Structural transformation and redox chemistry of Pd/CeO₂ during SO₂-induced sulfurization: an in situ XAFS study

Abstract

In situ X-ray absorption fine structure (XAFS) spectroscopy was employed to elucidate the structural evolution of Pd/CeO₂ catalysts during SO₂-induced sulfurization. Linear combination fitting of Ce L₃-edge XANES spectra quantified the Ce⁴⁺ → Ce³⁺ transformation, revealing fundamentally different sulfurization mechanisms for bare versus Pd-promoted CeO₂. At 500 °C, bare CeO₂ exhibited surface-limited sulfurization with Ce³⁺/(Ce³⁺ + Ce⁴⁺) reaching 0.18, while Pd/CeO₂ achieved 0.68, demonstrating bulk oxygen participation. Temperature-dependent measurements (200–500 °C) confirmed thermal activation of the sulfurization process, with higher temperatures enabling deeper lattice penetration. Complementary Pd L₃-edge XANES revealed that Pd maintained its oxidized state throughout SO₂ exposure, excluding PdS formation. S K-edge analysis confirmed exclusive SO₄²⁻ formation via direct oxidation without intermediate species. These findings establish that Pd catalyzes oxygen mobility within the CeO₂ lattice, transforming sulfurization from a surface-confined to a bulk-accessible process while preserving the fluorite structure. The resulting cerium sulfate oxide (Ce₂O₂SO₄) exhibits enhanced stability against re-oxidation in Pd/CeO₂, contrasting with the partial reversibility observed for bare CeO₂.

---

1. Introduction

Pd catalysts are key materials for purifying exhaust gases from factories and vehicles.^1–4 These gases contain air pollutants and harmful compounds, such as NOx, CO, and hydrocarbons. Regulations on these compounds are becoming increasingly stringent worldwide to achieve a sustainable society. Generally, platinum-group metal (PGM) catalysts are used to oxidize or reduce harmful compounds to non-toxic substances. There is a strong need to reduce the amount of PGM used because they are scarce and expensive. Therefore, the activity of PGM catalysts must be improved. It has been reported that Pd has better thermal durability than Pt.^1,5–7 Various oxides, such as Al₂O₃, SiO₂, and TiO₂, have been used as supports for Pd catalysts. CeO₂ has excellent redox properties and improves CO oxidation activity at low temperatures when used as a supporting material for Pd catalysts. Therefore, CeO₂-supported Pd catalysts (Pd/CeO₂) are frequently used for oxidation reactions.

Catalyst poisoning by sulfur-containing substances (e.g., SO₂ and H₂S) poses a critical challenge in exhaust gas purification. SO₂, ubiquitously emitted from both stationary and mobile sources, can be converted to H₂S under reducing conditions, creating multiple pathways for sulfur poisoning. This sulfur exposure severely deteriorates the performance of Pd catalysts.^8–12 Therefore, developing Pd catalysts with both high activity and enhanced sulfur tolerance is imperative. A thorough understanding of the sulfur poisoning mechanism is essential for the rational design of such S-tolerant catalysts.

It has been reported that treatment of supported Pd catalysts with SO₂ generally leads to a significant decrease in catalytic activity, primarily due to palladium sulfidation and a reduction in catalyst surface area.^11,13–16 The effects of SO₂ treatment on the properties of CeO₂ have also been investigated.^17–19 When CeO₂ undergoes sulfurization, oxygen vacancies are introduced, resulting in the reduction of Ce⁴⁺ to Ce³⁺. Sulfurized CeO₂ formed by SO₂ treatment exhibits enhanced oxygen storage capacity and high activity for the reduction of nitrogen oxides by ammonia.²⁰

Although Pd/CeO₂ catalysts suffer from activity deterioration in the presence of SO₂, Hilaire et al. reported that SO₂ poisoning, under alternating CO and O₂ pulse conditions, unexpectedly increased the amount of oxygen that could be transferred to and from the catalyst across the entire temperature range examined.²¹ These findings highlight the complexity of sulfur poisoning behavior in CeO₂-supported Pd catalysts. Despite these insights, most previous investigations have been conducted under ex situ conditions or have focused solely on catalytic performance evaluation. Consequently, the real-time structural evolution of Pd/CeO₂ during SO₂ exposure and the specific role of Pd in promoting CeO₂ sulfurization remain insufficiently understood. Addressing these knowledge gaps is essential for developing sulfur-tolerant Pd-based catalysts.

In this study, we employed in situ XAFS measurements under controlled SO₂ gas flow to capture the dynamic structural changes of both Pd and CeO₂ components in real time. This approach enables us to directly observe the sequential transformation of Ce oxidation states (Ce⁴⁺ to Ce³⁺) during SO₂ exposure, simultaneously monitor the chemical state of Pd species to clarify whether PdSO₄ or PdS formation occurs under our reaction conditions, quantitatively determine the Ce³⁺/Ce⁴⁺ ratio evolution as a function of time and temperature, and elucidate the promotional effect of Pd on CeO₂ sulfurization through comparative studies of Pd/CeO₂versus bare CeO₂. Our in situ approach, combined with complementary ex situ characterization (XRD, STEM-EDS, XPS), provides unprecedented insights into the sulfurization mechanism and helps explain the paradoxical enhancement of oxygen storage capacity reported by Hilaire et al. This mechanistic understanding is crucial for designing next-generation sulfur-tolerant catalysts.

Unlike previous studies that primarily relied on ex situ characterization or focused on catalytic performance under SO₂ exposure, our work provides the first comprehensive in situ XAFS investigation of Pd/CeO₂ sulfurization under controlled reaction conditions. This real-time approach enables direct observation of Ce oxidation state evolution and Pd speciation during SO₂ treatment, revealing the mechanistic role of Pd in promoting bulk oxygen mobility and irreversible sulfate formation. By combining Ce L₃-, Pd L₃-, and S K-edge analyses with complementary structural characterization, we uncover a fundamental shift in sulfurization behavior—from surface-limited in bare CeO₂ to bulk-accessible in Pd/CeO₂—thus offering unprecedented insights into the design of sulfur-tolerant Pd-based catalysts.

2. Experimental

2.1 Catalyst preparation

Pd/CeO₂ catalysts were prepared by the impregnation method. CeO₂ (JRC-CEO-5, Catalysis Society of Japan, 99.98% purity) was used as the support. The pore volume of JRC-CEO-5 CeO₂ determined by N₂ adsorption was 0.24 mL cm⁻³ g⁻¹. An aqueous solution of tetraamminepalladium(II) nitrate, [Pd(NH₃)₄](NO₃)₂ (10 wt% in H₂O, Sigma-Aldrich Japan, purity: 99.99%) was added dropwise to CeO₂ powder. The amount of [Pd(NH₃)₄](NO₃)₂ solution was adjusted to achieve a Pd loading of 1 wt%. The impregnation was carried out in a water bath at 60 °C, followed by thorough mixing and drying. The precursor was dried at 100 °C overnight, calcined in air at 400 °C for 2 h, and subsequently reduced under a 5% H₂–N₂ flow at 200 °C for 1 h. The final Pd loading was 1 wt%.

2.2 Catalyst characterization

X-ray diffraction (XRD) patterns were recorded using a RINT 2200 diffractometer (Rigaku, Japan) with Cu Kα radiation (λ = 1.54 Å) at 40 kV and 20 mA, with a step rate of 2° min⁻¹. Temperature-programmed reduction with H₂ (H₂-TPR) was performed using a BEL-CAT instrument (Microtrac BEL). Samples (0.10 g) were pretreated in air at 350 °C for 1 h, then heated from 50 to 950 °C at 5 °C min⁻¹. X-ray photoelectron spectroscopy (XPS) was carried out using a Kratos ESCA-3400 spectrometer (AXIS-165, KRATOS, Japan) with an Al Kα source. X-ray fluorescence (XRF) spectra were obtained using a ZSX Primus IV/RX9 instrument (Rigaku).

2.3 In situ XAFS measurements

XAFS measurements were conducted at the Kyushu Synchrotron Light Research Center (SAGA-LS), beamline BL06 (Saga, Japan). The storage ring energy was 1.4 GeV, and a Si(111) double-crystal monochromator was employed. Ce L₃-edge spectra were collected in transmission mode, while S K-edge and Pd L₃-edge spectra were recorded in fluorescence mode. For in situ measurements, catalysts were mixed with boron nitride, pressed into 10 mm diameter disks, and placed in a glass cell (Fig. S1). The samples were heated to the reaction temperature under a 500 ppm SO₂–N₂ flow (100 mL min⁻¹), then measured at room temperature. Subsequently, the catalysts were reoxidized in O₂ at 500 °C and re-measured. Additional measurements were performed at 200 °C and 400 °C under the same conditions. XAFS data analysis was carried out using the Athena software package.²² CeO₂ (JRC-CEO-5, Catalysis Society of Japan, 99.98% purity) and Ce(NO₃)₃ (Kojundo Chemical Lab. Co., Ltd, 99.9% purity) were used as reference samples.

3. Results and discussion

3.1 Structural changes in CeO₂ support induced by SO₂ treatment

Fig. 1 presents the in situ Ce L₃-edge XANES spectra of CeO₂ and Pd/CeO₂ during SO₂ treatment at 500 °C. In the fresh samples, characteristic peaks corresponding to quasi-atomic Ce 2p_3/2 → 5d_5/2,3/2 electronic transitions^23–25 were observed at 5730 and 5737 eV, confirming the predominance of Ce⁴⁺ in both materials. The similarity between the spectra of Pd/CeO₂ and bare CeO₂ indicates that Pd loading did not alter the initial oxidation state of Ce. Upon exposure to SO₂, both samples exhibited a gradual decrease in peak intensities at 5730 and 5737 eV, accompanied by a shift of the absorption edge (Fig. 1(b)). Notably, Pd/CeO₂ displayed more pronounced spectral changes, including the emergence of a distinct peak at 5726 eV characteristic of Ce³⁺, demonstrating that Pd significantly enhances CeO₂ reduction under SO₂. The presence of isosbestic points in both samples confirms direct conversion of Ce⁴⁺ to Ce³⁺ without detectable intermediates.

Fig. 1 In situ Ce L₃-edge XANES spectra of CeO₂ and Pd/CeO₂ at 500 °C. (a) Fresh samples before SO₂ exposure. (b) Under SO₂ flow (500 ppm, 100 mL min⁻¹) for up to 300 min. (c) After switching to O₂ flow at 500 °C following SO₂ treatment.

Quantitative analysis of the Ce oxidation states was performed using two complementary approaches. First, the XANES spectra were deconvoluted using arctangent and Gaussian functions (Fig. 2), yielding four components: peaks A, B, and D (Ce⁴⁺) and peak C (Ce³⁺). The Ce³⁺/(Ce³⁺+Ce⁴⁺) ratio was calculated from the integrated peak areas. In pristine CeO₂, Ce³⁺ arises from intrinsic oxygen vacancies; however, its absolute concentration depends significantly on preparation conditions and fitting procedures.^26–28 To validate these results, linear combination fitting (LCF) analysis was conducted using CeO₂ and Ce(NO₃)₃ as Ce⁴⁺ and Ce³⁺ references, respectively (Fig. 3). While this method cannot quantify intrinsic Ce³⁺ in pure CeO₂, it provides reliable assessment of SO₂-induced changes. The temporal evolution of Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratios (Fig. 4) was monitored throughout SO₂ exposure and subsequent O₂ reoxidation. For bare CeO₂, measurements were taken at 0, 20, 40, 60, and 80 min under SO₂ flow, followed by O₂ treatment at 500 °C. For Pd/CeO₂, data points were collected at 0, 20, 40, 60, 80, 150, 240, and 300 min during continuous SO₂ exposure, after which the gas was switched to O₂ for reoxidation. These time intervals correspond exactly to the horizontal axis in Fig. 4, ensuring direct comparison between the two catalysts. Bare CeO₂ reached a plateau at Ce³⁺/(Ce³⁺ + Ce⁴⁺) = 0.18 after 60 min, whereas Pd/CeO₂ continued reducing to Ce³⁺/(Ce³⁺ + Ce⁴⁺) = 0.68 after 300 min, highlighting the catalytic role of Pd in CeO₂ sulfurization.

Fig. 2 Ce L₃-edge XANES spectra and peak fitting for Ce oxidation states. (a) Fresh CeO₂. (b) CeO₂ after 80 min under SO₂ flow at 500 °C. (c) Pd/CeO₂ after 300 min under SO₂ flow at 500 °C.

Fig. 3 Linear combination fitting (LCF) of Ce L₃-edge XANES spectra using CeO₂ and Ce(NO₃)₃ references. (a) CeO₂ after 80 min SO₂ exposure at 500 °C. (b) Pd/CeO₂ after 300 min SO₂ exposure at 500 °C.

Fig. 4 Time-dependent evolution of Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratio for bare CeO₂ and Pd/CeO₂ during SO₂ exposure at 500 °C, followed by O₂ reoxidation. CeO₂ was measured at 0, 20, 40, 60, and 80 min under SO₂ flow, then switched to O₂ at 500 °C. Pd/CeO₂ was measured at 0, 20, 40, 60, 80, 150, 240, and 300 min under SO₂ flow, followed by O₂ treatment at 500 °C.

Upon switching from SO₂ to O₂ flow at 500 °C, the XANES spectra revealed distinct re-oxidation behaviors (Fig. 1(c)). Bare CeO₂ exhibited partial recovery of the Ce⁴⁺ state, as evidenced by the diminished Ce³⁺ peak intensity, indicating reversible reduction of surface sites. In contrast, Pd/CeO₂ showed no spectral changes during O₂ treatment, demonstrating that Pd stabilizes the sulfurized state and prevents re-oxidation. The absence of re-oxidation also excludes the formation of bulk Ce₂O₃ during SO₂ treatment, as Ce₂O₃ would readily re-oxidize to CeO₂ under these oxidizing conditions.

Fig. 5 presents the in situ Ce L₃-edge EXAFS spectra of CeO₂ and Pd/CeO₂ collected at 500 °C under SO2 flow. Both samples initially displayed characteristic coordination shells: Ce–O at 2.0 Å and Ce–(O)–Ce at 3.5 Å (phase uncorrected). Progressive SO₂ exposure induced time-dependent attenuation of both peaks, reflecting the disruption of local structural order in the CeO₂ lattice. Notably, after 40 min of SO₂ treatment, Pd/CeO₂ exhibited significantly greater structural disorder than bare CeO₂, as evidenced by the more pronounced reduction in Ce–O and Ce–(O)–Ce peak intensities. Subsequent O₂ treatment revealed contrasting structural recovery behaviors. While bare CeO₂ showed partial restoration of the Ce–O coordination shell, consistent with the Ce³⁺ → Ce⁴⁺ re-oxidation observed in XANES (Fig. 1), Pd/CeO₂ remained structurally unchanged. This irreversibility demonstrates that Pd not only accelerates CeO₂ sulfurization but also stabilizes the resulting disordered structure against re-oxidation, likely through the formation of stable sulfate species at the Pd–CeO₂ interface.

Fig. 5 In situ Ce L₃-edge EXAFS spectra of CeO₂ and Pd/CeO₂ during SO₂ exposure at 500 °C. CeO₂ measured at 0, 20, 40, 60, and 80 min; Pd/CeO₂ measured at 0, 20, 40, 60, 80, 150, 240, and 300 min under SO₂ flow (500 ppm, 100 mL min⁻¹).

The temperature dependence of SO₂-induced CeO₂ reduction was investigated systematically (Fig. 6). At 400 °C, the Ce L₃-edge XANES spectra exhibited a characteristic red-shift of the absorption edge accompanied by diminished Ce⁴⁺ peak intensity, confirming partial reduction to Ce³⁺. The presence of clear isosbestic points demonstrated that direct Ce⁴⁺ → Ce³⁺ conversion occurred without intermediates, similar to the 500 °C treatment. However, quantitative analysis revealed a lower Ce³⁺/(Ce³⁺+Ce⁴⁺) ratio compared to 500 °C, establishing a positive correlation between reaction temperature and reduction extent. At 200 °C, minimal spectral changes were observed even after prolonged SO₂ exposure, suggesting that sulfurization was confined to the outermost surface layer under these mild conditions. These results demonstrate that CeO₂ sulfurization exhibits strong thermal activation, with higher temperatures required to enable SO₂ penetration beyond the surface region into the bulk lattice.

Fig. 6 Ce L₃-edge XANES spectra of CeO₂ and Pd/CeO₂ under SO₂ flow (500 ppm, 100 mL·min⁻¹) at 200 °C and 400 °C. Measurements taken at 0, 20, 40, and 60 min.

To elucidate the role of Pd in CeO₂ reduction, comparative H₂-reduction experiments were conducted at 600 °C (Fig. 7(a)). This temperature was selected based on established H₂-TPR profiles showing distinct reduction regimes: surface oxygen removal (50–600 °C) and bulk oxygen extraction (600–950 °C).^29,30 Therefore, treatment at 600 °C enables selective monitoring of surface reduction processes. Both CeO₂ and Pd/CeO₂ exhibited comparable shifts in their absorption edges during H₂ treatment, with linear combination fitting revealing identical Ce³⁺ formation (15% conversion from Ce⁴⁺) in both samples. This equivalence contrasts sharply with the SO₂-treatment results, where Pd dramatically enhanced CeO₂ reduction. The absence of Pd promotion under H₂ indicates that the catalytic effect is specific to SO₂-mediated reduction pathways, suggesting fundamentally different mechanistic routes for H₂versus SO₂ reduction of CeO₂.

Fig. 7 (a) Ce L₃-edge XANES and (b) EXAFS spectra of CeO₂ and Pd/CeO₂ during H₂ reduction at 600 °C (5% H₂–N₂ flow, 100 mL min⁻¹). Measurements taken after 60 min of H₂ exposure.

The structural consequences of H₂ reduction were examined via in situ Ce L₃-edge EXAFS (Fig. 7(b)). Both CeO₂ and Pd/CeO₂ displayed comparable attenuation of Ce–O and Ce–(O)–Ce coordination peaks at 600 °C, indicating similar degrees of local structural disorder upon surface oxygen removal. The equivalent post-reduction peak intensities confirm that while Pd accelerates the kinetics of CeO₂ reduction under certain conditions, it does not alter the final structural state of H₂-reduced CeO₂. This observation further supports that the promotional effect of Pd is reaction-specific rather than a general enhancement of CeO₂ reducibility.

The comparative analysis of H₂ and SO₂ reduction provides critical mechanistic insights into Pd's catalytic role. For bare CeO₂, SO₂ treatment at 500 °C yielded Ce³⁺/(Ce³⁺+Ce⁴⁺) ratios identical to those achieved by H₂ reduction at 600 °C, demonstrating that SO₂-induced sulfate formation is confined to surface oxygen sites. This surface-limited reaction persisted even under prolonged SO₂ exposure, indicating an intrinsic barrier to bulk sulfurization in unpromoted CeO₂. In stark contrast, Pd/CeO₂ under SO₂ treatment surpassed the H₂-induced reduction threshold, unequivocally demonstrating SO₂ access to bulk oxygen sites. This fundamental difference reveals that Pd not only accelerates surface sulfurization kinetics but, more importantly, enables SO₂ penetration into the CeO₂ bulk lattice. The mechanism likely involves Pd-mediated oxygen mobility enhancement or the creation of preferential SO₂ diffusion pathways at the Pd–CeO₂ interface, fundamentally altering the sulfurization process from a surface-limited to a bulk-accessible reaction.

Surface chemical states were characterized by XPS to complement the bulk-sensitive XAFS measurements (Fig. 8). The Ce 3d spectra exhibited the characteristic multiplet structure with V (3d_5/2) and U (3d_3/2) components and their satellites (V′, V″, V‴, U′, U″, U‴).^31–34 Fresh samples displayed predominantly Ce⁴⁺-associated features, while SO₂-treated Pd/CeO₂ showed pronounced enhancement of the Ce³⁺ markers (U′ and V′ peaks).^31,34 Quantitative deconvolution yielded Ce³⁺/Ce⁴⁺ = 0.31/0.69, in reasonable agreement with the XANES-derived values. Given the XPS probing depth of approximately 2 nm in CeO₂,^38,39 these results confirm substantial Ce³⁺ enrichment within the near-surface region. The coexistence of Ce⁴⁺ and Ce³⁺ in this shallow sampling volume indicates incomplete surface sulfurization, suggesting either heterogeneous reaction sites or equilibrium between sulfurized and pristine surface domains.

Fig. 8 XPS spectra of Pd/CeO₂ before and after SO₂ treatment at 500 °C (500 ppm SO₂, 100 mL min⁻¹, 300 min). Spectra show Ce 3d multiplet structure and changes in Ce³⁺/Ce⁴⁺ ratio upon sulfurization.

XRD analysis was employed to assess crystallographic changes induced by SO₂ treatment (Fig. 9). Fresh Pd/CeO₂ exhibited characteristic reflections of the fluorite structure at 2θ = 28, 33, 47, 56, 59, 69, 76, 79, 88, and 95°, consistent with cubic CeO₂ (Fm3̄m).³⁵ The absence of Pd or PdO reflections confirms high metal dispersion on the CeO₂ support. Following SO₂ treatment, the fluorite reflections showed reduced intensities while maintaining their positions, with no emergence of new phases. This intensity attenuation without peak shifting or phase transformation indicates partial amorphization or increased structural disorder rather than formation of crystalline sulfate phases. The preservation of the fluorite framework despite SO₂-induced reduction aligns with the EXAFS observations of decreased local ordering while maintaining the basic CeO₂ structure.

Fig. 9 XRD patterns of Pd/CeO₂ before and after SO₂ exposure at 500 °C (500 ppm SO₂, 100 mL min⁻¹, 300 min). Fluorite structure reflections are retained with intensity attenuation, indicating increased disorder without phase transformation.

In this study, the Pd/CeO₂ catalyst was subjected to calcination in air at 400 °C, followed by reductive activation under 5% H₂/N₂ at 200 °C. To properly interpret the sulfurization behavior of Pd/CeO₂, it is important to consider the extent to which these thermal and reductive treatments may influence the CeO₂ support itself. Because the CeO₂ used in this work had already undergone high-temperature calcination at 600 °C prior to catalyst preparation, an additional calcination at 400 °C is unlikely to induce further structural modifications. Thus, the only treatment step that may affect the bare support is the mild reduction at 200 °C. According to previous H₂-TPR studies, CeO₂ exhibits negligible reduction at 200 °C, with only minimal Ce⁴⁺ → Ce³⁺ conversion and insignificant oxygen-vacancy formation.²⁹ These findings indicate that such mild reductive pre-treatment does not meaningfully alter the intrinsic redox properties of CeO₂. In contrast, H₂-TPR measurements demonstrate that the incorporation of Pd markedly enhances the reducibility of surface lattice oxygen in CeO₂. Therefore, the possibility that this Pd-induced increase in oxygen mobility facilitates the reaction with SO₂ cannot be excluded.

3.2 Changes in Pd on catalysts and the structure of sulfur species formed by SO₂

The Pd-CeO₂ interface plays a crucial role in catalytic performance, particularly for CO oxidation, where Pd deposition on CeO₂ dramatically enhances activity.^29,36,37 Since the Pd oxidation state directly governs interfacial reactivity, we monitored Pd speciation during SO₂ exposure using in situ Pd L₃-edge XANES (Fig. 10). The Pd L₃-edge spectra, originating from 2p_3/2 → 4d electronic transitions and thus sensitive to d-electron occupancy, remained invariant throughout SO₂ treatment at 500 °C. This spectral stability definitively excludes Pd reduction to metallic Pd⁰ or conversion to PdS, indicating that Pd maintains its oxidized state despite the strongly reducing SO₂ atmosphere. This resistance to reduction contrasts markedly with the facile CeO₂ reduction observed under identical conditions, suggesting that SO₂ preferentially interacts with the CeO₂ support rather than the supported Pd species.

Fig. 10 In situ Pd L₃-edge XANES spectra of Pd/CeO₂ during SO₂ exposure at 500 °C (500 ppm SO₂, 100 mL min⁻¹) for up to 300 min. Spectra confirm Pd remains in oxidized state throughout sulfurization.

STEM-EDS mapping provided direct visualization of Pd dispersion and sulfur distribution before and after SO₂ treatment at 500 °C (Fig. 11). Fresh Pd/CeO₂ exhibited uniformly dispersed Pd nanoparticles with a mean diameter of ∼3 nm and absence of large aggregates (>30 nm), consistent with the XRD results showing no detectable Pd or PdO reflections. Following SO₂ treatment, Pd maintained its original particle size distribution, demonstrating remarkable resistance to sintering under sulfurizing conditions. Critically, EDS mapping revealed homogeneous sulfur distribution across the entire catalyst surface, confirming uniform CeO₂ sulfurization rather than localized sulfate formation. This comprehensive surface sulfurization, combined with stable Pd dispersion, indicates that SO₂ interacts primarily with the CeO₂ support while leaving the metal nanoparticles structurally intact.

Fig. 11 STEM-EDS mapping of Pd/CeO₂ before and after SO₂ exposure at 500 °C (500 ppm SO₂, 100 mL min⁻¹, 300 min). (a) Fresh sample showing uniform Pd dispersion (∼3 nm). (b) After SO₂ treatment, Pd particle size remains unchanged; sulfur is homogeneously distributed across the surface.

Sulfur speciation during SO₂ treatment was monitored via in situ S K-edge XANES spectroscopy (Fig. 12). Both CeO₂ and Pd/CeO₂ developed a dominant peak at 2.482 keV, characteristic of SO₄²⁻ species, which intensified over 60 min before reaching steady state. For bare CeO₂, a transient feature at 2.479 keV indicated initial SO₂ adsorption, which subsequently oxidized to S⁶⁺ within 60 min. In contrast, Pd/CeO₂ exclusively formed SO₄²⁻ without detectable intermediate oxidation states, suggesting that Pd facilitates direct SO₂ → SO₄²⁻ conversion. These observations confirm that surface sulfate formation is the primary sulfurization pathway for both catalysts, with the key distinction being the reaction kinetics: Pd accelerates complete oxidation to sulfate while suppressing intermediate species formation. This direct sulfate formation mechanism explains the irreversible nature of Pd/CeO₂ sulfurization observed in the Ce L₃-edge XANES studies.

Fig. 12 In situ S K-edge XANES spectra of CeO₂ and Pd/CeO₂ during SO₂ exposure at 500 °C (500 ppm SO₂, 100 mL min⁻¹) for up to 60 min. Spectra show formation of SO₄²⁻ species; Pd/CeO₂ exhibits direct conversion without intermediate oxidation states.

Diffuse-reflectance FTIR spectra collected for CeO₂ and Pd/CeO₂ before and after sulfation exhibit a broad absorption band near 1200 cm⁻¹, which is indicative of bulk-like sulfate species incorporated beneath the ceria surface (Fig. S2). This interpretation is consistent with previous findings showing that subsurface sulfate formation on ceria produces a broad feature around 1200 cm⁻¹, whereas surface sulfate species generate sharper bands in the 1340–1400 cm⁻¹ region.³⁸ The predominance of the 1200 cm⁻¹ band under the present conditions suggests that sulfation proceeds through partial reduction of Ce⁴⁺ to Ce³⁺ and subsequent migration of sulfate into near-surface layers. The comparatively stronger intensity of this band on Pd/CeO₂ implies that palladium enhances the redox activity and oxygen mobility of the ceria support, thereby promoting the incorporation of sulfate species into subsurface sites.

Quantitative sulfur analysis by XRF provided further insights into the extent of sulfurization (Table 1). The Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratios were evaluated from these peak areas according to the reported literature. The Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratios were calculated from the integrated peak areas based on established methods reported in the literature.³⁹ The S/CeO₂ molar ratios were 0.24 for Pd/CeO₂versus 0.09 for bare CeO₂, demonstrating a 2.7-fold enhancement in sulfate formation with Pd promotion. This substantial difference correlates directly with the enhanced Ce⁴⁺ → Ce³⁺ reduction observed in XANES, confirming that Pd catalyzes both lattice oxygen extraction and subsequent sulfate formation. Notably, the Ce³⁺/S molar ratio was approximately 1.7 for both catalysts, suggesting a stoichiometric relationship independent of Pd presence. This invariant ratio indicates that each sulfate formation event consumes a consistent number of lattice oxygen atoms, likely following the reaction: 2CeO₂ + SO₂ → Ce₂O₃ + SO₄²⁻. The preservation of the fluorite framework despite extensive oxygen removal and sulfate incorporation demonstrates remarkable structural resilience, with sulfate species presumably occupying oxygen vacancy sites or surface positions without disrupting the underlying crystal lattice.

Table 1 Ce ion ratios calculated by linear combination fitting with standards and mol ratio of SO₃/CeO₂ calculated by XRF

	LCF fitting		Curve fitting		SO3/CeO2
	Ce^3+	Ce^4+	Ce^3+	Ce^4+
Fresh CeO2	0.00	1.00	0.11	0.89	—
CeO2 after SO2	0.15	0.85	0.14	0.86	0.09
CeO2 after H2	0.15	0.85	0.13	0.87	—
Fresh Pd/CeO2	0.02	0.98	0.12	0.88	—
Pd/CeO2 after SO2	0.40	0.60	0.28	0.72	0.24
Pd/CeO2 after H2	0.15	0.85	0.13	0.87	—

---

The sulfurized cerium compounds formed require careful consideration of thermodynamic stability and experimental evidence. While Ce(SO₄)₂ and Ce₂(SO₄)₃ represent the thermodynamically stable phases, cerium uniquely does not form oxysulfate phases observed in other rare-earth elements.⁴⁰ Previous studies have identified cerium sulfate oxide (Ce₂O₂SO₄) containing Ce³⁺ as the primary product of CeO₂ sulfurization by SO₂.^41,42 Our observed Ce³⁺/S ratio of ∼1.7 closely approximates the theoretical value of 2.0 expected for Ce₂O₂SO₄ formation, strongly suggesting this phase as the dominant sulfurization product. The presence of isosbestic points during in situ XANES measurements provides definitive evidence for direct CeO₂ → Ce₂O₂SO₄ transformation without detectable intermediates, supporting a concerted mechanism where lattice oxygen removal and sulfate incorporation occur simultaneously. This direct conversion pathway, accelerated by Pd but maintaining the same stoichiometry, underscores the fundamental nature of the CeO₂–SO₂ reaction.

4. Conclusions

This study employed comprehensive in situ XAFS spectroscopy combined with XRD and STEM-EDS characterization to elucidate the mechanistic pathways of SO₂-induced structural transformations in Pd/CeO₂ catalysts. SO₂ treatment of bare CeO₂ at 500 °C induces surface-limited sulfurization with Ce⁴⁺ → Ce³⁺ reduction reaching a plateau at Ce³⁺/(Ce³⁺ + Ce⁴⁺) = 0.18. The presence of Pd fundamentally alters this behavior, enabling SO₂ penetration into the bulk lattice and achieving a Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratio of 0.68 after 300 min. This sulfurized state proves remarkably stable, showing no re-oxidation upon O₂ exposure at 500 °C. Throughout the treatment, Pd remains oxidized and uniformly distributed without sintering, indicating that Pd promotes oxygen mobility rather than directly participating in sulfur chemistry. The primary product, Ce₂O₂SO₄, preserves the fluorite framework but exhibits diminished long-range order. These findings establish that Pd transforms the sulfurization mechanism from surface-limited to bulk-accessible while maintaining the structural architecture of CeO₂.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting the findings of this study are available within the article and its supplementary information files. Additional raw data, including XAFS spectra, XRD patterns, and XPS measurements, are available from the corresponding author upon reasonable request. No restrictions apply to data sharing.

Acknowledgements

The authors gratefully acknowledge the support of this research from JST SICORP JPMJSC18H3 and JSPS KAKENHI Grant Number 21H03635. Parts of the XAFS experiments were performed at Kyushu University Beamline (SAGA-LS / BL06) with the proposal No. 2020IIK006.

References

1. J. Wang, H. Chen, Z. Hu, M. Yao and Y. Li, Catal. Rev., 2015, 57, 79–144.
2. M. A. van Spronsen, J. W. M. Frenken and I. M. N. Groot, Chem. Soc. Rev., 2017, 46, 4347–4374.
3. D. Ciuparu, M. R. Lyubovsky, E. Altman, L. D. Pfefferle and A. Datye, Catal. Rev., 2002, 44, 593–649.
4. R. M. Heck and R. J. Farrauto, Appl. Catal., A, 2001, 221, 443–457.
5. Y. Ryou, J. Lee, H. Lee, J. W. Choung, S. Yoo and D. H. Kim, Catal. Today, 2015, 258, 518–524.
6. M. Kaneeda, H. Iizuka, T. Hiratsuka, N. Shinotsuka and M. Arai, Appl. Catal., B, 2009, 90, 564–569.
7. M. Chen and L. D. Schmidt, J. Catal., 1979, 56, 198–218.
8. C. H. Bartholomew, Appl. Catal., A, 2001, 212, 17–60.
9. P. Gélin and M. Primet, Appl. Catal., B, 2002, 39, 1–37.
10. S. Colussi, F. Arosio, T. Montanari, G. Busca, G. Groppi and A. Trovarelli, Catal. Today, 2010, 155, 59–65.
11. H. S. Gandhi and M. Shelef, Appl. Catal., 1991, 77, 175–186.
12. F. Pinna, F. Menegazzo, M. Signoretto, P. Canton, G. Fagherazzi and N. Pernicone, Appl. Catal., A, 2001, 219, 195–200.
13. M. Monai, T. Montini, M. Melchionna, T. Duchoň, P. Kúš, C. Chen, N. Tsud, L. Nasi, K. C. Prince, K. Veltruská, V. Matolín, M. M. Khader, R. J. Gorte and P. Fornasiero, Appl. Catal., B, 2017, 202, 72–83.
14. S. P. Chenakin, G. Melaet, R. Szukiewicz and N. Kruse, J. Catal., 2014, 312, 1–11.
15. N. M. Kinnunen, J. T. Hirvi, K. Kallinen, T. Maunula, M. Keenan and M. Suvanto, Appl. Catal., B, 2017, 207, 114–119.
16. T. Kolli, M. Huuhtanen, A. Hallikainen, K. Kallinen and R. L. Keiski, Catal. Lett., 2009, 127, 49–54.
17. H. Yang, G. Wei, X. Wang, F. Lin, J. Wang and M. Shen, Catal. Commun., 2013, 36, 5–9.
18. T. Luo and R. J. Gorte, Appl. Catal., B, 2004, 53, 77–85.
19. R. Flouty, E. Abi-Aad, S. Siffert and A. Aboukaïs, J. Therm. Anal. Calorim., 2003, 73, 727–734.
20. X. Liu, P. Wang, Y. Shen, S. Bi, W. Ren and D. Zhang, ACS Catal., 2022, 12, 11306–11317.
21. S. Hilaire, S. Sharma, R. J. Gorte, J. M. Vohs and H.-W. Jen, Catal. Lett., 2000, 70, 131–135.
22. B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12, 537–541.
23. A. V. Soldatov, T. S. Ivanchenko, S. Della Longa, A. Kotani, Y. Iwamoto and A. Bianconi, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50, 5074.
24. P. Nachimuthu, W.-C. Shih, R.-S. Liu, L.-Y. Jang and J.-M. Chen, J. Solid State Chem., 2000, 149, 408–413.
25. H. Dexpert, R. C. Karnatak, J.-M. Esteva, J. P. Connerade, M. Gasgnier, P. E. Caro and L. Albert, Phys. Rev. B:Condens. Matter Mater. Phys., 1987, 36, 1750–1753.
26. J. Ke, W. Zhu, Y. Jiang, R. Si, Y.-J. Wang, S.-C. Li, C. Jin, H. Liu, W.-G. Song, C.-H. Yan and Y.-W. Zhang, ACS Catal., 2015, 5, 4884–4892.
27. L. Sun, X. Huang, L. Wang and A Janotti, Phys. Rev. B, 2017, 95, 245101.
28. C. Macchi, A. M. S. Procópio, L. S. R. Rocha, P. P. Ortega, C. M. Aldao, L. I. Granone, H. M. S. Nascimento, M. S. Castro, F. D. Ivorra, A. Somoza, F. Moura and M. A. Ponce, ACS Omega, 2024, 9, 42172–42182.
29. S. Shigenobu, H. Hojo and H. Einaga, Ind. Eng. Chem. Res., 2022, 61, 15856–15865.
30. P. Fornasiero, G. Balducci, R. Di Monte, J. Kašpar, V. Sergo, G. Gubitosa, A. Ferrero and M. Graziani, J. Catal., 1996, 164, 173–183.
31. M. Romeo, K. Bak, J. El Fallah, F. Le Normand and L. Hilaire, Surf. Interface Anal., 1993, 20, 508–512.
32. P. W. Park and J. S. Ledford, Langmuir, 1996, 12, 1794–1799.
33. E. Bêche, P. Charvin, D. Perarnau, S. Abanades and G. Flamant, Surf. Interface Anal., 2008, 40, 264–267.
34. C. de Leitenburg, A. Trovarelli and J. Kašpar, J. Catal., 1997, 166, 98–107.
35. Y. Xu, M. Yamazaki and P. Villars, Jpn. J. Appl. Phys., 2011, 50, 11RH02.
36. S.-H. Oh and G. B. Hoflund, J. Phys. Chem. A, 2006, 110, 7609–7613.
37. M.-F. Luo, Z.-Y. Hou, X.-X. Yuan and X.-M. Zheng, Catal. Lett., 1998, 50, 205–209.
38. M. Waqif, P. Bazin, O. Saur, J. C. Lavalley, G. Blanchard and O. Touret, Appl. Catal., B, 1997, 11, 193–205.
39. A. Bianconi, A. Marcelli, H. Dexpert, R. Karnatak, A. Kotani, T. Jo and J. Petiau, Phys. Rev. B:Condens. Matter Mater. Phys., 1987, 35, 806–812.
40. R. K. Dwivedi and D. A. R. Kay, Metall. Trans. B, 1984, 15, 523–528.
41. M. Y. Smirnov, A. V. Kalinkin, A. V. Pashis, A. M. Sorokin, A. S. Noskov, K. C. Kharas and V. I. Bukhtiyarov, J. Phys. Chem. B, 2005, 109, 11712–11719.
42. M. Y. Smirnov, A. V. Kalinkin, A. V. Pashis, I. P. Prosvirin and V. I. Bukhtiyarov, J. Phys. Chem. C, 2014, 118, 22120–22135.

---