HIGH ELECTRON DENSITY, ATMOSPHERIC PRESSURE AIR GLOW
`DISCHARGES
`
`Frank Leipold, Abdel-Aleam H. Mohamed, and Karl H. Schoenbach
`Physical Electronics Research Institute, Old Dominion University, NorfiJIk, VA 23529
`
`Abstract
`
`The pulsed electron heating efi‘ect has been studied on an
`atmospheric pressure air glow discharge. Application of a
`high voltage pulse causes a shifi in the electron energy
`distribution function to higher energies. This causes a
`temporary increase of the ionization rate and consequently
`an increase of the electron density. The electron density
`after a 10 ns pulse application to a direct current glow
`dischargeincreased fromits dc value of 2 1013 cm'3 to 2.8
`10” cm'3 The average power density,
`required for
`sustaining the high pressure plasma with a given
`minimum electron density, was found to be lowered when
`the discharge was operated in a repetitive pulsed mode
`compared to a dc mode. For an atmospheric pressure air
`plasma, an average power density of 15 kW/cm3 and 50
`W/cm3 is required3for an average electron density of 1013
`cm'3 and [02 cm, respectively. This valueis less by a
`factor of three thari that required to sustain a dc plasma
`With the same base electron density
`
`required for reflection of microwave radiation of up to 30
`GHz [1]. The MCSG plasma was found to be scalable in
`size by extending the electrode gap and by placing the
`discharges
`in parallel
`[2]. Hewever,
`at
`equilibrium
`conditions,
`the power density required to sustain an
`atmospheric pressure air plasma of 10'3 cm‘3 electron
`density is approximately 5 kW/cm3 [3], a value which
`makes these equilibrium plasmas difficult to scale to large
`volumes.
`
`Pulsed electron heating has been shown to allow
`reduction of the electrical power, while keeping the
`average electron density at
`the
`required level
`for
`microwave reflection [4]. In order to explore the efi‘ect of
`pulsed electron heating on the temporal deveiopment of
`single discharges and discharge arrays we have measured
`the electrical and the optical reSponse to pulsed electron
`heating with a temporal resolution on the order of 10 ns.
`Laser interferometry, electrical conductivity and optical
`spectroscopy was used to determine
`the temporal
`development of electron density, and gas temperature.
`
`1. INTRODUCTION
`
`II. EXPERIMENTAL SETUP
`
`‘ Weakly ionized plasmas, generated in high pressure air
`g10w discharges,
`reflect
`or
`absorb electromagnetic
`radiation in the microwave range and consequently act as
`temporally controllable barriers for this radiation: as
`plasma ramparts. Direct current microhollow cathode
`sustained glow discharges (MCSG) have been showu to
`provide plasmas with an electron density of 1013 cm's,
`
`l. The
`shown in Fig.
`setup is
`experimental
`The
`atmospheric pressure air discharge(s) were operated in a
`direct current mode, with a 10 ns to 12 ns voltage pulse
`superimposed. The gap was set at 0.6 cm, the distance
`between
`discharge
`axes,
`for
`a
`three
`discharge
`arrangement, was 0.4 cm (Fig. 2).
`
`
`
`Line Type
`Pulse Generator
`_l_L
`
`
`
` Electrical Field:
`
`5 kV/cm to 10 kV/cm
`Gap Distance:
`2 mm to 10 mm
`
`Figure 1. Experimental setup
`
`serve as
`(MI-1CD)
`Microhollow cathode discharges
`plasma cathodes.
`In order to increase the size of the
`III-78034540810261 7.00©2002 IEEE
`
`plasma, three discharges were operatedin parallel [2]. The
`discharges
`can
`be
`operated
`either
`in DC with
`1 3O
`
`TSMC-1018
`
`TSMC v. Zond, Inc.
`
`Page 1 of 4
`
`TSMC-1018
`TSMC v. Zond, Inc.
`Page 1 of 4
`
`

`

`in pulsed mode only. Two
`superimposed pulses or
`line
`type
`pulse generators
`independent
`triggerable
`provided 10 ns pulses. The applied diagnostics are
`emission spectroscopy for gas temperature measurements,
`interferometry [l], and conductivity measurements for
`electron density measurements. High-speed photography
`was used to obtain the spatial plasma distribution.
`
`\ll l( 'l.)
`
`(mitotic-
`
`at
`parallel
`2. Three MCSGs operated in
`Figure
`atmospheric pressure in air. Electrode gap: 6 mm,
`distance between two discharges: 4 mm.
`
`III. EXPERIMENTAL RESULTS
`
`A. Gas temperature
`The gas temperature is obtained by comparison of a
`measured and a simulated spectrum of the 2"“ positive
`system of nitrogen. For a DC glow discharge, the gas
`temperature was feund to be 2200 K close to the cathode.
`The temperature in the plasma column reaches with
`increasing gap length a constant level of 2000 K (Fig. 3).
`
`2409
`
`M1 AnodeZ Anode3
`
`[K] § 1600
`GasTemperature
`
`
`l88
`
`88
`
`1400
`
`1200
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`Distance from MHCD [mm]
`
`Figure 3. Gas temperature on the axis of a DC MCSG for
`various electrode distances. The discharge current was 13
`m.
`
`For a DC operated MCSG with a superimposed pulse, an
`increase of the gas temperature by 300 K was measured
`10 ns afler pulse application (Fig. 4). Due to the low light
`intensity 25 as after pulse application, information on the
`
`decay of the temperature could not be obtained with this
`method.
`
`2400
`
`Exposure Time: 1 ns
`.0.
`
`O
`
`.0
`
`gig
`
`-
`
`DC Temperature (2000 K)
`
`1900 10m)
`
`[K] a
`GasTemperature
`
`
`13m
`
`0
`
`rHLLJHLLIQlJLJpJu‘L—ILIW
`5
`10
`15
`20
`25
`
`30
`
`Delay Time [ns]
`
`Figure 4. Temporal Development of the gas temperature
`in the center of a MCSG discharge operated DC with
`superimposed pulse. Electrode distance: 2 mm, IMCSG Dc =
`10 mA.
`
`8. Electron density
`and
`spatially
`The
`electron
`density was measured
`temporally resolved by means of infrared heterodyne
`interferometry. The radial profile was found to be time
`independent. It can be fit by a gaussian profile with a
`width of c = 0.056 mm. Due to the limited temporal
`resolution, the electron density cannot be measured during
`the pulse. However, a measurement at 22 ns alter pulse
`application provides an electron density of 2.8 10‘5 cm"
`for a discharge with a gap distance of 2 mm and an
`applied pulsed electrical field of 8 kV/crn. This indicates,
`that the electron density during the pulse is at least 2.8
`10" cm". The radial electron density distribution in the
`center plane of the discharge for difierent
`times alter
`pulse application is shown in Fig. 5.
`
`w
`
`cm“)
`ElectronDensity[10‘5
`
`
`~01
`
`0.0
`
`0.1
`
`Distance y from Center [mm]
`
`Figure 5. Radially resolved electron density in the center
`plane of the discharge for difi‘erent
`times alter pulse
`application. Electrode Distance: 2 mm, applied electrical
`field: 8 kVIcm.
`
`131
`
`TSMC-1018 / Page 2 of 4
`
`TSMC-1018 / Page 2 of 4
`
`

`

`The electron density after pulse application can also be
`obtained from measurements of the electrical field and
`current density. The relation between current density;
`electron density and electrical
`field is given by the .
`equation.
`g
`
`1 = n. e y(E/n)
`
`i
`(l)
`1
`~
`5
`.
`The electron density was calculated using average valug
`, of electric field (E,= V/d) and current density J. The
`current densityis given by the measured current and the
`spatial distribution profile of the electron density, whichis
`assumod to be represented by the optical emission profile.
`Photographs of the discharge plasma for difi‘erent times
`after pulse application are shown in Fig 6 The eimosure
`time is 5 us The drifi velocity v, which deepends on the
`reduced electrical field, varies between 2 10 m/s and 105 _
`m/s for reduced electrical fields between 10 ”Id and 200
`
`'
`
`.
`
`.
`
`.
`
`,
`
`Td 15]
`
`.
`
`--— Modeling
`Pulsed oc off
`(Conductivity)
`fl Pulsed, DC on
`(Conductivity)
`Pulsed. DC on
`(Interferometry)
`
`_
`
`-
`
`’
`
`(cm‘g)
`ElectronDensityDifferencenp—no
`
`
`
`
`10w
`
`,8
`
`to
`
`' 1°"
`
`1016
`
`1015
`10"
`
`‘
`
`. 1013
`
`1012
`
`1011
`
`1010
`300
`
`_
`
`_F
`l'l’llllt‘ lll
`\ik‘LlSLll'L'lllL'lll
`
`l‘
`l
`I"
`l ”l“ ‘1
`\iflmm‘mklm
`
`|
`
`Stills
`
`\lll‘l) .*\llHth'
`
`’
`i:
`
`,
`
`n -
`
`_
`
`I’lmlc (ll‘
`\lcllslll'cnll‘lll
`
`l’lmil‘ nl‘
`\lent-.llt'cltl-cnl.
`
`D
`lhll 11>
`
`\lllr'll
`
`|
`
`l
`
`100
`
`200
`
`.
`
`Reduced Electric Field. EIN (Td)
`
`7 Figure 7. Electron density difi‘erence alter a 10 ns pulse
`application versus the applied electrical field. The solid
`line represents modeling results [4].
`.
`.
`For high pulsed electrical field, the dc contribution to the
`electron density can be neglected, For investigation of the
`electron heating efi‘ect for low applied electrical fields, the
`current of the discharge has to be turned off before the
`pulsers applied. A typical temporal development of the
`- current is shown in Fig. 8.
`
`_8
`
`'
`
`.
`
`6
`5 ..
`H ‘
`if?
`’5'
`U 2
`
`Figure 6. Photographs of the MCSG for difi‘erem times ,
`after pulse application. ‘Electrode gap: 2min, applied
`electrical field: 10 kV/cm [6].
`,
`.
`_,
`
`‘
`
`0
`
`0.0
`
`field of 10 kV/crn and an
`For an applied electrical
`electrode distance of 2 turn,
`the FWHM of the radial
`profile was found to be 0.16 mm. The profile is tiine
`independent The electron density after pulse application
`versus the reduced electrical fieldls shownin Fig 7
`
`' 0.5
`
`1.0
`
`1.5
`
`2.0
`
`Time [ms]
`,
`'
`
`Figure 8. Temporal deveIOpment of the discharge current.
`Before the tuio voltage pulses were applied,
`the direct
`current was turned off.
`
`132
`
`TSMC-1018 / Page 3 of 4
`
`TSMC-1018 / Page 3 of 4
`
`

`

`C. Power Density
`The power density, P, for repetitive pulsed mode is:
`
`All factors'in this equation can be expressed'in terms of
`the electrical
`field intensity. The expression for
`the
`current density18 given in equ. 1. The drift velocityin this
`equation depends on the reduced electrical field [5]. The
`repetition time, thy, is the time, required for the electron
`density to decay fi'om the peak value to a minimum value.
`In our case, the major electron loss process is dissociative
`recombination. Attachment
`can
`be
`neglected
`in
`atmospheric pressure air plasmas with a gas temperature
`exceeding 1500 K [7]. The repetition time th, is therefore
`given by
`
`For an electron densityof 1013 cm"3,the minimum power
`density1S 085 kW/cma’. For an electron density of 10‘2
`cm‘3, 3the power consumption can be reduced to 18
`W/cma. The theoretical values are confirmed by the
`experimental results.
`
`IV. sirmnv
`
`Atmospheric pressure air plasmas could be generated with
`characteristic
`dimensions
`of
`centimeters
`at
`gas
`temperatures of 2000 K. A reduction in the power
`consumption compared to the DC glow discharge could
`be achieved by operating the discharges in pulsed mode.
`Minimum power densities required to sustain atmospheric
`pressure air plasmas with electron densities of 101
`cm
`and 1012 crn'3 are 850 W/cm3 and 18 W/cm3,respectively.
`
`tRep = (Hp-noflnp’mo B)
`
`(3)
`
`V. ACKNOWLEDGEMENT
`
`the peak electron density alter pulse
`is
`where tip
`application, no is the minimum electron density, and [i is
`the recombination coefficient. As shown in Fig. 7,
`the
`peak electron density is a function of the applied reduced
`electrical field.
`
`.
`
`The total power consumption for atmospheric pressure
`plasmas with minimum electron densities versus the
`applied electrical field is shown in Fig. 9. The solid lines
`represents the modeling results and the squares and circles
`the experimental results. If the applied electrical field is
`too low, a high repetition rate is required, a mode of
`operation which approaches direct current operation.
`Consoquently the power increases towards the dc value.
`With increasing electrical field, most of the energy is used
`to generate electron densities far exceeding the desired
`minimum value. This density decreases rapidly due to
`recombination and as shown in equ. 3, contributes only
`minimally to the repetition time. Consequently, there is an
`optimum electrical
`field
`for
`minimum power
`consumption, as shown in Fig. 9.
`
`ll?
`
`§
`
`[kW/curs]
`PowerDensity
`
`
`10°
`
`This work was supported by the US Air Force Ofiice of
`Scientific Research, and the National Science Foundation
`(Award # INT-0001438).
`
`VI. REFERENCES
`
`[1] Frank Leipold, Robert H. Stark, Ahmed El-Habachi,
`and Karl H. Schoenbach, “Electron Density
`Measurements in an Atmospheric Pressure Air
`Plasma by Means of IR Heterodyne Interferometry”
`I. Phys. D: Appl. Phys. 1;, 2268 (2000).
`[2] Abdel-Aleam H. Mohamed, Rolf Block, and Karl H.
`Schoenbach “Direct Current Glow Discharges in
`Atmospheric Air,” IEEE Trans. Plasma Science. :9,
`182 (2002).
`[3] Robert H. Stark and Karl H. Schoenbach, “Direct
`Current Glow Discharges in Atmospheric Air,” Appl.
`Phys. Lett. 11, 3770 (1999).
`[4] Robert H. Stark and Karl H. Schoenbach, “Electron
`Heating in Pulsed Atmospheric Pressure Glow
`Discharges,” J. Appl. Phys. 1L9, 3568 (2001).
`[5] A. V. Phelps, “Excitation and Ionisation
`Coefiicients”, in Mus Dielg'tn‘cs V, L. G.
`Christophourou, and D. W. Bouldin, eds, Pergamon
`Press, 1987, p. 1
`[6] Hisham Merhi, Masters Thesis, Old Dominion
`University, Norfolk, VA, August 2001.
`[7] IAux, C.O., Yu, L., Packan, 11M, Gessman, R1,
`Pierrot, L, Kruger, C.H., and late, R.N.,
`30th AIAA Plasmadynamics and Lasers Conference,
`Norfolk, VA, June 28-July 1, 1999
`
`Electrical Field [kV/cm]
`
`Figure 9. Power density versus the applied electrical field
`for diEerent electron densities. The solid lines represent
`the modeling results,
`circles and squares
`represent
`measured values.
`
`133
`
`TSMC-1018 / Page 4 of 4
`
`TSMC-1018 / Page 4 of 4
`
`