Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid

Transcription

1 Plasma Science and Technology, Vol.15, Vol.15, No.8, Aug Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid LIU Jingjing ( ), HU Xiao ( ) Department of Mechanical and Electrical Engineering, Guangzhou University, Guangzhou , China Abstract A comparative study of the needle-to-liquid plasma in the continuous mode with DC and AC excitations is detailed in this paper. All plasmas studied here are shown to be glow discharges. This study is based on measurements of several key parameters, including electrical energy, optical emission intensities of active species, rotational and vibrational temperatures, and temperatures of the needle and liquid electrodes. AC plasmas can produce times higher excited state active species than DC plasmas under the same dissipated power. AC excited liquid plasmas have the highest energy utilization efficiency among the three systems (AC excited plasmas, DC excited plasmas with water anode and DC excited plasmas with water cathode); most of the energy is used to produce useful species rather than to heat the electrodes and plasmas. Keywords: electrical energy, active species, rotational temperature, vibrational temperature, energy use efficiency PACS: Dg DOI: / /15/8/10 1 Introduction Discharges with liquid electrodes have been an active research area, because the presence of liquid introduces a complex cluster of physical and chemical aspects to the dynamics of plasmas [1,2]. Discharges between a metal pin and a liquid electrode have been mostly investigated due to their wide applications, including water sterilization [3,4], organic contaminant degradation [5], chemical analysis [6], microorganism inactivation [7], hydrogen peroxide [8] and nano-particle production [9]. DC excited plasmas with pin-liquid electrodes are at present the most commonly studied because they are relatively cheap and easy to realize. They are often used for water treatment and organic contaminants degradation. Electrical and optical characteristics are widely adopted for their characterisation. Physical properties such as gas temperatures and electron density have been obtained in some investigations [10 15]. AC excited plasmas in contact with liquid from Hz to MHz range are also widely investigated. Optical emission analysis of AC (60 Hz) excited pinwater discharges under He, Ar and N 2 was studied by PARK et al [16]. A 50 khz driven atmospheric cold plasma jet was adopted to produce Fe nanoparticles [9]. Phenol degradation in aqueous solution by a gas-liquid phase dielectric barrier discharge (DBD) reactor at both 1.5 khz and 15.6 khz was investigated by JARAMILLO-SIERRA et al [17]. An atmospheric pressure glow discharge (APGD) micro plasma was generated by applying a radio-frequency of MHz to a powered mesh electrode filled with solution, its optical emission analysis was conducted via optical emission spectroscopy [18]. It is well known that in the gas phase the physical property (e.g. electron/ion density and temperature) of AC plasmas is more advantageous than those of DC plasmas [18 20]. For DC gas-phase plasmas sustained between two metallic electrodes, each electrode can only act as either the cathode or the anode. Once the applied voltage is below the discharge sustaining value, a DC excited plasma becomes extinguished immediately because most electrons and ions are lost and there are no sufficient space charges left in the gap to maintain the discharge. For a low-frequency AC discharge, the drop of the applied voltage below the plasma-sustaining voltage causes the discharge to be extinguished as well. Because of their low excitation frequency, discharge events in each half cycle of the applied voltage are often independent of each other and the discharge plasma is similar to that in the DC case, behaving as a positive DC and a negative DC in alternation. When the frequency of the applied voltage is above a critical value (usually several tens of kilohertz) [21], the discharge will remain on throughout the excitation cycle. Although there is a period while the applied voltage is lower than the discharge sustaining value, the period is short due to the high frequency of the applied voltage, the ions and/or electrons in the gap may not totally decay before the next discharge. The space charges remaining in the gap can facilitate the regeneration of plasmas in the next cycle, or help sustain the discharge at lower electric field. In such cases, discharges during each half cycle become interdependent on each other. This leads to better energy efficiency of electrical energy consumption. From a practical and economic point of view, both supported by National Natural Science Foundation of China (No ) and Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry of China

2 LIU Jingjing et al.: Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid the production of plasma species and their production efficiency need to be considered. Hence, it is necessary to investigate the difference of energy usage in the production of active species with AC and DC excited plasmas in contact with liquid. In the low frequency (<10 khz) range, the difference between DC and AC excited plasmas may not be significant because discharge events in the low frequency AC case may be independent of each other [21]. To explore their possible difference, a higher excitation frequency of 20 khz is considered in this study. Electrical and optical characteristics of the 20 khz excited needle-to-liquid plasmas are investigated. It is shown that the discharge remains on throughout the whole cycle in a continuous mode, and a comparative study of AC excited plasmas with DC excited plasmas is conducted for this mode. 2 Experimental setup Schematic of the experimental setup is shown in Fig. 1. The geometry of the pin-water electrodes is similar to those used in the literature [11 14]. The needle electrode is a stainless steel rod with diameter of 5 mm. The counter water electrode is a stainless steel plate immersed in 50 ml tap water in a plastic container. The distance between the needle tip and the water surface is 2.5 mm, which is adjusted by a micrometer. For DC excited plasmas, a DC power supply (Glassman high voltage INC, PS/MRS, OPO 60-22, 5 kv, 60 ma) is connected to the electrodes in series with a 25 kω ballast resistor. A wideband high voltage probe (Tektronix P6015A) is used to measure the applied voltage, a Tektronix P2220 voltage probe and a 10 Ω resistor are used for the current measurement. For AC excited plasmas, a 20 khz power supply is connected to the electrodes. This power supply consists of a function generator (Tektronix AFG 3102), a power amplifier (Amplifier Research, Model 500A100A, 500 W, 10 khz 100 MHz) and a high-voltage transformer (FLYPVM400, Information Unlimited). The output of the transformer is directly connected to the electrodes. A Tektronix oscilloscope is used to record the voltage and current values. A Tektronix P6015A wideband high voltage probe and a Pearson current Fig.1 Schematic of the experimental setup (color online) monitor are used for the measurements of the applied voltage on the electrodes (v) and the discharge current through the electrodes (i). An Andor istar ICCD camera is used to capture nanosecond images of plasmas [22]. Andor Shamrock SR-303i and Ocean Optics spectrometers are used to take optical emission spectra [23]. A collimating lens is placed in front of the optical fiber to collect all light emission from the plasmas. The horizontal distance between the lens and the central line of the needle electrode is 4 mm. 3 Electrical and optical characteristics of AC excited plasmas The voltage and current waveforms of AC excited plasmas with increasing input power are shown in Fig. 2. In Fig. 2(a) and (b) and at any given input power, the voltage appears to rise to a maximum then drop rapidly accompanied with a sharp current pulse. The drop in the voltage comes from the fact that the gap becomes much more conductive with plasma generation. The current pulse is formed because the capacitive gas gap has been charged to several kv then discharges through the low impedance of the plasma. The voltage and current waveforms after the current spike are of similar shape and the voltage and current pulses occur approximately simultaneously, indicating a resistive plasma. Fig. 2(a) and (b) also show that the first voltage peak and the first current spike decrease in magnitude with increasing input power. However, the second voltage peak and the second current spike actually increase with the input power. Such pulsedlike waveforms of the applied voltage and the current suggest a pulsed plasma, and the region of the dissipated power P = T v(t)i(t)dt = W (v(t) is 0 the applied voltage across the electrode, i(t) is the current flowing through the electrode) is therefore referred to as the pulsed mode. In the pulsed mode, a weak plasma is formed in each cycle after breakdown, and a small number of residual charges remain either in the gas gap or on the liquid surface to help ignite the discharge in the next cycle. An increase in the input power results in more dissipated power, suggesting that more residual charges remain in the electrode structure for the next discharge event. This reduces the breakdown voltage for the gas gap. At a larger input power, the first peak voltage is lower and the first peak current is smaller. However, an increase in the input power leads to an increase in the second peak voltage and the second peak current. Since the dissipated power increases with the input power, the second voltage peak and the second current peak appear to contribute greatly to the dissipated power. With further increase in the input power, much more charges are left from previous discharge events so that the gap is more conductive. The plasma becomes more intense and persists for a longer duration. Eventually the plasma lasts for the entire duration of the applied voltage, as shown in Fig. 2(c) and (d). This is the continuous mode of the needle-to-water plasma. It is clear 769

3 Plasma Science and Technology, Vol.15, No.8, Aug from Fig. 2(c) and (d) that with increasing input power, the applied voltage decreases and the discharge current increases. This suggests a progressively more intense plasma with more charges in the current and greater conduction of the gas gap (lower voltage). Time-resolved nanosecond images with 2 ns exposure time in the pulsed mode and the corresponding voltage and current waveforms are shown in Fig. 3. It is clear that the plasma disappears when the voltage polarity changes (points A and I). This is consistent with the occurrence of the voltage and current spikes in Fig. 2(a) and (b) and confirms the suggestion that the plasma is pulsed with distinct plasma-off phase. The plasma is very intense at the current spike points (C and K). Fig. 3 suggests that the plasma has different appearances in the positive and negative half cycles. In the positive half cycle, a bright layer of optical emission is seen near the tip of the needle electrode. This is the anode glow. For non-thermal atmospheric pressure plas- mas sustained between two parallel plate electrodes, the anode glow is usually too weak to be visible. In the case of the needle-to-water plasma in Fig. 3, the electric field at the instantaneous anode (the metal needle) is significantly enhanced by the sharp tip of the needle. Therefore the anode glow is very bright. The bright layer near the water surface is due to the negative glow as the water acts as the cathode in the positive half cycle. In the negative half cycle, the only bright layer is observed near the needle tip. There is no obvious bright layer near the water surface, which is the instantaneous anode, because the flat water surface supports only a weak electric field there and so the anode glow is very weak. During the falling edge of the current pulse, the plasma starts to disappear from the middle part of the electrode gap, but remains visible near the electrodes. After the turning point (E and M), the plasma begins to re-develop from the cathode to the anode (G and O). Then they quench again with decreasing current. Fig.2 Waveforms of (a) the applied voltage, and (b) the current in the pulsed mode with increasing input power P as indicated with an arrow (the dissipated power is 14.7 W (black), 15.2 W (red) and 17.5 W (green) respectively). Waveforms of (c) the applied voltage and (d) the current in the continuous mode with increasing input power (the dissipated power in each case is marked on the graph) (color online) Fig.3 Time-resolved nanosecond images in the pulsed mode with a dissipated power of 12.2 W. The exposure time is 2 ns (color online) 770

4 LIU Jingjing et al.: Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid Fig.4 Time-resolved nanosecond images in the continuous mode with a dissipated power of 22.2 W. The exposure time is 2 ns (color online) A sequence of nanosecond images in the continuous mode are shown in Fig. 4. These are similar discharge appearances to those in the pulsed mode, for example, the anode glow is visible in the positive half cycle and the negative glow of the cathode fall is very strong in the negative half cycle. The most distinct difference is that the plasma in Fig. 4 appears to exist in the electrode gap throughout the whole cycle of the applied voltage. This is true even at the polarity transition point (i.e. point A and F). It means that there are always some residual charges left in the electrode gap at 20 khz. This confirms the suggestion of a continuous mode based on the current waveform of Fig. 2(d). These residual charges help sustain the discharge. 4 Operation range of stable plasmas Images of the DC excited plasmas in the steady state and AC excited plasmas in the continuous mode at different dissipated powers are shown in Fig. 5. The exposure time is 4 ms. It is shown that the diameter of plasma column increases with dissipated power in all three plasmas. There is also an increase in the diame- ter of the bright layers near the metal and water electrodes. In the case of DC excited plasmas with a water cathode, the anode glow near the needle tip is strong and the negative glow near the water surface is visible but less intense. This is due to strong electric field enhancement at the tip of the needle electrode. For DC excited plasmas with a water anode, the negative glow near the needle tip (cathode) is intense with a larger diameter than that in the DC plasma with water cathode in Fig. 5(a). A dark region appears above the water anode, and this is the anode dark space. The visible layer near the water surface is indicative of an anode glow but influenced by the reflection of the cathode glow. The appearance of the AC excited plasma seems to be the combination of those of the two DC excited plasmas. It is of interest to note that the width of the near water bright layer changes little with power, different from the case of Fig. 5(b). Schematic diagrams for the images taken at about 28 W of dissipated power are also shown on the right side in Fig. 5. These are used to estimate the volume of the plasma. The discharge appearance as shown in Fig. 5(b) at 27.6 W provides a clear evidence of a glow discharge structure which contains the negative glow (NG), the Faraday dark space (FDS), the positive column (PC), the anode dark space (ADS) and the anode glow (AG). Fig.5 Images and schematic diagrams of (a) the DC excited plasma with water cathode, (b) the DC excited plasma with water anode, (c) and the AC excited plasma. (The exposure time is 4 ms) (color online) 771

5 Plasma Science and Technology, Vol.15, No.8, Aug Electrical characteristics of AC and DC excited plasmas in the continuous mode are shown in Fig. 6. In general, the electrical characteristics of AC excited plasmas appear to fall between those of the two DC excited plasmas. For all three cases, the applied voltage slightly decreases with increasing discharge current, and the dissipated power increases approximately linearly with increasing current. The voltage of the AC excited plasma is higher than that of the DC excited plasma with water anode but lower than that of the DC excited plasma with water cathode. At any given dissipated power, the DC excited plasma with water anode has the highest discharge current and the lowest plasma-sustaining voltage. The first points of the three curves indicate the lowest discharge current for the plasma to work in the continuous mode. The last points of the three curves are the highest discharge currents our DC and AC power supply can support. the dissipated power (see Fig. 6). Our data match well with the power densities of DC excited plasmas given in Refs. [24] and [25]. The power densities in these previous studies were calculated by using the voltage, current and the plasma volume estimated from digital images. The difference of about 10% between our measurement and those in the literature is excellent, given the assumption of the plasma being a circular truncated cone. Fig.7 Power densities for AC and DC excited plasmas obtained in this work (solid curve) and reported in literatures. (Open symbols: 1-Ref. [24], 2-Ref. [25]. Red is used for DC excited plasma with water cathode, and green is used for DC excited plasma with water anode) (color online) 5 Fig.6 Dependence of the discharge current on (a) the voltage and (b) the dissipated power for the AC excited plasma and the two DC excited plasmas (color online) According to the images in Fig. 5, the discharge columns are drawn on the right side of Fig. 5, as suggested by their corresponding schematic diagrams, the discharge volumes are estimated by the dimensions marked on the schematic diagrams and this allows the dissipated power density to be calculated, as shown in Fig. 7. It is clear that the power density does not vary very much with the discharge current for each of the three plasmas. The power density of the AC excited plasma falls betweeen those of the two DC excited plasmas. Given the approximately linear relationship between the dissipated power and the current, this suggests a linear increase of the discharge volume with 772 Gas and water temperatures Gas temperature of non-thermal atmospheric plasmas may be approximated by rotational temperature [26,27]. In low temperature air plasmas, the emission spectrum of the OH(A-X) transition in the range of nm provides a convenient way to measure the rotational temperature since there is little scope for self-absorption, non-boltzmann rotational population distribution, and interferences with other species in the plasmas [28]. However, for plasmas in contact with liquid, water vapour results in electronic quenching of OH(A) and this tends to prevent thermalization of the rotational population distribution of OH(A), and thus the rotational population distribution of OH(A-X) is strongly influenced by the formation process [29,30]. It is suggested that the rotational temperature of N2 is a better estimate of the gas temperature than the rotational temperature of OH [31]. Here, rotational temperature is measured using both the OH band and the N2 band for comparison purpose. Rotational temperature using the OH band is obtained by fitting the experimental spectrum with a simulated spectrum obtained using Lifbase [32]. Rotational temperature and vibrational temperature using the N2 band are obtained by comparison to a simulated spectrum obtained using Specair [33]. Rotational and vibrational temperatures at different input powers for AC and DC excited plasmas are shown in Fig. 8. The rotational temperature obtained using the OH band increases with increasing input power for all three plasmas. But the rotational and vibrational

6 LIU Jingjing et al.: Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid temperatures obtained using the N2 band remains approximately constant over the range of the dissipated electrical power considered, which means that the gas temperature does not change significantly, and the excess dissipated power is not used to heat plasma. The rotational temperature obtained using the OH band is much higher than that using the N2 band. It is however interesting that measurement using both the N2 band and the OH band suggests that the rotational temperature of the AC excited plasma falls between those of the two DC excited plasmas. This is consistent with the comparison study of electrical properties in Fig. 6 and Fig. 7. Fig.8 Rotational and vibrational temperatures at different dissipated powers of (a) an AC excited plasma, (b) a DC excited plasma with water anode, and (c) a DC excited plasma with water cathode (color online) Fig.9 Temperatures of the needle and the water after 3 min of plasma treatment (color online) Temperatures of the needle and water electrodes at different input powers after three minutes of plasma treatment are measured using an electronic thermometer and the results are shown in Fig. 9. It is clear that the needle and water temperatures increase with increasing input power in all three plasmas. Again the needle and water temperatures of the AC excited plasma fall between those of the two DC excited plasmas. Interestingly however, the needle temperature of the DC excited plasma is the highest when the needle is the cathode, and the water temperature is the highest when the water is the cathode. It is known that the electric field is the largest near the cathode surface for glow discharges. This means that the electrons and ions gain their maximum kinetic energy there, and these energetic electrons and ions drive large currents through the cathode and lead to significant heat generation there. Meanwhile, most voltage drop across the cathode fall is expected to contribute to most energy dissipation in this region. 6 Production of active plasma species It is well known that chemically active species like O and OH are capable of degrading organic compounds and inactivating microorganisms [34,35]. It is therefore of great interest to consider the efficiency of production of these species. The optical emission intensities of excited atomic oxygen at 777 nm and 845 nm, excited OH(A-X) at 309 nm and excited atomic hydrogen Hα at 656 nm are measured at different dissipated powers for these three different plasma sources. Although these optical emission intensities cannot be used to directly represent the densities of their corresponding active species, they can be used as indicators of the production of active species with three different plasma sources for comparison purpose. To facilitate comparison with future studies, absolution emission measurement is made in the nm range and simultaneous relative emission measurement in the nm range. As the absolution emission measurement below 350 nm is unavailable at the point of this investigation, the relative intensity data provide an indirectly calibrated extrapolation of absolute emission intensities spanning across the visible range. OH(A-X) production efficiency is defined as the relative emission intensity per watt with the emission intensity measured at 309 nm (i.e. peak value). Hα, O(777 nm) and O(845 nm) production efficiency is defined as the absolute peak emission intensity per watt of dissipated electrical power, because for a particular emission line of Hα or O, the shapes of all the spectra at different dissipated powers are very similar, and the increment of the peak intensity increases with the dissipated power. The production efficiency of the excited state species (OH(A-X), Hα, O(777 nm) and O(845 nm)) with three plasma sources are compared at different dissipated powers (in Fig. 10). At each dissipated power, three measurements are made to indicate statistical variation. The Hα production efficiency of the AC excited plasma is about times higher than that of the two DC excited plasmas. The Hα production efficiency in the DC excited plasma with water cathode is slightly higher than that in the DC excited plasma with water anode. Similarly the AC excited plasma produces an O emission at 777 nm that is about times that of the DC excited plasma with water anode and about times that of the DC excited plasma with water cathode. For atomic oxygen emission at 845 nm, its emission intensity per watt in the AC excited plasma is times that of the excited plasma with water anode and times that of the DC excited plasma with water cathode. The OH(A-X) production efficiency of the AC excited plasma is the largest, being 1.2 times larger than 773

7 Plasma Science and Technology, Vol.15, No.8, Aug that in the DC excited plasma with water cathode and 1.7 times larger than the DC excited plasma with water anode. It is clear that the DC excited plasma with water cathode can produce more hydrogen related excited species and less oxygen related excited species than the DC excited plasma with water anode. that of the DC excited plasma with water anode, but higher than that of the DC excited plasma with water cathode. d. AC excited plasma can produce more excited state active species than the DC excited plasma for a given dissipated power. e. AC excited plasma has the highest energy utilization efficiency among the three liquid-containing plasmas, most of the energy is used to produce useful species rather than to heat the electrodes and plasmas. References 1 2 Fig.10 Emission intensity per watt at Hα(656 nm), O(777 nm), O(845 nm) and OH(309 nm) lines as a function of dissipated power (color online) OH(A-X) and Hα species are mainly produced from water vapour dissociation [36,37], and excited state O species are generated mainly from the O2 dissociation [38 42]. When water is the cathode, the water temperature is higher than that in the DC excited plasma with water anode, and so the water vapour in the electrode gap is likely to be more, thus leading to greater production of OH(A-X) and Hα. Water vapour can cause two effects on OH(A-X) and excited state H production: a. water vapour is the main source to produce excited state OH and H species; b. water vapour results in the excited state OH and H quenching [43]. Water vapour also has two effects on excited state O species production: a. a larger amount of water vapour reduces proportionally the oxygen gas concentration in the gap, thus reduction of excited state O atoms production; b. water vapour also causes excited state O species quenching [43]. These are the reasons why the DC excited plasma with water cathode has higher production efficiency for OH(A-X) and Hα and lower production efficiency for excited O species than the DC excited plasma with water anode. 7 Conclusion From the above discussions about the AC and DC excited liquid-containing plasmas in the continuous mode, the following conclusions can be drawn: a. AC and DC above-liquid plasmas are glow discharges, with a clear glow structure that contains the negative glow, the Faraday dark space, the positive column, the anode dark space and the anode glow. b. Rotational temperature of the AC excited plasma is higher than that of the DC excited plasma with water anode but lower than that of the DC excited plasma with water cathode. c. Twater of the AC excited plasma is lower than that of the DC excited plasma with water cathode and close to that of the DC excited plasma with water anode. Tneedle of the AC excited plasma is lower than Flannigan D J, Suslick K S. 2005, Nature, 434: 52 Kawashima A, Nomura S, Toyota H, et al. 2007, Nanotechnology, 18: Brisset J L. 1998, J. Trace and Microprobe Techniques, 16: 363 Locke B R, Sato M, Sunka P, et al. 2006, Industrial and Engineering Chemistry Research, 45: 882 Abdelmalek F, Ghezzar M R, Belhadj M, et al. 2006, Industrial and Engineering Chemistry Research, 45: 23 Jenkins G, Manz A. 2002, J. Micromechanics and Microengineering, 12: N19 Chen C W, Lee H M, Chang M B. 2008, IEEE Trans. Plasma Sci., 36: 215 Locke B R, Shih K Y. 2011, Plasma Sources Sci. Technol., 20: Zhang Y T, Guo Y, Wang D W, et al. 2010, Chin. Phys. Lett., 27: Cserfalvi T, Mezei P, Apai P. 1993, J. Phys. D: Appl. Phys., 26: 2184 Titov V A, Rybkin V V, Maximov A I, et al. 2005, Plasma Chemistry and Plasma Processing, 25: 503 Bruggeman P, Ribezl E, Maslani A, et al. 2008, Plasma Sources Sci. and Tech., 17: Bruggeman P, Liu J, Degroote J, et al. 2008, J. Phys. D: Appl. Phys., 41: Bruggeman P, Graham L, Degroote J, et al. 2007, J. Phys. D: Appl. Phys., 40: 4779 Bruggeman P, Leys C. 2009, J. Phys. D: Appl. Phys., 42: Park J Y, Kostyuk P V, Han S B, et al. 2006, J. Phys. D: Appl. Phys., 39: 3805 Jaramillo-Sierra B, Mercado-Cabrera A, LopezCallejas R, et al. 2011, Eur. Phys. J. Appl. Phys., 56: Baba K, Okada T, Kaneko T, et al. 2007, Thin Solid Films, 515: 4308 Liang Y, Bertschinger G, Biel W, et al. 2004, Radiation power profiles in the plasma with the Dynamic Ergodic Divertor on TEXTOR. 31st EPS Conf on Plasma Phys, London Park S Y, You S J, Kim C S, et al. 2010, J. Crystal Growth, 312: 2459 van Roosmalen A J, Baggerman J A G, Brader S J H. 1991, Dry etching for VLSI, Plenum press, New York Shi J, Zhong F, Zhang J, et al. 2008, Physics of Plasmas, 15: Shi J J, Kong M G. 2007, Appl. Phys. Lett., 90: Bruggeman P, Liu J, Degroote J, et al. 2008, J. Phys. D: Appl. Phys., 41:

8 LIU Jingjing et al.: Contrasting Behaviours of AC and DC Excited Plasmas in Contact with Liquid 25 Bruggeman P, Ribezl E, Maslani A, et al. 2008, Plasma Sources Sci. and Tech., 17: Staack D, Farouk B, Gutsol A F, et al. 2006, Plasma Sources Sci. and Tech., 15: Dieke G H, Crosswhite H M. 1962, J. Quantitative Spectro & Radiative Transfer, 2: Laux C O, Spence T G, Kruger C H, et al. 2003, Plasma Sources Sci. and Tech., 12: Bruggeman P, Iza F, Guns P, et al. 2010, Plasma Sources Sci. and Tech., 19: Bruggeman P, Verreycken T, Gonzalez MA, et al. 2010, J. Phys. D: Appl. Phys., 43: Bruggeman P, Schram D, Gonzalez MA, et al. 2009, Plasma Sources Sci. and Tech., 18: Luque J, Crosley D R. 1999, LIFBASE: Database and Spectral Simulation Program. SRI International, SRI report No. MP (last access date is on 1 June 2012) 34 Sato M, Sun B, Ohshima T, et al. 1999, J. Advanced Oxidation Tech., 4: Deng X, Shi J, Kong M G. 2006, IEEE Trans. Plasma Sci., 34: Beenakker C I M, Deheer F J, Krop H B, et al. 1974, Chemical Phys., 6: Rowe B R, Vallee F, Queffelec J L, et al. 1988, J. Chemical Phys., 88: Eliasson B, Kogelschatz U. 1986, J. Phys. B-Atomic Molecular and Opti. Phys., 19: Titov V A, Rybkin V V, Maximov A I, et al. 2005, Plasma Chemistry and Plasma Processing, 25: Niemi K, Reuter S, Graham LM, et al. 2009, Appl. Phys. Lett., 95: He J, Zhang Y T. 2012, Plasma Process. Polym., 9: Zhang Y T, He J. 2013, Phys. Plasmas, 20: Binns W R, Ahl J L. 1978, J. Chemical Phys., 68: 538 (Manuscript received 16 July 2012) (Manuscript accepted 18 March 2013) address of LIU Jingjing: jolerliu@163.com 775