Study on the Sustainable Degradation of Sulfur Hexafluoride by Thermal Plasma for Greenhouse Gas Abatement

Abstract

This study addresses the challenges of efficiency and cost in traditional sulfur hexafluoride (SF₆) degradation methods and the throughput limitations of common plasma technologies, with the aim of promoting sustainable treatment of potent greenhouse gases. A method of premixing SF₆ with plasma media before entering the plasma discharge region was employed to systematically investigate the effects of three atmospheres—nitrogen, air, and hydrogen—on the degradation efficiency, product distribution, and energy efficiency of SF₆. An experimental setup was constructed, and Gibbs free energy minimization simulations were conducted to analyze the degradation performance under different conditions. The results show that the premixed gas injection method achieves a degradation removal efficiency of over 99.84% when the SF₆ flow rate is lower than 4 slm, which is significantly better than the staged mixing method. When the discharge current increases from 40 A to 100 A, the degradation effect of SF₆ improves significantly, but the improvement becomes marginal when the current is further increased to 120 A. Compared with nitrogen, air and hydrogen atmospheres can effectively enhance the degradation removal rate, with the air atmosphere achieving the highest energy yield of 271 g/kWh. This research reveals the regulatory mechanism of medium components on SF₆ degradation, providing a theoretical basis for the sustainable, full-process treatment of industrial-scale reactors and contributing to the mitigation of greenhouse gas emissions.

1. Introduction

Sulfur hexafluoride (SF₆), with its excellent insulating properties and chemical stability, plays an indispensable role in fields such as gas-insulated switchgear in the power industry and plasma etching in semiconductor manufacturing [1]. The global warming potential (GWP) of SF₆ is as high as 25,900. This means that emitting 1 kg of SF₆ generates the same greenhouse effect as emitting 25,900 kg of carbon dioxide. It is one of the greenhouse gases with the highest GWP value and has thus been listed as a key-controlled greenhouse gas in the Kyoto Protocol of the United Nations Framework Convention on Climate Change and the Paris Agreement. From 1994 to 2020, the concentration of SF₆ in the global atmosphere surged from 3.67 ppt to 10.5 ppt [2]. With the increasing global attention to climate change issues, effectively controlling SF₆ emissions has become an urgent task.

However, traditional degradation methods are mired in the dilemma of efficiency and cost. Thermal catalytic degradation requires the use of catalysts such as CePO₄ and activates the reaction at temperatures above 800 °C. Although the degradation rate can be increased to 85% [3], The catalyst activity declines due to the phase transition of the support of the high-temperature catalyst, gamma-Al₂O₃ to alpha-Al₂O₃. Moreover, for every 1 kg of SF₆ treated, 0.5 kg of lime is consumed to adsorb by-products. Photocatalytic degradation can achieve a degradation rate of 60% by using ultraviolet excitation on the surface of polyisoprene [4]. However, the 4 h treatment duration and a single-batch processing capacity of 50 L fall considerably short of satisfying the annual demand within the power industry, which requires the treatment of tens of thousands of metric tons of waste gases.

Common plasma technologies used for the degradation of sulfur hexafluoride include radio-frequency plasma [5,6], microwave plasma [7,8], dielectric barrier discharge [9,10], etc. These plasma sources exhibit distinct generation methods and reaction activities, yet all of them fall within the category of non-thermal plasmas where the electron temperature significantly exceeds that of heavy particles. Non-thermal plasmas exhibit significant potential in laboratory settings. For instance, microwave plasma can achieve a degradation rate of 99.9% for 2.4% SF₆ at a power of 6 kW, and radio-frequency plasma can reach a degradation efficiency of 99.87% at a power of 50 W. However, due to their non-equilibrium characteristics and discharge power scales that are typically only in the tens to hundreds of watts, the treatment volume rarely exceeds 1 L/min. This makes it challenging to overcome the throughput limit, significantly impeding the transition of this technology to industrial applications.

Thermal plasma possesses unique advantages, including a local temperature exceeding 3000 K and a state of thermodynamic equilibrium. At these high temperatures, SF₆ molecules can be entirely dissociated into their constituent S and F atoms. By introducing reactive gases like nitrogen (N₂), hydrogen (H₂), water vapor (H₂O), or oxygen (O₂), the recombination of sulfur-fluorine (S-F) atoms can be effectively suppressed. The resulting products, including sulfuryl fluoride (SOF₂), thionyl tetrafluoride (SOF₄), sulfuryl difluoride (SO₂F₂), and hydrogen fluoride (HF), are acidic gases that are readily absorbed by alkaline solutions. Compared to non-thermal plasma degradation methods, the treatment volume of thermal plasma degradation methods has increased by more than an order of magnitude [11]. Ding and Sun investigated the impact of parameters such as reaction temperature, the mixing ratio of hydrogen or oxygen, and discharge power on the degradation of SF₆ through chemical kinetic simulations and experimental tests [11,12]. Under optimal conditions, the degradation removal efficiency (DRE) of SF₆ can reach 99%, and the energy yield can reach 206 g/kWh. However, in their study, the plasma working gas contained a significant amount of argon to stabilize plasma combustion, which led to increased costs for the carrier gas. Moreover, the engineering challenge of sulfur adhering to the reaction chamber, which is difficult to remove, is not favorable for direct engineering applications. The plasma wet scrubber, utilized for the treatment of perfluorinated compound (PFC) waste gas in the semiconductor industry [13], employs nitrogen plasma to mix with highly diluted (concentration ≤ 1%) perfluorinated compounds (PFCs) for degradation. It does not directly utilize plasma and still cannot achieve high-throughput degradation of SF₆.

In response to the aforementioned research gaps, the novelty of this study lies in employing a method that premixes SF₆ with plasma media prior to its entry into the plasma discharge region, which is distinctly different from the downstream staged-mixing methods commonly reported in the literature. This method capitalizes on the high temperature, high enthalpy, and high chemical reactivity of thermal plasma to enhance the treatment concentration, treatment volume, and degradation rate of SF₆. The paper systematically compares the influence of three atmospheres-nitrogen (neutral environment), air (oxidizing environment), and hydrogen (reducing environment)-on the degradation efficiency (DRE), product distribution, and energy efficiency (EY) of SF_6. By combining Gibbs free energy minimization simulations, the regulatory mechanism of medium components on SF₆ degradation is deeply elucidated, offering a robust theoretical foundation for the full-process treatment of industrial-scale reactors.

2. Experimental Setup and Evaluation Method

2.1. Experimental Setup

This experiment aims to investigate the degradation effect of thermal plasma on SF₆ when using nitrogen, air, and hydrogen as the plasma working gases. The experimental setup as shown Figure 1, mainly consists of a power supply, gas supply, a plasma generator, a reaction chamber, a gas absorption system (including spaying, washing and absorption). The power supply first applies an over-voltage between the cathode and the arc-starting anode to start the arc, and then maintains a stable current between the cathode and the main-arc anode to ensure the continuous generation of thermal plasma. The gas supply accurately regulates the flow rates of SF₆ and plasma media (air, nitrogen, and hydrogen) through mass flow meters. There are two intake methods: one involves pre-mixing SF₆ with the plasma media before entering the plasma generator, while the second method entails introducing SF₆ in stages through a second intake port into the plasma generator. The plasma generator is the core component, adopting a DC axial-type design with a double-anode structure. There is an additional inlet between the anodes. The tangential gas-inlet method allows gas to enter the torch body with both axial and tangential velocities. This prolongs the arc length, reduces local electrode ablation, and makes the arc temperature distribution more uniform, creating conditions for the efficient degradation of SF₆. After the mixed gas enters the reaction chamber, under the action of high-temperature thermal plasma, SF₆ molecules begin to decompose. The decomposition products enter the absorption and treatment system with the gas flow. The alkaline solution tower sprays sodium hydroxide (10% NaOH) solution to absorb most of the acidic gases generated during the degradation process. Then, activated alumina and 3X molecular sieves are used to absorb gases that do not hydrolyze and do not react with the alkaline solution, such as SOF₂ and SO₂F₂.

2.2. Evaluation Indexes

To accurately evaluate the degradation effect of SF₆ under different gas conditions, two key evaluation indexes, degradation removal efficiency (DRE) and energy yield (EY), are introduced. The degradation removal efficiency (DRE) is used to measure the degradation degree of SF₆, and its calculation formula is:

$$\mathrm{DRE}(\%)={\displaystyle \frac{C_{in}-C_{out}}{C_{in}}}\times 100\%$$

(1)

where $C_{in}$ is the input concentration of SF₆ before the reaction, and $C_{out}$ is the output concentration of SF₆ after the reaction. The higher the DRE value, the better the degradation effect. The energy yield (EY) reflects the energy utilization efficiency of the degradation process, and its calculation formula is:

$$\mathrm{EY}(\mathrm{g}/\mathrm{kWh})={\displaystyle \frac{{\dot{m}}_{\mathrm{sf}6}}{P_{plas}}}$$

(2)

In the equation, ${\dot{m}}_{\mathrm{sf}6}$ is the mass flow rate of SF₆ degraded in the reaction, and $P_{plas}$ is the input power of the plasma. A higher EY (Energy Yield) value indicates a more efficient use of energy for SF₆ degradation.

2.3. Analysis of Gas Degradation Products

In this work, a Panna Instrument A91-AMD10 Pro gas chromatography-mass spectrometry (GC-MS) instrument was used to analyze the degradation products under different gas conditions. When air is used as the plasma working gas, the degradation product components are complex. In addition to some unreacted SF₆, a variety of sulfur-oxygen-fluorine compounds are generated, such as sulfuryl fluoride (SO₂F₂), thionyl fluoride (SOF₂), and sulfur dioxide (SO₂). This is because SF₆ reacts with oxygen and nitrogen in the air, producing a variety of oxidation products. When nitrogen is used as the plasma working gas, the degradation products are mainly a mixture of sulfur-fluorine compounds, and some sulfur-fluorine compounds will also be formed with the OH of water vapor during the water-washing process. When hydrogen is used as the plasma working gas, the degradation products are mainly hydrofluoric acid and sulfur. After being washed and absorbed by the alkaline solution, by analyzing the retention time, peak intensity, and other information of each product peak in the gas chromatogram, the types and relative contents of different products can be determined, and thus the degradation degree of SF6 under different plasma medium conditions can be deeply understood. The results of mass spectrometry detection contain multiple substances, and potential components include SF₄, OF₂, SOF₄, SOF₂, SO₂F₂, SO₂, etc. The Figures S1–S3 show the detection results from the Panna A91-AMD10 Pro GC-MS, with SF₆ appearing at approximately 1.64. To further verify the accuracy of SF₆ content in exhaust gases, the Panna GC-4100-FPD method was also tested as shown in Figures S4–S6.

3. Results and Discussion

3.1. Chemical Equilibrium Calculation and Analysis Under Different Plasma Working Gas Atmospheres

Figure 2 shows the equilibrium amounts of various substances at different reaction temperatures in systems of nitrogen (neutral environment), air (oxidizing environment), and hydrogen (reducing environment) mixed with 5% SF₆ under a 1-atm working condition. The chemical equilibrium composition was calculated using the Gibbs free energy minimization module within the ChemKin-Pro 3.0 software package. The main calculation parameters were as follows: Pressure: 1 atmosphere; Temperature Range: 500 K to 4000 K, in increments of 100 K; Considered Species: neutral molecules, atoms, and radicals (e.g., SF₆, SF₄, SF₂, S, F, S₂, F₂, SO₂, SOF₂, SO₂F₂, HF, H, H₂, O, O₂, N₂, NF). Electrons and ionized particles were neglected. Only neutral molecules, atoms, and radicals were considered in the reaction components, and electrons and ionized-state particles were ignored.

When nitrogen is used as the working gas, it can be found that SF₆ molecules begin to decompose at around 1000 K. First, they lose F to form SF₅, SF₄, and F atoms. As the arc temperature rises, elemental S begins to form at around 2000 K. Overall, the temperatures required for the formation of SF₅, SF₄, and S gradually increase, and the maximum molar fraction of SF₄ is significantly greater than that of SF₅. The main reason is that the decomposition process of SF₆ is a process in which S-F bonds are gradually broken to form low-energy-level and low-fluorine sulfides, and the bond energy of SF₄ is higher than that of SF₅. When the arc temperature is higher than 2250 K, the molar fractions of all sulfur fluorides begin to decrease, and the decomposition products are mainly S and F atoms. However, when the temperature exceeds 2600 K, nitrogen molecules begin to be activated and dissociated, forming NF molecules. NF molecules are toxic substances that are difficult to for alkaline solutions, alumina, and molecular sieves to absorb. They should be avoided as much as possible during the degradation process. It should be noted that during the arc attenuation process, the arc temperature gradually decreases. When the arc temperature drops below 1000 K, all sulfur fluorides and atoms recombine into SF₆. The products are mainly SF₄ at temperatures below 2500 K, while at reaction temperatures higher than 2500 K under plasma working conditions, they mainly exist in the form of dissociated S and F atoms. At this time, spray-cooling and absorption with an alkaline solution can theoretically achieve a high degradation rate.

Secondly, when air is used as the working gas, SF₆ molecules also begin to decompose at around 1000 K, first forming SO₂F₂ and F atoms. As the arc temperature rises, SF₅, SF₄, and SOF₂ begin to form. The temperatures required for the formation of SF₅, SF₄, and SOF₂ gradually increase, and the maximum molar fraction of SOF₂ is significantly greater than that of SF₄ and SF₅. Different from the nitrogen plasma, there is no S atom in the oxidation condition. When the arc temperature is higher than 2000 K, the molar fraction of SF₆ in the system can be almost ignored. In the air atmosphere, the high-temperature products are mainly SOF_2. Since SOF₂ is easily absorbed by the alkaline solution, the reaction control temperature range needs to be guided by SOF₂ as the product. Theoretically, when air is used as the plasma medium, due to its oxidizing property, the degradation temperature can be reduced as a whole. However, during the discharge process of air plasma, high temperatures will generate NOx, which will cause additional pollution, and nitric oxide is not easily absorbed by the alkaline solution, causing secondary pollution.

Because the mixture of hydrogen and SF₆ as the working gas is extremely likely to cause discharge instability and has an explosion risk, only the 80%N₂ + 15%H₂ + 5%SF₆ system was subjected to chemical equilibrium calculation here. Different from using nitrogen and air as plasma media, the addition of hydrogen can greatly promote the degradation of SF6 to form HF and elemental S. As the arc temperature rises, HF will decompose into H, F atoms, and H₂ molecules. However, when the temperature exceeds 2600 K, nitrogen molecules begin to be activated and dissociated, forming NF molecules, which is not conducive to the conditions.

In summary, through the chemical equilibrium calculation and analysis of feasible reactive gases, it can be seen that the degradation product components of the H₂/SF₆ system are few, only HF and S elements. In the Air/SF₆ and N₂/SF₆ systems, the products are more complex, but they are all gases, and most of them can be absorbed by the alkaline solution. The addition of O₂ is beneficial to reducing the discharge power of the thermal plasma torch for degrading SF₆.

3.2. Influence of Mixing Methods on the Degradation Effect of Thermal Plasma

Two gas-inlet methods were used, namely directly discharging with N₂/SF₆ premixed as the plasma medium and downstream-staged mixing of SF₆ with N₂ as the plasma. The influence of the mixing method on the degradation effect of thermal plasma was compared. The discharge current was 60 A, the nitrogen flow rate was 40 slm, and the SF₆ flow rates were set as 1, 4, 8, 12, 16, and 20 slm, respectively. The results are shown in Figure 3. The premixed gas-inlet method can enter the plasma discharge region, which is different from the staged-mixing method. It can give full play to the energy and reaction activity of thermal plasma. Therefore, when the SF6 flow rate is lower than 4 slm, the degradation removal efficiency (DRE) of premixed SF₆ exceeds 99.84%. However, the staged-inlet method is difficult to maintain the degradation of relatively high-concentration SF₆. Only when the low-throughput SF₆ is 1 slm, its degradation removal efficiency (DRE) reaches 99.77%. As the SF₆ flow rate increases, the DRE values of both the premixed and staged-inlet methods will decrease significantly, but the premixed method always shows better degradation performance than the staged-mixing method. During the experiment, it was found that an increase in the SF₆ concentration can also lead to discharge instability, which has become one of the technical problems restricting the degradation of high-concentration SF₆.

3.3. Influence of Discharge Current on the Degradation Effect of Thermal Plasma

In order to compare the influence of the reaction temperature on the performance of thermal plasma in degrading SF₆, experimental analysis was carried out by changing the discharge current and adjusting the enthalpy value of the plasma. The premixed gas inlet mode was adopted, with the nitrogen flow rate being 40 slm, the SF₆ flow rate being 8 slm, and the discharge current set at 40~120 A. The results are shown in Figure 4. The discharge power increases with the rise in the discharge current, and the slope of the power increase is less than 1, which indicates that the arc exhibits a falling volt-ampere characteristic under this working condition. When the discharge current is lower than 40 A, the reaction temperature at the edge of the thermal plasma is not sufficient to fully degrade SF₆, resulting in a DRE of only 50.39%. When the discharge current increases from 40 A to 100 A, the degradation effect of SF₆ is significantly improved. By increasing the discharge power, the particle collisions will become more intense, and the temperature of heavy particles will also rise, which is conducive to promoting the decomposition of SF₆. When the discharge current is further increased to 120 A, the improvement of the degradation effect is relatively slight, and the DRE value only increases from 98.21% to 99.24%. This is similar to the trend of the simulation results: when the reaction temperature rises to a relatively high value, the degradation products no longer change, and further increasing the discharge power cannot improve the degradation effect.

3.4. Influence of Different Atmospheres on the Degradation Effect of Thermal Plasma

This study explores the influence of different atmospheres, including nitrogen (neutral environment), air (oxidizing environment), and hydrogen (reducing environment), on the performance of degrading SF₆. The discharge current is set at 60 A, and the flow rates of the plasma working gases N₂ and Air are both 40 slm, respectively. Since hydrogen is likely to cause discharge instability, when studying the hydrogen reducing environment, the plasma gas is set as a mixture of 12 slm H₂ + 28 slm N₂ to ensure the good operation of the device. The results are shown in Figure 5. Compared with the neutral environment of nitrogen, both the oxidizing environment of air and the reducing environment of hydrogen can improve the degradation removal rate of SF₆. This is consistent with the result analysis of chemical equilibrium calculations. The active gases H and O are more likely to combine with the dissociated S and F of SF₆, inhibiting the recombination of SF₆. Therefore, under the condition of a higher SF₆ flow rate of 8 slm, both the oxidizing environment of air and the reducing environment of hydrogen still have a degradation removal rate of over 96%, and the reducing environment of hydrogen is slightly higher than that of air, with the DRE reaching 98.13%. However, as the SF₆ flow rate continues to increase, the oxidizing environment of air shows a higher degradation removal rate than the reducing environment of hydrogen. The volume ratio of O₂ in air is approximately 21%, which is lower than the volume ratio of H₂ (30%) in the set hydrogen reducing environment. But the results of chemical equilibrium calculations indicate that the main product of SF₆ degradation by air plasma is SOF₂, and the main product of SF₆ degradation by hydrogen plasma is HF. The number of F atoms fixed by one O atom is twice that fixed by one H atom. Therefore, as the SF₆ flow rate increases, the molar ratio of H₂/SF₆ in the atmosphere may decrease, and the insufficient hydrogen content cannot meet the degradation of high-throughput SF_6. This phenomenon can be attributed to the different roles and fluorine fixation capacities of H and O atoms. At lower SF₆ flow rates, the H₂/SF₆ molar ratio is high. Hydrogen dissociates in the plasma to form highly reactive H radicals, which efficiently scavenge fluorine atoms to form stable HF, effectively suppressing SF₆ recombination and resulting in high DRE. In contrast, at higher SF₆ flow rates, the H₂/SF₆ ratio decreases, making the hydrogen content insufficient to capture all F atoms. Oxygen from air, however, has a higher fixation capacity for fluorine, as one O atom can fix two F atoms (e.g., in SOF₂ or SO₂F₂), whereas one H atom fixes only one F atom (in HF). Therefore, the oxidative pathway becomes more effective under high SF₆ loading, maintaining a higher degradation efficiency. Under the same current and gas flow rate, the plasma discharge power mainly depends on the medium composition. The trend of the discharge power is as follows: air < nitrogen < hydrogen. Regarding the energy yield EY, in addition to being affected by the degradation removal rate, it also depends on the discharge power. Since air not only shows excellent degradation performance for SF₆ but also has a relatively small discharge power, it exhibits the best energy yield EY among the three atmospheres, reaching up to 271 g/kWh at the highest. Although the hydrogen atmosphere also shows good SF₆ degradation performance, adding hydrogen to the plasma is likely to cause an increase in the discharge voltage and power. Therefore, its energy yield EY is not much different from that in the nitrogen atmosphere.

Conditions: Discharge current: 60 A; Plasma gas: N₂ (40 slm), Air (40 slm), H₂/N₂ (12/28 slm); Premixed mode.

4. Summary

In this study, an experimental device consisting of an intake air control system, an arc plasma torch generator, a reaction chamber, and a gas absorption system was constructed. Two key indicators, namely the degradation and removal efficiency (DRE) and the energy yield (EY), were introduced to evaluate the degradation effect. Additionally, a gas chromatograph-mass spectrometer was used to analyze the composition of the products. On this basis, the following significant results were obtained.

Chemical equilibrium calculations show that the degradation product composition of the H2 / SF₆ system is simple. The products of the Air/SF₆ and N₂/SF₆ systems are complex, but most of them can be absorbed by the alkaline solution, and the addition of O₂ helps to reduce the discharge power. The mixing method has a significant impact on the degradation effect. The premixed air intake method gives full play to the advantages of thermal plasma. When the flow rate of SF₆ is lower than 4 slm, the degradation and removal efficiency can exceed 99.8%, which is significantly better than the staged-mixing method. The discharge current has a threshold effect on the degradation effect. It increases significantly between 40 and 100 A, but the increase is limited when it exceeds 120 A. Among different atmospheres, the air atmosphere performs best in terms of energy yield, reaching up to 271 g/kWh. The superior EY of air plasma is attributed to a combination of its effective SF₆ degradation through oxidation, the higher fluorine fixation capacity of oxygen, and critically, its lower discharge power requirement compared to nitrogen and hydrogen-mixed plasmas under identical conditions. This result indicates that when using air as the plasma working gas in the industrial-grade degradation treatment of SF6, there are significant advantages in reducing energy consumption costs, providing important economic feasibility reference for large-scale industrial applications.

However, there are still certain limitations in this study. Experimental findings show that a high concentration of SF₆ can lead to unstable discharge, which restricts the degradation of high-concentration SF₆. This may be because the complex chemical reactions triggered by the high-concentration SF₆ in the plasma environment cause the disorder of the internal electric field distribution of the plasma, thus leading to unstable discharge. In addition, the mechanisms of some side reactions during the degradation of SF₆ by thermal plasma are still unclear. This may result in the inability to effectively control the generation of by-products, thereby affecting the stability and durability of the equipment and increasing the maintenance costs and treatment difficulties in industrial applications. Future work should therefore focus on: (1) optimizing reactor geometry and power supply characteristics to mitigate discharge instability at high SF₆ concentrations; (2) employing in situ diagnostic techniques to elucidate the pathways of uncertain side reactions; and (3) scaling up the reactor design for industrial pilot testing.