Revised

Gliding Arc Plasma in Cholorinated Methanes Antonius Indarto†, Jae-Wook Choi, Hwaung Lee and Hyung Keun Song Korea Institute of Science & Technology, Clean Technology Research Center, P.O. Box 131, Cheongryang, Seoul 130-650, Korea (Received 17 October 2004)

Abstract Plasma processing of the chloromethane compounds (methylene chloride (CH2Cl2), chloroform (CHCl3), and carbon tetrachloride (CCl4)) diluted in the atmospheric air using gliding arc has been studied. Various values of injected initial chloromethane concentrations, total gas flow rates, and power frequency were used as the variables to investigate their discharge characteristic. This paper evaluates the phenomena of chloromethane processing by gliding arc plasma

Key words: Plasma, Gliding Arc, chloromethane, AC wave form, equilibrium voltage, voltage breakdown

1. Introduction The plasma of gliding arc is widely used now to destruct toxic materials. Many dangerous emissions, such as H2S [1], N2O[2], CHCl3 and CCl4 [3-4] have been investigated and studied. Usually, high destruction efficiency can be achieved by using this method. The gliding arc consists of a pair of flat electrodes which are connected to the power supply. In operation, the arc starts at the narrowest part of electrodes gap. It starts immediately after breakdown, a process that takes place when the electric field in the gap is high enough to ignite the arc. The current of the arc increases very fast at moderate voltage, sufficient to create a powerful arc which expands upward on the surface of electrodes and elongate until it can no longer be †

Corresponding author: E-mail:[email protected], Tel:+82-19-352-1981

maintained. At this point, the arc goes out, and the process is repeated. The number of arcs that will be produced depends on many factors, such as the frequency of the power supply applied, flowing gas species, and the total gas flow rate. During this movement, molecule reaction simultaneously occurs. Plasma arcs usually have energy high enough to destruct strong molecule bond or initiate a reaction of stable gas material due to high temperature of flame, higher electron density, etc. However, the papers that discuss the the behaviour of gliding arc are few in number. The results of the theoretical and numerical studies performed with the use of many mathematical equations describing gliding arc have been published in literature [5-9].) In this paper, we tried to explain physical characteristics of the plasma of compressed air with the chloromethane compounds diluted in it. The experiment was carried out with two triangular stainless steel electrodes, which were electrically charged from an AC power supply. According to an US Environmental Protection Agency (EPA) report, the chloromethane to be destructed was categorized as a compound of high thermal stability [10]. An analysis was carried out, which was focused on discharge parameters, such as the equilibrium voltage, breakdown voltage, and voltage-current-power (V-I-W) profile as functions of different concentrations of chloromethane, total gas flow rate, and power frequency.

2. Experimental setup The schematic diagram of the experimental setup is shown in Fig.1. Chloromethane compounds and atmospheric air were used as an input gas. Each system and component of the setup are described in detail in the following section. 2.1. Plasma reactor and power supply Figure 1 The reactor was made from a quartz-glass tube of inner diameter 45 mm and length 300 mm. The upper part and the bottom of the reactor equipped with a teflon seal comprised two

electrodes made of stainless steel. The length of the electrodes was 150 mm. The separation of the electrodes in the narrowest section was 1.5 mm. The gas mixture was fed between the electrodes through a capillary (nozzle tube) of 0.8 mm inner diameter. A thermocouple, located 10 cm above the electrode, was provided to measure the temperature of outlet gas. A high-frequency AC power supply (the Auto electric, A1831) with a maximum voltage of 10 kV and a maximum current of 100 mA was connected to the gliding arc electrode to generate plasma. The frequency could be adjusted from 10 to 20 kHz.

2.2. Input gas Chlorinated methanes, used as the initial material, are: a. Methylene chloride: CH2Cl2, molecular weight 84.93, purity 99.0%, purchased from the Junsei Chemical Co., Ltd., concentration 1, 2, 3, 4 % (volume percent). b. Chloroform: CHCl3, molecular weight 119.38, purity 99.0%, purchased from the Junsei Chemical Co., Ltd., concentration 1, 3, 5, 8 % (volume percent). c. Carbon tetrachloride: CCl4, molecular weight 153.82, purity 99.5%, purchased from the Kanto Chemical Co., Inc., concentration 1, 3, 5, 8 % (volume percent). Atmospheric air was used as a carrier gas and was controlled by a calibrated mass flow controller (the Tylan, FC-280S). The flow rates were 3, 4, and 5 L/min. Before entering the reactor, atmospheric air first passed through a scrubber and then was mixed with chloromethane compound. The chloromethane compounds were injected by a syringe pump (the KD Scientific, Model 100). The temperature of input stream was maintained higher than the temperature of compounds vaporization by means of heating tape surrounding the stream line.

2.3. System of Measurements The power supplied and AC voltage-current (V-I) waveform were registered by a digital oscilloscope (the Agilent 54641A) with a high-voltage probe (the Tektronix P6015A) having analogue bandwidth of 350 MHz and a current monitor (the Pearson 4997). The power consumed was also calculated by a wattmeter (the Metex M-3860M). The power measured with the oscilloscope was the real value absorbed in the reactor only and was defined as:

Discharge power = ∫ (V (t ) × I (t ) ) dt × frequency Watt

(1)

In this study the experimental data were taken 30 minutes after the initiation of the plasma of gliding arc referred to the onset outlet temperature of the bulk gas measured by thermocouple. 3. Results and Discussion 3.1. Characteristics of Power Supplies Figure 2 Figure 3 The specific characteristic of the gliding arc is the initial breakdown of the moving gas, which initiates this arc. Initial breakdown voltage was higher than equilibrium voltage. Figure 2 shows the arc movement along the electrode plates. The number of arcs produced could be easily found from the waveform of voltage and current (Figure 3). Following Chiranjeeve et al work, over-current should be obtained at the shortest distance between a pair of electrodes which represents of breakdown state of arc production. In this study, the AC supply voltage applied and current of breakdown and equilibrium state were not manually adjustable. Required voltage and current value to achieve the initial arc production was fixed at the specific value which is determined by configuration of system, such as: gap distance of electrode, gas flow rate, etc. After achieving initial breakdown, supply voltage and current decreased in the equilibrium state to a stationary value which could not be adjusted or changed by varying the parameter of the power supply. The frequency of the power supply was the only adjustable independent parameter. However, the frequency played an important role in the amount of arcs produced.

3.1. Influence of chloromethane compounds Figure 4 Figure 5 The power consumed or applied plays the main role in holding the stability or instability of gliding plasma. Although the concentration and flow rate were kept the same, different compounds

of injected material gives gave different power consumption. Figure 4 shows the result of the measurements of voltage carried out by oscilloscope for different gases. Slight differences in voltage and positions of maxima occurred. With rising concentration of chloromethane in the inlet stream, the difference was getting more and more, which is clearly shown in Fig. 5. From Fig. 5, it can be concluded that the compound containing CCl4 consumed the highest discharge power. The power consumed by different compounds was as follows (in descending order): CCl4 > CH2Cl2 > CHCl3 A good analytical explanation can be given on the base of the Paschen’s Law, according to which the potential is a function of the product of pressure and gap length [11]. V = f ( p, d )

(2)

In this experiment the the interelectrode gap was kept constant, and the pressure could also be assumed constant. Although the potential was the function of p and d, in the real experiment, some coefficient must be introduced to match the results of the experiment and mathematical calculations [12]. Rearrangement of Eq. 2 which inserts some coefficient will give: V=

B pd  A pd  ln    ln(1 / γ ) 

(3)

where γ is the Townsend secondary emission coefficient of electrons of Townsend and followed, which is written as follows : 1 =∈αd γ

(4)

The differentiation of Eq. 2 and setting the derivation equal to zero will give: ( pd ) m =

e 1 2.718 1 ln = ln A γ A γ

(5)

The minimum/ maximum voltage was obtained by substituting Eq. 5 into Eq. 3: Vm = 2.718

B 1 ln A γ

(6)

The voltage given in Eq. 6 is usually called as a voltage of breakdown (Vbd). In case of gliding arc, Vm > V. Less information about of A and B constants is available under?? in the case of gliding arc plasma. The parameters A and B must be determined experimentally [13]. From Eqs. 3 and 6, it can be seen that there is a relation between V and Vbd. When chloromethane compounds were injected with different concentrations, the values of Vbd obtained from the experiment were different. In this study, to check the relationship between V and Vbd, we used the following algorithm: By rearranging Eq. (6) in the form: A=

2.718 1 B ln Vm γ

(7)

and substituting it into Eq. (3) we will have: V=

B pd  2.718 B  ln    Vm 

(8)

For two different concentrations of chloromethane compounds, we have: B1 p1 d 1  2.718 B1  ln   V m1  V1  = B2 p 2 d 2 V2  2.718 B 2  ln    Vm 2 

(9)

The experiment was carried out under the same pressure and gap distance: p1 = p2 and d1 = d2. Parameter B is the function of effective ionization potential (V*) and pressure. This potential ensures the transport electrons through the gap and thus ionization is produced. As we used the same gap distance, pressure, and concentrations of chloromethane compounds differ only slightly, it could be assumed that B1 ≈ B2.

In this case Eq. (9) can be written in the form:

1 ln   V1  = Vm 2  1  Vm1 ln   V2 

(10) Figure 6 The comparison between the calculation and experimental results is shown in Fig. 6. When the experiment was carried out with varying concentrations but with the same chloromethane compounds, the results were close. The satisfactory result was also achieved when the experiment was carried out with different total gas flow rates and fixed concentration and chloromethane species. However, this result could not be obtained when we applied the same rates with different chloromethane compounds. This means that parameters A and C have specific values for each chloromethane gas and play an important role in the initiation of arcs cycle production.

Molecule chemical stability has important role on the breakdown process into plasma by releasing energetic species, such as electron and ion. Radu, et.al. have studied and mentioned in their work the effect of electron on the initiation of breakdown. Lack of free electrons that are necessary to initiate the breakdown will lead to the over-voltage across the electrodes gap, which will result in larger magnitude of voltage and current amplitude and more rapid rising times [14-17]. Taylor et al. have made a comparison between these compounds and graded these compounds in stability under the conditions of oxidation [18]: CCl4 = CH2Cl2 > CHCl3 and in the absence of oxygen: CCl4 > CH2Cl2 > CHCl3 Stability depends on the structure and chemical bonds of compounds. High chemical bond stability compared to other chlorinated methanes could explain why CCl4 gave the highest value of

V and Vbd. This result also matched with other experimental results testifying that CCl4 has higher energy consumption than CHCl3 [4].

3.2. Influence of total gas flow rate Figure 7 After the initial breakdown of the discharge gap,we failed to control equilibrium voltage and current by varying the parameters of the power supply, and setting of specified voltage and current was difficult. In this case, the total gas flow rate also was aa factor to be counted as a variable. Figure 7 shows the effect of the total gas flow rate on power profile. It can be easily seen that at 3 L/min, the total discharge power that was supplied to the system was higher compared to its value at 4 and 5 L/min. Figure 8 To study thoroughly this effect, we have tried to capture- the real voltage-current profile at under the conditions of equilibrium. Figure 8 shows the behavior of voltage-current wave obtained by experiment. The calculations of both real and average values of voltage wave gave that the total voltage supplied would be lower at lower total gas flow rate. But this difference was not noticeable. This phenomenon can be also explained by the Paschen’s law [12]. Usually, rising flow rate increases the pressure in the system. Increasing pressure can increase the breakdown-voltage (Vb) to initiate the arc creation which is higher compared at the lower pressure system. Moreover, the voltage at the equilibrium condition will be higher also compared at the lower pressure. The effect of current waveform can be supposed to be the main reason for rising or reducing the value of total discharge power. The comparison between Fig.s 8 (b), (d), and (e), it shows that at 3 L/min the number of sudden-fluctuated pulses was higher that in the two others cases. This means that at 3 L/min the system produced more arcs compared to 4 and 5 L/min. As mentioned above, as the effect of rising flow rate results in higher pressure, therefore, the possibility to produce arcs was getting weaker . That is why the number of sudden-fluctuated pulses was smaller and smaller with

rising of the total gas flow rate. However, sudden-fluctuated pulses also gave significant contribution to the calculated average total current fed to the system. Compared to the average current when arc was not occurred, the value of the total average current was 5 ~ 10 times higher in case with plasma.

3.3. Effect of frequency Figure 9 Figure 10 The frequency of the power supply was an adjustable factor in this experiment. Figure 9 shows the effect of frequency on power profile. The integration of Eq. 1 showed that the total discharge power increased linearly with rising frequency. When the conditions were kept constant, the number of arcs also increased. Radu et. al. mentioned that a change in the frequency will change the basic mechanism of the Townsend breakdown [14]. Rising frequency will increase sudden fluctuations of current and voltage peak per cycle. Integration calculation of power waveform using Eq.1 shows that higher number of peaks per cycle will give higher energy supplied, Fig. 10.. The measurements carried out with a wattmeter showed the same tendency as the oscilloscope measurements but power value was a little bit higher. Oscilloscope was used to measure only the energy that was supplied to the plasma. On the other hand, wattmeter measured the total power needed for all instruments, including the total power to operate the power supply.

4. Conclusion . The power discharge characteristics of gliding arc plasma have been studied using chloromethane compounds. Different concentrations, total gas flow rate and frequency have been used to investigate the behaviour of voltage-current-power (V-I-W) characteristic. Different kinds of chloromethane compounds gave significantly different values of discharge power, equilibrium

voltage, and breakdown voltage; CCl4 gave their highest values. In case of different concentrations and total gas flow rate, the the behavior of chloromethane compounds followed the Paschen’s law, which gave the relation between equilibrium voltage and breakdown voltage. Higher total gas flow rate decreased the discharge power. It reduced the number of of arcs produced that, which would reduced the sudden fluctuations in the current wave. Discharge power also increased with higher frequency.

Acknowledgement This work was supported by the National Research Laboratory of the Ministry of Science and Technology of Korea.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

V. Dalaine, J. M. Cormier, and P. Lefaucheux, A gliding discharge applied to H2S destruction, J. Appl. Phys., 83 (5), 2435 (1998) K. Krawczyk and M. Młotek, Combined plasma-catalytic processing of nitrous oxide, Appl. Catal. B, 30, 233 (2001) K. Krawczyk and B. Ulejczyk, Decomposition of Chloromethanes in Gliding Discharges, Plasma Chem. Plasma Process., 23 (2), 256, 2003. K. Krawczyk and B. Ulejczyk, Influence of Water Vapor on CCl4 and CHCl3 Conversion in Gliding Discharge, Plasma Chem. Plasma Process., 24 (2), 155, 2004. A. Fridman, S. Nester, L. A. Kennedy, A. Saveliev, O. M. Yardimci, Gliding arc gas discharge, Prog. Energy Combust. Sci., 25, 211 (1999) O. M-Yardimci, A. V. Saveliev, A. A. Fridman, and L. A. Kennedy, Thermal and nonthermal regimes of gliding arc discharge in air flow, J. Appl. Phys., 87 (4), 1632 (2000) I. V. Kuznetsova, N. Y. Kalashnikov, A. F. Gutsol, A. A. Fridman, and L. A. Kennedy, Effect of "overshooting" in the transitional regimes of the low-current gliding arc discharge, J. Appl. Phys., 92 (8), 4231 (2002) F. Richard, J. M. Cormier, S. Pellerin, and J. Chapelle, Physical study of a gliding arc discharge, J. Appl. Phys., 79 (5), 2245 (1996) S. Pellerien, F. Richard, J. Chapelle, J-M Cornier, and K Musiol, Heat string model of bidimensional dc Glidarc, J. Phys. D: Appl. Phys., 33 (19), 2407 (2000) P. H. Taylor, B. Dellinger, and C. C. Lee, Development of a thermal stability-based ranking of hazardous organic compound incinerability, Environ. Sci. Technol. 24(3), 316 (1990). v. F. Paschen, Ueber die zum Funkenubergang in Luft, Wasserstoff und Kohlensäure bei verschiedenen Druken erforderliche Potentialdifferenz, Wied. Ann, 37, 69 (1889). J. D. Cobine, Gaseous Conductor Theory and Engineering Application (Dover Publications, Inc., 1958), pp. 160-177 J. R. Roth, Industrial Plasma Engineering Volume 1: Principles (Institute of Physic Publishing, 1995), pp.237-256. I. Radu, R. Bartnikas, and M. R. Wertheimer, Frequency and voltage dependence of glow and

[15] [16] [17] [18]

pseudoglow discharges in helium under atmospheric pressure, IEEE Trans. Plasma Sci., 31 (6), 1363 (2003) R. Bartnikas, Partial discharges. Their mechanism, detection and measurement, IEEE Trans. Dielect. Elect. Insulation, 9 (6), 763 (2002) J. P. Novak and R. Barnitas, Breakdown model of a short plane-parallel gap, J. Appl. Phys., 62 (9), 3605 (1987) R. Barnitas and J. P. Novak, Effect of overvoltage on the risetime and amplitude of PD pulses, IEEE Trans. Dielect. Elect. Insulation, 2 (4), 557 (1995) P. H. Taylor and B. Dellinger, Thermal degradation characteristics of chloromethane mixtures, Environ. Sci. Technol., 22 (4), 438 (1988)

Figure Captions Figure 1. Schematic diagram of the experimental setup Figure 2. Movement of the gliding arc along the electrode plate recorded by a high-speed camera. Figure 3. Typical waveform of the AC power supply. The phenomena of arc production could be clearly seen from the fluctuations of current waveform. Fig 4. Voltage profile. Figure 5. Effect of injected chloromethane compounds (species, concentration, and total gas flow rate) on discharge power. Figure 6. Comparison between the calculation and experimental value of Vbd. Figure 7. Power profile as a function of the total gas flow rate. The data were obtained using 1% of CCl4 as an injected compound at a power frequency of 20 kHz. Figure 8. Voltage-current behaviour at 1% of injected CCl4 and a power frequency of 20 kHz. (a) V-3 L/min (b) I-3 L/min (c) V-4 L/min (d) I-4 L/min (e) V-5 L/min (f) I-5 L/min Fig 9. Effect of the frequency of the power supply applied on the power profile. The data were obtained using 10% of CHCl3 at a total gas flow rate of 2.5 L/min Fig 10. Effect of the frequency of the power supply applied on the total discharge power. The data were obtained using 8% of CHCl3 at a total gas flow rate of 2.5 L/min