以實驗的方式初探在有無磁控管系統的直流電漿源氣體放電情況

國 立 交 通 大 學

機 械 工 程 學 系

碩 士 論 文

Preliminary Experimental Investigation of a DC Gas

Discharge With and Without Magnetron System

以實驗的方式初探在有無磁控管系統的直流電漿源氣

體放電情況

研究生：梁偉豪

指導教授：吳宗信 博士

以實驗的方式初探在有無磁控管系統的直流電漿源

氣體放電情況

Preliminary Experimental Investigation of a DC Gas Discharge With and

Without Magnetron System

研 究 生：梁偉豪

Student：Wei-Hao Liang

指導教授：吳宗信博士

Advisor：Dr. Jong-Shinn Wu

國 立 交 通 大 學

機 械 工 程 學 系

碩 士 論 文

A Thesis

Submitted to Institute of Mechanical Engineering Collage of

Engineering

National Chiao Tung University

In Partial Fulfillment of the Requirements

for the degree of

Master of Science

In

Mechanical Engineering

July 2006

Hsinchu, Taiwan, Republic of China

致

謝

在交大求學的這兩年來，感謝吳宗信老師悉心教導，讓我無論是在學習過程 中或是生活上都成長許多，在研究及做學問方面，因為老師不厭其煩的教導，而 使我的成長許多也學到許多，處理事情以及解決問題變得更有效率，並且也充實 了我許多日常生活經驗，感謝老師這些日子以來的照顧。同時也感謝口試委員傅 武雄教授、魏大欽副教授在口試時提供的寶貴意見，使得本論文更加充實完備， 在此一併致謝。 實驗室的氣氛融洽，使得我在良好的學習環境學習，學習過程中更有效率。 感謝學長(姐)邵雲龍、李允民、周欣芸、洪捷粲、許哲維、鄭凱文在研究上及生 活上的教導及幫助，感謝同學育進、孟樺、百彥在學習上的相互切磋砥礪，以及 學弟們生活上的幫助與鼓勵，使我這兩年的研究生活非常充實且溫馨，並能夠順 利完成學業。 此外，感謝在交大這二年的生活中陪伴著我的各位好友以及我的家人，有你 們的鼓勵，使得我更能堅持下去。在這離別的季節裡，大家各奔前程，去追求自 己心中的夢想。希望大家在不久的將來，我們都能擁有屬於自己的一片天空。 梁偉豪 謹誌 九五年七月于風城

以實驗的方式初探在有無磁控管系統的直流電漿源

氣體放電情況

學生：梁偉豪

指導教授：吳宗信 博士

國立交通大學機械工程學系

摘

要

近幾十年來，電漿在半導體工業的運用越來越廣泛。其中直流電漿源可以用 簡單的機制使得氣體能夠輕易的產生輕微的解離而形成電漿，並且直流電漿源被 普遍的運用在濺鍍金屬材料來形成薄膜沉積方面已經有很好的例子。 如果就直流放電的電漿，其電漿產生所需達到的崩潰電壓根據帕申(Paschen) 曲線，當氣體解離所需要達到崩潰電壓一般都需要相當高的氣體壓力。但這在這 壓力下，離子或分子的平均自由徑卻相當的小，要使得被轟擊出來的原子能夠黏 著在基材上也就相當不容易。所以為了要使直流電漿源能夠操作在較低的壓力 下，一般比較常見的作法會在靶材附近增加磁場，利用磁力來限制從靶材轟擊出 來的二次電子，用以增加在靶材附近的電漿密度。 此研究最主要的目的就是要了解電漿參數的變化，所以我們會利用蘭牟耳探 針(Langmuir probe)深入電漿內部，來量測電漿內的電漿特性變化，並比較在有 無外加磁場的情況下電漿特性的改變。

Preliminary Experimental Investigation of a DC Gas Discharge With

and Without Magnetron System

Student：Wei-Hao

Liang

Advisor：Dr. Jong-Shinn Wu

Department of Mechanical Engineering

National Chiao Tung University

Abstract

In the recent years, the field of gas discharge plasma applications has

rapidly expanded in semiconductor industry. The dc discharges has been applied

in generating weakly ionized plasmas and has been studied the properties of

plasma. The dc glow discharges have been used as the sputtering source to

product thin film and other specialized application for a long time.

As illustration in the Paschen curve, the pressure that discharges exceed

breakdown voltage must be high enough (p>30mTorr) and maintain in the

usually manner by the secondary electron emission from the cathode. These

pressures are higher than the optimum for deposition of sputtered atoms onto the

substrate. This results in sputtered atom poor adhesion for the sputtered film. It

is desirable to operate a sputtering discharge at lower pressures than to be

magnetic field at cathode to confine the secondary electrons.

This research will use Langmuir probe to measure the properties of dc

discharges in the condition with and without magnetron and compare the

INDEX

致 謝 ... I 摘 要 ...II ABSTRACT ... III INDEX ... V LIST OF TABLE ...VII LIST OF FIGURES... VIII

CHAPTER 1 INTRODUCTION...1

1.1BACKGROUND AND MOTIVATION...1

1.2PLASMA SOURCES AND DIAGNOSTICS...3

1.3LITERATURE SURVEY...5

1.3.1 The Development of Diagnostic Theory...5

1.3.2 Applications of DC Sputtering System ...7

1.4OBJECTIVES AND ORGANIZATION OF THE THESIS...9

CHAPTER 2 EXPERIMENTAL APPARATUS ... 11

2.1EXPERIMENTAL FACILITY... 11

2.1.1 Overview of the DC-magnetron sputtering system... 11

2.1.2 Pressure gauges ... 11 2.1.3 Plasma source ...12 2.1.4 Magnetron assembly ...13 2.1.5 DC Power Supply...13 2.2EXPERIMENTAL INSTRUMENTATION...13 2.2.1 Lamgmuir probe...13 2.2.2 Thermocouple...14

2.3EXPERIMENT METHODS AND STEPS...15

CHAPTER 3 LANGMUIR PROBE DIAGNOSTICS...16

3.1TYPICAL I-V...16

3.2THEORY OF LANGMUIR PROBE...18

3.2.1 Planar Probe with Collisionless Sheath...19

3.2.2 Cylinder probe with collisionless sheath...23

3.3THEORY OF ELECTRON ENERGY DISTRIBUTION FUNCTION (EEDF) ...25

3.4PARAMETER CALCULATION...28

4.1THE DISCHARGES WITH MAGNETRON...30

4.1.1 Plasma potential and floating potential ...31

4.1.2 Plasma density ...32

4.1.3 Electron temperature...34

4.2THE DISCHARGES WITHOUT MAGNETRON...35

4.2.1 Plasma potential and floating potential ...36

4.2.2 Plasma density ...37

4.2.3 Electron temperature...37

CHAPTER 5 CONCLUSIONS AND FUTURE WORK...39

5.1SUMMARY...39

5.2RECOMMENDATIONS FOR FUTURE WORK...40

List of Table

Table 1. The value of (Vp-Vf)/Te in different places in the chamber with working pressure 30 mtorr and operated with magnetron ...85 Table 2. The value of (Vp-Vf)/Te in different places in the chamber with working pressure 20

mtorr and operated with magnetron ...85 Table 3. The value of (Vp-Vf)/Te in different places in the chamber with working pressure 10

mtorr and operated with magnetron ...86 Table 4. The value of (Vp-Vf)/Te in different places in the chamber with working pressure 70

mtorr and operated without magnetron ...86 Table 5. The value of (Vp-Vf)/Te in different places in the chamber with working pressure 80

List of Figures

Figure 1.1. Breakdown voltages in various gases for the plane parallel electrode [Raizer, 1991]..44

Figure 1.2. Construction of a cylindrical probe ...44

Figure 2.1. The DC-magnetron Sputtering System ...45

Figure 2.2. The MKS self-tuning controller ...45

Figure 2.3. The Baratron capacitance manometer ...46

Figure 2.4. The Wide-Range Vacuum Gauge Controllers ...46

Figure 2.5. The ion gauge ...47

Figure 2.6. The thermocouple gauge...47

Figure 2.7. The plasma source ...48

Figure 2.8. The Magnetron System ...48

Figure 2.9. The LakeShore 421 Gaussmeter...49

Figure 2.10. The DC power control panel...49

Figure 2.11. Standard ESP analysis system with motorized Z-motion control ...50

Figure 2.12. RF-compensated probe ...50

Figure 2.13. Automatic Z-motion drivers ...51

Figure 2.14. The position of thermocouple in the chamber...51

Figure 2.15. Schematic of experimental chamber...52

Figure 3.1. Typical I-V curves for Langmuir probe...53

Figure 3.2. Qualitative behavior of the sheath and presheath in contact with a wall ...53

Figure 3.3. Ion orbital motion within the sheath of a cylindrical Langmuir probe ...54

Figure 4.1. The coordinate in the chamber where the origin is on the center of the target surface. (a) the coordinate is on the picture (b) the coordinate is on the schematic of experimental chamber ...55

Figure 4.2. The discharge was operated with the magnetron assembly in the pressure 10 mtorr.56 Figure 4.3. The horizontal distribution of plasma potential in the pressure 30 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...57

Figure 4.4. The horizontal distribution of plasma potential in the pressure 20 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...58

Figure 4.5. The horizontal distribution of plasma potential in the pressure 10 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...59

Figure 4.6. The horizontal distribution of floating potential in the pressure 10 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...60

Figure 4.7. The horizontal distribution of floating potential in the pressure 20 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...61

Figure 4.8. The horizontal distribution of floating potential in the pressure 30 mtorr and the distance between substrate and target is (a) 11 cm and (b) 8 cm. ...62

Figure 4.9. The vertical distribution of floating potential in the horizontal position x=-2, x=0 and x=2 and working pressure in (a) 30, (b) 20 and (c) 10 mtorr. ...64 Figure 4.10. The difference of plasma density difference between OML and EEDF; the working

pressure is (a) 30, (b) 20 and (c) 10 mtorr. ...65 Figure 4.11. The horizontal distribution of plasma density in the distance 11 cm between the

substrate and the target. The working pressure is (a) 10, (b) 20 and (c) 30 mtorr...67 Figure 4.12. The horizontal distribution of plasma density in the distance 8 cm between the

substrate and the target. The working pressure is (a) 10, (b) 20 and (c) 30 mtorr...68 Figure 4.13. The vertical distribution of plasma density in the horizontal position x=-2, x=0, x=-2

and x=4 and working pressure in (a) 10, (b) 20 and (c) 30 mtorr...70 Figure 4.14. The horizontal distribution of electron temperature in the distance 11 cm between

the substrate and the target. The working pressure is (a) 10, (b) 20 and (c) 30 mtorr...71 Figure 4.15. The horizontal distribution of electron temperature in the distance 8 cm between the substrate and the target. The working pressure is (a) 10, (b) 20 and (c) 30 mtorr...73 Figure 4.16. The difference of electron temperature between sheath theory and EEDF; the

working pressure is (a) 30 and (b) 20 mtorr...74 Figure 4.17. The vertical distribution of electron temperature in the horizontal position x=-2, x=0,

x=-2 and x=4 and working pressure in (a) 10, (b) 20 and (c) 30 mtorr. ...76 Figure 4.18. The changes of electron temperature in the different working pressure and in the

same vertical position z=4 ...77 Figure 4.19. The discharge was operated without the magnetron assembly in the pressure 90

mtorr ...78 Figure 4.20. The horizontal distribution of plasma potential in the distance between substrate

and target is 4 cm and working pressure (a) 70 mtorr and (b) 80 mtorr...79 Figure 4.21. The horizontal distribution of floating potential in the distance between substrate

and target is 4 cm and working pressure (a) 70 mtorr and (b) 80 mtorr...80 Figure 4.22. The horizontal distribution of plasma density in the distance between substrate and

target is 4 cm at working pressure (a) 70 mtorr and (b) 80 mtorr and (c) the distance between substrate and target is 5 cm at working pressure 70 mtorr. ...82 Figure 4.23. The horizontal distribution of electron temperature in the distance between

substrate and target is 4 cm at working pressure (a) 70 mtorr and (b) 80 mtorr and (c) the distance between substrate and target is 5 cm at working pressure 70 mtorr. ...84

Chapter 1 Introduction

1.1 Background and Motivation

Advances in microelectronics technology over the last two decades have

exceeded even the most optimistic exceptions. It depends on the application of

plasma, i.e. etching, deposition, sputtering, surface cleaning, ion implantation and

diamond film, etc. in semiconductor industry. We should know the circumstance of

plasma first and then we will know the design of the plasma source could meet our

needs or not. When talking to the development of plasma sources, it is quite important

to improve the diagnostics techniques and need the knowledge of mechanism to

generate plasma.

In the recent years, the field of gas discharge plasma applications has rapidly

expanded. This is due to the large chemical freedom offered by non-equilibrium

aspects of the plasma. This wide variety of chemical non-equilibrium condition is

possible, since (external condition) parameters can be modified, such as:

● The chemical input (working gas; this defines the different species in plasma:

electrons, atoms, molecules, ions, etc.)

● The pressure (ranging from 1 mTorr to atmospheric pressure; as mentioned

above, a higher pressure typically reduces the confinement and pushed the plasma

● The electromagnetic field structure ( typically externally imposed, but it can

also be modified by the plasma species; these electric and magnetic fields are used to

accelerate, heat, guide and compress the particles.)

● The discharge configuration (e.g. with or without electrodes; discharge

volume)

● The temporal behavior (e.g. pulsing the plasma)

In dc discharge, a few electrons are accelerated by the electric field in front of

the cathode; collide with the gas atoms and cause excitation and ionization. The

ionization collisions create new electrons and ions. The ions are accelerated by the

electric field toward the cathode, where they release new electrons by ion induced

secondary electron emission. The electrons give rise to new ionization collisions,

creating new ions and electrons. These processes of electron emission at the cathode

and ionization in the plasma make the glow discharge self-sustaining plasma.

The pressure is an important factor to reach the breakdown voltage in the dc

discharge and the minimum value of breakdown is relative to electrode distance and

pressure (Figure 1.1). These pressures are higher than optimum for deposition of

sputtered atoms onto the substrates due to scattering of sputtered atoms by argon

atoms. The sputtering efficiency of a dc discharge can be improved by a hollow

to the addition of a magnetic field. It is clearly desirable to operate a sputtering

discharge at higher current densities, lower voltages, and low pressures than can be

obtained in a conventional glow discharge. This has led to use of a dc magnetic

configuration. The permanent magnet placed at the back of the cathode target

generates magnetic field lines that enter and leave through the cathode plate. The

magnetic field could trap the electrons near the cathode in order to achieve the

intention had been mentioned above.

1.2 Plasma Sources and Diagnostics

There are various ways of transferring energy from fields to plasma discharges.

Utilizing different electron heating mechanism, it would influence the characteristics

of the plasma, i.e., capacitive discharges, inductive discharges, wave-heated

discharges and dc discharges. Capacitive discharges have been the most widely used

source for low-pressure materials processing. However, the limitation of capacitive

RF discharges and their magnetically enhanced variants have led to the development

of various low-pressure, high-density plasma discharges, i.e., inductively coupled

plasmas (ICP) and electron cyclotron resonance (ECR), etc. Besides preceding plasma

sources, a dc discharge has one obvious feature, its macroscopic time independence

both in application of weakly ionized plasma and in the studying the properties of the

plasma medium. DC discharge is also a fine source to begin plasma diagnostic

technique.

After introduction to the plasma sources briefly, we would be interesting in the

plasma properties. The purpose to diagnose is appreciating the basic plasma

properties and further we can control the fabrication of semiconductor devices to

detect the quality of the process of deposition and etching, etc. The basic plasma

properties that we are interesting are plasma density, electron temperature, electron

energy distribution, plasma potential, ion energy distribution and floating potential,

etc. How do we receive the above properties? Following will introduce some

diagnostics techniques shortly.

The diagnostics techniques could be generally divided into invasive and

non-invasive diagnostics techniques. The invasive diagnostics include microwave

diagnostics and optical methods. Each of above methods has different application

ways, i.e. interferometer, plasma induced emission spectroscopy, laser induced

fluorescence (LIF) and absorption spectroscopy, etc. All of these methods are

generally measure the entire averaged properties of the plasma. The invasive

diagnostics include Langmuir probe and emissive probe and the invasive diagnostics

would make perturbation in the plasma, it’s still a good studying tool. The Langmuir

probe specially can obtain more plasma properties than other diagnostic techniques

and its price also has a small advantage than others.

1.3 Literature Survey

1.3.1 The Development of Diagnostic Theory

One of the fundamental techniques (the first one, in fact) for measuring the

properties of plasmas is the use of electrostatic probes. This techniques was developed

by Langmuir and Mott-Smith [Langmuir and Mott-Smith, 1926] as early as 1924 and

consequently is sometimes called the method of Langmuir probes, basically an

electrostatic probe is merely a small metallic electrode, usually a wire, inserted into

plasma. The probe is attached to a power a supply capable of biasing it at various

voltages positive and negative relative to the plasma, and the current collected by the

probe then provides information about the condition in the plasma.

In 1950s, many scientists [Schneider, 1956; Boyd and Twiddy, 1959] proved that

the electron velocity distribution is not completely complying with Maxwellian

velocity distribution and not completely isotropic especially in low pressure, partially

1935] successfully proved, in 1930, any type of Electron Energy Distribution

Function (EEDF) could be directly received by the secondary differential of the I-V

curve of the probe. And EEDF could be employed directly to calculate the plasma

properties.

Johnson and Malter [Johnson and Malter, 1949, 1950] developed double probes

in 1949 and Yamamoto [Yamamoto and Okuda, 1956; Okuda and Yamamoto, 1960]

improved to triple probe in 1960. The triple probe measurement replaced the

sweeping bias method used in the single and double probe measurement. Both of

above development make the plasma diagnostics techniques more selectivity and

comprehensive applications.

There are lots of literatures about the Langmuir probe in the plasma diagnostic

techniques, and F. F. Chen [Chen, 1965] in 1965 gathered all of the literatures about

Langmuir probe together to supply the completely analyzing in all kind of modules

and theories. Except Chen, Swift, Schwan [Swift, 1971] and Schott [Schott, 1995]

supplied large contribution in the integrated of probe diagnostics techniques and

analysis methods.

Laframboise [Laframboise, 1966] especially do the discussion of the theory and

study to the spherical and cylindrical probe, and get the practical and important result.

the probe.

From 1924, Langmuir introduced the electrostatic probe to plasma diagnose, it

allowed people to understand the phenomenon in the plasma. Langmuir probe still

applies to the design and development of high density plasma source today.

1.3.2 Applications of DC Sputtering System

DC planar magnetron discharge are widely and maturely used for sputtering

deposition metallic thin films such as aluminum, zincoxide, gold, and various alloy

and have gained wide acceptance for high rate deposition of blanket metal films with

good film adhesion. A permanent, electromagnet, or rotating magnet assembly is often

used in the magnetron to confine energetic electrons. With the applied magnetic field,

the ionization efficiency is effectively increasing so that high deposition rate could be

achieved at a relatively low pressure.

In 1985, Rossnagel and Kaufman [Rossnagel and Kaufman, 1986] used

Langmuir probe to character the plasma properties of a magnetron sputtering system.

Langmuir probe presented to investigate more close to the cathode region. The

experiment presented the plasma sheath thicknesses were estimated from the potential

variations for the low-current discharge and compared the plasma potential, electron

and investigated a discharge in dual-side dc magnetron system in 2005, and he

explored the detail correlation between plasma density and discharge parameters. The

power input and gas pressure are two major parameters to tune the deposition process.

In Wu’s study it was shown that as these two parameters are varied, discharge

characteristic change correspondingly.

Pulsed-mode, non-continuous and continues are two plasma systems have widely

used for plasma-based processes. Pulsed dc magnetron sputtering technology has been

developed in the last 10 years as useful tool for deposition of high-quality dielectric

thon films. Typically, these sputtering sources are operated in the frequency range

from 10 to 100 kHz and the duty cycles from 50% to 90% with two operation

methods of the unipolar and bipolar modes. The characteristics of plasma during

pulses and between pulses are very important in plasma-based processes as well as

understanding plasma physics.

Seo [Seo et al., 2005] investigated the effect of the duty cycle of the cathode

pulse on the plasma parameters in the pulsed magnetron discharges of

constant-voltage mode, constant-power mode and constant-current mode. They

observed that as the duty cycle of the pulse is reduced, the average electron rapidly

increases irrespective of the operating mode, although the average electron density

operation with a short duty cycle can produce the high-temperature plasma that yields

improved films quality by achieving a high ionization rate of the sputtered atoms. The

time-resloved Langmuir probe technique has been developed to obtain the evolution

of the characteristics of the plasma versus time during high-voltage pulses and

between pulses. However, the secondary electron emission during high-voltage pulses

seriously interferes with plasma measurement. Qin and McTeer [Qin and Mcteer,

2005] used a time-delayed and time-resolved Langmuir probe measurement to

measure the evolution of the plasma densities versus time of a dc pulsed,

non-continuous plasma.

1.4 Objectives and Organization of the Thesis

The purpose of the present study is to use Langmuir probe to obtain the plasma

properties in the sputtering system. It will show the investigation of the plasma with

and without the magnetron assembly and apply in different theory ways, i.e. sheath

theory and electron energy distribution function method. And then we would construct

plasma diagnose the plasma properties of the system and to make the plasma

diagnostic becomes simple in the future.

The organization of the thesis would be stated as following. It would start from

introduction and introducing the experimental apparatus briefly. The analysis theory

obtained by the theories introduced in the following chapter. It would finally show the

Chapter 2 Experimental Apparatus

2.1 Experimental Facility

2.1.1 Overview of the DC-magnetron sputtering system

The figure 2.1 shows overview of the dc-magnetron sputtering system that

contains unit of mass flow controller, system control panel, valve controller, pressure

gauge, dc power control panel and the chamber.

2.1.2 Pressure gauges

1. MKS Pressure Controller and Capacitance Manometer

The MKS Type 651 instrument (figure2.2) is a self-tuning pressure controller for

throttle valves. The self-tuning feature of the Type 651 unit determines system

characteristics necessary for control. This feature takes into account time constants,

transfer functions of the valve and plumbing, valve gain, pump speed, and many other

important parameters when determining the system characteristics. The default

window display on the front panel shows the pressure readout and the valve position

(% open). The pressure readout can be displayed in units of Torr, mTorr, mbar, µbar,

Pascal, or kPa. Five reprogrammable set points are provided, each one having the

option of being setup for pressure or position control. Valve open, close, and stop

The instrument has a high-powered driver to operate MKS type throttle valves and

equip with Baratron transducers (Capacitance Manometer figure2.3) giving the unit a

control range from _{10 to 0.1 Torr.} -3

2. Thermocouple Gauge and Ion Gauge

The Model 934 Wide-Range Vacuum Gauge Controllers (figure2.4) provide

pressure measurements across a broad range of vacuum environments. It also provides

easy-to-use, intuitive front panel and large, bright, digital displays for ion gauge and

low-vacuum gauges. The Model 934 controller connected to a single Bayard-Alpert

ion gauge (figure 2.5) covers a pressure range that formerly required two instruments.

It allowed us to set gas factor in vacuum gauge controller to display corrected

readings if our system uses different gases. The gas factor is a multiplier that the

vacuum gauge controller applies to a reading before it shows the result on the digital

ion gauge display. The working range of the ion gauge is from _{2x10 Torr to 1} -10

Torr. The low-vacuum gauge is thermocouple gauge (figure 2.6) and the working

range of the thermocouple gauge is from _{10 Torr to 999 Torr.} -3

2.1.3 Plasma source

figure is the cathode region that we provide negative voltage source. And it is our

target in the centre of the picture that was made by aluminum.

2.1.4 Magnetron assembly

The figure 2.8 shows the top view and side view of the magnetron system. The

magnetic field is provided by a magnet assembly consisting of a diameter of 8.3 cm

iron plate on which is mounted a series of 1.1x0.6x2.6 cm magnets (see figure 2.8)

with the entire assembly mounted vertically under the target. The magnetic field was

measure by the Lakeshore 421 Gaussmeter (see figure 2.9); and the gaussmeter

indicated the maximum perpendicular magnetic field of 350G to the target surface at

the center and the maximum parallel magnetic field of 220G to the surface at 2.2cm

from the center.

2.1.5 DC Power Supply

The DC power supply may maximally provide 300 watts energy. The figure 2.10

shows the dc power control panel.

2.2 Experimental Instrumentation

2.2.1 Lamgmuir probe

(EPIU), a gas-cooled Radio-Frequency (RF) compensated electrostatic plasma probe

and all the required interconnection cabling, see Figure2.11.

The RF-compensated electrostatic plasma probe for fixed installations and

motorized Z-motion driver is shown in Figure 2.12. The tip can provide the potential

ramp start from -200V to 100V and resolution 0.025V minimum. This probe is

suitable for use with both DC and RF plasma; the maximum allowable temperature at

the probe tip is 2500_{C, the maximum allowable temperature at the mounting flange is}

700_{C. For high temperature plasma, cooling gas may be used to limit the temperature}

of the compensation components, which are mounted as close to the probe as

possible.

An automatic motor-driven Z-motion driver is an available as an option for the

probe system; see Figure 2.13.

A computer-controlled stepped motor, with maximum movement of 300mm,

powers the automatic Z-motion driver. The stepped motor is controlled by the PC via

the EPIU and the Linear Motion Driver. A lead screw drives the probe into the

chamber; vacuum sealing is provided by the flexible bellows unit.

2.2.2 Thermocouple

We could move the tip of thermocouple (see figure 2.14) to measure the

2.3 Experiment Methods and Steps

This research focuses on a spatial Langmuir probe survey of sputtering

deposition plasma and the response of this plasma to process condition. The DC

magnetron sputtering system employed in this investigation is depicted in figure 2.15.

The research is divided into with and without magnetron system conditions. Pressures

in the condition without magnetron system are set at 70 and 80 mTorr in the 99.99%

argon gas and the probe is scanned horizontally across the plasma at two different

cross-section in the constant distance between target and substrate and is also scanned

in the normal direction at the centre location above the target. Additionally, the

pressure in the condition with magnetron system are set at 10, 20 and 30 mTorr in the

Chapter 3 Langmuir Probe Diagnostics

A metal probe is one of the earliest and still the most useful tools for

diagnosing plasma. They are often used as plasma diagnostics because of their

apparent simplicity and they are constructed easily. They measure the electron

current which depends on their bias voltage with respect to the plasma potential.

Over a very range of situation, the details of the I-V curve characteristics can be

related to the plasma parameters. These probes, introduced by Langmuir and

analyzed in considerable by Mott-Smith and Langmuir are usually called

Langmuir probes.

3.1 Typical I-V

As the probe voltage is swept with respect to the plasma potential, electrons and

ions are either attracted or repelled and the net current collected by the probe changes.

From the interpretation of the I-V characteristic the electron distribution function,

electron temperature and plasma density can be obtained. A typical voltage-current

curve, in general, can be divided into three regions and two meaningful potential

value points [Chen, 1965; Schott, 1955; Hershkowitz, 1989; Chung et al., 1975]: 1.

VB<Vf: Ion saturation region. 2. Vf<VB<Vp: Transition region. 3. VB>Vp: Electron

saturation region. 4. VB=Vf: The floating potential. 5. VB=Vp: The plasma potential.

ion current and the magnitude of the ion saturation current is much smaller than the

electron current due to the much greater ion saturation mass.

1. Ion saturation region:

For the potential of probe is lower than floating potential VB<Vf, the current is

increasingly ion current tending to an ion saturation current and almost electrons are

repelled. It’s not all ions can be collected by the probe. According to Bohm criterion,

the velocity of ion should be accelerated over the Bohm velocity [Chen, 1965; Bohm,

1949; Lieberman and Lichtenberg, 1994] (

1/2 e B M KT u ⎟ ⎠ ⎞ ⎜ ⎝ ⎛

= ) so that the ion can be

collected by the probe.

2. Transition region:

For the probe potential is between plasma potential (Vp) and floating potential

(Vf), electrons are repelled according to the Boltzmann relation until at floating

potential.

3. Electron saturation region:

For increasing VB above the plasma potential, the current tends to saturate at the

electron saturation current. The probe potential is positive relative to the plasma

potential, electrons will be attracted and ions will be repelled. The electron sheath

form in the probe surface and the electron current is limited by the thermal motion of

raising potential in non-planar probe. The electron saturation current is always not a

constant and increase with the raising probe potential.

4. Floating potential:

The potential that the electron and ion currents cancel out and no net current flow

through the probe is called floating potential.

5. Plasma potential:

At the probe voltage VB=Vp, the probe is the same potential as the plasma and

draws mainly current flowing from the more mobile electron because of the electron

velocity is larger than ion velocity.

3.2 Theory of Langmuir Probe

The probe is immersing in the plasma though the structure is simple in Langmuir

probe. Its theory is quite complicated. In order to simplify the probe theory, we need

to do some assumptions in process of analyzing.

1. The plasma should be uniform and quasi-neutral in the condition without

probe.

2. Electrons and ions must obey Maxwellian velocity distribution.

3. Electron temperature is far larger than ion temperature.

4. The mean free path of electron and ion is far larger than any characteristic

5. Each charged particle hitting the surface of probe will be neutralized

completely and secondary electrons will not occur.

3.2.1 Planar Probe with Collisionless Sheath

Considering a flat plate probe with the physical probe area A>>s2_{, where s is the}

sheath thickness, such that the collecting area A is essentially independent of s.

A. Electron Current:

In thermal equilibrium and without perturbation in the plasma, the particle

current density is v N 4 1 J_r = (3.1)

where Jr is particle current density, N is particle number density, and

1/2

M 8KT v=_⎢⎣⎡ _⎥⎦⎤

π is

mean speed in thermal equilibrium.

Then the electron current Ie that the probe collects can be state as

e e e eN AV 4 1 I = (3.2)

WhenV_B ≥V_p, all electrons move into the sheath with thermal motion. The

property of plasma at the sheath edge is quasi-neutral, i.e.

∞ ∞

∞ ≈N ≡N

N_{e i}

whereN_e∞、N_i、_∞ N_∞ represent the density of electrons, ion and plasma with infinity

e e e sat e, eN AV 4 1 V A eN 4 1 I = _∞ = _∞ (3.3)

When VB<Vp, electrons are repelled by the probe. From the previous assumption,

electrons must obey Maxwellian velocity distribution function,

( )

_⎥ ⎦ ⎤ ⎢ ⎣ ⎡ ⎥⎦ ⎤ ⎢⎣ ⎡ = _∞ e 2 e 1/2 e e e 2KT mv exp m 2KT N v f

π . From Boltzmann relation, the electron density

becomes ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ = _∞ e p e e T V -V exp N N (3.4)

The electron current in the transition region and the ion saturation can be

represents as ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ = ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ = _∞ e p sat e, e e p e e T V -V I V A T V -V exp eN 4 1 I (3.5) B. Ion Current:

When VB<Vp, ions are attracted by the probe. Because ion temperature is low, ion

mass is large, the mean velocity of ion is far smaller than electrons in thermal

equilibrium. It’s more complicated than electrons in current analyzing.

The ion sheath will occur in the probe surface when the probe potential is lower

than plasma. The electron density would then decay to shield the electrons from the

wall. We will show that a transition layer or presheath must exist between the neutral

rise to an ion velocity at the plasma-sheath edge.

We define the zero of the potential V at x=0 and the ions have a velocity us

there.

Ion energy conservation (in collisionless) then gives

( )

Mu -eV

( )

x 2 1 x Mu 2 1 2 s 2 _{= (3.6)}

The continuity of ion flux (no ionization in the sheath) is

( ) ( )

is s i xu x n u

n = (3.7)

where nis is the ion density at the sheath edge. Solving for u(x) from equation (3.6)

and substituting in equation (3.7) we have

( )

₂ -1/2 s is i _Mu 2eV -1 n x n _⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ = (3.8)

The electron density is given by the Boltzmann relation can be stated as

( )

V( )x/Te

es e x n e

n = (3.9)

Setting N_es =N_is ≡N_sat the sheath edge and substituting N and _i N into _e

Poisson’s equation

( ) ( )

[

n x -n x

]

e dx V d i e 0 2 2 ε = (3.10) We obtain ( ) ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ = -1/2 2 s /T x V 0 s 2 2 Mu 2eV -1 -e en dx V d _e ε (3.11)

( )

∫

∫

_⎥⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ = ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ V 0 -1/2 2 s /T x V 0 s V 0 dx Mu 2eV -1 -e dx dV en dx dx dV dx d dx dV _e ε (3.12) ( ) ⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + = ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ e Mu -Mu 2eV -1 e Mu T -e T en dx dV 2 1 2 s -1/2 2 s 2 s e /T x V e 0 s 2 e ε (3.13)

It is apparent that the RHS of the equation (3.13) should be positive. We expect

this to be the problem only for small V; we expand the RHS of equation (3.13) to

second order in a Taylor series to obtain

0 Mu eV 2 1 -T V 2 1 2 s 2 e 2 ≥ (3.14) We see that B 1/2 e s u M eT u ⎟ ≡ ⎠ ⎞ ⎜ ⎝ ⎛ ≥ (3.15)

This result is known as Bohm sheath criterion [Langmuir, 1961] [Bohm, 1949]

[Lieberman and Lichtenberg, 1994]. There must be a finite electric field in presheath

region to give the ions the directed velocity uB.

From the Bohm sheath criterion, considering that the probe is biased sufficiently

negatively to collect only ion current which was collected by the probe is

B i sat

i, enAu

I = (3.16)

The potential drop across the presheath, which accelerates the ions to Bohm

velocity, is given by p 2 B eV Mu 2 1 ₌ (3.17)

2 T

V e

p = (3.18)

The ratio of the density at the sheath edge to that in the plasma is found from the

Boltzmann relation ∞ ∞ ≈ =n e 0.61n n -Vp/Te s (3.19)

Then we can obtain the ion current

1/2 e sat i, M eT A 0.61en I ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = _∞ (3.20)

When ion start to be repelled ( VB>Vp ), the ion current goes to zero rapidly and

it can be state as

(

)

⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ = i B p sat i, i KT V -V e exp I I (3.21)

3.2.2 Cylinder probe with collisionless sheath [Langmuir

Mott-Smith, 1926; Laframboise, 1966; Chung et al, 1975;

Lieberman and Lichtenberg, 1994]

In the previous section, we discuss the basic probe theory by the planar probe.

The planar probe would make too much perturbation in the plasma because of its

large assembling surface. In practically, we would choose the cylindrical type probe to

We consider first a thin wire probe for which the sheath thickness, s, is lager than

the probe radius, a. In the saturation condition, only a single species is collected. A

giving incoming particle in the attractive central force of the probe has initial velocity

components -ν_r and ν_φ in the radial and azimuthal at the edge of the sheath r=s.

At the probe radius r=a, the corresponding compounding are ' r

-ν and '

φ

ν . For a

collisionless sheath we require conservation of energy,

(

)

(

2 _'2

)

r ' B p 2 2 r m 2 1 V -V e m 2 1 φ φ ν ν ν ν + + = + (3.22)

and conservation of angular momentum

'

a

sν_φ = ν_φ (3.23)

where m is the mass of the attracted species, either electrons or ions. Then we obtain

m V -V 2e a s -1 2 p B 2 2 2 r 2 r ' ₊ ⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ + =ν ν_φ ν (3.24)

For ion to reach the probe, setting ν'2r =0

1/2 2 2 B p 2 r 0 _{s /a -1} /m V -V 2e ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ + = ν ν_φ (3.25)

The particle only reach the probe ifν_φ ≤ν_φ0.

The saturation current collected by the probe is found by integrating the radial flux

r s

n

- ν over the distribution function at the sheath edge, for those particles that reach

the probe:

(

)

φ ν ν

ν

ν

φ

ν

ν

ν

π

φ φ

d

,

f

d

sd

-e2

j

-eA

I

0 0 - r r 0 - r s

=

∫

_∞

∫

=

_(3.26)

where f is the normalized distribution function of electrons or ions. The distribution is

an isotropic Maxwellian, averaged over the third velocity coordinate, we have

(

)

(

)

⎥ ⎥ ⎦ ⎤ ⎢ ⎢ ⎣ ⎡ + = s 2 2 r s r 2eT m -exp eT 2 m , f _φ ν νφ π ν ν (3.27)

where Ts is the temperature of the collected species at the sheath edge. For large probe

voltage,V_b-V_p >> , we can simplify the evaluation of (equation 3.26) by assuming T_e

that 1 s a << (3.28) m V -V e _{p B} 2 r << ν (3.29) m eT_s 2 0 << φ ν (3.30)

Using equation (3.28) and (3.29) to evaluate equation (3.25), we obtain 1/2 B p 0 _m V -V 2e s a ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ = φ ν (3.31)

Using equation (3.27) and (3.31) in equation and condition equation (3.30), we

integral to find that

1/2 B p s

m

V

-V

2e

ad

2en

I

_⎟

⎟

⎠

⎞

⎜

⎜

⎝

⎛

=

_(3.32)

3.3 Theory of electron energy distribution function (EEDF) [Schott,

1955; Hershkowitz, 1989; Lieberman and Lichtenberg, 1994]

be written as

( )

v f d d d eA I _{z z e} - y - x e min ν ν ν ν ν

∫

∫

∫

∞∞ ∞ ∞ ∞ = (3.33) where

(

)

1/2 B p min _m V -V 2e ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ = ν

in the retarding potential region V_p-V_B > , is the minimum velocity for an electron 0

along z at the plasma-sheath edge to reach the probe. For an isotropic distribution we

can introduce spherical polar coordinates in velocity to obtain

( )

ν

θ

θν

φν

θ

ν

θ π ν e 2 2 0 0 e

eA

d

d

d

cos

sin

f

I

min min

∫

∫

∫

∞

=

_(3.34)

Where A is the physical collecting area of the probe and

ν ν

θ -1 min

min =cos The φ and θ are integrated, yielding

( )

ν

ν

ν

ν

ν

π

ν e 3 2 2 min e

eA

d

1

-

f

I

min

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

=

∫

∞ _(3.35)

Introducing the change of variable

e m 2 1 _ν2 ε = , then

( )

[ ]

⎭

⎬

⎫

⎩

⎨

⎧

⎟

⎠

⎞

⎜

⎝

⎛

=

∫

∞

ν

ε

ε

ε

ε

π

e V 2 3 e

f

V

-1

d

A

m

e

2

I

_(3.36) where _{( )} 1/2 m 2e ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = ε ε ν . Differentiating Ie we obtain

( )

[ ]

ν

ε

ε

π

e V 2 3 e

_A

_d

_f

m

e

2

-dV

dI

∫

∞

=

(3.37)

( )

[

V

]

Af

m

e

2

dV

I

d

e 2 3 2e 2

ν

π

=

_(3.38)

It is usually to introduce the electron energy distribution function (EEDF) g_e

( )

ε byg

( )

ε dε 4πν2f_e

( )

ν dν

e = .

Using the relation between ε and ν, we find

( )

ε π ε1/2 _e

[ ]

ν

( )

ε 3/2 e f m 2e 2 g ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = (3.39)

Using this to estimate fe from (equation 38), we obtain

( )

₂e 2 1/2 2 e dV I d m 2eV A e 2m V g ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = (3.40)

The electron energy probability function (EEPF)

( )

ε ε-1/2 _e

( )

ε

p g

g = is

sometimes introduced. For a Maxwellian distribution,

( )

_⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

=

e 3/2 -e e p

T

-exp

T

n

2

g

ε

π

ε

(3.41)

Obviously, we could obtain the EEDF and EEPF curves from the second order

differentiation of the electron saturation current. We also could use g_e

( )

ε to obtain the electron densityn_e, average energy and effective temperature by the followed

definition

∫

∞ = 0 e e g d n ε

∫

∞ = 0 e e d g n 1 _{ε ε} ε ε 3 2 T_eff =

3.4 Parameter Calculation

As the probe voltage is swept with respect to the plasma potential, electrons and

ions are either attracted or repelled and the net current collected by the probe changes.

From the interpretation of the I-V characteristic the electron distribution function,

electron temperature and plasma density can be obtained.

1. Plasma potential:

From I-V curve, we may obtain the plasma potential directly by max dV dI B = or 0 dV I d 2 B 2 = . 2. Plasma density:

In ion saturation region, the ion saturation current can be state as

1/2 B p s m V -V 2e ad 2en I ⎟ ⎟ ⎠ ⎞ ⎜ ⎜ ⎝ ⎛ =

And a plot of I2 versus VB should be linear, with ns2 by the slope of this line,

independent of Te and Ti.

3. Electron temperature:

In the transition and electron saturation region, the current can be state

respectively as e e p B e e _T AV V -V exp eN 4 1 I _⎥ ⎦ ⎤ ⎢ ⎣ ⎡ = _∞ and

e e sat e, eN AV 4 1 I = _∞

Comparing above equations and taking the logarithm, we have

e P B sat e, e T V -V I I ln _⎟⎟= ⎠ ⎞ ⎜⎜ ⎝ ⎛

From above equation we see that the electron temperature can be written as

( )

( )

-1 B e _dV lnI d eV T _⎟⎟ ⎠ ⎞ ⎜⎜ ⎝ ⎛ = 4. EEDF theory:

We can obtain the electron energy distribution function (EEDF), g_e

( )

ε , from the secondary differential of the electron current ₂e

2

dV I d

. And using g_e

( )

ε that we could obtain the electron density ne, average energy and effective temperature by the

followed definition

∫

∞ = 0 e e g d n ε

∫

∞ = 0 e e d g n 1 _{ε ε} ε ε 3 2 T_eff =

Chapter 4 Results and Discussion

We can obtain lots of properties of plasma from the I-V curve measured by

the electrostatics probe. The object we will measure is the DC sputtering chamber

in MuST laboratory. The probe was driven by the Z-motion will pass through

above the target from the left side to the right side. We started to measure the I-V

curve from the point 8cm besides left the center of the target to the point 8cm

besides right the centre of the target and we set the center of the target is x=0 (see

figure 4.1). We also set the distance of the horizontal cross section to the surface of

the target is Z and the surface of the target is z=0 (see figure 4.1). We will show the

I-V curves and the properties of plasma with and without magnetron in this chapter.

Then, we will show the spatial survey of the properties of a sputtering plasma

system between the substrate and the target.

4.1 The discharges with magnetron

An aluminum target is used with 99.99% purity argon gas at pressures of 10

mtorr, 20 mtorr and 30 mtorr and the magnetron assembly was installed under the

cathode. Figure 4.2 shows the picture of the discharge with magnetron assembly

and we can see the extreme glowing region above the target. The target voltage for

every scan was 310 V at 10 mtorr, 270 V at 20mtorr and 250 V at 30mtorr;

completed for each pressure at distances 4 and 6 cm from the cathode in the gap

distance 8 cm between the target and the substrate. Four horizontal surveys are

also completed for each pressure at distances 4, 6, 8 and 10 cm from the cathode in

the gap distance 11 cm between the target and the substrate.

4.1.1 Plasma potential and floating potential

The plasma potential of the horizontal surveys is presented from the figure

4.3 to the figure 4.5. There is a slight variation in plasma potential for each

horizontal and vertical survey. This weak change in plasma potential indicates that

there is a very weak electric field in the plasma.

The floating potential in contrast to plasma potential has a strong dependence

on the pressure and the spatial position. We could see horizontal distribution of

floating potential from figure 4.6 to figure 4.8 and vertical distribution in figure

4.9. We can find that in the outer edge of the cathode the floating potential is near

to 0 V but drops to minus when the probe moved into the center of the cathode. In

the vertical distribution, we can also find that the floating potential far away the

cathode is larger than the floating potential closing to the cathode.

The floating potential is dependent on the relative flux of electrons and ions

to the probe where the flux depends on the density and energy of the respective

the energy differences between the electron and positive ions. Duo to the great

difference in mass, the electrons are more mobile than the ions so the floating

potential is expected to depend on the electron energy. Hence, the changes in the

floating potential should be mirrored by changes to the electron energy or the

electron temperature.

From the concept of floating potential, we can equate the ion flux to the

electron flux, i.e. Φ /Te _e e s B s i n v e W 4 1 = u n =

Γ =Γ . Solving for Φw ( the potential of

the probe respect to the sheath-presheath edge), we obtain

1/2 e w m 2 M ln -T ⎟ ⎠ ⎞ ⎜ ⎝ ⎛ = π Φ .

For argon gas, we can calculateΦ_w =-4.7T_e. This means the argon ion with initial

energy ε_s =0.5T_efall through a collisionless dc sheath to a floating probe with

energy ofε_i ≈5.2T_e. We could preliminary estimate the value of e f p T V -V would

be around 5.2 ideally. Table 1~3 show the value of e f p T V -V in different places in

the chamber. The values are a little less than the ideal value. This means we could

overestimate the electron temperature or underestimate the plasma potential.

4.1.2 Plasma density

In the previous chapter, we talk about the sheath theory and electron energy

distribution function (EEDF). These two kinds of theories have different functions

plasma density and plasma potential via sheath theory. In the other hand, we could

obtain electron temperature and plasma density via electron energy distribution

function (EEDF) theory.

For the sheath theory, i.e. OML theory, we could obtain the plasma density and it

also required a thick sheath. On the other hand, the electron energy distribution

function is not affected by the thickness of the sheath and it also doesn’t require the

distribution of the electron energy obey the Maxwellian distribution. The theory of

electron energy distribution function is derived from the planar probe so it couldn’t

show the completely spatial symmetry. Figure 4.10 shows the differences of plasma

density between derived from sheath theory and the electron energy distribution

function theory. The plasma density derived from sheath theory is a little larger than

the plasma density from EEDF in the same position.

We also show the horizontal survey in the figure 4.11 and 4.12, and in these

figures the plasma density has larger value in the center of the target. In the outer field

of the cathode have lower value of the plasma density. In figure 4.13 with the probe

which was located above the center of the cathode, it shows the plasma density

decrease with the probe move up away the cathode. It makes sense that this location

close to the cathode (inside the magnetic trap) would record the highest electron

plasma diffuses outwards.

4.1.3 Electron temperature

There are two ways to obtain the electron temperature as we have remarked

before. We could obtain the electron temperature from the analysis of the curve of the

electron current but the electron energy distribution should be obeyed the Maxwellian

distribution first. Fortunately, most of the data are closing to the Maxwellian

distribution and we could receive the meaningful electron temperature.

In the theory of electron energy distribution function, it is using the method of

the averaged energy of distribution function as following equations. The equation 4.2

uses Teff to represent the effective electron temperature. The advantage of EEDF is

that the results are not influenced by the electron energy distribution and response the

instantaneous state of the plasma. We could find the difference of the electron

temperature derived from sheath theory and EEDF in the figure 4.16.

∫

∞ = 0 e e d g n 1 _{ε ε} ε (4.1) ε 3 2 T_eff = (4.2)

As the previous mentioned, the electron temperature is mirrored to the change of

the floating potential. In the figure 4.14 and 4.15, it shows the horizontal distribution

of electron temperature above the cathode. At the outer edges of the cathode the

2 eV above the center of the cathode. In figure 4.17 with the probe located above the

center of the cathode, it shows the electron temperature decrease with the probe move

up away the cathode. It makes sense that the plasma power source closes to the

cathode would generate the highest electron temperature and that at greater distances

vertical to the cathode the electron temperature would fall as more collisions occurred

outwards. We also can find the trend that the higher working pressure could lead to

lower electron temperature in figure 4.18. It could be realized that the more gas

molecules mean the more collisions occur and the electrons would lose its energy

after every collision.

4.2 The discharges without magnetron

In the discharges condition without magnetron, we also used an aluminum target

and 99.99% purity argon gas at pressures of 70 mtorr and 80 mtorr and removed the

magnetron assembly under the cathode. Figure 4.19 shows the picture of the discharge

without magnetron. According to the Paschen curve, we need the higher applied

voltage to make the gas discharge. The target voltage for every horizontal scan was

710 V at 70 mtorr and 700 V at 80 mtorr; currents at 0.01 A. Two horizontal surveys

are completed for each pressure at distances 2 and 3 cm from the cathode in the gap

4.2.1 Plasma potential and floating potential

The plasma potential of the horizontal surveys is presented from the figure

4.20. The variation in plasma potential is larger than the condition with magnetron

for each horizontal survey. This small change in plasma potential indicates that

there is a weak electric field in the plasma.

The floating potential in contrast to plasma potential has a strong dependence

on the spatial position. We could see horizontal distribution of floating potential in

figure 4.21. We can find that the floating potential decrease when the probe move

from the outer edge of the cathode to the center of the cathode.

The floating potential is dependent on the relative flux of electrons and ions to

the probe where the flux depends on the density and energy of the respective species

as we have mentioned before. Because of the plasma is quasineutral, any

inconsistency in flux will result the energy differences between the electron and

positive ions. Duo to the great difference in mass, the electrons are more mobile than

the ions so the floating potential is expected to depend on the electron energy. Hence,

the changes in the floating potential should be mirrored by changes to the electron

energy or the electron temperature.

As mentioned concept of floating potential in the section 4.1.1, we could state a

e f p T V -V

are larger than the ideal value, i.e. 5.2. This means the electron temperature is underestimated and plasma potential is overestimated in the condition operated

without magnetron.

4.2.2 Plasma density

We could obtain electron temperature and plasma density via electron energy

distribution function (EEDF) theory and sheath theory. The advantages and

disadvantages of the both methods had been mentioned before. We could find the

horizontal distribution of plasma density in figure 4.22. We also find out that the

plasma density is obviously quite smaller than the condition with magnetron even the

working pressure is larger than the condition with magnetron.

4.2.3 Electron temperature

There are two ways to obtain the electron temperature as we have remarked

before. We could obtain the electron temperature from the analysis of the curve of the

electron current but the electron energy distribution should be obeyed the Maxwellian

distribution first. Fortunately, most of the data are closing to the Maxwellian

distribution and we could receive the meaningful electron temperature by both of

above methods.

the floating potential. We also can find that the greater distances horizontal to the

center of the cathode and the electron temperature would fall down from figure 4.23

even though the range of decreasing is not large. The electron temperature did not rise

Chapter 5 Conclusions and Future Work

5.1 Summary

In the previous chapter, it could be found that we were dealing with two different

conditions. One condition is the system without the magnetron and another condition

is the system with the magnetron. The discharge system would need the higher

pressure and the system cannot be applied larger currents when system was operated

without the magnetron. The results of the distribution of plasma properties were not

symmetric as our expectation at beginning in both of conditions. We make the

coordinate move 4 cm right and the symmetry seems to be improved. It was found, in

general, that the pressure increased could lead density increasing at the same system

conditions. It also could be found that the electron temperature and the density

increased and floating potential decreased when thee probe moved forward into the

centre of the target.

The system with magnetron can generate the greater plasma density than the

system without magnetron, although the system with magnetron would be operated in

the lower pressure. The density and the electron temperature decrease progressively

with the distance of vertical the cathode increasing in the system with magnetron.

Because of the small variation of plasma potential, the electrical field in the plasma is

electron temperature, i.e. the floating potential decrease and the electron temperature

increase.

5.2 Recommendations for Future Work

This is a preliminary research to survey the special distribution of plasma

characteristics and it still has a lot of problems need to overcome. We could obtain the

different current values by applying a voltage bias when we install the electrostatic

probe into the plasma. However, it also disturbed the environment of the plasma. To

minimize the effects of the electrode that was on the probe was the following

problems needed to be solved.

When we put the magnetron into the system, the system can be operated in the

lower pressure. It also increased the sputtering phenomenon and it led the lots of

aluminum coated on the surface of the probe. Since the aluminum is the conductive

material, it seems to coat an electrode on the probe and increases the thickness of the

tip. It is also important to clean the probe after the measurement and decrease the time