Electrical discharge plasma characteristics in pure Ar gas...

Electrical discharge plasma characteristics in pure Ar gas

at multi-atmospheric pressure using the automatically

pre-ionized plasma electrode

July 2005

Department of Energy and Materials Science Graduate School of Science and Engineering

Saga University Sung-Ki Hong

- I -

ACKNOWLEDGEMENTS

I am thankful to my advisor, Professor Chobei Yamabe for his guidance, support

and cooperation throughout the course of my education. This work is the result of his

constant inspiration, encouragement and words of wisdom.

My sincere thanks to the members of my dissertation committee, Professor

Hiroharu Fujita, Professor Saburoh Satoh and Professor Kazuhiro Muramatsu for

their thoughtful comments.

I am grateful to the Ministry of Education, Science, Sports and Culture, Japan for

their financial support. Further, I am thankful to Dr. Satoshi Ihara and Dr. Nobuya

Hayahsi. Special thanks to Professor Sang-Bong Wee and Dr. Hee-Sung Ahn.

I am also very grateful to my family for their invaluable affections and

encouragement during our stay in Kyushu, Japan.

Finally, I thank my wife Yong-Jin Chun for her patience, devotion and

understanding for the past 3 years.

Saga University, Saga, Japan

June 20, 2005

Sung-Ki Hong

홍 성 기

- II -

ABSTRACT

Electrical discharge plasma characteristics in pure Ar gas at multi-atmospheric pressure using the automatically pre-ionized plasma electrode

Sung-Ki Hong

B.S., Korea University of Technology and Education

M.S., Korea University of Technology and Education

Ph.D., Saga University

In this study, for the investigation of the possibility of the Ar2* excimer laser

action and the development of the electrode structure for the efficient transversely

excited (TE) gas laser, a new pre-ionization electrode was designed using the

surface-corona pre-ionization method. It was investigated experimentally in terms of

generated charge density, electrical characteristics. These experimental results

suggest that a sharp edge of the ground electrode is possible to radiate the strongest

UV light. Therefore, the Automatically Pre-Ionized (API) plasma electrode system

was designed using a new pre-ionization electrode by the surface-corona pre-

ionization method. Characteristics of the main-discharge in multi-atmospheric (1~5

atm) pure Ar gas were investigated using the automatically pre-ionized plasma

electrode. The uniform main-discharge was formed and its volume and the

breakdown voltage increased with increasing Ar gas pressure. The instantaneous

maximum discharge electric power was 90 MW at 5 atm Ar gas and the maximum

energy deposition was 1.4 MW/cm3. It was demonstrated that the uniform pre-

ionization formed the uniform main-discharge by the control effect of Cpr and the

maximum energy deposition was increased.

- III -

The control effect of Cpr was examined the time dependent main-discharge from

two-dimensional simulation of electric field distribution of the automatically pre-

ionized plasma electrode discharge system. In addition, the light emission

characteristics of the discharge system using the automatically pre-ionized plasma

electrode were investigated by the measurement of Ar atomic line and Ar ionic line.

The intensity of the line at 427.8 nm increases proportionally to the approximate

cubic of the Ar gas pressure (Y = 0.22X3, where Y is intensity and X is the Ar gas

pressure). On the other hand, the line intensity at 696.9 nm shows saturation at

pressure above 4 atm. These experimental results suggest that the lasering band

Ar2*(1Σu) is enhanced due to the increase of the excited Ar* atom which is formed

by the electron collision reaction Ar + e→Ar* + e and the Ar2*(3Σu) is saturated by

the recombination process. On the basis of such experimental results, the new

resonator for the Ar2* excimer laser action was designed.

- IV -

TABLE OF CONTENTS

page ACKNOWLEDGEMENTS ....................................................................................... I ABSTRACT................................................................................................................II 1. Introduction............................................................................................................ 1

1.1 The electron beam pumped method ................................................................ 3 1.2 The discharge pumped method ....................................................................... 4 1.3 Aim of this study ............................................................................................. 8 Reference............................................................................................................... 9

2. Theoretical considerations ...................................................................................11

2.1 Rare gas kinetics considerations ................................................................... 12 2.2 Discharge formation and stability in high pressure gas ................................ 15 2.3 General characteristics of the glow (or uniform) discharge.......................... 22 2.4 Excitation circuits.......................................................................................... 26 2.5 The main discharge ....................................................................................... 28 Reference............................................................................................................. 31

3. The pre-ionization source for excimer laser ...................................................... 32

3.1 Experimental set-up and methods ................................................................. 33 3.2 Experimental results and discussions............................................................ 35 Reference............................................................................................................. 38

4. The Automatically Pre-Ionized plasma electrode discharge system ............... 39

4.1 Experimental set-up ...................................................................................... 39 4.2 Circuit characteristics.................................................................................... 43 4.3 Experimental results and discussions............................................................ 45

4.3.1 Discharge characteristic of the API plasma electrode ......................... 45 4.3.2 Discharge characteristic of the API discharge system without Cpr...... 48 4.3.3 Using the plate electrode as an anode.................................................. 52

- V -

Reference............................................................................................................. 54 5. The electric field of API plasma electrode system............................................. 55

Reference............................................................................................................. 66 6. Characteristics of discharge pumped Argon gas excitation............................. 67

6.1 Light emission characteristics of discharge pumped Ar gas ......................... 70 Reference............................................................................................................. 78

7. Conclusions........................................................................................................... 80 Appendix #A Necessary conditions for the discharge-pumped Ar2* laser ......... 82 Appendix #B Design of a new chamber and electrode ......................................... 84 List of publications................................................................................................... 92

Ch.1 Introduction 1

1. Introduction

There have been considerable demands for the development of compact short

wavelength lasers in the vacuum ultraviolet (VUV) spectral region. Such compact

short wavelength lasers would be applicable to various scientific and industrial fields,

such as photochemistry, biological science, and new types of materials processing.

Recently more attention is paid to short wavelength lasers in the VUV as coherent

light sources in the future optical lithography industry. Optical lithography is

considered the most desirable technique for mass fabrication of advanced

semiconductor devices. Currently available practical compact VUV lasers are the

ArF excimer laser at 193 nm and the F2 laser at 157 nm, both of which are excited by

a compact discharge device. Possible path of lithography technologies is shown in

Fig. 1.1.

KrF(248nm) ArF(193nm)

SCALPEL(Scattering with Angular Limitation

Projection Electron beam Lithography)

(100keV electron)

Ar2*(126nm)

EUV(Extreme UltraViolet Lithography)

(13nm)

XRL(X-ray Lithography) (1nm, 1x proximity)

F2(157nm)

IPL(Ion Projection Lithography)

(75keV He+ ion)

NGL(Next-Generation Lithography)

<50nm

~110 or 90nm ~90 or 70nm ~50nm

Fig. 1.1 Possible path of lithography technologies

Ch.1 Introduction 2

∑∑ ++− g

1u

1,3

In optical lithography, resolution is given by the equation

NAkR λ

1esolution = (1.1)

where λ and NA are the exposure wavelength and numerical aperturea of the optical

lithography tool, and k1 is a constant for a specific lithographic process. As the

wavelength becomes shorter, the light source becomes more complex and expensive.

Turning now to optical devices, rare-gasb excimer lasers are a vacuum ultra-

violet (VUV) laser, achieved laser action from the transitions of the excited

dimer Ar2*(around 126 nm), Kr2*(around 146 nm), and Xe2*(around 172 nm). The

emission wavelength of Ar2* is 126 nm which is long enough to use transmission

optical elements such as MgF and LiF. The Kr2* laser has an even longer emission

wavelength centered at 147 nm which relaxes the conditions for optics and would

become a competitor to the F2 laser at 157 nm. From the practical viewpoint, rare

gases are chemically inert which shows a contrast to chemically active fluorine used

in the F2 laser. Among these rare gas excimers, Ar2* excimer laser produces radiation

with the highest photon energy of 9.8 eV. An electron beam excitation has been the

only excitation method to oscillate the Ar2* excimer laser [1-4]. However, the

electron beam excitation method requires a rather large facility to be operated,

resulting in a low average power of the VUV laser with a low repetition rate. The

electron beam excitation method, therefore, may be unsuitable for certain

a Numerical Aperture: Describes the angle in a cone of light emitted by the condenser and accepted by the objective of a microscope; the index of refraction of the medium in which the image lies multiplied by the sine of the half angle of the cone of light. b Rare gas: Alternative name for noble gas. Any of a group of six elements (helium, neon, argon, krypton, xenon, and radon), originally named ‘inert’ because they were thought not to enter into any chemical reactions.

Ch.1 Introduction 3

applications such as optical lithography despite its high peak power operation in the

VUV.

These rare-gas excimer lasers have been realized only by e-beam excitation until

recently [1-6]. Recently, the Kr2* excimer laser (the maximum output laser energy

was 150 µJ at 148 nm in 10 atm) have been realized by discharge-pumped scheme

[7]. This is a very interesting result. Despite various new schemes such as a four-

stage discharge [8], gas-jet discharge [5], silent discharge excitation [9] have been

studied and developed to obtain rare-gas excimer lasers action by discharge-pumped

scheme until recently, the Kr2* excimer laser action have been realized by a

conventional UV pre-ionized transverse compact discharge device finally. In the light

of this result, it is reasonable to suppose that there is the possibility of the Ar2*

excimer laser action by discharge-pumped method.

1.1 The electron beam pumped method

Figure 1.2 shows the fundamental electron beam-pumped laser components [10].

The electron beam source consists of a high voltage generator such as a Marx bank

or pulse transformer, a pulse transforming line to produce ideally square pulses of

20-100 ns duration and a vacuum diode. Electron emission is from a cold cathode

which is constructed from graphite or sharp blades to enhance the local electric fields

and produce more efficient and uniform emission. The anode consists of a foil,

usually of titanium, aluminum, stainless steel or aluminized dielectric, which is

sufficiently thin to allow efficient penetration by electrons with energies of 200 keV

or greater. The beams pass through a thin foil that isolates the diode from the high-

Ch.1 Introduction 4

pressure laser gas. The two vessels are separated by a thin foil the electrons have to

penetrate the foil which causes losses. The foil is a weak link in the system, reducing

the reliability of the device.

Fig. 1.2 The fundamental electron beam-pumped laser components [10]

1.2 The discharge pumped method

These types of discharge pumped lasers can be broadly classified into two

categories depending on the methods of generation of pre-ionization, viz., UV pre-

ionized lasers and X-ray pre-ionized lasers.

Figure 1.3 shows the fundamental X-ray pre-ionized laser components [11]. X-

ray photons have a high energy and a large penetration depth, thus resulting in a

Ch.1 Introduction 5

homogeneous electron density. However, X-rays are generated outside the laser

chamber and have much longer laser systems, albeit with substantial technological

complexity.

Fig 1.3 The fundamental X-ray pre-ionized laser components [11]

UV preionized discharge pumped lasers have the advantage of relative simplicity

and convenience. These types of UV pre-ionized lasers also, can be broadly

classified into two categories depending on the electrode structure for generation of

pre-ionization, viz., double discharge lasers and corona pre-ionized lasers.

In the double discharge laser, although simple unsustained discharge are

fundamentally unstable, fast, transverse discharge have been very successfully to

Ch.1 Introduction 6

pump lasers and have the advantage of simplicity of construction. At atmospheric

pressures and high electric fields, the electrons multiply very rapidly, the plasma

impedance collapses and spatial non-uniformities grow to form arcs in times which,

at pressure of several atmospherics, can be as short as a few tens of nanoseconds.

Figure 1.4 shows the fundamental double discharge transversely excited (TE) laser

components, the array of pre-ionizing pins was placed on one side at a distance of

~10 mm from the centre of the electrodes resulting in the shifting of the glow

discharge towards the pre-ionizer. When the main discharge was initiated after a

suitable time delay with respect to the pre-ionizing discharge, its stable self-sustained

avalanche mode of operation lasting several microseconds was possible resulting in

the efficient operation of a gas laser.

Heat Exchanger

Fan

Spark

Pin-electrode (for pre-ionization)

Electrode

Laser Beam

ElectrodeMain-discharge

Laser GasRecirculator

Mirror

Fig. 1.4 The fundamental double discharge TE laser components

Ch.1 Introduction 7

The second method of electron seeding the gas volume prior to the application of

the main discharge is to use the ultraviolet light by initiating a corona discharge. This

technique, a schematic diagram of which is shown in Fig. 1.5, was first demonstrated

by Lamberton and Pearson [12] in the operation of a TEA (Transversely Excited,

Atmospheric pressure) CO2 laser. The corona pre-ionized method is the high pre-

ionization uniformity, determined by a very homogeneous UV radiation flux that can

be produced along the main-discharge electrodes during the entire stage of

development of the main-discharge, with no jitter and very little delay.

H.V H.V

Cathode

Anode

Dielectric

(a) Plate type (b) Tube type

Fig. 1.5 Schematic diagrams of the corona pre-ionization

The double discharge method, which are particularly suitable for pumping very

large volume systems, are relatively cumbersome and cannot be easily scaled down

to small dimensions. Corona pre-ionized TE lasers, on the other hand, are easily

amenable to miniaturization.

Ch.1 Introduction 8

1.3 Aim of this study

The aim of this study is the investigation of the possibility of the Ar2* excimer

laser action and the development of the electrode structure for the efficient TE gas

laser. A new pre-ionization electrode is designed using the surface-corona pre-

ionization method. It is investigated experimentally in terms of generated charge

density, electrical characteristics. The automatically pre-ionized (API) plasma

electrode system also is designed using a new pre-ionization electrode by the

surface-corona pre-ionization method. Characteristics of the main-discharge in multi-

atmospheric (1~5 atm) pure Ar gas have been investigated using the automatically

pre-ionized (API) plasma electrode. It is examined the time dependent main-

discharge from two-dimensional simulation of electric field distribution of the API

plasma electrode discharge system. In addition, the light emission characteristics of

the discharge system using the API plasma electrode are investigated by the

measurement of Ar atomic line and Ar ionic line. On the basis of such experimental

results, it is the purpose to design the new resonator for the Ar2* excimer laser action.

Ch.1 Introduction 9

Reference

[1] W. M. Hugdes, J. Shannon, and C.K. Rhodes, 126.1-nm molecular argon laser, Appl. Phys. Lett. 24, 488 (1974) [2] W-G. Wrobel, H. Röhr, and K-H. Steuer, Tunable vacuum ultraviolet laser action by argon excimers, Appl. Phys. Lett. 36, 113 (1980) [3] Y. Uehara, W. Sasaki, S. Kasai, S. Saito, E. Fujiwara, Y. Kato, C. Yamanaka, M. Yamanaka, K. Tsuchida, and J. Fujita, Tunable oscillation of a high-power argon excimer laser in the vacuum-ultraviolet spectral region, Opt. Lett. 10, 487 (1985)

[4] K. Kurosawa, Y. Takigawa, W. Sasaki, M. Okuda, E. Fujiwara, K. Yoshida, and Y. Kato, High-power operation of an argon excimer laser with a MgF2 and SiC cavity, IEEE J. Quantum Electron. QE-27, 71 (1991) [5] S. K. Searles, J. E. Tucker, B. L. Wexler, and M. F. Masters, VUV emission from discharge-pumped Ar supersonic jets under high-excitation conditions, IEEE J. Quantum Electron. QE-30, 2141 (1994) [6] T Sakurai, N Goto, and C E Webb, Kr2* excimer emission from multi-atmosphere discharges in Kr, Kr-He and Kr-Ne mixtures, J. Phys. D: Appl. Phys. 20, 709 (1987) [7] W. Sasaki, T. Shirai, S. Kubodera, J. Kawanaka, and T. Igarashi, Observation of vacuum-ultraviolet Kr2* laser oscillation pumped by a compact discharge device, Opt. Lett 26, 503 (2001) [8] H. Ninomiya, and K. Nakamura, Ar2* excimer emission from a pulsed electric discharge in pure Ar gas, Opt. Commun. 134, 521 (1997)

[9] S. Kubodera, M. Kitahara, J. Kawanaka, W. Sasaki, and K. Kurosawa, A vacuum ultraviolet flash lamp with extremely broadened emission spectra, Appl. Phys. Lett. 69, 452 (1996)

Ch.1 Introduction 10

[10] J. D. Sethian, M. Friedman, J. L. Giuliani, R. H. Lehmberg, S. P. Obenschain, P. Kepple, M. Wolford, F. Hegeler, S. B. Swanekamp, D. Weidenheimer, D. Welch, D. V. Rose, and S. Searles, Electron beam pumped KrF lasers for fusion energy, Phys. Plasmas. 10, 2142 (2003)

[11] L. Feenstra, O. B. Hoekstra, P. J. M. Peters, and W. J. Witteman, On the performance of an ArF and a KrF laser as a function of the preionisation timing and the excitation mode, Appl. Phys. B 70, 213 (2000)

[12] P.R. Pearson and H. M. Lamberton, Atmospheric Pressure CO2 laser giving high output energy per unit volume, IEEE J. Quantum Electron. QE-8, 145 (1972)

Ch.2 Theoretical considerations 11

2. Theoretical considerations

The Ar2* excimer laser, in the case of electron-beam pumping, the pumping rate

exceeds 100 MW/cm3, which is quite high compared with that of rare-gas-halide

excimer lasers in the ultraviolet region and the F2 molecular laser at 157 nm, which

are typically of the order of less than 10 MW/cm3. In Table 2.1, the values of the

stimulated emissiona cross section (σ) times the lifetime (τ) of the laser upper state

(στ) are summarized. In the case of Ar2*, the value of στ is smaller almost by two

orders of magnitude, due to its smaller stimulated emission cross section caused by

its wide fluorescence spectrum. Since the minimum pump rate to reach the threshold

is basically proportional to 1/στ, the Ar2* excimer needs a higher pumping rate than

the other lasers listed above. It is very difficult to sustain a stable electrical discharge

during high-pressure Ar gas pumping with such a high power density as 100

MW/cm3 [1].b

Table 2.1 Comparison of spectroscopic parameters

Lifetime τ (ns)

Stimulated emission cross section σ (cm2 )

στ (s cm2)

KrF

F2

Ar2*

6.5

3.7

4.2

2 × 10−16

6.8 × 10−16

8 × 10−18

1.3 × 10−24

2.5 × 10−24

3.4 × 10−26

a Stimulated Emission: in a quantum mechanical system, the radiation emitted when the internal energy of the system drops from an excited level (induced by the presence of radiant energy at the same frequency) to a lower level. b It was explained in Appendix #A.


2.1 Rare gas kinetics considerations

A schematic diagram is provided in Fig 2.1. The primary electrons deposit their

energy through ionization

e + X → X+ + 2e k1 (2.1)

The resulting secondary electrons cool through successive ionization steps and

atomic excitation

e + X → X* + 2e k2 (2.2)

until their energy drops below the excitation threshold. Subsequent elastic collisions

rapidly cool the electrons to a few tenths of 1 eV, where continued cooling is retarded

by the Ramsauer minimum in the scattering cross section.

The electron-ion recombination in these systems occurs mainly through

dissociative recombination with the molecular ion after it is formed by three-body

association:

X+ + 2X→ X2+ + X k3 (2.3)

X2+ + e → X** + X k4 (2.4)


X (1S)

e

e

e

ee

e

e

e

e

e

X+ (1PJ)

ENERGYTRANSFER

ENERGYTRANSFER

X*

X**

X* (1,3PJ)2X

2X

X2**

X2*

2X

X2+ (2Σ)

X2* (1,3Σ)

X2 (1Σ)

hv

hv

Fig. 2.1 Schematic energy level diagram for rare gas excimers

This dissociative recombination predominantly populates a second group of

atomic exited states, X**[np5(n+1)p]. (The double asterisk indicates excitation above

the first excited levels.) The p levels rapidly relax via the reactions

X** + 2X→ X2** + X k5 (2.5)

X2** + (X) → X* + X + (X) k6 (2.6)

to the s levels that finally populate the excimer levels

X* + 2X → X2*(1,3Σg+ ) + X k7. (2.7)

In the absence of quenching by added gases, the excimer levels may decay through


radiation:

X2* → 2X + hv k8. (2.8)

The fact that there are two excimer levels, the singleta and tripletb, greatly

complicates the interpretation of the fluorescence decay. The two excimer levels are

nearly coincident in energy, with the triplet lying less than 1000 cm-1 lower, but their

radiative lifetime differ greatly, as shown in Table 2.2. in the early afterglow, the

density of secondary electrons (at high pressures) can be sufficient to cause rapid

collision mixing of the tow levels, leading to a decay time representative of a

statistical 3-to-1 distribution of triplets over singlets. This leads to an affective

radiative lifetime approximately four times that of the singlet state (in the case Ar2*,

more five times), as shown in Table 2.2. In additional, the decay of the late afterglow

when the electron density is low is dominated by the triplets [2].

e + X2 (3Σu+ ) ↔ e + X2*(1Σu

+ ) k8. (2.9)

On the discharge pumped Ar gas excitation will be examined further in the 6 chapter.

a A singlet state is a state of an atom or molecule with zero net electronic spin(S=0). For two electrons, a singlet state is one with antiparallel (paired) spins, and is denoted ↑↓. b A triplet stats of an atom or molecule is a state in which the total spin quantum number is S=1. For two electrons, a triplet state corresponds parallel electron spins and is denoted ↑↑.


Table 2.2 Radiative lifetimes of the excimer states

1Σu+ (nsec) 3Σu

+ (nsec) Statistical Average (nsec) Reference

Ar2*

Kr2*

Xe2*

4.2

3.3

4.8

2880

265

100

16.7

12.7

16.8

2.1

2.2

2.3

2.2 Discharge formation and stability in high pressure gas [3]

As well known in the gas discharge literature, there are two type of electrical

breakdown which can convert an initially nonconducting high-pressure gas between

two parallel electrodes into a highly conducting plasma upon the application of a

high-voltage pulse. One is the classical Townsend breakdown [4,5] and the other is

the plasma streamer breakdown [6]. Even though the basic process for electron

multiplication is due to electron avalanche in both types of breakdown, the

conditions for occurrence and applicable ranges of field strength to gas density ratio

E/n are quite different, so that it is important to make a distinction between the two.

The well-known Townsend or Paschen breakdown mechanism is characterized by a

large number of successive electron avalanches that originate from secondary

electron generation. The space-charge field caused by differential motions between

the electrons and the positive ions is assumed to be so weak as to be completely

negligible. Continuous exponentiation of the electron current within the discharge

gap is assumed to be maintained by a positive feedback of the Townsend avalanche

process through secondary electron emission at the cathode surface. A self-sustained


discharge condition, corresponding to the onset of such positive feedback, is

accordingly given by an equation of the type

( ) ],/)1log[(/ γγα +=pdp (2.10)

where p is the gas pressure, d is the electrode gap, α is the first Townsend coefficient,

which measures the exponentiation rate of free electrons per unit mean drift distance

of the electrons under the influence of the constant applied electric field strength E

under consideration, and γ is the second Townsend coefficient, which measures the

total probability of secondary electron emission from all sources associated with a

single primary electron emission. If positive ion bombardment at the cathode surface

were the main source of secondary emission, the minimum time required for the

positive feedback mechanism to become effective after turning on the applied E field

at t=0 would then be some fraction of the ion transit time from anode to cathode,

,/ ii ud=τ (2.11)

where ui denotes the mean drift velocity of the positive ions. For He+ ions in Heat 1

atm pressure, ui is 5×102 m/sec at a typical breakdown field strength of 4×105 V/cm

[7]. Thus, for a single transit across a 4 cm electrode gap, τi ~10-4 sec. For heavier

ions across large gaps, the transit time would be correspondingly longer. On the other

hand, if photoelectric effects at the cathode were an important source of secondary

emission, the minimum time for positive feedback would then be governed either by

the characteristic time for generation of the appropriate excited molecular states

during the avalanche process or by the radiative lifetime of the excited molecule,


whichever is longer. In any case, a relatively long formative time delay (~10-6 sec) is

generally observed in a Townsend-type breakdown which leads to sparking across

the gap.

Anode

Cathode

Anode

Cathode

++

++

++

+

+

+

+

(a) (b) (c)

Anode

Cathode

++

+

Fig. 2.2 Schematic diagrams showing (a) streamer development around a single primary electron avalanche after its space-charge field has grown beyond a certain critical value, (b) continuous backward propagation of the cathode-directed plasma streamer after the arrival of the primary avalanche head at the anode, and (c) complete bridging of the electrode gap by the plasma streamer

In contrast to the Townsend breakdown model, “Kanal” or streamer breakdown

occurs as the result of a large space-charge field that develops from a single electron

avalanche process into a rapidly propagating plasma streamer. This form of

breakdown will therefore allow the sparking phenomenon to begin anywhere inside

the discharge gap without relying on the secondary electron generation processes at

the cathode surface. The large space-charge fields that develop are due to the


relatively low mobility of the positive ions as compared to that of the electrons. On

the time scale of interest in a typical short duration pulsed discharge, the electrons

are free to move toward the anode while the ions are essentially frozen in space. For

simplicity, the propagating avalanche head filled mostly with free electrons can be

idealized as a negatively charged sphere, behind which is the positive space-charge

[see Fig. 2.2(a)]. The shape of the avalanche cone is determined primarily by electron

diffusion [5]. At some critical point where the space-charge field of the avalanche

head becomes comparable in magnitude to that of the applied electric field E,

streamer development begins. At this point, secondary avalanches are initiated by

photo-ionization in front of and behind the head of the primary avalanche. Both an

anode- and cathode-directed streamer develop and move at a mush greater velocity

than the velocity of the avalanche head. The increased velocity of the anode-directed

steamer is due to the space-charge enhanced electric field on the anode side of the

avalanche head. The cathode-directed streamer is primarily the result of the positive

space charge left behind the avalanche head. In the surrounding gas, photoelectrons

are produced which initiate secondary avalanches directed along strong field lines

toward the stem of the primary avalanche. The greatest multiplication of these

secondary avalanches occurs along the axis of the primary avalanche where the

space-charge filed supplements the applied field. As the negatively charged

avalanche head propagates toward the anode and exponentiates, it also leaves behind

a positively-charged tail which continues to lengthen and intensify at an accelerating

pace until the anode and cathode are eventually connected by the self-propagating

plasma streamer. In Fig. 2.2, three successive stages of such streamer development


are schematically illustrated. Thus, according to this model, breakdown will occur

whenever a single primary electron avalanche is allowed to develop to the critical

point of streamer initiation anywhere within the electron gap. A breakdown criterion,

attributed to Raether [8], which corresponds to the condition that the critical track

length ξc for a primary electron avalanche developed under the influence of a

suddenly applied constant electric filed is equal to the electrode gap d can

accordingly be derived, such that for air, in mks units,

( ) ).log(20/ dpdp +=α (2.12)

According to the early experiments of Townsend [4], the second ionization

coefficient γ is typically of the order of 0.1, so that the total avalanche gain αd

corresponding to the breakdown condition (2.10) is about 2.4, which is more than a

factor of 6 smaller than the value of αd corresponding to the breakdown condition

(2.12) for all electrode gaps of the order of a few centimeters or greater. Thus the

applied filed strength required for observation of a Townsend-type breakdown is

generally much weaker than that required for observation of a streamer breakdown.

On the other hand, for large-volume high-pressure discharges at high values of E/n

corresponding to those of general interest in high-power excimer laser excitation as

mentioned earlier, the total avalanche gain across the discharge gap often exceeds a

numerical value of 20 so that a streamer-type breakdown can be initiated from points

far away from the cathode due to the relative shortness of the critical avalanche track

length ξc in comparison with the electrode gap d. Furthermore, due to the nonlinear

buildup of the space-charge filed after the primary avalanche track length has grown


beyond ξc within the electrode gap, a streamer-type breakdown can take place in a

time scale much shorter than the characteristic time

,/ ee ud=τ (4)

for the single transit of a primary electron from cathode to anode at a constant drift

velocity ue. For a fee electron in pure He or in a predominantly He gas mixture at 1

atm pressure under the influence of a typical breakdown filed strength E=4×105 V/m

[3], ue is about 2×104 m/sec, so that τe ~ 2×10-6 sec for a 4 cm electrode gap. This

explains why streamer breakdown has been observed with formative delay times as

shot as 10-9 sec in some fast pulse discharge at high values of E/n.

For large scale homogeneous discharge it is necessary to pre-ionize the gas so as

to form a homogeneous electron density distribution, from which a large number of

electron avalanches can start simultaneously. The resulting overlapping electron

clouds and ion-cones form homogenous plasma (see Fig 2.3).


Fig. 2.3 The effect of a homogeneous pre-ionization electron density: the electron clouds and the ion-cones overlap, yielding a homogeneous charge carrier density and thus homogeneous discharge plasma

Several authors have discussed the theoretical criteria for a minimum pre-

ionization electron density to ensure the good discharge homogeneity, e.g. see [3, 9].

In the calculations of Levatter and Lin, ref. [3], an average minimum pre-ionization

electron density of approximately 108 cm-3 and with a long voltage rise time a lower

electron density can be allowed. In order to start a discharge, the pre-ionization


electron density must be multiplied to the discharge level of ~1014 cm-3 by the

electric filed.

2.3 General characteristics of the glow (or uniform) discharge

If a glow discharge is maintained between two electrodes in a long glass tube,

several bright and dark regions will be observed as shown in Fig. 2.4(a). The

variation of the relative intensity of these regions is shown Fig. 2.4(d). It can be seen

that (i) the negative glow region is the brightest region of the glow discharge; (ii) the

positive column constitutes a uniform and large bright region; and (iii) the anode and

the Faraday dark spaces are not completely non-luminous and they are usually

difficult to observe experimentally. The dark region that is clearly observed

experimentally is the layer between the cathode surface and the negative glow which

is called the cathode-fall region. This region is of interest due to the fact that most of

the voltage maintained between the electrodes is dropped across this region. The

distribution of the potential along the tube is shown in Fig. 2.4(c). The potential rises

rapidly from zero at the cathode and reaches a value of cathode voltage Vc at the

edge of the negative glow. From this point onwards, the potential stays

approximately constant, or rises very gradually. The rise becomes slightly sharper

near the anode.


Fig. 2.4 Characteristics of a typical glow discharge


The physical appearance of the glow discharge has been found to be affected

by system parameters such as the gas pressure; the electrode separation; the type of

gas used; the current that flows and the cathode material. Their effects are described

briefly as follows:

a) Effect of the gas pressure

The pressure of the gas in the discharge tube has a significant effect on the

relative length of the various regions of the glow discharge. Generally, the glow

discharge is operated at a pressure of less than 100 torr. At the high pressure and, the

positive column is favoured. The negative glow and the cathode-fall region are

compressed at high pressure while positive column extends to fill the space between

the electrodes. However its radial dimension also shrinks and it becomes not in

contact with the wall of the glass tube anymore. On the other hand, low pressure

operation of the glow discharge results in the shrinking of the positive column in the

axial direction while the cathode-fall region and the negative glow extend to fill the

tube. At very low pressure, probably in the region of milli-torr, the positive column

may disappear completely.

b) Effect of electrode separation

Once a glow discharge has been initiated, the voltage required to maintain it

will only increase slightly if the distance between the electrodes is increased. The

main effect of increasing the electrode separation is on the length of the positive

column – it increases to fill the extra space created by increasing the separation while

the other regions remain practically undisturbed.


c) Effect of the type of gas used

The general description of the appearance of the glow discharge is valid for any type

of gas used. However, the colours of the luminous regions may vary for different

type of gas used. The characteristic colours for some commonly used gases are listed

in Table 2.3 below [10];

Table 2.3 Characteristic colours of glow discharge plasmas

Gas Cathode layer Negative glow Positive column

Ne

Air

H2

Yellow

Pink

Brownish-red

Orange

Blue

Pale blue

Brick red

Red

Pink

O2 Red Yellowish-white Pale yellow with

pink center

N2

Ar

He

Pink

Pink

Red

Blue

Dark blue

Green

Red

Dark red

Red to violet


2.4 Excitation circuits

H.V.Cm

Charging bypass (R or L)

Swich

V0

t0

-V0

-2V0

H.V.Cm

Charging bypass (R or L)

CpSwich

V0

(a) LC Circuit or RC Circuit

(a) and (b) Discharge Waveform

(b) Charge Transfer type Circuit

Fig 2.5 Typical excitation circuits and discharge waveform for a TE laser: (a)LC circuit or RC circuit and (b) Charge transfer type circuit

Typical excitation circuits and discharge waveforms for a TE laser are shown

in Figs. 2.5 and 2.6. In the case of an excitation circuit a Fig. 2.5(a), DC high voltage

supply charges up a condenser through a charging element, normally a resistance or

an inductance. The charging bypass provides a path for the charging current. Once

the condenser is charged to the required voltage, the rapid closure of the high voltage

and high current switch enables the condenser to deliver its stored energy into the

laser load before glow-to-arc transition can occur. The charging bypass must offer

impedance that is many times more than that of the laser load lest this should eat up a

significant fraction of the energy stored in the condenser lowering, thereby, the plug


in efficiency of the laser. At the same time its impedance should be much less than

that of the charging element so that the current flowing through the conducting

switch from the source following a discharge can be kept low for a given repetition

rate. In the single shot operation, the condenser is normally charged resistively and a

spark gap is traditionally used as a switch. For repetitive operation, however, more

efficient charging by means of inductance is employed and a thyratron replaces the

spark gap. In the case of an excitation circuit a Fig. 2.5(b), the capacitance Cm is

charged to ~10 kV and on closing the switch which can be either a spark gap or

thyratron, the voltage on Cp and therefore between the discharge electrodes rises

rapidly and breakdown occurs in the laser cavity. It is important to minimize the

inductance of the loop compressing Cp and the laser cavity and hence Cp generally

consists of a large number of low inductance capacitors distributed along and close to

the discharge channel.

Figure 2.6 shows several excitation circuits and discharge waveforms for a

TE laser: (a) LC inversion circuit (b) pulse forming circuit and (c) Blumlein circuit.

A circuit (a) can be used to supply a more high voltage across the load. Circuit (b)

and (c) can be used to supply a fast-rising voltage across the load.


H.V. Swich

Pulse Forming Line(PFL)

t0

-V0

-2V0

t0

-V0

-2V0H.V.

Charging bypass

(L)

Swich

V0

C1

C2

H.V.

Swich Blumlein

t0

-V0

-2V0

(a) LC inversion Circuit (a) Discharge Waveform

(b) Pulse Forming Circuit (b) Discharge Waveform

(c) Blumlein Circuit (c) Discharge Waveform

Fig 2.6 Several excitation circuits and discharge waveforms for a TE laser: (a) LC inversion circuit (b) pulse forming circuit and (c) Blumlein circuit

2.5 The main discharge

During stable period of the discharge a glow discharge is formed. A general

property of a glow discharge is its constant voltage across the electrode almost

independent of the current flowing through the discharge. At this voltage there is


equilibrium between the production and the loss of electrons. If the discharge

becomes unstable during the current pulse, i.e. if this equilibrium is disturbed locally,

streamers and eventually arcs will start to glow. Arcs shorten the life time of the

electrodes, so to prevent electrode wear the system has been optimized to prevent

arcs.

L

Cp

Vp Vss

Discharge

i

Fig 2.7 Simple LC model for the discharge circuit during steady state

During the steady state phase of the discharge, the electrical circuit can be described

with a simple LC-circuit as shown in Fig. 2.7. This simple circuit can only be used

for the first half cycle of the current. The voltage across the electrodes is kept

constant at the steady state voltage Vss. The differential equation of this circuit is

02

2

=+dt

idLCpi

(2.5)

with the initial conditions


LVV

ii ssp −=′= )0(0)0( (2.6)

The solution of this differential equation and its starting conditions is

)sin()()(LCptVV

LCpti ssp −= (2.7)

In these equations Vp is the charging voltage. Cp and L are the capacitance of the

capacitor bank and inductance of the circuit respectively.


Reference [1] H. Tanaka1, A. Takahashi1, T. Okada1, M. Maeda1, K. uchino1, T. Nishisaka, A. Sumitani, and H. Mizoguchi, Production of laser-heated plasma in high-pressure Ar gas and emission characteristics of vacuum ultraviolet radiation from Ar2 excimers, Appl. Phys. B 74, 323 (2002) [2] E. Zamir, D. Huestis, H, Nakano, R. Hill, and D. Lorents, Visible absorption by electron-beam pumped rare gases, IEEE J. Quantum Electron. QE-15, 281 (1979) [3] I. Jeffrey, Levatter, and S. C. Lin, Necessary conditions for the homogeneous formation of pulsed avalanche discharges at high gas pressures, J. Appl. Phys. 51, 210 (1980) [4] J. S. Townsend, Electricity in Gases, Oxford: Oxford University Press (1915) [5] H. S. W. Massey, Electronic and Ionic Impact Phenomena. Volume II: Electron Collisions With Molecules and Photo-Ionization, 2nd ed. Oxford: Oxford University Press (1969) [6] J. A. Rees, Electrical Breakdown in Gases, London: Macmillan (1973) [7] A. V. Phelps and S. C. Brown, Positive Ions in the Afterglow of a Low Pressure Helium Discharge, Phys. Rev. 86, 102 (1952) [8] H. Reather, Arch. Electrotech. (Berlin) 34, 49 (1940) [9] A. J. Palmer, A physical model on the initiation of the atmospheric-pressure glow discharges, Appl. Phys. Lett. 25, 138 (1974) [10] S.C. Brown, Introduction to Electrical Discharge in Gases, U.S.A: John Wiley (1966)

Ch.3 The pre-ionization source for excimer laser 32

3. The pre-ionization source for excimer laser

It is well known that an efficient laser excitation in pulsed transversely excited

(TE) gas lasers occurs in the glow discharges mode. The formation of streamers in

self-sustained discharges and their subsequent degeneration into arcs can be avoided

by pre-ionizing the discharge volume. The quality of the pre-ionization in terms of

density and uniformity of the produced photoelectrons is important because it

directly affects on the laser performance characteristics. By improving the pre-

ionization one can extend the laser operating conditions and increase the input energy

that can be deposited into the active medium while still maintaining a homogeneous

discharge. A variety of pre-ionization methods are being used [1,2], generally based

upon the generation of UV photons by discharge which are fired in advance of the

main excitation discharges. Among these methods, one of the most effective method

relies on corona discharges over the dielectric surface as UV radiation sources [3,4].

Pre-ionizing schemes which utilize this technique have proven to originate very

durable and reliable devices [5,6], and provide a viable means for reducing size, cost,

and complexity of TE gas lasers.

A main feature of the corona-discharge pre-ionization schemes is the high pre-

ionization with relative uniformly, determined by a very homogeneous UV radiation

flux that can be produced along the main discharge electrodes during the entire stage

of development of the main discharge, with no jitter and very little delay. Therefore,

to realize higher stability and efficiency of TE gas laser, we have been designed a

new pre-ionization electrode (see Fig. 3.1) for the generation of UV radiation from a


surface-corona discharge over the surface of a dielectric.

3.1 Experimental set-up and methods

Figure 3.1 shows a schematic diagram of surface-corona electrode for the

pre-ionization. A surface-corona electrode was consisted with a stick-shaped inner

conductor of 9.5 mm in diameter as a high potential electrode, which inserted in the

cylinder-shaped ceramic (Al2O3) tube of 10 mm in inside diameter and 13.4 mm in

outside diameter, and a 1/4 cylinder-shaped copper board as a ground electrode.

Figure 3.2 show an experimental set-up and an equivalent circuit. In this circuit, DC

high voltage supply charges up a condenser through a charging element, normally a

resistance or an inductance. The charging bypass provides a path for the charging

current. Once the condenser is charged to the required voltage, the rapid closure of

the high voltage and high current switch enables the condenser to deliver its stored

energy into the surface-corona electrode.


Rod electrode(H.V.)

Earth electrode

Ceramic tube

Surface-corona discharge

Fig. 3.1 Schematic diagram of a surface-corona electrode for the pre-ionization

Fig. 3.2 Experimental set-up and an equivalent circuit

A delayed pulse generator was used to adjust the timing between the surface-

corona discharge and the triggering of the ICCD camera. It was set that the operation

gas was Ne, Ar and F in 1000 Torr, applied voltage of 10 kV, and exposure time in 50

ns for the ICCD (Image Intensifier Charge-Coupled Device) camera. Special

attention was paid to charge the corona electrode set position, which can be described


by the deviation angle against the center of the slit to the edge of the 1/4 cylinder-

shaped copper broad electrode, and makes a minus for the clockwise, and counter

clockwise is a plus. The corrector electrodes shown in Fig. 3.2 measured charged

particle densities against the deviation angle [7].

3.2 Experimental results and discussions

Many pictures were taken by the ICCD camera to investigate the discharge

development with different delay time. Figure 3.3 shows the spatial-temporal

evolution of corona-surface discharge in Ne gas. Figure 3.4 shows the spatial-

temporal evolution of corona-surface discharge: (a) Ne gas and (b) mixed gas (Ne:

97.8 %, Ar: 2 %, F2: 0.2 %) at atmospheric pressure respectively. It knows that

corona discharge starts from the edge of the ground electrode and develops on the

surface of ceramic tube. Because a minus high voltage was applied to the inner

electrode, it should presume a positive streamer discharge, which crawled on the

surface of the ceramic tube. Moreover, both timing of the occurrence and the

development of the corona discharge were about simultaneous due to the uniformity

of pre-discharge along the axis direction of ceramic tube. Considering Thyratron

jitters in 10-20 ns, it was found that the development speed of the surface-corona

discharge was 107-108 cm/s in pure Ne gas.


Fig. 3.3 Spatial-temporal evolution of corona-surface discharge in Ne gas

Earth electrode

Observation region using an ICCD camera

Earth electrode

Time

(a)

10 ns 60 ns 130 ns

10 mm

(b)

30 mm

H.V

Fig. 3.4 Spatial-temporal evolution of corona-surface discharge: (a) Ne gas and (b) mixed gas (Ne: 97.8 %, Ar: 2 %, F2: 0.2 %)


ICCDCamera

-xO

Ceramic tube

Earth electrode

-60 -50 -40 -30 -20 -10 0

4

6

8

10

12

14

16

18

20

22

24

Den

sity

of c

harg

ed p

artic

le (

107 /c

m3 )

Angle ( xO)

8kV 10kV 12kV

Fig. 3.5 Charged particle density against the deviation angle in Ne gas

Figure 3.5 shows the charged particle density against the deviation angle.

The results show that the highest density of charged particles is obtained at the -30

degree deviation angle and density of charged particles increases with increasing

applied voltage. Therefore, a sharp edge of the ground electrode is possible to radiate

the strong UV light.


Reference

[1] M. Richardson, K. Leopold, and A. Alcock, Large aperture CO2 laser discharges, IEEE J. Quantum Electron. QE-9, 934 (1973) [2] V. Hasson and H. M. von Bergmann, Ultraminiature high-power gas discharge lasers, Rev. Sci. Instrum. 50, 59 (1979) [3] G. J. Ernst and A. G. Boer, Construction and performance characteristics of a rapid discharge TEA CO2 laser, Opt. Commun. 27, 105 (1978) [4] V. Hasson and H. M. von Bergmann, Simple and compact photopreionization-stabilized excimer lasers, Rev. Sci. Instrum. 50, 1542 (1979) [5] R. Sze, and E. Seegmiller, Operating characteristics of a high repetition rate miniature rare-gas halide laser, IEEE J. Quantum Electron. QE-17, 81 (1981) [6] R. Marchetti, E. Penco, and G. Salvetti, Compact sealed TEA CO2lasers with corona-discharge preionization, IEEE J. Quantum Electron. QE-19, 1488 (1983) [7] K. Fukuda, N. Hayashi, S. Satoh and C. Yamabe, Estimation of Corona Preionization for Excimer Laser, Rep. Fac. Eng. Saga Univ., 29, 35 (2000) (in Japanese)

Ch.4 The API plasma electrode discharge system 39

4. The Automatically Pre-Ionized plasma electrode

discharge system

The questions surrounding discharge stability became particularly relevant with

the advent of high-pressure electrical-discharge lasers in the 1970s, and a

considerable number of studies have been conducted on this topic. The mechanisms

of spark formation have been established [1-3]. The main factors limiting maximum

energy deposition are the formation of cathode spots during field emission, and

explosive processes occurring at the cathode in the strong field of the cathode

potential fall. As a result of electron emission fluctuations from the cathode, a highly

conductive channel forms near one of the spots, and the current from the entire

cathode surface is drawn into this channel. A plasma electrode has been developed to

control such cathode spots [4,5]. The disadvantages of such plasma electrodes, as

reported in the literature [6], include the deflection of the main-discharge and the

need of an auxiliary circuit to form the surface discharges channel. Therefore, the

API (automatically pre-ionized) transversely excited plasma electrode was designed

to solve these problems. The structure of this API plasma electrode has been derived

from the “double discharge” electrode structure [7,8].

4.1 Experimental set-up

Figure 4.1 shows a schematic diagram of a cross-sectional view of the API

plasma electrode discharge system used in this work. Figure 4.2 shows the structure


of the API plasma electrode. The surface-discharge plasma that forms on the

dielectric surface provides intense ultra-violet radiation, suitable for pre-ionization in

the main-discharge space and able to sustain a surface conductivity high enough to

serve as an electrode.

Fig. 4.1 Schematic diagram of a cross-sectional view of the API plasma electrode discharge system


Fig. 4.2 Structure of the API plasma electrode

Therefore, compared with the conventional flat-plate electrode, the API

electrode can supply relatively high-energy in the main-discharge region and can

persist for a long time. The discharge electrode consists of a 2 mm thick ceramic

(Al2O3: purity 99%) tube whose almost half of outside is covered with aluminum

plates and is clogged inside with an aluminum rod. The surface discharge area of the

API plasma electrode is 2.2 cm × 74 cm. The discharge gap between electrodes is 0.3

cm. As shown in Fig. 4.3, the equivalent circuit was the conventional charge transfer

type [9,10], but for two other points, i.e. the dielectric capacitance was varied by

using a special electrode structure, and the operation was performed with and without

capacitance, Cpr, which controls the pre-ionization. The circuit parameters indicated

in Fig. 4.3 and Table. 4.1 were estimated from the measurements of temporal

changes of the current which passes from Cm to Cp with a Rogowski coil and the

voltage applied across the electrodes was measured with a high-voltage probe (P6015,

Tektronix). Figure 4.4 shows a schematic describing observation of the discharge

Capacitors for pre-ionization, Cpr (on anode)

Peaking capacitance, Cp Dielectric (Al2O3 99.0%)

Cathode

Anode

Capacitors for pre-ionization, Cpr (on cathode) Aluminum Rogowski Coil


formation properties of the main-discharge. An ICCD camera with high-speed

electronic shutter (C5909, Hamamatsu, Japan) was used to observe the discharge of

~10-ns order. To observe the discharge form, quartz was used as the window material,

but an optical resonator was not used. In order to measure the emission characteristic

from the Ar discharge, a grating monochromator (CT-25C, JASCO) and a

photomultiplier tube (R372, Hamamatsu, Japan) were used. The gas used for these

measurements was pure Ar gas at longer than one atmospheric pressure. The

experiments were performed at a repetition rate of 1 Hz.

Fig. 4.3 Equivalent electrical circuit of the API plasma electrode discharge system


Cathode

Anode

CHAMBER

Power Circuit

Powercontroller

ICCDCamera

ICCD Controller

Filter

DELAYGENERATOR

DG 535

Fig. 4.4 Schematic for the observation of the discharge formation characteristics

4.2 Circuit characteristics

The driving parameters of the API plasma electrode discharge system are

shown in table.1. Behaviors of the API plasma electrode discharge system and the

conventional charge transfer system are almost same as shown in Fig. 4.3. Figure 4.5

shows the typical signal waveforms using the automatically pre-ionized charge

transfer system with the auxiliary pin electrode. The main capacitance Cm is charged

by the required voltage V0. As shown in Fig. 4.3, during the S/W closed, the electric

energy Ep = (1/2) CpVbr2 is transferred from Cm to Cp and also, the electric energy Epr

= (1/2) [(Cpr Cdi)/ (Cpr+ Cdi)]Vbr2 transferred from Cm to Cpr at the same time, where

Vbr is a breakdown voltage and Cdi is the dielectric capacitance between an inside

aluminum rod electrode(covered with Al2O3) and an outside aluminum electrode.


Table 4.1 Experimental results obtained by the API plasma electrode discharge system shown in Fig. 4.6

Parameter

Vo

Vbr

S/W

Rs

Rch

Cm

Cp

Cpr

Cdi

Lm

Lp

: Initial store voltage 20 kV

: Breakdown voltage

: Spark gap switch

: Resistance of the spark-gap S/W 0.5 Ω

: Charging bypass resistance 2.5 kΩ

: Main capacitance for energy store 36 nF

: Peaking capacitance 29.4 nF

: Capacitance for pre-ionization and surface plasma 350 pF

: Capacitance of ceramic tube for pre-discharge

: Circuit inductance for the current flow from Cm to Cp ~100nH

: Self-inductance of the main-discharge ~10nH

Electrode separation 0.3 cm

The Cpr is used to control the effect of the capacitance of ceramic tube Cdi on

the discharge system of the API plasma electrode and also for prevention from the

electrical breakdown of ceramic tube. Before the main-discharge is formed, the

electric energy Epr produces surface-corona discharge on the cathode surface, and

then the auxiliary plasma channel is generated. Consequently, the electric energy Ep

forms the main-discharge in the discharge space by the auxiliary plasma channel.


Initial discharge voltage VbrCm Cp Charging time

0Voltage

Time

Switch On

Pre-ionized discharge

Current

Fig. 4.5 Typical signal waveforms using the automatically pre-ionized charge transfer system with the auxiliary pin electrode (see Fig. 1.4)

4.3 Experimental results and discussions

4.3.1 Discharge characteristic of the API plasma electrode

Figure 4.6 shows that the variation of the main-discharge with Ar gas

pressure in the discharge system of the API plasma electrode. The main-discharge

volume is increased with increasing Ar gas pressure, and the form of the main-

discharge volume is distributed very uniformly but it is determined by the extent of

the surface-corona discharge.


1 atm 2 atm 3 atm 4 atm 5 atmWindowDielectricMetal

Main-discharge Fig. 4.6 Variation of the main-discharge with Ar gas pressure in the discharge system of the API plasma electrode (at the formed Vbr, V0 = 20 kV)

The latter is limited by the value of the voltage required for ignition of the

surface-corona discharge. The breakdown voltage Vbr increased with increasing Ar

gas pressure as shown in table. 4.2. It can be considered that the impedance of

discharge space increases with increasing Ar gas pressure. In addition, the increase of

impedance leads to the increase of the discharge power by improvement of an

impedance matching as shown the variation of the discharge power with Ar gas

pressure in Fig. 4.8. The peak power put into the main discharge is about 90 MW.

Figure 4.7 shows the variation of the discharge voltage and current at 5 atm Ar gas.

The oscillation of discharge current (dotted line circle) is observed before the main-

discharge is formed. This oscillation seems to be due to the surface-corona discharge

that is the pre-ionization. In addition, the main-discharge forms according to the

density and distribution of electric charges created by the surface-corona discharge.

Unfortunately, we are unable to observe the longitudinal uniformity of the main-

discharge, because copper plates were used to connect parallel capacitors of API

plasma electrode discharge system shown in Fig. 4.2. However, longitudinal


uniformity of the surface-corona discharge for the pre-ionization refers to a chapter 3

mentioned previously. Therefore for the experimental results shown in Fig. 4.6, it is

considered that the uniform main-discharge volume is formed because the surface-

corona discharge is formed uniformly.

Table 4.2 Experimental results obtained by the API plasma electrode discharge system shown in Fig. 4.6

1atm 2atm 3atm 4atm 5atm

Main-discharge volume [cm3] 13.7 38.9 47.8 62.8 67.4 Breakdown voltage Vbr [kV] 14.8 15.3 17.5 18.2 20

-40

-30

-20

-10

0

10

0 100 200 300 400 500 600 700 800 900 1000time[ns]

Dis

char

ge V

olta

ge [k

V]

-5

0

5

10

15

20

Dis

char

ge C

urre

nt [k

A]

Fig. 4.7 Variation of the discharge voltage and current with time at 5atm Ar gas (experiment of Fig. 4.3)


Fig. 4.8 Variation of the discharge power with Ar gas pressure (experiment of Fig. 4.3)

4.3.2 Discharge characteristic of the API discharge system without Cpr

To observe the control effect of the surface-corona discharge (i.e. the pre-

ionization), the capacitance Cpr has been removed from the circuit the main-

discharge for different Ar pressure without Cpr is shown in Fig. 4.9. All experimental

conditions except the capacitance Cpr were same shown in Fig. 4.3. Comparing with

Fig. 4.6, the main-discharge volume in Fig. 4.9 deflected perceptibly to right side.

1 atm2 atm

3 atm

4 atm 5 atm


Fig. 4.9 Main-discharge of API discharge system without Cpr for different Ar pressure

-40

-30

-20

-10

0

10

0 100 200 300 400 500 600 700 800 900 1000time[ns]

Dis

char

ge V

olta

ge [k

V]

-5

0

5

10

15

20

Dis

char

ge C

urre

nt [k

A]

Fig. 4.10 Variation of discharge voltage and current in the main-discharge with Ar gas pressure at API discharge system without Cpr (at the formed Vbr, V0 = 20 kV)

In addition, the oscillation of discharge current (dotted line circle in Fig.

4.10) is not observed before the main-discharge is formed. The main-discharge

volume was increased and the breakdown voltage Vbr was decreased, in comparison

with above experimental results. These results are due to the surface-corona


discharge which is the pre-ionization was not formed uniformly. The applied voltage

on dielectric tube (capacitance Cdi) is approximately Vbr/2 when Cpr is operated, but

when there is no Cpr, the capacitor Cdi is applied Vbr. The cathode spots are formed in

the edge of metal electrode if the very high-voltage is applied at Cdi. As a result of

fluctuations of the electron emission from the cathode, a highly conductive channel is

formed near one of the spots, and the current from the entire cathode surface is

drawn into this channel. This phenomenon appears in Fig. 4.9. Consequently, when

compared with Fig. 4.8, the maximum discharge power in Fig. 4.11 was decreased

due to these cathode spots which is the main factor of limitation for the maximum

energy deposition. Therefore, according to the comparison of above two

experimental results, it is considered that the uniform pre-ionization forms the

uniform main-discharge by the control effect of Cpr, and the maximum energy

deposition is also increased.

Table 4.3 Experimental results obtained by API plasma electrode discharge system without Cpr shown in Fig. 4.9

1atm 2atm 3atm 4atm 5atm

Main-discharge volume [cm3] 39.6 41.6 73.7 81.5 92.6

Breakdown voltage Vbr [kV] 11.9 15.1 15.4 15.8 18.5


Fig. 4.11 Variation of the discharge power with Ar gas pressure (experiment of Fig. 9)

1 atm

2 atm3 atm

4 atm 5 atm


4.3.3 Using the plate electrode as an anode

In order to investigate the effect of different electrode configuration used

such as the plate electrode as an anode, we performed the experiment using a circuit

shown in Fig. 4.12. The electrode material of the anode was used copper and

experimental conditions except the discharge gap (0.6 cm) were same as these

denoted above. This experimental result was found that the main-discharge was

formed splitting by two as shown in Fig. 4.13, and main-discharge was transferred to

the arc-discharge after several ten ns. The discharge current could not be measured

due to the arc discharge. Therefore, the electrode structure of Fig. 4.13 is not suitable

for the formation of the uniform main-discharge. Although the shape of the discharge

electrode and the length of the discharge gap are different in this experiment, it is

inferred from above experimental results that the capacitance Cpr on the anode was

required as well as the cathode plasma for the formation of the uniform main-

discharge.

Fig. 4.12 Equivalent electrical circuit of discharge system used copper electrode on an anode


Fig. 4.13 Variation of the main-discharge with Ar gas pressure of discharge system used copper electrode as an anode. (at the formed Vbr, V0 = 20 kV)

A copper electrode

After 70 ns

Clear arc discharge

5 atm 1 atm (b)1 atm (a)


Reference [1] T. E. Broadbent, The breakdown mechanism of certain triggered spark gaps, Br. J. Appl. Phys. 8, 37 (1957) [2] P. W. Chan, R. J. Churchill, and M. S. Gautam, Radial recovery of high current spark channels, Int. J. Electronics. 32, 745 (1973) [3] I. Jeffrey, Levatter, and S. C. Lin, Necessary conditions for the homogeneous formation of pulsed avalanche discharges at high gas pressures, J. Appl. Phys. 51, 210 (1980) [4] K. Nakamura, N. Yukawa, T. Mochizuki, S. Horiguchi, and T. Nakaya, Optimization of the discharge characteristics of a laser device employing a plasma electrode, Appl. Phys. Lett. 49, 1493 (1986) [5] A. R. Sorokin and V. N. Ishchenko, High-power discharge with a plasma cathode in dense gases, Tech. Phys. 42, 1249 (1997) [6] V. Yu. Baranov, V. M. Borisov, A. M. Davidovskii and O. B. Khristoforov, Sov. J. Quantum Electron. 11, 42 (1981) [7] Yu Li Pan, A.T. Bernhardt and J. R. Simpson, Construction and Operation of a Double-Discharge TEA CO2 Laser, Rev. Sci. Instrum. 43, 662 (1972) [8] R. V. Bobcock, I. Lberman and W. D. Partlow, Volume ultraviolet preionization from bare sparks, IEEE J. Quantum Electron. 12, 29 (1976) [9] A. J. Andrews, A. J. Kearsley, C. E. Webb and S. C. Haydon, A KrF fast discharge laser in mixtures containing NF3, N2F4 or SF6, Opt. Commun. 20, 265 (1977) [10] R. C. Sze and P. B. Scott, 1/4-J discharge pumped KrF laser, Rev. Sci. Instrum. 49, 772 (1978)

Ch.5 The electric field of API plasma electrode system 55

5. The electric field of API plasma electrode system

In comparison with the results of the prior experiments with and without Cpr,

the main-discharge volume was increased and the breakdown voltage Vbr was

decreased for without Cpr. These changes are due to the surface-corona discharge not

forming uniformly. Considering these results in terms of the equivalent circuit, when

Cpr is operating, the voltage applied across the dielectric capacitor Cdi is

approximately Vbr/2, but without Cpr operating, the voltage applied across Cdi is Vbr.

If a very high-voltage is applied at Cdi, the cathode spots are formed at the edge of

the metal electrode. As a result of fluctuations in electron emission from the cathode,

a highly conductive channel is formed near one of these spots, and the current from

the entire cathode surface is drawn into this channel. This is the phenomenon that

appears in Fig. 4.9. Consequently, when compared with Fig. 4.8, the maximum

discharge power in Fig. 4.11 was decreased due to these cathode spots, which is the

main factor of the limitation for the maximum energy deposition. Therefore, from the

above two experimental results, it is considered that the reduction of discharge power

is caused by the non-uniform main-discharge. In addition, the non-uniform main-

discharge exists over an expanded region. This can be clarified by measurement of

the electric field distribution in the discharge region.

However, precise measurement of the electric field in the discharge region suffered

from technical limitations. As such, calculation of the potential distribution was

performed using the finite element method (FEM) in order to determine the electric


field variation. We used the general Galerkin finite-element software FlexPDETM [1],

which provides great flexibility in boundary geometry and formulation of equations.

Figure 5.1 shows the FEM modeling of the API plasma electrode.

Fig. 5.1 Modeling of the FEM for the API plasma electrode

In order to determine the time dependent of the electrical field in the API

plasma electrode, variable boundary conditions were obtained from actual

experimental data (the discharge voltage data at 5 atm) for the electrode with Cpr and

without Cpr, respectively. For example, variable boundary conditions in Fig. 5.1 are

applied to a aluminum Al_b, Al_t, Al_rod_b and Al_rod_t. In the case with Cpr, actual

experimental data (the time dependent breakdown voltage Vbr at 5 atm with Cpr:


sampling 2.0 GS/s) is applied to boundary condition Al_b and Al_rod_t is applied a

half of Vbr. Figures 5.2 ~ 5.7 show the electric field distribution (log scale) and

potential distribution lines at 50 ns ~ 350 ns after initial discharge (with Cpr and

without Cpr). Figures 5.3(a) and 5.6(a) show the calculated electric field distribution

(log scale) at 150 ns after initial discharge.


with cpr1: Cycle=1 Time= 50.000 dt= 50.000 p2 Nodes=15863 Cells=7656 RMS Err= 0.0323

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

a

a

a

a

m

m

n

o

o

op p

q

qq

q q

q

qq

r

r

r

r rrr

r

r r

r

r

s

s

ss

s

s

ss

ss

ss

ss

ss

t

t

tt t

u

uuu u

v

w

only log

max 6.9e+10E : 1.e+11D : 2.e+10C : 1.e+10B : 2.e+9A : 1.e+9z : 2.e+8y : 1.e+8x : 2.e+7w : 1.e+7v : 2.e+6u : 1.e+6t : 200000.s : 100000.r : 20000.q : 10000.p : 2000.o : 1000.n : 200.m : 100.l : 20.0k : 10.0j : 2.00i : 1.00h : 0.20g : 0.1f : 0.02e : 0.01

d : 0.002c : 0.001b : 2.e-4a : 1.e-4min 6.94e-5

(a) Electric field distribution

with cpr1: Cycle=1 Time= 50.000 dt= 50.000 p2 Nodes=15863 Cells=7656 RMS Err= 0.0323

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a a

aa

a

a

aa

b

b

c

c

d

d

e

e

f

f

g

g

h

h

i

i

j

j

k

kl m

nop q r

st

u v

v

max 6.56v : 6.30u : 6.00t : 5.70s : 5.40r : 5.10q : 4.80p : 4.50o : 4.20n : 3.90m : 3.60l : 3.30k : 3.00j : 2.70i : 2.40h : 2.10g : 1.80f : 1.50e : 1.20d : 0.90c : 0.60b : 0.30a : 0.00min 0.00

Scale = E3

(b) Potential distribution lines

Fig. 5.2 The electric field distribution (log scale) and potential distribution lines at 50 ns after initial discharge (with Cpr)


with cpr1: Cycle=3 Time= 150.00 dt= 50.000 p2 Nodes=64193 Cells=31384 RMS Err= 8.2e-5

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

a

a

a

aa

m

m

mm

mn

nn

o

p

p

p

p

p

q

qq q

qq

q

q

q

q q

q

q q

q

qq

r

r

r

r

r

r

r

s

s

s

ss t

t

t

t

t

t

tt

u

u

only log

max 3.3e+11E : 1.e+12D : 2.e+11C : 1.e+11B : 2.e+10A : 1.e+10z : 2.e+9y : 1.e+9x : 2.e+8w : 1.e+8v : 2.e+7u : 1.e+7t : 2.e+6s : 1.e+6r : 200000.q : 100000.p : 20000.o : 10000.n : 2000.m : 1000.l : 200.k : 100.j : 20.0i : 10.0h : 2.00g : 1.00f : 0.20e : 0.1

d : 0.02c : 0.01b : 0.002a : 0.001min 3.32e-4



X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

aa a

a a a

a

a

aa

a

a

a

b

c

c

d

d

e

e

f

fg

g

h

h i

i

jklm

n o

pq

v

max 1.63q : 1.60p : 1.50o : 1.40n : 1.30m : 1.20l : 1.10k : 1.00j : 0.90i : 0.80h : 0.70g : 0.60f : 0.50e : 0.40d : 0.30c : 0.20b : 0.10a : 0.00min 0.00

Scale = E4





X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

a

a a

m

mm

m

m

n

n

n

nn

n

o

o o

o

o

o

p

p

pp

q q q

qq

q

q

q q

q q

q

q

qq

q r

r

r r

r

r

r

s

s

s

s

s s

t t

t t

t

u

v

only log

max 2.5e+10E : 1.e+11D : 2.e+10C : 1.e+10B : 2.e+9A : 1.e+9z : 2.e+8y : 1.e+8x : 2.e+7w : 1.e+7v : 2.e+6u : 1.e+6t : 200000.s : 100000.r : 20000.q : 10000.p : 2000.o : 1000.n : 200.m : 100.l : 20.0k : 10.0j : 2.00i : 1.00h : 0.20g : 0.1f : 0.02e : 0.01

d : 0.002c : 0.001b : 2.e-4a : 1.e-4min 2.49e-5



X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

aa a

a

aaa

a

b

c

d

e

f g

hijk lmn

n

o

o

p

pq qr

r

s

st

u

uv

w

x

x

y

y

zz

z z

z

v

max 0.00z : 0.00y : -0.05x : -0.10w : -0.15v : -0.20u : -0.25t : -0.30s : -0.35r : -0.40q : -0.45p : -0.50o : -0.55n : -0.60m : -0.65l : -0.70k : -0.75j : -0.80i : -0.85h : -0.90g : -0.95f : -1.00e : -1.05d : -1.10c : -1.15b : -1.20a : -1.25min -1.25

Scale = E3




without cpr3: Cycle=1 Time= 50.000 dt= 50.000 p2 Nodes=22351 Cells=10888 RMS Err= 0.0222

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

aa

a

a a

a

cd

d

d

d

d

dd

d

e e

e

e

f

f

f

f

g

g

g

g

g

gg

g g

gh

h

h

h

h

h

h

hh

hh

h

i

i

ij

j

jj

jj

j

jk k

k

k

only log

max 1.1e+11u : 2.e+11t : 1.e+11s : 2.e+10r : 1.e+10q : 2.e+9p : 1.e+9o : 2.e+8n : 1.e+8m : 2.e+7l : 1.e+7k : 2.e+6j : 1.e+6i : 200000.h : 100000.g : 20000.f : 10000.e : 2000.d : 1000.c : 200.b : 100.a : 20.0min 11.3


without cpr3: Cycle=1 Time= 50.000 dt= 50.000 p2 Nodes=22351 Cells=10888 RMS Err= 0.0222

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

bcd e

fg

h

ij

k ln

n

n

n o

p

p

q

q

r

rs

s

t

t

u

u

v

v

w

w

x

x y

zA

A

v

max 6.60A : 6.50z : 6.00y : 5.50x : 5.00w : 4.50v : 4.00u : 3.50t : 3.00s : 2.50r : 2.00q : 1.50p : 1.00o : 0.50n : 0.00m : -0.50l : -1.00k : -1.50j : -2.00i : -2.50h : -3.00g : -3.50f : -4.00e : -4.50d : -5.00c : -5.50b : -6.00a : -6.50

min -6.59

Scale = E3


Fig. 5.5 The electric field distribution (log scale) and potential distribution lines at 50 ns after initial discharge (without Cpr)


without cpr3: Cycle=3 Time= 150.00 dt= 50.000 p2 Nodes=95573 Cells=47092 RMS Err= 7.5e-5

X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

aaa

a

a

a

a

aa

a

a

b

b

b

c

c

c

c

dd

d

e

e

ee

f

f

f

g gg

g g

g

g

gg

g

g

g

h h

h

h h

h

h

ii

ii

i

i i

j

j

j

k k

only log

max 6.6e+11u : 1.e+12t : 2.e+11s : 1.e+11r : 2.e+10q : 1.e+10p : 2.e+9o : 1.e+9n : 2.e+8m : 1.e+8l : 2.e+7k : 1.e+7j : 2.e+6i : 1.e+6h : 200000.g : 100000.f : 20000.e : 10000.d : 2000.c : 1000.b : 200.a : 100.min 44.6



X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

bc

d

e

fg

h

i

ii

i

i ii

i

i

j

j

k

k l

l

m

m

n

n

op q

q

v

max 1.63q : 1.60p : 1.40o : 1.20n : 1.00m : 0.80l : 0.60k : 0.40j : 0.20i : 0.00h : -0.20g : -0.40f : -0.60e : -0.80d : -1.00c : -1.20b : -1.40a : -1.60min -1.63

Scale = E4





X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

a

a

a

a

a

a

aaaa

a

c

c

c

d

ddd

d

ee

e

e

e

ee

f

f

f

f

ff

ggg

g g

g

gg

g

gg

gg

h

hh h

h

h

h

i

i

i

ii

i

j

j

k k

only log

max 5.e+10u : 1.e+11t : 2.e+10s : 1.e+10r : 2.e+9q : 1.e+9p : 2.e+8o : 1.e+8n : 2.e+7m : 1.e+7l : 2.e+6k : 1.e+6j : 200000.i : 100000.h : 20000.g : 10000.f : 2000.e : 1000.d : 200.c : 100.b : 20.0a : 10.0min 2.95



X e-2

-3. -2. -1. 0. 1. 2. 3.

Y

e-2

-3.

-2.

-1.

0.

1.

2.

3.

a

a

b

b

c

c

d

d

e

e

f

f

g

g

h

h

ij

j

k

k

l

l

m

m

m

m

nsu

xy

v

max 1.26y : 1.20x : 1.10w : 1.00v : 0.90u : 0.80t : 0.70s : 0.60r : 0.50q : 0.40p : 0.30o : 0.20n : 0.10m : 0.00l : -0.10k : -0.20j : -0.30i : -0.40h : -0.50g : -0.60f : -0.70e : -0.80d : -0.90c : -1.00b : -1.10a : -1.20min -1.25

Scale = E3




To compare the previous two experimental results (Fig. 4.6: 5 atm, with Cpr;

and Fig. 4.9: 5 atm, without Cpr) we evaluate the simulation results (Fig. 5.3(a) and

Fig. 5.6(a)). In the simulation results, the main discharge forms around the electric

field region given by the distribution line t for the electrode with Cpr (Fig. 5.3(a)) and

around the line j for the electrode without Cpr (Fig. 5.6(a)). Quantitatively, the

electric field intensities of the distribution line t in Fig. 5.3(a) and line j in Fig. 5.6(a)

the same, representing values above 2 × 106 V/m. The area enclosed by the electric

field distribution line j in Fig. 5.6(a) for the electrode without Cpr is more extensive

than that enclosed by the line t in Fig. 5.3(a) for the electrode with Cpr. This result

seems to indicate an increase in the discharge region (see Fig 5.8). Therefore, in the

case for the electrode without Cpr, it can be assumed that the probability of cathode

spots forming at the edge of metal electrode is increased. For this reason, the main-

discharge volume in Fig. 4.9 deflected to the right side and the maximum energy

deposition was limited in the API plasma electrode discharge system without Cpr.

However, for the electrode with Cpr, a uniform main-discharge was formed by

controlling the effect of capacitance Cpr.


(a) With Cpr

(b) Without Cpr

Fig. 5.8 Compare the electric field distribution (log scale) with the experimental result at 150 ns after initial discharge


Reference

[1] FlexPDETM, Finite element software, PDESolutions Inc., Web page: http://www.pdesolutions.com

Ch.6 Characteristics of discharge pumped Argon gas excitation 67

6. Characteristics of discharge pumped Argon gas excitation

Most of the reactions due to electron-atom, ion-electron collisions and the

formations of ion or excited states etc. in high-pressure argon plasma have been

reported (e.g. [1-14]). The principal kinetic processes between electrons and heavy

particles, and the key reactions among the heavy particles are tabulated in tables 6.1

and 6.2 respectively. The important absorption, spontaneous emission and radiation

processes are listed in table 6.3. All of the rate constants and cross sections are

obtained from literature [1-14]. Generally speaking, argon atoms, such as those in

discharge Xe2* dimer kinetics [14], are ionized by multi-step processes to provide

the electron source. The energy of the electrons is acquired from the applied electric

field for the discharged pumped case. When an e-beam is used for pumping, the

argon atoms can be also directly ionized by energetic electrons of the e-beam. The

electrons decay through two-body or three-body electron-ion recombination

processes (see reactions 18 and 19 in table 6.1). The dimers are formed by three-

body collisions of Ar* with 2Ar (see reactions 5 and 6 in table 6.2). In this way, the

kinetic scheme is simplified considerably, but it is still sufficient for a detailed

description of the discharge characteristics [2]. Cross sections of these two effective

electron levels are equal to the sum of the cross sections with similar thresholds.


Table 6.1 Electron and Ar gas reactions.

I 1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17

II 18 19

III 20 21 22 23 24 25 26 27

Electron collisions Ar + e → Ar* + e Ar + e → Ar** + e Ar + e → Ar+ + e + e Ar* + e → Ar** + e Ar** + e → Ar+ + e + e Ar* + e → Ar+ + e + e Ar2** + e → Ar2

+ + e + e Ar2*(3Σu) + e → Ar2

+ + e + e Ar2*(1Σu) + e → Ar2

+ + e + e Ar2*(1Σu) + e → Ar + Ar + e Ar2*(3Σu) + e → Ar + Ar + e Ar2*(1Σu) + e → Ar* + Ar + e Ar2*(3Σu) + e → Ar* + Ar + e Ar2*(1Σu) + e → Ar2** + e Ar2*(3Σu) + e → Ar2** + e Ar2*(3Σu) + e → Ar2*(1Σu) + e Ar2

+ + e → Ar+ + Ar + e

Electron-ion recombination Ar3

+ + e → Ar** + Ar + Ar Ar2

+ + e → Ar** + Ar

Electron ion formation and destruction Ar+ + e → Ar+ (4s) + e Ar+ + e → Ar+ (4p) + e Ar+ + e → Ar+ (4d) + e Ar+ (4s)+ e → Ar+ (4p) + e Ar3

+ + e → Ar** + Ar + Ar + e Ar + e → Ar+ (4s) + e + e Ar + e → Ar+ (4p) + e + e Ar + e → Ar+ (4d) + e + e


Table 6.2 Key reactions among heavy species

Reactions Forward rate constant cm3s-1 or cm6 s-1

I 1 2

II 3 4 5 6

7 8

III 9

10 IV

11 12 13 14 15 16 17 18 19 20 21

V 22 23 24

Ion formation Ar+ + 2Ar → Ar2

+ + Ar Ar2

+ + 2Ar → Ar3* + Ar , F(T)=(300/Tg)1.5 Excited stats kinetics

Ar** + 2Ar →Ar2** + Ar Ar2** + Ar → Ar* + 2Ar Ar* + 2Ar → Ar2*(3Σu) + Ar Ar* + 2Ar → Ar2*(1Σu) + Ar

Ar2*(3Σu) + Ar → Ar2*(1Σu) + Ar

Ar** + Ar → Ar* + Ar Vibration-Translational relaxation Ar2*(3Σu) + Ar → Ar2*(3Σu)v=0 + Ar Ar2*(1Σu) + Ar → Ar2*(1Σu)v=0 + Ar Penning ionization Ar2*(1Σu) + Ar2*(1Σu) → Ar2

+ + 2Ar + e Ar2*(1Σu) + Ar2*(3Σu) → Ar2

+ + 2Ar + e Ar2*(3Σu) + Ar2*(3Σu) → Ar2

+ + 2Ar + e Ar* + Ar* → Ar+ + Ar + e Ar** + Ar** → Ar+ + Ar + e Ar* + Ar* → Ar2

+ + 2Ar + e Ar2*(1Σu) + Ar* → Ar2

+ + 2Ar + e Ar2*(3Σu) + Ar* → Ar2

+ + 2Ar + e Ar* + Ar** → Ar+ + Ar + e Ar* + Ar2* → Ar+ + Ar + e Ar** + Ar2** → Ar+ + Ar + e Excited ion formation and destruction Ar+(4s) + 2Ar → Ar2

+ + Ar Ar+(4p) + 2Ar → Ar2

+ + Ar Ar+(4d) + 2Ar → Ar2

+ + Ar

2.5 × 10-31(300/Tg)1.5

7.0 × 10-31(300/Tg)1.5

2.5 × 10-32

1.0 × 10-11

1.0 × 10-32/f(Ne/NAr)

3.0 × 10-34/f(Ne/NAr) with f(Ne/NAr)=1+Yr×(Ne/NAr)

and Yr=6×104, Ref [15] 4.0 × 10-14

5.0 × 10-12

9.0 × 10-11

9.0 × 10-11

5.0 × 10-10

5.0 × 10-10

5.0 × 10-10

5.0 × 10-10

5.0 × 10-10

5.0 × 10-10

6.0 × 10-10

6.0 × 10-10

5.0 × 10-10

5.0 × 10-10

5.0 × 10-10

2.5 × 10-31

2.5 × 10-31

2.5 × 10-31


Table 6.3 Absorption, spontaneous emission and laser radiation. Photon1-4 are photons released when ionic Ar in various states transits to a lower state. Laser radiation at 126nm is represented as rad in reaction 4 Reactions Rate constant or cross section

1 Ar2*(1Σu) → 2Ar + hv (λ = 126 nm) 2 Ar2*(3Σu) → 2Ar + hv (λ = 126 nm) 3 Ar** → Ar* + hv1 4 Ar2*(1Σu) + rad → Ar + Ar + rad

5 Ar* → Ar + hv1

6 Ar2*(3Σu) + hv (λ = 126 nm) → Ar2+ + e

7 Ar+(4s) → Ar+ + Photon1 8 Ar+(4p) → Ar+ (4s) + Photon2

9 Ar+(4d) → Ar+ (4p) + Photon3 10 Ar+(4d) → Ar+ + Photon4

2.38 × 108 Lifetime 4.2ns 3.13 × 105 Lifetime 3.2µs

1.40 × 107

1.24 × 10-17/W(Tg) with W(Tg) = 1+ 0.27 (Tg/300) exp[-0.24(Tg/300)2] 5.00 × 104

5.00 × 10-19

2.64 × 109

1.19 × 108

2.30 × 108

3.60 × 109

6.1 Light emission characteristics of discharge pumped Ar gas

In order to clarify the light emission characteristic of the API discharge system of

Fig. 4.4 with uniform main-discharge in Ar gas, the light emission with the

wavelength 427.8 nm and 696.5 nm are measured [1]. To measure the emission

characteristic from the Ar gas discharge, a grating monochromator (CT-25C, JASCO)

and a photomultiplier tube (R372, Hamamatsu, Japan) were used. A He-Ne laser as

shown in Fig. 6.1 was used for the arrangement of the electrode and the calibration of

a monochromator. The gas used for these measurements was pure Ar gas. The

experiments were performed at a repetition rate of 1 Hz.


CathodeAnode

CHAMBER

Power Circuit

Powercontroller

PMT

Oscilloscope

MONOCHROMATOR

He - Ne Laser

H.V. probe

Mirror

Fig 6.1 Schematic for the observation of the light emission characteristics in Ar gas discharge

Figure 6.2 shows the diagram of energy levels of major transitions in Ar gas and

the reaction processes of Ar2* excimer band (around 126 nm). Here, the wavelength

427.8 nm is Ar ionic line and the wavelength 696.5 nm is an Ar atomic line. The

kinetic reactions between the radiation processes of these two wavelengths and the

radiation processes of Ar2* excimer band (around 126 nm) can be considered simply

as tables 6.1 ~ 6.3. It might be inferred from radiation deexcitation reactions

Ar2*(3Σu) → hv(VUV) + 2Ar (Lifetime 3.2µs) and Ar2*(1Σu) → hv(VUV) + Ar

(Lifetime 4.2ns), mixing process reactions Ar2*(3Σu) + e ↔ Ar2*(1Σu) + e and

Ar2*(1,3Σu) ↔ Ar2** + e that Ar2* (around 126 nm) emission intensity as a function

of time has more than at least two peaks in the case of discharge-pumped schemes.


Ar 1s0

Ar+

Ar**

Ar*

ARGON ATOM

ARGON MOLECULE

+e

16

14

12

10

0

Ar2+

Ar2** + e

+e

+ 2Ar

+ e

+ e

+ e

+ 2Ar

+ e , Ar

+ e , Ar*

+ 2Ar Ar2*

+ e

ENER

GY[

eV]

+ e

+ e

+ e

45

37

34

+ e

Pumping

+e +e +e

Ar3+

427.8 nm

696.5 nm

126 nm

Ar++

Collisional relaxation or excitation with the particles indicated

Spontaneous dissociation

Radiative deexcitation

sec13.020.41 nu ±∑ +

sec3.02.33 µ±∑ +u

∑ +g

1

Fig 6.2 The diagram of energy levels of major transitions in Ar gas


Such experimental results are demonstrated in the literatures [1,15]. It is likely

that only the radiation deexcitation reaction Ar2*(1Σu) → hv(VUV) + Ar, leads to

Ar2* excimer laser action. However, the rate constant of Ar2*(3Σu) is similar to that

of Ar2*(1Σu) [16], the radiative lifetime of Ar2*(3Σu) is 3.2 µs which is much longer

than the case of Ar2*(1Σu). Therefore the energy level of Ar2*(3Σu) is easy to become

saturated. In addition, this energy level reacts with VUV (vacuum ultraviolet) light as

the photo-ionization reaction Ar2*(3Σu) + hv (VUV) → Ar2+ + e. Excimer formation

reactions Ar* + 2Ar → Ar2*(3Σu) + Ar and Ar* + 2Ar → Ar2*(1Σu) + Ar are most

important processes for the realize Ar2* excimer laser action. It is important to study

the generation process of excited Ar* (Ar atom of first excited level is most

important). The excited Ar* atoms were formed by the lower excited energy level

about 11.6 eV. i.e. the reactions Ar + e → Ar*(4s) + e, Ar2+ + e → Ar* +Ar, Ar2* + e

→ Ar* + Ar + e, Ar2** + Ar → Ar* + 2Ar and Ar3+ + e → Ar* + 2Ar. However,

excited Ar* atoms which were formed by the reactions Ar2+ + e → Ar* +Ar, Ar2* + e

→ Ar* + Ar + e, Ar2** + Ar → Ar* + 2Ar and Ar3+ + e → Ar* + 2Ar are not

suitable for the formation and enhancement of lasering band Ar2*(1Σu) because these

Ar* atoms were formed late comparatively. Therefore, for the formation and

enhancement of lasering band Ar2*(1Σu), it is suitable that it is increased the electron

impact excitation Ar + e → Ar*(4s) + e by very fast pumping energy deposition in

the uniform main-discharge condition.

Figure 6.3 illustrates the variation of the emission intensities of the Ar ionic line

at 427.8 nm and Ar atomic line at 696.9 nm with Ar gas pressure obtained from the

main-discharge. The obtained result is quite similar to the experimental results of the


literature [1]. According to the literature, the waveform of the ionic line at 427.8 nm

was observed in the discharge and the waveform of the atomic line at 696.5 nm was

observed both in the discharge and in the decaying plasma.


0 1 2 3 4 5 6 7 8 9 10Time[us]

Dis

char

ge v

olta

ge [K

V]

696.5nm

0

0

0IN

TEN

SIT

Y(ar

b. u

nits

)IN

TEN

SIT

Y(ar

b. u

nits

)

4atm5atm

2atm 3atm

1atm

-20

-10

4atm5atm

2atm3atm

1atm

427.8nm

1atm2atm

3atm

4atm

5atm

1.8 1.9 2.0 2.1 2.2 2.3 2.4Time[us]

1 2 3 4 5 6 7 8 9 10Time[us]

Fig. 6.3 Variation of the emission intensities of the Ar ionic line at 427.8 nm and Ar atomic line at 696.9 nm in Ar gas (experiment of Fig. 4.3)


These results are in agreement with the experimental results of Fig. 6.3. That is,

in the case of 1 atm, the waveform of the ionic line at 427.8 nm has the FWHM (full

width at half maximum) of about 700 ns and an effective lifetime of about 3 µs, and

the waveform of the ionic line at 696.5 nm has the FWHM of about 850 ns and an

effective lifetime of about 6 µs. It can be considered that the waveform of the ionic

line at 696.5 nm shows the formation of the excited Ar* atom from higher excited

state atoms or ions, by a radiative transition or a recombination process. Figure 6.4

shows the dependence of the light emission intensities of the Ar ionic line at 427.8

nm and Ar atomic line at 696.9 nm with the Ar gas pressure. The intensities of the

line at 427.8 nm and the line at 696.9 nm increase with increasing Ar gas pressure.

The intensity of the line at 427.8 nm increases proportionally to the approximate

cubic of the Ar gas pressure, i.e. a curve fitting Y = 0.22X3, where Y is emission

intensity and X is the Ar gas pressure (atm). On the other hand, the line intensity at

696.9 nm shows saturation at pressure above 4 atm. It can be considered that Ar

particles are pumped uniformly by the uniform main-discharge. In addition, it is

considered that the electron collision processes is enhanced due to the increase of the

pumping energy in the uniform main-discharge as shown in Fig. 4.8 which is

obtained experimentally.


0

10

20

30

1 2 3 4 5Ar gas pressure [atm]

INTE

NSI

TY [a

rb.u

nits

]

λ=427.8nm λ=696.5nm

Fig. 6.4 Dependence of the emission intensities of the Ar ionic line at 427.8 nm and Ar atomic line at 696.9 nm with the Ar gas pressure, which is derived from Fig. 6.14. Solid line denotes the fitting curve of Y =0.22X3, where Y is intensity and X is the Ar gas pressure (atm)

Consequently, these experimental results suggest that the lasering band

Ar2*(1Σu) is enhanced due to the increase of the excited Ar* atom which is formed by

the electron collision reaction Ar + e→Ar* + e and the Ar2*(3Σu) is saturated by the

recombination process.


Reference [1] H. Ninomiya and K. Nakamura, Ar2* excimer emission from a pulsed electric discharge in pure Ar gas, Opt. Commun. 134, 521 (1997) [2] K. S. Gochelashvily, A. V. Demyanov, I. V. Kochetov and L. R. Yangurazova, Fluorescence model of noble gas dimers in pulsed self-sustained discharges, Opt. Commun. 91, 66 (1992) [3] S Neeser, T Kunz and H Langhoff, A kinetic model for the formation of Ar2 excimers, J. Phys. D: Appl. Phys. 30, 1489 (1997) [4] H.A. Koehler, L.J. Ferderber, D.L. Redhead, and P.J. Ebert, Stimulated VUV emission in high-pressure xenon excited by high-current relativistic electron beams, Appl. Phys. Lett. 21, 198 (1972) [5] E Elson and M Rokni, An investigation of the secondary electron kinetics and energy distribution in electron-beam-irradiated argon, J. Phys. D: Appl. Phys. 29, 716 (1996) [6] Takefumi Oka, Masuhiro Kogoma, Masashi Imamura, and Shigeyoshi Arai, Energy transfer of argon excited diatomic molecules, J. Chem. Phys. 70, 3384 (1979) [7] J. W. Keto and Chien-Yu Kuo, Cascade production of Ar(3p54p) following electron bombardment, J. Chem. Phys. 74, 6188 (1981) [8] D. L. Turner and D. C. Conway, Study of the 2Ar+Ar2

+=Ar+Ar3+ reaction, J.

Chem. Phys. 71, 1899 (1979) [9] R. S. F. Chang and D. W. Setser, Radiative lifetimes and two-body deactivation rate constants for Ar(3p5, 4p) and Ar(3p5,4p) states, J. Chem. Phys. 69, 3885 (1978) [10] R. Sauerbrey, The photoionization cross sections of the Rg*2[3Σu

+] excimer


states for Ne2*, Ar2*, and Kr2*, IEEE J. Quantum Electron. 23, 5 (1987) [11] P. Dubé, M. J. Kiik, and B. P. Stoicheff, Spectroscopic study of vibrational relaxation and cooling of rare-gas excimers formed in a direct current discharge with supersonic expansion, J. Chem. Phys. 103, 7708 (1995)

[12] Wei-cheng F. Liu and D. C. Conway, Ion–molecule reactions in Ar at 296, 195, and 77 °K, J. Chem. Phys. 62, 3070 (1975)

[13] J. M. Hammer and C. P. Wen, Measurements of Electron Impact Excitation Cross Sections of Laser States of Argon(II), J. Chem. Phys. 46, 1225 (1967)

[14] I. V. Kochetov and Dennis Lo, Kinetics of a self-sustained discharge-pumped Xe*2 laser at 172 nm, Opt. Commun. 113, 541 (1995) [15] K. Nakamura, Y. Ooguchi, N. Umegaki, T. Goto, T. Jisuno, T. Kitamura, M. Takasaki, and S. Horiquchi (private communication) [16] W. Sasaki. Rev. Laser. Eng. 13, 912 (1985) (in Japanese)

Appendix #A 80

7. Conclusions

A new pre-ionization electrode was designed using the surface-corona pre-

ionization method. It was investigated experimentally in terms of generated charge

density, electrical characteristics. These experimental results suggest that a sharp

edge of the ground electrode is possible to radiate the strongest UV light. Therefore,

the API plasma electrode discharge system has been designed. The uniform main-

discharge was formed and its volume and the breakdown voltage Vbr increased with

increasing Ar gas pressure. The instantaneous maximum discharge electric power

was 90 MW at 5 atm Ar gas and the maximum energy deposition was 1.4 MW/cm3.

It was demonstrated that the uniform pre-ionization formed the uniform main-

discharge by the control effect of Cpr and the maximum energy deposition was

increased. It was examined the time dependent main-discharge from two-

dimensional simulation of electric field distribution of the API plasma electrode

discharge system. In the case of the using plate electrode on an anode, this electrode

structure is not suitable for the formation of the uniform main-discharge due to the

arc discharge with Ar gas.

It was measured light emission intensity of the Ar ionic line at 427.8 nm and Ar

atomic line at 696.9 nm with the Ar gas pressure. The intensity of the line at 427.8

nm increases proportionally to the approximate cubic of the Ar gas pressure (Y =

0.22X3, where Y is intensity and X is the Ar gas pressure). On the other hand, the line

intensity at 696.9 nm shows saturation at pressure above 4 atm. These experimental

results suggest that the lasering band Ar2*(1Σu) is enhanced due to the increase of the

Appendix #A 81

excited Ar* atom which is formed by the electron collision reaction Ar + e→Ar* + e

and the Ar2*(3Σu) is saturated by the recombination process.

Although these results are not enough to the discharge-pumped Ar2* excimer

laser action, it seems quite probable if a new chamber designed (i.e. it is more

compact and is operated more high pressure than 15 atm). Therefore, we will make a

new chambera and also extend our research for different gas lasers such as Xe2*,

Kr2* or F2.

a It was designed in Appendix #B.

Appendix #A 82

Appendix #A Necessary conditions for the discharge-pumped Ar2* laser

Necessary conditions for the discharge-pumped Ar2* laser, first, the necessary E/P is

given by the simple expressions: According to the gas law (at 0°C , 1atm) V0 = 2.241ⅹ10-2 m3/mole , N = NA/Vo = 2.687ⅹ1019 /cm3 atm , where, NA: Avogadro constant. Electron mean free path λe = (σesN)-1 = 4.652ⅹ10-3 cm atm , here, Ar2

* stimulated emission cross section σes: 8ⅹ10-18 cm2 [Ref. 9]. E/P ≈ (16)V/λe = 3439.38 V/cm atm , where, 16 eV is the energy of a Ar atom ionization

Second, the necessary pumping power for the discharge-pumped Ar2

* laser is given by the below expressions: Ar2

* saturation intensity Isat

,/46102.4108

1037.2106256.6/ 29218

11534

CmMWsCm

ssJhvI useulSAT =×××××⋅×

== −−

−−

τσ

where, 1159

8

1037.210126

/1099.2/ −− ×=

××

== sm

smcv ulul λ ,

c : the speed of light λul : the wavelength of Ar2

* emission h : Planck constant

512 ±≅⋅⋅∆⋅ LkNseσ

where, L: laser medium length ∆N: population density k: coefficient with optical gain

thus,

Necessary pumping power =)(

512CmkL

ISAT±

×

Appendix #A 83

Fig. a.1 Necessary condition for discharge-pumped Ar2* laser with breakdown

voltage Vbr

0

50

100

150

200

250

300

350

400

30 50 70 90 110 130 150 170 190

Laser gain medium length L [cm]

Pum

ping

pow

er [M

W/c

m3 ]

0

50

100

150

200

250

300

350

400

30 50 70 90 110 130 150 170 190

Laser gain medium length L [cm]

Pum

ping

pow

er [M

W/c

m3 ]

Fig. b.1 Necessary pumping power for discharge-pumped Ar2* laser with laser gain

medium length L (k=8%)

Appendix #B 84

Appendix #B Design of a new chamber and electrode

FRONT Unit [mm]

Appendix #B 85

ELECTRODE FRONT

Appendix #B 86

TOP

Appendix #B 87

SIDE

Appendix #B 88

ELECTRODE SIDE

Appendix #B 89

WINDOW

Appendix #B 90

Stre

ss n

umer

ical

cal

cula

tion

Stre

ss n

umer

ical

cal

cula

tion

Appendix #B 91

Stre

ss n

umer

ical

cal

cula

tion

List of publications 92

List of publications

1. S. K. Hong, N. Hayashi, S. Ihara, S. Satoh and C. Yamabe “Formation properties of the main-discharge in pure Ar gas using the automatically pre-ionized plasma electrode” The IEEE Transactions on Plasma Science, Vol.33. Apr (2005). pp.324-325.

2. S. K. Hong, N. Hayashi, S. Ihara, S. Satoh and C. Yamabe “Main-discharge formation and light emission in pure Ar gas at multi-atmospheric pressure using the automatically pre-ionized plasma electrode” Vacuum, Vol.79. Jun (2005), pp. 25-36

3. S. K. Hong, N. Hayashi, S. Ihara, S. Satoh, C. Yamabe and S. B. Wee “The discharge electrode for Ar2* excimer laser using plasma cathode” Optics Communications (in press)

4. J. Morida, S.K. Hong, N. Hayashi, S. Ihara, S. Satoh and C. Yamabe “真空紫外エキシマレーザ励紀用放電システムの基礎特性” Journal of Applied Plasma Science, Vol.11. Dec (2003). pp.129-134 (in Japanese)

Electrical discharge plasma characteristics in pure Ar gas...

Documents

Transcript of Electrical discharge plasma characteristics in pure Ar gas...