Oxygen vacancy-engineered titanium-based perovskite for boosting H₂O activation and lower-temperature hydrolysis of organic sulfur

Significance

Lower-temperature organic sulfur (COS and CS₂) hydrolysis is very critical to achieve high sulfur recovery efficiency and ultra-low emissions of the Claus process. However, hydrolysis activity at low temperature may be limited by inefficient activation of H₂O molecule. In this work, we revealed that a facile oxygen vacancy (V_O) engineering on titanium-based perovskite allowed to boosting H₂O activation and enables efficient COS and CS₂ hydrolysis at 225 °C, which attained to a favorable temperature for the Claus conversion. The in-depth understanding of the very nature for the promotion role of V_O in H₂O molecule activation and the correlation to catalytic activity can benefit the design of efficacious catalysts for H₂O-involved chemical reactions.

Abstract

Modulation of water activation is crucial to water-involved chemical reactions in heterogeneous catalysis. Organic sulfur (COS and CS₂) hydrolysis is such a typical reaction involving water (H₂O) molecule as a reactant. However, limited by the strong O-H bond in H₂O, satisfactory CS₂ hydrolysis performance is attained at high temperature above 310 °C, which is at the sacrifice of the Claus conversion, strongly hindering sulfur recovery efficiency improvement and pollution emissions control of the Claus process. Herein, we report a facile oxygen vacancy (V_O) engineering on titanium-based perovskite to motivate H₂O activation for enhanced COS and CS₂ hydrolysis at lower temperature. Increased amount of V_O contributed to improved degree of H₂O dissociation to generate more active -OH, due to lower energy barrier for H₂O dissociation over surface rich in V_O, particularly V_O clusters. Besides, low-coordinated Ti ions adjacent to V_O were active sites for H₂O activation. Consequently, complete conversion of COS and CS₂ was achieved over SrTiO₃ after H₂ reduction treatment at 225 °C, a favorable temperature for the Claus conversion, at which both satisfying COS and CS₂ hydrolysis performance and improved sulfur recovery efficiency can be obtained simultaneously. Additionally, the origin of enhanced hydrolysis activity from boosted H₂O activation by V_O was revealed via in-depth mechanism study. This provides more explicit direction for further design of efficacious catalysts for H₂O-involved reactions.

Water (H₂O) molecule activation on oxide surfaces is of great significance for both fundamental science and practical applications, especially for H₂O-involved chemical reactions in heterogenous catalysis, such as water gas shift, hydrolysis, steam reforming, and water splitting (1–5). Water is also employed as a key reactant to remove organic sulfur (CS₂ and COS) from the Claus reactor, in which the H₂S undergoes the Claus reaction to recover elemental sulfur, which is widely used to handle troublesome H₂S in the refinery and natural gas industries (6). The undesired organic sulfur would poison subsequent catalysts, and if unhandled, would be converted to harmful SO_x in the final incinerator, causing unnecessary loss of sulfur and severe environmental issues. With the increasingly stringent emission standards, it is urgent to improve the hydrolysis performance of CS₂ and COS to achieve higher sulfur recovery efficiency and ultra-low emissions. It is widely accepted that hydroxyl (-OH) from dissociated water on catalyst surface is pivotal active species for CS₂ and COS hydrolysis (7). However, owing to the high O-H bond energy (462 kJ/mol) of H₂O molecule, relative high temperature (above 310 °C) is employed for traditional catalysts to motivate efficient H₂O activation to gain satisfactory CS₂ hydrolysis performance, which is at the expense of the Claus conversion (8). Since the Claus reaction is exothermic, it is favored at lower temperature of 200 to 250 °C. Hence, enabling efficient H₂O molecule activation at such lower temperature is highly desired to promote CS₂ and COS hydrolysis, thereby maximizing sulfur recovery efficiency, reducing operational cost, and mitigating environmental issues.

Surface modification with defect engineering, such as the introduction of cation or anion vacancies, the creation of lattice distortion and so on, may tune the surface electronic properties to optimize the chemical adsorption and enrich the active sites, and thus enhance the catalytic activity (9–12). Among these strategies, oxygen vacancy (V_O) engineering with low formation energy attracts the most attention (13–16). For example, the introduction of V_O into CoFe₂O₄ nanosheet led to reduction of adsorption energy from −0.56 eV to −1.01 eV, thus enabling H₂O adsorption on the surface and improving overall water splitting activity (17). It was reported that doping transition metal (Fe, Co, Ni) can form asymmetric oxygen vacancies in CeO₂, which accelerated the dissociation of H₂O molecule and generation of active hydroxyls, contributing to promoted COS hydrolysis activity (18). Despite promising results of catalytic performance obtained with V_O engineering strategy, experimental proof and/or in-depth mechanistic understanding of the very nature for V_O promotion in H₂O molecule activation are still lacking, which is vital for further rational design of advanced catalysts.

Herein, considering the extraordinary tunability of perovskite and the superior sulfur resistance of titanium-based materials, titanium-based perovskite with abundant V_O was fabricated by adjusting A site cation with different radius (γ_A: Ca < Sr < Ba) and further vacancy engineering including plasma and H₂ reduction treatment. Experimental results determined that H₂O activation behavior was regulated by the different amount and type of V_O, in which the very nature for the promotion role of V_O was further interpreted via DFT calculations. As expected, enhanced lower-temperature hydrolysis performance was obtained over SrTiO₃ after H₂ reduction treatment, achieving 100% COS and CS₂ conversion at 225 °C, which attained to the optimized temperature for the Claus conversion. More importantly, the origin of enhanced hydrolysis activity from boosted H₂O activation by V_O was revealed via in-depth mechanism investigation. It is believed that with the nature for V_O promotion in H₂O activation elucidated, the design of efficacious catalysts for H₂O-involved chemical reactions would become more explicit.

Results

Characterization of MTiO₃ and Vacancy-Engineered SrTiO₃ Catalysts.

MTiO₃ (M = Ca, Sr, Ba) perovskite catalysts with A site cation having different radius (γ_A: Ca < Sr < Ba) were synthesized by a hydrothermal route, which are labeled as CTO, STO, and BTO, respectively. The perovskite phases are clearly identified for all catalysts via XRD and Raman spectra (SI Appendix, Figs. S1 and S2), and no diffraction peak nor Raman band of other species, such as MO_x, TiO₂, and MCO₃, is detected. Other structural and morphology features are provided and discussed in SI Appendix, Figs. S3 and S4 and Table S1). The surface chemistry of the catalysts was analyzed by XPS. As shown in Fig. 1A, the O 1s XPS spectra are characterized by three peaks including lattice oxygen (529.3 eV, O_lat: O²⁻), dissociative oxygen on the oxygen vacancies (531.0 eV, O_v: O^–, O₂^–, O₂²⁻), and surface-adsorbed molecular water (532.5 eV, O_ads) (19, 20). Considering that the presence of O_v is closely related to V_O, the existence of V_O (g = 2.003) is confirmed by EPR (Fig. 1B) (21, 22). Meanwhile, trace of Ti³⁺ with a weak signal at g = 1.976 is detected, which is induced by V_O (23, 24). Furthermore, the desorption peaks below 350 °C in O₂-TPD-MS profile (SI Appendix, Fig. S5) also provide worthy clues regarding V_O (25, 26). As depicted in EPR and O₂-TPD-MS profile, the amount of V_O varies with different A site cation. Knowing that XPS is a surface-sensitive technique, the V_O ratio = O_v/ (O_lat+ O_v+ O_ads) area ratio (SI Appendix, Table S2) is taken to further determine the amount of V_O. Accordingly, the V_O ratios are 0.15, 0.20, and 0.18 for CTO, STO, and BTO, respectively. The remarkably high V_O ratio of STO reveals that it has the highest amount of inherent oxygen vacancies. Hence, the above results prove that there are abundant surface oxygen vacancies in the synthesized perovskite catalysts, especially in STO.

Fig. 1.

Characterizations and identification of V_O. O 1s XPS spectra (A) and EPR spectra (B) of MTiO₃ catalysts; O 1s XPS spectra (C), FT-EXAFS spectra at Ti K-edge (D), and HRTEM images and the corresponding IFFT image (E) of STO catalysts.

Consequently, vacancy engineering including plasma treatment and H₂ reduction treatment were adopted to modify STO to create more V_O, denoted as STO-P and STO-R, respectively. The crystal structure of SrTiO₃ (SI Appendix, Fig. S7 and Table S3) is well maintained, neither plasma treatment nor H₂ reduction treatment has significant effect on SrTiO₃. According to XPS results (Fig. 1C and SI Appendix, Table S2), the V_O ratio calculated from the O 1s spectra follows the order of STO (0.20) < STO-P (0.37) < STO-R (0.41), implying that both H₂ reduction and plasma treatment are effective methods to create more V_O. This is also supported by remarkably enhanced EPR signal of V_O and Ti³⁺ in STO-R (SI Appendix, Fig. S8). Besides, compared with STO, an evident shift (~0.1 eV and 0.2 eV) to smaller binding energies is observed in Ti 2p spectra of STO-P and STO-R (SI Appendix, Fig. S9), respectively, which is due to the change of chemical environment of Ti atoms in subsurface, i.e., the number of coordinating oxygen atoms is reduced by V_O formation. XAFS spectra were employed to further figure out the chemical environment of Ti atoms. As shown in Fig. 1D, the EXAFS peak around 1.95 Å is attributed to Ti-O coordination and the second peak at 3.30 Å arises from Ti-Sr coordination. Moreover, based on EXAFS analysis (SI Appendix, Table S4), the coordination number of Ti-O decreases apparently in STO-P and STO-R, confirming the lower coordination of Ti atoms induced by V_O formation (27, 28). HRTEM images and corresponding inverse fast Fourier transform (IFFT) images were displayed to intuitively demonstrate the presence of V_O (Fig. 1E) (29). The interplanar lattice distances of ~0.276 nm are assigned to the (110) crystal plane of SrTiO₃. Visibly, some parts of the lattice images become blurred and some lattice fringes bend in STO-P and STO-R, which are ascribed to the local distortion of SrTiO₃ lattice structures deriving from the absence of oxygen atoms (30, 31). PALS was performed to gain more detailed information about the vacancy defects and the fitting results are summarized in SI Appendix, Table S5. Three positron lifetime components, τ₁, τ₂, and τ₃, are attributed to small size defects in the form of oxygen monovacancy (mono V_O), larger size defects formed by V_O clusters (V_O clusters), and large voids, respectively (32, 33). Obviously, the mono V_O accounts for more proportion than that of V_O clusters in all catalysts due to I₁> I₂. Besides, increased value of I₁/I₂ suggests that more mono V_O is generated after H₂ reduction and plasma treatment.

Promotion Role of V_O for H₂O Activation.

To evaluate the modulation of oxygen vacancies on H₂O molecule activation, in situ DRIFTS study (Fig. 2A) was conducted to investigate the H₂O molecule activation behavior on these catalysts. The bands at 1,625 cm⁻¹ and 3,691 cm⁻¹ correspond to the adsorbed H₂O molecule and surface -OH generated from H₂O molecule dissociation, respectively (34). The band intensity ratio of surface -OH/adsorbed H₂O (I_OH, Fig. 2B) is calculated to measure the H₂O dissociation degree, which follows the sequence of CTO (0.40) < BTO (0.54) < STO (0.71) < STO-P (0.99) < STO-R (1.03). Moreover, D₂O isotope exchange experiment was performed. Since H₂O would undergo adsorption, dissociation, and desorption on catalyst surface, the generation of DHO (m/z = 19) is intensively associated with the degree of H₂O dissociation. As recorded in Fig. 2C, the desorption amount of DHO follows the same variation order as the order of I_OH, wherein STO-P and STO-R release much more DHO, suggesting that V_O contributes to H₂O molecule activation. Furthermore, the V_O ratio obtained by XPS analysis is correlated with the I_OH. A linear correlation between the two parameters is observed both for A site cation adjusting and vacancy engineering (Fig. 2D), which perspicuously indicates that the V_O amount poses positive effect on H₂O molecule activation.

Fig. 2.

Promotion role of V_O in H₂O activation—experimental study. In situ DRIFTS spectra of H₂O adsorption (A), band intensity ratio of surface -OH/adsorbed H₂O (I_OH) (B), (C) MS signal of DHO, and (D) I_OH as a function of V_O ratio.

DFT calculations were employed to elucidate the nature of V_O promotion in H₂O molecule activation. Based on the above characterization results, SrTiO₃ (001) slabs via removing different amount of oxygen atoms were constructed to simulate catalyst surface with different amount and type of vacancy defect (Fig. 3 A–D) (35, 36). As shown in Fig. 3E, the V_O formation energy of generating second and third V_O to form V_O clusters is always higher than that of disperse mono V_O, indicating the easier formation of mono V_O, which is responsible for the predominance of mono V_O as detected by PALS. The interaction between a H₂O molecule and these SrTiO₃ (001) slabs and the corresponding energy change are illustrated in SI Appendix, Fig. S13 and Fig. 3F, respectively. As shown, H₂O molecule trends to adsorb and dissociate on low-coordinated Ti ions adjacent to V_O (like site 1), resulting in surface -OH and O_latH, wherein the former derives from H₂O trapped by V_O and the latter results from H atom transfering to adjacent O atom. The energy barrier for H₂O dissociation is 0.60, 0.47, 0.46, and 0.15 eV for Slab1, Slab2, Slab3, and Slab4, respectively, suggesting that surface rich in V_O, especially V_O clusters, is energetically more favorable for H₂O adsorption and dissociation. Additionally, the resulted surface -OH and O_latH would further accelerate subsequent H₂O adsorption and dissociation (SI Appendix, Fig. S14), which is verified by the energy change after introducing the second H₂O molecule on Slab3 and Slab4 (Fig. 3F). Moreover, the projected density of states was analyzed to further study the interaction mechanism of H₂O adsorption (SI Appendix, Fig. S15 and Table S6). O 2p orbitals of H₂O molecule is significantly overlapped with Ti 3d orbitals of the four slabs, and both the p-band center of O atoms and d-band center of Ti atoms shifts to lower energy level with the presence of more V_O, especially V_O clusters, which confirms the strong H₂O molecule adsorption on defective surface (37–39). Bader charge depicts a positive electron transfer from H₂O molecule to those slabs, indicating that there are electrons filling in the antibonding orbitals, which is not conducive to bonding. Hence, the minimum electron transfer in the Slab4 with V_O clusters is responsible for its strongest bonding with H₂O molecule. In short, these results illustrate that the more V_O contribute to the higher level of H₂O activation, which is expected to facilitate hydrolysis catalytic activity.

Fig. 3.

Nature of V_O promotion in H₂O activation—theoretical calculation. Constructed SrTiO₃ (001) Slabs: (A) Slab1 with perfect surface, (B) Slab2 with mono V_O (V_Ob1), (C) Slab3 with three disperse mono V_O (V_Oc1- V_Oc3), (D) Slab4 with V_O clusters (V_Od1- V_Od3). Sites 1 and 2 in the slabs represent adsorption sites. Green, blue, and red represent Sr, Ti, and O atoms, respectively. (E) Formation energy of different V_O. (F) The energy change for first and second H₂O molecule dissociation on slabs.

Enhanced Hydrolysis Activity and Claus Conversion at Lower Temperature.

Catalytic performance for COS and CS₂ hydrolysis was evaluated over MTiO₃ catalysts and SrTiO₃ catalysts. As shown in Fig. 4A, complete conversion of COS is obtained on all MTiO₃ catalysts in the investigated temperature range. Despite of difficulty for CS₂ hydrolysis, complete conversion (Fig. 4B) and long-time stability lasting for 40 h (SI Appendix, Fig. S17) is achieved at 250 °C on STO rich in V_O. Notably, a positive linear relationship (R² = 0.92) between hydrolysis rate at 200 °C and V_O ratio of MTiO₃ catalysts is present (Fig. 4C), which indicates the crucial role of V_O in the catalytic hydrolysis reaction. What’s more, the promoting role of V_O is further verified by catalytic activity over SrTiO₃ catalysts. Hydrolysis performance of CS₂ over STO-P and STO-R (Fig. 4D) is efficiently enhanced, particularly at lower temperature, wherein CS₂ conversion increases from 83.7 to 94.0% and 95.0% at 200 °C, respectively. More prominently, complete CS₂ conversion is attained at 225 °C. The correlation of hydrolysis rate with V_O ratio at 200 °C (Fig. 4E) confirms that the increased amount of V_O accounts for the improved CS₂ hydrolysis activity.

Fig. 4.

Enhanced hydrolysis activity at lower temperature. Catalytic performance for COS (A) and CS₂ hydrolysis (B), and (C) hydrolysis reaction rate as a function of V_O ratio over MTiO₃ catalysts. Catalytic performance for CS₂ hydrolysis (D) and hydrolysis reaction rate as a function of V_O ratio (E) over STO catalysts. (F) Catalytic performance for COS and CS₂ hydrolysis over STO-R under Claus condition.

More importantly, this appealing hydrolysis efficiency even maintains under the Claus condition, i.e., a harsh condition with co-existed H₂S and SO₂ (40). As depicted in Fig. 4F, complete hydrolysis of COS and CS₂ is realized at 225 °C on V_O-rich STO-R and no obvious decrease of CS₂ conversion is observed with a durability of 30 h (SI Appendix, Fig. S18). Particularly, this temperature is favorable for the Claus conversion, and thus both satisfactory COS and CS₂ hydrolysis performance and improved sulfur recovery efficiency can be obtained simultaneously at one condition, which would simplify the operation process and improve energy efficiency (as shown in SI Appendix, Fig. S19). A remarkable advance both in reaction temperature and long-time stability compares with the catalysts reported in the literature (SI Appendix, Table S7). Amazingly, apart from the advanced hydrolysis activity, the catalyst also reveals excellent Claus catalytic activity (SI Appendix, Fig. S20), which might be ascribed to the basic sites in the catalyst (SI Appendix, Fig. S21). These results sufficiently manifest the successful construction of efficient hydrolysis catalysts through facile vacancy engineering.

Origin of Enhanced Hydrolysis Activity from the Boosted H₂O Activation by V_O.

In situ DRIFTS-MS experiments were performed on STO to further unveil the role of surface -OH and the hydrolysis reaction mechanism in Operando. Reactants are introduced into the cell in sequence of CS₂→N₂→H₂O or H₂O→N₂ →CS₂, corresponding to adsorption, purge, and reaction stage, respectively. In the experiment of CS₂→N₂→H₂O (Fig. 5A), gaseous CS₂ (1,526 cm⁻¹ and 1,541 cm⁻¹) and xanthate (COS₂²⁻, 1,151 cm⁻¹ and 1,235 cm⁻¹) arose from dissociated adsorption CS₂ are appeared in CS₂ adsorption stage (41). Subsequently, gaseous CS₂ is swept by N₂ purging. Then, H₂O is added to react with the xanthate. Neither -SH (2,160 cm⁻¹) in infrared spectroscopy nor MS signal of H₂S in the outlet gas (Fig. 5C, Curve CS₂→H₂O) is detected, which implies that neither dissociative adsorbed -OH nor gaseous H₂O can react with dissociative adsorbed COS₂²⁻ (42).

Fig. 5.

Origin of enhanced hydrolysis activity from the boosted H₂O activation by V_O. In situ DRIFTS spectra of the successive adsorption and reaction for CS₂→N₂→H₂O (A) and H₂O→N₂→CS₂ (B); (C) The MS signal of H₂S (m/z 34) in the outlet gas during the reaction stage.

In another experiment (H₂O→N₂→CS₂, Fig. 5B), H₂O molecule is effectively activated and dissociated into surface -OH in H₂O adsorption stage. Similarly, gaseous H₂O is then purged by N₂. In reaction stage, xanthate gradually accumulates, while the surface -OH wears off and -SH groups arise in infrared spectroscopy. Meanwhile, the MS signals of H₂S are detected in the outlet gas (Fig. 5C, Curve H₂O→CS₂). Apparently, CS₂ hydrolysis is initiated from the reaction between dissociative adsorbed -OH and gaseous CS₂. Namely, CS₂ hydrolysis follows the Eley–Rideal mechanism, which is further confirmed by the uninfluential impact of excess H₂O concentration on CS₂ hydrolysis performance (SI Appendix, Fig. S24) (43). Therefore, the activation of H₂O molecule to generate -OH is the first and vital step of hydrolysis reaction, which is the origin of the enhanced hydrolysis activity from the boosted H₂O activation by oxygen vacancies.

Conclusions

In summary, titanium-based perovskite with boosted H₂O activation and enhanced hydrolysis activity were achieved via a facile V_O engineering strategy. Experimental results indicate that V_O amount poses positive effect on H₂O activation. DFT calculations demonstrate that low-coordinated Ti ions connected with V_O are active sites for H₂O activation, and surface rich in V_O, especially V_O clusters, is energetically favorable for H₂O dissociation to form surface -OH and O_latH, which would further accelerate subsequent H₂O activation. Thus, STO-R with the highest V_O amount enabled COS and CS₂ hydrolysis and achieved 100% conversion at 225 °C, which attained to optimized temperature for the Claus conversion, wherein both efficient COS and CS₂ hydrolysis and improved sulfur recovery efficiency can be obtained concurrently. Furthermore, the origin of the enhanced hydrolysis activity from the boosted H₂O activation by V_O was attributed to the key role of -OH in initiating hydrolysis reaction. This study provides in-depth insight into V_O promotion in H₂O activation, which is expected to inspire more research in development of efficient materials for H₂O-involved reactions.

Materials and Methods

Chemicals.

Chemicals including Ca(NO₃)₂·4H₂O, Sr(NO₃)₂, Ba(NO₃)₂, C₁₂H₂₈O₄Ti, and NaOH were purchased from Sinopharm Chemical Reagent Co., Ltd. All the reagents were of analytical grade and used without further purification.

Synthesis of MTiO₃ (M = Ca, Sr, Ba) Catalysts.

0.03 mol alkaline earth metal nitrate was firstly dissolved in 100 mL of deionized water. And then 0.03 mol C₁₂H₂₈O₄Ti and 0.06 mol NaOH were added orderly under vigorous stirring. After stirring for 1 h, the acquired solution was transferred into a 180-mL Teflon-lined stainless-steel autoclave. Next the autoclave was subjected to hydrothermal treatment at 200 °C for 24 h. The obtained precipitate was separated by centrifugation, washed with deionized water and ethanol several times, and then dried at 120 °C overnight. The ground precursor was finally calcined in air at 650 °C for 6 h. The obtained CaTiO₃, SrTiO₃, and BaTiO₃ catalysts were denoted as CTO, STO, and BTO, respectively.

Vacancy Engineering Treatment of SrTiO₃.

1g of the abovementioned STO was further etched with plasma (10% H₂/Ar, v/v) for 30 min and marked as STO-P. 1g of the abovementioned SrTiO₃ precursor (before calcination) was calcined in a tube furnace at 650 °C for 6 h under gas stream of 100 mL min⁻¹ 10% H₂/Ar (v/v) and labeled as STO-R.

X-ray Absorption Spectra Collection and Data Processing.

The X-ray absorption fine structure (XAFS) spectra over the Ti K-edge of SrTiO₃ catalysts were collected on the beamline 1W1B station in Beijing Synchrotron Radiation Facility at ambient temperature. The radiation was monochromatized by a Si (111) double-crystal monochromator. The EXAFS data were analyzed by Athena module implemented in the IFEFFIT software packages, and relevant quantitative structural parameters were obtained by Artemis module of IFEFFIT software packages.

In Situ DRIFTS-MS Experiments.

In situ DRIFTS experiments were performed in in situ cell with glass sheet made of BaF₂. The diffuse reflectance infrared spectra, from 1,000 cm⁻¹ to 4,000 cm⁻¹, were recorded via an infrared spectrometer (Bruker, TENSOR II) equipped with an Mercury-Cadmium-Telluride detector cooled with liquid nitrogen. The outlet gas of in situ cell was analyzed by a mass spectrometer (Hiden, HPR-20 R&D). Prior to each experiment, the catalysts were pretreated at 500 °C for 1 h in N₂ flow of 50 mL min⁻¹. After cooling to the desired temperature (200 °C for H₂O adsorption experiments and 300 °C for successive adsorption and reaction experiments), the spectrum was recorded as the background. Then introduce the reaction gas to test. Before the formal experiment, preliminary experiments about the separate adsorption of several representative reactant and product molecules were proceeded to accurately determine the peak positions of the adsorbed species that may appear during the hydrolysis reaction.

Method for DFT Calculations.

All the DFT calculations are performed by the Vienna Ab initio Simulation Package with the projector augmented wave method (44, 45). The exchange-functional is treated using the generalized gradient approximation with Perdew–Burke–Emzerhof functional (46, 47). The energy cutoff for the plane wave basis expansion was set to 400 eV. Partial occupancies of the Kohn−Sham orbitals were allowed using the Gaussian smearing method and a width of 0.2 eV. The Brillourin zone was sampled with Monkhorst mesh of 4 × 4 × 4 for the optimization for the bulk structure of SrTiO₃. According to the experimental results, bare SrTiO₃ (100) surface was built with the termination of Ti-O. The Monkhorst mesh of 2 × 2 × 1 was used in all the surface structure calculations. The energy convergence was set to 10⁻⁵ eV, and the force convergency was set to 0.05 eV Å⁻¹. The free energy corrections were considered at the temperature of 473 K, following Eq. 1:

$$\mathrm{\Delta G}=\mathrm{\Delta E}+{\mathrm{\Delta G}}_{\mathrm{ZPE}}+{\mathrm{\Delta G}}_{\mathrm{U}}-\mathrm{T\Delta S},$$ [1]

wherein the ΔE, ΔG_ZPE, ΔG_U, and ΔS refer to the DFT calculated energy change, the correction from zero-point energy, the correction from inner energy, and the correction from entropy, respectively (48).

Catalytic Performance Measurements.

The COS and CS₂ hydrolysis activity tests were evaluated independently with a fixed-bed quartz tube microreactor (inner diameter = 6 mm). 0.5 mL catalyst (40 to 60 mesh) was used to hydrolyze organic sulfur with water in 50 mL min⁻¹ reactant gas, giving a gas hourly space velocity of 3,000 h⁻¹. Water was brought in through N₂ bubbling, and its concentration was controlled via adjusting the flow rate of N₂. The reaction stream consisted of 0.2 vol.% CS₂ (or 0.5 vol.% COS), 1.2 times of H₂O (against to concentration of organic sulfur), and balanced with N₂. The catalytic performance was evaluated under various reaction temperature in sequence, and each temperature point was kept for 5 h. The average value of the experimental data obtained at last 1 h was taken as the activity data. For durability performance, the reaction temperature is fixed at a certain temperature. The composition and concentration of the effluent stream was analyzed by gas chromatography (Agilent, 7890B) with thermal conductivity detector and flame photometric detector. The tail gas of reaction evaluation system is introduced into an absorption bottle with ethanol and potassium hydroxide aqueous solution for harmless treatment.

Besides, for organic sulfur hydrolysis performance under Claus condition, the reactant stream was the mixture of 0.2 vol.% CS₂ (or 0.5 vol.% COS), 30 vol.% H₂O, 0.5 vol.% H₂S, 0.25 vol.% SO₂, and balance gas N₂ with a gas hourly space velocity of 1,500 h⁻¹. As for catalytic performance for the Claus conversion, the reactant was composed of concentration of 0.5 vol.% H₂S, 0.25 vol.% SO₂, and balance gas N₂ with a space velocity of 1,500 h⁻¹.

The COS conversion was calculated according to the following Eq. 2:

$$\begin{array}{c}\mathrm{COS} \text{conversion}=\frac{{\left[\mathrm{COS}\right]}_{\mathrm{In}}-{\left[\mathrm{COS}\right]}_{\mathrm{Out}}}{{\left[\mathrm{COS}\right]}_{\mathrm{In}}}\ast 100\%.\hfill \end{array}$$ [2]

The CS₂ conversion was calculated according to the following Eq. 3:

$$\begin{array}{c}{\mathrm{CS}}_2 \text{conversion}=\frac{{\left[{\mathrm{CS}}_2\right]}_{\mathrm{In}}-{\left[{\mathrm{CS}}_2\right]}_{\mathrm{Out}}}{{\left[{\mathrm{CS}}_2\right]}_{\mathrm{In}}}\ast 100\%.\hfill \end{array}$$ [3]

The Claus conversion was calculated according to the following Eqs. 4 and 5:

$$\begin{array}{c}{\mathrm{SO}}_2 \text{conversion}=\frac{{\left[{\mathrm{SO}}_2\right]}_{\mathrm{In}}-{\left[{\mathrm{SO}}_2\right]}_{\mathrm{Out}}}{{\left[{\mathrm{SO}}_2\right]}_{\mathrm{In}}}\ast 100\%,\hfill \end{array}$$ [4]

$$\begin{array}{c}{\mathrm{H}}_2\mathrm{S} \text{conversion}=\frac{{\left[{\mathrm{H}}_2\mathrm{S}\right]}_{\mathrm{In}}-{\left[{\mathrm{H}}_2\mathrm{S}\right]}_{\mathrm{Out}}}{{\left[{\mathrm{H}}_2\mathrm{S}\right]}_{\mathrm{In}}}\ast 100\%.\hfill \end{array}$$ [5]

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.