High-speed dispersed photographing of an open-air argon plasma plume by a grating–ICCD camera system

High-speed dispersed photographing of an open-air argon plasma plume by a grating–ICCD

camera system

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

2010 J. Phys. D: Appl. Phys. 43 415201

(http://iopscience.iop.org/0022-3727/43/41/415201)

Download details:

IP Address: 211.69.201.252

The article was downloaded on 17/12/2010 at 02:54

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

IOP PUBLISHING JOURNAL OF PHYSICS D: APPLIED PHYSICS

J. Phys. D: Appl. Phys. 43 (2010) 415201 (10pp) doi:10.1088/0022-3727/43/41/415201

High-speed dispersed photographing ofan open-air argon plasma plume by agrating–ICCD camera systemQ Xiong1,2, A Y Nikiforov2,3, X P Lu1 and C Leys2

1 College of Electrical and Electronic Engineering, HuaZhong University of Science and Technology,WuHan, Hubei 430074, People’s Republic of China2 Department of Applied Physics-Research Unit Plasma Technology, Ghent University, Rozier 44,Ghent B-9000, Belgium3 Institute of Solution Chemistry of the Russian Academy of Science, Academicheskaya St., 1,Ivanovo, 153045, Russia

E-mail: [email protected]

Received 16 June 2010, in final form 31 August 2010Published 28 September 2010Online at stacks.iop.org/JPhysD/43/415201

AbstractIn this paper, an open-air argon plasma plume is generated at atmospheric pressure by atwo-electrode jet device with sub-microsecond voltage pulses at a repetitive frequency of1 kHz. Optical emission spectroscopy measurements showed that spectral irradiance from OHand N2 bands, and Ar lines, characterized the spectrum of the open-air argon plasma plume.The rotational temperature estimation of UV OH band spectra indicated the gas temperature ofthe plasma plume to be as low as room temperature. A novel diagnostic method, based on twodispersion gratings and an ICCD camera, was designed for investigating the time- andspace-resolved propagation behaviour of the excited radicals in the plasma plume. Based onthe dispersion feature of gratings, a series of dispersed plasma optical emission volumes,which were formed by irradiance from different excited radiation emitters (excited species) inthe plasma plume, were captured in the form of high-speed images by the ICCD camera. Fromthe sequence of dispersed emission images, it is possible to observe the time- andspace-resolved behaviour of different excited species in the plasma, and meanwhile, tounderstand the propagation dynamics of the open-air argon plasma plume. It is found that theOH bands’ emission volume exhibited a propagation behaviour distinct from that of N2 and Aremission volumes. The OH emissions decayed immediately as soon as the plasma travelledout from the nozzle, but were able to last for a longer duration time inside the nozzle than bothN2 and Ar emissions. The N2 bands’ emission volumes propagated to a far distance andformed the whole length of the argon plasma plume in the surrounding air. The Ar emissionsdecayed rapidly for the plasma inside and outside the nozzle due to the adverse effect ofimpurities, in particular the large concentration of diffused air in the open space. Thesedistinct types of dynamic behaviour of the dispersed plasma emission volumes are attributed tothe different generation and quenching mechanisms of their corresponding excited species andthey shed light on the clear propagation dynamics of the argon plasma plume in open air.

(Some figures in this article are in colour only in the electronic version)

1. Introduction

Non-equilibrium atmospheric pressure plasma jets haverecently received increasing attention for their unique features

such as chamberless operation and enhanced chemistry withoutelevated gas temperatures. These attractive features makethem suitable for several emerging novel applications, such asbiomedical decontamination, material synthesis and surface

0022-3727/10/415201+10$30.00 1 © 2010 IOP Publishing Ltd Printed in the UK & the USA

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

fibre

Figure 1. (a) Schematic diagram of the experimental setup and (b) a photograph of argon plasma plume excited with +15 kV, 800 ns, 1 kHzpulses, and 140 sccm of argon gas. The position of the device nozzle exit and upper surface of the grounded electrode, are marked B and C,respectively.

modification [1–4]. In order to improve their applicationefficiency, many different electrically driven non-equilibriumatmospheric plasma jet devices have been designed and studiedrecently [5–10]. Most of these open-air plasma jet devices areable to generate a plasma plume of length of a few centimetreswith gas temperatures close to room temperature. This room-temperature feature is essential in plasma applications relatedto heat-sensitive materials, particularly in the fields of plasmahealthcare and biomedicine, where microorganisms, cells,living tissues, membranes, etc are very sensitive to the processtemperature [11–14]. For practical application purposes, thestability of the generated plasma jets is crucial. Recently,pulsed excitations have been employed and proved to becapable of improving the control of plasma instabilities andavoiding the glow-to-arc transition during plasma operation[15–17]. And not only the expensive helium gas, which iseffective in improving the plasma stability and used frequentlyas a working gas, but also other available low-cost gases, suchas argon, even air, are applicable in generating plasma plumeswith low gas temperatures by sub-microsecond high-voltagepulses [18–20]. This further improves the practicability andpotential of non-equilibrium atmospheric pressure plasma jetsin these novel applications.

As we know, reactive species play a key role in the plasmatreatment process [21–24]. A fundamental investigation on thedynamics of these reactive species is necessary and importantto understand the formation mechanism of open-air plasmaplumes and improve their treatment efficiency for applicationpurposes. Two main diagnostic methods are often appliedfor the temporal and spatial measurement of plasma plumes,i.e. optical emission spectroscopy (OES) and imaging byintensified charge-coupled device (ICCD) cameras [25, 26].In this paper, a novel experimental method was designedusing two different gratings (1200 and 600 g mm−1) and anICCD camera to study the time- and space-resolved behaviourof an open-air argon plasma driven by a homemade pulsed

excitation source at kilohertz repetition rate. Based on thedispersion feature of gratings, a series of dispersed plasmaoptical emission volumes formed by irradiance from differentexcited irradiative emitters in the plasma plume were capturedby the ICCD camera. From these high-speed dispersed images,the propagation behaviour of different excited radicals in theplasma plume was clearly observed. And finally, a detaileddynamic behaviour was obtained for the argon plasma plumein open air.

The rest of this paper is organized as follows. Theexperimental setup is described in section 2. Section 3 presentsthe experimental results related to the time- and space-resolveddynamics of several excited species in the open-air argonplasma plume through the dispersed photographing grating–ICCD system. A detailed discussion about the experimentalresults is given in section 4. Finally, section 5 presents a briefsummary of this work.

2. Experimental setup

The experimental arrangement used in this study isschematically shown in figure 1(a). The plasma jet devicestudied in this paper is almost the same as that we have reportedpreviously, except that a grounded electrode surrounding thenozzle [27]. It is composed of a syringe-like quartz tube, a10 cm long quartz tube with one end closed, a high-voltage(HV) electrode and a grounded electrode. The diameters ofthe syringe-like tube and the quartz tube are 8 mm and 4 mm,respectively. The HV electrode (made of a copper wire) isinserted into the quartz tube, and both together are arrangedinside and aligned along the axis of the syringe tube. Thegrounded electrode is made of a copper ring with thickness4 mm, outer and inner diameters 7 mm and 4 mm, respectively.The distance between the closed end of the quartz tube and thegrounded electrode was fixed at about 5 mm. The whole part

2

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

mentioned above is fixed by a mount with three channels forintroducing working gas flowing through the syringe tube.

The HV electrode is powered by a homemade high-voltagepulse generator. The pulse generator is excited by a dcpower supply (±40 kV, 30 mA), and triggered by a pulse-signal generator. The pulse system is able to excite theHV electrode with fast rise-and-fall high-voltage pulses, withvoltage amplitude up to 30 kV, pulse duration variable from50 ns to dc and repetition rate up to 2 kHz. Argon gas witha flow rate of 140 sccm was used as the working gas in thisstudy. When argon gas was introduced into the syringe tubeand high-voltage pulses were applied to the HV electrode, anon-uniform DBD discharge was ignited in the active spacebetween the HV and grounded electrodes, and meanwhile, ahomogeneous plasma plume was produced along the nozzleand formed in ambient air. Figure 1(b) shows a photograph ofthe atmospheric pressure argon plasma plume, with a lengthup to 2.8 cm. The exit end of the nozzle and the upper sideof the grounded electrode are marked B and C, respectively.It should be pointed out that the argon plasma plume canalso be generated without the grounded electrode. However,for the case without the grounded electrode, the emissionintensity and length of the generated argon plasma plumedecreased significantly. For all the experimental results in thiswork, the argon plasma plume was generated with a groundedelectrode surrounding the nozzle (as shown in figure 1). Theoperational parameters are fixed: applied voltage Va = 15 kV,pulse duration tp = 800 ns, pulse frequency fp = 1.0 kHzand argon gas flow rate g = 140 sccm. The argon gas has animpurity of H2O, about 0.5 ppm.

The applied voltages were measured by a TektronixP6015A 1000 : 1 high-voltage probe and the currents by an IPGcurrent transformer (CM-001-L). OES was used for obtainingthe emission spectrum of the argon plasma plume. A UV–visible lens and a fibre-optics system was used to focus theplasma emission into the entrance slit of the monochromator.For the large emission spectrum range (250–875 nm), a S2000Ocean Optics spectrometer with a resolution of 0.7 nm FWHMwas used. In order to obtain the average gas phase temperatureby estimating the rotational temperature of OH or N2, thehigh resolution spectra of OH bands (308 nm) and N2 bands(337 nm) were obtained by an Avantes 3648 spectrometer witha resolution of 0.05 nm FWHM. Two gratings, 1200 g mm−1

(blazed at 500 nm) and 600 g mm−1 (blazed at 1000 nm), wereapplied for obtaining the dispersed emission volumes of theargon plasma plume. The two gratings were arranged at thesame position individually with a distance of 10 cm betweenthe vertical central axes of the gratings and the plasma jetdevice. The dispersed time- and space-resolved emissionpattern images of the argon plasma plume were obtained by aHamamatsu ICCD camera (C8484). The ICCD camera gatingtime and the discharge initiation were synchronized using adelay generation (Stanford Instrument DG535). The exposuretime was fixed at 10 ns for all imaging measurements.

3. Experimental results

The typical voltage–current characteristics of the argon plasmaplume are shown in figures 2(a) and (b). Figure 2(a) shows

Figure 2. (a) Typical applied voltage Va, total current Ion (plasmaon) and displacement current Ioff (plasma off) of the argon plasmaplume. (b) The applied voltage Va and actual discharge current Idis

(difference of Ion and Ioff) of the argon plasma plume.

the applied voltage Va, total current Ion (plasma on) anddisplacement current Ioff (without argon flow, no plasma).The actual discharge current Idis is obtained by subtracting thedisplacement current Ioff from the total current Ion, as shownin figure 2(b). The peak value of the discharge current Idis isabout 0.6 A. As can be seen, two distinct discharge currentpulses (positive and negative discharges) are observed pervoltage pulse. The negative discharge is ignited by the chargesaccumulated on the surface of the quartz tube during thepositive discharge. The peak value of the negative dischargeis slightly smaller than that of the positive discharge, with amaximum of about 0.5 A.

To identify the various reactive species produced in theplasma plume, OES was applied in this work. Figure 3shows the emission spectrum of the argon plasma plume in thesurrounding air. The spectrum was detected from the plasmavolume about 0.5 cm in front of the nozzle exit. As can beseen, the plasma plume is characterized by optical emissionsfrom OH and N2 bands, and Ar lines. The observation of OHbands are due to the impurity of H2O in the argon gas, andN2 bands are attributed to the air diffusion into the argon flowstream in the open space. For the wavelength range from 250 to

3

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

Figure 3. Emission spectrum of the argon plasma plume in open airin the range 250–875 nm.

500 nm, the spectrum is dominant by the emission bands fromtransitions OH(A 2�+ → X 2�) and N2(C

3�u → B 3�g),and for the large wavelength range (650–875 nm) by the argonlines from Ar(4p → 4s).

In order to observe the temporally and spatially resolvedpropagation behaviour of the argon plasma plume in ambientair, two gratings (1200 and 600 g mm−1) were specially appliedto disperse the optical emissions of the plasma plume. A UV–visible optical lens was used to focus the dispersed emissionsfrom the gratings onto an ICCD camera. After that, a seriesof high-speed dispersed image sequences corresponding to thedynamics of different irradiative excited species in the plasmaplume were obtained. Figures 4 and 5 show the time- andspace-resolved image sequences of the OH and N2 bands,and Ar lines emissions for the (a) positive and (b) negativedischarges of the argon plasma, respectively. As can be seen,dispersed emission volumes, formed by radiative transitionsof corresponding excited species (OH(A), N2(C) and Ar(4p)),were observed while the plasma propagated from the nozzleinto the open space. The propagation behaviour of the OH andN2 emission volumes (figure 4) was dispersed by the grating1200 g mm−1, and Ar lines’ emission volumes (figure 5) bythe grating 600 g mm−1. The position of the nozzle exit B andthe upper surface of the grounded electrode C are also markedtherein, respectively. The delay of the time axis correspondedto the onset of the applied voltage Va. The reason for using the600 g mm−1 grating is the limited dispersion spectrum rangeof the 1200 g mm−1 grating (blazed at 500 nm), which is noteffective for dispersing the large wavelength range of Ar lines(>650 nm) in our case. However, the dispersion resolutiondecreased in the case of the 600 g mm−1 grating and the secondorder of the OH and N2 bands were observed and overlappedwith the Ar lines due to their same wavelength range (600–800 nm) in our case. In order to remove the second order ofthe OH and N2 bands, a high-pass filter for wavelength largerthan 500 nm was arranged between the argon plasma plumeand the 600 g mm−1 grating. After that, the time- and space-resolved propagation behaviour of optical emission volumesformed by Ar lines irradiance was obtained, as shown figure 5.

The resolutions of the two dispersion systems based on the1200 g mm−1 and 600 g mm−1 gratings are about 0.8 nm/pixeland 1.3 nm/pixel, respectively. As a result of the limitedresolution, overlapped effects were observed for the OH bandsat 308 nm and N2 bands at 315 nm, and several Ar lines, asshown in figures 4 and 5.

As can be seen, distinct types of time- and space-resolvedbehaviour were observed for the dispersed optical emissionvolumes while the plasma propagated inside and outside thenozzle. It should be noted that the emission volume, whichcontinued irradiating inside the nozzle, was formed by the OHbands’ emissions (308 nm) as shown in figure 4. And the fiveemission volumes propagating in the open air formed by the N2

bands’ emissions at different wavelengths. Emission durationtime τ (ns) and propagation distances d (cm) of the investigatedemission volumes were estimated for the plasma inside andoutside the nozzle. It should be mentioned that the emissionduration time τ was estimated according to the luminescenceperiod of the dispersed emission volumes propagating insideand outside the nozzle, as shown in figures 4 and 5. Andthe propagation distances d were estimated according to thedistance from the gas exit of the nozzle to the tip of thelast propagating luminescent emission volumes in the openspace. These estimated values are listed in tables 1 and 2for the positive and negative discharges of the argon plasma,respectively. As can be seen, for both the positive and negativeignited discharges, the duration time of OH bands’ emissionτOH was longer than τN2 and τAr inside the nozzle. However, theOH emission intensity decreased quickly as soon as the plasmatravelled out from the nozzle. Due to this, we did not estimateτOH and dOH for the plasma propagating in the surrounding air.Distinct from the OH bands, τN2 of N2 bands’ emissions wasshort (50–60 ns) when the plasma travelled inside the nozzle,but long (240–260 ns) for the plasma plume in the open space.The N2 emission volumes were able to propagate a largerdistance than the Ar emission volumes in open air. dN2 of theN2 emission volumes was estimated to be about 2.8 cm for thepositive ignited plasma, which forms the whole length of theargon plasma plume.

As shown in tables 1 and 2, the duration time of theinvestigated emission volumes showed almost no changes forboth the positive and negative ignited plasmas. However, dN2

and dAr are smaller for the negative ignited plasma plume.The observed shorter propagation distances of the investigatedemission volumes are attributed to the weak excitationprocesses and the low efficiency of plasma generation duringthe negative discharge, which are indicated by the smallerdischarge current and lower emission intensity of the negativedischarge. Distinct types of spatial behaviour between thepositive and negative ignited plasma plumes are also observedin other gas discharges, such as helium [26].

In order to estimate the average gas phase temperatureof the plasma discharges, rotational temperatures of severalmolecular species and molecular ions, such as N2, OH and N+

2 ,are often applied through spectrum fitting between syntheticand experimental spectral profiles. Due to the absence of theN+

2 emission spectrum in our case, rotational temperatures ofthe molecules N2 and OH were estimated and compared in

4

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

Figure 4. The time- and space-resolved image sequences of the OH and N2 bands’ emissions from the dispersed argon plasma plume inopen air by the 1200 g mm−1 grating, (a) positive and (b) negative discharges of the argon plasma. The images were acquired with a 10 nsgated ICCD camera at a time interval of 20 ns. The delay of time axis corresponds to the onset of the applied voltage Va. The position of thedevice nozzle exit B and upper surface of the grounded electrode C are also shown, respectively.

Figure 5. The time- and space-resolved image sequences of the Ar emissions from the dispersed argon plasma plume in open air by the600 g mm−1 grating, (a) positive and (b) negative discharges of the argon plasma. The images were acquired with a 10 ns gated ICCDcamera at a time interval of 10 ns. The delay of time axis corresponds to the onset of the applied voltage Va. The position of the devicenozzle exit B and upper surface of the grounded electrode C are also shown, respectively.

Table 1. Estimated emission duration time τ (unit: ns) of the investigated radiative transitions from the positive ignited plasma inside andoutside the nozzle. The propagation distances d (unit: cm) of their corresponding optical emission volumes in open air are also listed. Thespontaneous radiative lifetime (unit: ns) is shown as a comparison.

τ (ns)

Plasma inside Plasma outside RadiativeTransitions nozzle nozzle lifetime (ns) d ( cm) Ref

Ar(4p) → Ar(4s) + hν 120 90 40 1.6 [38]OH(A) → OH(X) + hν 420 — 800 — [45]N2(C) → N2(B) + hν 60 260 37 2.8 [47]

order to determine the gas temperature of the argon plasmaplume precisely. Figure 6 shows the best-fitting syntheticspectrum to the experimental spectrum of the C 3�u → B 3�g

(�υ = 0) band transition of nitrogen in the range from334 to 337.5 nm (figure 6(a)) and OH(A 2�+ → X 2�,�υ = 0) band transition from 306 to 312 nm (figure 6(b)).For comparison, both the experimental and calculated spectrawere normalized. As can be seen, the estimated rotational

temperatures of N2 and OH are distinct, with Trot = 600 Kof N2 being about twice as that of OH (330 K). This distinctdifference in the rotational temperatures are attributed to thegeneration mechanisms of the excited species N2(C

3�u)

and OH(A 2�). As is well known, the energy carried bythe argon metastables Ar(4s) (11.5–11.7 eV), which wereproduced effectively in our case indicated by the spectrumshown in figure 3, is slightly higher than that of the threshold

5

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

Table 2. Estimated duration time τ (unit: ns) and propagationdistances d (unit: cm) of the investigated radiative transitions fromthe negative ignited plasma.

τ (ns)

Plasma Plasmainside outside

Transitions nozzle nozzle d (cm)

Ar(4p) → Ar(4s) + hν 130 90 0.2OH(A) → OH(X) + hν 440 — —N2(C) → N2(B) + hν 50 240 0.9

Figure 6. Experimental and simulated rotational spectrum of (a) theC 3�u → B 3�g (�υ = 0) band transition of nitrogen in the rangefrom 334 to 337.5 nm and (b) OH(A 2�+ → X 2�, �υ = 0) bandtransition from 306 to 312 nm.

energy level of N2(C3�u) (∼11.1 eV) from ground states.

The argon metastables Ar(4s) can easily transfer energy tothe ground state nitrogen molecules to form the excited stateN2(C

3�u) at high rotational levels, resulting in a relativelyhigh rotational distribution of nitrogen. This observation hasalso been reported previously by other works [28–30]. Weconfirmed this by using a thermocouple inside a quartz tubeto measure the gas temperature of the plasma plume. A stabletemperature value 25 ◦C (303 K) was obtained after 8 min ofcontact between the quartz tube surface and the plasma plume.

Figure 7. Boltzmann plot of the OH(A 2�+ → X 2�, �υ = 0)band transition in the argon plasma plume.

The reason for using a thermocouple inside a quartz tube wasthat the generated plasma plume was not affected significantlycompared with the case without the tube. Although errorswould be induced by using the thermocouple inside a tube,we could still deduce that the gas temperature of the plasmaplume is not as high as 600 K by Trot of N2. Therefore, Trot

of N2 is not applicable for estimating the gas temperature ofthe argon plasma plume under our experimental conditions.It seems that the method by Trot of OH is more applicablebecause it is close to the temperature value measured by thethermocouple. And as reported previously, the UV OH bandspectrum is proved to be a good method for gas temperaturemeasurements of non-equilibrium plasmas [31]. In order tocheck this further, the Boltzmann plot method was applied todetermine the rotational temperature of OH(A 2�+ → X 2�,�υ = 0). The resolved lines of the Q1, Q2, P1, R1 and R2

branches of OH(A 2�+ → X 2�, �υ = 0) were used in thiswork. The basic parameters, i.e. energy level E, wavelength λ

and probabilities AJJ of transitions, are taken from [32, 33].Figure 7 shows the Boltzmann plot of the OH(A 2�+ →

X 2�, �υ = 0) transition. As can be seen, the OH(A 2�+ →X 2�, �υ = 0) transition exhibited a non-Boltzmannbehaviour and a two-temperature distribution was fitted to therotational population distribution. An overpopulation of highrotational levels was observed from rotational level J = 6 ofOH(A 2�+ → X 2�, �υ = 0). However, the temperatureof the Boltzmann plot for small rotational levels (J < 6),T1 = 300 ± 30 K, is taken as the Trot of OH, which isvery close to those obtained using the spectra fitting method(330 ± 30 K) and the thermocouple (303 K) mentioned above.Therefore, the actual gas temperature of the argon plasmaplume is determined to be about 300 K. The two-temperaturedistribution of OH(A 2�+ → X 2�, �υ = 0) transition isdue to the different production mechanisms of the excitedstate OH(A 2�+). The generation processes through electrondissociative excitation of water, and energy transfer fromargon metastables, all can induce different rotational statesof OH(A 2�+) and lead to multiple temperature distributions[34, 35]. Therefore, for the atmospheric pressure discharges

6

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

Table 3. Generation processes of investigated excited species in our case. The symbol σ(ε) is the reaction’s cross section, where ε is theelectron energy.

Excited species Generation reactions Rate Ref

Ar(4p) Ar + e + �E(∼13 eV) → Ar(4p) + e σ(ε) [37]Ar(4s) + e + �E(∼2 eV) → Ar(4p) + e σ(ε) [37]

Ar(4s) Ar + e + �E(∼11 eV) → Ar(4s) + e σ(ε) [37]Ar(4p) → Ar(4s) + hν 2.5 × 107 s−1 [38]

OH(A 2�+) H2O + e + �E(∼9.1 eV) → OH(A) + H + e σ(ε) [39]H2O + Ar(4p)/Ar(4s) → OH(A) + H + Ar ∼10−10 cm3 s−1 [40]

N2(C3�u) N2(X) + e + �E(∼11.1 eV) → N2(C) + e σ(ε) [41]

N2(A) + e + �E(∼4.8 eV) → N2(C) + e σ(ε) [41]N2(A) + N2(A) → N2(C) + N2(X) 1.5 × 10−10 cm3 s−1 [42]N2(X) + Ar(4p)/Ar(4s) → N2(C) + Ar 10−11–10−10 cm3 s−1 [40, 43]

Table 4. Quenching processes of investigated excited species under different gas background, Ar for the plasma inside the nozzle, and Armixed with N2 and O2 for the plasma in open air (plasma plume). Symbol Ar∗2 refers to an excited state of diatomic argon molecule.

Plasma inside nozzle Plasma outside nozzle

Excited species Quenching reactions Rate Quenching reactions Rate Ref

Ar(4p) Ar(4p) + Ar → products ∼10−11 cm3 s−1 Ar(4p) + N2/O2 → products 10−11–10−10 cm3 s−1 [38, 43]Ar(4p) → Ar(4s) + hν 2.5 × 107 s−1 Ar(4p) + Ar → products ∼10−11 cm3 s−1

Ar(4p) → Ar(4s) + hν 2.5 × 107 s−1

Ar(4s) Ar(4s) + Ar(4s) → Ar + Ar+ + e 6.4 × 10−10 cm3 s−1 Ar(4s) + N2/O2 → products ∼10−11 cm3 s−1 [40, 44]Ar(4s) + 2Ar → Ar∗

2 + Ar 1.4 × 10−32 cm6 s−1

OH(A 2�+) OH(A) → OH(X) + hν 1.25 × 106 s−1 OH(A) + N2/O2 → products 10−11–10−10 cm3 s−1 [45, 46]OH(A) + Ar → products ∼10−13 cm3 s−1

N2(C 3�u) N2(C) → N2(B) + hν 2.74 × 107 s−1 N2(C) + N2 → products ∼10−11 cm3 s−1 [47, 48]N2(C) + O2 → products ∼10−10 cm3 s−1

with Ar and impurities such as H2O and N2 in our case,it is not applicable to estimate the gas temperature throughdetermining the Trot of nitrogen. And for a precise estimation,Trot of OH is also not suggested to be applied due to itsoverpopulation of high rotational states. In this case, furtherconfirmation processes like the Boltzmann plot method shouldbe carried out, or Doppler temperatures of Ar atoms, rotationaltemperatures of other heavy species like N+

2 and Rayleighscattering of heavy particle can be used for obtaining amore precise gas temperature [28, 29, 36]. For vibrationaltemperatures of the plasma plume, the spectral profile ofthe nitrogen second positive system N2(C

3�u → B 3�g

(�υ = −2)) was applied through the spectrum fitting method,and was estimated to be about 1750 K.

4. Discussion

The distinct types of temporally and spatially resolvedpropagation behaviour of the dispersed emission volumes ofOH and N2 bands, and Ar lines, are attributed to the differentgeneration and quenching mechanisms of their correspondingexcited states, i.e. OH(A 2�+), N2(C

3�u) and Ar(4p) in theargon plasma plume. In order to obtain a good understandingabout the interesting experimental observations mentionedabove, a detailed discussion is presented in the followingsection focusing on the analysis of the dominant generationand quenching processes of the three excited species in theargon plasma.

The dominant generation and quenching processes of theinvestigated excited species in our case are listed in tables 3

and 4. Because of the absence of the emission spectrumcorresponding to the transitions from upper levels to Ar(4p),the effect of cascade transitions on the generation of Ar(4p) isnegligible in our case. The electron step-excitation and poolingreaction related to metastable N2(A

3�+u ) were considered and

may play a role in producing the excited state N2(C3�u)

while the plasma propagated in the surrounding air. Due tothe different gas background, i.e. argon gas for the plasmainside the nozzle, argon mixed with air molecules (nitrogen andoxygen) for the plasma in open air, the quenching mechanismsfor the three excited species are different and listed separately.Although air molecules were diffused into the nozzle, whichis indicated by the N2 bands’ emission volumes as shownin figure 4, the air concentration was considered to be verylow and would not play an important role in the quenchingprocesses of excited and metastable Ar for the plasma inside thenozzle. This is indicated by the shorter N2 emission durationtime τN2 than τAr and τOH for the plasma inside, as shownin tables 1 and 2. Due to the high quenching rates by N2

molecules (10−11–10−10 cm3 s−1), the excited states Ar(4p)and OH(A 2�+) will vanish quickly and result in a fast decayin the corresponding emission intensity if the N2 concentrationis high. Furthermore, the argon gas used in this work is of highpurity with only 0.5 ppm water vapour. The N2 impurity in theargon gas stream is due to the air diffusion effect. The innerdiameter of the nozzle is small, 0.8 mm, and the gas speedat the exit of the nozzle is estimated to be about 460 cm s−1.The N2 concentration inside the nozzle is reasonably low.However, a detailed fluent simulation is desired for gettingmore information about the exact N2 concentration in the argon

7

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

gas stream, which will be done in our future work and is notincluded in this paper. Therefore, the quenching processesinside the nozzle are dominated by collisional de-excitation byargon atoms and radiative decay. For the plasma plume in openair, the quenching processes are induced mainly by the largenumber of air molecules, i.e. nitrogen and oxygen molecules.

The long duration time τOH inside the nozzle is attributedto the quenching mechanisms of OH(A 2�+) by collisionaldeactivation with argon atoms and radiative de-excitation. Wedo not emphasize the collisional quenching by water moleculesdue to their very low concentration inside the nozzle (the waterimpurity in argon gas is 0.5 ppm). The characteristic times ofthe two quenching processes are estimated to be in the range100–800 ns, which is in the same order of magnitude withτOH of 420 ns as shown in tables 1 and 2. However, whenthe plasma propagated out from the nozzle, the immediatedecrease in OH emission intensity was attributed mainly tothe large amount of air diffusing into the argon gas stream.In this case, the excited and metastable argon atoms aredissipated significantly by the large number of nitrogen andoxygen molecules with a high reaction rate in the range 10−11–10−10 cm3 s−1. Although the quenching rate of the excitedand metastable argon atoms by water is high (10−10 cm3 s−1),the number density of water in the argon stream is low anddepends on the atmospheric moisture and air diffusion effect.The normal atmospheric moisture is in the range 0–4%, anorder of magnitude smaller than that of nitrogen and oxygen(78% and 21%) [49]. The significant quenching of excitedargon atoms was also indicated by the immediate increase inthe N2 emission intensity and rapid decay of Ar irradiance assoon as the plasma came out from the nozzle, as shown infigures 4 and 5. Therefore, the production of the excited stateOH(A 2�+) would be limited, as the energy transfer processbetween H2O and excited (or metastable) Ar is an importantgeneration mechanism for the hydroxyl radical at the A 2�+

state in our case. Furthermore, the collisional de-excitationrate of the excited state OH(A 2�+) by N2 and O2 is veryhigh. The excited OH(A 2�+) would be dissipated as soon asgenerated while the plasma propagated in open air. Becauseof these adverse effects induced by diffused air, the OH bands’emission underwent a very fast decay for the plasma plume inthe open space.

We attribute the short τN2 of N2 emission inside the nozzleto the two following facts. First, the N2 concentration isconsidered to be low for the gas stream inside the nozzle,which has been discussed above. The low density of N2 limitedthe generation of the excited state N2(C

3�u) through energytransfer collisions with excited and metastable argon atoms.The second reason for the short τN2 is the fast quenching ofthe excited state N2(C

3�u) through spontaneous radiativedecay. The spontaneous radiative lifetime of the transitionfrom N2(C

3�u) to N2(B3�g) is about 37 ns, in the same

order of magnitude with the emission duration (50–60 ns),as shown in tables 1 and 2. For the plasma propagating inopen air, the N2 emission could last for a longer duration timecompared with the case inside the nozzle. However, it shouldbe mentioned that the long τN2 is featured for the whole plasmaplume propagating in the open space. The plasma plume is

formed by a series of self-propagating plasma ionization frontswith a high propagation velocity (104–105 m s−1) in ambientair. Actually, τN2 of the plasma ionization front at each spatialposition along the gas stream is also short (<100 ns). Thiscan be seen from the sub-images shown in figure 4, whichform the time- and space-resolved propagation behaviour ofN2 emission volumes in open air. The situation is similar forthe spatially resolved Ar lines’ emission volumes, for whichthe emission intensity decayed quickly and lasted for a shorterduration time compared with the case of the plasma ionizationfront travelling inside the nozzle, as shown in figure 5. Theseobservations are understandable with the fast quenching effectsby the increasing concentrations of N2 and O2 along the argongas stream in the open space.

The absence of emissions from H and O lines in theargon plasma are mainly due to the following facts. First, theconcentration of reactive species, such as excited argon atomsat high levels, which carry a high energy sufficient to produceexcited states H∗ from ground state H2O, or O∗ from O2, isvery low in our case. The energy threshold of H(n = 3) fromH2O is about 17.2 eV, and 15.9 eV for O(5P) from O2. Bothenergy levels are higher than the energy level of the excitedstates Ar(4p) (13–14 eV). Therefore, the direct generation ofH∗ and O∗ from H2O and O2 through energy transfer processeswith excited argon atoms is ignorable in our case. The excitedspecies H∗ and O∗ can be produced from their ground stateatoms through energy transfer processes with excited Ar. Theenergy required to produce H(n = 3) from ground state H isabout 12.1 eV, and O(5P) from ground state O is about 10.7 eV.These two potential levels are in the range of the energy carriedby Ar(4p), or metastables Ar(4s) (for O(5P) from O). However,the quenching processes of H∗ and O∗ are very fast in ouratmospheric pressure plasma. The quenching rates of H∗ andO∗ by N2 and O2 are in the range 10−9–10−10 cm3 s−1, and10−11–10−10 cm3 s−1 by Ar [50]. The characteristic times ofthese quenching processes are estimated to be less than 1 nsunder our atmospheric pressure conditions. The spontaneousradiative lifetime of transition H(n = 3) → H(n = 2) isin the range 14–45 ns, and 27 ns for O(5P) → O(5S0); bothare higher than the characteristic de-excitation time by thegas background [38]. As a result of these mechanisms, theemissions from the excited H and O atoms are very weak andcannot be detected under our experimental conditions.

5. Conclusions

In summary, a novel emission-dispersion method was designedfor the time- and space-resolved dynamics study on an open-airpulsed argon plasma plume. Based on the captured dispersedemission volumes in the high-speed image sequences, thepropagation behaviour of different irradiative excited speciesin the argon plasma plume was clearly observed. It is found thatthe OH bands’ emissions were able to last for a long durationinside the nozzle, but decayed rapidly as soon as the plasmapropagated into open air. Distinct from the OH emissions, theN2 bands’ emissions faded away immediately while the plasmatravelled inside the nozzle. But for the plasma propagatingin open air, the N2 emission volumes were able to travel to

8

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

a longer distance and formed the whole length of the argonplasma plume. Due to the adverse effects by air diffusion,the dispersed Ar emission volumes exhibited a small emissionduration time and propagated to a short distance in the openspace. All these different types of time- and space-resolvedbehaviour of the investigated emission volumes are attributedto the different generation and quenching mechanisms of theircorresponding excited species in the argon plasma plume.

Acknowledgments

This work is supported in part by the China ScholarshipCouncil (CSC) and the Co-funding scholarship of GhentUniversity, and Interuniversity Attraction Poles Programmeof the Belgian Science Policy (Project PSI”-P6/08).

References

[1] Laroussi M 2009 Low-temperature plasma for medicine?IEEE Trans. Plasma Sci. 37 714–25

[2] Weltmann K-D, Kindel E, Brandenburg R, Meyer C,Bussiahn R, Wilke C and Woedtke T V 2009 Atmosphericpressure plasma jet for medical therapy: plasma parametersand risk estimation Contrib. Plasma Phys. 49 631–40

[3] Benedikt J, Raballand V, Yanguas G A, Focke K andKeudell A V 2007 Thin film deposition by means ofatmospheric pressure microplasma jet Plasma Phys.Control. Fusion 49 B419–27

[4] Noh J H, Baik H K, Noh I, Park J C and Lee I S 2007 Surfacemodification of polytetrafluoroethylene using atmosphericpressure jet for medical application Surf. Coat. Technol.201 5097–101

[5] Laroussi M and Akan T 2007 Arc-free atmospheric pressurecold plasma jets: a review Plasma Process. Polym. 4 777–88

[6] Laroussi M and Lu X P 2005 Room-temperature atmosphericpressure plasma plume for biomedical applications Appl.Phys. Lett. 87 113902

[7] Lu X P, Jiang Z H, Xiong Q, Tang Z Y, Hu X W and Pan Y2008 An 11 cm long atmospheric pressure cold plasmaplume for applications of plasma medicine Appl. Phys. Lett.92 081502

[8] Cao Z, Walsh J L and Kong M G 2009 Atmospheric plasma jetarray in parallel electric and gas flow fields forthree-dimensional surface treatment Appl. Phys. Lett.94 021501

[9] Hong Y C and Uhm H S 2006 Microplasma jet at atmosphericpressure Appl. Phys. Lett. 89 221504

[10] Teschke M, Kedzierski J, Finantu E G, Korzec D andEngemann J 2005 High-speed photographs of a dielectricbarrier atmospheric pressure plasma jet IEEE Trans.Plasma Sci. 33 310–11

[11] Fridman G, Friedman G, Gutsol A, Shekhter A B, Vasilets V Nand Fridman A 2008 Applied plasma medicine PlasmaProcess. Polym. 5 503–33

[12] Laroussi M 2005 Low temperature plasma-based sterilization:overview and state-of-the art Plasma Process. Polym.2 391–400

[13] Lee H J, Shon C H, Kim Y S, Kim S, Kim G C and Kong M G2009 Degradation of adhesion molecules of G361melanoma cells by a non-thermal atmospheric pressuremicroplasma New J. Phys. 11 115026

[14] Shashurin A, Keidar M, Bronnikov S, Jurjus R A andStepp M A 2008 Living tissue under treatment of coldplasma atmospheric jet Appl. Phys. Lett. 93 181501

[15] Liu S H and Neiger M 2001 Excitation of dielectric barrierdischarges by unipolar submicrosecond square pulsesJ. Phys. D: Appl. Phys. 34 1632–8

[16] Larroussi M and Lu X P 2004 Power consideration in thepulsed dielectric barrier discharge at atmospheric pressureJ. Appl. Phys. 96 3028–30

[17] Xiong Q, Lu X P, Ostrikov K, Xian Y B, Zou C L, Xiong Z Land Pan Y 2010 Pulsed dc- and sine-wave-excited coldatmospheric plasma plumes: a comparative analysis Phys.Plasmas 17 043506

[18] Walsh J L and Kong M G 2007 Room-temperature atmosphericargon plasma jet sustained with submicrosecondhigh-voltage pulses Appl. Phys. Lett. 91 221502

[19] Lu X P, Xiong Z L, Zhao F, Xian Y B, Xiong Q, Gong W W,Zou C L, Jiang Z H and Pan Y 2009 A simple atmosphericpressure room-temperature air plasma needle device forbiomedical applications Appl. Phys. Lett. 95 181501

[20] Park H S, Kim S J, Joh H M, Chung T H, Bae S H andLeem S H 2010 Optical and electrical characterization of anatmospheric pressure microplasma jet with a capillaryelectrode Phys. Plasmas 17 033502

[21] Iza F, Kim G J, Lee S M, Lee J K, Walsh J L, Zhang Y T andKong M G 2008 Microplasmas: sources, particle kineticsand biomedical applications Plasma Process. Polym.5 322–44

[22] Stoffels E, Sakiyama Y and Graves D B 2008 Coldatmospheric plasma: charged species and their interactionswith cells and tissues IEEE Trans. Plasma Sci. 36 1441–57

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

[24] Lim J P, Uhm H S and Li S Z 2007 Influence of oxygen inatmospheric-pressure argon plasma jet on sterilization ofBacillus atrophaeous spores Phys. Plasmas 14 093504

[25] Ye R and Zheng W 2008 Temporal-spatial-resolvedspectroscopic study on the formation of an atmosphericpressure microplasma jet Appl. Phys. Lett. 93 071502

[26] Xiong Q et al 2009 Temporal and spatial resolved opticalemission behaviors of a cold atmospheric pressure plasmajet J. Appl. Phys. 106 083302

[27] Lu X P, Jiang Z H, Xiong Q, Tang Z Y and Pan Y 2008 Asingle electrode room-temperature plasma jet device forbiomedical applications Appl. Phys. Lett. 92 151504

[28] Alexander A B, Brett A C and Surendra P S 2004Determination of gas temperature and thermometric speciesin inductively coupled plasmas by emission and diode laserabsorption Plasma Sources Sci. Technol. 13 691–700

[29] Fantz U 2004 Emission spectroscopy of molecular lowpressure plasmas Contrib. Plasma Phys. 44 508–15

[30] Jachson G P and King F L 2003 Probing excitation/ionizationprocesses in millisecond-pulsed glow discharges in argonthrough the addition of nitrogen Spectrochim. Acta B58 185–209

[31] Pellerin S, Cormier J M, Richard F, Musiol K and Chapelle J1996 A spectroscopic diagnostic method using UV OHband spectrum J. Phys. D: Appl. Phys. 29 726–39

[32] Dieke G H and Crosswhite H M 1962 The ultraviolet bands ofOH. Fundamental data J. Quant. Spectrosc. Radiat.Transfer 2 97–199

[33] Chidsey I L and Crosley D R 1980 Calculated rotationaltransition-probabilities for the A–X system of OH J. Quant.Spectrosc. Radiat. Transfer 23 187–99

[34] Bruggeman P, Schram D, Gonzalez M A, Rego R, Kong M Gand Leys C 2009 Characterization of a direct dc-exciteddischarge in water by optical emission spectroscopy PlasmaSources Sci. Technol. 18 025017

[35] Motret O, Hibert C, Pellerin S and Pouvesle J M 2000Rotational temperature measurements in atmospheric

9

J. Phys. D: Appl. Phys. 43 (2010) 415201 Q Xiong et al

pulsed dielectric barrier discharge—gas temperature andmolecular fraction effects J. Phys. D: Appl. Phys. 33 1493–8

[36] Iordanova E 2010 Poly-diagnostic validation of spectroscopicmethods PhD Thesis Eindhoven University of Technology,Netherlands

[37] Dasgupta A, Blaha M and Giuliani J L 1999 Electron-impactexcitation from the ground and the metastable levels of Ar I

Phys. Rev. A 61 012703[38] NIST Chemical Kinetics Database, Standard Reference

Database 17, Version 7.0 (web version), Release 1.4.2,Data Version 2009.01 (http://kinetics.nist.gov/kinetics/)

[39] Morgan L A 1998 Electron impact excitation of water J. Phys.B: At. Mol. Opt. Phys. 31 5003–11

[40] Velazco J E, Kolts J H and Setser D W 1978 Rate constantsand quenching mechanisms for the metastable states ofargon, krypton, and xenon J. Chem. Phys. 69 4357–73

[41] Teulet P, Sarrette J P and Gomes A M 1999 Calculation ofelectron impact inelastic cross sections and rate coefficientsfor diatomic molecules. Application to air moleculesJ. Quant. Spectrosc. Radiat. Transfer 62 549–69

[42] Piper L G 1988 State-to-state N2(A3�+

u ) energy-poolingreactions: I. The formation of N2(C

3�u) and the Hermaninfrared system J. Chem. Phys. 88 231–9

[43] Sadeghi N, Setser D W, Francis A, Czarnetzki U andDobele H F 2001 Quenching rate constants for reactions of

Ar(4p′[1/2]0, 4p[1/2]0, 4p[3/2]2, and 4p[5/2]2) atoms with22 reagent gases J. Chem. Phys. 115 3144–54

[44] Bogaerts A and Gijbels R 1995 Modeling of metastable argonatoms in a direct-current glow discharge Phys. Rev. A52 3743–51

[45] Leblond J B, Collier F, Hoffbeck F and Cottin P 1981 Kineticstudy of high-pressure Ar–H2O mixtures excited byrelativistic electrons J. Chem. Phys. 74 6242–5

[46] Tamura M, Berg P A, Harrington J E, Luque J, Jeffries J B,Smith G P and Crosley D R 1998 Collisional quenching ofCH(A), OH(A), and NO(A) in low pressure hydrocarbonflames Combust. Flame 114 502–14

[47] Lofthus A and Krupenie P H 1977 The spectrum of molecularnitrogen J. Phys. Chem. Ref. Data 6 113–307

[48] Pancheshnyi S V, Starikovskaia S M and Starikovskii A Y2000 Collisional deactivation of N2(C

3�u, v = 0, 1, 2, 3)states by N2, O2, H2 and H2O molecules Chem. Phys.262 349–57

[49] Pidwirny M 2006 Atmospheric composition Fundamentals ofPhysical Geography 2nd edn; http://www.physicalgeography.net/fundamentals/7a.html

[50] Bittner J, Kohse-hoinghaus K, Meier U and Just T 1988Quenching of two-photon-excited H(3s, 3d) and O(3p3P2,1,0) atoms by rare gases and small molecules Chem.Phys. Lett. 143 571–6

10