Study on microwave plasma source based on parallel stripline resonator

A Dissertation for the Degree of Doctor of PhilosophyStudy on microwave plasma source based on parallel stripline resonator Department of Materials Science and Engineering Graduate schoo

A Dissertation for the Degree of Doctor of Philosophy

Study on microwave plasma source based on parallel stripline resonator

Department of Materials Science and Engineering

Graduate school Chungnam National University

By

Tran Thanh Hai

February 2013

Study on microwave plasma source based on parallel stripline resonator

Submitted to the Graduate School in Partial Fulfillment of the Requirements for the Degree

of Doctor of Philosophy

October 2012

Department of Materials Science and Engineering

Graduate School Chungnam National University

By

Tran Thanh Hai

To approve the submitted Dissertation

for the Degree of Doctor of Philosophy

on parallel stripline resonator

Committee

Dr Seung-Kyu AhnKorea Institute of energy research

Graduate School Chungnam National University

I would like to express my sincere gratitude to my advisor Dr Shine- JaeYou who guided and supported me during the good and bad days of my PhD.His way of thinking, like a physicist and not like an engineer, has made me look

at problems with a new eye I would also like to thank him for the critical review

of the publications and this dissertation

I would also like to thank Prof Jong-Ruyl Jeong, my advisor in ment of Materials Science and Engineering who helped me enter the graduateschool of Chungnam National University, gave me useful classes and introduced

Depart-me to Dr Shine-Jae You He has always been available and this dissertationwould not have been possible without his support, suggestions and enlighteningdiscussions

I am grateful to my Ph.D committee members for their valuable commentsand suggestions on my thesis

I would also like to thank Dr D.j.Seong, Dr J.H.Kim in the Plasma Lab

of KRISS for their support and discussions

My thanks also go to Dr Min Park, B.H Seo, K.H You, D.W Kim fortheir help in my life and my work Whenever I needed their support, they werepleased to help me

Finally but not least, I would like to thank my parents and my wife Theyhave been a great source of encouragement during these years

Korea, October 2012Tran Thanh Hai

이 논문의 주제는 마이크로웨이브 병렬 스트립라인 공진기를 이용한 마이크 로플라즈마 소스에 관한 것이다 높은 작동 주파수 (~800~MHz)에서의 공진을 이용

하여 플라즈마의 사이즈를 줄이며 기본적인 장치의 동작구조는 마이크로스트림 구조

의 소스 (J Kim, K Terashima, Applied Physics Letters 86 (2005) 191504)와 마이크로스

트렘 분할원형 공진기 (F Iza, J.A Hopwood, Plasma Sources Science and Technology

14 (2005) 397) 이다 각각의 소스의 장점을 활용하여, 본 논문에서 제안한 마이크로 플라즈마 소스는 다른 플라즈마 소스에 비하여 효과적으로 강한 전기장을 발생시켜 플라즈마를 만들다

높은 주파수에서 작동하하여 RF 용량 결합성 플라즈마 소스에 비하여 효과

적으로 전자들에 에너지를 주게 된다

마이크로플라즈마 소스의 경우 헬륨과 아르곤 가스로 대기압내에서 방전이

되며 1W 이하늬 200mW 정도에서 유지됭다 대기 방전의 경우 5W 정도에서 방전이

되며 1.4W 정도에서 플라즈마가 유지된다 본 마이크로플라즈마의 경우 낮은 압력조 건에서는 더 낮은 전압과 전력으로 구동이 된다

마이크로플라즈마 소스이 플라즈마 변수들은 광검출 스펙트로스코피를 이용

하여 측정이 되었다 플라즈마 온도의 경우 OH의 회전 스펙트로스코피 (Rotational spectroscopy) 를 이용하여 측정되었으며, 전자 온도의 경우 스타크 확산 (Stark broadening) 분석을 통하여 얻어졌다 전자 밀도의 경우 중성 헬륨의 29-level 의 충

돌성 방출 (Collisional-radiative, CR) 모델을 이용하여 얻어졌다 측정을 통하여 가스

온도, 전자 온도, 전자 밀도 각 각은 2.5W 조건에서 400~K, 1~eV, 3}$ 에 있음이 확인되었다 이를 통하여 본 논문에서 연구한 마이크로플라즈마 소스

$10^{14}cm^{-가 비열원 (Non-thermal)의 전자 온도$10^{14}cm^{-가 $10^{14}cm^{-가스 온도보다 월등히 높은 소스임이 확인 되었다

이러한 특성은 높은 효율, 작은 크기, 낮은 구동 전력 틍성을 가지는 대기압 플라즈마 소스로의 탁월함을 보여준다

Study on microwave plasma source based on

parallel stripline resonator

Tran Thanh Hai

Department of Materials Science and Engineering, Graduate

school of Chungnam National University

Daejeon, Korea

(Advised by Professor Jong-Ryul Jeong)

Microplasma sources, with advantages of low temperature plasmas andsmall-scale discharge, have received much attention for development and appli-cation in a variety of fields including bio-medical applications (treatment of livingtissues, tissue sterilization and blood coagulation), dental treatment, displays, ra-diation sources, micro-chemical analysis systems, gas analyzers, photo-detectors

1 A dissertation submitted to the committee of Graduate School, Chungnam National versity in partial fulfillment of the requirements for the degree of Doctor of Philosophy conferred

Uni-in February 2013

i

Microplasmas can be generated in a wide range or pressure from a few Torrs

up to a few atmospheres In normally, operation of microplasmas at atmosphericpressure is more favorite because its size can be reduced due to elimination ofmicro-pump However, some problems, such as the electrode erosion due to en-ergetic ion bombardment, the difficulty of sustaining a glow discharge in the air,the higher voltages required for gas breakdown, the arcing at high pressure lead

to a new set of challenges in the research field To deal with these problems,several schemes have been devised However, they do not always provide perfectproperties

This dissertation present a microplasma source based on a microwaveparallel stripline resonator (MPSR) The design for high frequency operation(840 MHz) using resonance phenomenon allows its size to minimize The basicdesign and performance of the device is a hybrid type of microwave discharge be-tween microtrip structure source (MSS, J Kim, K Terashima, Applied PhysicsLetters 86 (2005) 191504) and microstrip split-ring resonator (MSRR, F Iza, J.A.Hopwood, Plasma Sources Science and Technology 14 (2005) 397) Therefore,

by virtue of combination effect of each discharge source advantage, the MPSRcan concentrate a strong electric field of which direction is rather parallel to thestripline around the small gap, compared with other plasma sources

Since the device is operated at high frequencies, the power is more ciently coupled to the electrons in the plasma and better performance than in a

effi-RF capacitively coupled plasma source is achieved The increasing sheath ages and collisionless nature of the sheath as pressure is decreased

volt-MPSR can self-ignited plasmas in helium and argon gases as input powerless than 1W and maintained plasma with as little as 200mW at atmosphericpressure For air gas they are 5W for self-ignition and 1.4W for maintainingplasma This design can also be operated at low pressure with lower ignitedinput voltage and power The low-power requirement allows for air-cooled oper-ation and the possibility of driving the device with low-cost off-the-self electronicscurrently used in telecommunication applications

The parameters of a helium plasma at atmospheric pressure generated byMPSR were measured with optical emission spectroscopy method The plasmagas temperatures was estimated by OH emission rotational spectroscopy, theelectron density was obtained from Stark broadening analysis of the hydrogen

Balmer - β line, and the plasma electron temperature was determined by using29-level collisional-radiative (CR) model for neutral helium The results showthat plasma gas temperature, electron temperature, and electron density were inthe order of 400 K, 1 eV, and 1014cm−3as the power less than 2.5W, respectively.This device is a non-thermal discharge, the electron temperature is much higherthan gas temperature.

With the advantages of high efficient, compact size, low input power andcapability of operation at atmospheric pressure, the MPSR plasma source is asuited device for portable systems

iii

TABLE OF CONTENTS

1.1 Plasma applications 1

1.2 Plasma and breakdown voltage 2

1.2.1 What is the plasma? 2

1.2.2 Breakdown voltage 4

1.3 Plasma loss versus operating frequency 6

1.4 Motivation and objectives of the project 7

2 THEORY OF TRANSMISSION LINE 10 2.1 Wave propagation equations 10

2.1.1 Lossless line with special length 15

2.1.2 The low-loss line 18

2.2 Stripline 19

2.2.1 Parameters of stripline operation in TEM mode 20

2.2.2 Empirical equations in practice 20

3 MICROWAVE PARALLEL STRIPLINE RESONATOR 22 3.1 Basics Source design 22

3.2 Principle of operation 23

3.3 Discharge parameters 25

3.4 MPSR versus other microplasma sources 26

4 SOURCE DESIGN ANALYSIS 31

4.1 Conformal mapping 32

4.2 Schwarz-Christoffel transformation 33

4.3 Capacitance calculation 34

4.3.1 Mutual capacitance 35

4.3.2 Discharge gap capacitance 42

4.3.3 Strip to ground capacitance 45

4.3.4 Resonator capacitance and characteristic impedance 46

4.4 Resonant frequency 46

4.5 Input impedance of the device 48

4.6 Matching network 50

4.6.1 Quater-wavelength transformer 51

4.6.2 Direction coupling 52

4.6.3 FABRICATION OF THE DESIGN 52

5 EXPERIMENT SETUP 57 6 DIAGNOSTIC OF PLASMA WITH OPTICAL EMISSION SPEC-TROSCOPY 59 6.1 Introduction 59

6.2 Diagnostic methods 60

6.2.1 Line intensity ratio 60

6.2.2 Plasma gas temperature 60

6.2.3 Plasma electron density 61

6.2.4 Plasma electron temperature 63

6.3 Results and discussions 65

6.4 Biological applications of microplasmas 66

6.4.1 Applications conditions 66

6.4.2 MPSR plasma for biological application 68

7 DEVICE PERFORMANCE 75 7.1 Discharge gap voltage 75

7.2 Power efficiency 75

7.2.1 Discharge gap voltage 75

7.2.2 Power efficiency 76

v

7.3 Result and discussion 79

List of symbols

MSRR-MIP Microstrip Split-ring Resonator-Microwave Induced PlasmaMSS Mirostrip Structure

MPSR Microwave Parallel Stripline Resonator

CR Collisional Radiative model

AC Altinative current

DC Direct current

TEM Transverse ElectroMagnetic wave

TE Transverse Electric wave

TM Transverse Magnetic wave

CST Computer Simulation Technology company

RF Radio Frequency

FWHW Full Width Half Maximum

CCP Capacitive Coupled Plasma

ICP Inductively Coupled Plasma

CTLR Coaxial Transmission Line Resonator

UV-C Ultraviolet of wavelength range 280nm-200nm

DNA Deoxyribonucleic Acid

ICNIRP International commission on Non-Ionizing Radiation Protection

vii

LIST OF FIGURES

1.1 Comparison of the process limits resulting from vapor pressure andeconomic constraints for both vacuum and atmospheric pressureplasma processing 21.2 Schematic view of (a) a plasma and (b) a discharge 31.3 Paschen curves for DC breakdown in various gases 41.4 Variation of breakdown voltage with gap length in atmosphericair at different frequencies [1] 62.1 Equivalent circuit for a a) transmission line and b) lumped-elementcircuit model for this line 122.2 Voltage, current and impedance variation along a open-circuitedtransmission line 162.3 Voltage, current and impedance variation along a short-circuitedtransmission line 172.4 Strip line a) geometry b) electric and magnetic field lines in aplane normal to stripline and c) voltage pattern along a stripline

of λ/2 long 193.1 Construction of MPSR design 223.2 Electric field vectors in a plane perpendicular to the ground planealong two legs of the resonator at 1W of input power as dischargegap width 100µm and dielectric thickness 3mm Simulation usingcommercial CST microwave studio 233.3 The electric field of MPSR at the center of the discharge gap versusgap width, simulation with CST microwave studio, as h=2mm,d=6mm, dielectric material is taconic εr = 2.55 24

3.4 Two designs of MPSP with characteristic impedance Z0 = 20Ωwere self-ignited plasma in He and air gas at atmospheric pressure

as a) VRM S = 9.2V for gap width of 80µm and b) VRM S = 11.6Vfor gap width of 192 µm 253.5 Design of MPSR with characteristic impedance of 40Ω and dishargegap 140µm, The discharge was ignited at input voltage VRM S =6.4V 263.6 Design of MPSR with characteristic impedance of 40Ω and dishargegap 32µm,at atmospheric pressure, this device was self-ignitedplasma in He and air gas at input voltage VRM S = 4.5V and in airgas at VRM S = 16V 263.7 Reflection coefficient of two design type quarter-wavelength match-ing type at different gap size and characteristic impedance withoutpresent of plasma in discharge gap 273.8 Sample for simulation with CST softwave, the conductor is copperand dielectric is taconic (εr = 2.55), resonant frequency of twosamples was 793MHz a)microstrip structure and b) split-ringresonator 283.9 a)microstrip structure and b) split-ring resonator c)MPSR 294.1 The reflection coefficient (S11) of the MPSR of figure.3.1 calculatedfrom the CST microwave studio 314.2 Schwarz-Christoffel conformal mapping for a half-plane 334.3 Sample of electrical field distributions along to two legs of theresonator in a) side view, b) front view, and (c) the correspondingcapacitance network for the resonator operated in odd resonantmode 354.4 Mapping sequential for calculating Cmincluding inside gap capac-itance (a) original structure (b) intermediate structure, upperhalf-plane; (c) intermediate structure with normalize axis and (d)parallel plane capacitor 364.5 Mapping sequential for calculating capacitance between two strips(a) original structure (b) intermediate structure, upper half-plane; (c) intermediate structure with normalize axis and (d) par-allel plane capacitor 41

ix

4.6 Edge effect contribution 424.7 Magnified view of discharge gap with simple capacitance modelfor calculation 424.8 Mapping sequential for calculating C2 with (a) original structure.(b) parallel plane capacitor 434.9 Gap capacitance with edge effect 444.10 Schematic of strip to ground capacitance (a) original structure;(b) a half of the strip to ground capacitance that is mapped into(c) upper-half plane and (d) parallel plane capacitor 454.11 Capacitance of one leg of the resonator versus stripline havingsame size The black curve are the result of the calculation in thiswork The red curve are result of stripline from empirical model 474.12 Characteristic impedance of one leg of the resonator versus striplinehaving same size The black curve are the result of the calculation

in this work The red curve are result of stripline from empiricalmodel 484.13 The black curve show the effect caused by lengthening of thestripline to the resonant frequency and the red curve show theeffect cause by both lengthening of stripline and discharge gapcapacitance under changing of discharge gap size 494.14 The equivalent circuit of MPSR discharge source operated in anodd resonant mode (not include the matching part) 504.15 Resonator input impedance versus input power position along theresonator at various gap width in vacuum as qualify Q=150, Z0 =40Ω Results were obtained from equation.(4.47) 514.16 Net impedance of MPSR versus input power position along theresonator at various gap width in vacuum as Q = 150, Z0 = 40Ω,power input through a quarter-wave length transformer Resultswere obtained from equation.(4.47) 524.17 Input impedance of the resonator versus feeding position and gapwidth The design parameter Q increases from 100 to 500 in thedirection of the arrow, and the gap width increases from 0.1 to1mm Results were obtained from equation.(4.47) 53

4.18 Position of feeding point power on the resonator versus qualify factor at different gap size as characteristic impedance of 40 Ω,

Results were obtained from equation.(4.47) 54

4.19 Schematic of matching impedance for MPSR a) Quater-wavelength transformer and b) direct coupling 54

4.20 Detach of Components of MPSR device and assembly process for b) quarter-wave transformer and c) direct coupling source 55

4.21 Resonator fabrication process 55

4.22 Discharge gap fabrication process 56

5.1 A schematic diagram of the experiment setup of MPSR discharge and OES diagnostics 57

6.1 Simulation results of light emission intensity of OH A-X(0,0) at different rotational temperature as FWHM of Gaussian broaden-ing is a) 0.05nm and b) 0.3nm 62

6.2 Comparison of the measured spectrum with the synthetic OH spectra yielding a rotational temperature of a) 480K, as input power 0.72W and b) 520K as input power 2.25W 69

6.3 Gas temperature with various discharge powers 70

6.4 Measured Hβ lineshapes and fitting curves of the Voigtian distri-bution at various powers: (a) 0.45 W, (b) 0.72 W, (c) 1.12 W, (d) 2.25 W 71

6.5 Plasma electron density at various discharge powers 72

6.6 Results from CR model for the optically thin line ratios (I728/I706) and I492/I504 with various electron temperatures and electron den-sities 73

6.7 Electron temperature with various discharge powers 74

7.1 Voltage pattern along two branches of the resonator 75

7.2 The power coupled to the plasma in the resonator 77

7.3 The signal flow graph represent of a two-port network 78

7.4 Gap voltage versus input power of the resonator as infinite dis-charge gap impedance a) and 100µm disdis-charge gap width 79

xi

7.5 The dependence gap width of the resonator at various istic impedance as input power 1W of a) discharge gap impedanceand b) characteristic impedance 807.6 The dependence gap width of the resonator at various characteris-tic impedance as input power 1W of a) gap voltage and b) electricfield strength 817.7 Power efficiency of the resonator versus normalized plasma impedance 827.8 Power efficiency of MPSR versus gap width at various character-istic impedance 837.9 Quality factor versus impedance lines of the resonator 84

character-LIST OF TABLES

3.1 Comparing electric field strength in the discharge gap of MPSRdesign to other designs that are also based on transmission line assupplied power 1W The results had simulated by CST microwavestudio 293.2 MPSR plasma source versus other microplasma sources 304.1 Parameters of MPSR designed sources at different gap size,characteristicimpedance and their matching efficient 546.1 MPSR plasma source parameters versus other microplasma sources 676.2 Effects of dose and dose rate to sample 687.1 Power coupled to the load of the resonator 77

xiii

includ-in figure.1.1 [16].

Together with large size plasmas, microplasmas have also received muchattention in recent years [17, 18, 19, 20] The combination from potential of lowtemperature plasmas with the advantages of being micro, the discharges create a

Figure 1.1: Comparison of the process limits resulting from vapor pressure andeconomic constraints for both vacuum and atmospheric pressure plasma process-ing

icals, and photons), and widely utilized in bio-medical applications (treatment

of living tissues, tissue sterilization and blood coagulation)[17, 21, 22], dentaltreatment [23], displays, radiation sources, micro-chemical analysis systems, gasanalyzers, photo-detectors Microplasmas can be generated in wide pressurerange from a few Torrs up to a few atmospheres However, operation at atmo-spheric pressure of microplasma sources is more preferred due to it eliminatesthe need of micro-pumps

1.2.1 What is the plasma?

Plasma is often called a forth state of matter that consists of positivelycharged ions with most or all of their detached electrons moving freely about(figure.1.2.a) In normally, plasmas are quasi- neutral

A plasma is created by applying energy to a gas in order to reorganize theelectronic structure of the species (atoms, molecules) and to produce excitedspecies and ions This energy can be thermal, or carried by either an electric

2

-Figure 1.2: Schematic view of (a) a plasma and (b) a discharge.

current or electromagnetic radiations A simple schematic for generating plasma

is shown in figure.1.2.b, two parallel plates (electrodes) are connected to a voltagesource By increasing the supplied voltage until the electric field strength betweentwo electrodes overcomes a threshold value, the breakdown occur and forming aplasma between two parallel plates

Plasma sheath

After the plasma was formed between two electrodes (walls), genetic electronsnear the walls, which thermal velocity is mush higher than ion velocity, willrapidly loss to walls, leading to a thin positively charged layers joined betweenwalls and plasmas, this charged layers are called sheaths The formation of asheath cause a potential with positive within plasma and zero near walls Thispotential will confine electrons in plasma and accelerate ions near sheaths towalls In most of cases, the voltage across the sheaths is much higher than voltageacross plasma Other word, most of potential voltage between two electrodes isfall on sheaths High sheath voltage will lead to high electrodes bombardmentthat is a big problem for most of plasma sources except a few special applicationsusing this property

1.2.2 Breakdown voltage

Breakdown voltage is dependent on not only ionization energy, electrodematerials and pressure but also driven frequency

Direct current breakdown voltage:

In direct current (DC) discharges, Townsend and Streamer are two lar theories to be used to explain for the breakdown phenomena under differentconditions[1] Pressure, temperature, electrode material and geometry are phys-ical conditions known to govern the ionization process of the gas and the onset

popu-of breakdown The Townsend theory agrees well with the experimentally served Paschen Law of gases which states that the voltage at which breakdownoccurs is a function of the product pressure (p) times electrode spacing (d) Anexample of the Paschen curve of DC breakdown voltages in different gases atatmospheric pressure is shown in figure.1.3 [1] The results show that the

Air Ar

Figure 1.3: Paschen curves for DC breakdown in various gases

breakdown voltage for each gas decreases as the product pd reduces from highvalue, reaches minimum at a value of pdmin of the order of unit Torr.cm, and then

4

-sharp increases in small value of pd The increasing of breakdown voltages inthe case product pd smaller than pdmin can be explained as there are not enoughcollisions happen before the electrons are lost to anode, while the higher range of

pd is that the electrons gain less energy in between collisions Therefore, to duce enough gas ionizations for maintaining the discharge, higher electric field isrequired Although the Townsend theory is in good agreement to most of exper-iment results special for breakdown at low pressure However, it can not explainthe filament appearing at some special values of pressure and electrode distancewhich can be adequately explained in the Streamer theory So, the combining ofboth above theories give us a comprehensive view of the breakdown phenomena

pro-AC breakdown voltage

The gas breakdown voltage of a gap in low frequency alternating electricfields (50Hz) is substantially the same as for DC fields as electrons and positiveions have enough time to cross the gap

High frequency breakdown voltage

At higher frequency, as positive ions have insufficient time to cross the gap

in half a cycle, the confining of ions inside gap leads to a positive space charge isbuilt up in the gap and cause field distortion, results a lowering in the breakdownvoltage from the DC value as observed in figure.1.4 The results show that thebreakdown voltages as driven frequency 300M Hz are less than half of that atdriven frequency 100M Hz, and are much smaller than AC and DC breakdownvoltage (figure.1.3)

If the frequency is so high that the amplitude of electron oscillations in thegap become comparing with the gap length, the electrons are now trapped in thegap without being lost in the electrodes In this case, cumulative ionization can

be produced in the gap by an electron traveling many times between electrodes,results for generating high density plasma

Figure 1.4: Variation of breakdown voltage with gap length in atmospheric air

at different frequencies [1]

There are three main energy loss processes in the plasma, they are sion energy loss per electron-ion pair created (ionization, excitation, and elasticscattering collision), electron and ion kinetic energy loss to discharge walls Thelosses of electrons and ions to walls due to drift by external electric field is muchhigher comparing to losses due to diffusion The drift flux of charge particles towalls is much dependent on the driven frequency

colli-• In DC and low driven frequency plasmas: the oscillation amplitudes

of electrons and ions are much large than the gap width, the ions andelectrons have significant time to drift to walls in half a cycle, leading tohigh energy loss of electrons and ions to walls

• At medium frequency: ions are confined between electrodes, in this caseions loss due to diffusion and electrons loss due to drift to walls

6

-• At very high frequency: electrons and ions confined between electrodes,

so the losses of ions and electrons to walls are only due to diffusion.Moreover, the calculation results by Lieberman M.A [2] for a RF capacitivelycoupled discharge show that the sheath voltage is inversely proportional to squarefrequency Thus, the increasing of driven frequency will reduce sheath voltage,leading to reduce the charge particles energy losses to the walls (bombardment

of ions to electrodes) Reduction ion bombardments to the electrodes lead toreducing the electrodes erosion, it help the plasma sources are more stable inoperation and the lifetime is longer

The advantages of high pressure plasmas compare to low pressure plasmashave been presented, however, the incumbrance of most plasma sources operating

at atmospheric pressure are the difficulty of sustaining a glow discharge in theair, the very high voltages required for gas breakdown as show in figure.1.3, theproblem of electrode erosion due to energetic ion bombardment [24, 25, 26], andthe arcing at high pressure lead to a new set of challenges in the research field Todeal with these problems, several schemes, such as dielectric-barrier discharges[27], plasma needle and plasma sources based on transmission line (stripline,microstripline, coaxial transmission line) have been devised [18, 19, 28, 29] Theplasma sources based on transmission line technology, which operated at veryhigh frequency, have many advantages compare to others, such as

• The microwave power precisely fed into the target area allowing the ation of high-density plasma with low loss at the rest of the device

gener-• The discharges can be sustained stably over a wide range of gas pressures

• Simple component for matching impedance,

• Low cost, easy to fabrication

• Low temperature plasma: with the gas temperature in the range (700 300K) and electron temperature in the range (2 - 1eV)

-• Wide range of applications: These plasma sources have been utilized insurface cleaning, activation, coating, biomedical application, enhance sur-face adhesion, etc.

However, these present plasma sources do not always provide perfect ties: it is hard to produce a large-scale plasma with high uniformity in the case ofthe splitring resonators [18], and the voltage across the gap was not sufficientlyhigh in the case of the microstrip structure devices [28] The need for a high effi-cient and flexible microplasma source In this paper, we propose a microplasmasystem based on microwave parallel stripline resonator (MPSR) The design forthe high frequency operation (840 MHz) using resonance phenomenon allows itssize to minimize The basic design and performance of the device are a hybridtype of microwave discharge between microwave stripline source (MSS) [20] andmicrowave split-ring resonator (MSRR)[30] Therefore, by virtue of combinationeffect of each discharge source advantage, the MPSR can concentrate a strongelectric field of which direction is rather parallel to the stripline around the smallgap, compared with other plasma sources The MPSR can be operated readily

proper-in low power level with high electron density because of its high strength tric field and low plasma loss rate stemming from the parallel direction of theelectric field to the stripline and the advantages of high frequency operation andtransmission line based With small size and low-power operation at atmosphericpressure, the MPSR plasma source is suitable for integration into microsystemsand portable device

elec-This dissertation is organized into 8 chapters Chapter 2 presents the basictheory about transmission line and stripline Chapter 3 presents the design ofMPSR, principle of operation and discharge parameters of the design at differentcharacteristic impedance and gap width The comparative study with the othermicrowave discharge sources is also performed in this chapter Chapter 4 presentsthe analysis of this design by using electrical circuit model to determining thekey parameters that decide the performance of the device Conformal mappingmethod was used to estimate the capacitance and characteristic impedance ofthe device Chapter 5 and 6 present the experiment set-up and helium plasmadiagnostics, respectively The plasma gas temperature, electron density and elec-tron temperature were measured by using optical emission spectroscopy method

8

-Based on the experiment results, the performance of the MPSR as a plasmasource is estimated in chapter 7 Finally, a summary is presented in chapter 8.The paper relate to this work

- T.H Tran, S.J You, M Park, J.H Kim, D.J Seong, Y.H Shin, and J.R.Jeong Atmospheric pressure microplasma source based on parallel stripline res-onator Current Applied Physics, 11, 2011 (published)

- S.J You, T.T Hai, M Park, D.W Kim, J.H Kim, D.J Seong, Y.H Shin,S.H Lee, G.Y Park,J.K Lee c, H.Y Chang Role of transverse magnetic field

in the capacitive discharge Thin Solid Films 519 (2011) 6981.6989

- T.H Trana,b, S.J You, J.H Kima, D.J Seonga, B.H.Seod, J.R Jeong.Measurement of the microplasma induced by microwave parallel stripline res-onator with optical emission spectroscopy (Submitted)

Chapter 2

THEORY OF TRANSMISSION LINE

In general, the phenomenon wave propagation along transmission line can

be acquired from a specialization of Maxwell’s equations If we assume that thetime-harmonic fields have the form

E(x, y, z) = [e(x, y) +bzez(x, y)]e−γz (2.1a)H(x, y, z) = [h(x, y) +bzhz(x, y)]eγz (2.1b)and waveguide region is source free, Maxwell’s equations can be written as [31]

where e(x, y) and h(x, y) are the transverse electric and magnetic field nents, respectively, and γ is is a complex propagation constant The conditionsfor special cases of wave mode (TEM, TE, and TM) propagation in the waveg-uides are

compo-• TEM mode: Ez = 0, Hz = 0

• TE mode: Ez = 0, Hz 6= 0

10

The result of (2.3b.a) shows that the transverse electric fields of a TEM wavesatisfy Laplace’s equation So, the transverse fields of a TEM wave are the same

as the static field By applying the boundary conditions to appropriate fieldcomponents, the electric and magnetic field distribution along the transmissionline can be obtained

However, the wave propagation along transmission line is normally in a smallspace and orientation Therefore, the solutions can be acquired with a simplecircuit model for voltage and current that represent for electric field and magneticfield, respectively, instead of solving the Maxwell equations But the conditionfor circuit analysis is voltages and currents phase that almost unchanged overits length or the dimensions of a network are much smaller than the electricwavelength, while voltages and currents along the stripline change in magnitudeand phase over its length, this can be solve if the circuit model is applied to

a small length on the line, as we can see on figure.2.4.a with considered length

∆z is much smaller than λ, and the equivalence circuit for this piece can beconsidered as 2.4.b [32],

where R, L, C, G are per unit length of resistance, inductance, capacitance,and conductance, respectively The resistance R results from finite conductivity

of the strip, and conductance G is due to loss in the dielectric between theconductors The solutions for current and voltage on the lump-circuit of figure.2.1can be found by applying Kirchhoff’s voltage laws and Kirchhoff’s current laws

to this circuit as following

v(z, t) = v(z + ∆z, t) + R∆z.i(z, t) + L∆z∂i(z, t)

i(z, t) = i(z + ∆z, t) + G∆z.v(z + ∆z, t) + C∆z∂v(z + ∆z, t)

∂t (2.4b)

+ _

After dividing equation.(2.4) by ∆z, taking limit as ∆z → 0, then transform

to frequency domain The two equations of 2.4 lead to the following form

d2V (z)

dz2 − γ2V (z) = 0 (2.5a)

d2I(z)

dz2 − γ2I(z) = 0 (2.5b)where

γ =pR + jωL)(G + jωC) = α + jβ (2.6)

is a complex propagation constant, the real α and imaginary β components arerepresent for attenuation and phase coefficients, respectively The solutions of(2.5) give the voltage and current at a point on the transmission line

V (z) = (V0+)e−γz+ (V0−)eγz (2.7a)I(z) = (I0+)e−γz+ (I0−)eγz (2.7b)

12

-V (z) and I(z) are the amplitude of voltage and current, -V0+,V0−,I0+,I0− are stants that can be determined by boundary conditions at two ends of the line.The equation.(2.7) shows that the wave propagate on the transmission line is thesuperposition of wave propagate in the positive direction (e−γz) and the negativedirection (eγz) of z The characteristic impedance of the line is defined as theratio amplitude of incident voltage to current, or minus ratio for correspondentreflected quantities (only applied for the line without load)

con-Z0 = V

+ 0

Transmission line with terminated load

When the stripline have a terminated load ZL place at z=0 as shown infigure.2.1, then the voltage reflection at load z=0 can be determined as followingexpression

Pav = 1

2Re[V (z)I(z)

∗

] = 12

|V+

0 |2

Z0 (1 − |Γ|

2) (2.11)

The present of reflection term (12Γ|V0 |

Z0 ) in equation (2.11) makes the power flow

to load is smaller than source capacity (12|V

+

0 | 2 Z0 ) The reflection phenomenon isnot only squandering power but also harmful to the power source Therefore,reduce the reflection wave is the prerequisite should be taken into account inoperation of the device It is noted that the equation(2.11) obtained with as-sumed that the source and transmission line are matched However, in normally,the internal source impedance is designed of 50Ω, and the transmission lines can

be designed at any value of characteristic impedance that depend on purposesand applications, so, the reflection coefficient at the power source is also takeninto account If we assume that the impedance at the input power of transmis-sion line (having a distance l=-z from load) including load is Z1, and the sourceimpedance is Zs then the reflection coefficient at source can be given as

a given operation frequency, therefore matching impedance between power sourceand transmission line can be done with the changing of characteristic impedance

or using an extra transmission line The wave length and phase velocity of thewave can be determined from the equation of voltage and current in time-domain,they are

v(z, t) = |V0+|.e−αzcos(ωt − βz + φ+

) + Γ.cos(ωt + βz + φ−) (2.14)I(z, t) = |V

+ 0

Z0 |.e−αzsin(ωt − βz + φ+) − Γ.sin(ωt + βz + φ−) (2.15)

14

-where φ+ and φ− are the incident and reflected phase angle, respectively Fromequation (2.14) and (2.15), the wavelength λ and phase velocity vp of wave prop-agation on the transmission line are

vp = ω

λ = vpT = 2π

2.1.1 Lossless line with special length

In normally, the energy losses come from conductor and dielectric is verysmall in a transmission line If the losses are neglected (R=0, G=0) then thecomplex propagation constant become

Half wavelength line

substitute l = λ/2 and β = 2πλ into (2.13) give

the input impedance is not independent on the line characteristic impedance, itvalue alway equal to load impedance

- Open circuit:

If the load impedance is very high (ZL → ∞) then the reflection coefficient

Γ = 1 (V0+ = V0−) Plot of voltage, current and impedance of the transmissionline in this cases is shown in figure 2.2 the voltages at two ends of the line are

Figure 2.2: Voltage, current and impedance variation along a open-circuitedtransmission line

maximum of magnitude and in 180 degrees out of phase This characteristic will

be applied in designing plasma source in next chapter

- Short circuit:

If the load impedance ZL → 0 Then the reflection coefficient Γ = −1(V0− = −V0+) Plot of voltage, current and impedance in the case of shortcircuited transmission line shown in figure.2.3

16

-Figure 2.3: Voltage, current and impedance variation along a short-circuitedtransmission line

the equation.(2.23) shows that the input impedance is dependent on the acteristic impedance of the quarter-wavelength line From this property, thequarter-wavelength line can be used as a impedance transformation circuit Aresistance load ZLcan be transformed into a impedance Zinat the input point ofthe line with a transmission line of characteristic impedance Z0 Therefore, thistransmission line can be used to match a resistance load ZL to a transmissionline with impedance Zs = Zin, if the characteristic of the quarter-wavelength is

char-Zλ/4=pZs.ZL, (2.24)

2.1.2 The low-loss line

It is known that all transmission lines have loss due to finite conductivity andlossy dielectric, but these losses are usually small In many practical problems,then, losses may be neglected, but sometimes the effect of loss may be interest.When the losses are small, the expressions of the complex propagation constantcan be approximated as



(2.26)

≈ 12

"

R

rC

L + G

rLC

#+ jω√

L + G

rLC

#

= 12

"rR

αc= Rs2Z0η

-characteristic impedance can be approximated as

con-of separation b, and region between the ground planes is filled with a dielectric

as shown in figure.2.4.a Since stripline has two conductors and a homogeneousdielectric, it can support a transverse electromagnetic (TEM) wave, and is a pre-ferred mode in operation In the other hand, it can also support the higher ordertransverse magnetic mode (TM) and transverse electric modes (TE) , but thesehigher modes can be suppressed with shorting screws between the ground planes

or restricting the ground plane spacing to less than λ/4 When the stripline

op-because the transverse component of electric field flux and potential of TEMmode satisfy the Laplace equation.

2.2.1 Parameters of stripline operation in TEM mode

-The phase velocity [32]

vp = √c

where εr is the relative permittivity, c is the velocity of light in vacuum it isseen in equation (2.32) that the speed of wave propagate along the stripline isslower than that in vacuum √

ε times Thus the wavelength is

C =

√LC

ca-2.2.2 Empirical equations in practice

Determining general and exactly solutions of stripline parameters is reallyhard So, the curve-fitting equation is often used to calculate the stripline pa-rameters, and the error in this method is less than 5%

20