Effect of Support on Complete Hydrocarbon Oxidation over Pd-Based Catalysts

Abstract

Developing efficient strategies for VOC emission abatement is an urgent task for protection of the environment and human health. Complete catalytic oxidation exhibits advantages, making it an effective, environmentally friendly, and economically profitable approach for VOC elimination. Pd-based catalysts are known as highly active for hydrocarbon catalytic oxidation. The nature of carrier materials is of particular importance because it may affect activity by changing physicochemical properties of the palladium species. In this work, Al₂O₃, CeO₂, CeO₂-Al₂O₃, and Y-doped CeO₂-Al₂O₃ were used as carriers of palladium catalysts. Methane and benzene were selected as representatives of two types of hydrocarbons. A decisive step in complete methane oxidation is the first C–H bond breaking, while the extraordinary stability of the six-membered ring structure is a challenge in benzene oxidation. The support effect was explored by textural measurements using XRF, XRD, XPS, EPR, and TPR techniques. Three ceria-containing samples showed superior CH₄ oxidation performance, achieving 90% methane conversion at about 300 °C and complete oxidation at 320 °C. Evidence for presence of Pd²⁺ species in all samples regarded as most active was provided by XP-derived analysis. Pd/Y-Ce/Al catalysts exhibited very high activity in benzene oxidation by reaching 100% conversion at 180 °C. The contributions of higher Pd and Ce³⁺ surface concentrations, the presence of O₂⁻-adsorbed superoxo species, and Pd⁰ ↔ PdO redox transfer were considered. The potential of a simple, environmentally friendly, and less energy demanding mechanochemical preparation procedure of mixed oxides was demonstrated.

1. Introduction

Nowadays, atmospheric pollution represents a serious concern with respect to human health and ecosystems. It is closely related to climate change and some effects of global warming that have occurred faster than scientists have predicted [1,2]. Volatile organic compounds (VOCs) are an important class of air pollutants in urban and industrial areas. Some of the main sources of VOC emissions are automobile exhaust, industrial and energetic processes, combustion by-products, and residential activities [3,4,5]. Among VOCs, benzene, toluene, ethylbenzene, and xylene belong to a group of compounds known as BTEX and are regarded as the most common aromatic VOCs, with significant contributions to industrial emissions. Using natural gas, primarily composed of methane, is an effective method to reduce the concentration of pollutants emitted by gasoline- or diesel-fueled vehicles and stationary power generation facilities. However, the release of unburned methane from vehicles and gas power plants, as well as leakage during production and transportation, is considered a major drawback due to methane’s stronger detrimental greenhouse effect compared to CO₂ [6]. In December 2019, the European Commission adopted the so-called Green Deal, also known as the Green Pact, which is a package of policies aimed at making Europe climate-neutral by 2050. EC strategies and initiatives target the reduction of greenhouse gas production at least by 50% compared to 1990 levels, as well as zero pollution by 2050 with respect to air, soil, and water. Horizon Europe, a key seven-year EU funding program for research and innovation, was established for this ambitious purpose for the period of 2021–2027. Climate technologies will receive 35% of the funding amount for Horizon Europe. In this context, VOC emissions abatement is a particularly pertinent problem for future research.

Catalytic oxidation is the most efficient method for VOC and methane abatement in exhaust gases. It is widely used, owing to some important advantages, such as high catalytic activity and treatment efficiency, low energy consumption, and a lack of secondary pollutants [7,8,9]. A significant number of research studies have been devoted to catalyst design and composition optimization because of the decisive role of the catalyst in achieving high efficiency in environmental pollutants abatement. Some recent reviews highlight the effectiveness of Pd-based catalysts, which are recognized as highly active for the catalytic oxidation of hydrocarbons [10,11,12]. Along with Pd dispersion and the oxidation state, the physicochemical properties of the support significantly affect catalyst performance. The selection of a support is of particular importance because in many cases, the metal oxide plays ab active part in the reaction mechanism instead of being merely a carrier of the active phase. One consolidated strategy to develop cost-effective and high-performing catalyst formulations is the use of appropriate and economically viable supports. Alumina is widely applied for catalyst support in industry, owing to its low cost, high specific surface area, high thermal stability, and mechanical strength. An effective approach to stabilize PdO against decomposition, as well as to diminish Pd content and increase PdO surface exposure, is alumina modification by metal oxides [11]. Ceria has a high oxygen storage and release capacity due to its fast Ce⁴⁺/Ce³⁺ redox cycle, motivating its use for the preparation of Ce-Al mixed oxide supports. Different compositions and synthesis methods have been reported. Wet impregnation of alumina with 1, 5, and 10 wt.% Ce favored coexistence between supported Pd⁰ and PdO species, which improved benzene oxidation on Pd/Ce-Al₂O₃ [13]. Ramírez-López et al. employed the sol–gel method for the preparation of alumina modified by 2 to 50 wt.% CeO₂ [14]. Palladium-containing samples with 5, 10, and 15 wt.% ceria achieved 100% methane conversion at temperatures below 550 °C. Alumina consecutive wetness impregnation by changing the CeO₂ (17.5 wt.%) and Pd (1.5 wt.%) loading sequences affected palladium dispersion, influencing catalytic activity [15]. Samples with Ce added over Pd and Al₂O₃ demonstrated higher propane, NO, and CO conversion. Fertal et al. reported opposite findings by studying the impact of sequential impregnation of PdO/CeO₂/Al₂O₃ catalysts with 4 wt.% Pd and 20 wt.% CeO₂ on methane combustion [16]. The lower activity of samples prepared by impregnation of cerium nitrate on Pd/Al₂O₃ was attributed to strong PdO–ceria interactions and diminished access of methane and oxygen to active PdO sites. Commercial γ-alumina-supported Pd/CeO₂ colloidal assembled spheres with higher amounts of interfacial Pd−O−Ce species and O_ads concentration than Pd/CeO₂ catalysts exhibited enhanced CH₄ combustion activity [17]. Recently, the effect of yttria modification on the activity of ceria-supported gold catalysts in complete benzene oxidation (CBO) was studied [18]. The authors reported that ceria doped with 1 wt.% Y₂O₃ demonstrated the best performance, which was attributed to the role of support in the benzene activation stage, along with enhanced surface reoxidation. Later, bearing in mind the suitability of Pd-based catalysts for low-temperature VOC oxidation, the same authors reported that the addition of 1 wt.% Y₂O₃ to ceria produced an active and stable Pd-based catalyst for benzene oxidation due to an increased oxygen vacancy concentration and improved Pd dispersion [19]. However, the effect of Y doping of ceria on the activity of supported palladium catalysts in methane oxidation has not been examined. Yttrium ionic radius is suitable for insertion into ceria lattice with 0.97 Å for Ce⁴⁺ and 1.02 Å for Y³⁺ in eight-fold coordination. Moreover, the oxidation state of yttrium ions is appropriate for the formation of oxygen vacancies in order to compensate for the negative charge induced by Ce⁴⁺→ Y³⁺ substitution. Along with yttrium abundance, these two features motivated the study of Y-doped, CeO₂-Al₂O₃-supported, Pd-based catalysts.

The aim of this work is to explore the role of support for complete hydrocarbon oxidation over Pd-based catalysts. Methane and benzene were selected as representatives of two types of very stable hydrocarbon molecules. A decisive step in complete methane oxidation is the first C–H bond breaking because of its higher activation energy than C–C bonds [20], while the extraordinary stability of the six-membered ring structure is a challenge in benzene oxidation. A cost reduction strategy for catalyst production stimulated palladium deposition on γ-alumina modified by well surface-dispersed ceria and Y-doped ceria using mechanochemical mixing. This approach enabled the achievement of a higher interfacial area between the palladium nanoparticles and alumina-supported ceria. It is not a commonly applied routine method and could be considered a novelty. From an economic point of view, modifying the alumina surface with highly dispersed ceria provided further advantages—namely, it decreased the amount of ceria in the catalyst formulation as compared to Pd/CeO₂. To the best of our knowledge, palladium promotion on CeO₂-Al₂O₃ and Y-doped CeO₂-Al₂O₃ mixed oxides prepared by simple, less polluting, and energy-saving mechanochemical methods has not been studied in the benzene and methane oxidation reactions.

2. Results

2.1. Catalytic Activity Measurements

Four metal oxides, namely Al₂O₃, CeO₂, CeO₂-Al₂O₃, and Y-doped CeO₂-Al₂O₃, were selected as carriers of palladium catalysts. The notations of the various samples are described in the Experimental section. Methane conversion as a function of the reaction temperature is presented in Figure 1a. The calculated value for the standard deviation was ±1.5% based on the average of six measurements at each experimental point. Two successive measurements were necessary to determine average values for the conversion degree. All samples demonstrated a progressive activity increase with increasing temperature. However, a favorable role of ceria as support was observed. Palladium on ceria and alumina-supported ceria, namely Pd/CeO₂, Pd/CeAl, and Pd/Y-CeAl, behaved significantly better as compared to Pd/Al₂O₃. In fact, Pd/CeO₂ exhibited higher conversions in a lower temperature range. The three ceria-containing samples reached about 90% CH₄ conversion at 300 °C and complete oxidation at 320 °C vs. 60% at 300 °C and 100% at 390 °C with Pd/Al₂O₃. A comparison of “light off” profiles demonstrated that yttrium-doped ceria had no effect on oxidation activity. Considering economic profitability, the best trade-off was achieved with the Pd/CeAl sample, which showed comparable activity for lower ceria content.

Table S1 presents comparative data about Pd-based catalysts used for complete methane oxidation. As can be seen, the catalysts developed in this work demonstrated high activity, achieving some of the lowest temperatures, such as 300 °C for 90% conversion. However, an assessment of the best-performing catalyst could not be conducted because of the diversity of catalyst compositions and experimental reaction conditions.

Long-term stability tests were not carried out; however, the establishment of steady-state activity for a certain reaction temperature took 1–2 h, while catalytic measurement of each sample continued for 3 days. Finally, the conversion remained complete within 2 h, and no activity decline was observed, indicating the stability of the catalysts.

Benzene oxidation activity is expressed as the benzene conversion degree. A quite different activity order was observed for benzene oxidation over the studied samples (Figure 1b). Contrary to the above-described results, Pd/Y-CeAl exhibited superior performance, achieving complete conversion at 180 °C. Products of incomplete oxidation were not registered within 24 h stability tests at this temperature. Preliminary measurements with some expected intermediate products of partial oxidation such as benzoquinone, phenol, maleic anhydride, and hydroxy-1,4-benzoquinone were carried out. These compounds were injected as evidence to detect the retention time under identical GC conditions. Pd/CeAl also showed high oxidation activity, achieving 96% conversion at 175 °C and complete conversion at 300 °C. The role of yttrium-doped ceria was clearly revealed by comparing the benzene conversion degree at 180 °C. An interesting finding was that Pd/Al₂O₃ performed better than Pd deposition on reducible CeO₂. A conversion of 85% was reached at 100 °C. However, it is worth noting that neither sample was able to achieve complete conversion, even at 300 °C. Table S2 presents a summary of the composition and catalytic performance of the Pd-based catalysts prepared in this work and reference literature. Pd/Y-CeAl significantly outperformed all catalysts, achieving the lowest T₉₀ of 100 °C. However, as noted above, catalyst composition and experimental conditions were quite different, and a reliable conclusion could not be made. The table also reports data about catalysts evaluated using the same experimental setup. The Pd/Y-CeAl catalyst developed in this work demonstrated superior performance. A 90% conversion at 180 °C and a beneficial impact of Pd deposition on Fe-doped ceria as compared with a matching Au-based sample (T₉₀ = 200 °C) were observed [21]. Both catalysts showed a lower T₉₀ relative to that of Pt loaded on Fe-modified fly ash zeolite X with T₉₀ = 250 °C [22].

2.2. Catalyst Characterization

2.2.1. Chemical Composition

The actual chemical compositions of the samples determined by X-ray fluorescence spectrometry (XRF) are reported in

Table 1. Catalyst compositions measured by XRF.

Sample	Chemical Composition (wt.%)
	Pd	CeO2	Y2O3	Al2O3
Pd/Al2O3	1.03	-	-	98.97
Pd/CeO2	1.10	98.90	-	-
Pd/CeAl	1.06	29.68	-	69.26
Pd/Y-CeAl	1.01	29.40	0.29	69.29

. A homogenous structure was verified by similar values measured at three spots. Analysis indicated values very close to the target values for all components, revealing the suitability of mechanochemical mixing for sample preparation with preset loading. In all catalysts, the palladium content was very close to 1 wt.%, which confirmed that palladium deposition–precipitation was carried out successfully.

2.2.2. Textural Characteristics

Textural features of the samples, namely specific surface area (S_BET), pore volume (V_pore), and average pore size (D_pore), are listed in

Table 2. Textural characteristics, average size (D_ceria), and lattice parameter (α) of ceria estimated by XRD.

Sample	SBET (m2 g−1)	Vpore (cm3 g−1)	Dpore (nm)	Dceria (nm)	αceria (nm)
γ-Al2O3	200.0	0.53	10.6	-	-
Pd/Al2O3	195.8	0.52	9.7	-	-
CeO2	63.0	0.29	20.3	6.3	0.5424 (2)
Pd/CeO2	67.6	0.28	16.5	6.2	0.5424 (4)
Ce/Al	165.0	0.39	9.5	6.1	0.5422 (1)
Pd/CeAl	161.2	0.45	9.5	5.8	0.5424 (1)
Y-Ce/Al	168.0	0.43	10.0	6.1	0.5421 (2)
Pd/Y-CeAl	166.4	0.42	9.6	5.8	0.5420 (1)

. Alumina-based samples are mesoporous materials with S_BET values close to those of γ-Al₂O₃ and pore sizes around 10 nm. In the case of modified alumina, namely Ce/Al and Y-doped Ce/Al, the S_BET and total pore volume showed a slight decrease that could be attributed to alumina mixing with ceria with a lower specific surface area, i.e., 63 m² g⁻¹. The same trend in S_BET values was reported by Yang et al., who reported values of 144 and 46 m² g⁻¹ and activity at T₉₀ = 410 and 535 °C for Pd/CeO₂/Al₂O₃ and Pd/CeO₂ catalysts, respectively [17]. The preparation of a 4Pd-20CeO₂/Al₂O₃ sample by impregnating Pd(NO₃)₂ aqueous solution to an aqueous slurry of 20CeO₂/Al₂O₃ produced a catalytic material with an S_BET value of 117 m² g⁻¹ [16].

The textural parameters of Pd-containing samples remained close to those of the used carriers. Recently, the textural properties of CeO₂–Al₂O₃ mixed oxides prepared by impregnation and mechanochemical mixing were studied [23]. A nearly two-fold lowering of S_BET and the pore volume of impregnated samples was observed in comparison with bare γ-Al₂O₃, which was attributed to the deep penetration of Ce³⁺ ions into alumina pores. The formation of ceria crystallites inside the pores was found to be responsible for the limited access of reactants to active sites and the lower catalytic activity. In the present case, employment of mechanochemical mixing is favorable for preserving beneficial textural features of alumina.

2.2.3. X-Ray Powder Diffraction (XRD) Measurements

XRD analysis was carried out for phase composition identification of catalysts. The XRD patterns of γ-Al₂O₃, CeO₂, and supported Pd samples are shown in Figure 2. The positions and intensities of the peaks are very similar. Clearly visible diffraction lines at 2θ of 28.5°, 33.1°, 47.5°, and 56.4° were attributed to the (111), (200), (220), and (311) planes of ceria phase in the cubic crystal structure of the fluorite type, respectively. A weak contribution of Al₂O₃ was distinguishable at about 2θ = 38°, as well as a ceria peak shoulder at 2θ = 47.5°. No diffraction peaks ascribed to the presence of yttria or any palladium-related phases were discernible. This finding may be associated with small loadings, suggesting high dispersion.

The average size of ceria (D_ceria) estimated according to Scherer’s equation using the FWHM of the peak at 2θ of 28.5° and the unit cell parameter (α_ceria) are reported in Table 2. All samples show very similar values. Analysis indicated that Pd deposition did not affect the particle size or the lattice parameter. Concerning yttrium doping, changes in the ceria unit cell parameter upon replacement of Ce with Y ions was not expected because of the similarity of the ionic radii: 0.97 Å for Ce⁴⁺ and 1.02 Å for Y³⁺ in eight-fold coordination.

2.2.4. X-Ray Photoelectron Spectroscopy (XPS) Measurements

The oxidation state at the surface of all samples was studied by XPS.

Table 3. X-ray photoelectron spectral analysis.

Catalyst	Ce 3d5/2 (eV)	Pd 3d5/2 (eV)	Y 3d5/2 (eV)	Y/Pd	Pd/(Ce + Al)	Ce3+/(Ce3+ + Ce4+)
Pd/Al2O3	-	336.7	-	-	0.017 0.048 *	-
Pd/CeO2	881.7	337.1	-	-	0.140 0.0163 *	0.32
Pd/CeAl	881.8	336.6	-	-	0.044 0.0062 *	0.33
Pd/Y-CeAl	881.8	336.7	157.8	0.21 0.28 *	0.041 0.0063 *	0.40

* Analytical ratio.

shows the results of quantitative XP spectral analysis. Al 2p binding energy (BE) is not listed because the peak was located at 75.1 ± 0.1 eV in the spectra of all samples. The BE of the Pd 3d_5/2 peak in the range of 336.6–337.1 eV indicated the presence of Pd²⁺ in all palladium-containing samples (Figure 3). Higher values of XPS-derived data on the Pd/(Ce + Al) atomic ratio, namely 0.044 and 0.041 for Pd/CeAl and Pd/Y-CeAl, respectively, as compared to the analytical value of 0.006 revealed a higher palladium dispersion on the surfaces of these samples. The highest atomic ratio of 0.14 calculated for Pd/CeO₂ reflects a lower ceria surface area, maximizing the Pd photoelectrons with respect to support-related photoelectrons. This value agrees well with the role of ceria in favoring a high dispersion of the supported noble metals.

Ceria oxidation states, namely Ce⁴⁺ and Ce³⁺, were detected in the Ce 3d spectra of all Ce-containing samples (Figure S1). Based on the consensus about the relationship between Ce³⁺ concentration and oxygen vacancies formed for charge compensation, special attention was paid to the Ce³⁺/(Ce⁴⁺ + Ce³⁺) atomic ratio. The values were in the interval of 0.25–0.38, i.e., higher than the typically calculated values of 0.06–0.10 for bulk ceria [24], highlighting the existence of a defective ceria structure with the formation of oxygen vacancies on the catalyst surface. The highest ratio values were calculated with the best-performing samples.

The Y 3d_5/2 peak was located at a BE of 157.8 eV and assigned to Y³⁺ ions. The calculated Y/Pd ratio of 0.21 provided important information, as it was very close to the analytical value (0.28), implying a probability of yttrium to occupy sites close to the palladium atoms. The contribution of Y doping to ceria defectivity was demonstrated. The effect was clearly visible when comparing Ce³⁺/(Ce⁴⁺ + Ce₃₊) values for Pd/Y-CeAl (0.38) and undoped Pd/CeAl (0.25).

2.2.5. Electron Paramagnetic Resonance (EPR) Measurements

EPR spectra of fresh Pd/CeO₂, Pd/CeAl, Pd/Y-CeAl, and Pd/Al₂O₃ samples are shown in Figure 4. There are two paramagnetic ionic states of palladium, namely Pd³⁺ and Pd⁺ with 4d⁷ and 4d⁹ configurations, respectively. Usually, the Pd³⁺ ions are produced upon oxidation of Pb-containing catalysts at high temperatures.

Table 4. Signal g values and identified paramagnetic species in fresh samples.

Sample	Chemical Composition (wt.%)
	Pd3+	Pd+	Pdx+−O2	O2−	Ce3+	Ce4+−O2−
Pd/Al2O3			2.0538 1.9974 1.9609	2.0812 2.0661 2.0270	-	-
Pd/CeO2	2.4558, 2.113, 2.0621, 1.9759	2.3485, 2.3050, 2.2405, 2.1872	2.0575		1.9575, 1.9395	2.0482, 2.0437, 2.0330, 2.0048
Pd/CeAl	1.9759				1.9576, 1.9383	2.0638, 2.0529, 2.0332, 2.0227, 2.0064
Pd/Y-Ce/Al	1.9770		2.0557		1.9573, 1.9379	2.0640, 2.0343, 2.0067

summarizes the g values of paramagnetic species in the different investigated samples. Several lines were observed in the EPR spectrum of Pd/CeO₂. Several low-intensity lines with g values ranging from 2.4558 to 2.1130 were recorded in the low-frequency part of the spectrum. These lines were attributed to different species of palladium. EPR lines at g values of 1.9759, 2.2405, and 2.4558 are due to Pd³⁺ ions of rhombic symmetry [25]. Signals with g values of 1.9759 and 1.977 were also recorded in the spectra of Pd/CeAl and Pd/Y-CeAl, respectively. Most probably, these signals can be ascribed to Pd³⁺ ions of different symmetries because the lines observed at a lower magnetic field in the Pd/CeO₂ spectrum with g factors of 2.2405 and 2.4558 were not registered.

According to the literature, two signals of axial symmetry at g_II = 2.3050 and g_⊥ = 2.1872 and g_II = 2.3485 and g_⊥ = 2.0482 in the Pd/CeO₂ sample originated from Pd⁺ ions [25]. The line at a g factor of 2.1130 is also related to the presence of Pd⁺. In the central part of the spectrum, several lines with g values ranging from 2.0621 to 2.0048 were recorded. These signals confirmed the presence of adsorbed superoxo species (O₂⁻), with one unpaired electron. Oxygen species can be adsorbed onto Ce⁴⁺ ions or can fill oxygen vacancies, thereby forming surroundings of different symmetries and resulting in different EPR signals [26,27]. Values of g_x for O₂⁻ are greater than those of the electron g_e factor. The difference is due to the formation of a covalent bond via the overlapping of the oxygen π_y orbital with the 4f orbital of the Ce cation, which has a greater spin orbital coupling constant λ. This kind of oxygen species was found in all spectra.

A signal at g_II of about 1.9573 and g_⊥ of about 1.9379 was found in all ceria-containing samples; exact values are reported in Table 4. According to Soria et al. [28], this signal could be assigned to Ce³⁺ ions associated with anion vacancies; however, it might also originate from F⁺ centers [29]. The Ce³⁺ signal intensity was lowest in the Pd/CeO₂ spectrum. It increased with the Pd/CeAl sample and was the most intense for Pd/Y-CeAl. Therefore, Y doping caused an increase in the amount of Ce³⁺ ions. Ceria entities in alumina support might be associated with more dispersed Ce particles at the surface, while O₂⁻–Ce⁴⁺ species are present as less dispersed or aggregate systems. EPR signals with g values of 2.0575, 2.0529, 2.0557, and 2.0538 were observed for Pd/CeO₂, Pd/CeAl, Pd/Y-CeAl, and Pd/Al₂O₃, respectively. Most probably, the signal is related to some palladium species because it is present in all spectra. According to available information in the literature, Pd^x+–O adducts yield g_ll = 1.99, 1.964 and g_⊥ = 2.05, 2.055 [30]. Therefore, these signals may arise from Pd(I)-O₂ adducts, especially in the Pd/Al₂O₃ sample, where, besides g = 2.0538, signals at g = 1.9974 and g = 1.9609 were also recorded.

Variation of temperature upon recording did not change the shapes of EPR spectra. EPR spectra of Pd/CeO₂ recorded at different temperatures are displayed in Figure 5. A temperature decrease caused only an increase in line intensity. The same behavior was observed with the other samples.

EPR spectra of Pd/CeO₂, Pd/CeAl, Pd/Y-CeAl, and Pd/Al₂O₃ after methane oxidation reactions are shown in Figure 6. As can be seen, the spectral shapes are almost the same as those of fresh samples. Small changes of the g values were registered, owing to changes in the surrounding atoms during reaction; however, the paramagnetic species causing these changes are the same. Line broadening of the signals positioned at g 2.0529 and 2.0557 in the spectra of Pd/CeAl and Pd/Y-CeAl, respectively, was observed. This finding implies the formation of palladium magnetic domains of slightly different sizes and quantities attributed to partial particle agglomeration. As a whole, the line intensity of the spectra related to oxygen species decreased, which indicates its depletion during reaction. In the EPR spectra of the fresh Pd/Al₂O₃ sample, the line recorded at a g factor of 1.9609 (Pd^x+–O₂⁻ adducts) disappeared, and a new one with a g factor of 1.9760 (Pd³⁺) was observed. This is another piece of evidence for oxygen species participating in the methane oxidation reaction. An obvious decrease in Ce³⁺ signal intensity in the Pd/Y-CeAl spectrum was also noticed.

EPR measurements were also performed on selected samples after CBO tests to shed more light on catalyst stability. The purpose was to detect coke formation on the catalyst surface. Figure 7 shows the spectrum of Pd/CeO₂ because this sample exhibited lower CBO activity in comparison with the other samples. A single EPR line at a g factor of 2.003, which is typical of a carbon-centered radical and coke, was recorded. The EPR spectrum of the best-performing Pd/Y-CeAl catalyst comprises another line at a g value of 2.0023 related to carbon-centered radicals. Two EPR lines with g values of 2.0249 and 2.0054 were also registered in the spectrum, revealing the presence of oxygen species. A well-defined line at a magnetic field of 334–335 mT with a g value of 1.964, assigned to Ce³⁺, was also observed, focusing on the role of Ce³⁺ for oxygen activation [31]. It should be underlined that such a line was not detected in the Pd/CeO₂ spectrum. This observation highlights a beneficial role of mechanochemical mixing in modifying commercial alumina using highly dispersed ceria nanoparticles.

2.2.6. Temperature-Programmed Reduction (H₂-TPR) Measurements

Catalyst reducibility was evaluated by TPR with hydrogen. Sample TPR profiles are presented in Figure 8. Hydrogen consumption below 100 °C was observed for all catalysts and attributed to PdO reduction (PdO→Pd⁰). This is in agreement with XPS analysis showing palladium presence as Pd²⁺. The peak is centered at 21 °C, with Pd/Al₂O₃ and shifts to higher values in the profiles of other catalysts indicating a stronger interaction between palladium oxide and the support [6].

TPR profiles of ceria-containing samples contain two peaks. Many works have demonstrated that ceria reduction proceeds in two steps: surface layer reduction at around 500 °C and bulk reduction over 800 °C [32]. Noble metals were reported to facilitate ceria surface layer reduction and boost the first reduction peak at much lower temperatures [33]. Narrow peaks in the temperature range of 44–60 °C included a PdO ↔ Pd⁰ redox transfer and oxygen reduction from ceria surface layers in close vicinity to Pd particles. Broad peaks in the temperature range of 300–500 °C could be related to further ceria surface reduction. The lowest peak intensity was registered for the Pd/CeAl catalyst. A favorable effect of Y doping on ceria surface oxygen mobility and improved reducibility of Pd/Y-CeAl was demonstrated. The highest intensity of the peak with T_max at 44 °C in the profile of Pd/CeO₂ could be reasonably explained by considering the highest ceria content. A very weak negative peak at 90 °C in the profile of Pd/Al₂O₃ was attributed to decomposition of Pd β-hydride [34]. Small features related to such decomposition could also be detected at 80−90 °C in the other catalysts; however, their intensity and temperature depend on the catalyst composition and, therefore, on different properties.

Catalyst reducibility was also analyzed based on experimental hydrogen consumption (HC). Values up to 300 °C were calculated because of their relevance to the present study (

Table 5. Experimental H₂ consumption during TPR up to 300 °C.

Catalyst	HC (mmol·g−1)
Pd/Al2O3	0.09
Pd/CeO2	0.43
Pd/CeAl	0.23
Pd/Y-CeAl	0.29

). The theoretically calculated HC for PdO reduction was 0.09 mmol·g⁻¹, considering that 1 wt.% Pd matches 1.15 wt.% PdO. After subtraction of this value, HC values for ceria surface layer reduction were 0.34, 0.14, and 0.20 mmol·g⁻¹ for Pd/CeO₂, Pd/CeAl, and Pd/Y-CeAl, respectively. In view of the ceria content, the resulting reduction degree was 12, 16, and 23%, accordingly. As already commented above, low-temperature TPR peaks appeared up to 100 °C, and ceria surface layer reduction occurred only close to Pd⁰ particles, while Ce⁴⁺→ Ce³⁺ transition for ions far from metallic palladium particles proceeded at 300–500 °C. A similar finding was reported by Gil et al. [35], who assigned a low-temperature TPR peak at 50 °C to Pd-affected ceria surface reduction, followed by further surface reduction above 300 °C.

3. Discussion

The chemical state of palladium is often discussed, owing to its critical role in catalytic methane oxidation activity. Many works have reported PdO as the most active phase and highlighted a favorable correlation between PdO content and catalyst performance [11,36]. However, along with palladium dispersion and oxidation state, the nature of support is of particular importance because it could affect activity by changing the physicochemical properties of palladium species [37]. Cooperation between palladium nanoparticles and the carrier during the catalytic reaction has often been claimed to explain variations in terms of catalytic performance of Pd-based catalysts using different supports. A rational strategy for synthesis of well-performing and economically profitable catalytic materials is based on the exploration of appropriate and cost-effective carriers. In general, a large specific surface area and an abundance of active sites are considered beneficial for high levels of activity of heterogeneous catalysts. Recently, our own studies revealed the suitability of mechanical mixing for the preparation of alumina-supported ceria or Y-doped ceria. In this work, Al₂O₃, CeO₂, CeO₂-Al₂O₃, and Y-doped CeO₂-Al₂O₃ were used as support for palladium catalysts. As shown in Figure 1, three ceria-containing samples showed superior methane oxidation performance, achieving T₉₀ at about 300 °C and complete oxidation at 320 °C. Recently, Colussi et al. reviewed advances with respect to the relationship between structural features and reactivity in Pd/CeO₂ catalysts for low-temperature methane combustion [38]. The review focused on the creation of specific structural arrangements, owing to the high degree of nanoscale interaction between Pd/PdO and ceria. A combination of Pd with different oxidation states, such as Pd and PdO, and ionic entities (PdO_x, Pd²⁺) was indicated to provide more active sites than bulk PdO. This finding corroborates the clearly high combustion activity exhibited by ceria-containing samples. Evidence for the presence of Pd²⁺ in all palladium-based samples was provided by XP-derived analysis (Table 3, Figure 3). Higher values of the Pd/(Ce + Al) atomic ratio calculated for Pd/CeAl and Pd/Y-CeAl imply a higher dispersion of palladium on the surface of these samples. Two paramagnetic ionic states of palladium, namely Pd³⁺ and Pd⁺, as well as signals attributed to adsorbed superoxo species (O₂⁻) with one unpaired electron, were registered by EPR, confirming the role of ceria in promoting oxygen mobility using Pd ions. The existence of a defective ceria structure with oxygen vacancies forming on the catalyst surface, as revealed by XP and EPR spectral analysis, contributed to significantly higher activity as well. It should be noted that the reciprocal redox behavior of Pd and ceria resulted in the promotion of ceria surface oxygen reduction and enhanced Pd oxidation by ceria. Ceria’s unique oxygen exchange properties were considered to explain PdO stabilization because of ceria’s ability to hinder PdO reduction and promote Pd reoxidation [11,38] (and references therein). Mechanochemical mixing allowed for the maximization of ceria dispersion on the alumina surface, achieving both a larger interfacial area between deposited Pd nanoparticles and alumina-supported mixed oxides and enhanced oxygen mobility. This close contact favors the generation and movement of active oxygen species on the surface, which are substantial in keeping Pd species in an active oxidized state, thereby facilitating effective CH₄ activation. Chen et al. elucidated active sites for CH₄ oxidation over Pd/CeO₂ and observed CH₄ decomposition over Pd²⁺ in the absence of gas-phase oxygen [39]. Through density functional theory calculations, they revealed the excellent activity of coordinatively unsaturated Pd²⁺ ions and adjacent oxygen atoms in C−H bond breaking. Analysis of the EPR spectra of samples tested in CH₄ oxidation indicated decreased Ce³⁺ signal intensity in the Pd/Y-CeAl spectrum. As described in Section 2.1, Y doping of ceria had no effect on methane conversion. Despite the favorable effect of Y doping on oxygen supply from ceria observed in fresh samples, a possible explanation could be the insignificant contribution of yttrium to oxygen mobility and the formation of vacancies under conditions of CH₄ oxidation.

Although alumina is an inert support and electronic interactions with Pd nanoparticles were not expected, high dispersion of palladium resulted in good performance. It could be hypothesized that lower activity was caused by the deactivation of the active phase, owing to water generated during reaction. As shown in Figure 1a, the T₅₀ of Pd/Al₂O₃ was 30 degrees lower than that of Pd/CeO₂, while differences in T₁₀₀ reached 70 degrees.

Some divergences exist in the literature about the role of the palladium oxidation state in benzene oxidation. Some authors have reported that metallic palladium particles enable the formation of more active sites for benzene oxidation relative to PdO [40]. According to Padilla et al., the addition of ceria to Pd/γ-Al₂O₃ resulted in the coexistence of PdO species and Pd⁰, acting as active sites for benzene oxidation [13]. In contrast to the high CH₄ oxidation activity of all ceria-containing samples, Pd/Al₂O₃ demonstrated higher CBO activity than Pd deposited on reducible CeO₂. Highly dispersed Pd on Al₂O₃ achieved 85% conversion at 100 °C. An explanation could be PdO’s ability to provide active oxygen.

The CBO performance of Pd/CeAl and Pd/Y-CeAl resulted from the contribution of both Pd located on ceria or a Y-doped ceria phase and Pd dispersed on alumina. In studying benzene oxidation over zeolite-supported Pd catalysts, He et al. have implied Pd²⁺O²⁻ reduction by benzene, followed by metallic Pd oxidation with gas-phase oxygen and recovery of the Pd²⁺O²⁻ species [41]. Such a redox process using gas-phase oxygen is relevant for irreducible supports like alumina; however, in the case of ceria as a reducible support, ceria surface oxygen could be involved in the Pd⁰ ↔ PdO redox transfer. The beneficial impact of Y doping on oxygen supply from ceria was disclosed by a higher XPS-derived Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratio (Table 3) and a higher Ce³⁺ signal intensity in the EPR spectra (Figure 4). Differences in CBO activity could be explained by the role of both palladium and the support for benzene and oxygen activation. Huang et al. reported that the high performance of Pd-based catalysts in VOC oxidation can be attributed to PdO’s ability to provide oxygen and the participation of metallic Pd sites in VOC decomposition [3] (and references therein). Benzene adsorption occurs at the Pd surface through π bonds and back donation from the metal to the π* hydrocarbon orbitals [33].

An explanation of the best performance of Pd-based samples on alumina-supported ceria, particularly on alumina-supported Y-doped ceria, should be based not only on the abovementioned redox Pd⁰ ↔ PdO transfer. Higher surface concentration of Pd and Ce³⁺, evidenced by XPS analysis, as well as presence of adsorbed O₂⁻ superoxo species registered in the EPR spectra should also be considered. Moreover, reflections due to the presence of PdO were not detected in XRD patterns of all Pd-containing samples, indicating high dispersion. The role of reducible oxide in improving combustion activity is also worth noting. In spite of the high benzene conversion degree over Pd/Al₂O₃, this catalyst did not achieve complete oxidation of benzene, even at 300 °C. Analyzing the behavior of Pd-based samples on ceria, EPR spectra indicated the formation of coke during the catalytic tests that also negatively affected total benzene conversion. The modification of an irreducible oxide surface using reducible metal oxide appeared favorable in the present case. The advantages of mechanochemical preparation of mixed oxides as carriers of Pd-based catalysts for complete hydrocarbon oxidation as an environmentally friendly and energy-saving synthesis approach were demonstrated.

4. Materials and Methods

4.1. Materials

Commercial γ-Al₂O₃ (Sasol, Brunsbüttel, Germany) was used as support. All reagents used in this study are products of Sigma-Aldrich^® solutions (Steinheim, Germany), namely Ce(NO₃)₃·6H₂O (cerium nitrate hexahydrate, ≥99.0%), Y₂O₃ (yttrium oxide, ≥99.99%), Pd(NO₃)₂·2H₂O (palladium nitrate dehydrate, ~40% Pd basis), and K₂CO₃ (potassium carbonate, ≥99%).

4.2. Catalysts Preparation

Commercial γ-Al₂O₃ and lab-prepared cerium hydroxide were used for CeO₂-Al₂O₃ synthesis. Precipitation of Ce(NO₃)₃·6H₂O with K₂CO₃ was carried out at 60 °C and a constant of pH = 9.0, followed by aging at the same temperature for 1 h, washing, and drying at 100 °C. Mechanochemical mixing was performed in an electric mortar (Mortar Grinder RM 200, Retsch GmbH, Haan, Germany) by grinding for 1 min of alumina and calculated Ce(OH)₃ amount to provide 30 wt.% CeO₂ in Ce-Al mixed oxide. Then, the mixture was calcined in air at 400 °C for 2 h. Alumina-supported Y-doped ceria support was prepared by the same procedure upon Y₂O₃ addition to the mixture before grinding by using 1 wt.% yttria with respect to the amount of ceria. The notation of mixed oxides is Ce/Al and Y-Ce/Al. Pure CeO₂ was obtained by calcination of cerium hydroxide in air at 400 °C for 2 h.

Pd-based catalysts (1 wt.% Pd) were prepared by deposition–precipitation on carriers suspended in water using ultrasound through the interaction between aqueous solutions of Pd(NO₃)₂ and K₂CO₃ at a temperature of 60 °C, a constant of pH = 7, and a stirring speed of 250 rpm. The next steps were aging within 1 h, filtering, washing until the removal of nitrate ions, and thermal treatment at 400 °C for 2 h.

4.3. Catalyst Characterization

Quantitative determination of sample chemical composition was carried out by X-ray fluorescence spectrometry. Analyses were accomplished on a Fischerscope XDAL instrument using WinFTM software (Sindelfingen, Germany) https://www.kks.com.au/helmut_fischer_fischerscope_xrf_x_ray_winftm_v6_software_952_050/, accessed on 12 October 2024.

Sample textural characteristics were determined by N₂ adsorption–desorption isotherms at the liquid nitrogen temperature using a Micromeritics ASAP 2020 apparatus (Norcross, GA, USA). Before measurements, the samples were outgassed at 200 °C for 1 h under vacuum. The specific surface area (S_BET) and pore volume (Vp) were calculated using the BET equation applied to the adsorption curve in the standard pressure range of 0.05–0.3 P/P₀. The pore size distribution was obtained by analysis of the desorption curve using the BJH method.

X-ray powder diffraction (XRD) was applied to determine phase composition by means of a Panalytical Empyrean apparatus (Panalytical B.V., Almelo, The Netherlands) equipped with a multichannel detector using Cu Kα 45 kV-40 mA radiation in the 2θ range of 20–115° at a scan step of 0.01° for 20 s.

X-ray photoelectron (XP) spectra were recorded on a VG Microtech ESCA 3000 Multilab instrument (VG Scientific, Sussex, UK) equipped with a dual Mg/Al anode. The experimental procedure is described elsewhere [19]. Sample constant charging was removed by referencing all the energies to the C 1s set at 285.1 eV. Peak analysis was carried out with Casa XPS software (version 2.3.17, Casa Software Ltd. Wilmslow, Cheshire, UK, 2009).

EPR spectra were recorded on a JEOL JES FA100 (JEOL, Tokyo, Japan) ESR spectrometer. The instrument operated in the X band and had a standard T₀₁₁ cylindrical resonator. An ES-DVT4 temperature controller was used to change the resonator temperature and observe EPR spectra at various temperatures. Samples of 100 mg each were loaded into the EPR tubes with the aim of better standardization, and the spectra were recorded under the same conditions: modulation frequency of 100 kHz, microwave power of 1 mW, modulation amplitude of 0.2 mT, time constant of 0.3 s, and sweep time of 4 min. The spectra were temperature-independent. Line intensity was the only variable parameter, which increased with a recorded decrease in temperature.

Temperature-programmed reduction (H₂-TPR) measurements were carried out as previously described [18] using 10% H₂ in Ar at a flow rate of 24 mL min⁻¹ and under a temperature increase of 15 °C min⁻¹. The sample quantity was 0.05 cm³ (particle size of 0.25–0.50 mm) based on the criterion proposed by Monti and Baiker. Hydrogen consumption was calculated by calibration of the thermal conductivity detector using stoichiometric NiO obtained via oxide calcination at 800 °C for 2 h.

4.4. Catalytic Activity Measurements

The samples were examined in the reaction of complete methane oxidation. Measurements were conducted in a flow-type glass reactor at atmospheric pressure (Figure S2) with 0.3 cm³ charge of the 0.63–0.8 mm fraction and a thin thermocouple in the catalyst bed. A COMECO RT 290 programmable controller (Stuttgart, Germany) monitored reactor temperature. The reaction gas mixture was composed of 1000 ppmv CH₄ and 25 vol.% O₂ and balanced up to 100% with nitrogen, flowing at a total rate of 20,000 h⁻¹. Allborg mass flow controllers were used to keep steady gas flow rates. Gas analysis of the reaction products was performed online using a SICK GMS84 gas analyzer (SICK AG, Waldkirch, Germany) for CO/CO₂/O₂/CH₄. Before the reaction, the catalysts were pretreated in a flow of nitrogen and oxygen at 300 °C for 1 h.

The catalytic activity in CBO is expressed as the benzene conversion degree. It was studied within the temperature range of 150–350 °C at atmospheric pressure using a microcatalytic continuous-flow fixed bed reactor connected to a Hewlett Packard 5890 series II gas chromatograph (Agilent, Germany, working with Agilent G2070 Chemstation Software B.04.01, Waldbronn, Germany). The instrument was supplied with a flame ionization detector and capillary column HP Plot Q. A scheme of the catalytic setup used for benzene oxidation measurements is shown in Figure S3. The measurements were carried out using catalyst samples with a 0.5 cm³ bed volume, a particle size of 0.25–0.50 mm, an inlet benzene concentration of 1314 ppm in air, and a space velocity of 4000 h⁻¹. The benzene conversion degree was calculated by the following equation:

C₆H₆ conversion (%) = {[(C₆H₆)_in − (C₆H₆)_out]/(C₆H₆)_in} × 100,

where (C₆H₆)_in and (C₆H₆)_out are the benzene concentration at the reactor inlet and outlet, respectively.

5. Conclusions

The carrier effect on complete methane and benzene oxidation over supported Pd catalysts was found to depend on the hydrocarbon type, owing differences in the mechanisms of interaction between the substrate and Pd²⁺, namely through the electron-rich π bonding for benzene and through a C−Pd²⁺ interaction in the case of methane. Palladium samples prepared by deposition–precipitation of palladium nitrate on CeO₂, CeO₂-Al₂O₃, and Y-CeO₂-Al₂O₃ exhibited higher CH₄ oxidation activity than Pd/Al₂O₃. Ceria-containing samples reached about 90% CH₄ conversion at 300 °C and complete oxidation at 320 °C vs. 60% at 300 °C and 100% at 390 °C with Pd/Al₂O₃. The adsorption of water generated by the reaction over acidic Al₂O₃ could inhibit CH₄ oxidation, probably by blocking the active sites or due to the formation of less active Pd(OH)₂. Pd-based catalysts using CeO₂-Al₂O₃ and especially Y-CeO₂-Al₂O₃ carriers demonstrated superior benzene conversion. Higher Pd and Ce³⁺ surface concentrations, the presence of adsorbed O₂⁻ superoxo species, and redox Pd⁰ ↔ PdO transfer contributed to the best performance. These results represent an excellent example of the application of mechanochemical preparation of mixed oxides as carriers of Pd-based catalysts for complete hydrocarbon oxidation and highlight the crucial importance of rational catalyst design for efficient catalyst performance.