Study of Non-Thermal DC Arc Plasma of CH 4 /Ar at Atmospheric Pressure Using Optical Emission Spectroscopy and Mass Spectrometry

Transcription

1 Study of Non-Thermal DC Arc Plasma of CH 4 /Ar at Atmospheric Pressure Using Optical Emission Spectroscopy and Mass Spectrometry LIAO Mengran ( ) 1, WANG Yu ( ) 2, WU Hanfeng ( ) 1, LI Hui ( ) 1, XIA Weidong ( ) 1 1 Department of Thermal Science and Energy Engineering, University of Science and Technology of China, Hefei , China 2 National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei , China Abstract Non-thermal C/H/Ar plasmas are widely applied to carbonaceous material production and processing. In this work, plasma parameters and gaseous species of the atmospheric non-thermal C/H/Ar plasmas produced by an atmospheric-pressure DC arc discharge generator in CH 4/Ar were investigated. The voltage-current characteristics were measured for different CH 4/Ar ratios. Optical emission spectroscopy was employed to analyze the electron excitation temperature, gas temperature and electron density under various discharge conditions. The hydrocarbon molecules produced in the CH 4/Ar plasmas were detected with photoionization mass spectrometry. The optical spectral results demonstrated that the electron excitation temperature was ev, the gas temperature was K and the electron density was in the range of (5-20) cm 3. The mass spectrum indicated that a variety of unsaturated hydrocarbons (C 2H 4, C 3H 6, C 6H 6, etc.) and several highly unsaturated hydrocarbons (C 4H 2, C 5H 6, etc.) were produced in the non-thermal arc plasmas. Keywords: non-thermal plasma, atmospheric pressure, hydrocarbon, methane, argon, electrical characteristic, optical emission spectra, photoionization mass spectrometry PACS: Kn, Kz, Mg DOI: / /17/9/05 (Some figures may appear in colour only in the online journal) 1 Introduction Over the last several decades, hydrocarbon/argon (C/H/Ar) plasmas have been widely studied and utilized in a variety of industrial applications, such as methane pyrolysis [1,2], diamond film deposition [3,4], synthesis of acetylene [5] and various carbon materials, such as nanotubes, C 60 etc [6,7]. More recently, atmospheric pressure C/H/Ar plasmas have attracted more and more interest for their high energy efficiency and economy without vacuum equipment. Various methods have been proposed to produce atmospheric pressure C/H/Ar plasmas, such as gliding arc [8], nonthermal DC/AC arc [9], glowing discharge [10], DBD discharge [11], RF discharge [12], etc. However, most of such plasmas deviate from thermal or chemical equilibrium and should be treated as non-equilibrium plasmas, which makes them more complex to investigate. Since masses of experiments on non-equilibrium C/H/Ar plasmas were performed with various devices in different conditions, it appears quite difficult to achieve definite and general conclusions from the diverse experiment results. Atmospheric pressure nonthermal DC arc discharge is a relatively convenient and direct method to generate stable non-equilibrium plasmas with gas temperature ( K) much lower than electron temperature (>10000 K). Moreover, the intermediate level of energy densities and concentrations of chemically reactive species in non-thermal arc exhibit great industrial value. In the long run, the utilization of C/H/Ar plasmas produced with atmospheric pressure non-thermal arc discharge will attract more and more attention due to the advantages mentioned above. At the current stage, there are some studies and data on atmospheric pressure C/H/Ar non-thermal arc discharge plasmas properties, e.g. gas temperature, electron temperature and density, molecular radical compounds, etc. Some experimental information concerned with C/H/Ar plasmas has been obtained in the investigation of diamond film deposition at atmospheric or close-to-atmospheric pressure. However, few reports supported by National Natural Science Foundation of China (Nos , , ) and USTC-NSRL Association Funding (No. KY ) 743

2 discussed both the plasma parameters and species information. Derkaoui [13] measured electron density and electron temperature in H 2 /CH 4 plasma by optical emission spectroscopy (OES) and MW interferometry, whereas the investigation lacked gas temperature and mass spectrum information. Farauk [14] obtained the methane conversion factor and C 2 H 2 density by mass spectrometry, but the electron density, electron temperature and gas temperature were deduced from a numerical model. Lombardi [15] et al. evaluated the gas temperature of Ar/CH 4 /H 2 plasma from a numerical simulation of the emission spectra of the (0-0) C 2 (d 3 Π g - a 3 Π u ) transition. Ramasamy [16] et al. calculated the gas temperature and electronic density by applying OES and numerical simulation for Ar/CH 4 and Ar/C 2 H 2 /H 2 plasmas at atmospheric pressure. In this work, we present a non-thermal DC arc plasma generator and diagnosis system for the purpose of characterizing the plasma properties and chemical processes of C/H/Ar plasmas. A mixture of methane and argon was used to produce C/H/Ar plasmas in our study considering the universality and practical value in industries. Optical emission spectroscopy was carried out for the determination of the CH 4 /Ar plasma properties. In particular, the electron density was extracted from the Stark broadening of H α line [17,18]. The electron excitation temperature was deduced from the ratios of two spectral lines of argon atoms (2λ method [19 21] ). The gas temperature was obtained using numerical simulation of the rotation spectra of C 2 Swan (d 3 Π g - a 3 Π u ) system. Meanwhile, a newly developed mass spectrometry technique, single-photon ionization time-of-flight mass spectrometry [22 24] (SPI- TOFMS), was used to analyze the complicated compounds of the CH 4 /Ar plasmas. 2 Experimental setup Fig. 1 shows the schematic of the experimental setup. The experiment was conducted using a 2 mm diameter tungsten cathode and a cylindrical hollow copper anode with 15 mm O.D. and 5 mm I.D. Arc discharge was generated in a quartz chamber with a fixed electrodes gap of 8 mm. Methane and argon were pre-mixed with different ratios before being introduced into the quartz chamber from the cathode surrounding. Then the gas flow was ignited under a high voltage provided by a DC power supply and an arc discharge plasma formed. The exhaust gas was expelled through the center channel of the anode, where a capillary was placed to introduce the products into the mass spectrometer. The arc discharge was maintained at low currents and voltages (l0-150 ma, V) in our study and the current-voltage data were recorded by a digital oscilloscope (Tektronix TDS 2014, not displayed on the diagram). To perform optical emission spectroscopy, a 1:1 image of the arc column was projected on a movable focus plate with fixed fibers by an achromatic lens (focus length of 300 mm). A 3/4 m spectrograph (Princeton Acton SP 2750) equipped with an intensified charge coupled device (ICCD, Princeton PI-MAX 2) and a computer were used for measuring and recording the emission spectra, while gratings of 1200 lines/mm and 1800 lines/mm were adopted for different demands of precision. The imaging size of the ICCD is pixels and µm/pixel, meanwhile the spectral resolution of the spectrometer is less than 0.06 nm within a wide spectral range from 200 nm to 1000 nm. Finally, the whole OES system was calibrated with a deuterium and tungsten halogen radiometric calibration source (Ocean Optics Inc. DH-2000-CAL). Fig.1 Schematic of the experimental setup For mass spectrometry analysis, the SPI-TOFMS was applied in this work. As shown in Fig. 1, the details of this setup can be found elsewhere [22], so only a simple description is given here. Gaseous samples from the capillary were introduced into the ionization chamber and ionized by the ultraviolet light emitted from a krypton lamp. Then the photon-generated ions were pushed into the mass spectrometer chamber and analyzed by a TOFMS. It has been proved that the SPI-TOFMS can ionize molecules softly with little or even no fragments [23,24], which makes it much easier to identify complicated compounds in plasmas without pre-separation. However, it must be pointed out that only molecules with ionization potential lower than 10.6 ev could be detected in view of the krypton lamp used in our experiment. 3 Results and discussion 3.1 Electrical characteristics Fig. 2 illustrates the voltage-current characteristics of the atmospheric pressure non-thermal arc plasmas with the CH 4 /Ar ratio of 0.74%, 1.48% and 2.20%. It shows that the voltage decreases with increasing discharge current, well known as the negative resistance characteristic. On the other hand, it can be seen from Fig. 2 that the voltage increases dozens of volts while the CH 4 /Ar percentage increases. 744

3 LIAO Mengran et al.: Study of Non-Thermal DC Arc Plasma of CH 4 /Ar at Atmospheric Pressure Using OES Fig.2 The voltage-current characteristics of the nonthermal CH 4/Ar DC arc discharge 3.2 Optical emission spectral characteristics and T exc, T g, n e measurements A typical emission spectrum ( nm) for the non-thermal DC arc plasma (the CH 4 /Ar ratio of 1.48% and current of 100 ma) is presented in Fig. 3. As can be seen, the infrared spectrum ( nm) is dominated by the atomic spectrum of Ar lines. Meanwhile the visible spectrum ( nm) is much more complicated but less intense than the infrared part. It includes the H-Balmer lines and vibrational-rotational spectrum of diatomic molecules such as C 2 Swan system (d 3 Π g - a 3 Π u ) and CH A-X system (A 2 Π - X 2 Π), which demonstrates the existence of the species of H, C 2, CH radicals. Fig.3 Typical emission spectra at discharge current of 100 ma and the CH 4/Ar ratio of 1.48% The 2λ method (i.e. the ratio of the intensities of two spectral lines) is used to approximately calculate the electron excitation temperature T exc in nonthermal plasma [19 21]. Furthermore, it has been proved that using spectral lines with a higher excitation energy difference can improve the accuracy for the 2λ method [25,26]. Thus the argon lines nm and nm are adopted in our experiment. As can be seen from Table 1, the line nm is produced by the transition from 4p to 4s and the line nm from 6s to 4p [27]. The difference of the excitation energy of upper state of and nm lines is much bigger than other pairs of lines which are produced by the transitions from 4p to 4s. On the other hand, the self absorption is ignored in our measurements, considering the optically thin assumption [25]. Thus the electron excitation temperature T exc can be deduced from the relation between two excitation levels of argon atoms, as described by: I 1 = A ( 1g 1 /λ 1 exp E ) 1 E 2, (1) I 2 A 2 g 2 /λ 2 kt exc where the subscript 1, 2 represent the nm and nm lines, respectively. I is the integrated intensity of lines, A is the transition probability, g and E are the degeneracy and excitation energy of the upper state, respectively. λ is the wavelength and k is the Boltzmann constant. The parameters E, g, A are taken from the NIST [27] and tabulated in Table 1. The measurements and calculations of the electron excitation temperature T exc in CH 4 /Ar plasmas were carried out as functions of discharge current and the CH 4 /Ar percentage. As presented in Fig. 4, T exc varies within the temperature range from 0.5 ev to 1 ev in our experiments, which is of the same order of magnitude as other atmospheric pressure non-thermal plasma [16,28,29]. Moreover, as can be seen from Fig. 4, T exc decreases with both discharge current and CH 4 /Ar percentage, which could be ascribed to the increase in the number of inelastic collisions between electrons and molecules [29,30]. The gas temperature T g of heavy species can be deduced from the rotational temperature T r due to the strong coupling between translational and rotational energy states [31]. In our experiment, two systems of diatomic molecular spectra, C 2 Swan system (d 3 Π g - a 3 Π u ) and CH A-X system (A 2 Π - X 2 Π), have been observed. However, the CH A-X band ( nm) is very weak and overlapped with the Ar lines (427.2 nm, nm and nm) and H γ line (434 nm). On the contrary, the C 2 Swan system is much more intense and Table 1. Parameters of the Ar nm and nm lines for the excitation temperature determination [27]. E and g are the excitation energy and degeneracy of the upper states, respectively, A is the transition probability λ (nm) Transition E (ev) g A (10 6 s 1 ) s 2 3p 5 ( 2 P 0 1/2)4p-3s 2 3p 5 ( 2 P 0 3/2)4s s 2 3p 5 ( 2 P 0 3/2)6s-3s 2 3p 5 ( 2 P 0 3/2)4p

4 Fig.4 The electron excitation temperature T exc and gas temperature T g versus the discharge current for different CH 4/Ar ratios. (The shape of square ( ), circle ( ), diamond ( ) denotes the CH 4 percentage of 0.74%, 1.48%, and 2.20%, respectively) little overlapped with other lines, so it is appropriate and adopted to analyze the rotational temperature by numerical simulation in our study [32 34]. The PGO- PHER [35 38] program is used for determining the values of rotational temperature by fitting user-specified molecule species, broadening width and rotational, vibrational and electronic temperature with the measured rotational spectrum of C 2 Swan band. A typical result of these simulations is presented in Fig. 5, which indicates T r 3170 K. Fig.5 Comparisons of the PGOPHER code-generated spectra and the experimental data, the simulated rotational temperature T r=3170 K As shown in Fig. 4, the gas temperature varies little in the range of K. It increases slowly with the current and CH 4 percentage due to the rise of the discharge power, opposite to the electron temperature. The electron density n e is derived from the Stark broadening of the H α line (656.3 nm). It should be pointed out that in many studies of hydrocarbon plasmas the H β line (486.1 nm) is more frequently used for Stark broadening analysis because of the large broadening width and well established theory [17,18,20]. However, in our experiment the H β line is much weaker than H α line due to the metastable Ar atoms and H 2 molecules [18]. Moreover, the H β line is slightly overlapped by the tail of the C 2 Swan band. As a consequence, the H α line is adopted for the Stark broadening analysis in our study. The spectral line broadening and shape involve several broadening mechanisms, including Doppler broadening, instrumental broadening and Stark broadening, etc [20,25,26]. The actual shape of the experimental spectral line can be described by a Voigt function [20,25], which is a convolution of the Gaussian profile (Doppler broadening and instrumental broadening) and the Lorentzian one (dominated by Stark broadening). The instrumental broadening due to the apparatus function ( λ app =0.05 nm) is determined by using a low-pressure Hg-lamp in this experiment, and the Doppler broadening, λ Doppler, is calculated by the following formula [39] : ( ) 1 2kT ln2 2 λ Doppler = 2 λ0 Mc 2, (2) where k is the Boltzmann constant, T is the gas temperature, M is the molecule mass, c is the speed of light and λ 0 is the wavelength (656.3 nm for H α ). Actually Doppler broadening is usually very weak ( λ Doppler =0.025 nm for T =3000 K), but it is still taken into account in this experiment for a higher precision. Thus, the Gaussian FWHM (full width at half-maximum) λ G is calculated by the relation, λ G = λ 2 app + λ 2 Doppler, then the Stark broadening λ Stark can be determined by fitting the experiment profile with the Voigt function with the fixed Gaussian part. Finally, the electron density n e is calculated by the relation for Stark broadening, as the following semiempirical formula [17,18] : n e [m 3 ] = ( FWHA [nm] ) (3) Here FWHA is the full width at half-area of H α line. However, since the Stark broadening profiles of H α line are close to Lorentzian, it is justified to use λ Stark = FWHW = FWHA in formula (3) [18]. Fig. 6 displays a typical measured H α line profile with a Voigt fitting function, which indicates n e cm 3. Fig.6 Typical Voigt fit of the recorded H α line broadening from the CH 4/Ar plasma. λ G=0.06 nm, λ Stark =0.091 nm, which indicates n e= cm 3 746

5 LIAO Mengran et al.: Study of Non-Thermal DC Arc Plasma of CH 4 /Ar at Atmospheric Pressure Using OES Fig. 7 shows the variation of the electron density with the discharge current and CH 4 /Ar percentage. It illustrates that the electron density n e increases with the discharge current, probably due to the rising gas temperature, whereas n e decreases with the rise of the CH 4 percentage, which could be ascribed to the higher ionization energy of methane molecules (12.61 ev) than the metastable argon atoms (11.7 ev) in CH 4 /Ar plasmas [27,40]. C 3 H 6, C 4 H 2 and C 6 H 6 increase fast with the CH 4 ratio, while the amounts of C 6 H 12 and C 10 H 6 increase slightly. Besides, the amount of C 2 H 4 increases with the CH 4 ratio at first but then decreases. Due to the limitations of the apparatus and theory, the detailed reaction mechanism in non-thermal CH 4 /Ar arc plasmas is still not very clear and much more work needs to be done for understanding the reaction mechanism in non-thermal CH 4 /Ar arc plasmas. Fig.7 The electron density n e versus the discharge current at CH 4/Ar ratio of 0.74%, 1.48% and 2.20% 3.3 Mass spectrum characteristics Fig. 8 shows a typical photoionization mass spectrum of the CH 4 /Ar plasma with the CH 4 percentage of 1.48% and the discharge current of 100 ma. Many kinds of hydrocarbons with mass from 28 to 126 have been generated in the arc plasma. As declared in section 2, molecules with ionization potentials higher than 10.6 ev (such as CH 4, C 2 H 2, and C 2 H 6 ) cannot be detected because of the limitation of the krypton lamp. Moreover, the molecules with very low mass (such as H 2 ) and short lifetime (such as radicals C 2, CH) cannot be detected because of the instrumental limitation [22]. Fig.8 Typical photoionization mass spectra of the gasphase components in the CH 4/Ar plasma for CH 4 percentage of 1.48% at 100 ma Fig. 9 presents the relative variation of 6 typical species (C 2 H 4, C 3 H 6, C 4 H 2, C 6 H 6, C 6 H 12, C 10 H 6 ) with the CH 4 /Ar ratio at the discharge current of 100 ma. As can be seen from Fig. 9, the relative amounts of Fig.9 Relative variation of 6 typical species (C 2H 4, C 3H 6, C 4H 2, C 6H 6, C 6H 12, C 10H 6) versus the CH 4/Ar ratio at discharge current of 100 ma 4 Conclusion In this study, plasma parameters (the electron excitation temperature, gas temperature, electron density) and gaseous species of atmospheric-pressure nonthermal C/H/Ar plasmas are investigated by a CH 4 /Ar DC arc discharge plasma. Electrical characteristic is measured at different CH 4 /Ar ratios. The OES results demonstrate that the electron excitation temperature is ev, the gas temperature is in the range of K and the electron density is in the range of (5-20) cm 3 under ma, 0.7%-3% of CH 4 /Ar. Besides, it also shows that T g and n e increase but T exc decreases with the discharge current. On the other hand, the increase of CH 4 percentage leads to the rise of T g and the decrease of T exc and n e. Photoionization mass spectrometry is used to detect the hydrocarbon molecules generated in the CH 4 /Ar plasma. A variety of unsaturated hydrocarbons (C 2 H 4, C 3 H 6, C 6 H 6, etc.) and several highly unsaturated hydrocarbons (C 4 H 2, C 5 H 6, etc.) have been identified in this experiment and some relative amount variations with the CH 4 percentage have been illustrated. Acknowledgments We would like to thank the National Synchrotron Radiation Laboratory for providing us with the SPI-TOF mass spectrometer. The authors are grateful to Dr. Y. Pan for his valuable discussions. 747

6 References 1 Dors M, Nowakowska H, Jasinski M, et al. 2014, Plasma Chemistry and Plasma Processing, 34: Khalaf P I, de Souza I G, Carasek E, et al. 2011, Quimica Nova, 34: Chen R F, Zuo D W, Li D S, et al. 2005, Acta Metallurgica Sinica, 41: Sun Z, Lin C H, Lee Y L, et al. 2000, Journal of Applied Physics, 87: Fincke J R, Anderson R P, Hyde T, et al. 2002, Plasma Chemistry and Plasma Processing, 22: Baldissarelli V Z, Benetoli L O D, Cassini F A, et al. 2014, Journal of the Brazilian Chemical Society, 25: Wang J C, Zhu Z P, Chang L P, et al. 2010, New Carbon Materials, 25: Bo Z, Yan J, Li X, et al. 2008, International Journal of Hydrogen Energy, 33: Ohtake N and Yoshikawa M. 1990, Journal of the Electrochemical Society, 137: Farouk T, Farouk B, Gutsol A, et al. 2008, Journal of Physics D: Applied Physics, 41: Kundu S K, Kennedy E M, Gaikwad V V, et al. 2012, Chemical Engineering Journal, 180: Longmier B W, Gallimore A D and Hershkowitz N. 2012, Plasma Sources Science and Technology, 21: Derkaoui N, Rond C, Gries T, et al. 2014, Journal of Physics D: Applied Physics, 47: Farouk T, Farouk B and Fridman A. 2010, IEEE Transactions on Plasma Science, 38: Lombardi G, Hassouni K, Benedic F, et al. 2004, Journal of Applied Physics, 96: Pothiraja R, Engelhardt M, Bibinov N, et al. 2012, Journal of Physics D: Applied Physics, 45: Gigosos M A, Gonzalez M A and Cardenoso V. 2003, Spectrochimica Acta Part B-Atomic Spectroscopy, 58: Konjevic N, Ivkovic M and Sakan N. 2012, Spectrochimica Acta Part B-Atomic Spectroscopy, 76: Sarani A, Nikiforov A Y and Leys C. 2010, Physics of Plasmas, 17: Griem H. 1997, Principles of Plasma Spectroscopy. Cambridge University Press, New York 21 van der Sijde B and van der Mullen J A M. 1990, Journal of Quantitative Spectroscopy and Radiative Transfer, 44: Zhu Z X, Wang J, Qiu K Q, et al. 2014, Review of Scientific Instruments, 85: Pan Y, Hu Y, Wang J, et al. 2013, Analytical chemistry, 85: Mülberger F, Streibel T, Wieser J, et al. 2005, Analytical Chemistry, 77: Ovsyannikov A and Zhukov M F. 2000, Plasma diagnostics. Cambridge Int Science Publishing, Cambridge 26 Konjević N. 1999, Physics Reports, 316: Zheng S J, Ding F, Xie X H, et al. 2013, Acta Physica Sinica, 62: 5 29 Jauberteau J, Cinelli M, Jauberteau I, et al. 1999, Journal of Physics D: Applied Physics, 32: Raizer Y P, Kisin V I and Allen J E. 1991, Gas Discharge Physics. Springer-Verlag, Berlin 31 Herzberg G. 1950, Molecular Spectra and Solecular Structure. Van Nostrand Reinhold, New York 32 Pellerin S, Musiol K, Motret O, et al. 1996, Journal of Physics D: Applied Physics, 29: Goyette A, Lawler J, Anderson L, et al. 1998, Plasma Sources Science and Technology, 7: Lange H, Saidane K, Razafinimanana M, et al. 1999, Journal of Physics D: Applied Physics, 32: Western C. 2010, University of Bristol, 7: Brooke J S, Bernath P F, Schmidt T W, et al. 2013, Journal of Quantitative Spectroscopy and Radiative Transfer, 124: Richley J C, Fox O J, Ashfold M N, et al. 2011, Journal of Applied Physics, 109: Wehres N. 2011, Optical Spectroscopy of Interstellar and Circumstellar Molecules: a Combined Laboratory and Observational Study. University of Groningen, Groningen 39 Ashkenazy J, Kipper R and Caner M. 1991, Physical Review A, 43: Auciello O and Flamm D L. 1989, Plasma Diagnostics: Discharge Parameters and Chemistry. Academic Press, California (Manuscript received 18 December 2014) (Manuscript accepted 6 February 2015) address of corresponding author LI Hui: huilli@ustc.edu.cn 748