CuO/ZnO Modified C₄N Monolayer for SF₆ Decomposition: Experimental Analysis and First-Principles Simulation

Abstract

Sulfur hexafluoride (SF₆) serves as an essential insulating and arc-extinguishing medium in power facilities, while it is an extremely potent greenhouse gas with an ultra-long atmospheric lifetime. To realize efficient medium-temperature harmless disposal of waste SF₆, two-dimensional stable C₄N monolayers were adopted as substrates to load ZnO and CuO nanoparticles for composite catalyst fabrication. XRD, TEM, and EDS characterizations confirmed the uniform dispersion of the metal oxides without destroying the C₄N two-dimensional skeleton. Comparative experiments proved that an NH₃ reducing atmosphere significantly accelerates SF₆ decomposition. The 3:1 CuO-modified catalyst achieved a maximum SF₆ conversion of 91%, and a CaO additive effectively restrained high-temperature sintering and irreversible HF halogen poisoning, boosting overall catalytic efficiency by approximately 30%. DFT adsorption simulations revealed totally different active centers: hollow sites dominate ZnO-C₄N with broad adsorption capacity for SO₂, SO₂F₂, and other fluorinated intermediates, whereas surface O sites of CuO-C₄N exhibit exclusive strong chemisorption toward SO₂. Combined with macroscopic kinetics and atomic-scale interfacial interaction rules, their distinct stepwise defluorination pathways were illustrated. This study offers solid experimental data and microscopic theoretical guidance for designing advanced C₄N-based catalysts for waste SF₆ abatement.

1. Introduction

Sulfur hexafluoride (SF₆) possesses outstanding insulation, arc-extinguishing, and heat conduction properties, which makes it widely deployed in high-voltage circuit breakers, transformers, and gas-insulated switchgear (GIS) [1,2,3]. Regulated by the Kyoto Protocol, SF₆ features an ultra-high global warming potential (GWP) of approximately 23,900 times that of CO₂ and an atmospheric lifetime exceeding 3000 years [4,5]. Massive waste SF₆ is generated as aging power equipment is retired globally, and its direct emission causes severe greenhouse pollution, making low-cost, high-efficiency disposal technology an urgent research target [6].

Multiple technical routes have been proposed for SF₆ degradation, including high-temperature pyrolysis, plasma treatment, photocatalysis, and thermal catalytic decomposition [7,8,9]. Simple thermal cracking requires temperatures above 1000 °C and produces disordered toxic fluorinated byproducts; plasma and photocatalytic approaches suffer high operating costs and limited processing throughput, restricting industrial application [10]. Thermal catalysis can lower the reaction temperature and regulate the product distribution, emerging as the most promising technical scheme for practical engineering [11]. A variety of catalytic materials including metal oxides, as well as phosphate and carbon-based composites have been explored for SF₆ abatement [12,13].

Two-dimensional C_x layered materials represented by C₃N and C₄N have attracted extensive attention in adsorption and catalysis due to their large specific surface area and abundant nitrogen active sites [14,15,16]. Compared with C₃N, C₄N exhibits superior thermal and corrosion resistance under medium-high temperature SF₆ reaction environments, yet pure C₄N can barely activate the stable SF₆ molecules [17,18]. Transition metal oxides ZnO and CuO possess rich surface-active sites and are widely used to modify carbon-based carriers to construct heterojunction catalysts [19,20,21,22]. Loading ZnO and CuO onto C₄N can build heterogeneous interfaces to modulate surface electron distribution and enhance SF₆ adsorption activation capacity [23,24].

Most existing works have only reported macroscopic catalytic behavior of single metal oxides or carbon carriers [25,26,27], lacking atomic-scale DFT investigations of SF₆ adsorption on CuO/ZnO-modified C₄N [28]. DFT serves as a standard first-principles method for interfacial simulation [29]. Its theoretical basis was laid by Hohenberg and Kohn, then extended via the Kohn–Sham equations [30,31]. The PBE functional [32] and DFT-D3 dispersion correction [33] are commonly used for catalytic surface calculations. The Delley-developed DMol³ module [34,35] is widely employed to simulate fluorine-containing gas adsorption on solid surfaces [36].

In this work, C₄N monolayer supported CuO/ZnO composite catalysts were synthesized, and multi-condition thermal catalytic experiments combined with XRD/TEM/EDS characterizations were carried out. DFT calculations via DMol³ based on classic Hohenberg–Kohn and Kohn–Sham theory were performed to systematically study the adsorption behaviors of SF₆, CO₂, SO₂, SOF₂, SO₂F₂ on different active sites. By correlating macroscopic conversion data with atomic interfacial rules, the intrinsic difference in SF₆ degradation pathways of two composite catalysts was clarified. This work fills the research gap on C₄N-based bimetallic catalysts for SF₆ treatment and provides theoretical support for subsequent catalyst optimization.

2. Results

2.1. Analysis of Catalytic Performance

2.1.1. Morphological Characterizations of Catalysts

XRD measurement (XRD-6100, Shimadzu Corporation, Kyoto, Japan) was conducted to analyze the crystal phase of the 3:1 CuO/ZnO-C₄N catalyst, as shown in Figure 1. The diffraction peaks match the standard PDF cards of CuO and ZnO, confirming the successful generation of metal oxide crystalline phases.

To quantify the crystallite dimension of metal oxide nanoparticles, we calculated grain sizes using the Scherrer equation:

$$D={\displaystyle \frac{K\lambda }{\beta cos\theta }}$$

(1)

where D stands for crystallite size with the unit of nm; K is the shape factor; λ is the wavelength of the X-ray source, and Cu Kα radiation with a wavelength of 1.5406 Å is utilized in the present test; β refers to the full width at half maximum (FWHM) of each diffraction peak; θ is half of the diffraction angle of the corresponding peak.

Based on the spectral data of the CuO (110) and ZnO (111) characteristic diffraction peaks, the average crystallite sizes of CuO and ZnO were solved to be 18–24 nm and 16–22 nm, respectively.

Figure 2 displays transmission electron microscopy (TEM) (JEM-F200, JEOL Ltd., Tokyo, Japan) images of CuO/ZnO-C₄N. It can be observed that the ZIF-8 organic framework was transformed into wrinkled carbon skeleton films after calcination, coating nanoparticles, and providing a highly tunable surface and pore structure for CuO and ZnO particles.

Figure 3 presents the energy-dispersive X-ray spectroscopy (EDS) (Sigma 300, Carl Zeiss AG, Oberkochen, Germany) mappings obtained from two randomly selected regions on the catalyst surface. Point scanning images of the two regions reveal that Cu, Zn, and O elements exhibit uniform and highly overlapping distribution on the catalyst surface. For Zn element, dense and homogeneous distribution without obvious aggregation or segregation is observed in the scanning regions of different scales. The distribution of Cu elements highly coincides with Zn, showing synchronous spatial uniformity, which proves uniform mixing of Cu and Zn in the catalyst. Oxygen elements cover the entire scanning area with a distribution range consistent with Cu and Zn, demonstrating homogeneous dispersion of oxygen and formation of uniform composite oxides with Cu and Zn.

2.1.2. Catalytic Performance Analysis

First, reaction gas screening experiments were carried out. Three typical industrial atmospheres including inert atmosphere (He), oxidizing atmosphere (air), and reducing atmosphere (NH₃) were selected, corresponding to three gas systems: 2% SF₆/98% He, 2% SF₆/95% air, and 2% SF₆/2% NH₃/96% Ar. All blank experiments were performed at 600 °C, 0.2 MPa, and a flow rate of 5 mL·min⁻¹, with the blank test results summarized in

Table 1. Blank SF₆ conversion rates under different reaction gas systems.

Reaction Gas System	Temperature	Pressure	Flow Rate	Maximum Conversion Rate
2% SF6/98% He	600 °C	0.2 MPa	5 mL·min−1	11%
2% SF6/98% Air	600 °C	0.2 MPa	5 mL·min−1	17%
2% SF6/2% NH3/96% Ar	600 °C	0.2 MPa	5 mL·min−1	31%

. Although previous research [4] reported trace SF₆ decomposition at 300 °C, this study found that the blank conversion rate of SF₆ under an inert atmosphere remained below 11% at 600 °C. This originates from the ultrahigh stability and recovery characteristics of SF₆ molecules, fully verifying the limitation of pure thermal treatment in breaking S–F bonds and the low efficiency of simple pyrolysis. The oxidation effect of O₂ in an air atmosphere raises the conversion rate to 17% with limited improvement, while the reducing atmosphere containing NH₃ delivers optimal performance with a conversion rate of 31%, nearly doubling that under a He atmosphere, which preliminarily confirms the mild promoting effect of NH₃ on SF₆ pyrolysis.

To analyze product differences under an oxidative atmosphere, concentration curves of each product at 700 °C and 0.2 MPa in air were plotted as in Figure 4. CO₂ was the dominant product generated by high-temperature oxidation of the C₄N carbon skeleton, which damaged the catalyst long-term stability. SOF₂ and SOF₄ were the main sulfur-containing intermediates, while almost no SO₂ could be detected. In contrast, the NH₃ reducing system delivered much higher SF₆ conversion and SO₂ selectivity, so an NH₃ atmosphere was adopted for subsequent catalytic tests.

It can be seen that the concentrations of the intermediates SOF₂ and SOF₄ fluctuate significantly in the initial stage. The SF₆ decomposition system involves various fluorinated species with complex mutual transformation, and such concentration fluctuation of fluorine-containing by-products is intrinsic to this reaction system. The observed concentration peaks stem from incomplete defluorination in the early reaction period. As the reaction proceeds, these intermediates further cleave S–F bonds and transform into stable SO₂, which gradually lowers the intermediate concentrations.

When catalysts were added under an air atmosphere at 700 °C, conversion varied greatly with pressure (

Table 2. SF₆ conversion rates under 2% SF₆/98% air atmosphere with varied operating conditions.

Temperature	Pressure	Flow Rate	Maximum Conversion Rate
700 °C	0.4 MPa	5 mL·min−1	38%
700 °C	0.6 MPa	5 mL·min−1	12%
700 °C	0.2 MPa	5 mL·min−1	26%

). The optimal pressure was 0.4 MPa with a conversion of 38%, while conversion dropped sharply to 12% at 0.6 MPa and 26% at 0.2 MPa. Such a trend can be attributed to the adsorption equilibrium of reactants and intermediates on catalyst surfaces. Elevated pressure leads to competitive adsorption between SF₆ and fluorinated intermediates, inhibiting the activation of SF₆ molecules. Under an NH₃ reducing atmosphere, catalysts with different Cu/Zn ratios exhibited distinct activity. The initial conversion of 3:1 CuO/ZnO-C₄N reached 39%, higher than 29% of the 1:1 counterpart. Time-resolved conversion curves of catalysts with/without CaO additive at 500 °C, 0.2 MPa are presented below. The 1:1 catalyst peaked at 85% at 7 h, whereas the 3:1 catalyst achieved a maximum conversion of 92%. The richer surface CuO active sites of the 3:1 sample enhanced SF₆ adsorption and activation capacity. Without CaO, the initial conversion of 1:1 catalyst was only 14%, rising to 56% at 1 h and stabilizing at 50–55%, around 30% lower than the CaO-containing group.

Product selectivity profiles under different catalyst formulations and temperatures are provided below. At 500 °C without CaO, SOF₂ and SOF₄ dominated the early-stage intermediates, and SO₂ selectivity only increased after 6 h. After adding CaO, SO₂ generation accelerated and its selectivity rose from 4 h with a peak value of 55%. For the 3:1 catalyst at 500 °C, the ascending trend of SO₂ selectivity was delayed until 10 h. At 600 °C with CaO, the product distribution became dispersed, and SO₂ selectivity peaked at 54% within 2 h. Compared with existing catalysts requiring 650–750 °C, the present system realized mild-temperature degradation, yet further optimization on sulfur product selectivity is required. From the valence variation of sulfur species before and after the reaction, we can infer the core function of NH₃. SF₆ features S in the highest +6 oxidation state, which is chemically stable and hard to reduce. NH₃ acts as the critical reducing agent that provides hydrogen radicals, lowering the valence of S atoms stepwise and facilitating S–F bond cleavage. Meanwhile, the redox reaction between NH₃ and CuO generates in situ H₂O, which serves as an oxygen source to produce target products like SO₂ and accelerates the continuous defluorination of low-fluoride intermediates. Without NH₃, SF₆ can barely be reduced to low-valence sulfur species, resulting in extremely low conversion efficiency.

To further assess the performance of CuO/ZnO-C₄N, we compared its catalytic activity and product selectivity with reported SF₆ decomposition catalysts. Typical γ-Al₂O₃, AlPO₄/γ-Al₂O₃ and FeOOH-SiC require temperatures above 650 °C for high SF₆ conversion [13,37,38]. Owing to the heterojunction between CuO, ZnO, and wrinkled C₄N monolayer, our catalyst activates SF₆ starting at 500 °C under milder conditions. At 500 °C in an NH₃ atmosphere, the 3:1 Cu/Zn catalyst delivers a peak SF₆ conversion of 91%, outperforming most phosphate and single metal oxide counterparts at the same medium temperature. In addition, most prior works lack direct GC–MS quantification of sulfur-containing products and rely on error-prone alkaline absorption. Here, online GC–MS confirmed the catalyst reaches a maximum SO₂ selectivity of 97% while suppressing toxic SO₂F₂. Though phosphate catalysts favor SO₃ production, their SF₆ conversion at moderate temperatures is low. Overall, CuO/ZnO-C₄N exhibits integrated strengths in mild-temperature SF₆ degradation and selective SO₂ generation.

2.2. Adsorption Simulation Analysis of SF₆ on ZnO-C₄N Surface

Comprehensive macroscopic thermal catalytic experimental results indicate that an NH₃ reducing atmosphere and a high Cu doping ratio can effectively improve SF₆ degradation capacity. Macroscopic catalytic performance is essentially determined by molecular adsorption behavior and interfacial interaction strength on the catalyst surfaces. Therefore, DFT first-principles simulations are adopted below to analyze the interaction laws between SF₆ and its decomposition species on catalytic substrates from microscopic parameters including adsorption configuration, adsorption energy, and charge transfer, revealing the underlying differentiated catalytic activity mechanisms of ZnO-C₄N and CuO-C₄N.

2.2.1. Adsorption Behavior of SF₆ on Various Sites

SF₆ molecules were placed at hollow, N, and O sites of ZnO-C₄N for geometric optimization, with the optimized adsorption configurations illustrated in Figure 5. Geometric configuration analysis reveals obvious repulsion between surface O atoms of the material and F atoms in SF₆ at O sites, accompanied by overall offset of SF₆ molecules and slight deformation of C₄N nanosheets. Molecular offset and interfacial repulsion are weakened at N sites, while SF₆ molecules bind most tightly with the substrate at hollow sites, presenting the most remarkable molecular configuration variation and strongest interfacial interaction. This phenomenon demonstrates inherent electrostatic repulsion between O and F atoms, which hinders stable chemical bond formation. SF₆ decomposition relies on S–F bond cleavage and F atom desorption, and hollow sites can effectively destabilize F atoms.

Spontaneous adsorption of SF₆ occurs on all sites of ZnO-C₄N. Hollow sites exhibit the largest absolute adsorption energy (−0.253 eV), proving the strongest adsorption stability and serving as the optimal active sites for SF₆ activation on ZnO-C₄N. O sites deliver the smallest absolute adsorption energy with the weakest adsorption effect, consistent with the morphological analysis results.

2.2.2. Adsorption Characteristics of SF₆ Decomposition Intermediates

Adsorption calculations were conducted for four typical SF₆ decomposition products including CO₂, SO₂, SO₂F₂, and SOF₂. The optimized adsorption configurations are shown in Figure 6, and the adsorption energies alongside Mulliken charge transfer quantities are listed in

Table 3. Adsorption energies and charge transfer quantities of various gas molecules on ZnO-C₄N.

Gas Molecule	Eads/eV	Qt/e
CO2	−0.254	−0.032
SO2	−2.790	0.087
SO2F2	−1.425	−0.477
SOF2	−0.068	−0.010
SF6	−0.045	−0.018

. From the perspective of molecular geometry, S atoms in SO₂ and SO₂F₂ molecules display strong oxygen affinity, which enables strong interfacial bonding with surface O atoms of ZnO, stabilizing the decomposition products and driving the forward progress of the SF₆ decomposition reactions. No obvious bond cleavage was observed after adsorption of all the gas molecules, yet strong adsorption effectively weakens the internal chemical bonds of molecules, providing energy foundations for subsequent thermal catalytic bond breaking.

Combining adsorption energy and charge transfer data shows that SO₂ delivers an adsorption energy of −2.790 eV with a far larger absolute value than other gases, accompanied by prominent positive charge transfer, which verifies strong chemical adsorption between SO₂ and ZnO-C₄N with intense interfacial electron exchange. SO₂F₂ exhibits an adsorption energy of −1.425 eV, also corresponding to strong adsorption with massive negative charge transfer. In contrast, CO₂, SOF₂, and SF₆ possess small absolute adsorption energies dominated by weak physical adsorption, with negligible charge transfer and faint interfacial interactions.

2.3. Adsorption Simulation Analysis of SF₆ on CuO-C₄N Surface

2.3.1. Adsorption Behavior of SF₆ on Various Sites

Figure 7 displays the optimized adsorption configurations of SF₆ at three types of sites on CuO-C₄N. Similar to ZnO-C₄N, electrostatic repulsion exists between surface O atoms of CuO and F atoms in SF₆, yet the electronic properties of Cu elements alter the overall interfacial interactions: SF₆ molecules show no obvious offset at O sites with weak adsorption attraction; interfacial interactions at N sites are moderate; hollow sites deliver the weakest adsorption effect. This originates from the affinity between Cu atoms and SF₆ molecules, which offsets partial O–F atomic repulsion and changes the optimal active sites of CuO-C₄N.

The O sites of CuO-C₄N possess the largest absolute adsorption energy (−0.862 eV), followed by the N sites, while the hollow sites present the lowest adsorption energy. This result is completely opposite to ZnO-C₄N, revealing that modification with different metal oxides induces significant variations in surface active site distribution and electronic structures of composite materials, directly altering the adsorption preference toward SF₆. The O and N sites constitute core regions for SF₆ activation on CuO-C₄N.

2.3.2. Adsorption Characteristics of SF₆ Decomposition Intermediates

Adsorption calculations were performed for CO₂, SO₂, SO₂F₂, and SOF₂. The optimized adsorption configurations are presented in Figure 8, and the adsorption energies with Mulliken charge transfer quantities are summarized in

Table 4. Adsorption energies and charge transfer quantities of various gas molecules on CuO-C₄N.

Gas Molecule	Eads/eV	Qt/e
CO2	−0.622	−0.019
SO2	−3.316	0.075
SO2F2	−0.479	−0.007
SOF2	−0.460	−0.007
SF6	−0.020	−0.024

. Configuration observation demonstrates that all decomposition species except SO₂ exhibit weak repulsion trends with CuO-C₄N surfaces, characterized by enlarged atomic distances and molecular geometries departing from the substrates. This indicates that CuO-C₄N delivers an extremely high selective adsorption capacity toward SO₂ to directionally enrich this product, yet it lacks sufficient adsorption capacity for other fluorinated decomposition intermediates.

From adsorption energy and charge transfer perspectives, SO₂ presents an adsorption energy of −3.316 eV on CuO-C₄N, representing the strongest adsorption interaction among all molecules with remarkable charge transfer corresponding to intense chemical adsorption. SO₂F₂ and SOF₂ deliver moderate adsorption energies with negligible charge transfer and weak interfacial interactions. CO₂ exhibits moderate adsorption affinity, while SF₆ itself shows the weakest adsorption capacity.

2.4. Comparative Analysis of Adsorption Performance Between ZnO-C₄N and CuO-C₄N

DFT results reveal completely different active centers for the two catalysts. Hollow sites dominate ZnO-C₄N and trap SF₆ more stably, while surface O sites act as core active regions on CuO-C₄N. ZnO-C₄N shows universal adsorption toward SO₂, SO₂F₂, and other fluorinated intermediates, supporting continuous multi-step defluorination. In contrast, CuO-C₄N only strongly adsorbs SO₂ with a weak affinity for other byproducts, enabling selective product enrichment. Both materials undergo intense electron exchange with SO₂ via orbital hybridization to weaken the S–F bonds. ZnO-C₄N suits deep SF₆ degradation, whereas CuO-C₄N fits targeted tail gas control.

2.5. Analysis of Catalytic Decomposition Mechanism

XRD, TEM, and EDS verify uniform CuO/ZnO loading on intact C₄N frameworks with abundant heterogeneous interfaces lowering the SF₆ dissociation barriers. Single thermal treatment barely decomposes SF₆, proving surface adsorption pre-activation is indispensable. SF₆ is captured at optimal sites, and O–F electrostatic repulsion plus interfacial charge transfer break the S–F bonds to generate low-fluorine intermediates. Under NH₃ reduction, the intermediates gradually defluorinate into SO₂ and SOF₂. CaO removes HF and inhibits catalyst sintering/halogen poisoning to sustain long-term activity. The divergent adsorption behaviors originate from intrinsic electronic differences between Zn and Cu, consistent with both experimental conversion data and DFT calculations.

3. Materials and Methods

3.1. Experimental Reagents

All chemical reagents used in experiments were analytical grade without further purification before use. Zinc nitrate hexahydrate, copper nitrate trihydrate, 2-methylimidazole, methanol, ethanol, and sodium borohydride were purchased from Sinopharm Chemical Reagent Co., Ltd (Beijing, China). High-purity SF₆, NH₃, and He (purity ≥ 99.999%) were obtained from Wuhan Special Gas Co., Ltd. (Wuhan, China). Quartz cotton and calcium oxide (CaO) were supplied by Aladdin Reagent Co., Ltd. (Shanghai, China).

3.2. Catalyst Preparation

MOF-derived materials have been widely adopted to fabricate supported metal oxide catalysts, in which ZIF-8 serves as a self-sacrificial template to construct carbon–nitrogen skeletons under inert atmosphere [39,40,41]. Benefiting from the inherent porous structure of ZIF-8, metal species can be homogeneously confined within the pore channels, which effectively suppresses particle aggregation during thermal treatment. Herein, CuO/ZnO-C₄N composite catalysts were synthesized via a ZIF-8 template route combined with liquid-phase reduction and subsequent calcination. The detailed preparation procedures are described as follows:

• Synthesis of ZIF-8 precursor. ZIF-8 was prepared via a typical solution assembly method. In brief, 3.36 g Zn(NO₃)₂·6H₂O (0.07 mol L⁻¹) and 7.40 g 2-methylimidazole (0.56 mol L⁻¹) were separately dissolved in 160 mL methanol. The zinc solution was slowly blended into the ligand solution under continuous stirring. After stirring for 30 min at room temperature, the mixture was statically aged for 24 h. The white precipitate was centrifuged, rinsed repeatedly with methanol, and dried at 70 °C for 12 h to obtain pure ZIF-8 powder.
• Fabrication of Cu-modified ZIF-8 intermediate. An amount of 1.0 g of ZIF-8 powder (5.56 mmol Zn²⁺) was uniformly dispersed in 80 mL methanol by ultrasonication for 30 min. Subsequently, 1.34 g Cu(NO₃)₂·3H₂O (5.56 mmol) was added and stirred for 1 h to achieve sufficient Cu²⁺ adsorption. A total of 1.06 g NaBH₄ (13.9 mmol) dissolved in 60 mL deionized water was added dropwise within 4 min for liquid-phase reduction. The reaction continued for another 1 h. The solid product was centrifuged, washed with deionized water and ethanol, and vacuum-dried at 50 °C for 12 h. For the 3:1 Cu/Zn sample, the dosages of copper salt and NaBH₄ were tripled while the ZIF-8 amount remained unchanged.
• Synthesis of CuO/ZnO-C₄N catalyst. The dried Cu@ZIF-8 precursor was calcined in a flowing nitrogen atmosphere. The temperature was increased to 350 °C at a heating rate of 1 °C min⁻¹ and maintained for 6 h. During pyrolysis, the ZIF-8 organic framework was thermally decomposed and reconstructed into a wrinkled C₄N carbon–nitrogen monolayer skeleton, while the adsorbed metal species were simultaneously converted into CuO and ZnO nanoparticles. After natural cooling and grinding, the final CuO/ZnO-C₄N composite catalyst was obtained.

3.3. Thermal Catalytic Experimental Procedure

A vertical fixed-bed thermal catalytic reaction platform was built to verify the macroscopic catalytic performance of the composite materials, as shown in Figure 9. Experimental steps: accurately weigh 1.0 g catalyst and 0.5 g CaO additive, mix uniformly, and fill into the quartz cotton support layer in the middle section of the reactor; repeatedly vacuumize and fill with nitrogen 3 times to replace the residual air in the pipelines; introduce the reaction gas, heat to the set temperature, and stabilize under constant temperature; collect the tail gas with Teflon sampling bags every 1 h, and conduct qualitative and quantitative detection of SF₆ and decomposition products through GC–MS (QP2010 Ultra, Shimadzu Corporation, Kyoto, Japan). After the experiments, nitrogen was introduced to purge the pipelines, and the tail gas was discharged after an alkali washing treatment. The mass ratio of catalyst to CaO additive (2:1) was determined based on our preliminary tests [26,27], which verified that 0.5 g CaO matched with 1.0 g catalyst achieves optimal anti-poisoning and anti-sintering effects.

3.4. First-Principles Computational Methods and Model Construction

3.4.1. Computational Parameter Settings

All simulation calculations in this work were performed based on density functional theory (DFT) using the Dmol³ module [30,31,35]. The exchange-correlation functional adopted the Perdew–Burke–Ernzerhof (PBE) functional within the generalized gradient approximation (GGA) [32]. The interaction between valence electrons and ionic cores was described by Kleinman–Bylander norm-conserving ultrasoft pseudopotentials [42]. A basis set cutoff of 400 eV was selected after systematic energy convergence tests to balance the computational accuracy and cost. The Brillouin zone was sampled using a 3 × 3 × 1 Monkhorst–Pack k-point mesh [43]. The convergence criteria for geometric structure optimization are as follows: atomic force less than 0.03 eV·Å⁻¹, total system energy convergence precision below 1.0 × 10⁻⁵ eV·atom⁻¹. A vacuum layer with a thickness of 15 Å was constructed to eliminate interlayer interactions. Van der Waals dispersion correction (DFT-D3) [33] was incorporated during calculations to accurately characterize the weak interactions between the gas molecules and the material surfaces.

3.4.2. Model and Adsorption Site Construction

A two-dimensional C₄N monolayer supercell model was built based on the C₄N crystal structure obtained from the experimental characterizations. ZnO and CuO clusters were loaded onto the substrate surface separately, and geometric optimization was conducted to acquire the ZnO-C₄N and CuO-C₄N composite models. Combined with the surface structures of the materials, three typical adsorption sites (hollow sites, N sites, and O sites) were selected for adsorption calculations of SF₆ and its decomposition species. Molecules including SF₆, SO₂, SOF₂, SO₂F₂, and CO₂ were placed above each site at initial positions for structural optimization and energy calculations. The initial structural models of SF₆ placed at three representative active sites on the C₄N monolayer before geometric relaxation are illustrated in Figure 10, containing both side and top views of the supercell slab.

3.4.3. Calculation of Adsorption Energy and Charge Transfer

Adsorption energy, E_ads, serves as the core index to evaluate the adsorption strength, defined by Equation (2):

E_ads = E_total − E_sub − E_gas

(2)

where E_total denotes the total energy of the whole system after the gas molecules are adsorbed on the material surface (eV); E_sub represents the energy of the pure catalytic substrate (eV); E_gas is the energy of the isolated gas molecules (eV). A negative E_ads indicates an exothermic adsorption process, and a larger absolute value corresponds to stronger adsorption and a more stable system structure.

Mulliken population analysis was employed to calculate the total charge transfer Q_t of the system, characterizing the electron transfer laws at the gas–solid interface and analyzing the types of interfacial interactions.

4. Conclusions

In this work, a two-dimensional C₄N monolayer was employed as a carrier to fabricate CuO/ZnO bimetallic composite catalysts for medium-temperature SF₆ abatement. Systematic catalytic experiments, XRD, TEM, and EDS microstructural characterizations, and DFT first-principles adsorption calculations were combined to reveal the macroscopic degradation performance and atomic-scale interfacial mechanism of the as-prepared materials. The core findings are summarized as follows:

(1)

The impregnation-calcination strategy enables uniform dispersion of CuO and ZnO nanoparticles on intact C₄N skeletons without structural collapse. Atmosphere screening verifies that NH₃ reducing conditions greatly promote SF₆ conversion, and the catalyst with a Cu/Zn molar ratio of 3:1 delivers a maximum SF₆ conversion efficiency of 91%. Incorporated CaO effectively alleviates catalyst sintering and irreversible HF-induced halogen poisoning, raising the overall catalytic activity by approximately 30%.

(2)

DFT calculations identified totally distinct intrinsic active sites for ZnO-C₄N and CuO-C₄N. Hollow sites dominate SF₆ activation on ZnO-C₄N, while surface O-top sites serve as the primary reactive centers on CuO-C₄N. Electrostatic repulsion between surface oxygen and fluorine atoms acts as the key driving force for stretching and breaking S–F bonds; strong chemical adsorption accompanied by intensive interfacial electron transfer occurs between both catalysts and SO₂ intermediates.

(3)

The two composites exhibit divergent catalytic selectivity. ZnO-C₄N shows universal adsorption affinity toward all fluorinated SF₆ decomposition species, making it suitable for thorough harmless disposal of waste SF₆. In contrast, CuO-C₄N possesses exclusive selective adsorption for SO₂, which can be exploited for targeted recovery of sulfur-containing products from exhaust streams. The DFT-derived interfacial interaction rules are highly consistent with the measured conversion and product distribution data obtained from thermal experiments.

(4)

Heterojunctions formed by loaded CuO/ZnO modulate the surface electronic distribution of C₄N substrates. The inherent electronic discrepancy between Zn and Cu species accounts for the different catalytic activity and selectivity of the two materials. This study establishes a fundamental atomic-level understanding of bimetallic C₄N-based catalysts and provides experimental and theoretical guidance for designing high-efficiency catalysts to mitigate SF₆ greenhouse gas emissions.

Nevertheless, the present investigation is mainly based on static DFT adsorption simulations and short-duration batch catalytic experiments. High-temperature dynamic evolution and long-term continuous performance remain to be explored. Complementary techniques including in situ characterizations and AIMD simulations could be implemented in future work to further corroborate the reaction pathway proposed herein.