Effect of oxygen on power frequency breakdown voltage and decomposition characteristics of the C₅F₁₀O/N₂/O₂ gas mixture

Abstract

Sulfur hexafluoride (SF₆) is widely used in the power industry because of its excellent insulation and arc extinguishing performance; however, as the global environment is deteriorating, the need to replace SF₆ is becoming significantly critical. In recent years, C₅F₁₀O has received extensive attention as a potential alternative to SF₆. In this study, a part of N₂ in C₅F₁₀O/N₂ was replaced by O₂, and the breakdown voltages of C₅F₁₀O/N₂/O₂ at different oxygen concentrations under a slightly uneven electric field were tested. The dispersion of breakdown voltage and the discharge decomposition components of C₅F₁₀O/N₂/O₂ with different oxygen concentrations were analysed. It was found that as the oxygen concentration increased, the breakdown voltage of C₅F₁₀O/N₂/O₂ with 15 kPa C₅F₁₀O at 0.2 MPa increased, and the dispersion of the breakdown voltage became worse. When 0.5% O₂ or more O₂ was added to the C₅F₁₀O/N₂ gas mixture, the carbon precipitates on the electrode surface disappeared. As the oxygen concentration continued to increase, another characteristic component, CF₂O, could be detected, whereas C₂F₄ and C₃F₆ disappeared. It is believed that O₂ can inhibit the formation of C₂F₆, C₃F₈, C₄F₁₀, and C₃F₇H. Therefore, it is recommended to use oxygen as the second buffer gas for the engineering applications of C₅F₁₀O. Moreover, the ratio of C₅F₁₀O to O₂ is recommended to be 1 : 1.

---

1 Introduction

The gas insulated switchgear (GIS) has become an indispensable part of the power system due to its advantages such as small size and high operational reliability.^1–5 Sulfur hexafluoride (SF₆) is widely used in the GIS as a gas insulating medium with high dielectric strength and excellent arc extinguishing performance.^6–8 However, the global warming potential (GWP) value of SF₆ is as high as 23 500, and its atmospheric lifetime is about 3200 years, SF₆ is one of the most greenhouse gases.^9–11 Therefore, the demand to limit the application of SF₆ in the power industry is urgent.^12–14 In this regard, researchers have paid significant attention to find alternative gases for SF₆ from the perspective of insulation strength, environmental characteristics, decomposition properties as well as safety. Over the past three years, fluorocarbon macromolecules, such as c-C₄F₈, CF₃I, C₄F₇N, C₅F₁₀O, and C₆F₁₂O, have attracted attention.^15–25 Among these molecules, C₅F₁₀O has shown significant advantages in terms of insulation strength, low GWP and reliable biosafety. The molecular structure of C₅F₁₀O is shown in Fig. 1; C₅F₁₀O is a non-toxic, colourless, and odourless substance.

Fig. 1 Molecular structure of C₅F₁₀O.

Its chemical properties are stable; according to the data provided by 3M™, the initial decomposition temperature of C₅F₁₀O is up to 600 °C, whereas the temperature of the hottest spot in the equipment is usually less than 200 °C.²⁶ In addition, C₅F₁₀O has the GWP of only 1, and its atmospheric lifetime is only 14 days. Moreover, the ozone depletion potential (ODP) value of C₅F₁₀O is zero; this means that C₅F₁₀O does not consume the ozone layer of the atmosphere and fully meets the requirements of the “Montreal Protocol on Substances that Deplete the Ozone Layer”.²⁶ The insulation strength of pure C₅F₁₀O is twice that of SF₆; however, the liquefaction temperature of pure C₅F₁₀O under normal pressure is high (i.e. 26.9 °C); therefore, it is usually mixed with a buffer gas that has very low liquefaction temperature during its research and practical application.^27,28

Recently, several studies have been carried on C₅F₁₀O gas mixtures. Saxegaard et al. have demonstrated that the use of C₅F₁₀O/air with the liquefaction temperature of −25 °C can increase the rated voltage of an air switchgear from 12 kV to 24 kV.²⁹ In addition, it has been reported that the AirPlus™ ring network unit of ABB in the Netherlands meets the insulation expectations well and has no impact on the equipment life in the first year.³⁰ Wang et al. tested the power-frequency (AC) breakdown characteristics of C₅F₁₀O/CO₂ under a non-uniform electric field. They believe that the insulation performance of the gas mixture can be effectively improved by increasing the content of C₅F₁₀O.³¹ Aints et al. found that an increase in the content of C₅F₁₀O in C₅F₁₀O/air would reduce the effective ionization coefficient.³² Chachereau et al. have found that there is a significant synergy between C₅F₁₂O and N₂ or CO₂, and the C₅F₁₂O gas mixtures can reach the insulation strength requirement of SF₆ by increasing the gas pressure.³³

In addition, it is important to study the decomposition characteristics of the C₅F₁₀O gas mixtures. Our team calculated the decomposition mechanism of C₅F₁₀O using density functional theory (DFT). It has been found that the activity of carbonyl in the C₅F₁₀O molecule (as shown in Fig. 1) is strong. And, C1–C2 and C2–C3 in the C₅F₁₀O molecule easily break to form CF₃CO˙ and C₃F₇˙ or C₃F₇CO˙ and CF₃˙.^34,35 Wang et al. have also explored the decomposition pathway of C₅F₁₀O and pointed out that the C2–C3 and C3–C4 (or C3–C5) bonds in the C₅F₁₀O molecule are more likely to break.¹⁸ The main decomposition components for the AirPlus™ switchgear after arc extinguishing are CO₂, CO, and HF, among others, and it has been considered that the toxicity of the C₅F₁₀O gas mixture is mainly determined by CO.³⁶ Hammer et al. detected the decomposition components of C₅F₁₀O/N₂ under a dielectric barrier discharge. Several by-products such as CO, C₂F₆, and C₃F₈ were detected.³⁷

Previous studies have shown that the addition of O₂ to a C₅F₁₀O/CO₂ gas mixture can improve the insulation strength of the C₅F₁₀O gas mixture.³⁸ However, at present, only few studies have been reported on the influence of O₂ on the insulation and decomposition properties of a C₅F₁₀O gas mixture. In this study, C₅F₁₀O was mixed with N₂, and O₂ was added as the second buffer gas. By adjusting the oxygen concentration, the breakdown voltage and decomposition components of a C₅F₁₀O/N₂/O₂ ternary gas mixture after AC breakdown tests were investigated. The influence of oxygen on the insulation and decomposition properties of the C₅F₁₀O/N₂/O₂ gas mixture was explored first. Relevant results can provide reference for the engineering application of a C₅F₁₀O gas mixture.

2 Test equipment, conditions and methods

2.1 Test equipment and conditions

Fig. 2 shows the gas insulation performance and composition analysis platform. The output voltage of the high-voltage test transformer (100 kV/0.5 A) was adjusted by an induction voltage regulator. The AC voltage applied on the electrodes was measured by a voltage divider capacitor. Moreover, protective resistance (10 kΩ) was used to limit the short-circuit current during gap breakdown to prevent damage to the test transformer. Gas chromatography and mass spectrometry (GC/MS) were used for the qualitative analysis of the decomposition components.

Fig. 2 Gas insulation performance test and gas composition analysis platform.

A slightly uneven electric field is the most common electric field in the SF₆ equipment. In the experiments to explore the insulation and decomposition characteristics of a potential SF₆ substitute gas, sphere–sphere electrodes with the gap distance of 2 mm were used to simulate the slightly uneven electric field (the non-uniform coefficient of the electric field was calculated to be 1.02 by COMSOL Multiphysics with E_max = 2.04 × 10⁷ V m⁻¹ and E_av = 2 × 10⁷ V m⁻¹). The sphere–sphere electrodes are made of brass, and their radius is 25 mm.

Although the insulation strength of the C₅F₁₀O gas mixtures can be increased by increasing the content of C₅F₁₀O in the gas mixture, high liquefaction temperature of C₅F₁₀O under normal pressure will severely limit the application of C₅F₁₀O in high-voltage gas insulation equipment.²⁹ Exploration of the application of C₅F₁₀O gas mixtures in medium-voltage or low-voltage gas insulated equipment (GIE) has become the focus of research in the industry. For this reason, all the experiments in this study were carried out under the condition of the absolute pressure of 0.2 MPa.

Considering the low liquefaction temperature of nitrogen and oxygen (−140.7 °C and −118.57 °C, respectively), the liquefaction temperature of a C₅F₁₀O gas mixture under normal pressure is determined by the partial pressure of C₅F₁₀O in the gas mixture.³⁹ In this study, the partial pressure of C₅F₁₀O was set to 15 kPa, and the corresponding liquefaction temperature of the gas mixtures was −15 °C. In addition, because the pressure remains unchanged, the addition of oxygen will reduce the nitrogen concentration accordingly.

To investigate the effect of oxygen concentration on the breakdown voltage and decomposition characteristics of C₅F₁₀O/N₂/O₂, herein, four groups of gas mixtures with the oxygen concentrations of 0%, 0.5%, 7.5% and 19.33% were tested. The corresponding partial pressure of oxygen was 1 kPa when the oxygen concentration was 0.5%. The oxygen concentration of 7.5% corresponded to the oxygen partial pressure of 15 kPa (i.e. C₅F₁₀O : O₂ was equal to 1 : 1), and when the oxygen concentration was 19.33%, the partial pressure of oxygen was 38.66 kPa (actually, it was a gas mixture of C₅F₁₀O and air).

2.2 Test methods

The gas chamber was vacuum-pumped and then filled with buffer gas until the absolute pressure was greater than 0.2 MPa. Then, the abovementioned steps were repeated three times to ensure that each test was unaffected by previous tests and other impurities in the chamber. Because of the low saturated vapor pressure of C₅F₁₀O at normal temperature, it should be filled into the chamber first followed by oxygen and nitrogen. It was necessary to stand the gas mixture for 24 hours to make it mix evenly. The power frequency voltage was applied to the discharge chamber by a step-up method with a gap of at least 5 minutes between every two breakdowns. A total of 60 breakdown tests were conducted for each group of gas mixture. The average value of the breakdown voltage of the first ten breakdowns was taken as the breakdown voltage of C₅F₁₀O/N₂/O₂ under each oxygen concentration condition.

The C₅F₁₀O/N₂/O₂ samples after 60 times breakdown were obtained and tested by GC/MS. Moreover, the qualitative analysis of the gas was carried out by comparing the scanned results of the standard gas chromatography with the standard chromatographic database of the National Institute of Standards and Technology (Nist14.0).

3 Test results and analysis

3.1 Effect of oxygen concentration on the breakdown voltage of C₅F₁₀O/N₂/O₂

As shown in Fig. 3, the breakdown voltage of C₅F₁₀O/N₂/O₂ increases with the increasing oxygen concentration. The electronegativity of oxygen is greater than that of nitrogen, and its ability to attract electrons is stronger; thus, oxygen is more likely to hinder the development of discharge. Therefore, in the case of constant pressure and C₅F₁₀O concentration, the dielectric strength of the C₅F₁₀O gas mixture will increase with the increasing oxygen content.

Fig. 3 Relationship between the oxygen concentration and breakdown voltage of C₅F₁₀O/N₂/O₂.

Fig. 4 shows the breakdown voltage values of C₅F₁₀O/N₂/O₂ for 60 breakdown tests under different oxygen concentration conditions. Due to the random nature of discharge, the standard deviation is defined to characterize the dispersion of the breakdown voltage of C₅F₁₀O/N₂/O₂ at different oxygen concentrations. The formula for the calculation is as follows:

where N is the number of samples (N = 60), x_i is the breakdown voltage value of the ith time, x̄ is the average value, and σ is the standard deviation of the breakdown voltage of the experimental set.

Fig. 4 60 times breakdown voltage of C₅F₁₀O/N₂/O₂ under different oxygen concentrations.

When the oxygen concentration was 0%, the breakdown voltages of C₅F₁₀O/N₂/O₂ for 60 times breakdown were around 12.8 kV with σ = 0.40. For gas mixture with 0.5% oxygen, a few data points deviated from 12.83 kV, and σ = 0.46; with an increase in the oxygen concentration to 7.5%, the amount of the breakdown voltage data deviating from 13.97 kV increased, and σ = 0.83. When the oxygen concentration gradually increased to 19.33%, most of the breakdown voltage values deviated from 16.08 kV, and σ = 1.08.

Overall, the breakdown voltage of C₅F₁₀O/N₂/O₂ increased as the oxygen concentration increased; however, most of the breakdown voltage data deviated from the average breakdown voltage values, and σ became larger; this indicated that the dispersion of gas insulation strength gradually became worse.

3.2 Effect of oxygen concentration on the decomposition characteristics of C₅F₁₀O/N₂/O₂

Fig. 5 and 6 show the detection results of the C₅F₁₀O/N₂/O₂ components before and after breakdown obtained by GC/MS in the SCAN mode. As shown in Fig. 5, there are no other impurities in the gas.

Fig. 5 Components before breakdown.

Fig. 6 shows the component detection results of the C₅F₁₀O gas mixture with different oxygen concentrations after 60 times breakdown. Since no substance is detected before 4.4 min and after 7 min, to facilitate the description of the discharge decomposition components of the C₅F₁₀O gas mixture, only the GC/MS scan results from 4.25 to 6.25 min are shown in Fig. 6. Considering that the breakdown voltage is positively related to the energy generated during breakdown, we have plotted the scanning results of the discharge decomposition components of the C₅F₁₀O gas mixture at different oxygen concentrations in the same figure in turn. Therefore, the yield of each component can be compared intuitively by the peak area of each component.

Fig. 6 Components after 60 times breakdown with different oxygen concentrations.

As shown in Fig. 6a, when the oxygen concentration was 0%, the characteristic peaks of CF₄, C₂F₆, C₃F₈, C₂F₄, C₄F₁₀, C₃F₆ and C₃F₇H were found. Moreover, the peak area of C₂F₆ is largest. The peak area of CF₄ is significantly less than that of C₂F₆. However, when 0.5% oxygen was added to the gas mixture, the content of C₃F₇H was reduced relative to the case when the oxygen concentration was 0%. The content of other components is close to that of the C₅F₁₀O gas mixture without oxygen.

As shown in Fig. 6b, when the oxygen concentration is increased from 0.5% to 7.5%, the content of C₂F₆ is greatly reduced, and the peak area of CF₄ is close to that of C₂F₆. Moreover, the contents of C₃F₈ and C₄F₁₀ are greatly reduced. In addition, C₂F₄ and C₃F₆ are not detected in the gas mixture with 7.5% oxygen, and a small amount of CF₂O is detected at 4.757 min. The content of CO₂ increased when compared with that of the gas mixture with 0.5% oxygen. The contents of most of the decomposition components of the C₅F₁₀O gas mixture with 7.5% oxygen are lower than those of the gas mixture with 0.5% oxygen, except for CF₂O and CO₂. Thus, the addition of oxygen inhibits the formation of macromolecules such as C₂F₆, C₃F₈, C₄F₁₀, and C₃F₇H.

As shown in Fig. 6c, when the oxygen content increases from 7.5% to 19.33%, the peak area of CF₄ is closer to that of C₂F₆. The peak areas of CF₄, C₂F₆, CF₂O, C₃F₈, CO₂, C₄F₁₀ and C₃F₇H increased to different extents. Among them, the peak area of CF₂O is significantly increased. Note that oxygen cannot be considered to promote the production of C₂F₆, C₃F₈, C₄F₁₀, etc. As the oxygen concentration increases, the breakdown voltage increases as well; thus, the greater energy generated during the discharge breakdown causes more C₅F₁₀O to decompose; this results in an increase in the decomposition of the C₅F₁₀O gas mixture.

To more clearly illustrate the inhibitory effect of oxygen on the formation of C₂F₆, C₃F₈, C₄F₁₀ and C₃F₇H, the results of the decomposition components of the C₅F₁₀O gas mixtures with the oxygen concentrations of 0% and 19.33% were compared, as shown in Fig. 7.

Fig. 7 Components after 60 times breakdown with different oxygen concentrations.

As shown in Fig. 6c, when the oxygen concentration increased from 7.5% to 19.33%, the breakdown voltage increased by 2.11 kV; this resulted in a significant increase in the contents of the decomposition components of the C₅F₁₀O gas mixture. However, as shown in Fig. 7, the breakdown voltage of C₅F₁₀O/N₂ increased from 12.8 kV to 16.08 kV (an increase by 3.28 kV) when the oxygen concentration was increased from 0% to 19.33%. However, the contents of C₂F₆, C₃F₈, C₄F₁₀ and C₅F₁₂ for the C₅F₁₀O gas mixture with 19.33% oxygen were not more than those of the C₅F₁₀O gas mixture without oxygen addition.

The electrode surfaces with and without oxygen were also compared. As shown in Fig. 8, it was found that a small amount of black substance was adhered to the surface for the gas mixture without oxygen addition after 60 times of breakdown (Fig. 8 left). When the oxygen concentration was 0.5% or more, only discharge ablation traces could be observed (Fig. 8 right). Therefore, it can be believed that oxygen can inhibit the precipitation of carbon black when C₅F₁₀O decomposes. This may be because the generated carbon particles in the discharge can react with oxygen to produce carbon dioxide in the presence of a high-energy electric field.

Fig. 8 (Left) Electrode surface after 60 times breakdown of C₅F₁₀O/N₂ (right) electrode surface after 60 times breakdown of C₅F₁₀O/N₂/O₂.

Moreover, when the oxygen concentration was 0.5%, C₂F₄ and C₃F₆ still existed, and CF₂O was not detected. When the oxygen concentration was increased to 7.5% or more, C₂F₄ and C₃F₆ were not detected in the decomposition components, whereas CF₂O began to appear, and its content increased with an increase in oxygen concentration.

Under certain conditions, O₂ will react with C₂F₄ and C₃F₆. Several studies have reported the reaction between O atoms and C₂F₄ or C₃F₆. In the literature,⁴⁰ a mercury lamp was used to irradiate N₂O to produce O atoms, which could react with C₂F₄ or C₃F₆ and produce CF₂O. Moreover, in literature,⁴¹ it has been mentioned that the carbon–carbon double bond (C=C) of C₂F₄ and C₃F₆ weakens after ionization. Another study⁴² has pointed out that the O atoms can attack C=C in C₃F₆ leading to the generation of CF₂O and CF₃CF:. Moreover, CF₃CF: will rapidly undergo spin relaxation and then form C₂F₄. Due to the difference in the experimental conditions, these studies can only provide reference for the mechanism of CF₂O production in our experiments. Based on these studies as well as our team's computational study on the mechanism of CF₂O production when studying the decomposition characteristics of CF₃I,⁴³ the following pathways can be obtained.

O₂ + e → O˙ + O⁻ (2)

C₃F₆ + O → CF₃CF: + CF₂O (3)

CF₃CF: → C₂F₄ (4)

C₂F₄ → 2CF₂: (5)

CF₂:+ O₂ → CF₂O + O˙ (6)

4 Discussion

The uniformity of the electric field is vulnerable to the conductive particles in the insulating gap and causes local aggrandizement of the electric field.²⁶ Although under our experimental conditions, the carbon black adhered on the electrodes had less effect on the dispersion of the breakdown voltage, with an increase in the operating voltage and time, the amount of carbon black increased, and when it accumulated to a certain extent, the insulation safety of the equipment was seriously threatened.

Note that CF₂O is highly toxic, and its LC50 (Lethal Concentration 50, 4 h) value is only 270 mg m⁻³ (inhaled by rats). According to the Globally Harmonized System of Classification and Labelling of Chemicals toxicity classification standard, the acute inhalation toxicity of CF₂O is level 1. It has a strong stimulating effect on respiratory mucosa and can cause chemical pneumonia, pulmonary edema, and acute poisoning. Inevitably, the equipment contains a small amount of water, and CF₂O itself is highly corrosive and easily reacts with water to release highly toxic corrosive gases such as HF, thereby affecting the operating life of the equipment. However, note that although CF₂O is highly toxic, CF₂O is irritating as well, and the occurrence of gas leakage is easily detected in time. In addition, CF₂O is soluble in water and ethanol; thus, the waste gas can be treated harmlessly. Therefore, the toxicity of the decomposition products of a C₅F₁₀O mixture will not affect its application in the power industry to a large extent.

Although the breakdown voltage of C₅F₁₀O/N₂/O₂ increases with an increase in oxygen concentration and there is no inflection point value in the curve of breakdown voltage and oxygen concentration, this does not mean that it is reasonable to increase the oxygen concentration or even replace all N₂ to increase the insulation strength of C₅F₁₀O/N₂/O₂. In fact, when the oxygen concentration is 7.5% (C₅F₁₀O : O₂ is 1 : 1), the breakdown voltage of C₅F₁₀O/N₂/O₂ reaches 53.20% that of SF₆ under the same conditions (the power frequency breakdown voltage of SF₆ tested herein is 26.26 kV). It is possible to achieve the insulation requirements of the SF₆ equipment by increasing the pressure of the C₅F₁₀O gas mixture.⁴⁴ In addition, a partial discharge will inevitably occur during the long-term operation of the equipment, which will lead to the decomposition of the insulation medium. If C₅F₁₀O/N₂/O₂ with higher insulation strength is pursued by further increasing the oxygen concentration, the service life of the equipment and the safety of the field personnel will be greatly threatened. For this reason, it is recommended to use oxygen as the second buffer gas for C₅F₁₀O; however, it is not recommended to increase the dielectric strength of C₅F₁₀O/N₂/O₂ by greatly increasing the oxygen concentration. In the practical application of C₅F₁₀O/N₂/O₂, the 1 : 1 ratio of C₅F₁₀O : O₂ can be considered.

5 Conclusion

In this study, the breakdown voltages of C₅F₁₀O/N₂/O₂ at different oxygen concentrations were tested by a gas insulation performance test and gas composition analysis platform. The dispersion of the C₅F₁₀O/N₂/O₂ discharge breakdown voltage was analyzed, and the decomposition components were compared. The following conclusions can be obtained:

(1) When the absolute pressure is 0.2 MPa and the partial pressure of C₅F₁₀O is 15 kPa, the breakdown voltage of C₅F₁₀O/N₂/O₂ and its dispersion increase with an increase in oxygen concentration under a slightly uneven electric field.

(2) The carbon precipitates on the electrode surfaces disappear after multiple discharge breakdowns when 0.5% or more O₂ is added to the gas mixture.

(3) With an increase in the oxygen concentration, the peak area of CF₄ in the C₅F₁₀O/N₂/O₂ gas mixture gradually approaches that of C₂F₆. When the oxygen content reaches 7.5%, C₂F₄ and C₃F₆ disappear due to reaction with oxygen to form CF₂O. The addition of oxygen inhibits the formation of C₂F₆, C₃F₈, C₄F₁₀ and C₃F₇H to a certain extent.

(4) CF₂O generated after the C₅F₁₀O/N₂/O₂ discharge breakdown is highly corrosive and extremely toxic, which is harmful to the equipment and personnel. Therefore, it is not preferable to increase the dielectric strength of C₅F₁₀O/N₂/O₂ by further increasing the oxygen concentration. The recommended C₅F₁₀O : O₂ ratio is 1 : 1.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The current work was supported by the National Natural Science Foundation of China (No. 51707137, 51877157).

References

1. D. Chen, X. Zhang and J. Tang, et al., Using Single-Layer HfS₂ as Prospective Sensing Device Toward Typical Partial Discharge Gas in SF₆-Based Gas-Insulated Switchgear, IEEE Trans. Electron Devices, 2019, 66(1), 689–695.
2. J. Chu, X. Wang and D. Wang, et al., Highly selective detection of sulfur hexafluoride decomposition components H₂S and SOF₂, employing sensors based on tin oxide modified reduced graphene oxide, Carbon, 2018, S0008622318303968.
3. H. Cui, X. Zhang and G. Zhang, et al., Pd-doped MoS₂ monolayer: a promising candidate for DGA in transformer oil based on DFT method, Appl. Surf. Sci., 2019, 470, 1035–1042.
4. D. Wang, X. Wang and A. Yang, et al., A first principles theoretical study of the adsorption of SF₆, decomposition gases on a cassiterite (110) surface, Mater. Chem. Phys., 2018, 212, 453–460.
5. D. Chen, X. Zhang and J. Tang, et al., Pristine and Cu decorated hexagonal InN monolayer, a promising candidate to detect and scavenge SF₆ decompositions based on first-principle study, J. Hazard. Mater., 2019, 363, 346–357.
6. A. V. Tatarinov, I. V. Bilera and S. V. Avtaeva, et al., Dielectric Barrier Discharge Processing of trans-CF₃ CH=CHF and CF₃C(O)CF(CF₃) ₂, Their Mixtures with Air, N₂, CO₂ and Analysis of Their Decomposition Products, Plasma Chem. Plasma Process., 2015, 35(5), 845–862.
7. X. Zhang, Z. Cui and Y. Li, et al., Abatement of SF₆ in the presence of NH₃ by dielectric barrier discharge plasma, J. Hazard. Mater., 2018, 360, 341–348.
8. P. Simka and N. Ranjan, Dielectric strength of C5 perfluoroketone,//Proceedings of the 19th International Symposium on High Voltage Engineering, Pilsen, Czech Republic, 2015, pp. 23–28.
9. X. Zhang, Y. Li and D. Chen, et al., Dissociative adsorption of environment-friendly insulating medium C₃F₇CN on Cu (111) and Al (111) surface: a theoretical evaluation, Appl. Surf. Sci., 2018, 434, 549–560.
10. Y. Deng, D. Xiao and J. Chen, Insulation performance of CF₃I-N₂ gas mixtures as alternative for SF₆ in GIS/C-GIS, High Voltage Eng., 2013, 39(9), 2288–2293.
11. D. Chen, X. Zhang and J. Tang, et al., Noble metal (Pt or Au)-doped monolayer MoS₂ as a promising adsorbent and gas-sensing material to SO₂, SOF₂ and SO₂F₂: a DFT study, Appl. Phys. A: Mater. Sci. Process., 2018, 124(2), 194.
12. L. G. Christophorou and R. J. Van Brunt, SF₆/N₂ mixtures: basic and HV insulation properties, IEEE Trans. Dielectr. Electr. Insul., 1995, 2(5), 952–1003.
13. H. Cui, X. Zhang and D. Chen, et al., Adsorption mechanism of SF₆, decomposed species on pyridine-like PtN₃, embedded CNT: a DFT study, Appl. Surf. Sci., 2018, S0169433218309280.
14. J. Xiong, X. Li and J. Wu, et al., Calculations of total electron-impact ionization cross sections for Fluoroketone C₅F₁₀O and Fluoronitrile C₄F₇N using modified Deutsch–Märk formula, J. Phys. D: Appl. Phys., 2017, 50(44), 445206.
15. Y. Li, X. Zhang and D. Chen, et al., Theoretical study on the interaction between C5-PFK and Al (1 1 1), Ag (1 1 1): a comparative study, Appl. Surf. Sci., 2019, 464, 586–596.
16. C. Zhang, H. Shi and L. Cheng, et al., First principles based computational scheme for designing new SF₆ replacements, IEEE Trans. Dielectr. Electr. Insul., 2016, 23(5), 2572–2578.
17. Y. Li, X. Zhang and S. Tian, et al., Insight into the decomposition mechanism of C₆F₁₂O-CO₂ gas mixture, Chem. Eng. J., 2019, 360, 929–940.
18. H. Katagiri, H. Kasuya and H. Mizoguchi, et al., Investigation of the performance of CF₃I Gas as a Possible Substitute for SF₆, IEEE Trans. Dielectr. Electr. Insul., 2008, 15(5), 1424–1429.
19. Y. K. Deng and D. M. Xiao, Analysis of the insulation characteristics of c-C₄F₈ and N₂ gas mixtures by Boltzmann equation method, Eur. Phys. J.: Appl. Phys., 2012, 57(2), 20801.
20. R. Ullah, A. Rashid and A. Rashid, et al., Dielectric characteristic of dichlorodifluoromethane (R12) gas and mixture with N₂/air as an alternative to SF₆ gas, High Voltage, 2017, 2(3), 205–210.
21. D. Chen, X. Zhang and J. Tang, et al., High selective SO₂ gas sensor based on monolayer β-AsSb to detect SF₆ decompositions, IEEE Sens. J., 2018, 1.
22. Y. Li, X. Zhang and Q. Chen, et al., Study on the thermal interaction mechanism between C₄F₇N-N₂ and copper, aluminium, Corros. Sci., 2019, 153, 32–46.
23. Y. Deng and D. Xiao, Analysis of the insulation characteristics of CF₃I gas mixtures with Ar, Xe, He, N₂, and CO₂ using Boltzmann equation method, Jpn. J. Appl. Phys., 2014, 53(9), 096201.
24. D. Chen, J. Tang and X. Zhang, et al., Detecting decompositions of sulfur hexafluoride using MoS₂ monolayer as gas sensor, IEEE Sens. J., 2018, 1.
25. Y. Li, X. Zhang and J. Zhang, et al., Assessment on the toxicity and application risk of C₄F₇N: a new SF₆ alternative gas, J. Hazard. Mater., 2019, 368, 653–660.
26. L. Niemeyer, A systematic search for insulation gases and their environmental evaluation,//Gaseous Dielectrics VIII. Springer, Boston, MA, 1998, pp. 459–464.
27. Y. Li, X. Zhang and S. Xiao, et al., Decomposition properties of C₄F₇N/N₂ gas mixture: an environmentally friendly gas to replace SF₆, Ind. Eng. Chem. Res., 2018, 57(14), 5173–5182.
28. Y. Fu, X. Wang and X. Li, et al., Theoretical study of the decomposition pathways and products of C5-perfluorinated ketone (C5 PFK), AIP Adv., 2016, 6(8), 085305.
29. M. Saxegaard, M. Kristoffersen and P. Stoller, et al., Dielectric properties of gases suitable for secondary medium voltage switchgear, CIRED Paper, 2015, p. 926.
30. M. Kristoffersen, T. Endre and M. Saxegaard, et al., Ring main units with eco-efficient gas mixtures: field experience, CIRED-Open Access Proceedings Journal, 2017, 2017(1), 412–415.
31. F. Wang, X. Fu and G. Han, et al., Insulation performance of C₅F₁₀O/CO₂ gas mixture, Gaodianya Jishu, 2017, 43(3), 715–720.
32. M. Aints, I. Jõgi and M. Laan, et al., Effective ionization coefficient of C5 perfluorinated ketone and its mixtures with air, J. Phys. D: Appl. Phys., 2018, 51(13), 135205.
33. A. Chachereau, A. Hösl and C. M. Franck, Electrical insulation properties of the perfluoroketone C₅F₁₀O, J. Phys. D: Appl. Phys., 2018, 51(33), 335204.
34. X. Zhang, Y. Li and S. Xiao, et al., Decomposition mechanism of C₅F₁₀O: an environmentally friendly insulation medium, Environ. Sci. Technol., 2017, 51(17), 10127–10136.
35. X. Zhang, Y. Li and S. Tian, et al., Decomposition mechanism of the C5-PFK/CO₂ gas mixture as an alternative gas for SF₆, Chem. Eng. J., 2018, 336, 38–46.
36. M. Hyrenbach, T. A. Paul and J. Owens, Environmental and safety aspects of airplus insulated GIS, CIRED-Open Access Proceedings Journal, 2017, 2017(1), 132–135.
37. Z. Guo, X. Li and B. Li, et al., Dielectric properties of C5-PFK mixtures as a possible SF₆ substitute for MV power equipment, IEEE Trans. Dielectr. Electr. Insul., 2019, 26(1), 129–136.
38. P. C. Stoller, C. B. Doiron and D. Tehlar, et al., Mixtures of CO₂ and C₅F₁₀O perfluoroketone for high voltage applications, IEEE Trans. Dielectr. Electr. Insul., 2017, 24(5), 2712–2721.
39. L. Zhong, M. Rong and X. Wang, et al., Compositions, thermodynamic properties, and transport coefficients of high-temperature C₅F₁₀O mixed with CO₂ and O₂ as substitutes for SF₆ to reduce global warming potential, AIP Adv., 2017, 7(7), 075003.
40. K. A. Kobe, The properties of gases and liquids, Phys. Today, 1959, 12(4), 38–40.
41. G. K. Jarvis, K. J. Boyle and C. A. Mayhew, et al., Threshold Photoelectron−Photoion Coincidence Spectroscopy of Perfluorocarbons. 2. Unsaturated and Cyclic Perfluorocarbons C₂F₄, C₃F₆, 2-C₄F₈, and c-C₄F₈, J. Phys. Chem. A, 1998, 102(19), 3230–3237.
42. D. Saunders and J. Heicklen, Some Reactions of Oxygen Atoms. I. C₂F₄, C₃F₆, C₂H₂, C₂H₄, C₃H₆, 1-C₄H₈, C₂H₆, c-C₃H₆, and C₃H₈1a, J. Phys. Chem., 1966, 70(6), 1950–1958.
43. S. Xiao, Y. Li and X. Zhang, et al., Formation mechanism of CF₃I discharge components and effect of oxygen on decomposition, J. Phys. D: Appl. Phys., 2017, 50(15), 155601.
44. X. Li, D. Yun and H. Zhao, et al., Insulation Performance and Application of Environment-friendly Gases Mixtures of C₄F₇N and C₅F₁₀O with CO₂, High Voltage Eng., 2017, 43(3), 708–714.

---