addition may increase the Lewis sites while reducing the size of Ni particles.^50,51 Gd had a strong interaction with Ni addition as spin-orbit interaction (4f-3d charge transfer) between Gd and Ni, which inhibited the formation of NiAl₂O₄ phases. Larger Gd³⁺ promotor doping in smaller Zr⁺⁴ has been found to create a good number of defects. Gadolinia-doped zirconia solid solutions are known for high sintering temperatures, and Gd₂O₃-ZrO₂ catalyst system is known for high chemical homogeneity.^52-54

After investigating a good deal of literature survey, we conclude that surface modification property carried out by Ca, Cr, Ga and Gd promoter over the zirconia-based system is markable and needs more investigation in hydrogen-rich syngas production through dry reforming of methane. Herein, we have deliberately studied Ca-, Cr-, Ga- and Gd-promoted lanthana-zirconia–supported Ni catalyst system (Ni-M/LaZr; M = Ca, Cr, Ga, Gd). Ni coordination environment, CO₂-bonding surface species, the presence of reducible NiO-interacting species (with support/promotor) and oxygen replenishment capacity of CO₂ to different promoted reduced catalyst systems are investigated by UV-Vis spectroscopy, infrared spectroscopy, CH₄-temperature programmed surface reaction and ‘H₂TPR-CO₂TPD-H₂TPR’ cycle experiment. Finally, a scientific correlation of characterization results and H₂ yield over Ni-M/LaZr (M = Ca, Cr, Ga, Gd) catalyst system was found.

EXPERIMENTAL Catalyst preparation

The support 9 wt% lanthanum-91 wt.% zirconia was gifted by Daiichi Kigenso Kagaku Kogyo Co., Ltd. Osaka-Japan. 5 wt% nickel nitrate hexahydrate solution was added over support 9 wt% lanthanum-91 wt% zirconia under stirring for 3 h. The solution was kept at 90°C. Then, the catalyst slurry dried at temperature 120°C overnight in an oven. The dried mixture was calcined under a programmable muffle furnace at 600°C at the rate of 3°C/min for 3 h. After calcination, 1 wt.% nitrate hexahydrate salt of promotors (calcium, chromium, gallium, gadolinium) was added with lanthanum-zirconia supported Ni catalysts by the co-impregnation method. Calcium-, chromium-, gallium-, gadolinium-promoted lanthana-zirconia–supported Ni catalysts are abbreviated as Ni-Ca/LaZr, Ni-Cr/LaZr, Ni-Ga/LaZr, Ni-Gd/LaZr, respectively.

Catalyst characterization

The catalysts were characterized by X-ray diffraction (XRD), Raman spectroscopy, IR (infrared spectroscopy), ultraviolet-visible spectroscopy (UV-vis), scanning electron microscopy energy-dispersive X-ray spectroscopy (SEM-EDS), CH₄-temperature programmed surface reaction (CH₄-TPSR), ‘H₂-temperature programmed reduction (H₂TPR)-CO₂-temperature programmed desorption (CO₂TPD)-H₂-temperature programmed reduction (H₂TPR) cycle’ and thermal gravimetric analysis (TGA). Instrument specification detail and characterization procedure are given in supporting information S1.

Catalyst activity test

Dry reforming methane experiment is carried out in a stainless tubular reactor (PID Eng. & Tech. Micro. Activity company) with 0.91 cm diameter and 30 cm length under one atmospheric pressure and at 700°C. 100 mg catalyst was charged into the reactor. Before the reaction, the reductive treatment of the catalyst is carried out under the flow of 20 ml/min hydrogen at 600℃ for 60 min. The mixture of gases CH₄, CO₂ and N₂ in the ratio of 3:3:1 is passed through the catalyst bed in the reactor with a total flow rate of 70 ml/min. The product was analysed by gas chromatography with TCD equipment. The following expression determines H₂ yield:[Image Omitted. See PDF]

RESULT Catalytic activity results

H₂ yield performance of catalyst without promoter and with promoter shown in Figure 2. Lanthana-zirconia–supported Ni catalyst gives 70.5% H₂ yield, which decreases continuously to 58% within 440 min. After incorporating various promoters such as Ga, Ca, Gd and Cr into lanthana-zirconia–supported nickel catalysts, the catalyst performance increases and remains almost constant for up to 440 min on stream. During the 440-min time on stream, the calcium-promoted catalyst shows 73.8%–72.2% H₂ yield, while the gallium-promoted catalyst shows 76.5%–73.7% H₂ yield. Gadolinium- and chromium-promoted catalysts with equal H₂ yield (80% and ~81%, respectively), but comparably chromium promotion brings more stable performance than gadolinium-promoted catalyst towards hydrogen production from dry reforming of methane. After 440 min, the H₂ yield dropped to 77.4% in Gd-promoted catalyst while the drop was limited to 79.5% in Cr-promoted catalyst. Even on longer TOS (33 h), the H₂ yield of chromium-promoted catalyst remained constant at about 79%.

FIGURE 2. (A) Catalytic performance of Ni/LaZr and Ni-X/LaZr (X = Ca, Cr, Ga, Gd) in a different catalytic systems for dry reforming methane (DRM). (B) Catalytic performance of Ni-Cr/LaZr catalytic system for dry reforming methane (DRM) up to 33 h

Catalyst characterization results

XRD diffractograms of promoted and unpromoted catalysts are shown in Figure 3. The support lanthana-zirconia has hexagonal ZrO₂ phase (at 2θ = 30.0, 34.9, 50.0, 59.9, 72.7°; JCPDS card reference number 00-037-0031). After 5 wt.% Ni loading, cubic NiO phase (at 2θ = 37.2, 43.2, 62.4°; JCPDS card reference number 01-073-1523) and tetragonal ZrO₂ phase are appeared (at 2θ = 29.9, 34.9, 43.2, 49.8, 59.9, 62.5, 74.0, 81.1, 83.3, 84.4°; JCPDS card reference number 01-071-1282). It indicates phase transfer of ZrO₂ from hexagonal ZrO₂ to tetragonal zirconia. Further, in promoted catalyst sample, the tetragonal phase is retained. In calcium-promoted catalyst sample, additional orthorhombic ZrO₂ diffraction lines (at 2θ = 30.0, 35.0, 43.3, 49.9, 59.9, 62.4, 73.9, 81.4, 84.4°; JCPDS card reference number 00-037-1010), in chromium-promoted sample lanthanum zirconium oxide phase (at 2θ = 29.9, 34.8, 49.9, 59.8, 62.4, 72.7, 73.8, 81.2, 84.3°; JCPDS card reference number 00-037-1413) and in gadolinium-promoted sample additional orthorhombic ZrO₂ phase (at 2θ = 29.9, 34.9, 43.3, 50.0, 59.8, 62.4, 73.9, 81.3, 84.4°; JCPDS reference number 00-037-1413) and zirconium gadolinium oxide phase (at 2θ = 29.9, 37.3, 73.9, 81.3, 84.4°;JCPDS reference number 01-080-0469) are found.

FIGURE 3. XRD profile of Ni/LaZr and Ni-M/LaZr (M = Ca, Cr, Ga, Gd) catalyst used for dry reforming methane (DRM)

The Raman profile of promoted and unpromoted catalysts is shown in Figure 4. The Raman band at 268, 457 and 630 cm⁻¹ is due to the tetragonal zirconia phase.¹⁴ The variance in peak intensity at 268 cm⁻¹ is clearly observed in all catalyst samples. Unpromoted catalyst has tetragonal zirconia phase in moderate amount. After the addition of Ga, the tetragonal zirconia phase remains unaffected. On addition of Cr, the tetragonal zirconia phase is intensified, whereas on Ca or Gd promotor addition, the tetragonal zirconia intensity is reduced. The lower frequency shifts of Raman peak (about 421 and 590 cm⁻¹) in Cr-promoted catalyst indicates incorporation of a foreign element into the lattice, causing perturbation of the M–O bond involving some kind of disordering or structural distortion or formation of mixed oxide. The XRD of chromium-promoted catalyst also showed the presence of lanthanum zirconium oxide in the sample.⁵⁵ Chromium-promoted catalyst has an additional peak at 847 cm⁻¹ which is attributed to bridging Cr-O-Cr vibration of polychromate species.⁵⁶

FIGURE 4. Raman spectra of Ni/LaZr and Ni-M/LaZr (M = Ca, Cr, Ga, Gd) in a different sample

IR (Infrared spectroscopy) spectra of catalyst samples are shown in Figure 5. IR was studied in the range of 400–4000 cm⁻¹. After Ni anchoring over lanthana-zirconia support, no new absorption band has appeared other than support lanthana-zirconia. However, IR absorption band intensity at 1628 cm⁻¹ (attributed to OH stretching vibration) and at 3428 cm⁻¹ (attributed to OH bending vibration) have decreased after Ni anchoring. Three absorption bands were found at $v_1{v}_1$- 849 cm⁻¹ for CO₃⁻² symmetric stretching (vibration belongs to the pure form of typical ionic anhydrous lanthanum carbonate La₂(CO₃)₃), $v_2{v}_2$- 1070 cm⁻¹ for the free ion of CO₃⁻² symmetric stretching vibration band (which is inactive for IR spectra, but the absorption band is active for lattice and change of D₃h symmetry of free ion CO₃⁻² to lower symmetry C_2v or Cs) and $v_3{v}_3$- 1367 cm⁻¹ for vibration of oxycarbonate strongly coordinated with La³⁺ (La₂O₃.CO₂) species. All are attributed to La₂O₂CO₃ species, and these IR absorption bands are present over the catalyst surface.^18,57 It indicates easy adsorption of CO₂ on the surface. Non-promoted catalyst has stretching vibration peak of C-H at 2850 cm⁻¹ and combination of asymmetric stretching of COO and bending vibration of C-H bond peak at absorption band at 2932 cm⁻¹.¹⁶ On Ca, Gd, Ga promotional addition, absorption bands about 2850 and 2932 cm⁻¹ are disappeared. On chromium promotional addition, the absorption bands at 1070 and 1368 cm⁻¹ disappeared, whereas the absorption band at 915 cm⁻¹ for stretching vibration of Cr=O,⁵⁸ 2850 and 2932 cm⁻¹ has appeared. It indicates that Cr presence brought major bonding patterns over the catalyst surface.^59,60

FIGURE 5. IR spectra of (A) Ni/LaZr and (B) Ni-M/LaZr (M = Ca, Cr, Ga, Gd) different catalytic systems used for dry reforming methane (DRM)

UV spectra of all catalyst samples are shown in Figure 6. The bandgap of lanthana-zirconia support is found to be 5.25 eV. The addition of nickel on the support shows an intense peak at 280 nm, which is attributed to the charge transfer band from O₂²⁻ to Ni²⁺ in an octahedral environment (O_h). The d-d transition UV band is also shown at 377 and 410 nm for ³A_2g(F) state to ³T_1g(P) state of Ni²⁺ (O_h) and at 718 nm for ³A_2g(F) state to ³T_1g(F) state of the Ni⁺² (O_h). UV peaks at 280, 377, 410, 718 nm justify the Ni²⁺ octahedral coordination environment. The bandgap of lanthana-zirconia–supported nickel catalyst is decreased to 3.45 eV. It indicates that the valance band and conduction band come nearer after nickel addition. After adding Ga, Cr, Gd and Ca promoters, bandgaps are modified to some extent (3.52, 2.70, 3.44 and 3.44 eV, respectively). Interestingly on the addition of Cr promotor, the d-d transition band at 377 and 410 nm (³A_2g(F) → ³T_1g(P) of Ni²⁺ (O_h)) has vanished completely, and a new peak at 368 nm for O-Cr(VI) electronic charge transfer of Cr₂O₇²⁻ species appeared.⁶¹ It indicates the interruption of d-d transition among Ni²⁺ states (³A_2g(F) → ³T_1g(P)) by the electronic transition of O-Cr(VI) of Cr₂O₇²⁻, implying that two charge clouds related to Ni and Cr centre respectively are under strong interaction.

FIGURE 6. Ultraviolet-visible spectroscopy (UV-Vis) and bandgap of Ni/LaZr (bandgap = 3.45 eV); and Ni-M/LaZr (M = Ga, Cr, Gd and Ca) (bandgaps = 3.52, 2.70, 3.44 and 3.44 eV, respectively) different catalytic samples

The SEM and EDX profiles of promoted and unpromoted catalysts are shown in Figure 7. All of the elements claimed in the catalyst are visible in the EDX patterns. For Ni/LaZr, Ni-Ca/LaZr catalysts, as Ca promoter is added, Ni element proportion decreases, and the EDX mapping in Figure 7B,C shows a decrease in the concentration of Ni from 4.5 to 3.5 wt.%. At the same time, the wt.% of Ni element decreases as follows: Ni-Cr/LaZr < Ni-Ga/LaZr ≈ Ni-Gd/LaZr catalysts Figure 7D–F.

FIGURE 7. SEM image of (A) LaZr (B) Ni/LaZr (C) Ni-Ca/LaZr (D) Ni-Cr/LaZr (E) Ni-Ga/LaZr (F) Ni-Gd/LaZr. EDX profile of (a) LaZr (b) Ni/LaZr (c) Ni-Ca/LaZr (d) Ni-Cr/LaZr (e) Ni-Ga/LaZr (f) Ni-Gd/LaZr

CH₄-temperature programmed surface reaction (TPSR) of promoted and unpromoted catalysts is carried out and shown in Figure 8A. The peak near 830°C temperature is attributed to the thermal decomposition of CH₄, whereas the peak at a lower temperature near 400°C is attributed to CH₄ decomposition at the nickel surface. The broad peak from 400 to 800°C is attributed to CH₄ decomposition at intimate nickel contact with the lanthana-zirconia interface. Among all, chromium-promoted catalyst had the widest range of CH₄-TSPR peaks, including a peak at about 400°C for CH₄ decomposition site at nickel surface, a broad peak in the range of 400–750°C for CH₄ decomposition peak at intimate nickel contact with lanthana-zirconia interface and diffuse peak centred about 830°C for thermal decomposition of CH₄ at lanthana-zirconia surface.

FIGURE 8. CH4-TPSR profile of (A) catalyst and H2TPR-CO2TPD-H2TPR cycle of (B)–(F) Ni/LaZr and Ni-M/LaZr (M = Ca, Cr, Ga, Gd)

‘H₂TPR-CO₂TPD-H₂TPR’ cycle for promoted and unpromoted catalysts shown in Figure 8B–F. H₂TPR profile of catalyst system can be generalized by low-temperature reduction peak centred about 300°C for ‘NiO species weakly interacted with the support’ and high-temperature peak approximately at 475°C for ‘NiO species strongly interacted with the support’. Lanthanum-zirconia supported Ni catalyst system has both low-temperature and high-temperature peaks. Ga, Gd and Cr promotional addition showed a more intense reduction peak for ‘NiO species strongly interacted with the support’ than ‘NiO species weakly interacted with the support’. In contrast, calcium promotional addition caused a more intense reduction peak for ‘NiO weakly interacted with the support’ than ‘NiO strongly interacted with the support’. The reduction profile on chromium promotion is specific. It has a reduction peak centred at 186, 338 and 442°C. The low-temperature peak at 186°C may be due to the reduction of ‘chromate species weakly interacted with support’⁶² (Cr⁺⁶ → Cr⁺³) O₃ species. The high-temperature peak in chromium-promoted catalyst is intense and shifted to a lower temperature concerning the unpromoted catalyst. It indicates the increasing reducibility of surface, as well as high density of ‘NiO species strongly, interacted with the support’. However, high-temperature reduction peaks due to the reduction of ‘chromate species strongly interacted with support’ cannot be neglected.^62,63

The CO₂TPD profile of the catalyst system can be generalized by low-temperature desorption peak attributed to weak basic site and intermediate temperature desorption peak attributed to moderate strength basic sites attributed to surface anion.³³ During H₂ treatment, surface oxygen species (which is known for moderate strength basic sites) may be removed, and therefore, surface oxide vacancies would be created. So, if CO₂TPD is carried out in continuation with H₂TPR, only weak basic sites (surface hydroxyl) are detected, whereas moderate strength basic sites (surface anion) are not detected in all catalyst samples. It should be considered that during the interaction of the CO₂ stream with a surface, surface oxide vacancy may be replenished by oxygen of CO_2, and the surface would be oxidized again. It will be interesting to observe the extent of oxygen replenishment (by CO₂) to surface oxide vacancy. So, if further H₂TPR is continued over ‘H₂TPR-CO₂TPD treated catalyst sample’, the comparison of reduction peak intensities of catalyst sample (with H₂TPR peak of fresh catalyst sample) shows the extent of replenishment of oxygen to the vacancy by CO₂. In a non-promoted catalyst, the lower temperature reduction peak is recovered and shifted to a lower temperature, whereas the higher temperature reduction peak is negatively affected as it is not recovered and shifted to the higher temperature. It indicates that CO₂ re-oxidizes the reduced surface (created after H₂TPR of the fresh sample) in the lower temperature range more easily than the higher temperature region. ‘NiO weakly interacted with the support’ (in low-temperature range) is recovered more easily but not ‘NiO strongly interacted with the support’.

In Ca-promoted catalyst, CO₂ re-oxidizes the reduced surface such that both ‘NiO interacted weakly with support’ as well as ‘NiO interacted strongly with support’ are recovered to a certain extent (Figure 8C). The latter one is recovered more greatly as well as shifted to a higher temperature which indicated higher metal-support interaction. Ga-promoted catalyst shows reoxidation of a reduced surface by CO₂ such that both types of reducible NiO species are remarkably recovered and shifted towards higher temperature (Figure 8D). It also shows additional diffuse reduction peaks in a very low-temperature region (about 250°C), which may be attributed to reducing free NiO species.⁶⁴ The shift of major reduction peak towards higher temperature showed increased metal-support interaction again. In Gd- and Cr-promoted catalyst, CO₂ not only re-oxidizes the reduced surface (created after H₂TPR of the fresh sample) up to the optimum level (Figure 8E,F) but also the peak patterns are shifted to a higher temperature. It indicates the reducing-oxidizing capability as well as increased metal-support interaction of catalyst surface.

The Raman of spent catalysts has shown F supporting information Figure S1. All spent catalysts showed ‘defects carbon band’ (I_D) at 1340 cm⁻¹ and ‘graphite band’ (I_G) at 1560 cm⁻¹ and 2D band at 2673 cm^{−16,28,29,31} and intensity of defects carbon band are less than graphite band. Raman spectra of ‘spent Ni-Cr/LaZr catalyst obtained after 33-h reaction’ showed less intensity peak than ‘spent Ni-Cr/LaZr catalyst obtained after 440-min reaction’ (Figure 9). It indicates that carbon deposits of both types are also removed during the long run reaction.

FIGURE 9. Raman spectra of fresh and spent Ni-Cr/LaZr catalyst

Interestingly, graphitic carbon is removed more pronouncedly than defect carbon. Apart from oxidizing action of CO₂, the oxidizing role of chromate species at high temperatures may also have a crucial role in oxidizing carbon deposits [10.1016/j.matchemphys.2019.122571]. TGA profile of spent Ni-Cr/LaZr also showed prominent weight loss (Figure S2). In the long run, the constant catalyst performance (33 h) and coke deposition over chromium-promoted catalyst indicate that coke is oxidizable and removable and does not affect the catalytic performance even in longer TOS. On comparing the Raman spectra of fresh and spent Ni-Cr/LaZr catalyst, it is found that peaks due to tetragonal zirconia and polychromate species were suppressed after the reaction. It indicates the active role of ZrO₂ and polychromate species in the dry reforming of methane.

DISCUSSION

Interestingly, Ni dispersion over lanthana-zirconia support stabilizes the tetragonal ZrO₂ phase and brings valance band and conduction band nearer (bandgap 3.45 eV in Ni/LaZr than 5.25 eV in LaZr). Lanthana-zirconia–supported nickel catalyst has additional amorphous La₂O₂CO₃ species and format species which indicates wide interaction of CO₂ over the catalyst surface. It had reducible NiO species interacted either weakly or strongly with the support. The reduced catalyst is re-oxidizable by CO₂ more pronouncedly in the lower temperature regions. NiO species that are weakly interacted with the support are regenerated exclusively than NiO species that strongly interact with the support. In the mean of CH₄ decomposition sites, lanthana-zirconia–supported nickel catalyst facilitates the CH₄ decomposition pronouncedly over nickel in intimate contact with lanthana-zirconia interface. Overall, this catalyst system can produce H₂ with a 70% yield from dry reforming of methane. However, the activity drops rapidly to 58% within 440 min of the stream test time.

On calcium promotional addition, the intensity of the tetragonal ZrO₂ phase is decreased, and an additional orthorhombic ZrO₂ phase has appeared, whereas IR absorption bands for format species disappeared. The bandgap between conduction and valance band is modified nominally. It had reducible NiO species that interacted weakly and strongly with the support. The intensity of the earlier one is greater than the latter. The reduced surface is re-oxidizable by CO₂ comparably in both low- and high-temperature regions. That is, NiO species that interacted either weakly or strongly with the support were regenerated to a certain extent. Relatively ‘NiO species interacted strongly with support’ is generated to a great level, and the concerned reduction peak is shifted to higher temperature indicating reducing-oxidizing capacity as well as stronger metal-support interaction during the reaction. The Ni-Ca/LaZr catalyst's wide range of reducible NiO, strong metal-support interaction and high re-oxidizing capability result in a higher and more consistent H₂ yield (73.8%–72.2%) than an unpromoted catalyst in 440 min on stream.

When gallium is added, the tetragonal zirconia phase is retained, the format species disappears, and there is more ‘reducible NiO interacted strongly with the support’ than ‘reducible NiO interacted weakly with the support’. It demonstrates great oxygen replenishment by CO₂ with substantial metal-support interaction with both types of reduced NiO species. Altogether, it had a high and constant H₂ yield (76.5–73.7) in 440 min of time on stream.

On account of Gd and Cr promotion, the catalyst has enriched zirconia phases and mixed oxide phases. As Gd-promoted lanthana-zirconia–supported Ni catalyst has additional orthorhombic zirconia and zirconium gadolinium oxide phases, Cr-promoted lanthana-zirconia–supported the Ni catalyst contains high-intensity tetragonal zirconia and lanthana-zirconia oxide phase. Both catalyst systems had excellent oxygen replenishment (by CO₂) such that the reduced catalyst system was regenerated up to optimum level with increased metal-support interaction. So, both are highly performed (hydrogen yield 80%–80.9%) catalyst systems initially. It should be noted that the chromium-promoted catalyst system had a minimum energy gap (between valance band and conduction band), an additional format species for enhanced CO₂ adsorption and Ni-Cr interaction for the felicitating widest range CH₄ decomposition sites and additional chromate species for carbon deposit oxidation.⁶⁵ As a result, the Cr-promoted catalyst system outperforms the Gd-promoted catalyst system in the longer run. After 440 min of time on stream, Gd-promoted catalyst has 77.4% H₂ yield, while Cr-promoted catalyst has 79.5% H₂ yield. The hydrogen yield remained constant up to tested 33-h time on stream.

During reduction in H₂, ‘NiO species strongly interacted with support’ turns to metallic nickel under ‘strong metal-support interaction’ over which CH₄ decomposition (into carbon deposit and H₂) is felicitated in the majority. It is generally accepted that CO₂ has an oxidizing role in DRM, oxidizing carbon deposits. In our H₂TPR-CO₂TPD-H₂TPR cyclic experiment, it is observed that the reduced ‘NiO species interacted with support’ is oxidized to different extents by oxygen replenishment by CO₂. It can be argued that the oxidized NiO species (from metallic Ni) may cause a decrease in DRM activity as CH₄ decomposition happens over metallic Ni. Nevertheless, in our catalytic activity data of H₂ yield, it was found that those catalysts which had more amount of reducible ‘NiO species strongly interacted with support’ or which had shown the great extent of the re-oxidizing capability of such reduced nickel-species (through oxygen replenishment by CO₂) had good H₂ yield. So, it can be concluded that in the presence of all three streams CH₄, H₂, and CO₂ over the catalyst surface, a dynamic equilibrium among ‘reduction of NiO into Ni by H₂’, ‘decomposition of CH₄ over metallic Ni (into carbon deposit and H₂)’ and ‘oxygen replenishment by CO₂ to the reduced metallic Ni followed by oxidation of carbon deposit by such replenished oxygen’ are required for maximum H₂ production.

CONCLUSION

Limitation of the re-oxidizing capacity of reduced ‘NiO species strongly interacted with support’ (by CO₂) in lanthana-zirconia–supported Ni catalyst causes lower catalytic activity. Ga-, Gd- and Cr-promoted lanthana-zirconia–supported Ni had more amount of reducible ‘NiO species strongly interacted with the support’, which makes it ahead than Ca-promoted catalyst in hydrogen production through DRM. The presence of mixed oxide and regeneration of reduced catalyst up to optimum level (by CO₂) in both Gd-promoted catalyst and Cr-promoted catalyst made cause optimum ~80% H₂ yield initially. Cr-promoted catalyst system outperforms the Gd-promoted catalyst system in the long run where it maintained the constant ~79% conversion up to 33 h due to the smallest energy gap between valance and conduction band, Ni-Cr interaction species for wide range CH₄ decomposition and chromate species for profound carbon deposit oxidation.

ACKNOWLEDGMENTS

The authors would like to extend their sincere appreciation to Researchers Supporting Project number (RSP-2021/368), King Saud University, Riyadh, Saudi Arabia. Dr Ahmed Osman and Professor David Rooney would like to acknowledge the support given by the EPSRC project ‘Advancing Creative Circular Economies for Plastics via Technological-Social Transitions’ (ACCEPT Transitions, EP/S025545/1) and the support of The Bryden Centre project (Project ID VA5048), which was awarded by The European Union’s INTERREG VA Programme, managed by the Special EU Programmes Body (SEUPB), with match funding provided by the Department for the Economy in Northern Ireland and the Department of Business, Enterprise and Innovation in the Republic of Ireland. J K acknowledges the administration of Sankalchand Patel University. R K and V K S acknowledge the administration of Indus University.

DISCLAIMER

The views and opinions expressed in this paper do not necessarily reflect those of the European Commission or the Special EU Programmes Body (SEUPB).