Optimizing barium promoter for nickel catalyst supported on yttria-stabilized zirconia in dry reforming of methane

Abstract

Barium doping effect on the activity and stability of nickel-based catalysts, supported on yttria-stabilized zirconia (Ni-YZr), was investigated in dry reforming of methane. Catalysts were characterized by several techniques (nitrogen sorption, X-ray diffraction [XRD], scanning electron microscopy with energy dispersive X-ray, transmission electron microscopy [TEM], thermogravimetric analysis [TGA], temperature programmed oxidation, CO₂-TPD, H₂-TPR) and were tested in a fixed-bed reactor at 800°C and 42,000 mL/h g_cat. Barium played a crucial role in enhancing catalyst reducibility and CO₂ adsorption at high temperatures, as indicated by the activity and stability of the Ni-YZr catalyst. The addition of 4.0 wt% of barium appeared to be the optimal loading, allowing for CH₄ conversion of 82%, which remained constant for 7 h of reaction, compared with 72% of barium-unpromoted Ni-YZr at 800°C. TEM images of the spent catalysts revealed the formation of multiwalled carbon nanotubes on all samples. The TGA analysis showed, however, that an increase in baria loading significantly reduced the coke formation amount, indicating the inhibition of coke formation and the enhancement of the catalytic activity. Such improvement in activity and stability was attributed to the incorporation of barium into YZr support, as revealed by XRD analysis, which inhibited the sintering of the catalysts support.

INTRODUCTION

The worldwide demand for a clean environment and energy has become a driving force for using CO₂ as a starting material for chemical synthesis and a major building block in the chemical industry.^1–4 Since the industrial revolution, CO₂ emissions have increased primarily due to the various transportation means, power plants, and industrial manufacturing. Exploiting this gas for reforming processes may aid in the control of local and global CO₂ emissions while simultaneously producing hydrogen and syngas for industrial use.⁵ In this sense, dry reforming of methane (DRM) is a process for converting methane and carbon dioxide greenhouse gases into synthesis gas (syngas; a mixture of carbon monoxide and hydrogen), where hydrogen is the essential raw material for all petrochemical industries and renewable energy source. In addition, syngas is used in the Fischer–Tropsch process for generating hydrocarbon fuels, alcohols, and dimethyl ether. Besides, DRM represents a unique means of utilizing biogas as a feedstock for the petrochemical industry.⁶

The DRM is an energy-demanding process (Equation 1)⁶: [Image Omitted. See PDF]

It usually goes concurrently with reverse water–gas shift (RWGS) reaction (Equation 2): [Image Omitted. See PDF]DRM suffers from carbon deposition via the decomposition of methane (Equation 3) and/or the Boudouard process (Equation 4). [Image Omitted. See PDF] [Image Omitted. See PDF]

Generally, noble metals,¹ such as Pt,^7,8 Ru,⁹ and Rh,⁸ have high activity and stability in methane reforming. Despite this fact, they are not commercially desirable due to their scarcity and expense. Nickel-based catalysts are favored for industrial applications due to their low cost and high efficiency as noble metals.⁴ At high reaction temperatures, carbon deposition is an inevitable process. This implies that a catalyst must tolerate activity loss due to carbon deposition and agglomeration. Unfortunately, nickel-based catalysts had all of these disadvantages.^4,10 The carbon is deposited over the active sites of nickel, which blocks the catalyst bed, hence increasing the pressure drop along the catalyst bed, clogging the reactor, and shutting down the plant.^10,11 Consequently, preventing carbon deposition and prolonging the catalyst life are essential.¹² It is well known that particle size, surface area, surface acidity and basicity, metal dispersion, and thermal stability of the support and its interaction with metal impact global catalyst efficiency.⁵ Various modifications, such as the formulation of bimetallic catalysts by doping nickel with noble metals and the support doping with lanthanides, alkali, or alkaline-earth metals, can increase the catalyst's resistance to carbon deposition.¹³ Incorporating these metals into the support and their interactions with the nickel-metal are crucial in determining the identity of the DRM support.¹⁴ The combination of ZrO₂ and SiO₂ indicated that acidity was not the only determining factor in dry reforming catalysts, but it also affected the active sites, accessibility to surface oxygen for the oxidation of deposited carbon, metal dispersion, and their dimensions. Besides, promoters such as potassium (K) and magnesium oxide (MgO) doping have been developed.⁵ Due to its effect on the mechanism of carbon oxidation, the support plays a crucial role in preventing carbon deposition during the DRM reaction course. Due to their high basicity, ceria, lanthana, boria, and baria are advantageous supports for Ni-based catalysts for reducing carbon deposition.^15–17

In contrast, few studies have incorporated barium into perovskite structures, as this structure induces the separation of barium and barium carbonate. However, the dispersion of barium species within the NiLa-based catalyst may effectively contrast coking kinetics⁶ due to its inherent properties of high dielectric constant, proton conductivity, and the ability to produce hydrogen by chemical reactions due to its high thermal stability. Barium enhances CO₂ adsorption by increasing the basicity of the BaZrO₃ catalyst.^18–22 In addition, barium creates oxygen vacancy and accelerates the carbon dioxide reaction. The performance of oxygen carriers in the partial oxidation of the methane stage is crucial for the composition of syngas products. It maintains the ideal crystal structures, the macroporous layer of BaCoO₃ adhered to the monolith achieved a good energy supply and a high gas–solid contact area. The study of Ding et al. suggested that the CeO₂-supported Ba–Co-based perovskite has a good reaction selectivity in chemical-looping steam methane reforming (CL-SMR).²³

Adding 3.0% zirconia to a catalyst prolongs its life and inhibits methane decomposition. However, this improvement diminished activity and decreased the H₂/CO molar ratio. The modification with 5% zirconia produced adequate activity, but the stability was poor.^24,25 Thus, stabilizing the catalytic system with zirconia (5 wt%) was a compromise between activity and stability. The ZrO₂–Y₂O₃ solid solution was investigated in various reforming processes.^26–29 In recent years, ZrO₂ has been tested in methane-reforming reactions because its high ionic conductivity and thermal stability are very useful for these reactions. The tetragonal phase of ZrO₂ is stabilized by incorporating cations, like, Mg⁺², Ca⁺², Sr⁺², La⁺³, C^e+4, and Y⁺³, that form solid solutions resulting in oxygen vacancies in the zirconia lattice.³⁰ These oxygen vacancies have a very important role in gasifying carbon deposits and in the activation of CO₂ molecules, as has been demonstrated in previous reports on the reforming reactions of methane and ethanol.^31–33 Yttrium oxide is a white solid with a high melting point. The protective oxide coating of yttrium develops its remarkable resistance against air up to 1000°C.³⁴ The vacancies in zirconia can be easily compensated for by substitution with yttrium.³⁵ The addition of yttria to zirconia prevents the phase transition of zirconia. Yttrium in ZrO₂ is also known for high oxygen ion conductivity, improvement in bending strength, and shock resistance.^36,37

Altogether, the yttria-stabilized zirconia system appears to be good support for high-temperature DRM reaction. The catalytic role of yttria-stabilized zirconia is well recognized. The acidic and basic sites of yttria-stabilized zirconia were found to be responsible for olefination and aldol condensation of alcohol, respectively.³⁸ The fast diffusion of oxygen in conjunction with rapid oxygen activation at extrinsic oxygen vacancies in the yttria–zirconia surface induces partial oxidation of methane.³⁹ Yttria can produce oxygen vacancies in ZrO₂-based supports.^40,41 Catalytic deactivation, owing to agglomeration and metal particle sintering, is a serious challenge in the DRM and encourages researchers to design anticoke catalysts in a suitable route. Therefore, a series of yttrium oxide (Y₂O₃)-promoted Ni catalysts supported on magnesium oxide (MgO)-modified MCM-41 were successfully synthesized by the one-pot method.⁴² On the basis of reforming results, the catalytic performance depended on yttria contents, as a catalyst containing 2 wt% yttria possessed the highest NiO dispersion, oxygen vacancies, reducibility, CH₄ conversion (79%), H₂/CO ratio (0.85), and stability in time-on-stream (TOS) (20 h).

Baria is basic oxide, which reacts with carbon dioxide to form barium carbonate, according to Equation (5): [Image Omitted. See PDF]

Barium carbonate may react with carbon to regenerate baria and to produce carbon monoxide as shown in Equation (6): [Image Omitted. See PDF]

Therefore, baria is very useful in endowing the catalysts with carbon resistance properties.^6,8,15,43,44

Herein, several nickel-based, barium-promoted, and yttria-stabilized zirconia catalysts were prepared via the incipient wetness impregnation method to optimize the barium loading and its influence on the activity and stability of NiZrY catalysts.

EXPERIMENTAL SECTION Materials

All chemicals were obtained commercially and were used without further purification. Barium nitrate [Ba(NO₃)₂, 99.999% trace metals basis] was purchased from Sigma-Aldrich, while nickel nitrate hexahydrate [Ni(NO₃)₂·6H₂O, 98%] and mesoporous 8.0 wt% yttria-stabilized zirconia (meso-8Y₂O₃–ZrO₂ and meso-YZr) were bought from Alfa Aesar. A Milli-Q water purification system (Millipore) was used to generate ultrapure water (18.2 MΩ cm).

Catalyst synthesis

The catalysts 5Ni + xBa/YZr (x = 0.0, 1.0, 2.0, 3.0, 4.0, or 5.0 wt%) were synthesized by dry impregnation of the required amounts of Ni(NO₃)₂·6H₂O to have 5.0 wt% loading of Ni, Ba(NO₃)₂ to obtain (0.0, 1.0, 2.0, 3.0, 4.0, or 5.0) wt% loading of Ba, and of meso-YZr. Afterwards, a green mixture was produced after mixing and grinding these constituents. Onto this solid matrix, drops of distilled water were poured to obtain a thick paste, which was subsequently agitated routinely until complete dryness. The wetting, stirring, and drying processes were repeated three times to ensure a uniform distribution of loaded materials within the matrix of the meso-YZr support. The green solid product was then weighed and was calcined at 600°C for 3 h. The catalysts are cited herein as Ni-xBa-YZr (x = 0, 1, 2, 3, 4, 5).

Catalyst characterization X-ray diffraction (XRD)

Powder XRD patterns of the prepared catalysts were recorded on a Miniflex Rigaku diffractometer, which was equipped with CuK_α, X-ray radiation. The device was run at 40 kV and 40 mA. The Joint Committee on Powder Diffraction Standards (JCPDS) database was used to identify the different phases of the diffractograms.

Nitrogen physisorption

The surface area and distribution of pore sizes of the catalysts were measured by N₂ adsorption–desorption at −196°C by using a Micromeritics Tristar II 3020 surface area and porosity analyzer. The Barrett, Joyner, and Halenda (BJH) method was employed to calculate pore size distribution.

Transmission electron microscopy (TEM)

Fresh and spent samples were characterized using a Philips CM12, provided with a Gatan high-resolution camera. Samples were prepared using a drop of catalyst dispersion in ethanol for deposition on a carbon-coated copper grid.

Scanning electron microscopy with energy dispersive X-ray (SEM-EDX)

SEM analysis was carried out to study the topology of the prepared catalysts by using a Philips XL-30-FEG instrument (Eindhoven). An EDX analyzer (Oxford, model 6587) was used to determine the atomic composition with specimens of catalyst samples deposited as powders on pin flat stubs.

Thermogravimetric analysis (TGA)

The quantity of carbon deposits on the spent catalysts was measured by means of TGA. A platinum pan was filled with 10–15 mg of the used catalysts and was carefully positioned inside the device. Heating was done from room temperature up to 1000°C at a 20°C/min temperature ramp. The change in mass was continuously monitored as the heating progressed.

Temperature programmed reduction (TPR)

Seventy milligrams of the sample were loaded inside the TPR sample holder of a Micromeritics Auto Chem II apparatus. The sample was flushed with argon gas for 30 min at 150°C. After that, the sample was cooled to room temperature. The next step involved heating by the furnace up to 900°C, ramping at a rate of 10°C/min, in an atmosphere of an H₂/Ar mixture (1:9 vol%), flowing at 40 mL/min. The thermal conductivity unit recorded the consumption of H₂ during the operation.

Carbon dioxide temperature-programmed desorption (CO₂-TPD)

The samples (ca. 100 mg) were washed in helium flow and were then saturated for 1 h at 200°C in an atmosphere of 20 vol% CO₂/He. A carrier flow of helium at 25 mL/min was used in the temperature range of 25–720°C (heating rate, 12°C/min) for CO₂ desorption.

X-ray fluorescence (XRF) spectroscopy

The elemental composition of the calcinated samples was determined by XRF spectroscopy on an XRF-1800 spectrometer (Shimadzu). The samples were diluted in boric acid and pressed into a pellet before measurement.

Catalyst performance test

The dry reforming reactions were performed by using stainless-steel fixed-bed reactor (PID Eng. & Tech Micro activity Reference Company) having 9.1 mm internal diameter and 300 mm height, operated at 1.0 bar. An activation of 0.1 g of catalyst was carried out under 1200 mL/h H₂ flow for 1.0 h at 700°C. Then after eliminating the physiosorbed H₂, the N₂ treatment was performed for 15 min. The proportion of pure feed gases of CO₂/CH₄/N₂ was set to a 3/3/1 volume ratio for generating 42 L/h g_cat of gas hourly space velocity. To determine feed and output gas compositions, a thermal conductivity detector of “GC-2014 SHIMADZU” was used. After cooling the reactor by N₂ gas, the characterizations of the spent catalysts were performed. The expressions for conversions and H₂ to CO mole ratio were given as [Image Omitted. See PDF] [Image Omitted. See PDF] [Image Omitted. See PDF]

Catalyst constituent

For preparing 1.0 g of catalyst with 1.0 wt% Ba promoter, the amounts used were 0.194 g of Ni(NO₃)₂·6H₂O, 0.017 g of Ba(NO₃)₂, and 0.943 g of meso-8Y₂O₃–ZrO₂ support.

RESULTS AND DISCUSSION

Before starting with the evaluation of doped catalysts, a blank experiment was performed. In this blank test, an empty stainless-steel reactor was used without catalysts under the same feed ratio and the same reaction temperature of 800°C. The CH₄ and CO₂ conversions at 800°C of the blank resulted in 1.65% and 0.36%, respectively, with an H₂/CO ratio close to 0.12, deducing that the influence of the metallic reactor can be considered negligible.

Methane and CO₂ conversions, along with H₂/CO ratio profiles, are shown in Figure 1A–D. Results clearly indicated that the incorporation of Ba clearly affected the performance of the NiZrY catalyst in terms of both CH₄ and CO₂ conversions as well as the H₂/CO ratios. By increasing the barium loading from 1.0 up to 4.0 wt%, CH₄ conversion increased from 75% up to 82%. However, a further increase in barium loading up to 5.0 wt% lowered the CH₄ conversion to 70%, with a progressive decrease with the TOS test for up to 420 min. What import was the stability of the barium-doped catalysts with TOS. In particular, the Ni-4Ba-YZr catalyst showed the highest performance in activity and stability. The CO₂ conversion was always higher than CH₄ due to the influence of the RWGS reaction (Equation 2), in which the CO₂ was further consumed by the produced hydrogen. The CO₂ conversion of unpromoted NiYZ catalyst (see Figure 1B) was much lower than that of barium-promoted catalysts. The addition of Ba affected the reversed water–gas shift reaction due to its basic nature, favoring the adsorption of CO₂.⁴⁵

Figure 1. (A) CH4 conversion as a function of TOS, (B) CO2 conversion as a function of TOS, (C) variation of the conversion of reactants with the loading of Ba content at 270-min TOS, and (D) H2/CO as a function of TOS (reaction conditions: CH4/CO2/N2 = 3/3/1 [v/v/v]; GHSV = 42,000 mL/gcat/h; Mcat = 0.1 g; t = 800°C). GHSV, gas hourly space velocity; TOS, time-on-stream.

In terms of the H₂/CO ratio (Figure 1D), the Ni-4Ba-YZr catalyst resulted in the best H₂/CO ratio, close to 0.94. As it is well known, the values of the H₂/CO ratios were affected by the contribution of the RWGS (Equation 2) and Boudouard reaction (Equation 4).

To compare our best catalyst with the literature, it was reported in Table S1.

To try to understand in which terms the presence of Ba influences the performance of the Ni-YZr catalysts, several characterizations of both fresh and used catalysts have been made.

Figure 2A shows the N₂ adsorption–desorption isotherms of all prepared catalysts. On the basis of the IUPAC classification, the isotherms can be classified as IV-type with an H1-type hysteresis loop, indicating that all catalysts were mesoporous.⁴⁶ Due to the capillary phenomenon at relatively high pressures, condensation and evaporation occurred in the H1-type hysteresis loop, which exhibited tubular and cylindrical interconnected pores with a relatively narrow distribution. The addition of the active metal and promoter had no effect on the framework of the support. Notwithstanding, a significant increase in the region of relative pressure between 0.65 and 0.95 was observed due to the unification of N₂ condensation in the mesopores. Figure 2B depicts the pore size distribution, determined by the BJH formula. Catalysts exhibited pore diameter distribution of a bimodal type, with the peak ranging between 0 and 500 Å, reinforcing the experimental results of the N₂ physisorption, and a small peak in the macropore range (>1000 Å). These macropores would have a negligible effect on the textural properties.

Figure 2. (A) N2 adsorption–desorption isotherms of the fresh catalysts and (B) BJH pore diameter distributions for the fresh catalysts. BJH, Barrett, Joyner, and Halenda method.

The nitrogen sorption (BET) results for the catalysts are presented in Table 1. The Ni-YZr had a smaller BET surface area than YZr (30.8 and 62.6 m²/g, respectively). Incorporating Ni into yttria-stabilized zirconia blocked the pores accessible to nitrogen adsorption, resulting in a decrease in the BET surface area and the pore volume of the support. Incorporating barium in the matrix of the catalyst at loadings of 1.0, 2.0, and 3.0 wt% decreased the surface area slightly, but increasing the loading to 4.0 and 5.0 wt% maintained the surface area as that of a barium-free catalyst. On the other hand, the incorporation of barium had a negligible effect on pore volume and pore size compared with those of an unpromoted catalyst, indicating that barium did not block the pores of the support, but rather diffused in the lattice of the support.

Table 1 Nitrogen physisorption results for the prepared catalysts.

Catalyst	BET (m2/g)	Pore volume (cm3/g)	Pore size (Å)
YZr	62.6	0.29	179
Ni-YZr	30.8	0.19	248
Ni-1Ba-YZr	24.8	0.16	273
Ni-2Ba-YZr	24.3	0.17	280
Ni-3Ba-YZr	24.8	0.14	244
Ni-4Ba-YZr	30.2	0.18	244
Ni-5Ba-YZr	30.2	0.18	246

Abbreviation: BET, nitrogen sorption.

The XRD patterns of the fresh Ni-xBa-YZr (x = 0.0, 1.0, 2.0, 3.0, 4.0, 5.0) catalysts and YZr support are shown in Figure 3A. The XRD patterns of Ni-YZr and Ba-promoted Ni-YZr catalysts displayed diffraction peaks at 2θ of ~30°, ~35°, ~50°, ~60°, ~63°, ~74°, ~82°, ~84°, and ~94°, respectively, which corresponded to the (111), (200), (220), (311), (222), (400), (331), (420), and (422) crystallographic planes of the cubic phase of yttria-stabilized zirconia (JCPDS No. 49-1642). On the other hand, the diffraction peak at 2θ of ~43° (200) could be ascribed to the cubic phase of nickel oxide (PDF 00-044-1159). This diffraction peak of nickel oxide was more apparent in the barium-promoted catalysts than in the barium-unpromoted Ni-YZr catalyst. Increased intensities of NiO diffraction peaks in the patterns of Ba-promoted catalysts might be attributable to a reduction in NiO dispersion over the support. The diffraction peak at 2θ of ~28° (110) could be attributed to the cubic phase of baria (PDF 00-001-0746). The addition of barium promoter shifted the diffraction peaks of YZr support slightly to a higher 2θ angle, that is, caused a slight reduction in the d-spacing parameter, implying the incorporation of barium into the lattice of the YZr support, as shown in Table 2.

Table 2 Shift in the 2θ angle and change in the d-spacing of (111) and (220) crystallographic planes of cubic yttria-stabilized zirconia phase for fresh catalysts.

Catalyst	Ba (wt%)	2θ (°)	d-Spacing for (111) (Å)	2θ (°)	d-Spacing for (220) (Å)
Ni-YZr	0.0	30.04010	2.97214	50.08210	1.81994
Ni-1Ba-YZr	1.0	30.05120	2.97124	50.09310	1.81951
Ni-2Ba-YZr	2.0	30.06228	2.97017	50.11239	1.81885
Ni-3Ba-YZr	3.0	30.06396	2.97001	50.11531	1.81875
Ni-4Ba-YZr	4.0	30.06540	2.96987	50.11782	1.81867
Ni-5Ba-YZr	5.0	30.06831	2.96959	50.12288	1.81850

Figure 3. XRD patterns of (A) the fresh catalysts and (B) the spent catalysts, where the black labels for yttria-stabilized cubic zirconia (JCPDS No. 49-1642), the green label for cubic baria (PDF 00-001-0746), and the blue label for cubic nickel oxide (PDF 00-044-1159). JCPDS, Joint Committee on Powder Diffraction Standards; XRD, X-ray diffraction.

Figure 3B presents the XRD patterns of the spent Ni-xBa-YZr (x = 0.0, 1.0, 2.0, 3.0, 4.0, 5.0) catalysts. The XRD patterns of Ni-YZr and Ba-promoted Ni-YZr catalysts showed diffraction peaks at 2θ of ~30°, ~35°, ~50°, ~60°, ~63°, ~74°, ~82°, ~84°, and ~94°, respectively, corresponding to the (111), (200), (220), (311), (222), (400), (331), (420), and (422) crystallographic planes of the cubic phase of yttria-stabilized zirconia (JCPDS No. 49-1642). The diffraction peak at 2θ of ~28° could be ascribed to the cubic phase of baria (PDF 00-001-0746). The intensity of this peak was reduced in comparison to that of the fresh catalysts, probably due to carbon deposition. The shifts in YZr peaks of the spent catalysts by incorporating barium as shown in Table 3 were higher than their corresponding shifts of the fresh catalysts. Table 3 displays the enhancement of barium incorporation into the YZr support after the reaction.

Table 3 Shift in the 2θ angle and change in the d-spacing (111) and (220) crystallographic planes of cubic yttria-stabilized zirconia phase for spent catalysts.

Catalyst	Ba (wt%)	2θ (°)	d-Spacing for (111) (Å)	2θ (°)	d-Spacing for (220) (Å)
Ni-YZr	0.0	30.15761	2.96100	50.27839	1.81324
Ni-1Ba-YZr	1.0	30.15761	2.96100	50.27839	1.81324
Ni-2Ba-YZr	2.0	30.10279	2.96627	50.18292	1.81646
Ni-3Ba-YZr	3.0	30.10658	2.96590	50.18953	1.81624
Ni-4Ba-YZr	4.0	30.07809	2.96865	50.13992	1.81792
Ni-5Ba-YZr	5.0	30.13207	2.96345	50.23391	1.81474

To determine the crystallite size, Scherrer's equation was utilized, which can be written as follows: [Image Omitted. See PDF]in which D_p is the crystallite size in nanometers, K is the shape factor which is 0.94, λ is the wavelength of X-ray and equals to 1.5406 Å, β is the full width at half maximum of the diffraction peak of the sample, and θ is the diffraction angle in degrees. The crystallite sizes of all fresh yttria-stabilized zirconia-supported Ni catalysts (Ba wt%: 0.0, 1.0, 2.0, 3.0, 4.0, 5.0) were determined by XRD using the most intense peaks at 30.06° and 50.17°. It was observed that crystallite sizes were identical for all fresh catalysts (17.2 nm), regardless of Ba loading, implying that the product of $$(\beta \times \cos \unicode{x0200A}\theta )$$ was constant (Equation 10).

The crystallite sizes of the spent catalysts were also determined by using the most intense peaks at 30.04° and 50.13°, where the crystallite sizes were the same for all spent catalysts (17.4 nm). The crystallite size was almost identical to that of the fresh catalyst, implying that barium inhibited the catalyst sintering, and hence, catalyst deactivation would be due to only carbon deposition.

From the TPR profiles shown in Figure 4, it was evident that catalyst reduction was primarily characterized by a broad peak in the range of 250−550°C. This characteristic was likely attributable to a moderate interaction between nickel oxide species and the support. However, adding barium enhanced nickel's reducibility, as evidenced by a larger reduction area relative to H₂ consumption. All the NiO particles were completely reduced at 700°C.

Figure 4. H2-TPR profiles of fresh Ni-xBa-YZr (x = 0.0, 1.0, 2.0, 3.0, 4.0, 5.0 wt%) catalysts. TCD, thermal conductivity detector; TPR, temperature programmed reduction.

SEM-EDX analysis was performed to examine the topology and distribution of catalyst components. Figure 5 shows the images obtained at various magnifications, along with the corresponding elemental mapping and surface content of various elements (see the table inserted).

Figure 5. SEM images for the promoted Ni-xBa-YZr (x = 0.0, 1.0, or 4.0 wt%) fresh catalysts. SEM, scanning electron microscopy.

The NiYZr catalyst appeared quite homogeneous, with agglomerated particles of size ranging between 100 and 300 nm. The surface concentration of each element is very close to the anticipated composition during the preparation phase. The addition of 1.0 wt% Ba did not significantly alter the topology of the catalyst, and Ba was uniformly distributed on the surface (see the corresponding elemental mapping image). The further addition of Ba (4.0 wt%) resulted in larger particles than those of Ni-YZr catalyst with nonuniform distribution of small Ba islands.

To verify the chemical composition of catalysts, also XRF analysis has been performed. The results, reported in Table S2, clearly indicated that the concentration of all elements was very close to that of the EDAX analysis in Figure S2.

A series of characterizations were conducted to justify the observed positive effect of adding Ba. Because CO₂ adsorption capacity plays a fundamental role in promoting CH₄ conversion in DMR, the effect of Ba on CO₂ adsorption capacity was the first parameter considered.¹⁵ Figure 6A demonstrates that the addition of Ba did not effectively promote the CO₂ adsorption at temperatures below 450°C when compared with the Ni-YZr catalyst without barium promoter. At temperatures higher than 700°C, however, massive CO₂ desorption was observed only for the catalyst promoted with 4.0 wt% of Ba, which explained the superior activity and stability of this catalyst. From a literature survey, it was found that the high-temperature CO₂ peak may be the result of the decomposition of barium carbonate at ~400°C, followed by probable CO₂ adsorption at high temperatures on strong basic sites, promoted by the presence of Ba.⁴⁷

Figure 6. (A) CO2-TPD profiles of different catalysts. (B) TGA profiles for the Ni-xBa-YZr (x = 0, 1.0, 2.0, 3.0, 4.0, 5.0 wt%) spent catalysts after TOS of 420 min in DRM, at 800°C and one atmosphere. DRM, dry reforming of methane; TGA, thermogravimetric analysis; TOS, time-on-stream; TPD, temperature-programmed desorption.

Whereas coke formation is one of the main problems in DRM, the characterization of spent catalysts was performed by TGA, TEM, Raman spectroscopy, and temperature-programmed oxidation (TPO) measurements.

As shown in Figure 6B, TGA analysis revealed that the highest coke content formed during the reaction was found on an unpromoted Ni-YZr catalyst. The progressive addition of Ba positively affected the coke formation. In particular, 3–7 wt% of coke was deposited on catalysts containing 4–5 wt% of BaO, whereas 15 wt% of coke was deposited on baria-unpromoted Ni-YZ catalyst. A comparison between coke formation and catalytic activity helps understand the smaller coke amount on the Ni-5Ba-YZr due to its lower activity with respect to Ni-4Ba-YZr. Overall, Ni-4Ba-YZr exhibited the highest resistance to coke formation, although it is the highest activity.

Regarding the topology of the formed coke, SEM images of the spent catalysts (see Figure 7A) showed the presence of carbon nanotube filaments deposited on the surfaces of the catalysts. EDX analysis of the Ni-2Ba-YZr used catalyst (Figure 7B) revealed that overall, the surface composition did not change significantly after the reaction. Unfortunately, the SEM investigation did not allow one to understand if different topologies of coke were formed during the reaction or if the filaments dimension changed on different catalysts.

Figure 7. (A) SEM images for the promoted Ni-xBa-YZr (x = 0, 1.0, 2.0, 3.0, 4.0, 5.0 wt%) spent catalysts and (B) The EDX spectrum of the spent Ni-2Ba-YZr catalyst. EDX, energy dispersive X-ray; SEM, scanning electron microscopy.

To obtain more information on the nature and dimension of coke formed during the reaction, TEM investigation was performed. TEM images of fresh Ni-YZr and Ni-4Ba-YZr catalysts are shown in Figure 8A. Both catalysts were characterized by irregular particle shape.

Figure 8. (A) TEM images of fresh catalysts and (B) spent catalysts. TEM, transmission electron microscopy.

The correspondent spent catalysts (see Figure 8B) clearly revealed the prevalent form of carbon filaments on both catalysts. However, as confirmed by TGA analysis, a large amount of coke was formed on Ni-YZr catalyst. Two different typologies of filaments were formed during the reaction: linear nanotubes 30–50 nm in size and irregular multiwalled nanotubes of similar dimensions. By considering the high temperature of the reaction (800°C), coke formation by the Boudouard reaction (2CO → CO₂ + C) can be considered negligible.⁴⁸ Therefore, methane decomposition occurring on metallic Ni surface would be the main source for coke formation. In this context, several studies have demonstrated that the growth of filaments strongly depends on the interaction of Ni with the carrier: the stronger the interaction is, the lower the probability of filaments growth is to detach the Ni particle from the catalyst surface.⁴⁹ In addition, the strong basicity of barium contributes to promoting the interaction of Ni with the YZr support, as reflected by the catalyst stability and the reduction of coke formation. To verify the catalyst stability, a long-time test over Ni-4Ba-YZr catalyst was performed. In addition, the Raman spectroscopy and TPO analysis (Figure S1) were performed to explore more information about carbon deposition. The stability test indicated that the catalyst was very stable because both CO₂ and CH₄ conversions remained almost constant for up to 50 h of TOS, as shown in Figure 9.

Figure 9. CH4 and CO2 conversions as a function of time-on-stream.

Raman analysis of spent catalysts confirmed that the nature of coke formed was mainly of graphitic nature. This result was confirmed by TPO analysis (see Figure S1), which showed a single combustion peak with a maximum at 500°C, indicating that, during the reaction, only filamentous coke was formed. The amount of formed coke was low by considering the reaction long time of 50 h.

Reaction mechanism

Zirconium oxide (ZrO₂) is an example of oxygen-deficient material. Thus, it can dissociate carbon dioxide into carbon monoxide and oxygen radical, as illustrated in Equation (11): [Image Omitted. See PDF] $${\square }_{\text{Zr}}$$ represents oxygen vacancy, while O_Zr refers to oxygen on the surface of ZrO₂ support.

In addition, carbon monoxide might be generated via the dissociation of bicarbonate intermediate, as shown in Equations (12) and (13): [Image Omitted. See PDF] [Image Omitted. See PDF] CO₂│_Zr represents adsorbed carbon dioxide and OH_Zr is hydroxyl species on ZrO₂ surface.

Yttria-stabilized ZrO₂ possesses higher oxygen deficiency and more basic sites, which improves the dissociation of carbon dioxide according to Equation (11).⁵⁰

Barium oxide is basic and interacts with carbon dioxide to form barium carbonate (BaCO₃), which in turn reacts with carbon deposition to produce carbon monoxide and barium oxide, as shown before in Equations (5) and (6).

In this scenario, barium is helpful to reduce the amount of carbon deposition on the surface of the catalysts during the DRM reaction course.^6,8,15,43,44

Furthermore, oxygen radical, generated by the decomposition of the adsorbed carbon dioxide over the oxygen vacancies of yttria-stabilized ZrO₂ support, takes part in the processes of regenerating metallic nickel sites by the reaction with the deposited carbon and adsorbed hydrogen atom, produced by the methane decomposition.⁵⁰

CONCLUSIONS

This paper examined the effect of Ba incorporation into nickel/yttria–zirconia catalyst for the DRM. Catalysts were prepared by incorporating various amounts of Ba to determine the optimal loading for enhancing the activity and stability of nickel/yttria–zirconia catalyst. The characterization results confirmed that the addition of Ba did not significantly alter the morphology in terms of BET and porosity. In contrast, the dominant effect of Ba was associated with its capacity to absorb CO₂ at high temperatures (700–800°C) and to improve catalyst reducibility. These observations, in our opinion, were the keys to justify the observed highest CH₄ and CO₂ conversions over the catalyst containing 4.0 wt% Ba. The presence of Ba improved the performance in terms of activity, inhibited catalyst sintering, and reduced coke formation, which had a significant positive effect on the stability of the catalyst.

ACKNOWLEDGMENTS

The authors would like to extend their sincere appreciation to the Researchers Supporting Project number (RSP2023R368), King Saud University, Riyadh, Saudi Arabia Dr. Ahmed I. Osman wishes to acknowledge the support of The Bryden Centre project (Project ID VA5048). The Bryden Centre project is supported by the European Union's INTERREG VA Programme, managed by the Special EU Programmes Body (SEUPB). The authors would like to thank Charlie Farrell for proofreading the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.