Introduction

Gas phase per- and polyfluoroalkyl substances (PFAS) are among the most persistent environmental pollutants, owing to their exceptional chemical inertness, high ecological risk, and substantial climate impact1,2. Among them, tetrafluoromethane (CF4) is particularly concerning due to its extremely high global warming potential (GWP), ~7390 times higher than that of CO2, coupled with an exceptionally long atmospheric lifetime exceeding 50,000 years3,4. Major CF4 emissions arise from industrial activities such as aluminum electrolysis and semiconductor manufacturing, where it is an unavoidable byproduct5,6,7,8. In response to the environmental risks posed by CF4, the European Union implemented the Carbon Border Adjustment Mechanism (CBAM) in 2022, mandating stringent emission control9. Therefore, developing efficient, energy-saving, and sustainable technologies for CF4 decomposition is crucial for achieving global climate mitigation goals.

Catalytic hydrolysis has emerged as a highly promising approach for CF4 decomposition, offering notable advantages in terms of high reaction rates, industrial scalability, and compatibility with existing infrastructure3. Nevertheless, the challenge lies in breaking the exceedingly strong C–F bonds, which have an extremely high bond dissociation energy of approximately 543 ± 4 kJ mol−1 and confer exceptional stability to CF44,10,11. Significant efforts have been dedicated to develop catalysts capable of activating and cleaving these robust bonds under relatively mild conditions. A growing body of evidence underscores the critical role of surface hydrogen species (i.e., protons) in this process12,13,14. For instance, Chen et al. demonstrated that the interaction of surface protons with C–F bonds significantly lowered the activation barrier and reaction temperature for CF4 decomposition15. Zhang et al. advanced this approach by employing Ga–OH groups as defluorination sites, achieving remarkable stability over 1000 h16. Similarly, Luo et al., utilizing constrained ab initio molecular dynamics (cAIMD), confirmed the essential role of surface hydroxyl groups in CF4 hydrolysis17. Collectively, these findings emphasize that proton availability not only facilitates C–F bond activation but also plays a key role in regenerating fluorinated active sites, thus ensuring sustained catalytic performance. Despite these advances, enhancing proton supply during hydrolysis remains a significant bottleneck. Traditional proton sources often suffer from thermal instability or rapid desorption at elevated temperatures, especially under industrial reaction conditions. This highlights the urgent need for new strategies that can enable persistent proton availability while maintaining high-temperature stability.

Sulfur dioxide (SO2), typically regarded as a catalyst poison due to its strong adsorption affinity, is frequently co-emitted with CF4 in industrial flue gases, particularly from aluminum electrolysis processes5,6,7. Intriguingly, under hydrolysis-relevant conditions, SO2 readily forms strongly acidic and thermally stable surface species, such as hydrogen sulfite or bisulfate (–HSO3/–HSO4)18,19. These SO2-derived surface species possess high acidity, robust thermal stability, and low volatility, distinguishing them significantly from conventional proton sources which typically suffer from rapid desorption or thermal decomposition at elevated temperatures20,21,22,23. Therefore, the formation of these species on catalyst surfaces could substantially increase local proton concentrations near active catalytic sites, thus potentially overcoming existing limitations related to proton scarcity in conventional CF4 catalytic hydrolysis process.

In this work, we report a novel strategy that leverages SO2-driven in situ formation of surface proton-supplying sites to significantly enhance proton availability and thus catalytic CF4 hydrolysis. Through detailed in situ spectroscopic analyses and X-ray photoelectron spectroscopy (XPS) measurements, we identify the formation of Al–HSO4 and Ga–HS surface sites, which substantially enhance H2O dissociation and proton availability by factors of 6 and 10, respectively, compared to systems without SO2. Furthermore, comprehensive theoretical and experimental investigations reveal that these proton-supplying sites not only reduce the activation energy for C–F bond cleavage but also accelerate defluorination of active sites, thereby mitigating fluorine poisoning and enhancing catalyst durability. As a result, our optimized system achieves complete CF4 decomposition at a notably low temperature of 550 °C (comparing to the normal temperature of 700 °C) with exceptional operational stability exceeding 2500 h4. This work introduces a novel framework for in situ proton regulation and opens new possibilities for the efficient, long-term catalytic degradation of PFAS under practical industrial conditions.

Results

CF4 catalytic hydrolysis performance

To evaluate the promotional effect of SO2 on CF4 catalytic hydrolysis, we selected a Ga/θ-Al2O3 catalyst with 30 mol% Ga doping, chosen for its structural stability and abundance of active sites16. Detailed procedures for catalyst synthesis and characterization are provided in the Methods section.

We first examined the influence of SO2 concentration by co-feeding CF4 with SO2 at varying levels to simulate realistic industrial exhaust conditions (Supplementary Fig. 1). The reaction temperature was maintained at 550 °C, significantly lower than that used in most previously reported CF4 decomposition system. Remarkably, the introduction of SO2 dramatically enhanced CF4 decomposition efficiency, increasing from ~60% (without SO2) to ~90% across an ultra-broad SO2 concentration range (250–20,000 ppm, Fig. 1a and Supplementary Fig. 2). This robust performance across more than three orders of magnitude demonstrates the catalyst’s wide applicability to various industrial SO2/CF4 ratios. Notably, at 5000 ppm SO2, complete CF4 decomposition was achieved at 550 °C, a record low temperature for 100% CF4 decomposition. Hence, 5000 ppm SO2 was chosen as the standard condition for subsequent evaluations.

Fig. 1: Catalytic stability and performance characterizations.
figure 1

a CF4 decomposition (%) during the CF4 and SO2 synergistic reaction at different SO2 concentration (250–20,000 ppm). b CF4 decomposition (%) during synergistic/separate reaction at different reaction temperatures, respectively. c Arrhenius plots for CF4 and SO2 during synergistic/separate reaction (“With SO2” and “Without SO2” denote CF4 hydrolysis performed under conditions with and without SO2, respectively. “With CF4” and “Without CF4” denote SO2 oxidation performed under conditions with and without CF4, respectively). d CF4 decomposition (%) during catalytic reaction at different experimental conditions. e Stability test under 550 °C for the CF4 and SO2 synergistic reaction. f Comparison of the CF4 complete decomposition temperature and stability with the reported results.

Full size image

The effect of temperature was further investigated across a range from 425 to 600 °C (Fig. 1b and Supplementary Fig. 3). In all cases, SO2 addition significantly boosted the reaction rate compared to CF4 decomposition alone, with the most pronounced enhancement observed at lower temperatures. Arrhenius analysis revealed that SO2 addition lowered the apparent activation energy for CF4 decomposition from 137.2 kJ/mol to 114.4 kJ/mol, a 17% reduction (Fig. 1c and Supplementary Figs. 4, 5). Interestingly, CF4 also promoted SO2 oxidation, reducing its apparent activation energy from 23.5 kJ/mol to just 5.3 kJ/mol, a 77% decrease. These results highlight a mutual and synergistic promotion between CF4 decomposition and SO2 oxidation pathways.

To further probe this synergistic effect, we conducted transient feed experiments involving three sequential gas feeding protocols at 550 °C (Fig. 1d). When CF4 was introduced alone, the decomposition efficiency remained below 60%. Strikingly, upon addition of SO2, CF4 decomposition rapidly increased to >99% (Fig. 1d, bottom), clearly demonstrating the promotional effect of SO2. In contrast, SO2 oxidation alone exhibited a modest conversion efficiency of ~75%; yet, co-feeding CF4 led to full conversion of both gases (>99%, Fig. 1d, top and Supplementary Figs. 6–8). Notably, this synergistic enhancement was equally evident when CF4 and SO2 were introduced concurrently from the outset (Fig. 1d, middle), confirming the strong and consistent mutual promotion effect between the two species.

To assess the long-term practical viability of this system, we conducted extended stability tests at 550 °C (Fig. 1e and Supplementary Table 1). Remarkably, the catalyst exhibited outstanding durability, maintaining nearly complete CF4 and SO2 conversion over 2500 h without any detectable deactivation, demonstrating the robustness and long-term operational viability of the system for potential industrial applications. Further, the calculated deactivation constant was 3.97 × 10−4 h−1, corresponding to an expected catalyst lifetime of ~2522 h, validating the long-term practical viability of this system. A comprehensive comparison with previously reported CF4 decomposition systems (Fig. 1f and Supplementary Table 2) confirms the superior performance of this system, which achieves the lowest operational temperature for full CF4 decomposition and the longest recorded lifetime under continuous flow conditions15,16,24,25,26,27,28,29,30,31,32. These advantages, high efficiency, excellent durability, and compatibility with SO2-containing gas streams, highlight the promise of this system for scalable and environmentally sustainable degradation of gas phase PFAS.

In situ formation of proton-supplying sites

To elucidate the role of SO2 in enhancing proton availability during CF4 catalytic hydrolysis, we employed a suite of characterization techniques to identify the in situ formation of proton-supplying species on the Ga/θ-Al2O3 catalyst surface. X-ray diffraction (XRD) patterns of both fresh and used Ga/θ-Al2O3 catalysts (Fig. 2a) displayed only the characteristic reflections of θ-Al2O3 (JCPDS#35-0121), with no additional diffraction peaks observed16, indicating no bulk sulfur species formed during the reaction. Transmission electron microscopy (TEM) and energy-dispersive X-ray (EDX) mapping of the used Ga/θ-Al2O3 catalyst (Fig. 2b) further confirmed this observation. The used catalyst retained its nanosheet structure, and sulfur species were uniformly distributed across the surface, verifying the absence of excessive sulfur accumulation. To probe the nature of sulfur-containing species formed under reaction conditions, in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was conducted (Fig. 2c). The results revealed new infrared features of –HSO4 (1000–1250 cm−1) and –HS (688 cm−1) species under anhydrous CF4 and SO2 co-feeding15,33,34. This confirmed the in situ formation of proton-supplying sites (–HSO4 and –HS) during the reaction.

Fig. 2: In situ formation of dual Brønsted acid sites.
figure 2

a XRD patterns of the fresh and used Ga/θ-Al2O3 catalysts (testing under 5000 ppm SO2 and 2500 ppm CF4 for 10 h). b TEM images and EDX mapping of the used Ga/θ-Al2O3 catalyst. c In situ DRIFTS of CF4 and SO2 synergistic reaction without H2O over Ga/θ-Al2O3 catalyst under 550 °C as a function of time. d, e XPS spectra of S 2p for the used Ga/θ-Al2O3 catalysts during the CF4 and SO2 synergistic reaction under different SO2 concentration (250–20,000 ppm) and after different treatments (“-used” represents the catalyst obtained after testing under SO2 and CF4; “-SO2 treated” represents the catalyst obtained after testing under SO2 only; “H2SO4-modified” represents the catalyst modified by 20% mol ratio of H2SO4). f The proportion of –HS and –HSO4 species, as well as the total amount of sulfur species on the catalysts surface as a function of SO2 concentration.

Full size image

To analyze the surface elemental composition, X-ray photoelectron spectroscopy (XPS) was performed on catalysts for CF4 and SO2 reactions with different SO2 concentrations (250–20,000 ppm, Fig. 2d and Supplementary Fig. 9). The results revealed clear S 2p signals characteristic of Ga–HS (~161.6 eV) and Al–HSO4 (~169.6 eV) species (Supplementary Fig. 10)35,36,37. The assignments were validated by comparison with standard samples and supported by thermogravimetric analysis (TG, Supplementary Fig. 11)15,38. Further comparison among catalysts with different treatments (Fig. 2e) revealed the following order in Al–HSO4 content: H2SO4-modified > SO2 treated > used catalyst, confirming the participation of –HSO4 species in the catalytic hydrolysis process. The absence of Ga–HS signals in the θ-Al2O3-used samples reinforced the conclusion that Ga–HS sites were formed exclusively in situ during the CF4 and SO2 synergistic reaction.

Figure 2f summarizes the quantifications of Ga–HS and Al–HSO4, along with the total sulfur content on the catalyst surface as a function of SO2 concentration. The ratios stabilized at ~75% (Ga–HS), ~25% (Al–HSO4), and ~3% (total S), respectively, confirming the stable in situ formation of proton-supplying sites without excessive sulfur species accumulation during the reaction.

To investigate the impact of proton-supplying sites on the catalyst’s performance, XPS and in situ DRIFTS analyses were conducted. F 1s spectra (Supplementary Fig. 12) revealed that fluorine species (AlFx), typically indicative of fluorine poisoning, were present only in CF4-alone-treated samples. In contrast, samples from CF4–SO2 reactions showed negligible AlFx signals, suggesting that proton-supplying sites effectively mitigated fluoride accumulation on active sites16. O 1s spectra revealed two major components: lattice oxygen (Olat, ~531.4 eV) and chemisorbed oxygen (Oads, ~533.0 eV)39,40. The CF4 + SO2-treated sample exhibited the highest Oads content, demonstrating minimal surface oxygen damage. Additionally, the Al 2p peak at ~75.9 eV, attributed to Al–OH (a key proton donor in CF4 hydrolysis), decreased significantly under CF4 or SO2 alone but increased significantly under the synergistic reaction, confirming the restoration of proton-donating groups via SO2-derived species.

In situ DRIFTS analysis (Supplementary Fig. 13) provided further insights into the reaction dynamics. During SO2 pre-adsorption, the depletion of Al–OH groups (3650–3750 cm−1) and the formation of HSO3− species (1200 and 966 cm−1) indicated that SO2 was activated via interaction with surface hydroxyls. When CF4 was co-fed with SO2, the intensities of both adsorbed SO2 and HSO3− bands decreased due to competitive adsorption. As the reaction temperature increased above 250 °C, the transition of HSO3− to SO42− (1371 and 995 cm−1) and HSO4− (1190 cm−1) confirmed further oxidation of sulfur intermediates. Concurrently, a monotonic decline in CF4 signals and a transient rise-and-fall of HSO4− bands were observed, demonstrating the dynamic role of HSO4− in facilitating C–F bond activation during hydrolysis.

Promotion effect of proton-supplying sites

Previous studies have demonstrated that protons can effectively promote C–F bond activation through strong interactions with C–F bond, thereby facilitating hydrolysis41,42,43,44. In conventional CF4 hydrolysis, protons are primarily supplied by the dissociation of H2O. To investigate the mechanism role of SO2 in enhancing proton availability, we performed time-resolved in situ DRIFTS on Ga/θ-Al2O3 catalyst at 550 °C under various reaction conditions (Fig. 3). In the case of SO2 alone (Supplementary Fig. 14), H2O dissociation was significantly inhibited. Similarly, for CF4 decomposition alone (Fig. 3a), no significant H⁺ bands were detected, suggesting that limited H+ availability hindered effective C–F bond cleavage. In contrast, when SO2 and CF4 were co-fed (Fig. 3b), distinct peaks corresponding to gaseous SO2 (1379 cm−1) and CF4 (1279 cm−1) appeared initially, followed by prominent H2O dissociation and generation of surface sulfate species. Notably, vibrational features in the 2900–3400 cm−1 region (Fig. 3c), assigned to protonic species, emerged rapidly, confirming in situ H+ generation from H2O dissociation in the presence of SO219,20.

Fig. 3: Al–HSO4 sites promote C–F activation.
figure 3

In situ DRIFTS of a CF4 hydrolysis and b CF4 and SO2 synergistic reaction over Ga/θ-Al2O3 catalyst under 550 °C with a function as time. c Comparison of H+ peak between solo and synergistic reaction. d The absorbance of H+ generating with a function as time.

Full size image

The temporal evolution of H+ generation, H2O dissociation and sulfate species formation are shown in Fig. 3d and Supplementary Fig. 15. During the SO2 and CF4 synergistic reaction, H⁺ generation and H2O dissociation rates were ~10 and 6 times higher, respectively, than in CF4-only hydrolysis. Moreover, sulfate formation was enhanced by ~3 times compared to SO2 oxidation alone. These results demonstrated that proton-supplying sites not only accelerate H2O activation but also enhance SO2 oxidation, thereby reducing the activation barrier for CF4 hydrolysis.

To evaluate the effect of surface Al–HSO4 sites on the CF4 adsorption, DFT calculations were performed for CF4 adsorbed at different sites on the θ-Al2O3 (010) and Ga/θ-Al2O3 (010) surface (Supplementary Figs. 16, 17 and Supplementary Table 3). These results indicated that the AlIII site was the primary adsorption site for CF4, and the effect of Ga doping on this site was negligible. The intrinsic stability of Al–HSO4 sites, as well as the influence of SO2 introduction on its structural stability, was further evaluated (Supplementary Figs. 18–20). These results demonstrated that Al–HSO4 sites was intrinsically stable, and the introduction of SO2 does not compromise its structural integrity. The CF4 adsorption energy (Eads) on θ-Al2O3–HSO4 was −0.50 eV, significantly stronger than that on θ-Al2O3–OH (−0.15 eV), confirming that Al–HSO4 sites significantly enhanced the CF4 adsorption affinity (Supplementary Fig. 21 and Supplementary Table 4).

To investigate the role of Ga–HS sites in defluorination and active sites regeneration, we first studied the defluorination kinetics of AlIII active sites using constrained ab initio molecular dynamics (cAIMD) simulations (Fig. 4a, b). The energy barrier for HF elimination from fluorinated Ga sites was dramatically reduced from 2.34 eV (with Ga–OH) to 0.32 eV with Ga–HS sites, demonstrating the critical role of Ga–HS proton-supplying sites in promoting defluorination and overcoming fluorine poisoning. In addition, the effect of SO2 introduction on the stability of the Ga–HS site and the regeneration of the Ga–HS site was further analyzed by DFT calculations (Supplementary Figs. 19, 22). The results indicated that the introduction of SO2 does not disrupt the structural integrity of the Ga–HS site, and the regeneration of the Ga–HS structure is feasible (an energy barrier of 0.98 eV).

Fig. 4: Ga–HS sites promote regeneration of active sites.
figure 4

a Time-dependent evolution and b reaction free energy profile for the regeneration of active sites with assistance of Ga–OH and Ga–SH sites. In situ DRIFTS of regeneration of active sites for c CF4 and SO2 synergistic reaction and d CF4 solo over Ga/θ-Al2O3 catalyst under 550 °C, respectively. e Al–OH relative intensity for CF4 solo and CF4 and SO2 synergistic reaction as function with time. In situ Raman spectra testing for f CF4 solo and g CF4 and SO2 synergistic reaction over Ga/θ-Al2O3 catalyst under 550 °C, respectively. h The regeneration rate of active sites on –OH sites and –SH sites.

Full size image

Furthermore, the regeneration behavior of active sites was investigated via in situ DRIFTS (Fig. 4c, d). In CF4-alone reactions (Fig. 4d), a rapid consumption of Al–OH was observed over time, indicating poisoning of Al active sites45. Upon H2O introduction, only partial signal recovery occurred. In contrast, under SO2 and CF4 conditions (Fig. 4c), Al–OH signals were similarly consumed, but fully regenerated after switching to H2O. Quantitative analyses of poisoning and regeneration of active sites (Fig. 4e) showed that Ga–OH sites facilitated only 49% regeneration of the active sites, while Ga–HS sites achieved complete regeneration. These results demonstrated that the in situ formed Ga–HS proton-supplying sites significantly enhanced the regeneration of fluorine-poisoned active sites.

In situ Raman spectroscopy testing was conducted to further evaluate the role of the proton-supplying sites in active site regeneration (Fig. 4f, g). For the fresh catalyst, a sharp peak at 248 cm−1 and a broad peak between 500 and 950 cm−1 were clearly observed, corresponding to the vibrations of Ga–O bonds and Al–O bonds, respectively46,47. Upon CF4 exposure, the Al–O signal diminished, indicating poisoning of Al active sites. Subsequent H2O exposure led to only partial recovery. In contrast, during the simultaneous introduction of SO2 and CF4, full regeneration of the Al–O signal was observed upon switching to H2O. Quantitative analysis of active site regeneration (Fig. 4h) revealed that complete regeneration was achieved with the assistance of Ga–HS proton-supplying sites, significantly outperforming Ga–OH (37% regeneration). These results confirmed that Ga–HS proton-supplying sites markedly enhanced the regeneration of fluorine-poisoned active sites.

Promotion mechanism

The remarkable enhancement in catalytic performance is attributed to a novel promotion mechanism involving the in situ formation of proton-supplying sites, as illustrated in Fig. 5. Upon introduction, SO2 is activated by H2O on the catalyst surface to form HSO3−, which subsequently undergoes a disproportionation reaction to generate sulfur species in +6 and −2 oxidation states. The +6 sulfur species bind with Al sites to form Al–HSO4 proton-supplying sites, while the −2 species coordinate with Ga to generate Ga–HS proton-supplying sites. For the energetically challenging activation of C–F bond, the in situ formed Al–HSO4 proton-supplying sites significantly promote C–F bond cleavage by enhancing the interactions between protons and C–F bonds, thereby boosting CF4 decomposition activity. Simultaneously, the in situ formed Ga–HS proton-supplying sites effectively assist in defluorination and the regeneration of Al–F sites, leading to the complete regeneration of fluorine-poisoned Al active sites and greatly improving catalyst stability.

Fig. 5: Schematic diagram.
figure 5

SO2-driven proton supply enables efficient and stable catalytic hydrolysis of CF4 via in situ formed proton-supplying sites.

Full size image

Thus, by introducing SO2 as a promoter, the formation of proton-supplying sites is induced in situ, which greatly enhances both H2O dissociation and proton availability within the reaction system. This dual functionality results in dramatically improved activity and stability for CF4 decomposition.

Discussion

In summary, we developed a novel catalytic strategy that leverages SO2-driven in situ formation of proton-supplying sites to achieve efficient and stable CF4 decomposition under low-temperature conditions. The introduction of SO2 into the reaction system leads to the formation of Al–HSO4 and Ga–HS species on the catalyst surface, which significantly enhance H2O dissociation and proton availability by factors of 6 and 10, respectively. These proton-supplying sites not only lower the energy barrier for C–F bond activation but also facilitate active site regeneration through defluorination, effectively overcoming catalyst deactivation caused by fluorine poisoning. As a result, complete CF4 decomposition was achieved at a record-low temperature of 550 °C with long-term stability exceeding 2500 h, substantially surpassing current state-of-the-art systems. Importantly, the CF4 and SO2 concentrations adopted in this study are consistent with those encountered in industrial aluminum electrolysis processes (Supplementary Table 4), highlighting the practical applicability of this strategy (Supplementary Fig. 23). This work provides fundamental insights into the proton regulation mechanism and offers a generalizable approach for the catalytic degradation of gas phase PFAS in industrial conditions.

Methods

Chemicals

All chemicals were obtained commercially and used as received. Aluminium isopropoxide (Al(C3H7O)3, 98.5%), Isopropyl alcohol (99.0%) and Gallium nitrate (Ga(NO3)3·xH2O, 99.9%) were purchased from Aladdin. Sulfuric acid (H2SO4) was purchased from Sinopharm.

Preparation of θ-Al2O3

θ-Al2O3 nanosheets were synthesized using a hydrothermal strategy. Specifically, 40.0 g of aluminum isopropoxide (AIP) was dissolved in 400 mL of isopropanol under continuous stirring to obtain a clear solution. Subsequently, 40.0 mL of deionized water was added dropwise to initiate the hydrolysis of AIP, followed by an additional 30 min of stirring. The resulting mixture was then divided equally into four 150 mL Teflon-lined stainless-steel autoclaves and subjected to hydrothermal treatment at 110 °C for 1 h. After naturally cooling to ambient temperature, the precipitates were dried at 80 °C for 12 h. The obtained solids were finely ground and calcined in a muffle furnace at 900 °C for 4 h with a heating rate of 1 °C min−1 to yield the final θ-Al2O3 product.

Preparation of Ga/θ-Al2O3

Ga-doped θ-Al2O3 (Ga/θ-Al2O3) catalysts were prepared via a conventional wet impregnation approach. In a typical procedure, 10.0 g of the previously synthesized θ-Al2O3 was dispersed in 400 mL of deionized water and subjected to ultrasonic agitation for 30 min to ensure uniform dispersion. An appropriate amount of gallium nitrate, corresponding to a Ga mol content of 30%, was then introduced into the suspension. The resulting mixture was ultrasonicated for an additional 30 min and subsequently concentrated using a rotary evaporator at 75 °C for 2 h. The obtained solid was dried and then calcined at 600 °C for 4 h in a tubular furnace to afford the final Ga/θ-Al2O3 catalyst.

Resource recycling

Specifically, the tail gas generated from the CF4 + SO2 reaction was first passed through 500 mL of an alkaline absorption solution containing 10 g L−1 NaOH. The concentrations of F- and SO42− ions in the resulting solution were determined by ion chromatography. To recover sulfate species, a small amount of dilute HCl solution was added to adjust the pH of the absorption solution to 7, followed by the addition of BaCl2 at a stoichiometric ratio of Ba:SO42− = 1:1 under ambient conditions. After stirring for 30 min, the resulting white precipitate was separated by centrifugation, washed three times with deionized water and three times with ethanol, and identified as BaSO4. The remaining solution was then heated in a water bath at 60 °C, and an excess amount of BaCl2 was added under stirring for 30 min. The newly formed white precipitate was collected and identified as BaFCl. The XRD patterns confirmed that both BaSO4 and BaFCl were obtained as nearly pure phases, verifying the effectiveness of the by-product recovery process.

Characterizations

XRD patterns were recorded on a Bruker D8 Focus diffractometer with Ni-filtered Cu-Ka (λ = 1.540598 Å) (40 kV, 40 mA) radiation in the 2θ range of 10–90° with a scan rate of 1°/min. TG was obtained on Perkin-Elmer Pyrisis from ambient temperature to 1000 °C under air. XPS measurements were obtained on Thermo Fisher Scientific Escalab 250 XI, and all the binding energies were calibrated by the C 1s peak at 284.8 eV. TEM images were obtained from FEI Tecnai G2 F20 field emission transmission electron microscope operated at 200 kV.

Catalytic activity evaluation

A self-made fixed-bed reactor (Supplementary Fig. 1) was used to evaluate the CF4 and SO2 synergistic removal. The CF4 and SO2 synergistic removal reaction was conducted in a continuous flow reaction system with a quartz fixed-bed reactor (20 mm i.d.) under atmospheric pressure. The gas mixture was composed of 2500 ppm of CF4, 250–20,000 ppm of SO2 and balance in Air. The total flow rate was 33.3 mL min−1, and the gas hourly space velocity (GHSV) was about 1000 mL g−1 h−1. The SO2 and CF4 conversion rates were calculated by the following equation:

$${{\mbox{CF}}}_{4}{{{\rm{decomposition}}}}(\%)=\frac{{\left[{{\mbox{CF}}}_{4}\right]}_{{\mbox{in}}}-{[{{\mbox{CF}}}_{4}]}_{{\mbox{out}}}}{{\left[{{\mbox{CF}}}_{4}\right]}_{{\mbox{in}}}}\times 100\%$$
(1)
$${{\mbox{SO}}}_{2}{\mbox{conversion}}(\%)=\frac{{[{{{{{\rm{SO}}}}}_{4}}^{2-}]}_{{\mbox{out}}}}{{\left[{{\mbox{SO}}}_{2}\right]}_{{\mbox{in}}}}\times 100\%$$
(2)

Where [CF4]in, [CF4]out, [SO2]in and [SO42−]out indicates the corresponding inlet and outlet gas concentrations and mass, respectively. SO32− and SO42− were detected by using a Thermo Scientific ICS-600 ion chromatograph system.

To quantitatively evaluate the catalyst stability, we calculated the deactivation constant (kd) based on the time-on-stream conversion data according to the first-order deactivation model:

$${k}_{d}\left({h}^{-1}\right)=\frac{{{\mathrm{ln}}}\left(\frac{1-{X}_{f}}{{X}_{f}}\right)-{{\mathrm{ln}}}(\frac{1-{X}_{i}}{{X}_{i}})}{t}$$
(3)

Xi is the initial decomposition (%) of CF4. Xf is the final decomposition (%) of CF4. t is the time throughout the reaction. τ (h): the catalyst life, τ = 1/kd.

In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS)

A Thermo Fisher iS50 spectrometer was employed for recording FTIR spectra. Catalysts were pretreated by heating to 550 °C (heating rate of 20 °C min−1) and holding at 550 °C for 2 h under Air flow (30 mL min−1). Water vapor was introduced via passing through a deionized water bottle. In situ DRIFTS was used to analyze the catalytic reaction mechanism.

In situ Raman spectroscopy

The Raman was conducted on inVia Reflex (Renishaw, UK), using a 532 nm laser at a power of 50 mW. Catalysts were pretreated by heating to 550 °C (heating rate of 20 °C min−1) and holding at 550 °C for 1 h under Air flow (30 mL min−1). For the poisoning process, the catalyst was treated with reaction gas (CF4 or CF4 + SO2) for 30 min; For the regeneration process, the deactivated catalyst was treated with water vapor for 30 min.

DFT computational details

All the first-principles calculations were performed using DFT as implemented in the Vienna ab initio simulation package (VASP. 5.4.4). The exchange–correlation potential is treated with the Perdew–Burke–Ernzerhof (PBE) formula using the projected augmented wave (PAW) method within the generalized gradient approximation (GGA). The cutoff energy for all calculations is set to 450 eV. All positions of the atoms were fully relaxed until the Hellmann–Feynman forces on each atom were less than 0.01 eV Å−1, thus ensuring that the atomic positions in the atomic model are optimized to the state with the smallest energy deviation. Meanwhile, a k-point Γ-centered mesh is generated for Brillouin zone samples for geometry optimization. The DFT-D3 method proposed by Grimme is applied to model the van der Waals interactions, and has been shown to accurately describe chemisorption properties. Throughout the geometry optimization process, the top two atomic layers of the supercell were permitted to relax, ensuring a realistic representation of the catalyst surface while maintaining computational efficiency. A vacuum region of 15 Å is employed to decouple the periodic replicas. In addition, the VESTA package was used to visualize the atomic structure and charge density.

The adsorption energy of –OH is defined as:

$${E}_{{ads}}={E}_{{total}}-{E}_{{slab}}-{E}_{{OH}}$$
(4)

where Etotal is the total energy of OH adsorbed on the surface, Eslab is the energy of clean surface, EOH is the energy of the –OH.

The adsorption energy of –HSO4 is defined as:

$${E}_{{ads}}={E}_{{total}}-{E}_{{slab}}-{E}_{{HSO}4}$$
(5)

where Etotal is the total energy of –HSO4 adsorbed on the surface, Eslab is the energy of clean surface, EHSO4 is the energy of the –HSO4.

The adsorption energies of H atom adsorption is defined as:

$${E}_{{ads}}={E}_{{total}}-{E}_{{slab}}-0.5*{E}_{H2}$$
(6)

where Etotal is the total energy of H atom adsorbed on the surface, Eslab is the energy of clean surface, EH2 is the energy of the H2 molecule.

cAIMD simulations

AIMD simulations are carried out via the Nose–Hoover thermostat using the canonical ensemble (NVT) at 550 °C, with a time step of 1 fs. Constrained ab initio molecular dynamics (cAIMD) simulations with a SG sampling approach as implemented in VASP (SG-AIMD) are performed to evaluate the kinetic barriers of Al ‒ F bond defluorination. In this method, one suitable collective variable (CV, namely ξ) can be defined as the reaction coordinate, which is linearly changed from the initial state to the final state with a transformation velocity ξ̇. The work required to perform the transformation from initial to final states can be computed as:

$$W={\int }_{\xi ({initial})}^{\dot{\xi }({final})}(\frac{\partial F}{\partial \xi })\cdot \xi {dt}$$
(7)

where F is the computed free energy, which is evolving along with t, can be computed along cAIMD using the blue-moon ensemble with the SHAKE algorithm. At the limit of infinitesimally small ∂ξ, the needed work (Winitial-to-final) corresponds to the free-energy difference between the final and initial states. In the SG sampling, a value ∂ξ of 0.001 Å is used for each cAIMD step after testing the shorter step size for the “slow-growth”.