at SciVerse ScienceDirect

Journal of Electrostatics 71 (2013) 93e101

Contents lists available

Journal of Electrostatics

journal homepage: www.elsevier .com/locate/elstat

Characteristic study of cold atmospheric argon plasma jets withrod-tube/tube high voltage electrode

Yi Hong a,b, Na Lu a, Jing Pan b, Jie Li a,*, Yan Wu a, Ke Feng Shang a

a Institute of Electrostatics and Special Power, Dalian University of Technology, No. 2 Linggong Road, Ganjingzi District, Dalian 116024, PR ChinabDepartment of Physics and Electrical Engineering, Weinan Teachers University, Weinan 71400, PR China

a r t i c l e i n f o

Article history:Received 21 March 2012Received in revised form1 November 2012Accepted 7 December 2012Available online 20 December 2012

Keywords:Rod-tube/tube high voltage electrodeElectronic excitation temperatureAtomic oxygen densityMolecular nitrogen densityAverage electronic density

* Corresponding author.E-mail address: lijie@dlut.edu.cn (J. Li).

0304-3886/$ e see front matter � 2012 Elsevier B.V.http://dx.doi.org/10.1016/j.elstat.2012.12.009

a b s t r a c t

Atmospheric argon plasma jets are generated with the rod-tube/tube high voltage electrode and a ringground electrode at 8 kHz sinusoidal excitation voltage. It is found that the vibrational temperature,electronic excitation temperature, atomic oxygen density and spectral intensity with the rod-tube highvoltage electrode are enhanced significantly than that with the tube high voltage electrode. The atomicoxygen density, molecular nitrogen density, and average electronic density are about magnitude of1016 cm�3, 1015 cm�3, and 1012 cm�3 respectively, and the excited Ar, N2, OH and O are presented in theplasma plume with the rod-tube/tube high voltage electrode.

� 2012 Elsevier B.V. All rights reserved.

1. Introduction

Low temperature atmospheric plasma jets have received muchattention recently from several emerging novel applications such asmaterials processing [1], thin film deposition [2e4], etching [5],sterilization [6e10], and surface modification [11]. The plasma jetsare usually formed in open space surrounding air and have chem-ically active species such as OH and O.

Such an atmospheric plasma jets have been generated by thevarious electrode configurations and the generators. For example,Erdinc Karakas and Mounir Laroussi [12] used the plasma pencil tomeasure the plasma bullet lifetime and its velocity. The plasmapencil was made of hollow dielectric tube with two copper ringelectrodes attached to the surface of centrally perforated dielectricdisks separated by 5 mm. A unipolar square high voltage (HV) pulsein the order of 4.0e7.5 kVwas applied to the high voltage electrode.Chiang et al. [13] presented the nitrogen-based dielectric barrierdischarge plasma jet device, which was driven by a quasi-pulsedbipolar power supply (Model Genius-2, EN Technologies, Inc.) ata fixed frequency of 60 kHz. The plasma jet device was made of twoparallel copper plate electrodes with embedded cooling water andthe copper plate electrodes were coveredwith quartz plate. Lu et al.

All rights reserved.

[14] presented an RC plasma jet, which could be touched by barehands and could be directed manually by a user to place it into rootcanal for disinfection without causing any painful sensation. Themain body of plasma jet device was made of a medical syringe anda needle. The needle was used for guiding the gas flow and alsoserved as the electrode, whichwas connected to a high voltage (HV)submicrosecond pulsed direct current power supply througha 60 kU ballast resistor R and a 50 pF capacitor C, where both theresistor and the capacitor were used for controlling the dischargecurrent and the voltage on the needle. Deng et al. [15] investigatedbacterial inactivation by atmospheric pressure dielectric barrierdischarge plasma jet. Themain body of plasma jet reactor wasmadeof a quartz condenser tube used as the dielectric layer. The sodiumchloride (NaCl) solution in the outer layer of the reactor was actedas the external electrode to ground. The internal electrode wasa copper rod to connect the alternating current (AC) power supplyoperated between 16 and 20 kHz.

It is well known that dielectric barrier discharge (DBD) config-uration is to prevent the transition to spark and to homogenize thedischarge. In this paper, cold atmospheric argon plasma jets aregenerated with the rod-tube/tube high voltage electrode and a ringground electrode in DBD, which are powered by a sinusoidalexcitation voltage at 8 kHz. The aim of this work is the dischargecharacteristic study of the plasma jets generated with the rod-tube/tube high voltage electrode, such as the applied voltage, conductioncurrent, and displacement current waveform are determined by

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e10194

digital oscilloscope and equivalent circuit, and the gas temperature,excitation temperature, atomic oxygen density, molecular nitrogendensity, and average electronic density are obtained by opticalemission spectroscopy.

2. Experimental setup

A schematic of experimental apparatus and discharge photo-graph are shown in Fig. 1. In Fig. 1(a), first high voltage electrode isa stainless steel tube with outer diameter of 8 mm, inner diameterof 6 mm, and length of 128 mm. The second high voltage electrodeis a stainless steel rod with diameter of 2 mm, length of 200 mm,and a pencil-shaped tapered end, which is inserted into stainlesssteel tube. The tapered end of rod protrudes 7 mm from the bottomend of tube electrode and the rod-tube/tube (without the rod) highvoltage electrode is powered by a sinusoidal excitation voltage at8 kHz. The tube electrode is tightly covered by a quartz glass tubewith outer diameter of 10 mm, inner diameter of 8 mm, and lengthof 100 mm. The open end of quartz glass tube is a pencil-shapedtapered end as the gas outer hole, the diameter of which is2 mm. A copper foil of 20 mm length is wrapped tightly on theoutside of the quartz glass tube as the ring ground electrode, andwhich away from the gas outer hole is 10 mm. The bottom end oftube electrode is at the same height with the top end of groundelectrode. A quartz glass plate (thickness of 1 mm) is placed in theposition of 5 mm away from the gas outer hole.

The working gas of Ar (99.999%) is injected through the tubepower electrode controlled by mass flow controller at Ar flow rateof 1 lpm. The applied voltages are measured by using a high voltageprobe (Tektronix P6015A) and the currents are measured througha 50 U resistor in series with the ground electrode, and the elec-trical signals are recorded via a digital oscilloscope (TektronixTDS2012B). An optical fiber located in the position of 2 mm awayfrom the quartz glass plate is used to collect the optical emission ofthe plasma plume and the signals are recorded by a spectrometer(Acton INS-300-122B) with a grating of 1200 grooves per milli-meter and slit width of 20 mm. The spectrometer allows that thespectrum can be measured by hardware model and softwaremodel, respectively. In hardware model, the noise signal of spec-trum is small, but the spectrum is easily saturated and can be solvedby adjusting the pixels. Due to the difference of the pixels, thespectral intensity can not be compared between that with the rod-tube high voltage electrode and with the tube high voltage elec-trode. Accordingly, the spectral intensity is compared by software

Fig. 1. Experimental setup of the device (a) and discharge photograph at peak appliedvoltage of 6.2 kV and Ar flow rate of 1 lpmwith the rod-tube high voltage electrode (b).

model, in which the saturation problem of spectrum is notappeared, but the noise signal of spectrum is relatively larger thanthat in hardwaremodel. In this paper, all calculating and simulatinginvolved spectrum is obtained by hardware model and thedischarge images are taken by a Nikon digital camera COOLPIXS600.

3. Experiment results and discussion

3.1. Electrical discharge characteristics

It is well known that applied voltage and gas flow rate havea great influence on the length of atmospheric pressure plasma jet[16]. In this work, the plasma jet length versus Ar flow rate at peakapplied voltage of 6.2 kV and versus peak applied voltage aremeasured by discharge images, as shown in Fig. 2. As regardsFig. 2(a), the lengths of the plasma jets increase with the Ar flowrate from 0.5 lpm to 1 lpm and decrease quickly with the Ar flowrate from 1 lmp to 3 lpm. The lengths of plasma jets saturate whenthe Ar flow rate exceeds 3 lmp. The peak values of plasma jetlengths appear at Ar flow rate of 1 lpm and the peak value with therod-tube high voltage electrode is 2.3 mm larger than that with thetube high voltage electrode. In Fig. 2(b), the plasma jet lengthsincrease with the peak applied voltage from 4.6 to 6.2 kV andsaturate when the peak applied voltage exceeds 6.2 kV with therod-tube/tube high voltage electrode.

The conduction current is very important for dielectric barrierdischarge, because the effective power and the voltage of gas gapand barrier can be determined by the conduction current [17].Accordingly, the conduction current in this work is determined bythe equivalent circuit diagram shown in Fig. 3. According to theschematic of an asymmetric single dielectric barrier plasma actu-ator reported by Singh and Roy [18], we have imposed a virtualelectrode parallel to the ring ground electrode over the dielectricsurface. There are two capacitive components: one is between therod-tube high voltage electrode and the ring ground electrode Cd(which involves the two capacitive components: one is betweenthe rod high voltage electrode and the ring ground electrode andanother is between the tube high voltage electrode and the ringground electrode, and they are in parallel) and another is betweenthe virtual electrode and the ring ground electrode Cdv. The Cp andRp are the equivalent capacitor and resistor of the plasma, respec-tively. In fact, the Cp and Rp also involves two components: one isthe equivalent capacitor and resistor of the plasma generated by therod high voltage electrode and another is the equivalent capacitorand resistor of the plasma generated by the tube high voltageelectrode, and they are in parallel. The equivalent circuit of theplasma jet device is essentially a capacitor Cp in series witha resistor Rp and a capacitor Cdv, and a capacitor Cd parallel to the fullcircuit with the rod-tube/tube high voltage electrode. In Fig. 3,Utot(t), Itot(t), Idc(t) and Idp(t) are the total applied voltage, totalcurrent, conduction current and displacement current, respectively.Using Kirchoff’s theorem for the equivalent circuit given in Fig. 3,following equations are obtained:

dUtotðtÞdt

¼ IdpðtÞCd

; (1)

ItotðtÞ ¼ IdpðtÞ þ IdcðtÞ: (2)

where, Eqs. (1) and (2) are used to determine the conduction anddisplacement current when peak applied voltage is higher than3 kV. Fig. 4(a) shows the waveforms of applied voltage anddisplacement current in peak applied voltage of 2 kV. As regarding

Fig. 2. Plasma jet length versus Ar flow rate at peak applied voltage of 6.2 kV (a) and versus peak applied voltage (b) with the rod-tube/tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e101 95

the results shown in Fig. 4(a), the displacement current is measuredwithout discharge in the reactor when the air replaces the argon asthe working gas at peak applied voltage of 2 kV. The displacementcurrents are almost same between that with the rod-tube highvoltage electrode and with the tube high voltage electrode. It isindicating that the equivalent capacitor between the rod highvoltage electrode and the ring ground electrode is much smallerthan the equivalent capacitor between the tube high voltage elec-trode and the ring ground electrode.

The capacitance waveforms of capacitor Cd for two periods areshown in Fig. 4(b), which are deduced from the experimental datashown in Fig. 4(a) using Eq. (1). The average values of the Cd curveswith the rod-tube/tube high voltage electrode are determinedwithout considering the effect of peaks, which arises due tonumerical singularities occurring during the zero crossing of thedenominator [17,19].

As the peak applied voltage exceeds 3 kV, the displacementcurrent can not be measured, because the discharge occurs in thereactor when the air replaces the argon as the working gas.Accordingly, the conduction current and displacement current inpeak applied voltage of 6.4 kV are determined by Eqs. (1) and (2).Fig. 4(c) shows the waveforms of applied voltage, total current,conduction current, and displacement current. As presented inFig. 4(c), the amplitude of conduction current is much close to thetotal current, because the displacement current takes up a smallproportion of the total current. It is indicating that the averagepower can regard as the effective power when the applied voltageis adequately large. The waveforms of conduction currents aresimilar between that with the rod-tube high voltage electrode andwith the tube high voltage electrode, and the conduction currentwith the rod-tube high voltage electrode is a little larger than thatwith the tube high voltage electrode. It is found that the dischargeis dominated by the tube high voltage electrode even if the rod highvoltage electrode is added. In other words, the conduction current

Fig. 3. Equivalent circuit diagram of the plasma jet device.

generated by the rod high voltage electrode is much smaller thanthat generated by the tube high voltage electrode. Fig. 5 shows thevariations in effective power versus peak applied voltage. Theeffective power is determined by Lissajous plot through a 100 nFcapacitor in series with the ground electrode. As presented in Fig. 5,the effective power with the rod-tube high voltage electrode islarger than that with the tube high voltage electrode. As the peakapplied voltage increases from 4.6 to 7 kV, the effective powersincrease from 3.1 to 6.5Wwith the rod-tube high voltage electrodeand from 2.6 to 5.4 W with the tube high voltage electrode,respectively.

3.2. Gas temperature and electronic excitation temperature

The rotational temperature Trot(K) plays the role of plasmathermometer, the value of which is close to gas temperature[5,20,21], whereas the vibrational temperature Tvib(K), due to itsadiabatic character can trap energy and plays the role of energyreservoir [22]. To determine the rotational and vibrationaltemperatures of the nitrogen, the emission spectrum of secondpositive system 0e2 transition of nitrogen is used. By comparingthe simulated spectrum of C3Pu/B3PgðDv ¼ �2Þ band transitionof nitrogen with the experimental measured spectrum, the rota-tional and vibrational temperatures are obtained when the best fitis achieved. The variations in rotational and vibrational tempera-tures versus peak applied voltage with the rod-tube/tubehigh voltage electrode are shown in Fig. 6, and the inset graph inFig. 6(a) is the experimental and simulated spectrum ofC3Pu/B3PgðDv ¼ �2Þ band transition of nitrogen at peak appliedvoltage of 6.2 kV with the rod-tube high voltage electrode. InFig. 6(a), the rotational temperatures with the rod-tube highvoltage electrode are a little larger than that with the tube highvoltage electrode. It is may be due to the fact that the differences ofconduction current and effective power are very small betweenthat with the rod-tube high voltage electrode and with the tubehigh voltage electrode. When the difference of mean square rootvalue of total current was about 1.8 mA between that with andwithout addition of a ring ground electrode, the difference of gastemperature was smaller than 5 K in atmospheric helium plasmajet [23]. A larger power density led to a higher gas temperature inatmospheric argon plasma jet [24]. In this work, the difference ofpower density is very small between that with the rod-tube highvoltage electrode and with the tube high voltage electrode, becausethe volumes of plasma column are almost same in the both case.Therefore, the difference of rotational temperatures is much

Fig. 4. Typical waveforms of applied voltage and displacement current (a), capacitance waveforms of capacitor Cd for two periods (b) in the peak applied voltage of 2 kV, and typicalwaveforms of applied voltage, total current, conduction current, and displacement current in peak applied voltage of 6.4 kV (c) with the rod-tube/tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e10196

smaller. As shown in Fig. 6(b), the vibrational temperature with therod-tube high voltage electrode is obviously higher than that withthe tube high voltage electrode. It is may be due to the fact that theeffective power with the rod-tube high voltage electrode is a littlelarger than that with the tube high voltage electrode. As a functionof radio-frequency power, the vibrational temperature increasedfrom 2520 K at 20 W to 2940 K at 60 W (at 400 torr) [25]. It isindicating that the small variation of power causes the large

Fig. 5. Effective power versus peak applied voltage with the rod-tube/tube highvoltage electrode.

variation of vibrational temperature. In addition, the axial electricfield is enhanced with adding the rod high voltage electrode, so thegained energy of electron in plasma plume with the rod-tube highvoltage electrode is higher than that with tube high voltageelectrode. This causes that the vibrational temperature with therod-tube high voltage electrode is higher by electronemoleculescollision. It is may be another reason. As the peak applied voltageincreases from 4.6 to 7 kV, the vibrational temperatures increaselinearly from 1298 to 1342 K with the rod-tube high voltage elec-trode and from 1260 to 1296 Kwith the tube high voltage electrode,respectively.

The optical emissions spectrum obtained by hardware model atpeak applied voltage of 4.6 kV are shown in Figs. 7 and 8, respec-tively. It is shown that the spectrum is dominated by the excitedOH, N2, Ar, and O species with the rod-tube/tube high voltageelectrode. The neutral argon atomic emission lines 415.9 nm for thetransition (3s23p5)5p / (3s23p5)4s and 706.7, 714.7, 738.4 and751.5, 794.8, 800.6 nm for the transition (3s23p5)4p / (3s23p5)4sare chosen to determine the excitation temperature Texc(K) undera Boltzmann approximation [5,24,26]. Fig. 9 shows the variations inexcitation temperature as a function of peak applied voltage andthe inset graph is the Boltzmann plot of excitation temperature atpeak applied voltage of 6.2 kV with the rod-tube high voltageelectrode. The excitation temperature with the rod-tube highvoltage electrode is higher than that with the tube high voltageelectrode at same peak applied voltage. It is may be due to the factthat the effective power with the rod-tube high voltage electrode isa little larger than that with the tube high voltage electrode. Whenthe difference of average power was about 0.7 W between the Ar

Fig. 6. Rotational temperature versus peak applied voltage (the inset graph is the experimental and simulated spectra of second positive system 0e2 transition of nitrogen at peakapplied voltage of 6.2 kV with the rod-tube high voltage electrode) (a) and vibrational temperature versus peak applied voltage (b) with the rod-tube/tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e101 97

flow rate of 2 lpm and 4 lpm, the difference of excitation temper-ature was smaller than 50 K in atmospheric argon plasma jet [24].In addition, the excitation temperatures grow up linearly with theapplied voltage with the rod-tube/tube high voltage electrode,which is in well agreement with the conclusion reported by Qianet al. [26].

Fig. 10 shows the optical emission spectrum obtained by soft-ware model in peak applied voltage of 5.4 kV. As the results shownin Fig. 10, the spectral intensity with the rod-tube high voltageelectrode is much higher than that with the tube high voltageelectrode. This shows that the electrons with the rod-tube highvoltage electrode are at the higher energy state than that with thetube high voltage electrode. It is in well agreement with theaforementioned results that the effective power, vibrationaltemperature, and excitation temperature with the rod-tube highvoltage electrode are higher than that with the tube high voltageelectrode.

3.3. Atomic oxygen density and molecular nitrogen density

It is well known that oxygen atoms play an important role inplasma biological application [8e10]. Harshbarger et al. [27] firstsuggested the use of optical emission spectroscopy to monitor thevariation of atomic oxygen concentration by comparing the emis-sion intensity from the argon and the oxygen atoms. Katsch et al.

Fig. 7. Optical emission spectrum 300e650 nm (a) and 650e900 nm (b) in hardware model425 nm with the rod-tube high voltage electrode.

[28] chose the atomic oxygen line l ¼ 844.6 nm and the argon linel ¼ 750.4 nm to determine the density of ground state oxygenatoms and gave the condition of applicability:

(1) The excitation cross-sections have the same shape, particularlyclose to the threshold.

(2) The population of excited levels fromhigher levels is negligible.(3) Two-step excitation, e.g., via metastables, is negligible.(4) There is no population of atomic levels via dissociation.(5) Radiationless de-excitation (quenching) of excited levels is

negligible.

Qayyum et al. [29] chose the argon line l ¼ 419.8 nm and thenitrogen line l ¼ 337.1 nm to determine the molecular nitrogendensity, because the excitation thresholds for the two radiativespecies were reasonably close (13.2 and 11.1 eV) and the shapes ofthe electron-impact cross-sections were also rather similar.

In this work, the dissociative electron-impact excitation process(eþ O2 / O(3p3P) þ Oþ e) have a higher threshold of 16.3 eV thanthe direct excitation process (e þ O / O(3p3P) þ e), so the disso-ciative electron-impact excitation process are neglected [30].Besides, in this argon plasma plume the nitrogen and oxygenmolecular density comparing to the argon density are very small, sothe collisional induced quenching of the upper Ar(2p1), O(3p3P),andΝ2ðC3Pu; n ¼ 0Þ in the nitrogen and oxygen are also neglected.

at peak applied voltage of 4.6 kV and the inset graph is optical emission spectrum 410e

Fig. 8. Optical emission spectrum 300e650 nm (a) and 650e900 nm (b) in hardware model at peak applied voltage of 4.6 kV and the inset graph is optical emission spectrum 410e425 nm with the tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e10198

Considering the collisional induced quenching of t the upperAr(2p1), O(3p3P), and Ν2ðC3Pu; n ¼ 0Þ in the argon, the intensity ofatomic oxygen line l ¼ 844.6 nm can be expressed as

IðΟÞ ¼ COhgΟAðΟÞij nek

ðΟÞe nΟ

�AðΟÞi þ kΟqnAr

��1: (3)

The intensity of argon line l ¼ 750.4 nm is given by

IðArÞ ¼ CArhgArAðArÞij nek

ðArÞe nAr

�AðArÞi þ kArqnAr

��1: (4)

The intensity of nitrogen line l ¼ 337.1 nm is given by

IðN2Þ ¼ CN2hgN2

AðN2Þij nek

ðN2Þe nN2

�AðN2Þi þ kN2qnAr

��1: (5)

where the quantities C contain all optical and geometrical Param-eters; hg is the photon energy of respective transition; Ai denotesthe sum over all optical transition probabilities Aij for the excitedstate, and is equal to the reciprocal of the natural lifetime s0; Aij isequal to Ai ¼ 1/s0, since the only optically allowed transition fromeach of the excited states is observed, respectively [30,31]. kΟq, kArqand kN2q denote the quenching rate coefficient of the upper Ar(2p1),O(3p3P), and Ν2ðC3Pu; n ¼ 0Þ in argon, respectively. Table 1

Fig. 9. Excitation temperature as a function of peak applied voltage with the rod-tube/tube high voltage electrode and the inset graph is Boltzmann plot of excitationtemperature at peak applied voltage of 6.2 kV with the rod-tube high voltageelectrode.

summarizes the data taken from the literature [32e35]. ne, nO, nArand nN2

denote the electron density, atomic oxygen density, argondensity, and molecular nitrogen density. The argon density nAr isdetermined by the ideal gas equation n ¼ p/kTg, where the gaspressure p(Pa) takes the value of 1.01 � 105 Pa and the gastemperature Tg(K) is obtained from Fig. 6(a).

The electron excitation rate coefficient ke is determined byintegrating the cross sections over an assumed Maxwelliandistribution

ke ¼ZN0

sðεÞf ðεÞffiffiffiffiffiffiffi2eme

sεdε; (6)

where me is the electronic mass, s(ε) is the electron collision crosssection for excitation [36e38], and f(ε) is the electron energydistribution function. Fig. 11 shows the electron-impact excitationcross sections for Ar(2p1), O(3p3P), and Ν2ðC3Pu; n ¼ 0Þ. As regardsthe results in Fig. 11, the cross section shapes of nitrogen linel ¼ 337.1 nm and atomic oxygen line l ¼ 844.6 nm are similar, andthe excitation threshold of the both spectral line are much close,which are 10.98 eV and 11.1 eV, respectively. Therefore, themolecular nitrogen density can be determined by the intensity ratioof nitrogen line l ¼ 337.1 nm to atomic oxygen line l ¼ 844.6 nm.Taking the intensity ratio of Eq. (3) to Eq. (4) and Eq. (5) to Eq. (3),can obtain

nΟ ¼CArhgArIΟk

ðArÞe aðArÞij

CΟhgΟIArkðΟÞe aðΟÞ

ij

nAr; (7)

nN2¼

CΟhgΟIN2kðΟÞe aðΟÞ

ij

CN2hgN2

IΟkðN2Þe aðN2Þ

ij

nΟ (8)

with

aðΟÞij ¼ Aij

AiþkΟqnAr; aðArÞij ¼ Aij

AiþkArqnAr; aðN2Þ

ij ¼ Aij

AiþkN2qnAr: (9)

The atomic oxygen density and molecular nitrogen density aredetermined by Eqs. (7)e(9), as shown in Fig. 12. As the resultspresented in Fig. 12, the difference of atomic oxygen densityincreases with peak applied voltage between that with the rod-tube high voltage electrode and with the tube high voltage

Fig. 10. Optical emission spectrum 300e650 nm (a) and 650e900 nm (b) in software model at peak applied voltage of 5.4 kV with the rod-tube/tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e101 99

electrode, but the difference of molecular nitrogen density decreasewith peak applied voltage. It is may be due to the fact that theelectrons at high energy state are more with the rod-tube highvoltage electrode than that with the tube high voltage electrode, sothe generated oxygen atoms are more with the rod-tube highvoltage electrode, whereas the electrons at high energy state arerelatively less with the tube high voltage electrode, and hence thegenerated nitrogen molecules are relatively more with the tubehigh voltage electrode. As the peak applied voltage increases from4.6 to 7 kV, the atomic oxygen density and molecular nitrogendensity increase from 2.95 � 1016 to 3.33 � 1016 cm�3 and from5.57 � 1015 to 8.49 � 1015 cm�3 with the rod-tube high voltageelectrode, and from 2.54 � 1016 to 2.69 � 1016 cm�3 and from3.80 � 1015 to 8.12 � 1015 cm�3 with the tube high voltage elec-trode, respectively.

3.4. Average electronic density

The density of argon plasma column in quartz tube was deter-mined by the energy balance equation [39]. In this study, theplasma column is dominated by the species Ar, Ar*, Arþ, N2* and O*,so the energy balance equation is reduced as follows:

WeffznenArk1izε1iz þ nenArk1exε*1ex þ nenArkel;e�a

3me

MTe

þ nenN2k2exε2ex þ nenΟk3exε3ex:

(10)

where Weff (W cm�3) is the effective power density determinedby Peff/V, Peff (W) is the effective power and V (cm3) is the volumeof plasma column. The volumes of plasma column with the rod-tube/tube high voltage electrode are estimated by the expressionsV1 ¼ pR21L1 þ pR22L2 þ 1=3phðR21 þ R22 þ R1R2Þ � pR23L3 andV2 ¼ pR21L1 þ pR22L2 þ 1=3phðR21 þ R22 þ R1R2Þ, here R1 is the innerradius of quartz glass tube, L1 is the length of ground electrode, R2 is

Table 1Radiative lifetimes and room temperature quenching coefficients for Ar(2p1),O(3p3P), and Ν2ðC3Pu; n ¼ 0Þ in argon.

Radiative lifetime (ns)O(3p3P) Ar(2p1) N2 (C3Uu, v ¼ 0)34.7a 24b 42c

Reagent Quenching coefficients (10�10 cm3 s�1)O(3p3P) Ar(2p1) N2 (C3Uu, v ¼ 0)

Ar 0.14a 0.16b 0.008d

a [32].b [33].c [34].d [35].

the radius of gas outer hole, L2 is the distance from the gas outerhole to the quartz glass plate, h is the distance from the groundelectrode to the gas outer hole, R3 is radius of rod electrode, and L3is the distance from the pencil-shaped tapered end of rod to thebottom end of tube electrode. k1iz, k1ex and kel,e�a are the ionizationrate coefficient, excitation rate coefficient, and electron-ion colli-sion rate coefficient for argon, respectively. ε1iz and ε1ix

* are the firstionization energy (eV) from the ground state and the first excitationenergy (eV) for argon [40]. M is the argon mole mass equal to4� 10�2 kg. Te is the electron temperature, which takes the value of1 eV, according that the electron temperature of plasma column inatmospheric argon plasma jet was in the range of 1.12e1.14 eV [39].k2ex and k3ex are the excitation rate coefficient for nitrogen mole-cules and oxygen atoms, respectively. The rate coefficients arefunction of the electron temperature and in unit of cm3 s�1. ε2ex andε3ex are the excitation threshold of nitrogen line l ¼ 337.1 nm(11.1 eV) and the excitation threshold of atomic oxygen linel ¼ 844.6 nm (10.98 eV), respectively. nΟ and nN2

respectively arethe density of atomic oxygen andmolecular nitrogen obtained fromFig. 12 and the argon density nAr is determined by the ideal gasequation.

Fig. 13 shows the variations in average electronic density versuspeak applied voltage. The average electronic density with the rod-tube high voltage electrode is larger than that with the tube highvoltage electrode, it is in well agreement with the aforementionedresults that the conduction current with the rod-tube high voltageelectrode is larger than that with the tube high voltage electrode.

Fig. 11. Electron-impact excitation cross section of molecular nitrogen se3337.1, atomic

oxygen se844.6, and argon se

750.4.

Fig. 12. Atomic oxygen density (a) and molecular nitrogen density (b) as a function of peak applied voltage with the rod-tube/tube high voltage electrode.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e101100

Beside, the average electronic density increases lineally withapplied voltage in the both case.

In addition, the electron density of the argon plasma jet can beestimated by the expression ne ¼ J=ðemeEÞ, where J and e are thecurrent density and electron charge, respectively, me is the electronmobility, and E is the electric field sustained in the discharge region.The current density is obtained by the discharge cross-section todivide the current and the electric field is obtained by the distanceof discharge gap to divide the applied voltage. In this work, thedischarge is dominated by the tube electrode, so the discharge crosssection and the electric field are considered approximately equalbetween that with the rod-tube high voltage electrode andwith thetube high voltage electrode. Accordingly, as following equation isobtained: ne, with rod/ne, without rod ¼ I1, with rod/I2, without rod, where Irepresents the effective value of conduction current with the rod-tube/tube high voltage electrode and the electronic density ne isobtained from Fig. 13. Under the condition in Fig. 4(c) (I1, with rod is7.56 mA and I2, without rod is 6.55 mA), I1, with rod/I2, without rod ¼ 1.15and ne, with rod/ne, without rod ¼ 1.17. It is indicating that the value ofaverage electronic density estimated by the energy balance equa-tion is reasonable for this plasma jet.

Fig. 13. Average electronic density as a function of peak applied voltage with the rod-tube/tube high voltage electrode.

4. Conclusions

Cold argon plasma jets are generated with rod-tube/tube highvoltage electrode and a ring ground electrode in atmosphericpressure. It is found that the vibrational temperature, excitationtemperature, atomic oxygen density, and spectral intensity withthe rod-tube high voltage electrode are enhanced significantly thanthat with the tube high voltage electrode, and the effective power,vibrational temperature, excitation temperature, atomic oxygendensity, molecular nitrogen density, and average electronic densitywith the rod-tube/tube high voltage electrode increase linearlywith the applied voltage. The atomic oxygen density, molecularnitrogen density and average electronic density are in order ofmagnitude of 1016 cm�3, 1015 cm�3, and 1012 cm�3 with the rod-tube/tube high voltage electrode, respectively. Besides, the opticalemission spectrum shows that the active species such as OH and Oare presented in the plasma plume. Therefore, the plasma jet devicecan be used for applications such as etching, sterilization, andsurface modification, etc.

Acknowledgments

The authors thank the National Natural Science Foundation, P. R.China (Project No. 51177007) and the Ministry of Science andTechnology, P. R. China (Project No. 2009AA064101-4).

References

[1] A. Anders, Fundamentals of pulsed plasmas for materials processing, Surf.Coat. Technol. 183 (2004) 301e311.

[2] J. Benedikt, K. Focke, A. Yanguas-Gil, A. von Keudella, Atmospheric pressuremicroplasma jet as a depositing tool, Appl. Phys. Lett. 89 (2006) 251504.

[3] J. Benedikt, V. Raballand, A. Yanguas-Gil, K. Focke, A. von Keudell, Thin filmdeposition by means of atmospheric pressure microplasma jet, Plasma Phys.Control. Fusion 49 (2007) B419eB427.

[4] M. Leduc, S. Coulombe, R.L. Leask, Atmospheric pressure plasma jet depositionof patterned polymer films for cell culture applications, IEEE Trans. Plasma Sci.37 (2009) 927e933.

[5] J.L. Walsh, M.G. Kong, Contrasting characteristics of linear-field and cross-fieldatmospheric plasma jets, Appl. Phys. Lett. 93 (2008) 111501.

[6] X.P. Lu, T. Ye, Y.G. Cao, Z.Y. Sun, Q. Xiong, Z.Y. Tang, Z.L. Xiong, J. Hu, Z.H. Jiang,Y. Pan, The roles of the various plasma agents in the inactivation of bacteria,J. Appl. Phys. 104 (2008) 053309.

[7] X. Lu, Z. Xiong, F. Zhao, Y. Xian, Q. Xiong, W. Gong, C. Zou, Z. Jiang, Y. Pan,A simple atmospheric pressure room-temperature air plasma needle devicefor biomedical applications, Appl. Phys. Lett. 95 (2009) 181501.

Y. Hong et al. / Journal of Electrostatics 71 (2013) 93e101 101

[8] J. Huang, W. Chen, H. Li, X.Q. Wang, G.H. Lv, M.L. Khosa, M. Guo, K.C. Feng,P.Y. Wang, S.Z. Yang, Deactivation of A549 cancer cells in vitro by a dielectricbarrier discharge plasma needle, J. Appl. Phys. 109 (2011) 053305.

[9] G. Fridman, A.D. Brooks, M. Balasubramanian, A. Fridman, A. Gutsol,V.N. Vasilets, H. Ayan, G. Friedman, Comparison of direct and indirect effectsof non-thermal atmospheric-pressure plasma on bacteria, Plasma Process.Polym. 4 (2007) 370e375.

[10] X.T. Deng, J.J. Shi, G. Shama, M.G. Kong, Effects of microbial loading andsporulation temperature on atmospheric plasma inactivation of Bacillussubtilis spores, Appl. Phys. Lett. 87 (2005) 153901.

[11] R. Dorai, M.J. Kushner, A model for plasma modification of polypropyleneusing atmospheric pressure discharges, J. Phys. D: Appl. Phys. 36 (2003) 666.

[12] E. Karakas, M. Laroussi, Experimental studies on the plasma bullet propaga-tion and its inhibition, J. Appl. Phys. 108 (2010) 063305.

[13] M.H. Chiang, K.C. Liao, I.M. Lin, C.C. Lu, H.Y. Huang, C.L. Kuo, J.S. Wu, Modifi-cation of hydrophilic property of polypropylene films by a parallel-platenitrogen-based dielectric barrier discharge jet, IEEE Trans. Plasma Sci. 38(2010) 1489e1498.

[14] X.P. Lu, Y.G. Cao, P. Yang, Q. Xiong, Z.L. Xiong, Y.B. Xian, Y. Pan, An RC plasmadevice for sterilization of root canal of teeth, IEEE Trans. Plasma Sci. 37 (2009)668e673.

[15] S.X. Deng, C. Cheng, G.H. Ni, Y.D. Meng, H. Chen, Bacterial inactivation byatmospheric pressure dielectric barrier discharge plasma jet, Jpn. J. Appl. Phys.47 (2008) 7009e7012.

[16] Q. Li, J.-T. Li, W.C. Zhu, X.M. Zhu, Y.K. Pu, Effects of gas flow rate on the lengthof atmospheric pressure nonequilibrium plasma jets, Appl. Phys. Lett. 95(2009) 141502.

[17] U.N. Pal, A.K. Sharma, J.S. Soni, Kr Sonu, H. Khatun, M. Kumar, B.L. Meena,M.S. Tyagi, B.-J. Lee, M. Iberler, J. Jacoby, K. Frank, Electrical modellingapproach for discharge analysis of a coaxial DBD tube filled with argon, J. Phys.D: Appl. Phys. 42 (2009) 045213.

[18] K.P. Singh, S. Roy, Impedance matching for an asymmetric dielectric barrierdischarge plasma actuator, Appl. Phys. Lett. 91 (2007) 081504.

[19] E. Panousis, F. Clement, J.-F. Loiseau, N. Spyrou, B. Held, M. Thomachot,L. Marlin, An electrical comparative study of two atmospheric pressuredielectric barrier discharge reactors, Plasma Sources Sci. Technol. 15 (2006)828e839.

[20] X.P. Lu, Z.H. Jiang, Q. Xiong, Z.Y. Tang, Y. Pan, A single electrode room-temperature plasma jet device for biomedical applications, Appl. Phys. Lett.92 (2008) 151504.

[21] Q. Xiong, X.P. Lu, K. Ostrikov, Y. Xian, C. Zou, Z. Xiong, Y. Pan, Pulsed dc- andsine-wave-excited cold atmospheric plasma plumes: a comparative analysis,Phys. Plasmas 17 (2010) 043506.

[22] N.U. Rehman, F.U. Khan, M.A. Naveed, M. Zakaullah, Determination of exci-tation temperature and vibrational temperature of the N2 (C3Pu, n’) state inNeeN2 RF discharges, Plasma Sources Sci. Technol. 17 (2008) 025005.

[23] S.J. Kim, T.H. Chung, S.H. Bae, Characteristic study of atmospheric pressuremicroplasma jets with various operating conditions, Thin Solid Films 517(2009) 4251e4254.

[24] G.D. Wei, C.S. Ren, M.Y. Qian, Q.Y. Nie, Optical and electrical diagnostics of coldAr atmospheric pressure plasma jet generated with a simple DBD configura-tion, IEEE Trans. Plasma Sci. 39 (2011) 1842e1848.

[25] N. Masoud, K. Martus, M. Figus, K. Becker, Rotational and vibrationaltemperature measurements in a high-pressure cylindrical dielectric barrierdischarge (C-DBD), Contrib. Plasma Phys. 45 (2005) 32e39.

[26] M.Y. Qian, C.S. Ren, D.Z. Wang, J.L. Zhang, G.D. Wei, Stark broadeningmeasurement of the electron density in an atmospheric pressure argonplasma jet with double-power electrodes, J. Appl. Phys. 107 (2010) 063303.

[27] W.R. Harshbarger, R.A. Porter, T.A. Miller, P. Norton, A study of the opticalemission from an RF plasma during semiconductor etching, Appl. Spectrosc.31 (1977) 201e207.

[28] H.M. Katsch, A. Tewes, E. Quandt, A. Goehlich, T. Kawetzki, H.F. Döbele,Detection of atomic oxygen: improvement of actinometry and comparisonwith laser spectroscopy, J. Appl. Phys. 88 (2000) 6232e6238.

[29] A. Qayyum, Shaista Zeb, M.A. Naveed, S.A. Ghauri, M. Zakaullaha, A. Waheed,Diagnostics of nitrogen plasma by trace rare-gas-optical emission spectros-copy, J. Appl. Phys. 98 (2005) 103303.

[30] J. Waskoenig, K. Niemi, N. Knake, L.M. Graham, S. Reuter, V. Schulz-von derGathen, T. Gans, Diagnostic-based modeling on a micro-scale atmosphericpressure plasma jet, Pure Appl. Chem. 82 (2010) 1209e1222.

[31] K. Niemi, S. Reuter, L.M. Graham, J. Waskoenig, N. Knake, V. Schulz-von derGathen, T. Gans, Diagnostic based modelling of radio-frequency drivenatmospheric pressure plasmas, J. Phys. D: Appl. Phys. 43 (2010) 124006.

[32] K. Niemi, V. Schulz-von der Gathen, H.F. Döbele, Absolute atomic oxygendensity measurements by two-photon absorption laser-induced fluorescencespectroscopy in an RF-excited atmospheric pressure plasma jet, PlasmaSources Sci. Technol. 14 (2005) 375e386.

[33] N. Sadeghi, D.W. Setser, A. Francis, U. Czarnetzki, H.F. Döbele, Quenching rateconstants for reactions of Ar (4p [1/2], 4p [1/2], 4p [3/2], and 4p [5/2]) atomswith 22 reagent gases, J. Chem. Phys. 115 (2001) 3144e3154.

[34] S.V. Pancheshnyi, S.M. Starikovskaia, A.Yu. Starikovskii, Collisional deactiva-tion of N2 (C3Pu, n ¼ 0, 1, 2, 3) states by N2, O2, H2 and H2O molecules, Chem.Phys. 262 (2000) 349e357.

[35] J.F. Loiseau, P. Plgnolet, B. Held, Numerical simulation of AreN2 excitationtransfer in flowing afterglow, J. Phys. D: Appl. Phys. 25 (1992) 745.

[36] M. Hayashi, 2003, online available: http://jila.colorado.edu/wavp/collision_data/electronneutral/hayashi. txt.

[37] R.R. Laher, F.R. Gilmore, Updated excitation and ionization cross sectionsfor electron impact on atomic oxygen, J. Phys. Chem. Ref. Data 19 (1990)277e305.

[38] Y. Itikawa, Cross sections for electron collisions with nitrogen molecules,J. Phys. Chem. Ref. Data 35 (2006) 31e53.

[39] S.Z. Li, W.T. Huang, J.L. Zhang, D.Z. Wang, Discharge characteristics of anatmospheric-pressure argon plasma column generated with a single-electrodeconfiguration, Phys. Plasmas 16 (2009) 073503.

[40] S.Z. Li, J.P. Lim, J.G. Kang, H.S. Uhm, Comparison of atmospheric-pressurehelium and argon plasmas generated by capacitively coupled radio-frequency discharge, Phys. Plasmas 13 (2006) 093503.