分子发射光谱法测量微波等离子体气体温度
邓磊, 张贵新*, 刘程, 谢宏
清华大学电机工程与应用电子技术系, 北京 100084
摘要

通过OH自由基 A2 Σ+ X2 Πr电子带系分子发射光谱测温法, 实现了对氩气、 氮气、 空气三种大气压微波等离子体气体温度的测量。 探究了不同微波功率、 不同气体流量下气体温度的变化规律, 测量了氮气、 空气微波等离子体羽流的轴向温度分布。 实验结果表明, 不同工作条件下微波等离子体核心温度普遍超过2 000 K, 空气微波等离子体可超过6 000 K; 同样工作条件下三种微波等离子体气体温度满足: TAr<TN2< TAir; 气体温度总体上随微波功率增加而小幅增加, 随气体流量下降而小幅降低; 氮气与空气等离子体羽流温度沿轴向迅速降低。 为验证分子发射光谱测温法的准确性, 以热电偶测温作为比对, 对温度较低的介质阻挡放电氩气等离子体进行了温度测量, 实验表明, 分子发射光谱法与热电偶所测结果十分接近。

关键词: 分子发射光谱法; 微波等离子体; OH自由基; 气体温度
中图分类号:O657.3 文献标志码:A
Measurement of the Gas Temperature in Microwave Plasma by Molecular Emission Spectrometry
DENG Lei, ZHANG Gui-xin*, LIU Cheng, XIE Hong
Department of Electrical Engineering, Tsinghua University, Beijing 100084, China
*Corresponding author e-mail: guixin@mails.tsinghua.edu.cn
Abstract

In this study, gas temperature measurements of argon, nitrogen, and air microwave plasmas are achieved by the molecular emission spectrometry of the A2 Σ+ X2 Πr electronic system of OH radical. The gas temperatures at different microwave powers or gas flow rates were explored, and the axial temperature distributions of nitrogen and air microwave plasma plumes were measured. The experimental results showed that temperatures in the core region of microwave plasma were higher than 2000 K at different working conditions, even up to over 6 000 K in air microwave plasma. At the same working condition, the three kinds of microwave plasma gas temperatures meet: TAr<TN2< TAir. The gas temperature increased slightly with the increase of microwave power, and decreased slightly with the decrease of gas flow rate overall. The gas temperatures of nitrogen and air microwave plasma plumes reduced quickly along the axial direction. In order to verify the accuracy of molecular emission spectrometry, the thermocouple was used as a comparison to take the temperature of the dielectric barrier discharge argon plasma. Experiments showed that the temperature measurements of molecular emission spectrometry and thermocouple are fairly consistent.

Keyword: Molecular emission spectrometry; Microwave plasma; OH radical; Gas temperature

Introduction

In recent years, researches about the microwave discharge plasma have been developed rapidly, and remarkable achievements have been made in many fields including material modification, gas synthesis, volatile organic compounds degradation and so on[1, 2, 3]. The microwave plasma is excited by the microwave field injected into the waveguide. Compared with dielectric barrier discharge (DBD) and pulse corona discharge plasmas, the microwave plasma has many advantages, such as without electrode, high energy density, high temperature, easy to produce plasma of large volume at atmosphere pressure[1, 4], so it has great practical value and extensive applied foreground.

The gas temperature of microwave plasma is an important physical parameter that affects plasma chemical reactions. Therefore, the measurement of gas temperature is very important for the practical applications of material modification, pollutant degradation and so on. Nevertheless, microwave plasma’ s ionization degree is high, and thereby the gas temperature is generally more than thousands of degrees[4]. It’ s too high to be measured by contact measurement methods such as the thermocouple.

For non-contact methods, the atomic emission spectrometry can take the electron excitation temperature of plasma[5], but can’ t take the translational temperature (i.e. gas temperature), the rotational temperature and the vibrational temperature.

Since the end of the last century, the method of molecular emission spectrometry has been gradually developing based on the emission spectra of OH radical, CN radical, N2 molecule, N2+ ion and so on. Some researchers have conducted measurements of various plasma temperatures[6, 7, 8].

H2O will be ionized to produce OH radicals in the discharge and high temperature environment. The reaction equation is as follows[9]

e-+H2OOH+H+e-(1)

A few scholars have researched spectral mechanism of the OH radical[6, 10, 11]. Based on the theory of molecular spectroscopy, OH radical’ s A2Σ +X2Π r electronic bands emission spectra are able to be accurately calculated. The relative intensity of spectral lines is determined by the rotational temperature Tr and vibrational temperature Tv of OH radical. Especially the OH radical’ s spectra of 306~310 nm interval are almost entirely decided by the rotational temperature[12], and this spectral interval is less interfered by other common molecular and atomic spectra. Therefore, the rotational temperature can be taken by measuring the 306~310 nm interval of OH radical’ s emission spectra and can be compared with the theoretical spectra at different rotational temperatures. Under the atmosphere of high pressure and high temperature, collisions between molecules are frequent, and exchanges of rotational energy and translational energy are sufficient. Therefore, according to the theoretical analysis, the molecular rotational temperature can be approximately regarded as the gas temperature[13].

In this context, simulation software LIFBASE[14] is used to calculate the theoretical spectra of the 306~310 nm interval of OH radical’ s A2Σ +X2Π r electronic bands emission spectra, and the spectra of argon, nitrogen and air microwave plasma are measured by a grating spectrometer and a charge coupled device (CCD). The gas temperatures are obtained by comparing the theoretical spectra with the spectra measured. The gas temperatures at different microwave power or different gas flow rates are explored, and the axial temperature distributions of nitrogen and air microwave plasma plumes are measured. In addition, in order to verify the accuracy of the molecular emission spectrometry, an experiment is designed to measure the temperature of DBD plasma by the molecular emission spectrometry and thermocouple.

1 Experimental setup

The experimental setup is shown in figure 1, which is divided into three parts: a microwave conduction system, a gas flow system and a spectra measurement system. In the course of experiment, the ambient temperature is 24 ℃and the humidity is 10%.

The system of microwave conduction showing in Fig.1 (a) consists of a 2.45 GHz microwave generator with the output power range of 0~3 000 W, WR430 waveguide components, including an isolator, a directional coupler, and a 3-stub tuner. The microwave is transmitted along the WR430 rectangular waveguide with the inner dimension of 86 mm width and 43 mm height. The waveguide is tapered to a shorted cross-section of 86 mm× 20 mm to increase the electric field intensity in the discharge tube. This design doesn’ t change the conduction mode of microwave, and is conducive to stimulate and maintain the plasma[15].

Fig.1 Schematic diagram of the experimental system
(a): The system of microwave conduction and gas flow; (b): The spectra mesurement system inner the reactor; (c): The spectra mesurement system of plasma plume along the axial direction

In order to avoid the contacts of plasma cavity in the reaction region, the fused quartz tube with 30 mm diameter and total length of 20 cm is used as the plasma discharge tube. The centre axis of the quartz tube is located one-quarter wavelength from the shorted end of the waveguide. Four flows of working gases are injected into the discharge tube from the tangential direction, creating a vortex flow, which is convenient for the heat dissipation of the quartz tube and stablity of the torch flame. In the experiment, the mass flow controller is used to control the flow rate of working gases of argon, nitrogen and air respectively. The adjustable range is 0~30.0 lpm. In order to make more OH radicals in the plasma, deionized water is added to the gas flow system.

In order to make the microwave plasma more easily excited, DBD is used to provide pre-ionization function.Argon gas is used as axial pre-ionization gas which is injected into a fine quartz tube with a diameter of 6 mm. Alternating voltage with a frequency of 44 kHz and adjustable amplitude of 0~20 kV is generated by a high frequency and high voltage power supply. The alternating voltage acts between a bottom ground metal plate and a copper foil electrode on the outer surface of the fine quartz tube, making the excitation of DBD argon plasma jet inner the fine quartz tube. The plasma jet flows into the microwave discharge tube along the axial direction, so the discharge tube has a large number of highly active particles, which is convenient for microwave plasma excitation. The alternating voltage is stopped after the excitation of microwave plasma.

The measurement system of spectra inner the waveguide is showed in Fig.1(b). After the excitation of microwave plasma, the fine quartz tube is replaced with optical fiber attached self collimating mirror, in order to measure the spectra inner the waveguide. The self collimation mirror is to reduce the divergence angle of the fiber. Spectra are measured by a Acton S2500 grating spectrometer and a PIXIS 400 CCD. The 300 and 2 400 lines· mm-1 gratings are used respectively for measuring large wavelength range spectra and small wavelength range spectra.

2 Results and discussion
2.1 Spectrum analysis

By the experimental setup, large volume of microwave plasmas are able to be generated at different kinds of working gases, as shown in figure 2. The plasma is blown out of the waveguide because of the gas flow. The photos are taken under the condition that the microwave power is 800 W and the working gas flow rate is 5 lpm.

Fig.2 Plasma pictures of different kinds of working gases with the microwave power of 800 W and the gas flow rate of 5.0 lpm

The 240~870 nm intervals of spectra of three kinds of microwave plasmas are obtained at the microwave power of 800 W and the gas flow rate of 5 lpm by means of the grating spectrometer and CCD with the method in figure 1(b). The results are showed in figure 3. The grating is 300 lines· mm-1 with the resolution of 0.13 nm, the exposure time of 100 ms and the instrument broadening of 0.6 nm measured by the 546.07 nm line of a mercury lamp using the linear fitting of Gaussian profile.

Fig.3 The spectra of argon, nitrogen, and air microwave plasmas with the grating resolution of 0.13 nm, equipment broadening of 0.6 nm and exposure time of 100 ms

As shown in figure 3, in the microwave argon plasma, there are a large number of Ar atom spectra in 700~870 nm interval, and the OH radical’ s spectra appear in 306~320 nm interval. The 300~450 nm interval shows the first negative system of N2+ ion, and the second positive system of N2 molecule. In addition, we also found Balmer lines of H atoms at 486.2 and 656.2 nm, and the O atomic lines at the 777.2 and 844.5 nm. The atomic spectral lines are determined by means of the NIST atomic spectra database[16]. The spectral lines of N2+ ions, N2 molecules and O atoms appear because the plasma is opened to the atmosphere. The appearance of the H atomic lines is due to the formations of H atoms by the decomposition of H2O molecules.

Compared with the microwave argon plasma, the spectra of nitrogen and air microwave plasmas are mainly shown as the continuous spectra probably caused by the common effection of molecular spectra, bremsstrahlung and compound radiation spectra.

2.2 Temperature measurements

2.2.1 Method

By the molecular spectra simulation software LIFBASE, the spectra of the (306~310) nm interval of OH radical’ s A2Σ +X2Π r electronic bands can be simulated. The spectral distribution of this interval is almost entirely decided by Tr, so Tv is set equal to Tr in software settings. The theoretical spectra of partial temperatures simulated by LIFBASE are shown in figure 4, with the spectral line broadening of 0.07 nm which is the instrument broadening of the 2 400 lines· mm-1 grating measured by the 546.07 nm line of the mercury lamp using the linear fitting of Gaussian profile.

Fig.4 The theoretical spectra of OH radical in 306~310 nm interval at different Tr with the broadening of 0.07 nm

The 306~310 nm intervals of spectra of three kinds of microwave plasmas are obtained by means of the grating spectrometer and CCD with the method in figure 1(b). The grating is 2 400 lines· mm-1 with the resolution of 0.01 nm, the exposure time of 1 000 ms, and the instrument broadening of 0.07 nm. After calibration and normalization, the spectra measured are compared with simulation spectra at different temperatures (step of 10 K). By the calculation of the deviation at all data points, the simulation spectra of the minimum deviation can be obtained.Hence, the temperature corresponding to the simulation spectra is the gas temperature of the plasma inner the waveguide.

A set of spectra measured of the argon, nitrogen and air microwave plasmas and their nearist simulation spectra are shown in figure 5. The temperatures measured of argon, nitrogen, air microwave plasma are 2 550, 3 870, 6 330 K respectively. The profiles of spectra measured and spectra simulated are considerably close.

Fig.5 Comparisons of theoretical and experimental spectra of argon, nitrogen and air microwave plasmas in 306~310 nm interval

2.2.2 Mesurements at different microwave powers and gas flow rates

Changing the microwave power and gas flow rate, the temperatures of three kinds of microwave plasmas are measured respectively. The results are shown in figure 6 and figure 7. It can be observed that the temperatures of three kinds of microwave plasmas at different working conditions meet TAr< TN2< TAir. As shown in figure 6, with the increase of microwave power, the temperatures show a slightly upward trend, which is due to the enhancement of the energy density caused by the increase of input power in the plasma region. It can be observed from figure 7 that with the increase of the gas flow, the temperatures show a slightly downward trend, which is caused by the reduction of the time of the gas passing through the microwave region.

Fig.6 Measured gas temperatures of argon, nitrogen and air microwave plasmas as a function of microwave power with the gas flow rate of 5.0 lpm

Fig.7 Measured gas temperatures of argon, nitrogen and air microwave plasmas as a function of the gas flow rate with the microwave power of 800 W

2.2.3 Measurement of the temperature distribution of plume along the axial direction

As shown in figure 1(b), the optical fiber mainly receives the plasma emission spectra inner the waveguide, so the temperatures measured mainly represent the temperatures inner the waveguide. Since both nitrogen and air plasma have plumes outside the waveguide, the optical fiber is moved along the axis of the plumes from the side of the quartz tube to measure the distributions of plasma plumes as shown in figure 1(c). Figure 8 shows that the temperatures of two kinds of plumes decrease rapidly with the increasing of distance from the waveguide. It’ s due to the plasma merely obtain energy inside the waveguide.

Fig.8 Measured gas temperatures along the axial direction in nitrogen and air microwave plasma plumes

2.3 Validation of accuracy by the mesurement of DBD argon plasma temperature

In order to verify the accuracy of the OH radical emission spectrometry, an experiment is designed to take the temperature simultaneously by emission spectra and a thermocouple. Because the measurable temperature range of the thermocouple is lower than 2 000 ℃, the DBD argon plasma whose temperature is usually relatively low is chosen as the measurement object. The experimental setup is shown in figure 9.

Fig.9 Schematic diagram of the gas temperature measurement ofDBD argon plasma

Argon gas is axially injected into a fine quartz tube with a gas flow rate of 1.5 lpm, and an argon plasma jet can be formed by DBD. The thermocouple model is WRNK-191, with the measurable temperature range of 0~1 100 ℃, the maximum relative error of 0.75%, and the diameter of 1 mm. The collimator is aligned with the head of the temperature-measuring rod to take the temperature of nearby plasma. The comparison results at different conditions of the two methods are obtained, as shown in figure 10.

Fig.10 Comparisons of measurements by the thermocouple and emission spectra

As shown in figure 10, temperature measurements of the OH molecular emission spectrometry and the thermocouple are very consistent, which demonstrates that the OH molecular emission spectrometry is reliable and fairly accurate.

3 Conclusions

In this study, by the OH radical molecular emission spectra, the gas temperatures of argon, nitregon and air microwave plasmas have been measured and researched. Through experiment, the conclusions are as follows:

(1) The gas temperatures of argon, nitregon and air microwave plasmas have been measured successfully. The spectra measured and spectra simulated are considerably close which proves the accuracy of simulation.

(2) The temperatures of the three kinds of microwave plamas are all high. They are generally more than 2 000 K at different conditions and the temperatures of air microwave plasma are able to be more than 6 000 K. At same condition, the temperatures of the three kinds of microwave plasmas meet: TAr< TN2< TAir.

(3) The gas temperature increases slightly with the increase of microwave power, and decreases slightly with the decrease of the gas flow rate overall. The gas temperatures of nitrogen and air microwave plasma plumes reduce quickly along the axial direction.

(4) The thermocouple is used as a comparison with the OH radical emission spectra to take the temperature of the DBD argon plasma. Experiments showed that temperature measurements of the molecular emission spectrometry and the thermocouple are fairly consistent which demonstrates that the OH molecular emission spectrometry is reliable and fairly accurate.

The authors have declared that no competing interests exist.

参考文献
[1] Uhm H S, Hong Y C, Shin D H. Plasma Sources Science Technology, 2006, 15: S26. [本文引用:2]
[2] Choi D H, Chun S M, Ma S H, et al. Journal of Industrial and Engineering Chemistry, 2016, 34: 286. [本文引用:1]
[3] XIE Hong-duan, SUN Bing, ZHU Xiao-mei. Journal of Hazardous Materials, 2009, 168: 765-769. [本文引用:1]
[4] Tendero C, Tixier C, Tristant P, et al. Spectrochimica Acta Part B, 2006, 61: 2. [本文引用:2]
[5] Green K M, Borras M C, Woskov P P, et al. IEEE Transactions on Plasma Science, 2001, 29(2): 399. [本文引用:1]
[6] PENG Zhi-min, DING Yan-jun, YANG Qian-suo, et al. Acta Physica Sinica, 2011, 60(5): 053302. [本文引用:2]
[7] Hong Y C, Chun S M, Cho C H, et al. IEEE Transactions on Plasma Science, 2015, 43(2): 696. [本文引用:1]
[8] Su Liu, Kumar R, Ogungbesan B, et al. Energy Conversion and Management, 2014, 78: 695. [本文引用:1]
[9] Li D, Nikiforov A, Britun N, et al. Journal of Physics D: Applied Physics, 2016, 49: 455202. [本文引用:1]
[10] Dieke G H, Crosswhite H M. Journal of Quantitative Spectroscopy & Radiative Transfer, 1962, 2(2): 97. [本文引用:1]
[11] Luque J, Crosley D R. Journal of Chemical Physics, 1998, 109(2): 439. [本文引用:1]
[12] ZHAI Xiao-dong, DING Yan-jun, PENG Zhi-min, et al. Journal of Tsinghua University·Science and Technology, 2012, 52(7): 980. [本文引用:1]
[13] Bruggeman P J, Sadeghi N, Schram D C, et al. Plasma Sources Science and Technology, 2014, 23: 023001. [本文引用:1]
[14] Luque J, Crosley D R. SRI International Report MP 99-009 (1999). [本文引用:1]
[15] Kim J H, Hong Y C, Kim H S, et al. Journal of the Korean Physical Society, 2003, 42: S876. [本文引用:1]
[16] NIST Atomic Spectral Database Lines Data (ASD). http://physics.nist.gov/PhysRefData/ASD/lines_form.html. [本文引用:1]