University of Southern California Dissertations and Theses

A comparative study of plasma conditions, microstructure and residual stress in sputtered thin films

(USC Thesis Other)

A comparative study of plasma conditions, microstructure and residual stress in sputtered thin films

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

A COMPARATIVE STUDY OF PLASMA CONDITIONS, MICROSTRUCTURE
AND RESIDUAL STRESS IN SPUTTERED THIN FILMS

by
Anahita Afshin Navid

A Dissertation Presented to the
FACULTY OF THE USC GRADUTE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MECHANICAL ENGINEERING)

May 2011
Copyright 2011 Anahita Afshin Navid
ii

Dedication

To the love of my life, Babak, and my sister, Atoosa

iii

Acknowledgments
First and foremost, I would like to thank my advisor Professor Andrea Hodge for her
advice and encouragement throughout this work. I truly appreciate her dedication,
patience and unconditional support. This dissertation would not have been possible
without her thoughtful vision.
I would also like to thank Professor Michael Kassner and Professor Hai Wang for their
support and guidance during tough times in my Ph.D. pursuit. I would like to express my
gratitude to Professor Steven Nutt, Professor Florian Mansfeld and Professor Edward
Goo who accepted to serve on my committee.
Many thanks in particular to Timothy Furnish for his technical assistance and support in
setting up the sputtering chamber. It was pleasure working with him and I hope we can
collaborate again in the near future. I would like to thank Mr. Ilya Golosker for his help
in acquiring Langmuir probe data. I am also grateful to all my friends in the
Nanotechnology Lab for creating a productive working environment.
I am grateful to KNI (Kalvi Nanoscience Institute) Center at Caltech and specially Mrs.
Carol Garland for her assistant with TEM imaging. I would like to thank LLNL
(Lawrence Livermore National Lab) for their financial and technical support. This
research is partially funded by the U. S. Department of Energy by Lawrence Livermore
National Laboratory under Contract DE-AC52-07NA27344 and NSF grant number NSF-
EEC-0824059.
iv

Finally special thank to my sister, Atoosa, for her unconditional love and support. I
dedicate this thesis to the memory of my parents, Ali and Nahid, who raised me with
love, and supported me in my educational pursuits. Most of all, I am truly grateful to the
love of my life and my best friend, Babak, for his support, encouragement, and patience
in the last few years. I am very fortunate to have him in my life.
v

Table of Contents
Dedication ................................................................................................................... ii
Acknowledgments ............................................................................................................ iii
List of Tables ................................................................................................................ viii
List of Figures .................................................................................................................. ix
Abstract ................................................................................................................ xiv
Chapter 1. Introduction ............................................................................................. 1
Chapter 2. Background .............................................................................................. 6
2.1 Thin film processing............................................................................................. 6
2.1.1 Chemical Vapor Deposition (CVD).............................................................. 6
2.1.2 Physical Vapor Deposition (PVD) ................................................................ 7
2.2 Sputtering physics .............................................................................................. 15
2.2.1 Plasma diagnostic........................................................................................ 16
2.3 Thin film characterization .................................................................................. 17
2.3.1 Microstructure in sputtered films ................................................................ 17
2.3.2 Residual stress in sputtered films ................................................................ 19
2.4 Sputter deposition of copper and tantalum films ............................................... 25
2.4.1 Copper thin films ........................................................................................ 25
2.4.2 Tantalum thin films ..................................................................................... 27
Chapter 3. Experimental procedures and materials ............................................. 32
3.1 Sputtering chamber ............................................................................................ 32
3.1.1 Set up the sputtering chamber ..................................................................... 32
3.1.2 Plasma characteristic ................................................................................... 34
3.2 Characterization methods ................................................................................... 36
3.3 Mechanical and electrical testing ....................................................................... 39

vi

Chapter 4. Plasma characteristics in sputtering chamber .................................... 43
4.1 Introduction ........................................................................................................ 43
4.2 Experimental details ........................................................................................... 43
4.3 Results and discussion ........................................................................................ 44
4.3.1 Ion flux and energy ..................................................................................... 44
4.3.2 Microstructure and residual stress .............................................................. 51
4.4 Summary ............................................................................................................ 54
Chapter 5. In-situ stress evolution in sputtered copper and tantalum films ....... 55
5.1 Introduction ........................................................................................................ 55
5.2 Experimental Details .......................................................................................... 57
5.3 Results and discussion ........................................................................................ 59
5.3.1 Effects of sputtering pressure ...................................................................... 61
5.3.2 Effects of sputtering power ......................................................................... 67
5.3.3 Thermal and relaxation stress evaluation .................................................... 69
5.3.4 Total residual Stress Evaluation .................................................................. 76
5.4 Summary ............................................................................................................ 78
Chapter 6. Synthesis of nano-structured alpha and beta tantalum ..................... 80
6.1 Introduction ........................................................................................................ 80
6.2 Experimental Details .......................................................................................... 82
6.3 Results and discussions ...................................................................................... 83
6.3.1 Phase formation as a function of sputtering conditions .............................. 83
6.3.2 Effect of internal parameters on phase formation ....................................... 97
6.4 Summary .......................................................................................................... 104
Chapter 7. Controlling residual stresses in sputtered alpha-Tantalum ............ 106
7.1 Introduction ...................................................................................................... 106
7.2 Experimental methods ...................................................................................... 107
7.3 Results and discussions .................................................................................... 108
7.3.1 Stress development as a function of sputtering pressure .......................... 108
7.3.2 Role of underlayer ..................................................................................... 112
7.3.3 Alpha tantalum stress state ........................................................................ 117
7.4 Summary .......................................................................................................... 118
Conclusions ............................................................................................................... 119
References ............................................................................................................... 122
Alphabetized Bibliography .......................................................................................... 130
vii

Appendices ............................................................................................................... 140
Appendix A: Summary of Aluminum samples ........................................................... 140
Appendix B: Summary of copper samples ................................................................. 141
Appendix C: Summary of multilayer copper samples ................................................ 143
Appendix D: Summary of tantalum samples .............................................................. 144
Appendix E: Summary of tantalum sample with underlayers .................................... 147

viii

List of Tables
Table 1: Thermal-mechanical properties of alpha and beta tantalum .............................. 28
Table 2: Proposed models for beta tantalum structure ...................................................... 30
Table 3: Summary of stress measurements for copper and tantalum films
deposited at 100 watts sputtering power with thicknesses of ~ 500 nm ................... 71
Table 4 : Summary of substrate temperature increase for copper and
tantalum films deposited at 100 watts sputtering power with
thicknesses of ~ 500 nm ............................................................................................ 72
Table 5: Total energy supplied to the film during deposition at a constant
sputtering power (100 watts) calculated from Ref. [1]. ............................................ 73
Table 6: Summary of the processing conditions and phase formation of
samples deposited at different powers and pressures................................................ 89
Table 7: Summary of the hardness, residual stress and texture for samples
processed at 100 watts power and sputtering pressures ranging from
0.3 Pa to 1.4 Pa. ........................................................................................................ 94
Table 8: Calculated argon impurities in the Ta films as a function of
sputtering pressure .................................................................................................. 103
ix

List of Figures
Figure 1 : Schematic of the correlation between internal and external
chamber conditions to the films residual stress and microstructure ........................... 2
Figure 2: The sequence of gas transport and reaction process to deposit the
film by CVD technique. .............................................................................................. 7
Figure 3 : Schematic of Pulsed Laser Deposition technique .............................................. 9
Figure 4 : Schematic of Molecular beam epitaxy chamber .............................................. 10
Figure 5: Schematic of planer diode sputtering system ................................................... 13
Figure 6 : Schematic of magnetron sputtering process ..................................................... 14
Figure 7 : Schematic of balance and unbalance sputtering system ................................... 15
Figure 8 : Schematic of the influence of substrate temperature and argon
pressure on the microstructure .................................................................................. 18
Figure 9: Schematic representation of thermal and intrinsic stress .................................. 23
Figure 10 : Effect of working gas pressure on the stress condition ................................. 23
Figure 11: The effect of bias voltage on the residual stress on aluminum ...................... 24
Figure 12: The texture evolution as a function of bias voltage for sputter
copper thin films ...................................................................................................... 26
Figure 13 : In-situ stress development in copper thin films as a function of
sputtering pressures. ................................................................................................. 27
Figure 14 : XRD pattern of the alpha tantalum (foil and powder) and beta
tantalum .................................................................................................................... 29
Figure 15 : Schematic of sputtering chamber ................................................................... 32
Figure 16 : Racetrack formation during sputtering of copper ........................................... 33
Figure 17 : Schematic of substrate holder......................................................................... 34
x

Figure 18 : Langmuir Probe inside the chamber ............................................................... 35
Figure 19 : Steps for TEM sample preparation using FIB liftout technique .................... 38
Figure 20: Before and after profile for tantalum sample at 100 watts and 2
mtorr .......................................................................................................................... 39
Figure 21: Schematic of MOSS system ............................................................................ 41
Figure 22 : Schematic of I-V graph for single probe ....................................................... 45
Figure 23 : Ion density as a function of sputtering pressure. Notice that
ion density increase almost linearly as the pressure increase. .................................. 48
Figure 24 : The ratio of metal flux vs. ion flux as a function of sputtering
pressures. ................................................................................................................... 49
Figure 25 : The plasma potential as a function of sputtering pressure.
Notice that in the absence of bias voltage the plasma potential is
equal to ion energy. ................................................................................................... 50
Figure 26 : Plan view TEM images for the copper films deposited at 100
watts sputtering power and pressures ranging from (a) 0.3 Pa (b) 0.7
Pa (c) 1.4 Pa (d) 2.8 Pa. ............................................................................................ 51
Figure 27 : Residual stress as a function of sputtering pressure for copper
films. Notice that at 1.4 and 2.8 Pa sputtering pressure the films are
cracked. ..................................................................................................................... 53
Figure 28: Growth and cooling graph for copper samples deposited at 0.3,
0.7 and 1.4 Pa sputtering pressure. The cooling graph represents the
thermal stress contribution and temperature rise during growth. ............................. 58
Figure 29: Film stress vs. thickness plots at 100 watts sputtering power
calculated using MOSS for (a) copper and (b) tantalum samples
sputtered at 0.3, 0.7 and 1.4 Pa. Insets present a magnified view of
the film stress vs. thickness for the first 20 nm. ........................................................ 61
Figure 30: Scanning electron micrographs of the surface morphology for
copper films processed at 100 watts sputtering power as a function of
sputtering pressure at (a) 0.3, (b) 0.7 and (c) 1.4 Pa, (d) XRD spectra
of the films as a function of sputtering pressures...................................................... 63
xi

Figure 31: Scanning electron micrographs of the surface morphology for
tantalum films processed at 100 watts sputtering power as a function
of sputtering pressure at (a) 0.3, (b) 0.7 and (c) 1.4 Pa , (d) XRD
spectra of the films as a function of sputtering pressure, noting α
(cubic) and β (tetragonal) phases. ............................................................................. 66
Figure 32: (a) Film stress vs. thickness plots for copper samples deposited
at 1.4 Pa sputtering pressure measured with MOSS technique as a
function of power including surface morphology at (b) 100 watts (c)
200 watts sputtering power. ...................................................................................... 67
Figure 33: Film stress vs. thickness plots for tantalum samples deposited at
0.7 Pa sputtering pressure measured with MOSS technique as a
function of sputtering power including surface morphology at (b) 100
watts (c) 200 watts sputtering power. ....................................................................... 69
Figure 34: Focus Ion Beam (FIB) cross section images of copper samples
deposited using 100 watts sputtering power at (a) 0.3, (b) 0.7 and (c)
1.4 Pa sputtering pressure and. Note the grain size changes as a
function of sputtering pressure and thickness. .......................................................... 75
Figure 35 : Comparison between growth stress measured in-situ using
MOSS technique and the total residual stress measured ex-situ
utilizing stylus profilometer. The difference represents the thermal
relaxation for copper sample deposited at 0.7 Pa sputtering pressure
as a function of film thickness. ................................................................................. 77
Figure 36: XRD spectra of Ta films deposited at 100 watts sputtering
power and different sputtering pressures ranging from 0.3 Pa to 1.4
Pa. Note the formation of (110) BCC Ta at 0.7 Pa sputtering
pressure. .................................................................................................................... 84
Figure 37: SEM images of the surface morphology of tantalum films
synthesized at 100 watts sputtering power and various sputtering
pressures ranging from 0.3 Pa to 1.4 Pa, respectively. ............................................. 85
Figure 38: (a) XRD spectra of the Ta films deposited at 0.7 Pa sputtering
pressure and different powers ranging from 50 watts to 200 watts (b-
d) corresponding surface morphology of the Ta films. Note formation
of BCC Ta at all powers. ........................................................................................... 88
Figure 39: Hardness values for nanocrystalline alpha (d ~ 45nm) and beta
(d ~ 41-51nm) Ta as a function of sputtering pressure (0.3 Pa to 1.4
xii

Pa). Lower hardness values at 0.5 and 0.7 Pa sputtering pressure are
due to the formation of alpha Ta. .............................................................................. 90
Figure 40: TEM cross-section images of Ta films at 0.3, 0.7 and 1.4 Pa
sputtering pressure, respectively. Note formation of the interface
layer for the sample processed at 0.7 and 1.4 Pa sputtering pressure.
The inset shows the SAD pattern of the films. ......................................................... 92
Figure 41: Plane-view TEM micrograph of the nanocrystalline Ta
processed at 100 watts sputtering power and 0.3 Pa sputtering
pressure. .................................................................................................................... 93
Figure 42: XRD spectra of the tantalum films as a function of thickness
(10 nm to 1000 nm) and at different sputtering pressures of 0.3, 0.7
and 1.4 Pa sputtering pressure. ................................................................................. 96
Figure 43: Calculated ion energy for tantalum and copper samples
measured by Langmuir probe at 100 watts power and sputtering
pressures ranging from 0.3 to 1.4 Pa. A small peak is observed at 0.7
Pa sputtering pressure. .............................................................................................. 98
Figure 44: XRD spectra of Ta films deposited using (a) 2×10
-3
Pa base
pressure (low vacuum) and (b) 2×10
-5
Pa base pressure (high
vacuum) at 100 watts power and 0.3, 0.7 and 1.4 Pa sputtering
pressures. Note formation of highly textured beta Ta at all of the
processing conditions for the sample deposited using 2×10
-3
Pa base
pressure (low vacuum). ........................................................................................... 101
Figure 45: (a) Selected X-ray spectra of Ta films sputtered at a power of
100 watts, with increasing sputtering pressures from 0.3 Pa to 1.4 Pa.
Notice the formation of alpha Ta at 0.7 Pa as shown by XRD and
TEM in (b-d). Where b) is the cross-sectional image, arrow denotes
growth direction, (c) Top-view image and (d) plane-view image of
the interface (i) layer showing the formation of ultra small grains at
the interface between film and substrate. ................................................................ 109
Figure 46: Plot of the residual stress as a function of sputtering pressures
from 0.3 Pa to 1.4 Pa. Arrow denotes alpha phase at 0.7 Pa. All
other pressures have a pure beta or a mixture of alpha and beta
structures. ................................................................................................................ 110
Figure 47 : The effect of (a) Aluminum and (b) Niobium underlayers on
the phase formation of tantalum.............................................................................. 114
xiii

Figure 48: (a) the XRD spectra of the films processed at 0.3, 0.7, 1.4Pa,
respectively, using an alpha Ta underlayer. (b)-(d) SEM images of
the surface morphology of Ta samples synthesized at 100 watts using
three different sputtering pressures of 0.3, 0.7 and 1.4 Pa deposited
on an alpha Ta underlayer surface. ......................................................................... 116
Figure 49: Plot of the residual stress of alpha Ta deposited at different
sputtering pressures (0.3, 0.7, and 1.4 Pa) using an alpha Ta
underlayer. Note the formation of alpha Ta at all three different
sputtering pressures. ................................................................................................ 117

xiv

Abstract
The correlation between external (e.g. power, pressure) and internal (e.g. plasma
conditions, impurities) chamber conditions to the film residual stress and microstructure
has been studied for copper and tantalum sputtered thin films. To investigate the effect of
external parameters on the film residual stress and microstructure, the stress evolution in
copper and tantalum films at various sputtering powers and pressures was measured in-
situ using Multi-beam Optical Stress Sensor (MOSS) technique. The results demonstrate
that for both materials, during deposition, a compressive stress initially develops at low
sputtering pressures, while at the highest sputtering pressures the stress is always tensile.
In addition, profilometry measurements are compared to the final instantaneous stress
value at the end of deposition measured by MOSS. Comparing the total residual stress
measured ex-situ with the growth stress measured in-situ, emphasizes on the significant
contribution of the relaxations mechanism after deposition. The measurements
demonstrate that the contribution of the relaxation mechanism is quite significant and can
be as large or larger than the growth stress and might change the stress state on the film.
Meanwhile, the effect of internal chamber conditions in relation to the phase formation
and stress state of tantalum films was investigated as a function of external chamber
parameters. The formation of stable body centered cubic (BCC) and metastable
tetragonal tantalum was observed in films deposited on Si substrates at pressures between
0.3 and 1.4 Pa. The results demonstrate the formation of low resistivity BCC tantalum at
0.7 Pa sputtering pressure and various powers ranging from 50 to 200 watts. The
xv

variation in hardness, residual stress and texture development are investigated for all
deposition conditions. In addition, the effect of internal chamber conditions such as the
plasma kinetics and impurities levels was studied in respect to the phase formation at a
given sputtering condition. Results show that the formation of alpha tantalum was related
to the lower amount of impurities during deposition as well as to an increase in the ion
bombardment energy. Therefore, the internal chamber conditions have a significant
effect on the tantalum film’s microstructure. Moreover, the results demonstrated that
alpha tantalum films can be deposited at room temperature having either compressive or
tensile stress states. Hence, the stress evolution in tantalum films can be independent of a
particular phase formation and is directly related to the kinetics of ions and atoms inside
the chamber.
1

Chapter 1. Introduction
Thin films have a wide range of applications from aerospace to semiconductor industries
and play an important role in the development of nanometer materials. However, the
study of thin films is very challenging since it is highly dependent on the materials and
the processing techniques; therefore, physical understanding of the processing methods is
very important. Among thin film deposition techniques, magnetron sputtering is one of
the most flexible methods to synthesize any type of material ranging from high purity
thin films to multilayers. The main advantage of this method is that during sputtering,
the target composition is maintained whether it is a single element or an alloy, and it can
also provide a high deposition rate and uniform film thickness.
Sputtered thin films are usually under a tensile or compressive stress, depending on the
deposition conditions, and their microstructure can be quite different than the bulk
material with respect to grain size and texture. Therefore, one has to consider the role of
background pressure, kinetic energy of the incident particles, deposition rate, the
substrate material and temperature. However, these parameters are often coupled each
other and are also affected by chamber geometry, source dimensions and target wear.
Consequently, due to the existence of numerous parameters that affect the properties of
the films, the processing of materials using magnetron sputtering is still mostly based on
trial and error techniques.

2

To visualize the correlation between the internal and external chamber conditions to the
films residual stress and microstructure, Figure 1 is presented. In this figure the relation
between the external chamber conditions, such as power and pressure, to both the internal
conditions and the film characteristics are summarized.

Figure 1 : Schematic of the correlation between internal and external chamber conditions to the films
residual stress and microstructure

It is observed that the internal chamber conditions such as plasma characteristics and
impurity levels are directly influenced by the external conditions such as sputtering
power and pressure. Meanwhile, the film stress and microstructure also depend on both
the external and internal conditions. During the 1970’s and 1980’s, Thornton et al. [5]
developed a comprehensive study on magnetron sputtering, relating the external
External Conditions
(Power & pressure)
Internal Conditions
(Plasma parameters & impurities)

Films characteristics
(Stress & microstructure)
(1)
(2)
(3)
3

parameters to the film microstructure and residual stress (arrow 3 in Figure 1). However,
in this study we attempt to investigate the physics behind this processing method, rather
than only the final outcome, by considering the effects of internal chamber conditions. In
order to do so, we studied the correlation of internal conditions to both the external
conditions (arrow 1 in Figure 1) and the films’ microstructure and residual stress (arrow 2
in Figure 1). The combination of these factors will define the dominate parameters in
the processing of sputtered thin films.
Copper and tantalum have been selected in this study representing FCC (Face Centered
Cubic) and BCC (Body Centered Cubic) structures. We chose copper since it is a
standard material and sputtering of copper has been extensively documented [5, 11-13]
and therefore, chamber to chamber results can be compared. In addition, acquiring stable
plasma is fairly straightforward since copper has lower atomic mass. On the other hand,
synthesizing tantalum films is challenging due to formation of multiple phases (alpha and
beta) and due the fact that film properties are highly affected by the processing
conditions. To date, the controlling mechanism for processing a single phase tantalum is
not well understood and further research is essential to understand the dominant factors in
tantalum phase formation.
In order to take into account the effects of internal chamber conditions, we measured the
plasma characteristics (e.g. ion flux, electron temperature, electron density) during
deposition using a Langmuir probe. The aim is to build a bridge between plasma physics
and material science, with the purpose of predicting materials behavior from plasma
4

parameters. By studying the effects of plasma conditions inside the chamber and their
relation to the films residual stress and microstructure, we will be able to identify the
controlling parameters during sputter deposition.
With the goal of investigating the effects of external chamber conditions, such as
sputtering power and pressure on the film properties, we studied the in-situ stress
evolution and microstructure development as a function of time. To date, significant
progress has been made towards understanding real-time stress formation during film
growth using in-situ techniques, with emphasis in evaporation methods [14-25]. In
contrast, for magnetron sputtering, which has different and more complex kinetics than
evaporation, only a few studies have been performed in-situ focusing on tensile to
compressive state transitions [8, 26-29]. In-situ stress measurement is a promising
technique to investigate the growth mechanism and is directly related to the films
morphology; therefore, it is a valuable method to expand our knowledge in relation to the
film growth.
As the next step, we incorporated our synthesis expertise to process single phase alpha
tantalum. In general, during magnetron sputtering deposition, metastable beta Ta
develops [7, 30-50], while the stable alpha phase is typically acquired by applying
external parameters such as heating the substrate [38, 51], adding a bias [31, 39, 43, 52,
53] or by changing the substrate material [33, 38, 54-57]. However, the dominant factors
promoting the formation of alpha Ta are not yet understood and the results can change
from system to system. Therefore, it is essential to investigate the kinetics of the atoms
5

and ions inside the sputtering chamber, and specifically in the case of BCC materials, it is
important to study the effects of low level impurities during deposition. To this extent,
understanding the processing of single phase alpha Ta at room temperature without
applying external parameters could lead us to recognize both the effects of external and
internal chamber conditions to a particular phase formation during deposition. Finally,
we studied the variation in residual stress development in respect to the phase formation
as a function of sputtering conditions (without heating or bias); specifically with
emphasis on controlling the residual stress in single phase alpha Ta at different sputtering
pressures.

6

Chapter 2. Background
2.1 Thin film processing
Many techniques are utilized to deposit thin films; each allows control over composition,
thickness and microstructure in order to meet the requirements for any specific
applications. Physical vapor deposition (PVD) and chemical vapor deposition (CVD) are
the most common methods which involve transferring a material atom by atom from the
source to the substrate [58]. PVD method is based on the creation of a materials vapor by
means of evaporation, sputtering or condensing of the vapor onto a substrate to form a
film. In a CVD method, volatile precursors react on the substrate surface and produced
solid films [58]. The following sections present a short summary of these techniques
with emphasis on PVD methods.
2.1.1 Chemical Vapor Deposition (CVD)
This method is based on the chemical reaction between volatile compounds of the
materials with gases to assist the atomic deposition of non-volatile solid films on the
substrate. There are different kinds of reactions including Pyrolysis, Reduction,
Oxidation and Compound formation [3]. The sequence of gas transport and reactions
during CVD deposition is shown in Figure 2. High growth rate is the main advantage of
the CVD methods. However, this technique is not applicable to all compounds and a set
of chemical reactions has to be indentified prior to deposition. In addition, this method is
not always the most favorable technique since large amount of chemicals must be used
during CVD deposition.
7

2.1.2 Physical Vapor Deposition (PVD)
PVD methods are based on the creation of a materials vapor which subsequently
condenses on the substrate to create a film. There are different techniques to create the
vapors which are mainly divided into vacuum evaporation and physical sputtering
methods. In the following sections, the most common techniques for creating the vapor
with evaporation and sputtering are summarized.
2.1.2.1 Vacuum evaporation:
In the vacuum evaporation method, materials are heated in the vacuum until the vapor
pressure exceeds ambient pressure [2]. This method consists on transferring atoms in a
control way from the heated source to a substrate located a distance away, where the film
formation and growth proceed atomistically. This process has to be performed in a
vacuum better than 10
-4
Torr. In order to prevent film contamination a vacuum of 10
-7

Figure 2: The sequence of gas transport and reaction process to deposit the film by CVD technique.
[3]
8

Torr is desired. The main advantages and disadvantages of this method, compared to
other PVD methods, are as follows [59]:
 High deposition rate can be obtained
 Deposition monitoring is relatively easy
 High purity film can be deposited from high purity sources
 Technique is relatively cheap
The main disadvantages are as follows:
 Poor surface coverage and uniformity
 Poor ability to deposit many alloys and compounds
 Non-optimal film properties- e.g. pinholes, high residual stress
 High radiant heat loads during the processing
There are different evaporation methods such as thermal evaporation which use different
sources such as sublimation sources, crucible, electron beam heated and resistance heated
sources and also various techniques such as arc evaporation, pulsed laser deposition
(PLD) and molecular beam epitaxy (MBE). The last two techniques are the most
common methods and are presented in more details.
Pulsed Laser Deposition (PLD): PLD method is commonly used for growing highly
crystalline dielectric films, compound semiconductor epitaxial layers and
superconducting films. The schematic of this technique is shown in Figure 3. In this
method, material is vaporized and ejected from the surface of a target as it irradiated by a
9

laser beam [2]. Subsequently, a thin layer of materials are formed by condensing the
vaporized material from the target onto a solid substrate. Absorption characteristics of
the material to be evaporated determine the laser wavelength to be used [59].

The main advantages of this technique are as following [2]:
 Deposited film has very high quality.
 Excellent transfer of stoichiometry between the target and the film.
The main disadvantages of this technique are:
 The wavelength of the laser should be compatible with the absorption characteristics
of the material to be evaporated.
 Energy conversion efficiency is very low.
 The size of the deposited film is small (10 to 20 mm or 0.4 to 0.8 in. diameter).

Figure 3 : Schematic of Pulsed Laser Deposition technique [6]
10

Molecular beam epitaxy (MBE): Molecular beam epitaxy is a technique for epitaxial
growth via the interaction of one or several molecular or atomic beams [59]. The vacuum
better than 10
-9
Torr is necessary for MBE deposition. The schematic of the MBE
chamber is shown in Figure 4.

The solid sources materials are placed in evaporation cells to provide an angular
distribution of atoms or molecules in a beam. The substrate is heated to the necessary
temperature and continuously rotated to improve growth homogeneity [2]. In this
technique, the substrate is pre-sputtered with the low energy ion source to remove any
contamination on the surface and then annealed at high temperature in order to relax any
damage that may occur during preparation. The substrate is then cooled to the growth
temperature and growth started [59]. The beams are emitted from crucible, which were
heated to temperature above the substrate to induce evaporation and condensation. This

Figure 4 : Schematic of Molecular beam epitaxy chamber [4]
11

is a good method for growth of multilayer structures having different compositions and
doping levels. The main advantage of this method is processing high quality film;
however, the rate of growth is very low (~ 2 nm/s) compare to the other methods. [60]
2.1.2.2 Physical sputtering:
Physical sputtering involves the physical vaporization of atoms from the surface by
momentum transfer of bombarding energetic particles [2]. During sputtering deposition,
ions of sputtering gas are accelerated toward the cathode using a large electrical field and
strike the target surface with high energetic particles. As a result, secondary electrons
and neutral gas atoms are released. When these electrons and ions reach critical numbers,
the gas begins to glow and the discharge becomes self-sustaining [3]. The main
advantages of sputtering depositions are as follows:
 It is possible to sputter any type of materials.
 It is possible to use several targets simultaneously or sequentially to acquire
different compositions.
 The sputtering yield is independent of the temperature.
 During sputtering of alloys and other complicated materials, stoichiometry is
maintained.
The main disadvantages are as following:
 Sputter vaporization rate are low compared to those that can be achieved by
thermal vaporization.
12

 Sputtering is not energy efficient.
 Sputtering is an expensive method.
 Substrate heating from electron bombardment can be high in some configurations.
There are different sputtering methods including cathodic sputtering, diode sputtering,
Direct current and Radio frequency magnetron sputtering, ion-beam sputtering and
reactive sputtering [2]. However, the most common methods are Planar diode,
magnetron sputtering and balance and unbalance sputtering system and are presented in
more detail.
Planar Diode Glow Discharge Sputter Deposition: This configuration consists of the
cathode (the target) and anode facing each other (see Figure 5). The target is usually
water-cooled during deposition and it works both as the source of coating material and
also as the source of secondary electrons to sustain the plasma. The distance between the
cathode and anode is usually about 5 to 10 cm. The deposition rate is usually change as a
function of the power density at the target surface, size of the erosion area, source-
substrate distance, source material, and working gas pressure. Some of these factors are
interrelated, such as pressure and power density [2]. Therefore, the optimum operating
condition is obtained by controlling the parameters to get the maximum power which can
be applied to the target without causing cracking, sublimation or melting [2].
This technique is widely used due to its simplicity and also it is relatively easy to
fabricate targets for a wide range of materials. However, it has couple if disadvantages
13

such as low deposition rate, substrate heating due to the bombardment of high energy
particles, and relatively small deposition surface areas. [59]

Magnetron sputtering: Magnetron sputtering is a high deposition rate vacuum coating
technique which allows deposition of different metals, alloys and compounds with
different thicknesses. The schematic of a magnetron sputtering system is shown in
Figure 6. In magnetron sputtering, an applied magnetic field parallel to the cathode
surface traps the primary electrons around the cathode and prevents them from losing
their energy to the chamber walls. The magnetic field strength is in the range of few
hundred gauss and therefore, it can influence the plasma electrons but not the ions [2].
The electrons trapped on a given field line can advance across the magnetic field to an
anode or walls by making collisions. Hence, the number of primary electrons which hit
the substrate without experiencing collision is decreased. Consequently, ionization
efficiencies are high and this allows processing at low pressure (as low as 0.5 mtorr)
compared to planar diode sputtering which has to be performed at pressures higher than
20 mtorr [61]. The primary advantages of magnetron sputtering are (1) high deposition

Figure 5: Schematic of planer diode sputtering system [2]
14

rate (2) ease of sputtering any metal, alloys or compound (3) high purity film (4)
excellent coverage of steps (5) excellent uniformity on a large area substrate.

Balanced and Unbalanced Magnetrons: In the balanced magnetron system the magnetic
field is designed to keep the electrons around the substrate. The balanced magnetron was
developed mainly for microelectronic applications, since it is not favorable to
bombarding the growing film by energetic particles during deposition (see Figure 7). On
the other hand, for hard coatings it is necessary to bombard the growing film with
energetic particles. For this purpose, in unbalanced magnetron, the films are bombarded
by energetic ions with setting up the secondary plasma near the substrate as a source of
energetic particles or using a separate ion gun. In some cases multiple cathodes sources
may be used to ensure that the plasma covers the entire deposition volume [59].

Figure 6 : Schematic of magnetron sputtering process
15

2.2 Sputtering physics
The sputtering process is based on the movement and collisions between the ions,
electrons and neutrals. The energetic particles inside the plasma penetrate into the target
surface as a result of an electrical field and create the collision cascade. As a result of
these collision cascades, some of the momentum energy is transferred to the target atoms
and they are ejected from the surface and the rest are reflected as high energy neutrals
implanted onto the surface. The main part of the transferred energy will change to heat
on the target surface [58, 59].
At low sputtering pressure, the high energy reflected neutrals and atoms bombard the film
and results in the formation of dense films. On the other hand, at high pressures, because
of the high rate of physical collisions and charge exchange collisions, less sputtered
atoms may reach the film surface; therefore the voided microstructure film may develop.

Figure 7 : Schematic of balance and unbalance sputtering system [2]

16

In the latter case, the deposition rate decreases while the sputtering rate increases due to
higher collisional scattering [59].
Plasma conditions inside the chamber can change with different parameters such as
pressure, power and geometry of the chamber. The purpose of the next section is to
evaluate the plasma conditions and different methods for plasma diagnostic during film
deposition.
2.2.1 Plasma diagnostic
Plasma is a gaseous mixture of the electrons, ions and neutrals [62]. Partially ionized and
low temperature plasma (cold plasma) is usually used for thin film processing. In cold
plasma, electrons are not in the thermal equilibrium with ions therefore, they are able to
reach higher temperature. Electron temperature is usually between 1 to 10 eV with 2 eV
being a typical value [3]. Ions have much less energy usually around 0.04 eV and
neutrals have energies of 0.025eV [63, 64].
For evaluating plasma condition such as electron and ion temperature and density, a
Faraday cup or a Langmuir probe can be used. A Faraday cup is a conductive cup
designed to catch particles inside the vacuum. When the ions in the plasma hit the cup,
the metal acquires a charge while the ions are neutral. By measuring the electrical
current of the charged particles, the number of ions can be determined [65].
A Langmuir probe is an electrically biased wire inserted in the plasma. The data consists
of a set of measurements of the probe current versus the probe voltage which is referred
as I-V characteristic graph [62, 66]. The wire is usually made from high melting point
17

materials like tungsten and tantalum covered with an alumina tube. There are different
kinds of Langmuir probes such as single probe, double probe or triple probe [66]. The
most straightforward probe is the single probe, where the tungsten tip is biased with a
voltage ramp relative to the chamber. When the bias is sufficiently negative relative to
the floating potential, the current collected by the probe is the ion saturation current since
no electrons can reach the probe. Since the Faraday cup installation is more difficult
compared to the Langmuir probe, the latter has been chosen in this study.
2.3 Thin film characterization
Controlling over microstructure and stress state is usually achievable by changing
sputtering conditions such as power, pressure, substrate temperature, bias voltage,
deposition rate and working gas species. However, there is not a unique parameter to
control the film properties and to date most of the studies are based on the trial and error.
In this section, the influence of these parameters on the film characterization will be
discussed in detail.
2.3.1 Microstructure in sputtered films
The effect of different parameters on the microstructure of the sputtered thin films was
studied in detail by Thornton et al [5, 13, 67]. However, they did not present any unified
parameter to characterize the microstructure during deposition. Specifically, they have
studied the effect of substrate temperature, gas pressure and apparatus geometry on the
microstructure for both thin and thick films deposited by magnetron sputtering and
proposed a three-zone model to present the relationship between the microstructure and
18

the gas pressure (see Figure 8). It has been shown that substrate temperature is the most
important factor in the thin film growth given the fact that at high substrate temperature,
bulk chemistry and diffusion dominate all other factors like chamber geometry and
working gas pressure [13].
It is observed from Figure 8 that at low deposition temperature (T/T
m
< 0.3 , the
columnar structure with voided grain boundaries develops which is the consequence of
atomic shadowing (Zone 1). By increasing the gas pressure, more collisional scattering
happens which results in extending Zone 1 to the higher temperature due to more oblique
component of the flux [13]. At higher temperature (0.3 < T/T
m
< 0.5) the Zone 2 structure
consist of columnar grains separated by true grain boundaries growth. The source of
activation energy for zone 2 is the surface diffusion and recovery usually starts in this
(T/T
m
> 0.5) zone. At very high temperature, bulk diffusion dominates the deposition
parameters and equiaxed grains start to growth. Recrystallization and recovery occurs in
this phase [5].

Figure 8 : Schematic of the influence of substrate temperature and argon pressure on the
microstructure [5]
19

Zone T structure is characterized by dense fibrous with the smooth surface. This occurs
at low temperature when the flux arrives in the normal direction to the substrate surface,
therefore, minimizes the effect of shadowing. In general, the zone T structure is the same
as zone 1 but with fibrous crystals and dense grain boundaries which can improve the
materials properties. This structure is characterized by compressive internal stress
compare to Zone 1 structure which present the tensile stress [13].
2.3.2 Residual stress in sputtered films
Thin films can contain many defects which are the result of the growth process and act as
the sources of stress. The major objective in thin film research is to decrease the residual
stress and synthesize stress free films. The residual stresses in the film are composed of
intrinsic stress and extrinsic stress [68]. Intrinsic stress is the measured stress within the
film such as grain boundaries, recrystallization, voids and impurities, while extrinsic
stress is the combination of intrinsic and thermal stress [69].
I. Extrinsic stress:
Extrinsic stress which can be measured after deposition is analogous to a “total stress”
and contains the thermal stress and intrinsic stress. Thermal stress depends on different
parameters such as processing method, materials melting point and the difference
between the thermal expansion coefficients of the film and substrate. In general, the
substrate temperature usually increase during the film deposition, therefore, the thermal
stress which is due to the difference between the thermal expansion coefficient of the film
20

and substrate will be imposed to the film [16]. The biaxial strain due to this difference
between thermal expansion coefficients is as follows:

≅ ∆ ∆   2.1
And the thermal stress associated with it is as follows:

1

2.2
Where is the zero stress-temperature, T is the actual temperature and and are the
thermal expansion coefficients of the film and substrate, respectively [11, 16]. When the
thermal expansion coefficient of the film is higher than the substrate, a tensile stress is
expected in the film during cooling. It has been shown that when the thermal stress
exceeds the elastic limit; relaxation occurs by various methods including dislocation
glide, dislocation climb and grain boundaries diffusional creep. However, the strain
relaxation rate depends on the film microstructure, applied stress and temperature
II. Intrinsic stress:
Stress contributions within the film are from different sources such as small angle grain
boundaries, recrystallization processes, impurities and lattice expansion mechanism. A
short description of these factors is explained in this section.
Small angle grain boundaries: Polycrystalline films with columnar grains usually
demonstrate high numbers of small-angle grain boundaries which separate the randomly
21

oriented grains. These grain boundaries acts as a source of intrinsic stress since the inter-
atomic force at the grain boundaries has the tendency to close the existing gap on the
neighboring crystals, therefore, straining the film in tension [16, 70].
Recrystallization processes: If the self-diffusion of the deposited film material is high
enough recrystallization may occur during and after depositions. During
recrystallization, defects such as grain boundaries, voids and point defects are annealed
which results in film densification and consequently tensile stress contribution in the film
[16].
Impurities: It is known that the growth and morphology of thin films are affected by the
vacuum condition during thin film deposition. At low partial pressure the impurities such
as oxygen or argon are incorporated into the film at interstitial sites or at grain boundaries
and contributed to a compressive stress development. Moreover, increasing amount of
oxygen in the system may increase the chance of oxide formation for some metallic films
[16].
Lattice expansion mechanism: It is well known that due to the capillary effect the lattice
parameter of insolated metal particles is smaller than their respective bulk phase.
However, during the growth their lattice parameter increases and approaches their bulk
value. [16]. In the case of thin film, this expansion is inhibited due to adhesion to the
substrate. Therefore during the crystal growth a compressive stress develop inside the
particles. Though due to the weak interaction between film and substrate adhesion is
usually not strong enough and relaxation occurs during the growth.
22

Effect of deposition parameters on the intrinsic stress:
It is observed that the stress mode in the film changes during and after deposition due to
the different factors. However, there are different external parameters during film
processing which can be employed to change the stress trend in the film to attain the
desirable condition. In the next section, the effect of different deposition parameters such
as substrate temperature, working gas pressure and substrate bias on the film stress is
discussed.
Substrate temperature: Substrate heating during deposition is due to four different sources
including heat of condensation, kinetic energy of sputtered atom, plasma radiation and
energetic neutrals and reflected atoms [12]. Thornton et al [71] investigated the effect of

⁄ on the intrinsic stress and demonstrated that at low
⁄ the intrinsic stress
dominates the thermal stress development while for low melting point materials or high
substrate temperatures ⁄ 0.250.3 ) the thermal stress is the dominant factor.
Low melting point materials have high thermal expansion coefficients and low activation
energy and consequently low intrinsic stress as shown in Figure 9.

23

Working gas pressure: The effect of working gas pressure on the film stress was
investigated and the results are shown in Figure 10 [71-74]. It is observed that the low
gas pressure results in the formation of compressive stress while the tensile stress
developed at higher pressures. The compressive stress at low gas pressure is due to the
bombardment of sputtered atoms which peen the film surface [71, 72]. Alternatively, at
high pressure, tensile stress develops due to higher rate of scattering collisions.

Figure 9: Schematic representation of thermal and intrinsic stress [5]

Figure 10 : Effect of working gas pressure on the stress condition [5]
24

Substrate bias: The effect of increasing the bias voltage on the residual stress has been
reported for different materials [9, 75]. It is observed that by increasing the bias voltage
higher rate of argon ions are directly accelerated toward the substrate, therefore
compressive stress develops in the film. However, this transition in stress state from
tensile to compressive mode depends on different factors such as sputtering gas pressure
and material mass. For example, it is shown in Figure 11 that raising the bias voltage
develops more compressive stress in an aluminum film and transition from compressive
to tensile mode occurs at 50 volts bias voltage at 1 mtorr working gas pressure. Though,
in the absence of bias voltage this transition occurs at much higher working gas pressure
(40 mtorr).

Figure 11: The effect of bias voltage on the residual stress on aluminum [9]
25

2.4 Sputter deposition of copper and tantalum films
Nano-structured materials can be produced with a variety of processing methods;
however, bulk nano-structured materials can be made with different techniques such as
ball milling, electrodeposition, sever plastic deformation, PVD and CVD techniques.
The main difficulty in processing of nano-structured materials is that these materials are
not typically stable at high temperature and in some cases the grain growth starts at room
temperature.
Specifically, processing of nano-structured BCC materials are even more challenging
since these materials are very sensitive to the various processing parameters such as
impurities. In this study copper and tantalum films are used, since they present very
different structures (FCC vs. BCC) and characteristics, therefore, we can expand our
knowledge to a wide range of materials.
2.4.1 Copper thin films
The effect of external chamber conditions on the films microstructure and residual stress
of sputtered copper films has been extensively documented [5, 11-13]. Schell et al. [10]
demonstrated that sputter copper films always grow in (111) direction and in contrast to
other metallic films such as silver, aluminum, niobium and tantalum, the texture is
consistent with changing the external sputtering conditions such as power and pressure or
applying the bias (see Figure 12).

26

In addition, the residual stresses evolution in sputter copper films were studied and the
results showed that the in-situ stress changed from compressive to tensile mode as the
pressure increase. However, copper thin films usually deposited in the low stress state as
shown by Pletea et al [8]. It is observed from Figure 13 that the amount of residual
stresses changes from -250 MPa to 250 MPa as the pressure increase. Therefore, in
contrast to other metallic thin films such as tantalum or tungsten, copper films present
low stress state during sputter deposition.

Figure 12: The texture evolution as a function of bias voltage for sputter copper thin films
[10]
27

In this study, copper films are used as a base material to study the dominant factor in
sputter deposition technique and also to investigate the behavior of FCC crystalline
materials. The effect of external chamber parameters on the microstructure and stress
state of copper films are well understood, therefore, we can employ this knowledge to
investigate the effect of the internal chamber conditions such as plasma characterization
(ion flux, ion energy, electron density) and chamber geometry on the film microstructure
and stress state.
2.4.2 Tantalum thin films
Tantalum thin films are attractive in a number of applications such as protective coatings
for corrosion, heat, and wear resistance due to its high melting point, toughness and
immunity to chemical attack. However, the main applications of Ta films are in the
electronics industry such as capacitors, resistors and more recently as diffusion barrier

Figure 13 : In-situ stress development in copper thin films as a function of sputtering pressures. [8]
28

between copper interconnects and silicon substrate [41, 43, 76-78]. This wide range of
applications is due to its different phases and their characteristics. Bulk tantalum always
exits in BCC structure; however, sputter tantalum thin films exhibit two different phases,
body center cubic ( α-Ta) and tetragonal ( β-Ta) which differ in mechanical and electrical
properties. The thermal and mechanical properties of alpha and beta tantalum are
summarized in Table 1.
Table 1: Thermal-mechanical properties of alpha and beta tantalum [41]

Although alpha phase is the stable phase during magnetron sputtering deposition usually
tetragonal or mixed phases are developed. The typical XRD patterns of the alpha and
beta tantalum are shown in Figure 14 [7]. For β-tantalum, only three peaks are observed
which indicate the high degree of texture for this phase ; however, randomly textured
beta tantalum films has been observed in some cases which could be due to the different
structure of beta phases.
Phase Alpha Tantalum Beta Tantalum
Structure BCC Tetragonal
Lattice parameters a=b=c= 0.3305 nm a= b= 1.0194nm, c=0.5313nm
Hardness 200-400KHN 1000-1300KHN
Ductility Ductile Brittle
Resistivity 15-60 µ Ω cm 170-210 µ Ω cm
Thermal stability T
melting point
= 2996°C T
Beta-alpha
= 750-775°C
29

As stated by Jiang et al [7] the structure of beta tantalum is not well characterized,
although various models are proposed to determine the most accurate structure (see Table
2). It has been reported that the most probable structure for beta tantalum is similar to
beta-uranium with density of 16.33 g/cm
3
, conversely, as shown by Catania et al [53]
tantalum can be proceed with distorted A15 structure with density of 16.96 g/cm
3
. The
theoretical calculation by Jiang et al [7] is based on the formation of highly textured
(002) tantalum, though in some cases the texture of beta tantalum film changes
significantly with changing the sputtering conditions [77]. Therefore, it is likely that the
structure of beta Ta varies as a function of sputtering conditions.

Figure 14 : XRD pattern of the alpha tantalum (foil and powder) and beta tantalum [7]
30

To date, processing of alpha tantalum on silicon substrate has been studied extensively,
however, the dominate factors in controlling the phase formation is still not well
understood. It has been reported that the properties and crystal structures of the tantalum
thin films are related to the both processing techniques [79] and processing conditions
such as gas pressure, substrate temperature, bias voltage and substrate materials. Among
these conditions the role of substrate temperature [38, 51, 80, 81] , the bias voltage [31,
39, 43, 52, 53] and the substrate materials [33, 38, 54-57] has been studies, however,
there is not much attention on the processing of alpha tantalum with changing the
deposition parameters such as sputtering power and pressure.
It is shown that increasing the substrate temperature promotes the formation of the alpha
tantalum and phase transformation from beta to alpha occurs at lower temperature in the
absence of oxygen in the system [78]. However, this is not the favorable method since by
increasing the substrate temperature relaxation occurs due to grain growth.
On the other hand, a great effort has been made to synthesize the alpha tantalum by
changing the substrate materials. It has been shown that utilizing different underlayer
Table 2: Proposed models for beta tantalum structure
Models Space group Atom/unit cell Density (g cm
-3
)
β-Uranium P4
2
/mnm 30 16.33
Distorted A15 Pm(-3)n 16 16.96
Supper lattice P4/mmm 56 16.70
Hexagonal P6
3
/mmc 2 16.23

31

with the same lattice match and texture alignment such as aluminum, niobium, titanium
and tantalum nitride would promote the formation of the alpha tantalum. However, the
exact mechanism that leads to the formation of alpha tantalum with changing the
substrate is not clearly indentified [33, 38, 54-57].
Alternatively, the effect of bias voltage and specifically the ion bombardment on the
properties of the deposition film was studied and it was illustrated that increasing the ion
energy promotes the growth of alpha tantalum [39, 43, 53, 82, 83]. Though, there is a
controversy on the amount essential threshold energy to transit from beta to alpha phase.
Meanwhile, surface morphology and stress state of the film were strongly influenced by
applying a bias during magnetron sputtering deposition.
Therefore, the main challenge is to process alpha tantalum without applying external
parameters and to identify the dominate factors in processing of the alpha tantalum with
magnetron sputtering system. This result would lead us to have better understanding
about the microstructure and stress development as a function of sputtering conditions
and consequently enable us to process tantalum films with high strength and low stress
state.
32

Chapter 3. Experimental procedures and materials
This chapter presents the thin films synthesis method and the technique for analysis of the
internal chamber conditions during deposition. Characterization tools to study the films
microstructure, stress development, hardness and resistivity are also presented.
3.1 Sputtering chamber
3.1.1 Set up the sputtering chamber
The first step in processing films is setting up the sputtering chamber. The schematic of
the chamber is shown in Figure 15. This particular high-vacuum chamber has three main
ports which can be used for sputtering sources or for plasma characterization tools such
as a Langmuir probe. Through the combination of a diaphragm pump and a turbo pump,
chamber base pressures of 1×10
-6
and 1×10
-7
Torr can be achieved within 1 to 5 days
pumping, respectively.

Figure 15 : Schematic of sputtering chamber
33

A 33 mm sputtering gun is used in this study with a maximum power of 300 watts and 1
amp current. Many types of materials from metals, ceramics and semiconductors can be
deposited with this type of source with the maximum target thickness of 5.70 mm. The
surfaces of most target materials are oxidized during exposure to air; therefore, pre-
sputtering is performed before each run to prevent film contamination. During sputtering
deposition the geometry of the target changes due to the racetrack formation which may
affect the sputtering rate. However, no significant changes in the sputtering rate have
been observed in this study. Figure 16 shows the racetrack formation in our sputtering
chamber during copper sputtering.

A water cooling system has been installed in order to cool down the gun and the turbo
pump during sputtering. A minimum flow rate of 2.5 liters/min is required for the gun
and 3.7 liters/min for the turbo, both at 16°C temperature. A flow switch has been
installed in the system for safety matter; this allows the system to shut off automatically
whenever the water flow is stopped.

Figure 16 : Racetrack formation during sputtering of copper
30 mm
34

A substrate holder with high accuracy for preventing the formation of a non-uniform film
(shadowing effect) has been designed and built for the sputtering machine. Figure 17
shows the schematic of the substrate holder. This design enables the rotation of the
substrate with a motor during the sputtering which allows the processing of multilayer
and nanostructure materials.

3.1.2 Plasma characteristic
A Single Langmuir probe has been built and placed inside the chamber. This part of the
project included two supervised undergraduate students. The main goal for the probe
construction is to minimize the plasma perturbation and to acquire data accurately. The
smaller the probe tip, the less the plasma perturbation may occur during sputtering. The
tungsten tip with the diameter of 0.175 mm and a tip length of 12.5 mm was chosen for
this study since it provides us with the capability to handle the probe and clean it
frequently between the runs. Figure 18 shows the probe installed inside the sputtering
chamber.

Figure 17 : Schematic of substrate holder
35

During data acquisition, the probe is coated with target material, which results in an
increase in the probe surface area; therefore, the probe tip has to be cleaned after each run
with a diluted HNO
3
solution and rinsed with water. However, this procedure does not
work for the materials which have the similar properties as tungsten tip, such as tantalum.
In the later case, the tip has to be changed frequently to decrease the possible error due to
an increase in tip area.
In order to apply a bias to the Langmuir probe, an adjustable power supply is used. The
probe current is calculated by measuring the voltage drop across a 1k Ω resistor using a
digital Multi-meter. The bias voltage is changed from 40 to -100 V to achieve a steady
state condition which enables us to calculate the ion current saturation. Using a Lab view
interface, the I-V graphs were measured automatically and the plasma characteristic such
as ion flux, electron density and plasma temperature was calculated according to two-
tangent method. Details about data calculation are presented in the chapter four.

Figure 18 : Langmuir Probe inside the chamber
30 mm
36

3.2 Characterization methods
XRD (X-Ray Diffraction)
Thin film diffraction allows the precise determination of the phase formation, lattice
constants as well as the strain state of the thin films. Ex-situ XRD measurement is a
valuable tool for the study of thin films and multilayers while the films are not oxidized
or can be capped with a protective layer which prevents the deterioration of the films in
contact with air.
In the current study, Rigaku Ultima-IV X-ray diffraction machine was utilized to identify
the phase formation and texture development of the sputtered film. Thin film diffraction
was performed using Cu K α radiation with the wavelength of 1.5418Å. A series of θ-2θ
scans from 10 to 90 ° with the rate of 1°/ min were performed to provide the XRD
patterns.
SEM/FIB (Scanning Electron Microscope/Focus Ion beam)
The surface morphologies of the samples were characterized using a high resolution SEM
in a Nova 600 Dual beam FIB (FEI, Tokyo, Japan). To investigate the films
microstructure, cross-sections of the specimens were prepared by focus ion beam (FIB).
The FIB instrument is similar to a scanning electron microscope (SEM), except that the
beam that raster over the sample is an ion beam rather than an electron beam. Secondary
electrons are generated by the interaction of the ion beam with the sample surface and
can be used to obtain high-spatial-resolution images.
37

However, imaging with ion beam usually provides higher contrast compare to electron
beam. Since in this study we were trying to image the grains on the films cross-section,
ion beam was utilized for sample imaging process, though, the images have lower
resolution compare to images taken with electron beams.
TEM (Transmission Electron Microscope)
In order to study the effect of sputtering conditions on the film microstructure,
observations with TEM were conducted to investigate the grain size, the growth
mechanism and the film/substrate interface. A Tecnai F30 TEM (FEI, Tokyo, Japan) was
utilized to investigate the film microstructure.
TEM samples were prepared with FIB in-situ lift up technique for studying the film’s
cross sections. The steps for making the TEM sample are shown in Figure 19. First, the
thin layer of protective layer is deposited with electron beam to protect the surface during
ion beam deposition. Then, a thick layer (~ 1 µm) of tungsten, platinum or carbon is
deposited on the films surface to protect film during the milling process. Next, two
trenches are milled on either side of the tungsten using a large beam current for fast ion
milling. Using the Omniprobe the free lamellae is lifted up and attached to the TEM grid.
The final thinning process will perform on the free standing lamellae.
To study a plane-view microstructure of the samples, a thin layer of the films were
deposited on a 10 nm thick amorphous Si
3
N
4
TEM substrate grid directly. This method
facilitates the TEM imaging and provides us with detail information regarding the
film/substrate interface.
38

Figure 19 : Steps for TEM sample preparation using FIB liftout technique
5 µ
10 µ
39

3.3 Mechanical and electrical testing
Residual stress
Thin film residual stress can be measured with both an ex-situ and in-situ methods. In
this study the ex-situ and in-situ stress was measured with stylus profilometer and Multi-
beam optical stress sensor (MOSS) system, respectively.
Ex-situ stress measurement: Ex-situ stress was measured using an XP-2 stylus
profilometer (Ambios, Santa Cruz, California). The curvatures of sample before and
after deposition were obtained (see Figure 20) and the specimen curvature ( ∆K) is related
to the film stress according to the modified Stoney’s formula [84] as follows:

.
6 .

. 1
.
3.1
Where
8.
and R is the curvature of the radius and L is the length of the
coating. This equation is valid while the ratio of
is less than 5%.

Figure 20: Before and after profile for tantalum sample at 100 watts and 2 mtorr
40

In-situ stress measurement: In order to measure the real-time stress during deposition, a
Multi-beam Optical Stress Sensor (MOSS) system (k-space associates Inc., Ann Arbor,
Michigan) was used (see Figure 21). In this system, parallel beams strike the surface of
the sample and the Charge Couple Device (CCD) camera images the reflected beam as
shown by Ryu et al. [15]. As the film grows, the distance between the adjacent laser
beams changes, due to the more divergent or convergent reflected beams, therefore, the
film curvature is determined from changes in the spacing of the beams. Using a modified
Stoney’s formula [84, 85] , the specimen curvature ( ∆K) is related to the film stress as
follows:

Where < σ> is the total film stress at a given thickness which is calculated as an average
from the values of the displacement points used to track the change in curvature by
MOSS; h
f
and h
s
are the film and substrate thicknesses respectively, and M
s
is the biaxial
modulus of the substrate. Several tests were performed at each processing condition and
at different thicknesses in order to investigate the repeatability of this technique.

f
s s
h
K h M


 
6
2
 (3.2)
41

Nano-indentation
All hardness tests were performed using a Hysitron TriboIndenter (Hysitron,
Minneapolis, Minnesota). The maximum load P
max
, the maximum displacement h
max
and
the stiffness S (dP/dh) which is defined as a slope of upper portion of unloading curve
were calculated from load-displacement curve.
The hardness value derived based on the following formulas:

3.3
Which A represents the area function and determined based on a series of indents at
various contact depths for the sample of known elastic modulus and the contact area A is
calculated using the polynomial equation proposed by Oliver and Pharr for a Berkovich

Figure 21: Schematic of MOSS system
42

tip [86]. In the present study the hardness values were measured at constant loading rate
of 1000 µN/s with the maximum load of 5000 µN using a Berkovich tip.
43

Chapter 4. Plasma characteristics in sputtering chamber
4.1 Introduction
As discussed in chapter 2, during the 1970’s and 1980’s Thornton et al. [5] developed a
comprehensive study on magnetron sputtering, relating the microstructure and residual
stress to the external parameters , however, they did not present unifying parameter to
evaluate stress. Later on, Kelly et al. [87] proposed a three dimensions structure zone
model correlating the ion to atom ratio to the substrate temperature and the bias voltage
in unbalance magnetron sputtering system.
In this chapter, the goal is to incorporate the study of internal plasma conditions (e.g. ion
flux, electron temperature, electron density) and correlate them to external factors and
also film microstructure and residual stress. By studying the effects of plasma conditions
inside the chamber at different locations and its relation to the film residual stress and
microstructure, we will be able to apply our deposition parameters to different chambers
with different conditions. Since copper is a well known material, in this study we used
copper films as a base material to study the plasma characteristics. In the future we can
extend this knowledge to process more complex materials.
4.2 Experimental details
Using a Langmuir probe is a straightforward method to diagnose the particle distribution
functions within the plasma. The probe is usually placed close to the target but it should
be far enough to prevent it from the melting. As discussed in chapter 3, a single
44

Langmuir probe was built and installed inside the sputtering chamber. The base pressure
of the vacuum chamber was lower than 2 ×10
-3
Pa and 99.999% argon was used as a
sputtering gas. To study the plasma conditions inside the chamber 99.995% copper
targets were used. The plasma characteristic such as ion and electron density, flux ratio
and plasma potential have been calculated according to I-V graph derived at different
sputtering pressures ranging from 0.3 Pa to 2.4 Pa while the sputtering power was kept
constant at 100 watts.
To study the films microstructure, 70nm thickness copper films were deposited on the 10
nm thickness amorphous Si
3
N
4
substrate TEM grids and the microstructures were studied
with FEI Tecnai F30 TEM. In addition, the residual stresses of the 500 nm thick copper
films were measured with XP-2 stylus profilometer.
4.3 Results and discussion
4.3.1 Ion flux and energy
Figure 22 shows a typical I-V curve for a single Langmuir probe. When the bias voltage
on the probe is adequately negative with respect to the floating potentials, the sheath
around the probe will be saturated with ions so the probe reaches the ion saturation
current. In the transition region, however, the probe starts to attract to the electrons and
more energetic electrons reach the probe. Eventually when the bias voltage reaches the
plasma potential, the probe shows the electron saturation current.

45

The ideal plasma conditions to use a Langmuir probe is when the electrons and ions are
fully thermalized at equal temperatures and when the particles mean free paths exceed the
Debey length. Under this condition the electron-electron collision is small and the
electron-gas cannot be expected to be in thermal equilibrium. Therefore, the electrons in
the plasma can be approximately described by an electron temperature. Meanwhile, in
magnetron sputtering system, the concern is whether the probe causes major perturbation
of the plasma or not. It has been shown that because of the non-local nature of the
magnetic source, the local perturbation of plasma by probe is minimal and can be ignored
[62].
In an unperturbed plasma, the plasma characteristics can be found by different methods
including graphical methods such as two tangents (TT) and the derivate method and
computational methods such as Orbit motion limited (OML), Allen-Boyd-Reynolds
(ABR) and Bernstein-Rabinowitz-Laframboise (BRL) theory which are based on

Figure 22 : Schematic of I-V graph for single probe [1]
46

different assumption about electron distribution, sheath thickness and ion and electron
density [88, 89].
The two main graphical methods are the two tangents (TT) method and the derivate
method [90, 91]. It has been shown that the electron temperature measured by using the
two graphical methods, is considerably overestimated as compared to analytical methods
[92]. Though for ion density measurement, the graphical method is more reliable since
no assumption is necessary about the density of the ions in the plasma.
In contrast, OML theory is based on the assumption that the gas pressure is low and the
collisions between the electrons and ions are negligible. Consequently, the ions in the
sheath are moving in the free orbits and some of them would end on the surface of the
collector [88]. This theory is based on the isotropic Maxwellian distribution of the
velocities of the ions and atoms with the assumption that ions attract the probe from
infinity. The application of this theory is very restricted since it is based on the
collisionless plasma. This theory can be applied to the conditions of low ion density or
very thin probe radius. In the ABR theory, the potential has been derived everywhere
from probe surface to infinity without dividing the plasma to the sheath and presheath
regions. However, this theory is based on the assumption that the ions move radially
toward the probe and they are not following the orbital motion movement, therefore the
absorption radius is at infinity. Eventually, Bernstein, Rabinowitz and Laframboise have
introduced the BRL theory which accounted for both sheath formation and orbit limit
motion and extended this calculation to the Maxwellian ion density distribution. ARB
and BRL theory are valid for higher ion density or larger probe radii [63, 89].
47

In this study we calculated the data based on the graphical method since this method is
valid for both high and low ion plasma density conditions [90]. The two tangent method
involves taking the semilog plot of I-V graph and fitting a straight lines through the
retardation and electron saturation regions. From the intersection of these two straight
lines we can determine the plasma potential and electron saturation current. Also we can
obtain the ion saturation current by making a straight line fit to the left side of the graph.
Ion density can be derived from following formula having the ion saturation current:

Where I
is
is ion saturation current e is electron charge, A
p
is probe area and v
i
is equal to:
8
4.2
This calculation is based on the assumption that we have Maxwellian distribution and the
ion temperature is comparable to electron temperature [93].
The measured variation in ion density as a function of sputtering pressure for our copper
samples is shown in Figure 23.

p i i is A v en I
4
1

(4.1)
48

Meanwhile, in order to study the effect of material characteristics such as deposition rate
and molar mass on the plasma condition, the ratio of the metal flux to the ion flux was
calculated. Ion flux can be derived according to following formula including the ion
saturation current:
1
4
4

  4.3
Where I
is
is the ion saturation current, A
p
is the probe area, e is the electron charge; while,
metal flux is defined by following formula:

4.4
Where R
d
is deposition rate, N
A
is Avogadro number, ρ is density and M is atomic mass.
Therefore, the flux ratio is obtained according to following formula:

Figure 23 : Ion density as a function of sputtering pressure. Notice that ion density increase
almost linearly as the pressure increase.
49

4.5
where M
u
is equal to 1/N
A
[88, 94]. The flux ratio for copper sample as a function of
sputtering pressure is shown in Figure 24.

It is observed from the above figures that both the ion density and flux ratio increased
linearly with increasing pressure. This trend is expected since the rate of ionization
increased due to the higher volume of the gas inside the chamber. However, the
deposition rate is not increasing at the same ratio as the ionization increase due to the
higher rate of collisional scattering as the pressure increase.

Figure 24 : The ratio of metal flux vs. ion flux as a function of sputtering pressures.
50

In addition, in order to study the kinetics inside the chamber the plasma potential is
measured from the I-V graph. In the absence of the bias, the plasma potential represents
the ion energy inside the chamber during sputtering.
It is observed from Figure 25 that the plasma potential shows a nonlinear behavior as a
function of sputtering pressure. It is expected that by increasing the pressure, the
bombarding energy decreases due to the higher gas scattering; however, at 0.7 Pa
sputtering pressure, a small increase in the ion energy observed and it reaches a steady
state condition thereafter. This behavior could be related to the specific sputtering
chamber geometry and the peak might shift to a different pressure depending on the
chamber size and geometry.

Figure 25 : The plasma potential as a function of sputtering pressure. Notice that in the
absence of bias voltage the plasma potential is equal to ion energy.
51

4.3.2 Microstructure and residual stress
In order to have better understanding regarding the effect of internal chamber conditions
to the film properties, we studied the copper film’s microstructure and residual stress as a
function of sputtering gas. The samples were deposited at 100 watts sputtering power
and different pressures ranging from 0.3 to 2.8 Pa and the plan view TEM images are
shown in Figure 26.

Figure 26 : Plan view TEM images for the copper films deposited at 100 watts sputtering
power and pressures ranging from (a) 0.3 Pa (b) 0.7 Pa (c) 1.4 Pa (d) 2.8 Pa.
52

We noticed that the film microstructure at 0.3 Pa sputtering pressure consists of the large
grains (~ 300-400nm) and also small grains (~10-20nm). This wide range of grain sizes
could be due to the different effects such as recrystallization and grain growth during the
cooling process. The more detail discussion about the effect of substrate heating and
grain growth is presented in chapter 5. At 0.7 Pa sputtering pressure, the microstructure
is very uniform with the average grain size of ~ 20nm. By increasing the sputtering
pressure to 1.4 and 2.8 Pa the films start to crack while the average grain size increased to
~40 nm.
Comparing the films microstructure with the plasma characteristics, we noticed that
increasing the ion energy (plasma potential) at 0.7 Pa sputtering pressure may stabilized
the microstructure and results in the formation of dense film. However, with decreasing
the ion energy at 1.4 Pa sputtering pressure the films start to crack due to the effect of
less atomic peening.
Meanwhile, the total stresses of the films after deposition with the maximum thickness of
500 nm were measured and the results are shown in Figure 27. It is observed that by
increasing the pressure to 0.7 Pa, higher tensile stress develops in the film, yet by
increasing the pressure to 1.4 Pa and 2.8 Pa a compressive stress trend is observed. To
date, no single model has been proposed that can be account for developing tensile stress
during sputtering deposition. However, it has been shown that the main factors could be
kinetics and thermal energy.

53

It is generally observed that in the absence of ion bombardment or low deposition
temperature which favors low mobility of condensing atoms, the film is characterized
with voided microstructure and tensile stress state. By increasing the ion bombardment, a
compressive stress develops depending upon the ion to atom arrival ratio with a dense
microstructure. Therefore, it is expected that by increasing the sputtering pressure, the
tensile stress increase due to more collisional scattering and less ion bombardment. It is
observed from Figure 27 that the residual stress decreases by increasing the pressure to
1.4 Pa. This stress relaxation could be due to the formation of cracks (see Figure 26),
though, it is not clear whether these cracks develop due to the high tensile stress state or
thermal relaxation after growth. More detail discussion in regard to crack formation is
presented in chapter 5.

Figure 27 : Residual stress as a function of sputtering pressure for copper films. Notice that at
1.4 and 2.8 Pa sputtering pressure the films are cracked.
54

Overall, comparison between plasma conditions, microstructure and residual stress
demonstrates that at 0.7 Pa sputtering pressure films with dense microstructure without
cracks or any other defects develop which could be due to the small increase in the
plasma potential. This change in plasma trend is related to the chamber geometry and
should change from chamber to chamber since the bombarding energy and scattering of
the ions and atoms depend on the volume of the chamber and the its specific geometry.
4.4 Summary
The correlation between the plasma characteristics such as ion density and plasma
potential with film microstructure and residual stress has been studied. It is observed that
films properties are highly correlated to the kinetics inside the chamber. Specifically,
increasing the ion energy at 0.7 Pa sputtering pressure in our sputtering chamber
stabilized the microstructure and results in the formation of dense film without few
defects. Meanwhile, the kinetics inside the chamber are related to the chamber geometry
and it changes from system to system. This study provides us with valuable information
regarding our chamber geometry and therefore we can extend our knowledge to process
more complex materials such as alpha tantalum (see chapter 6).
55

Chapter 5. In-situ stress evolution in sputtered copper and tantalum
films
5.1 Introduction
Residual stress development during thin film growth limits the range of available
deposition methods due to crack formation (see chapter 4) and/or de-adhesion of the
deposited film. As discussed in chapter 2, the residual stress evolution strongly depends
on the sputtering conditions such as sputtering gas pressure, deposition rate and bias
voltage since each of these has a significant effect in the stress mode and in the film
characteristics [71, 95-101]. The investigation of stress evolution in thin films is
commonly performed by ex-situ techniques such as profilometry [9, 72, 102, 103];
however, in-situ stress measurements can provide valuable information about the effect
of sputtering conditions on the stress state as a function of the coating thickness and time
[8, 15, 28, 29, 104]. The measurement of stresses using in-situ techniques can be divided
in two major groups: 1) diffraction methods such as X-ray or electron diffraction which
characterize the crystalline structure of materials using the position, broadening and
intensity of the diffraction peak [105, 106], and 2) techniques that measure the bending of
the substrate or changes in the film curvature such as Cantilever or Multi-beam Optical
Stress Sensor (MOSS) techniques [22].
The residual stresses in sputtered thin films are composed of growth and thermal stresses.
In this study, the growth stress represents the real-time stress evolution during the
56

deposition, such as intrinsic stress, thermal stress and other stress contribution relieved by
relaxation mechanisms such as for instance mechanical annealing. The growth stress
must be measured in-situ in real-time during deposition, so that it can be separated from
effects such as thermal relaxation due to cooling of the sample after deposition. The total
residual stress is the combination of both the growth stress and thermal stress after
deposition and can be measured both with in-situ and ex-situ techniques. [12, 68, 71, 73,
107].
A wide range of sources for the development of residual stress have been suggested [12,
16, 71, 108-110]. Thornton et al. [71] described the residual stress during the growth as a
combination of intrinsic and thermal stress, where the intrinsic stress reflects the
measured stress within the film due to growth-related processes, e.g., recrystallization,
grain boundary formation, dislocations and defects such as voids and impurities [8, 16,
23, 111]. Koch [16] describes the thermal stress as an extrinsic stress which usually
arises when there is a change in temperature after film deposition. For real-time stress
evolution, previous studies by Pletea et al. [8, 28, 29] calculated the intrinsic stress of
various films by subtracting the effect of temperature measured by a thermocouple. In
their study, the measured temperature increase was estimated to be less than 14°C;
however, the in-situ stress-thickness vs. time curves, appeared to have a thermal stress or
thermal contribution values which are much higher than those reported by the authors.
In this chapter, in-situ (MOSS) measurements are used for determining the stress during
growth and the possible thermal contributions due to temperature changes after
57

deposition. First, the stress evolution is studied during deposition as a function of
thickness and time for tantalum and copper samples, which represent high and low
melting point materials as well as a low and high adatom mobility, respectively. In-situ
(MOSS) measurements are used for determining the total residual stress during
deposition at a given thickness while profilometry is used to determine the total residual
stress after deposition at the final film thickness (~ 400 nm). The surface morphology
and microstructure are also investigated in order to study the correlation between the
residual stress and film structure.
In addition, the contribution of thermal and relaxation mechanism after growth to the
total value of the residual stress was studied. We report the results using in-situ and ex-
situ techniques for tantalum and copper samples. Specifically, the growth stress
developed as a function of thickness and sputtering pressure is studied and compared to
the total residual stress measured by ex-situ techniques.
5.2 Experimental Details
Copper and tantalum films, 0.4 and 0.5 microns thick, were deposited on a 250 μm thick
silicon (100) grounded substrate using a DC- magnetron sputtering system. The base
pressure of the vacuum chamber was lower than 2 ×10
-5
Pa and 99.999% argon was used
as a sputtering gas. The sputtering pressure was increased from 0.3, 0.7 to 1.4 Pa, and the
sputtering power was varied from 100 to 200 watts, while the distance between the target
and sputtering source was kept constant at 15 cm. The summary of the samples is shown
in appendix B and D.
58

Using an XP-2 stylus profilometer (Ambios, Santa Cruz, California) the substrate and
film thickness were measured. In order to measure the real-time stress during deposition,
a Multi-beam Optical Stress Sensor (MOSS) system (k-space associates, Inc., Ann Arbor,
Michigan) was used. Using a modified Stoney’s formula [84, 85] , the specimen
curvature ( ∆K) is related to the film stress. Several tests were performed at each
processing condition and at different thicknesses in order to investigate the repeatability
of this technique. The film temperature at the end of the growth run was calculated from
the change in curvature as the sample was cooled down inside the vacuum chamber,
assuming no other relaxation mechanism occurs after the film deposition stops. Figure
28 illustrates the growth and cooling graph for copper samples deposited at 0.3 and 1.4
Pa sputtering pressures. The dashed line represents the time at which the sputtering
source was turned off, thus noting the end of the stress evolution during the growth and
beginning of the cooling period.

Figure 28: Growth and cooling graph for copper samples deposited at 0.3, 0.7 and 1.4 Pa
sputtering pressure. The cooling graph represents the thermal stress contribution and
temperature rise during growth.
59

The substrate temperature increase ( ∆T) during deposition was calculated according to
following formula, by using the change in the stress-thickness ( ∆σh
f
) during the cooling
process,

where ∆α is the difference of the thermal expansion coefficients of the film and
substrate. By using the stylus profilometer the curvature of samples before and after
deposition were obtained and the total residual stress after deposition was calculated
according to the modified Stoney’s formula. The surface morphology was studied using
SEM. Furthermore, sample cross sections were prepared and investigated using high
resolution scanning electron microscopy in a FEI Nova 200 Dual beam FIB. The phase
compositions were studied using Rigaku Ultima-IV X-ray diffraction (XRD) using Cu K
α

radiation with a wavelength of 1.5418 Å. The microstructural analysis was performed on
samples obtained at different sputtering pressures as well as various sputtering powers.
5.3 Results and discussion
As presented in the introduction, understanding the stress evolution during film growth
for magnetron sputtering still lacks sufficient studies. For example, in Figure 28 during
growth, the stress is highly compressive for the sample sputtered at 0.3 Pa, however,
when the deposition is stopped and the sample begins to relax, the growth stress becomes
tensile. Therefore, in-situ measurements techniques provide a completely different view
α M h
h σ
T
f f
f
Δ
) ( Δ
= Δ (5.1)
60

of the stress during growth as compared to ex-situ measurements of the total stress after
growth.
In Figure 29, the stress evolution during and after deposition measured in-situ using
MOSS is presented. Both copper and tantalum films (Figure 29 a and b, respectively)
were sputtered at 100 watts at various sputtering pressures. These measurements allow
for the comparison of the stress evolution of two sputtered materials have very different
melting points (T
m Cu
= 1083°C, T
m Ta
= 3017°C) and adatom mobility. The stress values
presented in Figure 29 are the total stress ( 
t
) values at a given thickness, where 
t
is
composed of growth ( 
g
) and thermal stress ( 
th
) components.

61

5.3.1 Effects of sputtering pressure
For the copper samples (Figure 29a), beyond the complexity in the initial stage of
deposition, the growth stress generally changes from compressive to tensile with

Figure 29: Film stress vs. thickness plots at 100 watts sputtering power calculated using MOSS for
(a) copper and (b) tantalum samples sputtered at 0.3, 0.7 and 1.4 Pa. Insets present a magnified
view of the film stress vs. thickness for the first 20 nm.
62

increased pressure. For the lowest sputtering pressure (0.3 Pa), the stress value has a
tensile maxima around 10 nm (see inset) which is typical for Volmer-Weber growth after
which the stress remains compressive over the entire growth run. The corresponding
surface morphology after growth is shown in Figure 30a. At a sputtering pressure of 0.7
Pa, the film is initially compressive reaching a tensile stress maxima at around 100 nm
after which it continues to growth under a tensile stress; a smooth surface was observed
after growth (Figure 30b). At the highest sputtering pressure (1.4 Pa), a tensile stress is
observed and maintained throughout the entire thickness reaching a maxima at around 50
nm thickness, the final surface morphology is cracked (Figure 30c) indicating a high
residual stress state [112]. The XRD spectrum shown in Figure 30d does not demonstrate
a significant change in texture as a function of sputtering pressure. By increasing the
pressure, less energetic particles strike the film surface due to more scattering and this
effect is believed to contribute to the increased tensile stress that develops in the film
[71].
The observed tensile stress of the copper films with increased sputtering pressure is not
in agreement with previous results on evaporated copper films [25] demonstrating a
compressive stress trend for low melting point materials, but do appear to follow
somewhat previous magnetron sputtering data [8]. In this study, the sputtering rates for
the three sputtering conditions are similar but not identical since the power was kept
constant; this is in comparison to the study by Pletea et al. [8] for which the power was
modified in order to maintain the same sputtering rate.
63

Figure 30: Scanning electron micrographs of the surface morphology for copper films
processed at 100 watts sputtering power as a function of sputtering pressure at (a) 0.3, (b) 0.7
and (c) 1.4 Pa, (d) XRD spectra of the films as a function of sputtering pressures.
64

The in-situ stress curves after deposition were taken inside the chamber for a total of 10-
20 minutes after deposition was stopped. These curves show a tensile trend for all three
samples; however the trend is more pronounce for the sputtered films at 0.3 and 0.7 Pa.
As described by Koch [113] we also observed that for the sample sputtered at 1.4 Pa the
trend is minimal since the sample was always under tensile stress. However, it appears
that the stress curve for this samples 10 min after deposition is turning to compressive
stress mode, but since the film cracked and we are unable to confirm the final stress state
by profilometry.
The tantalum samples represent a high melting point material and in contrast to the
copper samples, showed significant changes in stress evolution as the pressure increased.
At the low sputtering pressure (0.3 Pa), a tensile maxima is observed at around 2 nm film
thickness followed by a high compressive stress with a maxima of -1.2 GPa. At 0.7 Pa
sputtering pressure, a tensile maxima is observed at around 8 nm thickness followed by a
tensile stress trend. At the highest sputtering pressure (1.4 Pa), a high tensile stress is
observed with a maxima at around 50 nm film thickness. Thornton et al. [71]
demonstrated that the transition from compressive to tensile stress mode for tantalum
should occur at high sputtering pressures (>2.8 Pa); our results demonstrate a similar
transition though it occurs at a lower sputtering pressure (0.7 Pa) which could be
attributed to different phase formations in tantalum (alpha or beta phase).
The in-situ stress curves after deposition show a tensile trend for the sample deposited at
the lowest sputtering pressure (0.3 Pa) while smaller stress changes were noted for both
65

the 0.7 and 1.4 Pa films. It should be noted that the sample sputtered at 1.4 Pa seems to
go towards a compressive, however; the film also cracked and we are unable to confirm
the final stress state.
SEM images of the tantalum films shown in Figure 31 demonstrated a smooth
morphology at low sputtering pressure and similar to the Cu sample, the film cracked at
highest sputtering pressure (1.4 Pa). However, in contrast to the Cu films, the XRD
spectrum of the Ta films (Figure 4d) demonstrates a phase composition change as the
sputtering pressure is increased. Previous studies have noted that sputtering tantalum on
the silicon substrates results in the formation of the highly textured (002) beta phase and
that increasing the pressure does not have major influence on the phase formation [56].
However, our results show the formation of mixed phases as a function of pressure. For
example, the surface morphology of tantalum deposited at a low sputtering pressure (0.3
Pa) shown in Figure 31a, contains mixed phases of (002) texture beta phase with
spherical particles containing (110) texture alpha phase facets [43, 114]. At 0.7 Pa
sputtering pressure the needle-like shapes disappear and a conversion from mix phases to
single beta phase is observed with a high (002) texture (see Figure 31b). By increasing
the sputtering pressure to 1.4 Pa, the microstructure is characterized by randomly oriented
beta phase with (400) and (202) texture (Figure 31c). These morphology results agree to
some extent with a previous publication by Zhou et al. [50].

66

Figure 31: Scanning electron micrographs of the surface morphology for tantalum films
processed at 100 watts sputtering power as a function of sputtering pressure at (a) 0.3, (b) 0.7
and (c) 1.4 Pa , (d) XRD spectra of the films as a function of sputtering pressure, noting α
(cubic) and β (tetragonal) phases.

67

5.3.2 Effects of sputtering power
In order to further understand how the initial stress stage affects the overall and final
stress of the film, copper was deposited at 1.4 Pa sputtering pressure at two different
powers (see Figure 32).
It is observed that the tensile stress increased in the film by increasing the power from
100 to 200 watts. For both powers, the samples have a trend towards a tensile regime
which continuously increased during the early growth stage and appeared to plateau after
the film reaches a thickness greater than 200 nm. It is interesting to note that for both
pressures the tensile maxima occurs at around 50 nm thickness suggesting similar growth
mechanisms.

Figure 32: (a) Film stress vs. thickness plots for copper samples deposited at 1.4 Pa sputtering
pressure measured with MOSS technique as a function of power including surface morphology at
(b) 100 watts (c) 200 watts sputtering power.
68

The surface morphology of the samples (Figure 32b and c), showed that cracks are
present for the film deposited at 100 watts, while, the sample processed at 200 watts is
characterized by having a very dense microstructure with large particles. As discussed
earlier the sample sputtered at 1.4 Pa the film also cracked and we are unable to confirm
the final stress state, but, in the case of the film deposited at 200 watts, for which sample
did not break, we were able to confirm a total stress after deposition from profilometry at
around 286 MPa which indicates that the film compressive stresses relaxed since the this
value is more tensile than at the instantaneous stress measured at the end of deposition (~
200 MPa). It should also be noted that by doubling the sputtering power, the deposition
rate increases from 1.7 to 3.6 Å/s which could account for the drastic changes in
morphology.
For comparison to the copper films, tantalum samples were deposited at different powers
at a 0.7 Pa sputtering pressure as shown in Figure 33a. This pressure was selected since
it presented a single beta phase composition (see Figure 31d). At 100 watts the film
showed a growth stress transition from tensile to compressive while at 200 watts the
opposite transition is observed, thus leading to a convergence of the stress values at a
thickness of ~ 400 nm. The surface morphology of the tantalum samples deposited at
100 and 200 watts (Figure 33 b-d) showed no significant changes with increasing power;
the deposition rate of tantalum films increased from 1.2 to 2.3 Å/s as the power increased.
It should be noted that both sputtering powers resulted in the same texture alignment and
phase.
69

5.3.3 Thermal and relaxation stress evaluation
As mentioned in the introduction, the residual stress in thin films during deposition is the
combination of growth stress and thermal stress and can be expressed as follows:

5.2

Figure 33: Film stress vs. thickness plots for tantalum samples deposited at 0.7 Pa sputtering pressure
measured with MOSS technique as a function of sputtering power including surface morphology at (b)
100 watts (c) 200 watts sputtering power.

70

The residual stress after growth, can be presented as a modified Koleshko et al. equation
[115] shown in eqn. 5.2:
,,
, ,, (5.3)
Where T is temperature during cooling and T
0
is the deposition temperature, σ is stress
and is the amount of stress relaxation after a given time t. Therefore, after growth is
stopped, the total residual stress in the film narrows down to two components, thermal
stress due to the substrate temperature change, and stress relaxation , which is
changing as a function of time. The amount of stress relaxation during and after film
growth is due to two major mechanisms including plasticity such as dislocation glide,
dislocation climb, diffusional creep and grain boundary sliding, and changes in the
microstructure after deposition such as recrystallization and grain growth.
5.3.3.1 Thermal relaxation
In this study the thermal stress has been calculated according to the changes in the film
curvature after growth is stopped and with the assumption that the stress relaxation is
mainly due to the thermal stress relieve. The role of other relaxation mechanisms will be
discussed extensively later. Given that the tantalum films presented in Figure 31a and c,
show multiple phases (alpha and beta) with different thermal expansion coefficients ( α
phase
= 6.9 ± 0.9×10
-6
/K , β
phase
=6.0 ± 0.8×10
-6
/K) [116], the assessment of the
contribution of thermal stress to the total residual stress is very challenging for the
tantalum samples. Therefore, this discussion will focus on thermal stress evolution in
copper films, exclusively.
71

The thermal stress is estimated by monitoring the change in curvature in-situ when the
film growth is terminated as shown in Figure 28, and the substrate temperature change
was calculated according to eqn. 5.1. In general, measuring the temperature rise during
deposition is not trivial for magnetron sputtering. Previous studies measured the
temperature rise during the deposition by attaching the thermocouple directly to the
substrate [8, 28]. However, these measurements can be problematic due to thermocouple
contact issues and possible coating of the thermocouple.

Table 3 presents a summary of the stress measurements for both materials (copper and
tantalum) at 0.3, 0.7 and 1.4 Pa sputtering pressures at a final thickness of ~400-500 nm.
The growth stress reported in Table 3 is the value measured in-situ using MOSS at the
maximum thickness before deposition is stopped. The thermal stress can be measured
from the difference between the growth stress and the total stress measured in-situ after
the sample cools. We used eqn.5.1 to calculate the temperature rise during film
deposition from the measured thermal stress. It is observed from the results in Table 4
Table 3: Summary of stress measurements for copper and tantalum films deposited at 100 watts
sputtering power with thicknesses of ~ 500 nm
Materials Pressure (Pa)
Growth stress
calculated from
MOSS(MPa)
Thermal stress
(MPa)
calculated from
MOSS
Total stress
calculated from
profilometer
(MPa)
Copper
0.3 -195 262 151
0.7 197 144 347
1.4 63 16 —
Tantalum
0.3
-1253 25 -1208
0.7
41 19 131
1.4 433 5 —

72

that the temperature changed significantly with increasing the pressure for copper
samples; however, tantalum samples do not demonstrate significant changes.

The higher thermal stress observed in the copper samples is partially due to a larger
difference between the thermal expansion coefficients of copper and silicon as compared
to tantalum and silicon. On the other hand, the thermal stress decrease with increased
sputtering pressure for both copper and tantalum samples as can be seen in Table 3. In
order to understand this result, the total energy supplied to the surface of the film during
deposition was calculated according to the analytical model introduced by Detor et al. [1]
(see Table 5). It is observed, that the total energy of the sputtered atoms and gas neutrals
decreases with increasing pressure, thus, this could explain the decrease in thermal stress.
This observation is consistent with previous deposition theories proposed by Thornton
[12].

Table 4 : Summary of substrate temperature increase for copper and tantalum films deposited at
100 watts sputtering power with thicknesses of ~ 500 nm
Materials Pressure (Pa)
Temperature
increase
calculated from
MOSS (°C)
Copper
0.3 98
0.7 54
1.4 6*
Tantalum
0.3
24
0.7
20
1.4 5*
*The calculated temperature increase is based on the assumption that the cracks develop after cooling.
73

Other stress relaxation mechanism
As mentioned earlier, the thermal stress calculation presented is based on the assumption
that there are no other relaxation mechanisms during the cooling process. However, it is
important to discuss the contribution of other possible relaxation mechanism. Pletea et al
[8] proposed two different mechanisms for stress relaxation after deposition is ceased: 1)
the short-term reversible stress relaxation which is due to the increase in surface chemical
potential as observed by Chason et al [23] and 2) the long-term stress relaxation which is
due to the plastic deformation or changes of the microstructure after growth. Short-term
stress relaxation is not expected in this study since the samples cooled down in the
chamber for long periods of time. Nevertheless, the long-term stress relaxation might
develop in the film during cooling.
To further investigate the relationship between microstructure and the total stress
measured during deposition, cross sectional FIB images of the copper samples as
function of pressure are shown in Figure 34. The samples have a thickness of ~ 3µm to
Table 5: Total energy supplied to the film during deposition at a constant sputtering power
(100 watts) calculated from Ref. [1].
Materials Pressure (Pa)

Deposition rate
(Å/s)

Voltage (V)
Total energy
(eV/atom)
Copper
0.3 1.94 479 60
0.7 1.89 437 44
1.4 1.7 430 28
Tantalum
0.3 1.18 275 79
0.7 1.08 225 45
1.4 0.97 210 22
74

facilitate the imaging process and the investigation of the growth mechanisms. It is
observed that for the samples deposited at 0.3 and 0.7 Pa (Figure 34 a-b), the grains grew
significantly with some twin formation at the interface between the silicon substrate and
copper film. However, as the thickness increases, columnar grain growth is observed
which remains constant until the final film thickness is reached. In contrast, for the
sample processed at 1.4 Pa sputtering pressure (Figure 34 c), smaller grains formed at the
interface and followed by columnar growth as the thickness increases. The formation of
large grains at the substrate/film interface could be due to grain growth during deposition
[117]. By calculating the total energy supplied to the surface of the film during
deposition according to the analytical model introduced by Detor et al. [1] , it was
determine, that the total energy of the sputtered atoms and gas neutrals increases as the
pressure decrease. This energy increase could lead to an increase in substrate
temperature during deposition which could explain the formation of larger grains at lower
sputtering pressures and formation of smaller grains at higher sputtering pressures. In
electroplated copper thin films it is expected that significant coarsening occurs at room
temperature after deposition [118]; in contrast, sputtered copper films generally have
stable microstructures at room temperature and to the authors’ best knowledge there is
hardly any information on the room temperature grain growth for sputtered copper.
Therefore, formation of large grains at the interface of the substrate and film could be due
to the effect of a temperature increase during deposition or to room temperature grain
growth which results in the large amount of stress relaxation at low pressure for the
75

copper samples. Since no final conclusion on the mechanism could be obtained in this
study, further investigations on the grain growth are part of ongoing research.

In the case of relaxation due to plastic deformation, Murakami [119] showed that the
stress relaxation such as dislocations glide or grain boundary diffusional creep develops

Figure 34: Focus Ion Beam (FIB) cross section images of copper samples deposited using 100 watts
sputtering power at (a) 0.3, (b) 0.7 and (c) 1.4 Pa sputtering pressure and. Note the grain size
changes as a function of sputtering pressure and thickness.
76

when the thermal stress exceeds the elastic limit of the film. However, for copper
samples with a grain sizes similar to our films, the yield stress is expected to be higher
than 300 MPa as presented by Carlton et al. [120]. Given the values of the total stress
after deposition, independent of whether we assume all thermal stress or other relaxation
mechanism, the maximum stress after growth is well below the reported yield limit as can
shown in Table 3.
5.3.4 Total residual Stress Evaluation
As presented in last section, the contribution of stress relaxation after deposition is
significant and has to be considered. However, studying the residual stress utilizing the
in-situ techniques such as MOSS system is challenging especially for monitoring the
residual stress in thick films. Therefore, the ex-situ technique was employed to calculate
the total residual stress at different thicknesses and to investigate the contribution of the
stress relaxation to the total residual stress. In order to validate the accuracy of this
measurement, the total stress evolution in films were studied with ex-situ and in-situ
techniques and the results presented in Table 3. As previously discussed, total residual
stress were not measured in samples which showed cracking since as discussed by Nix
[121] the curvature of a film/substrate changes due to the crack formation.
The total stress from the profilometer measurements should be similar to the value of the
growth stress plus the thermal stress measured in-situ, if there were no other sources of
relaxation such as surface oxidation after exposing the sample to the atmosphere. Li et al
[122] have shown that the rate of copper film oxidation is not significant at temperature
77

below 150°C and it is depending on the time that the samples are exposed to air. Similar
to Cu, Ta has a thin native oxide which is stable below 300°C [123]. Since the ex-situ
stress measurements in this study were completed immediately after the samples were
removed from the chamber, surface oxidation is not expected to be a significant
contribution. However, as shown in Table 3, a slight difference between the ex-situ
measurement and total in-situ measurement for 0.3 and 0.7 Pa sputtering pressure is
observed. A possible source for this small discrepancy between the values presented
could be due to the much higher sensitivity of the MOSS technique compared to the
profilometer. The MOSS system has a very high sensitivity and the standard deviation of
the spacing between adjusted laser points is usually less than ~5 %; the average stress
measured with profilometer has standard deviation of ~10%.

Figure 35 : Comparison between growth stress measured in-situ using MOSS technique and the
total residual stress measured ex-situ utilizing stylus profilometer. The difference represents the
thermal relaxation for copper sample deposited at 0.7 Pa sputtering pressure as a function of
film thickness.
78

As mentioned earlier the ex-situ techniques present a good estimate of the total residual
stress evolution after deposition. To further study the expected thermal relaxation as a
function of thickness, Figure 35 compares the stress measured in-situ during growth to
measurements made ex-situ by profilometry for samples processed to different
thicknesses. The solid line in Figure 35 is a representative real-time stress curve for a
copper film deposited at 0.7 Pa; the squared symbols represent the data from profilometry
on samples at a given thickness. At a thickness of 100 nm there is not much deviation
between the growth and the total residual stress values; whereas, at 200 nm the difference
between these two values increases and can be attributed to an increase in the thermal
relaxation as a function of deposition time. The microstructure of the films (see Figure
34) demonstrates no significant changes in grain size at film thicknesses greater than
200nm, thus, suggesting that the thermal relaxation during deposition reaches a steady
state condition and does not continue to increase with film growth. This observation
validates the assumption that grain growth at the interface is due to the thermal effects.
5.4 Summary
The changes in the stress and surface morphology as a function of pressure and power
were investigated for high and low melting point materials having a high and low adatom
mobility. It was shown that by increasing the sputtering power, a higher tensile stress
develops in the copper films; while, the tantalum films demonstrate the opposite
transition from tensile to compressive stress states. For Cu we observed a decrease in
grain size and a shift to higher thicknesses for the tensile stress maxima as the sputtering
pressure is increase. For Ta the tensile stress maxima shift to higher thicknesses as the
79

sputtering pressure is increase; however, the microstructure comparison is inconclusive
since the three sputtering pressures lead to three different phase compositions (beta and
mixtures of alpha and beta).
By using in-situ monitoring, the stress during growth and after deposition can be
measured. The stress after deposition was determined to be mostly from thermal
contributions, other relaxations mechanisms appeared to be minimal. Agreement
between the values of the total residual stress measured by profilometry and the in-situ
values for growth stress plus thermal stress, demonstrate the major contribution of the
relaxation stresses after deposition. It was concluded that for both materials the main
relaxation mechanism after growth was thermal; however, the effect was more
pronounced for the lower melting temperature material (Cu). Overall, these results
emphasize the value of in-situ stress monitoring for understanding the competing stress
contributions from growth, thermal and other relaxation mechanisms after deposition.
80

Chapter 6. Synthesis of nano-structured alpha and beta tantalum
6.1 Introduction
In this chapter we emphasized on the processing of alpha tantalum by taking into account
the effect of internal and external chamber conditions on the film characteristics. To date,
synthesis of Ta thin films has attracted considerable interest due to its varied properties
and applications in semiconductor industries and for microelectronic devices [1-3].
Magnetron sputtering is one of the most common processing techniques for Ta thin films;
however, films grown by this method can have two different crystal phases: alpha or beta.
The stable body centered cubic (BCC) alpha phase is commonly used for thin film
semiconductor interconnects and most recently as a diffusion barrier in integrated circuits
[1-4] , while the tetragonal metastable beta phase is used for heaters and resistors [5, 6].
In general, magnetron sputtering deposition leads to metastable beta Ta formation [1-5, 7-
23], while the stable alpha phase is typically acquired by heating the substrate [12, 24],
by adding a bias [1, 13, 18, 25, 26] or by changing the substrate material[8, 12, 27-30].
The most trivial method to synthesize alpha Ta is by heating the metastable beta phase
above 750° C where a transformation from beta to alpha occurs [7, 31]. Applying a bias
voltage also promotes the growth of alpha Ta by increasing the ion bombarding energy
[9, 13, 18, 26, 32] . Alpha Ta formation has also been reported by changing the substrate
materials or utilizing different underlayers such as aluminum, niobium or titanium [8, 12,
27-30]. However, the exact mechanism that lead to the formation of alpha Ta by this
81

method is not clearly indentified and has been mainly attributed to lattice match or
texture alignment [8, 12, 27].
Although alpha Ta has been processed by the above methods, these procedures are not
always favorable approaches for understanding Ta film deposition. For example, the
addition of a heating step during deposition could lead to significant amounts of grain
growth which is likely to affect the overall stress state. Clevenger et al [7] showed that
heating a Ta film during magnetron sputtering causes a drastic change towards tensile
stresses, while Detor et al [33] demonstrated that surface morphology and stress state
were strongly influenced by applying a bias during deposition in magnetron sputtered Be.
To date, the dominant factors promoting the formation of alpha Ta are not yet
understood. Current studies are typically not reproducible and depend on many
parameters which can change from system to system. Therefore, it is important to relate
sputtering conditions such as pressure and power to their effect on the kinetics of atoms
and ions inside the chamber and their correlation to the chamber geometry. In the case of
BCC materials, it is also crucial to investigate the effect of plasma kinetics and low level
impurities during deposition. To this extent, understanding the processing of alpha Ta at
room temperature without applying external parameters could lead us to recognize the
main factor or factors leading to a particular phase formation during deposition.
In this study, we present a comprehensive study about the processing of alpha Ta at room
temperature while controlling the sputtering conditions such as power and pressure. The
influence of sputtering kinetics on the phase formation was investigated as a function of
82

sputtering pressure. The impurities inside the chamber were calculated as a function of
sputtering conditions and their effect on the phase formation was studied. Additionally,
the relationship between surface morphology, texture alignment, hardness and residual
stress were investigated.
6.2 Experimental Details
Ta thin films with the maximum thickness of 0.5 micron were deposited on a 250 μm
thick Si (100) grounded substrate using a DC- magnetron sputtering system. Prior to the
film deposition, the substrates were only cleaned with isopropyl alcohol. The base
pressure for the sputtering chamber was kept less than 2×10
-5
Pa and argon gas with a
purity of 99.999% was utilized during sputtering. The Ta target (99.95%) with 32 mm
diameter was used and the distance between the target and sputtering source was kept
constant at 12.5 cm for all conditions. The sputtering was carried at 100 or 200 watts
sputtering powers with varying sputtering pressures ranging from 0.3 to 1.4 Pa. Samples
having thicknesses of 10, 50, 100, 500 and 1000 nm were sputtered at constant power of
100 watts using 0.3, 0.7 and 1.4 Pa sputtering pressures. A summary of the samples is
shown in appendix D.
The phase formation of as-deposited Ta was identified by Ultima-IV X-ray diffraction
(Rigaku, Tokyo, Japan) using Cu K α radiation with the wavelength of 1.5418 Å. The
surface morphologies of the samples were characterized using a high resolution SEM in a
Nova 600 Dual beam FIB (FEI, Tokyo, Japan). To investigate the films microstructure,
cross-sections of the specimens were prepared by focus ion beam (FIB) liftout technique
83

and observed with a Tecnai F30 TEM (FEI, Tokyo, Japan). All hardness tests were
performed using a Hysitron TriboIndenter (Hysitron, Minneapolis, Minnesota). The
hardness values were measured at constant loading rate of 1000 µN/s with the maximum
load of 5000 µN. Residual stresses were measured using an XP-2 stylus profilometer
(Ambios, Santa Cruz, California). The curvatures of the samples before and after
deposition were obtained and the specimen curvature ( ∆K) was related to the film stress
according to the modified Stoney’s formula [34]. The plasma conditions inside the
chamber were measured using a single Langmuir probe made of tungsten with a 0.17 mm
diameter and 10mm in length [35]. The Langmuir data was analyzed and the plasma
potential was calculated according to two tangent (TT) method [35]. The ion current was
calculated by a graphical method and ion density was derived based on the Bohm sheath
theory [35, 36].
6.3 Results and discussions
6.3.1 Phase formation as a function of sputtering conditions
6.3.1.1 Effect of sputtering pressure
The results of XRD spectra of 500 nm Ta thin films deposited at 100 watts sputtering
power and sputtering pressures ranging from 0.3 to 1.4 Pa are shown in Figure 36. It is
observed that the texture of the films changed significantly as the pressure increased. At
0.7 Pa sputtering pressure the formation of pure (110) alpha Ta is observed, while at 0.5
Pa sputtering pressure the microstructure is characterized by having a 96% alpha Ta
84

composition with a (211) texture alignment. All other processing conditions yielded
either a randomly orientated beta Ta or mixed phases.

Given that the XRD peaks for (110) alpha Ta (2-theta = 38.47) and (202) beta Ta (2-theta
= 38.20) are almost identical, surface morphology analysis was employed to further
identify and differentiate between these two phases since the surface morphology of
alpha Ta is characterized by needle shape particles while beta Ta is described as having
spherical shaped particles [11, 18, 23].

Figure 36: XRD spectra of Ta films deposited at 100 watts sputtering power and different
sputtering pressures ranging from 0.3 Pa to 1.4 Pa. Note the formation of (110) BCC Ta at 0.7
Pa sputtering pressure.
85

Figure 37: SEM images of the surface morphology of tantalum films synthesized at 100 watts
sputtering power and various sputtering pressures ranging from 0.3 Pa to 1.4 Pa,
respectively.
86

Figure 37 (a-i) shows the surface morphologies of the films at different processing
conditions ranging from 0.3 to 1.4 Pa sputtering pressures. At low sputtering pressures
(0.3 and 0.4 Pa), surface morphologies are characterized by a fully dense film having a
low surface roughness; at 0.5 Pa pressure, alpha phase facet develops having a small
percentage of spherical shape beta particles ~50 nm diameter. At 0.7 Pa pressure the
surface morphology validates the XRD results and shows (110) textured alpha phase
facets. By increasing the sputtering pressure to 0.8 Pa, a typical beta phase surface
morphology is observed similar to the 0.3 and 0.4 Pa conditions; however, it appears to
have some crack-like features. The XRD spectra for the samples deposited at 0.8 Pa as
compared to samples deposited at 0.3 and 0.4 Pa shows an additional peak at 2-theta=
56°, representing a (432) texture which could be related to the difference in surface
morphology. At 0.9 and 1.1 Pa sputtering pressure the surface morphology is
characterized by having very small spherical particles and is consistent with XRD spectra
noting a beta phase formation. At 1.2 Pa sputtering pressure the film shows a high
surface roughness and appears to be dense with some elongated particles with sharp
edges. These particles could have either a (202) or (330) texture. At 1.4 Pa the amount
of elongated particles decreased and the film is characterized by having spherical
particles.
In general, the surface morphology of beta Ta can change significantly with pressure as
observed from the wide range textures. Jiang et al [14] reported that beta Ta can have
four possible configurations with the most probable structure being similar to beta-
uranium which has 30 atoms/cell. However, the theoretical calculations by Jiang et al
87

[14] are based on the formation of highly textured (002) Ta. Hence, it is likely that the
structure of beta Ta varies at different sputtering pressures and therefore the
corresponding surface morphology of beta phases also changes.
6.3.1.2 Effect of sputtering power
To further investigate the effect of processing conditions on the phase formation, Ta films
were deposited at 0.7 Pa sputtering pressure and different sputtering powers. Figure
38(a) shows the XRD results of the Ta films at sputtering powers of 50, 100 and 200
watts. It is observed that alpha Ta develops at different sputtering powers with the same
texture alignment, however, the intensity of the (110) peak decreases with increasing
power. Figure 38 (b-d) shows the corresponding surface morphologies of the films
deposited at 0.7 Pa sputtering pressure and sputtering powers of 50, 100 and 200 watts,
respectively. For all the processing conditions, surface morphologies are characterized
with dense regularly spaced grain facets which become elongated as the power increase.
88

Figure 38: (a) XRD spectra of the Ta films deposited at 0.7 Pa sputtering pressure and different
powers ranging from 50 watts to 200 watts (b-d) corresponding surface morphology of the Ta
films. Note formation of BCC Ta at all powers.
89

The summary of the processing conditions is presented in Table 6. The deposition rates
of the samples processed at different sputtering pressures do not significantly change,
however, increasing the deposition power results in higher deposition rates. Kim et al
[37] proposed that lower sputtering rates allow atoms to migrate over larger areas through
the crystals before being quenched in a specific position which leads to a strongly
textured orientation; however, at higher sputtering rates, the adatoms may be buried by
additional depositing material before they reach the lower surface energy sites in the
lattice. Therefore, by decreasing the deposition rate (decreasing the power) at 0.7 Pa,
adatoms are less likely to be buried by further incident of adatom flux and consequently
atoms diffuse over longer distance leading to stronger (110) texture alpha Ta.

Table 6: Summary of the processing conditions and phase formation of samples deposited at
different powers and pressures.
Pressure (Pa) Power (Watts) Phase Deposition rate (Å/s)
0.3 100 Beta 1.9
0.7 100 Alpha 1.8
1.4 100 Mix 1.8
Power (Watts) Pressure (Pa) Phase Deposition rate (Å/s)
50 0.7 Alpha 0.9
100 0.7 Alpha 1.8
200 0.7 Alpha 3.2

90

6.3.1.3 Phase Characteristics
A. Hardness, texture and residual stress
Even though a specific phase was identified for each of the sputtering conditions and
powers used, there was not a direct correlation that could lead us to understand the
formation of alpha vs. beta phase. Therefore, the relationship between texture, hardness
and residual stresses was further investigated.
The hardness values of the Ta films were measured using nanoindentation for samples
deposited at different sputtering pressures (0.3 to 1.4 Pa) and results are shown in Figure
39.

It is observed that the hardness of the alpha Ta films processed at 0.5 and 0.7 Pa
sputtering pressure is ~ 10 GPa, while the hardness of the beta phases varied between

Figure 39: Hardness values for nanocrystalline alpha (d ~ 45nm) and beta (d ~ 41-51nm) Ta as a
function of sputtering pressure (0.3 Pa to 1.4 Pa). Lower hardness values at 0.5 and 0.7 Pa
sputtering pressure are due to the formation of alpha Ta.
91

12.5 to 17.5 GPa. Zhang et al [22] demonstrated that the hardness of nanocrystalline beta
Ta (d~76 nm) is much higher than alpha Ta and they attributed this effect to the
complicated four layer stacking fault arrangement in (002) beta Ta which makes
dislocation motion more difficult. To investigate the effect of grain size on the hardness,
the microstructure of the films deposited at 0.3, 0.7 and 1.4 Pa sputtering pressure were
studied using TEM cross-sectional images as shown in Figure 40 (a-c).
92

The average column width of the Ta films increased from ~ 41 nm to ~51 nm as the
pressure was increased. Although, a slight change in grain size values is observed with
pressure this variation is not significant enough to explain the considerable changes in the

Figure 40: TEM cross-section images of Ta films at 0.3, 0.7 and 1.4 Pa sputtering pressure,
respectively. Note formation of the interface layer for the sample processed at 0.7 and 1.4 Pa
sputtering pressure. The inset shows the SAD pattern of the films.
93

hardness values. Average grain size values were also verified using top view TEM
images (see Figure 41).
Furthermore the hardness of beta Ta changed in the range of 12.5 to 17.6 GPa as the
pressure was increased. A wide range of hardness values has been reported for beta Ta
films changing from 15 GPa for (002) texture with the grain size of 32.3 nm [22] to 16-23
GPa for the beta Ta with the grain size ranging from 5-15 nm[38]. This variation in
hardness values could be related to different factors such as texture development or
different beta microstructures. Table 7 shows the summary of hardness, residual stress
and texture alignment as a function of sputtering pressures.

Figure 41: Plane-view TEM micrograph of the nanocrystalline Ta processed at 100 watts
sputtering power and 0.3 Pa sputtering pressure.
94

Results demonstrate that the hardness measurements do not reflect the variation in the
texture development; however, for films having the same texture alignment the hardness
value decreases as the residual stress becomes more tensile. A compressive stress
typically develops at low sputtering pressures due to the higher atomic bombardment and
leads to the formation of denser microstructures [39] (see Figure 40a). On the other hand,
by increasing the pressure, tensile stresses are generated at the grain boundaries which
can lead to less dense voided microstructures (see Figure 40c). Therefore, the decrease in
hardness values can be related to the increase in tensile stress which leads to the
formation of less dense structures. This result is in agreement with the Mani et al [40] ,
Table 7: Summary of the hardness, residual stress and texture for samples processed at 100 watts
power and sputtering pressures ranging from 0.3 Pa to 1.4 Pa.
Pressure (Pa) Hardness (GPa) Residual stress
(MPa)
Phase Texture
0.3 17.7 -1518 Beta (400) or (410)
0.4 14.7 -1271 Beta (400) or (410)
0.8 13.6 -708 Beta (400) or (410)
0.9 15.9 -214 Beta (212)
1.1 13.9 426 Beta (400) or (002)
1.2 13.0 616 Beta (202) or (303)
1.4 12.4 1131 Beta (202) or (303)
0.5 10.2 -1125 alpha (211)
0.7 10.9 -1039 alpha (110)

95

who showed that the compressive stress is directly correlated to higher hardness values
for TiC thin films.
B. Film Thickness
TEM cross-sectional images shown in Figure 40 demonstrated that the film
microstructure changed as a function of thickness. At 0.3 Pa sputtering pressure the film
microstructure is characterized by having a very dense columnar growth throughout. In
contrast, at 0.7 and 1.4 Pa sputtering pressure, a non-columnar layer is detectable at the
Ta/substrate interface. At 0.7 Pa sputtering pressure, this layer is less than 100 nm and
columnar growth starts thereafter. By increasing the sputtering pressure to 1.4 Pa the
thickness of the interface layer increase to ~150 nm followed by a columnar voided
structure. As recently shown by Navid and Hodge [41], the microstructure of this
interface layer is characterized by having an almost amorphous layer with very small
grain sizes (2-3nm). To study the effect of this interface layer on the phase formation, Ta
films were deposited at different thicknesses ranging from 10nm to 1000nm. The XRD
spectra of the samples deposited at sputtering pressures ranging from 0.3, 0.7 and 1.4 Pa
and different thicknesses are shown in Figure 42.
96

Figure 42: XRD spectra of the tantalum films as a function of thickness (10 nm to 1000 nm) and
at different sputtering pressures of 0.3, 0.7 and 1.4 Pa sputtering pressure.
97

For the 10nm thickness films, a low intensity peak is detected for all processing
conditions due to the formation of very small grain size at the Ta/substrate interface;
while as the film thickness increases, higher intensity XRD peaks were observed at all
sputtering pressures. These results are in agreement with Lee et al [42] to some degree.
They proposed that the structure of Ta films deposited on Si substrates is determined by
the thin layer of non-crystalline Ta at the interface with the silicon substrate. However,
in this study, this thin layer changes the structure but does not appear to have a direct
correlation to a given phase. Alpha Ta is always formed at 0.7 Pa even though the
texture changed from (211) to (110) as the thickness increased. Therefore, even though
this interface does appear to have an effect on the overall stress state and column growth,
at this stage it is preliminary to directly related to a given phase formation.
6.3.2 Effect of internal parameters on phase formation
So far we correlated a specific phase and its characteristics to a given sputtering
condition; however it is still not clear why a specific phase forms at a given condition.
Therefore, one must search internal factors for controlling the phase formation such as
plasma kinetics during the sputtering deposition as well as the role of impurities.
A. Ion flux and energy
The energy and direction of sputtered particles arriving at the substrate affect the final
microstructure and texture evolution of a sputtered film [43, 44]. To date, most studies
were emphasized on the effect of ion flux and energy on the nucleation of the beta and
alpha Ta during bias sputter deposition [1, 9, 13, 18, 26, 28]. It was demonstrated that
98

increasing the ion bombardment energy by applying a bias promotes the formation of
alpha Ta. Meanwhile, formation of a stable alpha phase during sputtering with low ion
bombardment energy (no bias or low bias voltage) has also been reported by Catania et al
[26] and Ino et al [13]. Their results show that there is a critical energy threshold
necessary to synthesize alpha Ta. However, the value of the threshold energy needed to
transition from beta to alpha phase is still under debate [13, 18, 26].

As it has been mentioned in chapter 4, in the absence of substrate bias, ion bombardment
energy is defined by the plasma potential. The substrate was grounded in all the
deposition processes, therefore, the ion bombardment energy during deposition is

Figure 43: Calculated ion energy for tantalum and copper samples measured by Langmuir probe
at 100 watts power and sputtering pressures ranging from 0.3 to 1.4 Pa. A small peak is observed
at 0.7 Pa sputtering pressure.

99

describe by eV
p
, where V
p
is the plasma potential derived from I-V graphs. Similar to
the method mentioned in chapter 4, the bombardment energy measurements are
calculated from Langmuir probe data for Ta films deposited at 0.3, 0.7 and 1.4 Pa
sputtering pressure at 100 watts sputtering power (Figure 43). It is expected that by
increasing the pressure, the bombarding energy decreases due to the higher gas
scattering; though, a small peak is observed at 0.7 Pa sputtering pressure which is the
pressure that leads to alpha formation. The same peak was present at 0.7 Pa sputtering
pressure for copper samples (see Figure 25) and it is shown again in Figure 43 for the
comparison. Therefore, the formation of alpha Ta at 0.7 Pa sputtering pressure could be
due to the higher ion bombarding energy observed at this particular sputtering condition.
This effect could be related to the specific chamber geometry which changes the kinetics
inside the chamber and can lead to a shift in the critical threshold energy at a given
sputtering pressure.
B. Impurities
The level of impurities during sputtering affects the film’s crystalline orientation, grain
size and electrical properties [9, 20, 23, 31, 45-49]. To investigate the effect of impurities
on the phase formation in the present study, Ta films deposited at a lower vacuum (2×10
-3
Pa) are compared to films deposited at a higher vacuum (2×10
-5
Pa). Figure 44a and b
show the XRD spectra of the Ta films deposited at 0.3, 0.7 and 1.4 Pa sputtering
pressures at the two different base pressures. It is observed that for the samples deposited
at lower vacuum (2×10
-3
Pa), highly textured (002) beta Ta films developed for all
sputtering pressures. In contrast, the films deposited at the high vacuum (2×10
-5
Pa)
100

have randomly textured beta phases at 0.3 and 1.4 Pa sputtering pressures while 0.7 Pa
sputtering pressure leads to (110) alpha phase . Chemical analysis by Cathode Emission
was performed on the samples processed at 0.7 Pa sputtering pressure for both base
pressures. Results show identical level of impurities for both samples, ranging from
0.001 to 0.0001 w.t. %. It should be noted that low level, low weight impurities such as
O and H are beyond the detection limits of this technique.
101

However, the contribution of low level impurities such oxygen in the films can be
estimated. For example, a rough calculation of the time that it would take to from a
monolayer of the oxygen at different base pressures can be derived using the following
formulas:

Figure 44: XRD spectra of Ta films deposited using (a) 2×10
-3
Pa base pressure (low vacuum)
and (b) 2×10
-5
Pa base pressure (high vacuum) at 100 watts power and 0.3, 0.7 and 1.4 Pa
sputtering pressures. Note formation of highly textured beta Ta at all of the processing
conditions for the sample deposited using 2×10
-3
Pa base pressure (low vacuum).
102

1.99 10

/

  1
7.929   10

/
(2)
Where M is the molar mass, T is the temperature, P is the pressure, is the molecular
diameter and b is a constant. Therefore, the time to form a monolayer of oxygen at
2×10
-3
Pa (low vacuum) is ~ 0.23 second, while at 2×10
-5
Pa (high vacuum) this time is
increased to ~23 seconds. Consequently, increasing the amount of oxygen in the system
appears to promote the formation of highly textured beta Ta. It is observed that once the
level of oxygen in the film decreases, alpha Ta can be deposited by optimizing the
sputtering pressure. These results are in agreement with Knepper et al [31], who
investigated the effect of oxygen on the thermomechanical behavior of Ta films at
different oxygen pressures and showed that the phase transformation from beta to alpha
occurs at higher temperatures by increasing the oxygen content. However, this effect
alone does not explain the formation of alpha Ta at 0.7 Pa sputtering pressure.
Other impurities such as argon or nitrogen can affect the surface mobility of the Ta atoms
and can result in the nucleation of a large number of small crystallites, which can lead to
a change in film texture from a preferred orientation to a more randomly oriented
structure [20, 47-49]. For instance, the amount of argon trapped inside the film has been
shown to change at different processing conditions [20]. The concentration of impurities
(e.g. argon) trapped inside the film can be calculated based on the following calculations
[50] as shown by Knepper et al [31]:
103


  3

√2
4
Where α
c
and α
s
are the condensation and the sticking coefficient, respectively, Z is the
impingement rate and j
j
is the incident flux (taken from the Langmuir data). It has been
reported that the condensation coefficient is often practically equal to one during thin
film deposition [20].

On the other hand, the value of the sticking coefficient changes as a function of trapping
probability (experimentally it varies in the range of 0.001 and 0.9) and fractional surface
converge. In this study the amount of argon trapped inside the film (f) has been
calculated based on the assumption that the sticking coefficient is equal to one (highest
possible value). It is observed from the results (see Table 8) that the amount of argon
Table 8: Calculated argon impurities in the Ta films as a function of sputtering pressure
Pressure (Pa) Incident Flux*
(m
-2
s
-1
)
f (trapped
impurity)
Phase
0.3 9.75 * 10
20
0.0468 Beta
0.7 1.35* 10
21
0.0810 Alpha
1.4 1.57*10
21
0.1321 Beta

104

trapped changed linearly as the pressure increased and no peak was observed at 0.7 Pa
sputtering pressure. A similar slope trend is expected for nitrogen.
Hence, it is observed that formation of alpha Ta is probably due to the mutual effect of
kinetics and impurity levels inside the chamber. Beyond some level of impurities, highly
textured beta Ta develops for all conditions (lower vacuum). At high vacuum the
amount of impurities decreases and randomly oriented beta Ta is formed at all sputtering
conditions, except at 0.7 Pa sputtering pressure for which the small increase in the ion
energy seems to stabilize the microstructure and leads to alpha Ta formation. Further
experiments at higher vacuum could further elucidate on this result and will be the
subject of a future study.
6.4 Summary
In the present study, alpha Ta has been synthesized on Si substrates using DC- magnetron
sputtering at room temperature by changing the sputtering pressures. The XRD results in
combination with SEM surface morphology images and hardness values verified
formation of alpha Ta at 0.7 Pa sputtering pressure and randomly textured beta phase or
mix phases at any other sputtering pressures. Comparison between hardness, residual
stress and texture alignment demonstrates no significant correlation between hardness
values and texture development; however, for films having the same texture alignment,
the hardness value decreases as the residual stress becomes more tensile. This result
could be due to the formation of the loosely packed grain boundaries as the pressure
increases. In addition, processing of alpha Ta at 0.7 Pa sputtering pressures at different
105

powers ranging from 50 watts to 200 watts demonstrates that sputtering pressure is the
dominant factor in alpha formation when compared to the sputtering power. From this
analysis, the effect of plasma kinetics and also impurities at different pressures were
discussed. The kinetics of ion bombardment were investigated at different sputtering
pressures and the formation of alpha Ta at 0.7 Pa sputtering pressure was related to the
increase in ion energy at this specific sputtering conditions. Meanwhile, it is observed
that the level of impurities in the sputtering chamber could affect the film texture and
therefore the phase formation during deposition.
106

Chapter 7. Controlling residual stresses in sputtered alpha-Tantalum
7.1 Introduction
As discussed in chapter 6, we were able to process BCC tantalum at room temperature
without applying external parameters. In addition, it was observed that nanostructured
tantalum films were sputtered at room temperature having either high compressive or
tensile stress states (-1500 GPa to 1000 GPa). However, it is not clear whether any
particular Ta phase is favored at a given stress state.
As it has been mentioned before, the most common strategies to control a specific phase
formation in Ta thin films are heating the substrate during deposition [4, 13-15], applying
a bias [3, 16-18] and changing the substrate material [5, 13, 19-22]. However, the
addition of a heating step during deposition could lead to significant amounts of grain
growth which is likely to affect the overall stress state. For example, Clevenger et al. [4]
showed that heating the sample causes a drastic change towards tensile stresses which is
accompanied by a phase transformation from beta to alpha; though, the contribution of
the stress state on the phase formation was not studied. In addition, Detor et al. [23]
demonstrated that stress state was strongly influenced by applying a bias during
deposition in magnetron sputtering.
Since previous measurements have been performed either with samples that were heated
[4, 13-15] or biased during sputtering [3, 16-18], they were not able to study the
correlation between phase formation and stress state during tantalum film deposition. In
107

this chapter, we present a study beyond current research and focus on the correlation
between alpha phase formation and stress state as a function of sputtering conditions
(without heating or bias); specifically with emphasis on controlling the residual stress in
single phase alpha Ta at different sputtering pressures.
7.2 Experimental methods
Ta thin films were deposited by DC- magnetron sputtering using sputtering pressures
ranging from 0.3 to 1.4 Pa at a power of 100 watts on 250 μm thick Si (100) substrates.
The base pressure was maintained at less than 2×10
-5
Pa using argon (99.999%) as the
sputtering gas. Films were processed with the maximum thickness of 500 nm using a
99.995% Ta target. In addition, to investigate the effect of underlayer, the sputtering
machine was equipped with multiple targets which allow the processing of different
layers without exposing to the air. A 100 nm thickness of aluminum, niobium and alpha
tantalum underlayer was deposited on the silicon substrate and tantalum processing was
continued by rotating the substrate holder to the tantalum target at different sputtering
pressures of 0.3, 0.7 and 1.4 Pa. Residual stresses were determined from wafer curvature
measurements using an XP-2 stylus profilometer (Ambios, Santa Cruz, California). The
surface morphologies of the samples were characterized using a high resolution SEM in a
Nova 600 Dual beam FIB (FEI, Tokyo, Japan). To investigate the films microstructure,
cross-sections of the specimens were prepared by focus ion beam (FIB) liftout technique
while the plan view of the samples were prepared with traditional method.

108

7.3 Results and discussions
7.3.1 Stress development as a function of sputtering pressure
In Figure 45(a) the phase formation is presented as a function of sputtering pressure for
selected pressures ranging from 0.3 Pa to 1.4 Pa. The X-ray spectra of the films
demonstrate the formation of (110) alpha Ta at 0.7 Pa sputtering pressure. As shown in
Figure 36 under any other sputtering condition (in the 0.3 to 1.4 Pa range) either randomly
oriented beta Ta or a mixture of alpha and beta phases are indentified. In order to
investigate the alpha Ta structure in detail, TEM images of a typical alpha Ta
microstructure sputtered at 0.7 Pa were obtained using a Tecnai F30 TEM (FEI, Tokyo,
JAPAN) and are presented in Figure 45b-d. Figure 45(b) shows a TEM cross-sectional
image; the sample was prepared by focus ion beam (FIB) lift-out technique. It is
observed that a very dense thin layer of Ta (less than 100 nm thick) with ~2-3 nm grain
size forms starting at the film/substrate interface, and a columnar grain structure with an
average column width of ~ 50 nm is deposited thereafter. The top view of the films is
presented in Figure 45(c) confirming an average grain size of ~ 50nm. The
microstructure of the thin interface layer was studied by depositing a 70nm thick Ta film
on a 10 nm thick amorphous Si
3
N
4
TEM substrate grid. The TEM plan view image of the
interface layer is shown in Figure 45(d) and confirms the formation of very small grain
sized Ta film on the silicon substrate.
109

Figure 45: (a) Selected X-ray spectra of Ta films sputtered at a power of 100 watts, with
increasing sputtering pressures from 0.3 Pa to 1.4 Pa. Notice the formation of alpha Ta at 0.7 Pa
as shown by XRD and TEM in (b-d). Where b) is the cross-sectional image, arrow denotes
growth direction, (c) Top-view image and (d) plane-view image of the interface (i) layer showing
the formation of ultra small grains at the interface between film and substrate.

110

The corresponding residual stress values as a function of sputtering pressure are shown in
Figure 46. It is observed that the stress increased almost linearly while increasing the
sputtering pressure and no peak in the stress trend is observed at 0.7 Pa sputtering
pressure.
A high compressive stress (~ -1500 MPa) develops at low sputtering pressures and as the
pressure is increased to 1.4 Pa, the film grows in a high tensile stress mode (~1000 MPa).
The transition from compressive to tensile stress occurs at around 1 Pa sputtering
pressure. This transition in stress state while increasing the pressure has been previously
reported for Ta [4, 24]; however, it typically occurred at higher sputtering pressures

Figure 46: Plot of the residual stress as a function of sputtering pressures from 0.3 Pa to 1.4 Pa.
Arrow denotes alpha phase at 0.7 Pa. All other pressures have a pure beta or a mixture of alpha
and beta structures.
111

(ranging from 3 to14 Pa) as compared to the present results. The lower transition
pressure from compressive to tensile stress may arise from the formation of the ultra
small grain size at the interface layer (see Figure 1d) which could promote higher tensile
stress values at lower pressures [25]. The correlation between tensile stress development
and grain size was studied by Doljack and Hoffman [24, 26] and later on modified by Nix
and Clemens [25]. They derived the upper bound estimate of the stress in the film by the
following formula,

1
2

/
7.1
where E is the Young’s modulus, ν is the Poisson’s ratio, γ
sv
and γ
gb
are the surface and
grain boundary energies of the film and d is the grain size. It is observed from the
formula that high tensile stress is developed in the film with decreasing the grain size.
From the graph presented by Nix and Clemens [25] the residual stress for the film with
the grain size around 2-3 nm expected to be between 10-15 GPa, however, our
experimental results demonstrated maximum stress of ~ 2 GPa during deposition.
Although, it is well known [23, 27] that the above equation overestimates the value of the
tensile stress evolution this formula provides some insights on the correlation between
grain size and tensile stress. Therefore, in our results, tensile stress development arising
from the formation of the ultra small grain size interface layer and the lower threshold
pressure could be related to the formation of this layer.
112

Our initial results demonstrate that alpha Ta developed only in a compressive mode,
while beta Ta could be formed in both compressive and tensile states as the pressure is
increased. At this stage it is still not obvious whether a compressive state is a
prerequisite for alpha phase formation.
7.3.2 Role of underlayer
To further investigate the relationship between phase formation and stress development,
we process alpha tantalum at different sputtering conditions using underlayer materials. A
great effort has been made to synthesize the alpha tantalum by changing the substrate
materials and it has been shown that utilizing different underlayers such as aluminum,
niobium, titanium and tantalum nitride promotes the formation of alpha tantalum [5, 13,
19-22]. However, the exact mechanism that lead to the formation of alpha tantalum with
changing the substrate is not clearly indentified.
It is well known that certain substrates with the same lattice parameter (e.g. niobium) or
texture alignment (e.g. aluminum) promote the growth of alpha tantalum independent of
sputtering conditions. Sato [21] showed that alpha tantalum always nucleated on the
titanium underlayer, however, mix phases developed on the aluminum and platinum
underlayer. In contrast, formation of alpha tantalum on (111) aluminum underlayer has
been demonstrated by Hoogeveen et al. [13] and Morohashi et al. [28]. The formation of
alpha tantalum on the FCC aluminum was explained by approximate matching of some
of the atoms on the (110) plane of alpha tantalum with the atoms on the (111) plane of
FCC aluminum. However, this matching is only along the <110> axis of the Al (111)
113

plane aligned with the <111> axis of the (110) tantalum plane but it appears to be
sufficient for growth of alpha tantalum. On the other hand, Face et al. [5] demonstrated
that depositing a layer of niobium on the Si substrate before tantalum deposition
promotes the growth of alpha tantalum independent of the processing condition in ion
beam sputtered deposition. They proposed that growth of alpha tantalum is due to the
very close lattice match (< 0.1 %) of alpha tantalum and niobium. Afterward, Sajovec et
al. [20] employed the same techniques for developing the alpha tantalum on the fused
silica substrate and they showed that the thickness of niobium underlayer in very crucial
with respect to the nucleation of the alpha tantalum. However, in both cases the
sputtering performed using high beam energies between 500 and 1500 eV. It is well
known [3, 16-18, 29, 30] that the phase development is highly dependent on the energies
of the particles inside the chamber, therefore, alpha tantalum development during the ion
beam sputtering could be due to the mutual effect of the substrate material and the
energetic particles inside the chamber.
In this study, we investigated the effect of aluminum and niobium underlayers on the Ta
phase formation. The XRD spectra of the Ta thin films deposited on the aluminum and
niobium underlayer at different sputtering pressures of 0.3, 0.7 and 1.4 Pa are shown in
Figure 47. The 500 nm thickness Ta films were deposited on the 100 nm thickness of
aluminum and niobium underlayer without exposing underlayers to air. It is observed
from Figure 47 that utilizing aluminum and niobium underlayers at 0.3 and 1.4 Pa
sputtering pressures results in the formation of mixed phases, however, at 0.7 Pa
114

sputtering pressure alpha tantalum develops on the both niobium and aluminum
substrates with the (110) and (211) texture, respectively.

Figure 47 : The effect of (a) Aluminum and (b) Niobium underlayers on the phase formation of
tantalum.
115

The XRD spectra of the underlayers in the present study demonstrate formation of (211)
textured niobium and (200) textured aluminum, therefore, the effect of an underlayer
cannot be confirmed with the current underlayers. However, growth of alpha tantalum at
0.7 Pa sputtering pressure on both aluminum and niobium underlayers, demonstrates that
the processing condition is the dominant factor as compare to the underlayer material.
To further investigate the effect of lattice match and texture alignment, we deposited
alpha tantalum on (110) texture alpha tantalum underlayer. Ta films were deposited at 0.7
Pa sputtering pressure to a thickness of 500 nm. Immediately after the first 500 nm were
deposited, the source shutter was closed for 5 minutes, and then sputtering continued at
either 0.3 or 1.4 Pa sputtering pressures until an additional 500 nm were deposited. It
should be noted that both 0.3 and 1.4 Pa sputtering pressures had led to beta Ta formation
as shown in Figure 45. The thickness of the underlayer was kept at 500 nm to avoid any
probable effect of the interface layer as shown in Figure 45(d). The XRD results shown
in Figure 48a demonstrate formation of alpha Ta at 0.3 and 0.7 Pa sputtering pressures.
The surface morphologies of the samples with an alpha Ta underlayer (Figure 48b-d)
were characterized using a high resolution SEM to verify the formation of alpha Ta. For
the samples processed at 0.3 and 0.7 Pa sputtering pressure, the surface morphologies are
smooth which; while by increasing the pressure to 1.4 Pa, a higher surface roughness is
observed. Therefore, growth of alpha tantalum at different sputtering pressures on the
alpha tantalum underlayer would support the effect of underlayer on the nucleation of
alpha tantalum. Though, this effect could be due to the mutual effect of lattice match and
texture alignment.
116

Figure 48: (a) the XRD spectra of the films processed at 0.3, 0.7, 1.4Pa, respectively, using an
alpha Ta underlayer. (b)-(d) SEM images of the surface morphology of Ta samples synthesized
at 100 watts using three different sputtering pressures of 0.3, 0.7 and 1.4 Pa deposited on an
alpha Ta underlayer surface.
117

7.3.3 Alpha tantalum stress state
In order to further elucidate about the effect of the alpha underlayer on the film texture,
phase formation and overall stress, Figure 49 is presented.

It should be noted that the stress values for the samples sputtered using an alpha
underlayer were calculated by taking into account the stress values from the alpha Ta on
Si presented in Figure 46. Overall, Figure 49 shows that alpha Ta can be synthesized in
both, a compressive mode at low sputtering pressures (0.3 Pa) as well as in a tensile mode

Figure 49: Plot of the residual stress of alpha Ta deposited at different sputtering pressures (0.3,
0.7, and 1.4 Pa) using an alpha Ta underlayer. Note the formation of alpha Ta at all three
different sputtering pressures.
118

at high sputtering pressures (1.4 Pa); thus, indicating that no particular phase is favored
for a given stress-state. These results are mainly attributed to the use of an underlayer
with a (110) texture which serves as a seed for further (110) growth. Previous studies
have demonstrated a similar growth effect due to a (110) texture alignment with the
underlayer material given that (110) is the lowest energy lattice site configuration for a
BCC material [13, 28]. However, no previous studies have demonstrated the relationship
between texture, phase and stress state. In general, this study has shown that having the
alpha underlayer is a key parameter to be able to synthesize alpha Ta having a wide range
of stress values. However, which sputtering conditions can always lead to the formation
of the initial alpha Ta underlayer is still under debate in the community and will be the
focus of a future study.
7.4 Summary
In summary, this study has shown that having an alpha Ta underlayer plays a dominant
role in developing an alpha structure (without heat or bias); since it has the same lattice
parameter and the lowest surface energy texture alignment (110). It is observed that the
stress evolution in Ta films can be independent of a particular phase formation.
Therefore, by manipulating the film growth by the selection of an adequate underlayer,
nanostructured alpha Ta can be synthesized in both compressive and tensile states, thus
allowing maximum flexibility for future Ta thin film applications.
119

Conclusions
Processing of thin films using magnetron sputtering method has been widely investigated
in many studies with emphasis on the correlation between film properties and external
chamber conditions such as power and pressure [5, 12, 13, 52, 67, 124-126]. However,
the effect of internal chamber conditions (e.g. ion and atom kinetics and impurities) on
the film microstructure and residual stress needs further investigation.
In this thesis, the correlation between the plasma conditions such as ion density and
plasma potential, the film microstructure as well as the residual stress have been
investigated. It has been shown that the film properties are directly affected by internal
chamber conditions such as the particles bombarding energies which are related to the
specific chamber geometry. In addition, the study of the in-situ stress measurement and
the correlation of the external chamber conditions to the films residual stress and
microstructure demonstrated the importance of the kinetics inside the chamber.
Taking into account the importance of internal chamber conditions, nanostructured alpha
tantalum films were synthesized on silicon substrates at room temperature. Processing of
nanocrystalline BCC materials is very challenging due their sensitivity to the processing
conditions. However, by controlling the level of impurities and changing the kinetics
inside the chamber, dense nanostructured BCC tantalum was processed successfully.
The main contributions of this study are summarized as following:
120

 The correlation between plasma characteristics (e.g. ion density, ion and atom
energy) with the film’s microstructure and residual stress demonstrated that film
properties are strongly related to the kinetics inside the chamber. For our specific
chamber configuration, the plasma potential shows a peak at 0.7 Pa sputtering
pressure which results in the formation of dense films with a uniform
microstructure.
 In-situ stress measurement for both copper and tantalum films demonstrate
compressive stress development in the films deposited at low sputtering pressure,
however, increasing the pressure results in tensile stress development and higher
surface roughness.
 Results show that the thermal stress contribution is quite significant for materials
with low melting point (high adatom mobility) and might change the stress state
of the film. In addition, the thermal stress increases significantly in the initial
growth stage and then remains relatively constant until the final thickness is
reached.
 Nanostructured BCC tantalum films were synthesized on silicon substrates at
room temperature without the addition of external conditions such as substrate
heating or bias application. Alpha tantalum films were deposited at 0.7 Pa
sputtering pressure and different powers ranging from 50 watts to 200 watts. No
phase transformation was observed as a function of film thickness.
121

 Formation of alpha tantalum appears to be related to a combination of factors
such as lower level of impurities (e.g. oxygen, nitrogen and argon) and the
kinetics of the sputtered atoms and ions.
 An alpha tantalum underlayer promotes the formation of alpha tantalum
independent of the sputtering conditions due to the common effect of lattice
match and orientation alignment. However, using an underlayer with either the
same lattice coefficient (e.g. Nb) or texture alignment (e.g. Al) did not result in
the formation of alpha tantalum.
 It is observed that the stress evolution in tantalum films can be independent of a
particular phase formation and is highly correlated to the kinetics inside the
chamber. Therefore, by manipulating the sputtering conditions, a low stress film
can be synthesized, allowing maximum flexibility for future thin film
applications.

122

References
[1] A.J. Detor, A.M. Hodge, E. Chason, Y.M. Wang, H.W. Xu, M. Conyers, A.
Nikroo, A. Hamza, Acta Mater. 57/7 (2009) 2055.
[2] D.M. Mattox, Handbook of physical vapor deposition (PVD) processing: film
formation, adhesion, surface preparation and contamination control Westwood, N.J.,
1998.
[3] M. Ohring, Materials science of thin films deposition and structure., Academic
press, San Diego, 2002.
[4] "MBE", photonics.ethz.ch, http://www.photonics.ethz.ch/research (20 Feb. 2009).
[5] J.A. Thornton, D.W. Hoffman, Thin Solid Films 171/1 (1989) 5.
[6] "PLD", Photo. physandtech.net, http://www.physandtech.net/4.htm (14 Nov.
2001).
[7] A.Q. Jiang, A. Yohannan, N.O. Nnolim, T.A. Tyson, L. Axe, S.L. Lee, P. Cote,
Thin Solid Films 437/1-2 (2003) 116.
[8] M. Pletea, W. Bruckner, H. Wendrock, R. Kaltofen, J. Appl. Phys. 97/5 (2005) 7.
[9] S.P. Kim, H.M. Choi, S.K. Choi, Thin Solid Films 322/1-2 (1998) 298.
[10] N. Schell, K.P. Andreasen, J. Bottiger, J. Chevallier, Thin Solid Films 476/2
(2005) 280.
[11] C.V. Thompson, R. Carel, Journal of the Mechanics and Physics of Solids 44/5
(1996) 657.
[12] J.A. Thornton, Thin Solid Films 54/1 (1978) 23.
[13] J.A. Thornton, Journal of Vacuum Science & Technology a-Vacuum Surfaces and
Films 4/6 (1986) 3059.
[14] A.L. Shull, F. Spaepen, J. Appl. Phys. 80/11 (1996) 6243.
[15] S. Ryu, K. Lee, S. Ma, Y. Kim, Amer Scientific Publishers, 2007, p. 4081.
123

[16] R. Koch, J. Phys.-Condes. Matter 6/45 (1994) 9519.
[17] K. Kinosita, Thin Solid Films 50/MAY (1978) 205.
[18] K. Kinosita, Thin Solid Films 12/1 (1972) 17.
[19] C. Friesen, C.V. Thompson, Phys. Rev. Lett. 89/12 (2002).
[20] C. Friesen, S.C. Seel, C.V. Thompson, J. Appl. Phys. 95/3 (2004) 1011.
[21] J.A. Floro, S.J. Hearne, J.A. Hunter, P. Kotula, E. Chason, S.C. Seel, C.V.
Thompson, J. Appl. Phys. 89/9 (2001) 4886.
[22] J.A. Floro, E. Chason, S.R. Lee, R.D. Twesten, R.Q. Hwang, L.B. Freund, Journal
of electronic materials (1997) 969.
[23] E. Chason, B.W. Sheldon, L.B. Freund, J.A. Floro, S.J. Hearne, Phys. Rev. Lett.
88/15 (2002) 4.
[24] R. Abermann, R. Koch, Thin Solid Films 129/1-2 (1985) 71.
[25] R. Abermann, Pergamon-Elsevier Science Ltd, 1990, p. 1279.
[26] G. Abadias, P. Guerin, Appl. Phys. Lett. 93/11 (2008) 3.
[27] R.J. Drese, M. Wuttig, J. Appl. Phys. 99/12 (2006) 5.
[28] M. Pletea, W. Bruckner, H. Wendrock, R. Kaltofen, R. Koch, J. Appl. Phys. 99/3
(2006) 5.
[29] M. Pletea, H. Wendrock, R. Kaltofen, O.G. Schmidt, R. Koch, J. Phys.-Condes.
Matter 20/25 (2008) 6.
[30] P.N. Baker, Thin Solid Films 14/1 (1972) 3.
[31] P. Catania, J.P. Doyle, J.J. Cuomo, Journal of vaccum science & technology A-
vaccum surface and films 10/5 (1992) 3318.
[32] L.A. Clevenger, A. Mutscheller, J.M.E. Harper, C. Cabral, K. Barmak, J. Appl.
Phys. 72/10 (1992) 4918.
[33] D.W. Face, D.E. Prober, J. Vac. Sci. Technol. A-Vac. Surf. Films 5/6 (1987)
3408.
124

[34] C.C. Fang, F. Jones, V. Prasad, J. Appl. Phys. 74/7 (1993) 4472.
[35] B.L. French, J.C. Bilello, Thin Solid Films 446/1 (2004) 91.
[36] M. Grosser, U. Schmid, Thin Solid Films 517/16 (2009) 4493.
[37] K. Holloway, P.M. Fryer, Appl. Phys. Lett. 57/17 (1990) 1736.
[38] R. Hoogeveen, M. Moske, H. Geisler, K. Samwer, Thin Solid Films 275 (1996)
203.
[39] K. Ino, T. Shinohara, T. Ushiki, T. Ohmi, J. Vac. Sci. Technol. A-Vac. Surf.
Films 15/5 (1997) 2627.
[40] R. Knepper, S.P. Baker, Appl. Phys. Lett. 90/18 (2007) 181908
[41] S.L. Lee, M. Doxbeck, J. Mueller, M. Cipollo, P. Cote, Surface & Coatings
Technology 177 (2004) 44.
[42] S.L. Lee, D. Windover, Surface & Coatings Technology 108/1-3 (1998) 65.
[43] H. Ren, M. Sosnowski, Thin Solid Films 516/8 (2008) 1898.
[44] Y.M. Wang, A.M. Hodge, J. Biener, A.V. Hamza, D.E. Barnes, K. Liu, T.G.
Nieh, Appl. Phys. Lett. 86/10 (2005) 101915.
[45] W.D. Westwood, R.J. Boynton, P.S. Wilcox, Thin Solid Films 16/1 (1973) 1.
[46] C. Yong, B.C. Zhang, C.S. Seet, A. See, L. Chan, J. Sudijono, S.L. Liew, C.H.
Tung, H.C. Zeng, Journal of Physical Chemistry B 106/48 (2002) 12366.
[47] Z.L. Yuan, D.H. Zhang, C.Y. Li, K. Prasad, C.M. Tan, L.J. Tang, Thin Solid
Films 434/1-2 (2003) 126.
[48] M. Zhang, B. Yang, J. Chu, T.G. Nieh, Scr. Mater. 54/7 (2006) 1227.
[49] M. Zhang, Y.F. Zhang, P.D. Rack, M.K. Miller, T.G. Nieh, Scr. Mater. 57/11
(2007) 1032.
[50] Y.M. Zhou, Z. Xie, H.N. Xiao, P.F. Hu, J. He, Vacuum 83/2 (2008) 286.
[51] M. Shiojiri, S. Shinkai, K. Sasaki, H. Yanagisawa, Y. Abe, Japanese Journal of
Applied Physics Part 1-Regular Papers Short Notes & Review Papers 42/7A (2003) 4499.
125

[52] J. Alami, P. Eklund, J.M. Andersson, M. Lattemann, E. Wallin, J. Bohlmark, P.
Persson, U. Helmersson, Thin Solid Films 515/7-8 (2007) 3434.
[53] P. Catania, R.A. Roy, J.J. Cuomo, J. Appl. Phys. 74/2 (1993) 1008.
[54] L. Gladczuk, A. Patel, J.D. Demaree, M. Sosnowski, Thin Solid Films 476/2
(2005) 295.
[55] F. Sajovec, P.M. Meuffels, T. Schober, Thin Solid Films 219/1-2 (1992) 206.
[56] S. Sato, Thin Solid Films 94/4 (1982) 321.
[57] J.Q. Zhang, Y.M. Huai, L.F. Chen, J. Zhang, JOURNAL OF VACUUM
SCIENCE & TECHNOLOGY B 21/1 (2003) 237.
[58] L.B. Freund, S. Suresh, Thin Film Materials: stress, defect formation and surface
evolution, Cambridge University Press, Cambridge, 2003.
[59] R.F. Bunshah, Handbook of hard coatings : deposition technologies, properties
and applications, Noyes Publications Park Ridge, N.J. , 2001.
[60] M. Ohring, Materials science of thin films deposition and structure, Academic
press, San Diego, 2002.
[61] J.A. Thornton, Journal of Vacuum Science & Technology 15/2 (1978) 171.
[62] I.H. Hutchinson, Princicles of plasma diagnostics, cambridge university press,
cambridge, 1987.
[63] D. Fang, R.K. Marcus, Spectroc. Acta Pt. B-Atom. Spectr. 45/9 (1990) 1053.
[64] W.M. Posadowski, Thin Solid Films 392/2 (2001) 201.
[65] K.L. Brown, G.W. Tautfest, Rev. Sci. Instrum. 27/9 (1956) 696.
[66] R.J. Shul, S.J. Pearton, Handbook of advanced plasma processing techniques,
springer, berlin, 2000.
[67] J.A. Thornton, Journal of Vacuum Science & Technology 11/4 (1974) 666.
[68] A.M. Hodge, R. Foreman, G.F. Gallegos, J. Nucl. Mater. 342/1-3 (2005) 8.
[69] R. Koch, D. Winau, K.H. Rieder, Royal Swedish Acad Sciences, 1993, p. 539.
126

[70] R. Abermann, Pergamon-Elsevier Science Ltd, 1990, p. 1279.
[71] J.A. Thornton, D.W. Hoffman, Thin Solid Films 171/1 (1989) 5.
[72] H. Ljungcrantz, L. Hultman, J.E. Sundgren, S. Johansson, N. Kristensen, J.A.
Schweitz, C.J. Shute, J. Vac. Sci. Technol. A-Vac. Surf. Films 11/3 (1993) 543.
[73] Y.G. Shen, Y.W. Mai, Q.C. Zhang, D.R. McKenzie, W.D. McFall, W.E.
McBride, J. Appl. Phys. 87/1 (2000) 177.
[74] S. Tamulevicius, Vacuum 51/2 (1998) 127.
[75] R.O.E. Vijgen, J.H. Dautzenberg, Elsevier Science Sa Lausanne, 1995, p. 264.
[76] S.L. Lee, D. Windover, T.M. Lu, M. Audino, Thin Solid Films 420 (2002) 287.
[77] H. Lee, S.S. Wong, S.D. Lopatin, J. Appl. Phys. 93/7 (2003) 3796.
[78] R. Knepper, B. Stevens, S.P. Baker, J. Appl. Phys. 100/12 (2006) 123508
[79] J. Narayan, V. Bhosle, A. Tiwari, A. Gupta, P. Kumar, R. Wu, Journal of Vacuum
Science & Technology A 24/5 (2006) 1948.
[80] L. Liu, H. Gong, Y. Wang, J.P. Wang, A.T.S. Wee, R. Liu, Mater. Sci. Eng. C-
Biomimetic Supramol. Syst. 16/1-2 (2001) 85.
[81] L. Liu, Y. Wang, H. Gong, J. Appl. Phys. 90/1 (2001) 416.
[82] G.M. Turner, I.S. Falconer, B.W. James, D.R. McKenzie, J. Vac. Sci. Technol. A-
Vac. Surf. Films 10/3 (1992) 455.
[83] K. Valleti, A. Subrahmanyam, S.V. Joshi, Surface & Coatings Technology 202/14
(2008) 3325.
[84] S. Timoshenko, Journal of the Optical Society of America and Review of
Scientific Instruments 11/3 (1925) 233.
[85] G.G. Stoney, Proc. R. soc. Lond. Ser. A-Contain. Pap. Math. Phys. Character
82/553 (1909) 172.
[86] W.C. Oliver, G.M. Pharr, Journal of Materials Research 19/1 (2004) 3.
[87] P.J. Kelly, R.D. Arnell, Journal of Vacuum Science & Technology a-Vacuum
Surfaces and Films 16/5 (1998) 2858.
127

[88] H.M. Mott-Smth, I. Langmuir, Phys. Rev. 28/4 (1926) 0727.
[89] I.D. Sudit, R.C. Woods, Journal of Applied Physics 76/8 (1994) 4488.
[90] D.J. Field, S.K. Dew, R.E. Burrell, Journal of Vacuum Science & Technology a-
Vacuum Surfaces and Films 20/6 (2002) 2032.
[91] P. Spatenka, I. Leipner, J. Vlcek, J. Musil, Plasma Sources Sci. Technol. 6/1
(1997) 46.
[92] M. Spolaore, V. Antoni, M. Bagatin, A. Buffa, R. Cavazzana, D. Desideri, E.
Martines, N. Pomaro, G. Serianni, L. Tramontin, Surface and coating technology (1999)
1083.
[93] B. Panda, A. Dhar, G.D. Nigam, D. Bhattacharya, S.K. Ray, Journal of Applied
Physics 83/2 (1998) 1114.
[94] C.E. Sosolik, A.C. Lavery, E.B. Dahl, B.H. Cooper, Rev. Sci. Instrum. 71/9
(2000) 3326.
[95] C.A. Davis, Thin Solid Films 226/1 (1993) 30.
[96] A. Misra, M. Nastasi, Appl. Phys. Lett. 75/20 (1999) 3123.
[97] A. Misra, M. Nastasi, Nuclear Instruments and Methods in Physics Research
Section B: Beam Interactions with Materials and Atoms (2001) 688.
[98] Y.G. Shen, Y.W. Mai, Surf. Coat. Technol. 127/2-3 (2000) 239.
[99] H. Windischmann, J. Appl. Phys. 62/5 (1987) 1800.
[100] G. Janssen, Thin Solid Films 515/17 (2007) 6654.
[101] F. Spaepen, Acta Mater. 48/1 (2000) 31.
[102] T. Hanabusa, K. Kusaka, O. Sakata, Elsevier Science Sa, 2004, p. 245.
[103] R. Abermann, R. Kramer, J. Maser, Thin Solid Films 52/2 (1978) 215.
[104] S.H. Lee, J.K. Park, Appl. Surf. Sci. 253/23 (2007) 9112.
[105] Y. Kuru, M. Wohlschlogel, U. Welzel, E.J. Mittemeijer, Surf. Coat. Technol.
202/11 (2008) 2306.
128

[106] Z.B. Zhao, Z.U. Rek, S.M. Yalisove, J.C. Bilello, Surf. Coat. Technol. 185/2-3
(2004) 329.
[107] R.O.E. Vijgen, J.H. Dautzenberg, Thin Solid Films (1995) 264.
[108] M. Itoh, M. Hori, S. Nadahara, J. Vac. Sci. Technol. B 9/1 (1991) 149.
[109] G. Janssen, A.J. Dammers, V.G.M. Sivel, W.R. Wang, Appl. Phys. Lett. 83/16
(2003) 3287.
[110] G. Janssen, J.D. Kamminga, Appl. Phys. Lett. 85/15 (2004) 3086.
[111] L.B. Freund, E. Chason, J. Appl. Phys. 89/9 (2001) 4866.
[112] J.L. Beuth, Int. J. Solids Struct. 29/13 (1992) 1657.
[113] R. Koch, Surf. Coat. Technol. 204/12-13 (2010) 1973.
[114] L. Gladczuk, A. Patel, C.S. Paur, M. Sosnowski, Thin Solid Films 467/1-2 (2004)
150.
[115] V.M. Koleshko, V.F. Belitsky, I.V. Kiryushin, Thin Solid Films 142/2 (1986)
199.
[116] R. Knepper, S.P. Baker, Appl. Phys. Lett. 90/18 (2007) 3.
[117] Chaudhar.P, Journal of Vacuum Science & Technology 9/1 (1972) 520.
[118] J.M.E. Harper, C. Cabral, P.C. Andricacos, L. Gignac, I.C. Noyan, K.P. Rodbell,
C.K. Hu, J. Appl. Phys. 86/5 (1999) 2516.
[119] M. Murakami, Journal of vacuum science and technology A- vacuum surfaces
and films (1991) 2469.
[120] C.E. Carlton, P.J. Ferreira, Acta Mater. 55/11 (2007) 3749.
[121] W.D. Nix, Metallurgical Transactions a-Physical Metallurgy and Materials
Science 20/11 (1989) 2217.
[122] J. Li, J.W. Mayer, E.G. Colgan, J. Appl. Phys. 70/5 (1991) 2820.
[123] H.J. Mathieu, M. Datta, D. Landolt, J. Vac. Sci. Technol. A-Vac. Surf. Films 3/2
(1985) 331.
129

[124] S.D. Dahlgren, Journal of Vacuum Science & Technology 11/4 (1974) 832.
[125] J.A. Thornton, Journal of Vacuum Science & Technology 15/2 (1978) 171.
[126] J.A. Thornton, Journal of Vacuum Science & Technology 15/2 (1978) 188.
130

Alphabetized Bibliography
Abadias, G. and P. Guerin (2008). "In situ stress evolution during magnetron sputtering
of transition metal nitride thin films." Applied Physics Letters 93(11): 3.
Abermann, R. (1990). Measurement of the intrinsic stress in thin metal-films, Pergamon-
Elsevier Science Ltd.
Abermann, R. (1990). Measurement of the intrinsic stress in thin metal films, Pergamon-
Elsevier Science Ltd.
Abermann, R. and R. Koch (1985). "The internal stress in thin silver, copper and gold
films." Thin Solid Films 129(1-2): 71-78.
Abermann, R., R. Kramer, et al. (1978). "Structure and internal stress in ultra thin silver
films deposited on silicon substrate." Thin Solid Films 52(2): 215-229.
Alami, J., P. Eklund, et al. (2007). "Phase tailoring of Ta thin films by highly ionized
pulsed magnetron sputtering." Thin Solid Films 515(7-8): 3434-3438.
Baker, P. N. (1972). "Preparation and properties of tantalum thin-films." Thin Solid
Films 14(1): 3-25.
Beuth, J. L. (1992). "Cracking of thin bonded films in residual tension." International
Journal of Solids and Structures 29(13): 1657-1675.
Brown, K. L. and G. W. Tautfest (1956). "Faraday-cup monitors for high-energy electron
beams." Review of Scientific Instruments 27(9): 696-702.
Bunshah, R. F. (2001). Handbook of hard coatings : deposition technologies, properties
and applications. Park Ridge, N.J. , Noyes Publications
Carlton, C. E. and P. J. Ferreira (2007). "What is behind the inverse Hall-Petch effect in
nanocrystalline materials?" Acta Materialia 55(11): 3749-3756.
Catania, P., J. P. Doyle, et al. (1992). "Low resistivity body-centered cubic tantalum thin-
films as diffusion-barriers between copper and silicon." Journal of vaccum
science & technology A- vaccum surface and films 10(5): 3318-3321.
Catania, P., R. A. Roy, et al. (1993). "Phase-transformation and microstructure changes
in tantalum thin-films induced by bias sputtering." Journal of Applied Physics
74(2): 1008-1014.
131

Chason, E., B. W. Sheldon, et al. (2002). "Origin of compressive residual stress in
polycrystalline thin films." Physical Review Letters 88(15): 4.
Chaudhar.P (1972). "Grain growth and stress relief in thin films." Journal of Vacuum
Science & Technology 9(1): 520-&.
Clevenger, L. A., A. Mutscheller, et al. (1992). "The relationship between depostion
conditions, the beta to alpha-transformatiom, and stress-relaxation tantalum thin-
films." Journal of Applied Physics 72(10): 4918-4924.
Dahlgren, S. D. (1974). "Columnar grains and twins in high-purity sputter-deposited
copper " Journal of Vacuum Science & Technology 11(4): 832-836.
Davis, C. A. (1993). "A simple model for the formation of compressive stree in thin films
by ion-bombardment." Thin Solid Films 226(1): 30-34.
Detor, A. J., A. M. Hodge, et al. (2009). "Stress and microstructure evolution in thick
sputtered films." Acta Materialia 57(7): 2055-2065.
Drese, R. J. and M. Wuttig (2006). "In situ stress measurements in zirconium and
zirconium oxide films prepared by direct current sputtering." Journal of Applied
Physics 99(12): 5.
Face, D. W. and D. E. Prober (1987). "Nucleation of body-cubic tantalum films with a
thin niobium underlayer." Journal of Vacuum Science & Technology a-Vacuum
Surfaces and Films 5(6): 3408-3411.
Fang, C. C., F. Jones, et al. (1993). "Effect of gas impurity and ion-bombardement on
stresses in sputter-deposited thin-films- a molecular-dynamics approach." Journal
of Applied Physics 74(7): 4472-4482.
Fang, D. and R. K. Marcus (1990). "Use of cylindrical Langmuir probe for the
characterization of charged-particle populations in a planer, diode glow-discharge
device." Spectrochimica Acta Part B-Atomic Spectroscopy 45(9): 1053-1074.
Field, D. J., S. K. Dew, et al. (2002). "Spatial survey of a magnetron plasma sputtering
system using a Langmuir probe." Journal of Vacuum Science & Technology a-
Vacuum Surfaces and Films 20(6): 2032-2041.
Floro, J. A., E. Chason, et al. (1997). "Real-time stress evolution during Si1-xGex
heteroepitaxy: Dislocations, islanding, and segregation." Journal of electronic
materials: 969-979.
132

Floro, J. A., S. J. Hearne, et al. (2001). "The dynamic competition between stress
generation and relaxation mechanisms during coalescence of Volmer-Weber thin
films." Journal of Applied Physics 89(9): 4886-4897.
French, B. L. and J. C. Bilello (2004). "In situ observations of the real-time stress-
evolution and delamination of thin Ta films on Si(100)." Thin Solid Films 446(1):
91-98.
Freund, L. B. and E. Chason (2001). "Model for stress generated upon contact of
neighboring islands on the surface of a substrate." Journal of Applied Physics
89(9): 4866-4873.
Freund, L. B. and S. Suresh (2003). Thin Film Materials: stress, defect formation and
surface evolution. Cambridge, Cambridge University Press.
Friesen, C., S. C. Seel, et al. (2004). "Reversible stress changes at all stages of Volmer-
Weber film growth." Journal of Applied Physics 95(3): 1011-1020.
Friesen, C. and C. V. Thompson (2002). "Reversible stress relaxation during
precoalescence interruptions of Volmer-Weber thin film growth." Physical
Review Letters 89(12).
Gladczuk, L., A. Patel, et al. (2005). "Sputter deposition of bcc tantalum films with TaN
underlayers for protection of steel." Thin Solid Films 476(2): 295-302.
Gladczuk, L., A. Patel, et al. (2004). "Tantalum films for protective coatings of steel."
Thin Solid Films 467(1-2): 150-157.
Grosser, M. and U. Schmid (2009). "The impact of sputter conditions on the
microstructure and on the resistivity of tantalum thin films." Thin Solid Films
517(16): 4493-4496.
Hanabusa, T., K. Kusaka, et al. (2004). Residual stress and thermal stress observation in
thin copper films, Elsevier Science Sa.
Harper, J. M. E., C. Cabral, et al. (1999). "Mechanisms for microstructure evolution in
electroplated copper thin films near room temperature." Journal of Applied
Physics 86(5): 2516-2525.
Hodge, A. M., R. Foreman, et al. (2005). "Residual stress analysis in thick uranium
films." Journal of Nuclear Materials 342(1-3): 8-13.
Holloway, K. and P. M. Fryer (1990). "Tantalum as a diffusion barrier between copper
and silicon." Applied Physics Letters 57(17): 1736-1738.
133

Hoogeveen, R., M. Moske, et al. (1996). "Texture and phase transformation of sputter-
deposited metastable Ta films and Ta/Cu multilayers." Thin solid films 275: 203-
206.
Hutchinson, I. H. (1987). Princicles of plasma diagnostics. cambridge, cambridge
university press.
Ino, K., T. Shinohara, et al. (1997). "Ion energy, ion flux, and ion species effects on
crystallographic and electrical properties of sputter-deposited Ta thin films."
Journal of Vacuum Science & Technology a-Vacuum Surfaces and Films 15(5):
2627-2635.
Itoh, M., M. Hori, et al. (1991). "The origin of stress in sputter deposited tungsten films
for X-ray masks." Journal of Vacuum Science & Technology B 9(1): 149-153.
Janssen, G. (2007). "Stress and strain in polycrystalline thin films." Thin Solid Films
515(17): 6654-6664.
Janssen, G., A. J. Dammers, et al. (2003). "Tensile stress in hard metal films." Applied
Physics Letters 83(16): 3287-3289.
Janssen, G. and J. D. Kamminga (2004). "Stress in hard metal films." Applied Physics
Letters 85(15): 3086-3088.
Jiang, A. Q., A. Yohannan, et al. (2003). "Investigation of the structure of beta-tantalum."
Thin Solid Films 437(1-2): 116-122.
Kelly, P. J. and R. D. Arnell (1998). "Development of a novel structure zone model
relating to the closed-field unbalanced magnetron sputtering system." Journal of
Vacuum Science & Technology a-Vacuum Surfaces and Films 16(5): 2858-2869.
Kim, S. P., H. M. Choi, et al. (1998). "A study on the crystallographic orientation with
residual stress and electrical property of Al films deposited by sputtering." Thin
Solid Films 322(1-2): 298-302.
Kinosita, K. (1972). "Recent development in study of mechanical properties of thin
films." Thin Solid Films 12(1): 17-&.
Kinosita, K. (1978). "Mechanical properties of vacuum deposited films." Thin Solid
Films 50(MAY): 205-210.
Knepper, R. and S. P. Baker (2007). "Coefficient of thermal expansion and biaxial elastic
modulus of beta phase tantalum thin films." Applied Physics Letters 90(18): 3.
134

Knepper, R. and S. P. Baker (2007). "Coefficient of thermal expansion and biaxial elastic
modulus of beta phase tantalum thin films." Applied Physics Letters 90(18):
181908
Knepper, R., B. Stevens, et al. (2006). "Effect of oxygen on the thermomechanical
behavior of tantalum thin films during the beta-alpha phase transformation."
Journal of Applied Physics 100(12): 123508
Koch, R. (1994). "The intrinsic stress of polycrystalline and epitaxial thin metal-films."
Journal of Physics-Condensed Matter 6(45): 9519-9550.
Koch, R. (2010). "Stress in Evaporated and Sputtered Thin Films - A Comparison."
Surface & Coatings Technology 204(12-13): 1973-1982.
Koch, R., D. Winau, et al. (1993). Intrinsic stress of epitaxial thin-films, Royal Swedish
Acad Sciences.
Koleshko, V. M., V. F. Belitsky, et al. (1986). "Stress relaxation in thin aluminum films."
Thin Solid Films 142(2): 199-212.
Kuru, Y., M. Wohlschlogel, et al. (2008). "Coefficients of thermal expansion of thin
metal films investigated by non-ambient X-ray diffraction stress analysis."
Surface & Coatings Technology 202(11): 2306-2309.
Lee, H., S. S. Wong, et al. (2003). "Correlation of stress and texture evolution during
self- and thermal annealing of electroplated Cu films." Journal of Applied Physics
93(7): 3796-3804.
Lee, S. H. and J. K. Park (2007). "In situ measurement of the surface stress evolution
during magnetron sputter-deposition of Ag thin film." Applied Surface Science
253(23): 9112-9115.
Lee, S. L., M. Doxbeck, et al. (2004). "Texture, structure and phase transformation in
sputter beta tantalum coating." Surface & Coatings Technology 177: 44-51.
Lee, S. L. and D. Windover (1998). "Phase, residual stress, and texture in triode-sputtered
tantalum coatings on steel." Surface & Coatings Technology 108(1-3): 65-72.
Lee, S. L., D. Windover, et al. (2002). "In situ phase evolution study in magnetron
sputtered tantalum thin films." Thin Solid Films 420: 287-294.
Li, J., J. W. Mayer, et al. (1991). "Oxidation and protection in copper and coper alloy thin
films." Journal of Applied Physics 70(5): 2820-2827.
135

Liu, L., H. Gong, et al. (2001). "Annealing effects of tantalum thin films sputtered on
[001] silicon substrate." Materials Science & Engineering C-Biomimetic and
Supramolecular Systems 16(1-2): 85-89.
Liu, L., Y. Wang, et al. (2001). "Annealing effects of tantalum films on Si and SiO2/Si
substrates in various vacuums." Journal of Applied Physics 90(1): 416-420.
Ljungcrantz, H., L. Hultman, et al. (1993). "Residual-stresses and fracture properties of
magnetron-sputtering Ti films on Si microelements." Journal of Vacuum Science
& Technology a-Vacuum Surfaces and Films 11(3): 543-553.
Mathieu, H. J., M. Datta, et al. (1985). "Thickness of natural oxide films determined by
XPS." Journal of Vacuum Science & Technology a-Vacuum Surfaces and Films
3(2): 331-335.
Mattox, D. M. (1998). Handbook of physical vapor deposition (PVD) processing: film
formation, adhesion, surface preparation and contamination control Westwood,
N.J.
Misra, A. and M. Nastasi (1999). "Limits of residual stress in Cr films sputter deposited
on biased substrates." Applied Physics Letters 75(20): 3123-3125.
Misra, A. and M. Nastasi (2001). "Intrinsic residual stresses in metal films synthesized by
energetic particle deposition." Nuclear Instruments and Methods in Physics
Research Section B: Beam Interactions with Materials and Atoms: 688-693.
Mott-Smth, H. M. and I. Langmuir (1926). "The theory of collectors in gaseous
discharges." Physical Review 28(4): 0727-0763.
Murakami, M. (1991). "Deformation in thin-films by thermal strain." Journal of vacuum
science and technology A- vacuum surfaces and films: 2469-2476.
Narayan, J., V. Bhosle, et al. (2006). "Methods for processing tantalum films of
controlled microstructures and properties." Journal of Vacuum Science &
Technology A 24(5): 1948-1954.
Nix, W. D. (1989). "MECHANICAL-PROPERTIES OF THIN-FILMS." Metallurgical
Transactions a-Physical Metallurgy and Materials Science 20(11): 2217-2245.
Ohring, M. (2002). Materials science of thin films deposition and structure. San Diego,
Academic press.
Ohring, M. (2002). Materials science of thin films deposition and structure. San Diego,
Academic press.
136

Oliver, W. C. and G. M. Pharr (2004). "Measurement of hardness and elastic modulus by
instrumented indentation: Advances in understanding and refinements to
methodology." Journal of Materials Research 19(1): 3-20.
Panda, B., A. Dhar, et al. (1998). "Relationship between plasma parameters and film
microstructure in radio frequency magnetron sputter deposition of barium
strontium titanate." Journal of Applied Physics 83(2): 1114-1119.
Pletea, M., W. Bruckner, et al. (2005). "Stress evolution during and after sputter
deposition of Cu thin films onto Si(100) substrates under various sputtering
pressures." Journal of Applied Physics 97(5): 7.
Pletea, M., W. Bruckner, et al. (2006). "In situ stress evolution of Co films sputtered onto
oxidized Si (100) substrates." Journal of Applied Physics 99(3): 5.
Pletea, M., H. Wendrock, et al. (2008). "Stress evolution during and after sputter
deposition of thin Cu-Al alloy films." Journal of Physics-Condensed Matter
20(25): 6.
Posadowski, W. M. (2001). "Plasma parameters of very high target power density
magnetron sputtering." Thin Solid Films 392(2): 201-207.
Ren, H. and M. Sosnowski (2008). "Tantalum thin films deposited by ion assisted
magnetron sputtering." Thin Solid Films 516(8): 1898-1905.
Ryu, S., K. Lee, et al. (2007). Stress Behavior of FCC metallic thin films during thermal
evaporation, Amer Scientific Publishers.
Sajovec, F., P. M. Meuffels, et al. (1992). "Structural and electrical properties of ion
beam sputter deposited tantalum films." Thin Solid Films 219(1-2): 206-209.
Sato, S. (1982). "Nucleation properties of magnetron-sputtered tantalum." Thin Solid
Films 94(4): 321-329.
Schell, N., K. P. Andreasen, et al. (2005). "On the dependence on bias voltage of the
structural evolution of magnetron-sputtered nanocrystalline Cu films during
thermal annealing." Thin Solid Films 476(2): 280-287.
Shen, Y. G. and Y. W. Mai (2000). "Effect of deposition conditions on internal stresses
and microstructure of reactively sputtered tungsten nitride films." Surface &
Coatings Technology 127(2-3): 239-246.
137

Shen, Y. G., Y. W. Mai, et al. (2000). "Residual stress, microstructure, and structure of
tungsten thin films deposited by magnetron sputtering." Journal of Applied
Physics 87(1): 177-187.
Shiojiri, M., S. Shinkai, et al. (2003). "Preparation of low-resistivity alpha-Ta thin films
on (001) Si by conventional DC magnetron sputtering." Japanese Journal of
Applied Physics Part 1-Regular Papers Short Notes & Review Papers 42(7A):
4499-4500.
Shul, R. J. and S. J. Pearton (2000). Handbook of advanced plasma processing
techniques. berlin, springer.
Shull, A. L. and F. Spaepen (1996). "Measurements of stress during vapor deposition of
copper and silver thin films and multilayers." Journal of Applied Physics 80(11):
6243-6256.
Sosolik, C. E., A. C. Lavery, et al. (2000). "A technique for accurate measurements of ion
beam current density using a Faraday cup." Review of Scientific Instruments
71(9): 3326-3330.
Spaepen, F. (2000). "Interfaces and stresses in thin films." Acta Materialia 48(1): 31-42.
Spatenka, P., I. Leipner, et al. (1997). "A comparison of internal plasma parameters in a
conventional planar magnetron and a magnetron with additional plasma
confinement." Plasma Sources Science & Technology 6(1): 46-52.
Spolaore, M., V. Antoni, et al. (1999). "Automatic Langmuir probe measurement in a
magnetron sputtering system." Surface and coating technology: 1083-1088.
Stoney, G. G. (1909). "The tension of metallic films deposited by electrolysis."
Proceedings of the Royal Society of London Series a-Containing Papers of a
Mathematical and Physical Character 82(553): 172-175.
Sudit, I. D. and R. C. Woods (1994). "A study of the accuracy of various Langmuir
probes theories." Journal of Applied Physics 76(8): 4488-4498.
Tamulevicius, S. (1998). "Stress and strain in the vacuum deposited thin films." Vacuum
51(2): 127-139.
Thompson, C. V. and R. Carel (1996). "Stress and grain growth in thin films." Journal of
the Mechanics and Physics of Solids 44(5): 657-673.
138

Thornton, J. A. (1974). "Influence of apparatus geometry and deposition conditions on
structure and topography of thick sputtered coatings." Journal of Vacuum Science
& Technology 11(4): 666-670.
Thornton, J. A. (1978). "Diagnostic methods for sputtering plasmas." Journal of Vacuum
Science & Technology 15(2): 188-192.
Thornton, J. A. (1978). "Magnetron sputtering - basic physics and application to
cylinderical magnetrons." Journal of Vacuum Science & Technology 15(2): 171-
177.
Thornton, J. A. (1978). "Magnetron sputtering - basic physics and application to
cylinderical magnetrons " Journal of Vacuum Science & Technology 15(2): 171-
177.
Thornton, J. A. (1978). "Substarte heating in cylinderical megnetron sputtering sources."
Thin Solid Films 54(1): 23-31.
Thornton, J. A. (1986). "The microstructure of sputter deposited coatings." Journal of
Vacuum Science & Technology a-Vacuum Surfaces and Films 4(6): 3059-3065.
Thornton, J. A. and D. W. Hoffman (1989). "Stress-related effects in thin-films " Thin
Solid Films 171(1): 5-31.
Thornton, J. A. and D. W. Hoffman (1989). "Stress-related effects in thin-films." Thin
Solid Films 171(1): 5-31.
Timoshenko, S. (1925). "Analysis of bi-metal thermostats." Journal of the Optical
Society of America and Review of Scientific Instruments 11(3): 233-255.
Turner, G. M., I. S. Falconer, et al. (1992). "Monte-carlo calculation of the sputtered
atoms at a substrate surface in a magnetron discharge." Journal of Vacuum
Science & Technology a-Vacuum Surfaces and Films 10(3): 455-461.
Valleti, K., A. Subrahmanyam, et al. (2008). "Growth of nano crystalline near alpha
phase tantalum thin films at room temperature using cylindrical magnetron
cathode." Surface & Coatings Technology 202(14): 3325-3331.
Vijgen, R. O. E. and J. H. Dautzenberg (1995). "Mechanical measurement of the residual
stress in thin PVD films." Thin solid films: 264-269.
Vijgen, R. O. E. and J. H. Dautzenberg (1995). Mechanical measurement of the residual
stress in thin PVD films, Elsevier Science Sa Lausanne.
139

Wang, Y. M., A. M. Hodge, et al. (2005). "Deformation twinning during nanoindentation
of nanocrystalline Ta." Applied Physics Letters 86(10): 101915.
Westwood, W. D., R. J. Boynton, et al. (1973). "Effects of argon pressure on properties
of sputtered tantalum films." Thin Solid Films 16(1): 1-25.
Windischmann, H. (1987). "An intrinsic stress scaling law for polycrystalline thin films
prepared by ion beam sputtering.” Journal of Applied Physics 62(5): 1800-1807.
Yong, C., B. C. Zhang, et al. (2002). "Cool copper template for the formation of oriented
nanocrystalline alpha-tantalum." Journal of Physical Chemistry B 106(48):
12366-12368.
Yuan, Z. L., D. H. Zhang, et al. (2003). "A new method for deposition of cubic Ta
diffusion barrier for Cu metallization." Thin Solid Films 434(1-2): 126-129.
Zhang, J. Q., Y. M. Huai, et al. (2003). "Formation of low resistivity alpha Ta by ion
beam sputtering." JOURNAL OF VACUUM SCIENCE & TECHNOLOGY B
21(1): 237-240.
Zhang, M., B. Yang, et al. (2006). "Hardness enhancement in nanocrystalline tantalum
thin films." Scripta Materialia 54(7): 1227-1230.
Zhang, M., Y. F. Zhang, et al. (2007). "Nanocrystalline tetragonal tantalum thin films."
Scripta Materialia 57(11): 1032-1035.
Zhao, Z. B., Z. U. Rek, et al. (2004). "Phase formation and structure of magnetron
sputtered chromium nitride films: in-situ and ex-situ studies." Surface & Coatings
Technology 185(2-3): 329-339.
Zhou, Y. M., Z. Xie, et al. (2008). "Effects of deposition parameters on tantalum films
deposited by direct current magnetron sputtering." Vacuum 83(2): 286-291.
140

Appendices
Appendix A: Summary of Aluminum samples

Sample power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometer)
D*
(inches)
Al-04 50 10 10 40 7
Al-05 50 10 40 128 7
Al-06 100 10 60 355 7
Al-07 100 10 90 624 7
Al-08 50 10 40 133 7
Al-09 50 10 40 85 7
⃰ D= the distance between source and substrate

141

Appendix B: Summary of copper samples

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometer)
D*
(inches)
Cu-01 50 10 10 89 7
Cu-02 100 10 60 1085 7
Cu-03 100 10 90 1527 7
Cu-04 50 10 40 362 7
Cu-06 150/200 10 10/10 780 7
Cu-07 200 10 30 1178 7
Cu-A-01 50 10 384 60 6.5
Cu-A-02 50 10 559 90 6.5
Cu-A-03 50 5 934 90 6.5
P-Cu-01 100 2 405 30 7
P-Cu-02 100 5 301 30 7
P-Cu-04 100 5 342 33 7
P-Cu-05 100 10 240 33 7
P-Cu-07 100 10 389 30 6.5
P-Cu-10 100 10 466 30 6
P-Cu-11 100 10 505 30 6
P-Cu-12 100 10 418 25 6
P-Cu-13 100 5 450 25 6
P-Cu-14 100 2 25 565 6
142

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometer)
D*
(inches)
P-Cu-15 100 2 20 424 6
P-Cu-16 100 2 20 481 6
Cu-P4-1 100 2 140 3000 6
Cu-P4-2 100 5 140 2800 6
Cu-P4-3 100 10 140 2300 6
Cu-30-P1 30 10 60 510 6
Cu-200-P1 200 10 20 750 6
Cu-P5 100 10 30 430 6
Cu-P5-1 100 10 75 900 6
Cu-P5-3 50 10 170 920 6
Cu-P5-4 100 10 130 1620 6
Cu-P5-5 200 10 65 1730 6
Cu-P5-6 100 2 95 1750 6
Cu-P5-7 200 2 45 1835 6
Cu-P5-8 50 2 170 1654 6
Cu-P5-9 50 10 270 1546 6
Cu-P5-10 100 2 60 1260 6
Cu-P5-11 100 2 60 1242 6
Cu-P5-12 100 2 60 1245 6
⃰ D= the distance between source and substrate

143

Appendix C: Summary of multilayer copper samples

Sample Power
(watts)
Pressure
(mtorr)
Number of
layers
Thickness
(nanometers)
D*
(inches)
Cu-ml-50-1 50 2 100 435 6
Cu-ml-50-2 50 2 150 650 6
Cu-ml-50-3 50 2 200 860 6
Cu-ml-100-1 100 2 163 1128 6
Cu-ml-30-1 30 2 203 510 6
Cu-Cu-6 100 2 206 1670 6
⃰ D= the distance between source and substrate

144

Appendix D: Summary of tantalum samples

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometers)
D*
(inches)
Ta-01 50 10 40 212 6
Ta-03 100 10 30 246 6
Ta-05 100 5 70 720 6
Ta-A-01 100 2 50 532 6
Ta-A-02 100 5 50 514 6
Ta-A-06 100 10 50 477 6
Ta-A-07 100 10 50 540 6
Ta-A-08 100 5 50 558 6
Ta-A-09 100 5 50 450 6
Ta-A-10 100 2 50 457 6
Ta-A-11 100 2 50 520 6
Ta-A-12 100 10 50 530 6
Ta-A-13 100 10 50 450 6
Ta-A-14 100 5 50 540 6
Ta-A-15 100 5 50 510 6
Ta-A-16 200 5 50 485 6
Ta-A-17 200 5 50 455 6
Ta-A-18 50 5 50 525 6
Ta-A-19 100 3 50 522 6
145

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometers)
D*
(inches)
Ta-A-20 100 4 50 530 6
Ta-A-21 100 6 50 560 6
Ta-A-22 100 7 50 510 6
Ta-A-23 100 8 50 410 6
Ta-A-24 100 9 50 500 6
Ta-A-25 50 5 2 10 6
Ta-A-26 50 5 20 92 6
Ta-A-27 50 2 2 10 6
Ta-A-28 50 2 20 95 6
Ta-A-29 50 10 2 10 6
Ta-A-30 50 10 20 92 6
Ta-A-33 50 5 40 233 6
Ta-A-34 50 5 10 48 6
Ta-A-35 50 2 10 52 6
Ta-A-36 50 10 10 5 6
Ta-A-37 100 5/2 50/50 980 6
Ta-A-44 100 2/5 50/50 1020 6
Ta-A-45 100 5/10 50/50 968 6
Ta-A-46 100 5 50 472 6
Ta-A-47 100 5 50 485 6
Ta-A-48 100 5 50 510 6
146

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Thickness
(nanometers)
D*
(inches)
Ta-A-49 100 5 50 565 6
Ta-A-51 100 5 90 950 6
Ta-A-52 100 10 90 940 6
Ta-A-53 100 2 90 980 6
Ta-A-54 100 2 90 1020 6
Ta-A-55 100 2 90 1032 6
Ta-A-56 100 5/10 50/50 1010 6
Ta-A-60 100 5/10 20/50 820 6
Ta-A-61 100 5/2 50/50 993 6
Ta-A-62 100 5/10 50/50 960 6
Ta-A-63 100 5/10 50/50 985 6
Ta-A-65 100 5/10 50 560 6
Ta-A-66 100 4 50 495 6
Ta-A-67 100 9 50 480 6
Ta-A-70 50 5 20 105 6
Ta-A-71 100 2/5 50/50 1075 6
Ta-A-72 100 3 50 510 6
Ta-A-81 100 5 50 588 6
Ta-A-82 100 5 50 576 6
Ta-A-83 100 5 50 534 6
Ta-A-84 100 5 50 520 6
⃰ D= the distance between source and substrate
147

Appendix E: Summary of tantalum sample with underlayers

Sample Power
(watts)
Pressure
(mtorr)
Time
(minutes)
Underlayer D*
(inches)
Ta-Ti-Si 100 10 30 Ti 6
Nb-r-Ta-1 100 10 20 Nb 6
Nb-r-Ta-1 100 10 20 Nb 6
Ta-A-31 100 2 20 α-Ta 6
Ta-A-32 50 10 20 α-Ta 6
Ta-A-38 100 5 50 Nb 6
Ta-A-57 100 10 50 Al 6
Ta-A-58 100 10 50 Nb 6
⃰ D= the distance between source and substrate

Abstract (if available)

Abstract The correlation between external (e.g. power, pressure) and internal (e.g. plasma conditions, impurities) chamber conditions to the film residual stress and microstructure has been studied for copper and tantalum sputtered thin films. To investigate the effect of external parameters on the film residual stress and microstructure, the stress evolution in copper and tantalum films at various sputtering powers and pressures was measured in-situ using Multi-beam Optical Stress Sensor (MOSS) technique. The results demonstrate that for both materials, during deposition, a compressive stress initially develops at low sputtering pressures, while at the highest sputtering pressures the stress is always tensile. In addition, profilometry measurements are compared to the final instantaneous stress value at the end of deposition measured by MOSS. Comparing the total residual stress measured ex-situ with the growth stress measured in-situ, emphasizes on the significant contribution of the relaxations mechanism after deposition. The measurements demonstrate that the contribution of the relaxation mechanism is quite significant and can be as large or larger than the growth stress and might change the stress state on the film.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Tailoring compositional and microstructural complexity in nanostructured alloys

PDF

Microstructural evolution and mechanical properties of a copper-zirconium alloy processed by severe plastic deformation

PDF

Understanding the formation and evolution of boundaries and interfaces in nanostructured metallic alloys

PDF

Long range internal stresses in deformed single crystal copper

PDF

Electrodeposition of platinum-iridium coatings and nanowires for neurostimulating applications: fabrication, characterization and in-vivo retinal stimulation/recording EIS studies of hexavalent a...

PDF

Fabrication, deposition, and characterization of size-selected metal nanoclusters with a magnetron sputtering gas aggregation source

PDF

The role of nanotwins and grain boundary plane in the thermal, corrosion, and sensitization behavior of nanometals

PDF

Chemical vapor deposition of graphene and two-dimensional materials: synthesis, characterization, and applications

Asset Metadata

Creator Afshin Navid, Anahita (author)

Core Title A comparative study of plasma conditions, microstructure and residual stress in sputtered thin films

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Mechanical Engineering

Publication Date 11/03/2012

Defense Date 03/21/2011

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Copper,magnetron sputtering,microstructure,nanostructure,OAI-PMH Harvest,plasma characteristics,residual stress,tantalum,thin films

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hodge, Andrea M. (committee chair), Goo, Edward K. (committee member), Mansfeld, Florian B. (committee member), Nutt, Steven R. (committee member)

Creator Email afshinna@usc.edu,anahitaa@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3865

Unique identifier UC1334975

Identifier etd-Navid-4533 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-462249 (legacy record id),usctheses-m3865 (legacy record id)

Legacy Identifier etd-Navid-4533.pdf

Dmrecord 462249

Document Type Dissertation

Rights Afshin Navid, Anahita

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu