You are on page 1of 200

Growth, Characterization and Luminescence and

Optical properties of Rare-Earth elements and


Transition Metals doped in Wide Bandgap Nitride
Semiconductors.

A dissertation presented to
the faculty of
the College of Arts and Sciences of Ohio University

In partial fulfillment
of the requirements for the degree
Doctor of Philosophy

Muhammad Maqbool
August 2005

This dissertation entitled


Growth, Characterization and Luminescence and
Optical properties of Rare-Earth elements and
Transition Metals doped in Wide Bandgap Nitride
Semiconductors.
By
Muhammad Maqbool
has been approved
for the Department of Physics and Astronomy
and the College of Arts and Sciences by

Martin E. Kordesch
Professor of Physics

Benjamin M. Ogles
Interim Dean, College of Arts and Sciences

MAQBOOL, MUHAMMAD. Ph.D. August 2005. Physics and Astronomy

Growth, Characterization and Luminescence and Optical properties of RareEarth elements and Transition Metals doped in Wide Bandgap Nitride
Semiconductors. (200pp)
Director of Dissertation:

Martin E. Kordesch

Rare-earth element and transition metals doped AlN, GaN and BN films were
successfully grown using reactive magnetron sputtering. The structural, optical and
luminescence properties of these nitride films were then studied using Scanning Electron
Microscopy, X-rays diffraction, Cathodoluminescence and Tube furnace. Both
amorphous and crystalline films were obtained depending on the substrate temperature
during the deposition.
Cryogenically grown amorphous films were the principal focus of this research.
The substrate were cooled using liquid nitrogen during the growth and pure amorphous
films were obtained. Crystalline films were also obtained using an electric heater to keep
substrates at high temperature. X-ray diffraction analysis was used to confirm the
structure of films.
Rare-earth elements Ho, Gd, Pr, Tm and Sm and transition metals W and Y were
doped into the nitride films by co-sputtering. The optical and luminescence properties of
these nitride materials were studied using Cathodoluminescence. Characteristic light
emissions related to these Ho+3, Gd+3, Pr+3, Tm+3 , Sm+3 , W+3 and Y+3 ions were
observed. The results show the suitability of these materials for potential applications of
light-emitting devices.

Luminescence enhancement in the nitride materials was studied by co-doping Gd


with Ho, Pr, Sm and W in nitride materials. Stripes of these materials were also prepared
and studied for luminescence enhancement. It was observed that not only the presence of
Gd but also some interference phenomena enhance luminescence in these materials. More
than 100% enhancement in luminescence shows that these techniques used for
luminescence enhancement are successful and useful for future applications.
Stopping power of AlN for electrons and depth penetration of electron were
studied by making bilayers of AlN doped with Tm+3 and Ho+3 ions. Electron beams of
different energies were allowed to penetrate in the known thickness of the
AlN:Tm/AlN:Ho bilayer. Stopping power of AlN was calculated using the numerical
values of the energies of electron beams and the thickness of the bilayers.

Approved:
Martin E. Kordesch
Professor of Physics.

DEDICATION

This work is dedicated to my late and great mother who was the fundamental
source of my achievements in life. The whole building of my achievements, honors,
distinctions and knowledge is visible and observable because of the strong base provided
by her in the form of love, affection, prays, advices, guidance and brought up. Mother if I
keep saying thanks to you for the rest of my life that is not even enough to compensate
your love. I would have never got this place without you. You are really the greatest
mother. God Bless You.

ACKNOWLEDGEMENT

On the completion of this hard but interesting journey it will be injustice not to
mention those people who helped me in completing this journey and were available for
me whenever I needed them.
First and foremost, many thanks to my advisor Dr. Martin E. Kordesch, for
providing me the opportunity to conduct this research. The knowledge, guidance, and
support you offered to me at every stage of this work put me on the path of success.
Thank you so much for your unlimited help and cooperation. I would have never
completed it without you.
My sincere expression of gratitude also goes to Dr. Hugh Richardson of the
Department of Chemistry, Ohio University, for his help in many aspects of my research. I
learnt a lot from him, especially how to work on Cathodoluminescence and Microscopy.
I am also thankful to Dr. P. G. Van Patten of the Department of Chemistry, for his
cooperation and helpful suggestions.
Many thanks to Dr. Wojciech Jadwisienczak of School of Electrical Engineering
and Computer Science for helping in the life time measurement
My thanks also go to Ryan Higgins, Randy Mulford, Roger Smith and my friend
John Peters. You guys helped me a lot during my work.
At last but not least I am thankful to my parents, wife and brothers and sisters for
their

well

wishes,

pray

and

support

throughout

this

time.

vii

Table of Contents

Abstract............................................................................................................................. iii
Dedication ...........................................................................................................................v
Acknowledgement ............................................................................................................ vi
Table of Contents ............................................................................................................ vii
List of Tables ................................................................................................................... xii
List of Figures................................................................................................................. xiii
List of Abbreviations ................................................................................................... xviii
Introduction........................................................................................................................1
Chapter 1

III-Nitride Semiconductors. ....................................................................11

1.1

Crystal Structure of III-Nitride Semiconductors........................................11

1.2

Aluminum Nitride (AlN). ..........................................................................16


1.2.1

Electrical Properties. ......................................................................16

1.2.2

Thermal and Chemical Properties..................................................17

1.2.3. Mechanical Properties....................................................................17


1.2.4. Optical Properties. .........................................................................17
1.3.

Gallium Nitride (GaN)...............................................................................19


1.3.1. Electrical Properties. ......................................................................20
1.3.2. Thermal and Mechanical Properties. .............................................21
1.3.3. Chemical Properties. ......................................................................22
1.3.4. Optical Properties...........................................................................22

1.4.

Indium Nitride (InN)..................................................................................24

viii
1.4.1. Electrical Properties. ......................................................................25
1.4.2. Mechanical and Thermal Properties. .............................................25
1.4.3. Optical Properties...........................................................................26
1.5.

Boron Nitride (BN). ...................................................................................28


1.5.1. Electrical Properties. ......................................................................28
1.5.2. Thermal and Mechanical Properties. .............................................29
1.5.3. Optical Properties...........................................................................30

Chapter 2

RF Magnetron Sputtering.......................................................................33

2.1.

Sputtering...................................................................................................33
2.1.1. RF Power. ......................................................................................35
2.1.2. Gas Pressure...................................................................................35
2.1.3. Target Material...............................................................................35
2.1.4. Target Structure and Topography. .................................................36

2.2.

Reactive Sputtering....................................................................................36

2.3.

RF Magnetron Sputtering. .........................................................................37

Chapter 3

Experimental Setup and Techniques. ....................................................40

3.1.

Film Growth and Treatment.......................................................................40


3.1.1. Growth Conditions and Sample Preparation..................................40
3.1.2. Nature of Film Growth and Substrate Temperature. .....................41
3.1.3. Thermal Annealing of the Films. ...................................................43

3.2.

Characterization Techniques......................................................................45
3.2.1. Scanning Electron Microscopy (SEM). .........................................45
3.2.2. X-ray Diffraction (XRD). ..............................................................45

ix
3.2.3. Cathodoluminescence. ...................................................................46
Chapter 4

Holmium and Gadolinium. .....................................................................50

4.1.

Growth. ......................................................................................................51

4.2.

Characterization. ........................................................................................53
4.2.1. X-Rays Diffraction.........................................................................53
4.2.2. Cathodoluminescence. ...................................................................57
4.2.3. Thermal Activation. .......................................................................64
4.2.4. XRD analysis of AlN:Ho films after Thermal Activation. ............67

4.3.

Co-Sputtered Ho and Gd Films in AlN host..............................................67


4.3.1. Deposition. .....................................................................................67
4.3.2. Energy Dispersive X-rays Spectroscopy (EDX) ...........................68
4.3.3. Cathodoluminescence and Luminescence Enhancement...............70
4.3.4. Picture of Enhanced Luminescence on Stripes..............................79

Chapter 5

Praseodymium..........................................................................................81

5.1.

Growth. ......................................................................................................82

5.2.

Characterization. ........................................................................................84
5.2.1. X-Rays Diffraction.........................................................................84
5.2.2. Cathodoluminescence of Amorphous Praseodymium. ..................87
5.2.3. Thermal Activation. .......................................................................90
5.2.4. X-rays analysis of AlN:Pr after Thermal Activation. ....................92
5.2.5. Cathodoluminescence of Crystalline Praseodymium. ...................92

5.3.

Co-Sputtered Pr and Gd Films in AlN host. ..............................................97


5.3.1. Sputtering Deposition. ...................................................................97

x
5.3.2. Cathodoluminescence and Luminescence Enhancement in Pr......97
5.3.3. Picture of Enhanced Luminescence on Stripes............................100
Chapter 6

Thulium and Samarium. .......................................................................103

6.1.

Growth. ....................................................................................................104

6.2.

Characterization. ......................................................................................104
6.2.1. X-Rays Diffraction (XRD) of Tm and Sm. .................................104
6.2.2. Cathodoluminescence of Tm. ......................................................105
6.2.3. Thermal Activation of Tm. ..........................................................111
6.2.4. Cathodoluminescence of Sm........................................................114
6.2.5. Thermal Activation of Sm. ..........................................................116
6.2.6. Addition of Gd to Sm in AlN host. ..............................................119

Chapter 7

Tungsten and Yttrium. ..........................................................................121

7.1.

Growth. ....................................................................................................122

7.2.

Characterization. ......................................................................................122
7.2.1

X-Rays Diffraction (XRD) of AlN:W films. ...............................122

7.2.2. Cathodoluminescence of AlN:W. ................................................123


7.2.3. X-Rays Analysis of AlN:Y ..........................................................131
7.2.4. Cathodoluminescence and Thermal Activation of AlN:Y...........131
Chapter 8

Stopping Power of AlN and electron depth penetration. ...................136

8.1.

Bilayer Films Growth. .............................................................................137

8.2.

Basic Idea Behind the experiment. ..........................................................137

8.3.

Cathodoluminescence. .............................................................................139

8.4.

Stopping Power of AlN............................................................................141

xi

Chapter 9.

Discussion and conclusion. ....................................................................145

9.1.

Discussion. ...............................................................................................145

9.2.

Conclusion. ..............................................................................................169

References.......................................................................................................................172
Appendix..........................................................................................................................178

xii

List of Tables

Table 1.1

Basic Properties of AlN at 300 K ..............................................................18

Table 1.2

Basic Properties of GaN at 300 K..............................................................23

Table 1.3

Basic Properties of InN at 300 K ...............................................................26

Table 1.4

Basic Properties of BN at 300 K................................................................30

Table 3.1.

Application scope and uses of CL in various fields...................................48

Table 4.1.

Summary of emissions from Ho3+ and Gd+3 ions ......................................58

Table 4.2.

Effect of Gd concentration on the luminescence of Ho 5S2 5I8


transition at 549 nm ...................................................................................77

Table 5.1.

Growth conditions for AlN:Pr, GaN:Pr and BN:Pr films ..........................83

Table 5.2.

Summary of Pr3+ ion emissions from Pr doped AlN, GaN and BN ..........87

Table 5.3.

Effect of Gd concentration on the luminescence of Pr transitions.............98

Table 6.1.

Summary of Tm3+ ion emissions from Tm doped AlN ...........................108

Table 6.2.

Summary of Sm3+ ion emissions from Sm doped AlN............................119

Table 7.1.

Effect on Gd concentration on the W and Gd luminescence ...................128

xiii

List of Figures

Figure 1.1.

Wurtzite crystal structure. White atoms represent Al/Ga/In/B


and black atoms represent nitrogen............................................................13

Figure 1.2.

Zincblende crystal structure.......................................................................14

Figure1.3.

Rocksalt or FCC crystal structure ..............................................................15

Figure 2.1.

Diagrammatical set-up of RF Magnetron sputtering deposition................38

Figure 3.1.

Picture of the RF Magnetron Sputtering system used for film


deposition...................................................................................................42

Figure 3.2.

Picture of the tube furnace 21100 used for thermal activation


of films .......................................................................................................44

Figure 3.3.

Scanning Electron Microscope used in the present work ..........................47

Figure 4.1.

Diagrammatical arrangement of film deposition on optical fiber,


thermally in contact with the holder containing liquid nitrogen................52

Figure 4.2a.

XRD spectrum of low temperature deposited AlN:Ho films.....................54

Figure 4.2b.

XRD spectrum of AlN:Ho films deposited at 700 C ...............................55

Figure 4.3.

XRD spectrum of low temperature deposited AlN:Gd..............................56

Figure 4.4a.

Energy levels diagram of Ho3+ in amorphous AlN host ............................59

Figure 4.4b.

Energy levels diagram of Gd3+ in amorphous AlN host ............................60

Figure 4.5.

CL emission from amorphous AlN:Ho in 300-750 nm range ..................61

Figure 4.6.

CL emission from amorphous AlN:Gd in 300-800 nm range ...................62

Figure 4.7.

CL emission from Holmium doped amorphous AlN


optical fiber ................................................................................................63

xiv
Figure 4.8.

Cathodoluminescence spectrum of thermally activated


AlN:Ho film ...............................................................................................65

Figure 4.9.

Comparison of thermally activated AlN:Ho film with


inactivated film ..........................................................................................66

Figure 4.10.

XRD analysis of AlN:Ho films after thermal activation. ..........................69

Figure 4.11.

EDX spectrum of AlN:Ho .........................................................................71

Figure 4.12a. EDX spectrum of AlN:1Ho1Gd.................................................................72


Figure 4.12b. EDX spectrum of AlN:1Ho2Gd.................................................................73
Figure 4.12c. EDX spectrum of AlN:1Ho3Gd.................................................................74
Figure 4.12d. EDX spectrum of AlN:1Ho4Gd.................................................................75
Figure 4.13.

Effect of Gd concentration on Ho luminescence .......................................76

Figure 4.14.

Increasing pattern of Ho luminescence with Gd concentration .................78

Figure 4.15.

CL picture of the enhanced luminescence in Ho by Gd ............................80

Figure 5.1.

XRD spectrum of AlN:Pr films deposited at liquid nitrogen


temperature ................................................................................................85

Figure 5.2.

XRD spectrum of AlN:Pr films deposited at 700 C.................................86

Figure 5.3.

Energy levels diagram of Praseodymium in amorphous


AlN host .....................................................................................................89

Figure 5.4.

CL spectra of praseodymium doped amorphous AlN, GaN and BN.........91

Figure 5.5.

Comparison of CL spectra from thermally activated and


inactivated AlN:Pr .....................................................................................94

Figure 5.6.

XRD spectrum of amorphous AlN:Pr after thermal activation. ................95

Figure 5.7.

Room temperature Cathodoluminescence of crystalline AlN:Pr...............96

xv
Figure 5.8.

Effect of Gd concentration on the luminescence of AlN:Pr films ............99

Figure 5.9.

Geometry of the AlN:1Pr4Gd/AlN:1Pr films deposited using a mask....101

Figure 5.10.

CL picture of the enhanced luminescence in Pr by Gd............................102

Figure 6.1.

XRD spectrum of AlN:Tm deposited at low temperature .......................106

Figure 6.2.

XRD spectrum of AlN:Sm deposited at low temperature .......................107

Figure 6.3.

Cathodoluminescence of amorphous AlN:Tm.........................................109

Figure 6.4.

Energy levels diagram of Tm in amorphous AlN host ............................110

Figure 6.5.

CL spectra from thermally activated AlN:Tm films................................112

Figure 6.6.

Comparison of CL spectra from thermally activated AlN:Tm


films .........................................................................................................113

Figure 6.7.

CL spectra from amorphous AlN:Sm films .............................................115

Figure 6.8.

Comparison of CL spectra from amorphous AlN:Sm films


exposed to 0.280 mA and 0.320 mA currents..........................................117

Figure 6.9.

CL spectra from amorphous AlN:Sm films before and after


thermal activation.....................................................................................118

Figure.6.10.

Effect of Gd concentration on the luminescence of AlN:Sm


Films ........................................................................................................120

Figure 7.1.

XRD analysis of W doped AlN films ......................................................124

Figure 7.2.

CL spectrum of amorphous AlN:W.........................................................125

Figure 7.3.

Effect of Gd concentration on the luminescence of AlN:W films...........127

Figure 7.4.

The effect of thermal activation and W concentration on Ho


Luminescence ..........................................................................................130

xvi
Figure 7.5.

X-rays analysis of the structure of low temperature deposited


AlN:Y films .............................................................................................132

Figure 7.6.

CL spectrum of amorphous AlN:Y..........................................................133

Figure 7.7.

CL spectrum of thermally activated amorphous AlN:Y films.................135

Figure 8.1.

A. Penetration of low energy electron beam in the AlN:Tm/AlN:Ho


bimetal target.
B. Penetration of high energy electron beam in the AlN:Tm/AlN:Ho
bimetal target. ....................................................................................140

Figure 8.2.

CL spectra of AlN:Tm/AlN:Ho bilayer at different electron beam


energies. ...................................................................................................142

Figure 8.3.

Variation in the intensity of 549 nm AlN:Ho peak with beam


energy.......................................................................................................143

Figure 9.1.

CL picture of the enhanced luminescence in Ho by Gd. .........................147

Figure 9.2.

CL picture of three layers of AlN:Ho deposited upon each other using a


mask. ........................................................................................................148

Figure 9.3.

Intensity of luminescence from AlN:Ho film with the angle of


laser beam. ...............................................................................................152

Figure 9.4.

Ray-diagram of the path followed by laser beam giving interference.....153

Figure 9.5.

Comparison of EDX spectra of AlN:Ho and AlN:1Ho.xGd. ..................158

Figure 9.6.

Effect of Gd concentration on Ho luminescence. ....................................159

Figure 9.7.

Excitation spectrum of AlN:Ho, using Xenon lamp................................160

Figure 9.8.

Effect of Gd dopant on the luminescence of Pr. ......................................161

Figure 9.9.

Picture of luminescence enhancement in Pr due to Gd dopant...............162

Figure 9.10.

Spectrum of the enhanced luminescence in Sm.......................................163

xvii
Figure 9.11.

Effect of Gd concentration on the luminescence of Tungsten.................164

Figure 9.12.

Energy levels shift in W due to Ho addition............................................167

Figure 9.13.

Comparison of AlN, GaN and BN hosts for the observation of


335 nm and 385 nm peaks in Pr...............................................................168

Figure A1.

Top. SEM picture of the cross-section of AlN:Ho film around fiber.


Bottom.

Figure A2.

Picture of the same fiber from a side ......................................179

Cross-sectional images of AlN:Ho films deposited around fiber............180

xviii

List of Abbreviations.

RE.

Rare-Earth

RE+3.

Trivalent rare-earth ions

AlN.

Aluminum Nitride

GaN.

Gallium Nitride

BN.

Boron Nitride

InN.

Indium Nitride

RF.

Radio Frequency

CL.

Cathodoluminescence.

SEM.

Scanning Electron Microscopy.

XRD.

X-rays diffraction.

Ho.

Holmium.

Gd.

Gadolinium.

Cr.

Chromium.

Tm.

Thulium

Sm.

Samarium.

Pr.

Praseodymium.

UV.

Ultraviolet.

IR.

Infrared.

W.

Tungsten.

Y.

Yttrium.

eV.

Electron Volt.

xix
nm.

Nanometer.

III.

Group three.

V.

Group five.

HCP.

Hexagonal closed packed.

FCC.

Face Centered Cube.

NaCl.

Sodium Chloride.

Al.

Aluminum

Ga.

Gallium.

B.

Boron.

In.

Indium.

Si.

Silicon.

GaAs.

Gallium Arsenoide.

AlGaN.

Aluminum Gallium Nitride.

AlGaN/GaN. Aluminum Gallium Nitride and Gallium Nitride Multilayers.


.

Ohm.

Thermal conductivity.

Thermal expansion coefficient.

GPa.

Gega Pascal.

LED.

Light emitting diode.

MODFET.

Modulation Doped Field Effect Transistors

SiC.

Silicon Carbide.

SiO2

Silicon Dioxide.

Sputter yield

xx
UHV.

Ultra high vacuum.

Micrometer

TEM.

Transmission Electron Microscopy.

AlN:Ho.

Holmium doped aluminum nitride.

AlN:Gd.

Gadolinium doped aluminum nitride.

Ho3+

Trivalent holmium ion.

Gd3+

Trivalent gadolinium ion.

Cr3+

Trivalent chromium ion.

Pr3+

Trivalent praseodymium ion.

Sm+3

Trivalent samarium ion.

a.u.

Arbitrary Units.

INTRODUCTION

Rare earth elements are composed of the fifteen lanthanides and fifteen actinides.
Lanthanides are usually famous for their unique optical and luminescence properties
when incorporated in a glass, ceramic or semiconductors. Wide bandgap III-V
semiconductors are amongst the favorite hosts for the development of photonic and
optoelectronic devices exploiting the optical and luminescence properties of lanthanides
[1]

. When these materials are excited by various means intense sharp-line emission is

observed due to intra-4fn-shell transitions of the rare-earth ion core

[2,3]

. Spectroscopic

properties of trivalent rare-earth ions (RE+3) result from their atomic structure. The
neutral atom electron configuration and the electron configuration in RE+3 are shown in
appendix A. All the lanthanides have configuration like Xenon with additional electrons
in 6s, 5d and 4f orbital. Since the 5s and 5p shells efficiently shield the 4f shell, the host
environment has only a weak influence on the 4f electrons. This weak perturbation is in
fact responsible for the spectral fine structure and the intensities of spectral transitions.
These transitions are responsible for the emission of light in a very wide range from
ultraviolet to far infrared region. But how energy is transferred from the host matrix to
RE ion when it is doped to the host? There is a good explanation for this question.
Emission from RE ions originates from excited states of the 4f-electron shells. Since
these electrons do not hybridize with the sp3 orbitals of the III-V semiconductor host,
there is no overlap between electron wave functions of the host the RE dopant. Therefore,
the energy of a recombining electron-hole pair optically (or electrically) excited in the
host cannot be transferred to the RE ion core by exchange interaction. Consequently, the

2
excitation process has to rely on dipole-dipole (or higher order poles) interactions that are
significantly less efficient [4].
Efficiency of the energy transfer process is enhanced when an impurity state is
created within the energy bandgap of the host by the RE dopant. In that case, energy
released by nonradiative recombination of an electron-hole pair localized at such a RErelated level can easily be taken over by the 4f-electrns, giving rise to core excitation.
In principle, for RE ions embedded in a semiconductor matrix, three different
excitation paths are possible [4].
1.

Direct energy absorption by the RE ion.


This process is identical to that for RE ions in insulators and can be realized only

under resonant conditions: the light quantum must precisely match the energy difference
between the ground and one of the excited states of the RE ion core. Since transitions
within the 4f-electron states are parity forbidden, the cross-section for such a direct
absorption is very small.
2.

Band-to-and host absorption with subsequent energy transfer to the RE ion.


Cross-section of the band-to-band absorption in a semiconductor is very large,

exceeding by many orders of magnitude as compared to the resonant core transitions. In


this case, the energy transfer between the semiconductor and the optical dopant limits the
excitation efficiency.
3.

Activation of specific RE-related levels in the bandgap of the host.


If the RE ion excitation process involves formation of one of multiple

intermediate stages, these can be directly activated by (resonant) photons with energies
lower than that of the bandgap of the host.

3
The energy transfer leading to activation of the optical dopant can be revealed by
excitation spectroscopy. For the direct core absorption (process 1), the excitation
spectrum reflects the discrete energy levels of the 4f-electron shell. The indirect
excitation process (2) is only possible for photons with energy quantum exceeding the
bandgap energy of the host. If the energy transfer process between the host and the RE
ion involves intermediate stages (3), these appear as specific peaks in the excitation
spectrum for the sub-bandgap energy range.
G. Ajithkumar

[5]

and E. D. Rosa-Cruz

[6]

established relationships for the

probability of energy transfer (Pet) from host to the active ion and its efficiency (et).
They can be estimated from the expressions given below.
Pet = (0 - d) / 0d

(1)

et = (0 Pet) / (1+0 Pet) = 1- d / 0

(2)

Using and playing with these transitions of RE ions and lanthanides, people are
widely fabricating light emitting devices, used in optical displays, electroluminescence
devices, optical communications and optoelectronic devices

[7-12]

. Work is still in

progress for advanced technology development and applications of these lanthanides,


which makes it important to investigate further these rare-earth elements.
The present work mainly reports the growth, surface characterization and optical and
luminescence properties of rare-earth elements (Holmium, Gadolinium, Praseodymium,
Thulium and Samarium) doped in wide bandgap III-nitride semiconductors (AlN,GaN
and BN). Two members of transition metals (Tungsten and Yttrium) doped in these
semiconductors are also investigated for the same properties. Holmium emits in a wide
range from ultraviolet to infrared with the most intense emission in the green.

4
Praseodymium is famous for its red and blue emission mainly. However it also emits in
the violet and infrared regions. Thulium is usually used for its blue and infrared
luminescence with some emission in ultraviolet and relatively weak emissions in red.
Samarium gives intense peaks in the red and yellow regions with some less intense
transitions in infrared as well. Both tungsten and yttrium of the transition metals group
provide very wide and broad emission in the blue region of electromagnetic spectrum.
The unique aspect of this work is the luminescence enhancement in Ho+3, Pr+3,
Sm+3 and W+3 ions by co-doping Gadolinium to these ions in an amorphous AlN host.
Gd+3 is a strong ultraviolet emitter in 314 nm while Ho+3, Pr+3, Sm+3 and W+3 ions emit
longer wavelengths as compared to Gd

+3

when exposed to an electron beam. This

relatively high energetic and strong emission from Gd+3 leads us to think in a new
direction. Does adding gadolinium as a co-dopant to lanthanide ions increase the
luminescence from the lanthanide ions? What could be the possible mechanism of
luminescence enhancement if answer to the first question is yes? Obtaining an answer to
these interesting questions are the main objectives of this work. Two mechanisms are
tested for this purpose

Direct UV excitation of lanthanide ions by the sharp and intense UV


emission from Gd+3.
Energy transfer from Gd+3 to lanthanide ions.

People have studied these two processes and have found that the chances of direct
excitation are very rear because the transitions within 4f-electron states are parity
forbidden

[4]

. The energy transfer and fluorescence lifetime are studied in detail and

5
different theories, to give proper explanation to such kind of transfer processes, are
proposed.
Two main theories of energy transfer commonly employed for rare-earth and
transition metal ions are those of Forster and Dexter and of Burshtein

[13-16]

. In both

theories, energy transfer processes depend strongly on the overlap of emission crosssection em and absorption cross-section abs, expressed by a critical radius Rsx in
equation (3), where s stands for sensitizer and x stands either for s in case of energy
migration between sensitizers or for acceptor in case of energy transfer from sensitizers
to acceptors.
R6sx = (3cs / 84n2) sem()xabs()d

(3)

Where sem and xabs are the emission and absorption cross-sections respectively, s is the
fluorescence decay time of the unperturbed sensitizer, c is the speed of light and n is the
refractive index.
The classic Forster-Dexter theory assumes that sensitizers and acceptors are
randomly and uniformly distributed and that sensitizers do not interact at all.
Fluorescence decay following the Forster-Dexter law obeys the form exp(- t) with
na resulting in a nonexponential decay with no influence of sensitizer concentration on
decay time. However since the overlap integral in equation (3) is not zero for some some
interactions (like praseodymium-erbium and ytterbium-ytterbium interactions) there is a
probability for energy transfer between the two ions. The nonradiative dipole-dipole
energy transfer rate varies[17] with (Rsx / R)6. So the energy transfer changes drastically
when the average distance between the ions comes close to Rsx. Mathematically, the
transfer probability, according to Forster-Dexter theory, can be expressed as [5,18]

6
W = s=6,8,10 [1 / s (R0 / R)s ]

(4)

Where s = 6,8,10 stands for dipole-dipole, dipole-quadrupole and quadrupole- quadrupole


interactions respectively, R is the separation between donor and acceptor and R0 is the
critical separation.
Forster also obtained a relation for donor emission surrounded with randomly
distributed acceptors. This relation is given by [19]
(t) = (0) exp[-t / s- na(163Ct/9)1/2]

(5)

Where na is the acceptor concentration and C is the dipole-dipole interaction parameter


between energy donor and acceptors.
The Burshtein model includes the effect of energy migration among sensitizers.
The physical picture is that energy migrates among sensitizers until it finds acceptor ions.
With the Burshtein model, fluorescence decays as exp(-Wt), with
W = ([ (2/3)5/2R3saR3ss] / s ) ns na ,

(6)

leading to a single exponential decay with a time constant depending on sensitizer as well
as on acceptor concentrations

[14]

. ns and na gives the concentrations of sensitizers and

acceptors respectively.
Another important and valuable section of this dissertation is the exploitation of
the luminescence from lanthanide ions to calculate the stopping power of amorphous AlN
exposed to electron beam. The Stopping power of a material for an electron beam is the
energy lost by electrons per unit length of the material when electron beam pass through
it. If we know the energy of an electron beam and the penetration depth of that beam in
AlN, then the stopping power can be easily calculated. This procedure is followed in this
work in calculating the stopping power of AlN using electron beam obtained from the

7
electron gun in cathodoluminescence apparatus. Ho+3 and Tm+3 ions are doped into
amorphous AlN to trace the penetration depth of electron beam and the exact thickness of
AlN crossed by electron beam when it penetrates in AlN.
Chapter 1 provides information about the structure and properties of aluminum
nitride (AlN), gallium nitride (GaN), indium nitride (InN) and boron nitride BN. Crystal
structure description, electrical, optical, mechanical, thermal and chemical properties of
these semiconductors are discussed in this chapter. Tables are given, providing detailed
information about these materials in a more compact form.
Chapter 2 gives information about RF Magnetron Sputtering method of film
growth and deposition. Basic concepts and ideas about sputtering and its use in the
deposition of thin films are explained in this section. Factors affecting the rate of film
deposition and film quality are also parts of this chapter.
Chapter 3 gives a brief description of the experimental techniques used in this
work. The film growth mechanism, conditions and substrates used for the growth are
discussed. Brief discussion about the Cathodoluminescence (CL), Photoluminescence
(PL), Scanning Electron Microscopy (SEM), X-rays diffraction (XRD) and Thermal
Annealing for the characterization of thin films is covered this chapter.
Chapter 4 gives a detailed analysis of holmium (Ho) and gadolinium (Gd). This
chapter opens a discussion of the thin film deposition of these elements in AlN host on
different

substrates

followed

by

films

analysis

using

Cathodoluminescence.

Cathodoluminescence is widely used to investigate these materials and to identify the


transitions, luminescence and intensities of different transitions in these materials. The
co-doped films of Ho and Gd together in AlN host and the effect of Gd concentration on

8
the luminescence from Ho is studied using CL. Thermal annealing is performed to see the
response of these elements to high temperatures. XRD and SEM characterization
provides information about the structure of the films.
A detailed discussion about praseodymium (Pr) is given in chapter 5. This chapter
provides information about the growth of Pr in AlN, GaN and BN hosts and the response
of Pr to luminescence in different hosts. Cathodoluminescence is performed at room
temperature for Pr deposited in each of the three hosts and a comparison is made in the
CL spectra. Both amorphous and crystalline grown films are discussed and compared in
this chapter. XRD spectra of all films are taken to characterize the structure of these films
and its effect on the luminescence of the material. Effect of Gd on the luminescence of Pr
is also studied in this chapter.
Thulium and Samarium are the parts of chapter 6. Growth thin films and their
surface, optical and luminescence characterization are the parts of this chapter. This films
of these elements in AlN host are grown in the same way using r.f. magnetron sputtering
and

characterized

for

their

structures

using

x-rays

diffraction

analysis.

Cathodoluminescence gives details about the luminescence from these elements and
identifies their suitability for use in UV, visible or IR applications. Thermal annealing
helps in understanding the temperature effect on the luminescence and is a part of this
chapter. Effect of Gd on the luminescence of these elements is also included in this
section.
Chapter 7 gives information about the two transition elements, tungsten (W) and
yttrium (Y), investigated in the present work. Thin films of these elements are deposited
in AlN host using the same conditions as maintained for the other films. XRD is used for

9
surface characterization and structure analysis. Cathodoluminescence provides data about
the luminescence from these materials. The effect of W on Ho luminescence and the
effect of Gd on W luminescence are also parts of this section. W is co-sputtered with Ho
and Gd for this purpose. Gd enhances the luminescence from W but W does not have any
effect on Ho luminescence.
Chapter 8 discusses the stopping power of AlN and the penetration depth of
electron beam when passes through a bilayer of AlN:Tm and AlN:Ho film. The energies
of electron beam and the thickness of AlN:Tm and AlN:Ho layers are known during the
penetration process. AlN:Tm film is deposited on the top of AlN:Ho film. Low energetic
beam does not have enough energy to cross the AlN:Tm film and hence below a specific
energy of the beam we should get the luminescence spectrum of AlN:Tm only. Increasing
the energy of the electron beam makes it to penetrate deeply inside AlN. And hence a
stage should reach that for a certain energy the electron beam should just cross the
AlN:Tm layer and touch the AlN:Ho layer. At this stage the spectrum should not only
give the luminescence from AlN:Tm but the spectrum of AlN:Ho should also start
appearing. This method is followed to calculate the stopping power of AlN in this
chapter.
Chapter 9 is the discussion and conclusion section of this dissertation. The results
obtained from different parts and chapters of this work are discussed in this chapter. The
luminescence enhancement in the rare-earth ions doped in amorphous AlN is the main
topic of discussion. The results are tested using different experimental techniques. The
increasing concentration of Gd+3 in the films is verified by energy disperssive X-rays
spectroscopy (EDX). The chances of direct excitation of Ho+3 by the UV emission from

10
Gd+3is checked by measuring the excitation spectrum of Ho+3 in UV region. The
probability of energy transfer from Gd+3 into Ho+3 is studied by measuring the life time of
Ho+3 before and after the doping of Gd+3.
AlN is the principal host used in all materials because its wide bandgap of 6.2 eV
makes it transparent to the electromagnetic spectrum above 200 nm. GaN and BN hosts
are also used. References are given at the end of the whole work.

11

Chapter 1

III-NITRIDE SEMICONDUCTORS
(AlN, GaN, InN And BN)
Nitride semiconductors represent one of the most exciting research areas in the
semiconductor field, having excellent materials, optical and electrical properties [13, 20-24].
Because of these properties they are used as hosts for a number of luminescent materials.
Despite their wide use there are some applications for which they are not suitable. Those
having a small band gap cannot be used for short wavelength emitter applications, such
as laser printers, high density optical storage, full-color display, and underwater
communications. III-V nitrides overcome many of these deficiencies but still fall short in
the area of high temperature devices. Some III-nitrides, however, do not share these
problems. They are suitable in the short wavelength and visible applications due to their
large band gaps [13, 20-27]. Their large bond strengths, especially AlN and GaN have made
them very well suited for high temperature applications
ability to form alloys

[13, 22,27]

. They also have the

[28,16]

. This chapter provides information about the properties of

these III-nitride semiconductors (AlN, GaN, InN and BN).

1.1 Crystal Structure of III-Nitride Semiconductors:


There are three common crystal structures shared by the group-III nitrides:
wurtzite, zincblende, and rocksalt structures. At normal growth conditions AlN, GaN, and
InN thermodynamically favor the wurtzite structure, while BN is zincblende

[13,22,25]

12
GaN, AlN and InN can be grown in both the zincblende and rocksalt structures under
very high pressure.
Figure 1.1 shows the wurtzite structure. The wurtzite structure has a hexagonal
unit cell and consists of two interpenetrating hexagonal closed packed (HCP) sublattices,
each with one type of atom, offset along the c-axis by 5/8 of the cell height.
The zincblende structure has a cubic unit cell, consisting of two interpenetrating
face centered cubic (FCC) lattices with the positions of the atoms the same as in
diamond. Figure 1.2 represents the zincblende structure. Each atom in the structure may
be viewed as positioned at the center of a tetrahedron, with its four nearest neighbors
defining the four corners of the tetrahedron.
Rocksalt or NaCl, structure is also cubic in structure with two interpenetrating
FCC structures. Each atoms has six nearest neighbors located at the corners of an
octahedron, as shown in figure 1.3.

13

Fig. 1.1. Wurtzite crystal structure. White atoms represent Al/Ga/In/B and black
atoms represent nitrogen.

14

Fig. 1.2. Zincblende crystal structure.

15

Fig.1.3. Rocksalt or FCC crystal structure.

16

1.2 Aluminum Nitride (AlN):


AlN was first synthesized in 1877. It posses wide range of mechanical and
electronic properties and is one of the most important materials for use in the
optoelectronic devices. Its hardness, high thermal conductivity, resistance to high
temperature and a reasonable thermal match to silicon (Si) and gallium arsenide (GaAs)
make it an attractive material for electronic packaging applications. The wide bandgap is
also a reason for AlN to be touted as an insulating material in semiconductor device
application. Piezoelectric properties make AlN suitable for surface-acoustic-wave device
applications [13,22,29] . However, the majority of interest in this semiconductor stems from
its ability to form alloys with GaN producing AlGaN and allowing the fabrication of
AlGaN/GaN based electronic and optical devices. Some of the useful properties of AlN
are given below.

1.2.1. Electrical Properties.


The electrical characterization of AlN has usually been limited to resistivity
measurements due to the low intrinsic carrier concentration, and the deep native defect
and impurity energy levels. For pure AlN these measurements are consistently found to
be = 1011-1013 *cm by different people

[30-33]

. However impure crystals have much

lower resistivity = 103-105 *cm.


AlN is presently grown with much improved quality and shows both n- and ptype conductions. This has made it possible to measure both the electron and hole
mobilities. Edwards et al

[30]

and Kawabe et al

[34]

carried out some Hall measurements

in p-type AlN which gave an estimate of the hole mobility p= 14 cm2 / Vs at 290 K.

17

1.2.2. Thermal and Chemical Properties.


AlN is an extremely hard ceramic material with a melting point higher than 2000
C. The thermal conductivity

[13,22]

of AlN at 300K is 2.85 W. cm

-1

K-1. However it

varies with temperature. Thermal expansion of AlN is isotropic with a room temperature
value

[13,22,25]

of a = 4.2 10-6 K-1 and c = 5.3 10-6 K-1. AlN exhibits inertness to

many etches. A number of AlN etches have been reported in the literature. However,
none of these etches have been performed on high quality single crystal AlN.

1.2.3. Mechanical Properties.


The bulk properties of AlN were performed on single crystalline AlN. The
measured Bulk Modulus B is ranging from 160 GPa to 201 GPa while the Youngs
Modulus E is found to be 193-208 GPa. The hardness of AlN has been measured
[13,22,35,36]

to be 12 GPa on the basal plane (0001). Some anisotropy in hardness has been

observed with the direction perpendicular to the c-axis with measured values in the range
of 10 14 GPa.

1.2.4. Optical Properties.


Optical and luminescence properties of AlN are studied by many people in
different kinds of AlN films and they found that AlN lattice has very large affinity to
oxygen and it is almost impossible to eliminate oxygen contamination in AlN. Currently,
commercially available AlN contains about 1 1.5 at. % oxygen. However Yim et al

18
[13,22]

have characterized high- quality AlN by optical absorption and determined the room

temperature bandgap to be 6.2 eV.


Measurements of the refractive index of AlN have been carried out for
amorphous, polycrystalline and single epitaxial thin films. The values of refractive index
are in the n = 1.99 2.25 range with the most reported value of n = 2.15 0.05. These
values are found to increase with increasing structural order, varying between 1.8 to 1.9
for amorphous films, 1.9 to 2.1 for polycrystalline films and 2.1 to 2.2 for single crystal
epitaxial films. Some of the basic properties of AlN are given in table 1.1

Table1.1.

Basic Properties of AlN at 300 K [13,22,35].

Basic Parameters:
Crystal Structure
Number of atoms per cm3
Debye Temperature (K)
Density (g cm-3)
Dielectric Constant
static
high frequency
Effective electron mass (in units of m0)
Effective hole mass (in units of m0):
Heavy
for kz direction mhz
for kx direction mhx
Light
for kz direction mlz
for kx direction mlx
Split-off band
for kz direction msoz
for kx direction msox
Lattice constant (A)
Electron affinity (eV)
Optical phonon energy (meV)
Energy gap (eV)

Wurtzite
9.58 1022
1150
3.23
8.5 0.2
4.68
0.4
3.53
10.42
3.53
0.24
0.25
3.81
a = 3.112
C = 4.982
0.6
99
6.2

19
Table1.1. continued.
Energy of spin-orbital splitting Eso (eV)
Effective conduction band density of states (cm-3)
Effective valance band density of states (cm-3)

0.019
6.3 1018
4.8 1020

Electrical properties:
Resistivity (.cm)
Breakdown field (V cm-1)
Mobility (cm2 V-1 S-1)
electrons
holes
Diffusion coefficient (cm2 s-1)
electrons
holes
Electron thermal velocity (m s-1)
Hole thermal velocity (m s-1)

1011-1013
(1.2 1.8) 106
300
14
7
0.3
1.85 105
0.41 105

Optical properties:
Refractive index
Radiative recombination coefficient (cm3 s-1)

2.15 0.05
0.4 10-10

Thermal and Mechanical properties:


Bulk modulus (dyn cm-2)
Melting point (C)
Specific heat (J g-1 C-1)
Thermal conductivity (W cm-1 C-1)
Thermal diffusivity (cm2 s-1)
Thermal expansion, Linear (C-1)

21 1011
2750 (between 100
and 500 atm of
nitrogen).
0.6
2.85
1.47
a = 4.2 10-6
c = 5.3 10-6

1.3. Gallium Nitride (GaN):


GaN was produced by Juza and Hahn [13,22] in 1938, by passing ammonia over hot
gallium. In 1959 Grimmeiss et al [13,22] measured the first PL spectra. Pankove et al [13,22]
fabricated a blue LED from GaN for the first time in 1972. After these initial attempts

20
extensive work was done in GaN by many people [13,22, 35,38]. But its growth often suffered
from large-background n-type carrier concentrations due to native defects and possibly
impurities [13,22]
These together with the difficulties in obtaining p-type doping, catalyzed the early
investigations and hence information available in the literature regarding many of the
physical properties of GaN is still in the process of evolution. Recent interests have led to
an improvement in the crystal growth and processing and many difficulties have been
overcome.
The room temperature electron concentrations of GaN is found to be 5 1016 cm-3.
The successful attempts to obtain p-type GaN has led to the demonstration of excellent pn-junction LEDs in the violet, blue, green and yellow bands of the visible spectrum with
brightness suitable for outdoor displays, lasers, power Modulation Doped Field Effect
Transistors (MODFETs) and UV detectors. Useful properties of GaN obtained from
literature are given below.

1.3.1. Electrical Properties.


The electrical characterization of GaN has been done by different researchers. The
electron mobility in wurtzite and Zinc Blende structures is found to be 1000 cm2 V-1s-1.
However the hole mobility in GaN is different in different crystalline structures. In
Wurtzite structure its value is 200 cm2 V-1s-1 while in Zinc Blende structure its 350
cm2 V-1s-1. The diffusion coefficient for electrons is 25 cm2 s-1 in both types of structures
and for holes its 5 cm2 s-1 in Wurtzite and 9 cm2 s-1 in Zinc Blende structure. The thermal
velocity for electrons is reported to be 2.6 105 m s-1 in Wurtzite and 3.2 105 m s-1 in

21
Zinc Blende structure. This value for holes is 9.4 104 m s-1 and 9.5 104 m s-1 in
Wurtzite and Zinc Blende structures respectively.

1.3.1. Thermal and Mechanical Properties:


The lattice parameters of GaN at room temperature are a = 3.1892 0.0009 A
and c = 5.1850 0.0005 A. However for the polycrystalline type zinc blende the
calculated lattice constant, based on the measured GaN bond distance in wurtzite GaN, is
a = 4.503 A . The measured value for this poly type varies between 4.49 and 4.55 A,
indicating that the calculated results lies within acceptable limits

[13]

. Overall, the lattice

constant of GaN is a function of growth conditions, impurity concentrations and film


stoichiometry

[13]

. For example, when doped heavily with Zn and Mg

[13]

a lattice

expansion occurs because, at high concentrations, the group-II element begins to occupy
the lattice sites of the much smaller nitrogen atom.
Information is also available about the coefficient of thermal expansion () of
GaN over 300 - 900 K with a mean coefficient of thermal expansion of GaN in the c
plane to be 5.59 10-6 K-1. Further investigations show that in the c-direction the mean
value of is 3.17 10-6 K-1 and 7.75 10-6 K-1 in the 300 700 K and 700 900 K
respectively. Sheleg and Savastenko reported a thermal expansion coefficient
perpendicular and parallel to the c-axis at 600 K, of (4.52 0.5) 10-6 K-1 and (5.25
0.5) 10-6 K-1 respectively.
Thermal Conductivity , of GaN was measured and reported by Sichel Pankove
[13]

in the temperature range of 25 360 K. The room temperature value of thermal

22
conductivity = 1.3 W/cm.K is a little smaller than the predicted value of 1.7 W/cm.K
[13]

. The Debye temperature D of GaN at 0 K was calculated to be D = 600 K

1.3.2. Chemical Properties:


A large set of information, after Johnson et al [13] first synthesized it in 1932, has
repeatedly indicated that GaN is a stable compound and exhibits significant hardness. It is
this chemical stability combined with its hardness that has made GaN an attractive
material for prospective coatings. Further, due to its wide bandgap, GaN is also an
excellent candidate for device operation in high temperature and caustic environment.
Since the materials characteristics depend, to a large extent, on the growth conditions and
hence different characteristics of GaN, studied in different laboratories, were obtained
from different sources that led to inconsistent results. For example some experimental
studies of the stability of GaN conducted at high temperatures suggested that significant
weight loss occur at temperatures as low as 750 C. Others contradicted this proposal and
suggested that no significant weight loss should occur even at a temperature of 1000 C.
Various spectroscopic techniques, such as Auger electron spectroscopy, X-ray
photoemission spectroscopy and electron energy loss spectroscopy have been very useful
for the study of the surface chemistry of GaN.

1.3.3. Optical Properties:


GaN grown over SiO2 on SiC substrates was studied for its optical properties by
John Torvik et al. Strong band-edge luminescence was observed at 3.40 eV on both SiO2
/ SiC and SiC. However A. Koo et al

[13]

have report that the optical absorption

23
measurements show a bandgap of 3 eV for amorphous GaN. The refractive index of GaN
was measured at 632.8 nm using ellipsometric techniques and it was found to be substrate
dependent. It was observed that the refractive index of GaN was 2.22 and 2.24 over SiO2
/ SiC and SiC respectively [13,22].
Table 1.2 gives some of the basic parameters of GaN at 300 K.

Table 1.2.

Basic parameters of GaN at 300 K [13,22].

Basic Parameters at 300 K:


Crystal Structure
Number of atoms per cm3
Debye Temperature (K)
Density (g.cm-3)
Dielectric constant
static
high frequency
Effective electron mass (in units of m0)
Effective hole mass (in units of m0)
heavy
light
split-off band
Electron affinity (eV)
Lattice constant (A)
Optical phonon energy (meV)
Energy gap (eV)
Effective conduction band density of states (cm-3)
Effective valance band density of states (cm-3)
Energy of spin-orbital splitting Eso (eV)
Energy of crystal-field splitting Ecr (eV)

Wurtzite
Zincblende
8.9 1022
600
6.15
8.9
5.35
0.20

9.7
5.3
0.13

1.4
0.3
0.3

1.3
0.2
0.3

4.1
a = 3.189
c = 5.186
91.2
3.39

87.3
3.2

2.3 1018

1.2 1018

4.6 1019
0.008
0.04

4.1 1019
0.02

~ 5 106

~ 5 106

1000

1000

4.52

Electrical Properties:
Breakdown field (V cm-1)
Mobility (cm2 V-1 S-1)
electrons

24
Table 1.2. continued.
holes
Diffusion coefficient (cm2 s-1)
electrons
holes
Electron thermal velocity (m s-1)
Hole thermal velocity (m s-1)

200

350

25
5
2.6 105
9.4 104

25
9
3.2 105
9.5 104

Optical Properties:
Refractive index
Radiative recombination coefficient (cm3 s-1)

2.3
10-8

Thermal and Mechanical Properties:


Bulk modulus (dyn cm-2)
Melting point (C)
Specific heat (J g-1 C-1)
Thermal conductivity (W cm-1 C-1)
Thermal diffusivity (cm2 s-1)
Thermal expansion, Linear (C-1)

20.4 1011
2500
0.49
1.3
0.43
a = 5.59 10-6
c = 3.17 10-6

1.4. Indium Nitride (InN):


InN is another important member of III-Nitride semiconductors family. Work has
been done to study InN but it has not received the experimental attention given to GaN
and AlN. This is probably due to difficulties in growing high-quality crystalline InN
samples and because of the existence of alternative, well-characterized semiconductors
such as AlGaAs, which have energy bandgaps close to that of InN. It is difficult to
obtain good quality of singlecrystal InN because it can not be grown at high
temperatures. It undergoes rapid dissociation at high temperatures, even as low as 600 C
and an extraordinary high nitrogen overpressure would be required to stabilize the

25
material up to the melting point, which is practically impossible

[13]

. The large disparity

of the atomic radii of In and N is another factor that enhances the difficulty in obtaining
InN of good quality. Some of the basic properties of InN are given below

1.4.1 Electrical Properties.


A large disparity of the atomic radii of In and N makes it difficult to obtain a good
quality InN films. Further, InN suffers from the lack of a suitable substrate material and
high native-defect concentrations that limits its quality and hence no reliable
experimental data for the electron mobility in InN have yet to be obtained. Different
people have attempted to obtain electron mobility for InN but the results vary widely. At
room temperature it can be as high as 3000 cm2 / Vs. It is also reported that the electron
mobility of InN varies with the substrate used for the deposition of InN films. A recent
study of the electron mobility of InN as a function of growth temperature indicates that
the mobility of Ultra-High-Electron Cyclotron Resonance-Radio-Frequency Magnetron
Sputtering (UHV-ECR-RMS) grown InN can be as much as four times the mobility of
conventionally grown InN.

1.4.2. Mechanical and Thermal Properties.


The density of InN at room temperature deduced from Archimedean-displacement
measurements is 6.89 g / cm3. This is comparable to 6.81 g / cm3 estimated from X-ray
data. The bulk modulus of InN determined from first principles calculations is B = 165
Gpa.

26
The thermal expansion coefficients of InN are also reported by different people. It
is measured at five different temperatures between 190 K and 560 K and the results
indicate that both along the parallel and perpendicular directions to the c-axis of InN
these coefficients increase with increasing temperature. Thermal conductivity of InN is
about 0.80 0.20 W.cm-1.K-1 while its heat capacity is (9.1 2.9) 10-3 cal/mol.K at
temperatures between 298 and 1273 K.

1.4.3. Optical Properties.


The optical properties measurements performed on InN are described by various
groups. The optical bandgap of InN at room temperature ranges from 1.7 2.07 eV with
a value of 1.89 eV reported by Tansley and Foley

[13]

who also measured the infrared

absorption of InN and observed an unidentified donor level approximately 50 60 meV


below the conduction band edge. However it has been reported recently that the bandgap
of InN is 0.7 eV [34-38].
Tyagai performed reflection and transmission measurements of InN and was able
to estimate an effective mass me* = 0.11 m0 and an index of refraction n = 3.05 0.05.
Data obtained about InN so far is listed in table 1.3.

Table 1.3.

Basic parameters of InN at 300 K [11,14].

Basic Parameters at 300 K:


Crystal structure
Number of atoms per cm3
Debye temperature (K)
Density (g cm-3)

Wurtzite
6.4 1022
660
6.81

27
Table 1.3. continued
Dielectric constant
static
high frequency
Effective electron mass (in units of m0)
Effective electron mass (in units of m0)
heavy
light
split-off band
Lattice constant (A)
Optical phonon energy (meV)
Energy gap (eV)
Energy of spin-orbital splitting Eso (eV)
Energy of crystal-field splitting Ecr (eV)
Effective conduction band density of states (cm-3)
Effective valance band density of states (cm-3)

15.3
8.4
0.11
1.63
0.27
0.65
a = 3.533
c = 5.693
73
1.89 2.05
0.7 (recent results)
0.003
0.017
9 1017
5.3 1019

Electrical Properties:
Mobility (cm2 V-1 S-1) electrons
Diffusion coefficient (cm2 s-1) electrons
Electron thermal velocity (m s-1)
Electron thermal velocity (m s-1)

3200
< 80
3.4 105
9 104

Optical Properties:
Refractive index
Radiative recombination coefficient (cm3 s-1)

2.9
2 10-10

Thermal and Mechanical Properties:


Bulk modulus (GPa)
Melting point (C)
Specific heat (J g-1 C-1)
Thermal conductivity (W cm-1 C-1)
Thermal diffusivity (cm2 s-1)
Thermal expansion, Linear (C-1)

140
950
0.32
0.45
0.2
a = 3.8 10-6
c = 2.9 10-6

28

1.5. Boron Nitride (BN):


Boron Nitride is another important member of the III-nitride semiconductors
family. It has at least four crystal structures known as wurtzite, zinc blende, hexagonal
and rhombohedral [14].
Wurtzite structure (also called -BN) was first synthesized in 1963. Typically, BN
crystals with wurtzite symmetry are very small (fraction of microns), are highly defective
and contain other phases.
Zinc blende modification of BN (also known as cubic or sphalerite or -BN) was
first synthesized in 1957 using the technique similar to that used for diamond growth.
Now crystals with a few millimeter sizes are commercially available.
Hexagonal BN (also known as -BN) with the structure similar to graphite has
been known for more than a century. Many properties of hexagonal BN are highly
anisotropic and depend on the growth method. In many cases, the different values of BN physical parameters given in this section reflect the differences in material properties
of hexagonal BN grown by different methods.

1.5.1. Electrical Properties.


The combination of high dielectric breakdown strength and volume resistivity
lead to h-BN being used as an electrical insulator. However its tendency to oxidize at
high temperatures often restrict its use to vacuum and inert atmosphere operation The
volume resistivity of hexagonal BN lies in a wide range from 1108 to 11013 .cm

[25]

The study of break down field for zinc blende and hexagonal BN shows that these two
structures have different break down field strengths. The value for this breakdown field is

29
2 - 6 106 V cm-1for zincblende structure and 1 - 3 106 V cm-1 for hexagonal BN
[11,21,22]

. Information about the electron and hole mobility in the zinc blende structure BN

has also obtained. Data shows that that electron mobility is 200 cm2 V-1 S-1. On the
other hand holes have a mobility of 500 cm2 V-1 S-1 which is pretty high as compared
to the electron [13,22,25].

1.5.2 Thermal and Mechanical Properties.


BN exist in different crystalline forms and hence its density varies with its
structure. Wurtzite, zinc blende and hexagonal BN have densities 3.48 gm/cm3 , 3.45
gm/cm3and 2.0 2.28 gm/cm3 respectively. The bulk modulus is 400 GPa for wurtzite
and zinc blende structures and however the bulk modulus of hexagonal BN is very
different from wurtzite and zinc blende structures. Green et al

[11]

reported a bulk

modulus of 36 GPa for hexagonal BN.


The experimental value of Thermal conductivity of BN reported is 7.4 Wcm-1K-1.
However its value is different for different crystalline structures in different directions.
Specific heat of BN is also reported by different researchers and it was found that its
value not only depends upon the crystal structure but also varies with temperature.
Hexagonal structure has a maximum value of specific heat (0.8 J.g-1.C-1) at 300 K
followed by 0.75 J.g-1.C-1 for wurtzite and 0.6 J.g-1.C-1 for zinc blende structures
respectively.

30

1.5.3. Optical Properties.


The optical properties of BN varies with its crystalline structure. A. El-Yadouni
and J. Boudiombo report the bandgap of BN to be 5.8 eV. However, for wurtzite
structure the band gap ranges from 4.5 5.5 eV. In zinc blende structure its value varies
between 6.1 eV and 6.4 eV while in hexagonal structure the BN band gap falls in 4.0
5.8 eV. The refractive index of wurtzite BN is 2.05, for zinc blende structure its 2.1 and
for hexagonal structured BN the refractive index is 1.8. Some of the basic parameters of
BN are given in table 1.4.

Table 1.4.

Basic parameters of BN at 300 K [13,22].

1.5.3 Basic Parameters at 300 K:


Crystal Structure
Debye Temperature (K)
Density (g cm-3)
Dielectric Constant
static parallel

perp.
high frequency

Wurtzite
1400
3.48
5.1
6.8
parallel = perp.
4.2 4.5

Effective electron mass


(in units of m0)
Longitudinal ml
0.35
Transversal mt
0.24
In the M direction
In the ML direction
Effective hole mass
(in units of m0)
In the MK
0.88
direction
In the A
direction
In the M

1.08
1.02

Zinc Blende
1700
3.45

7.1
4.46

Hexagonal
400
2.0 2.28
5.06
6.85

parallel = 2.2
perp. = 4.3

1.2
0.26
0.26
2.21
m1 3.16
m2 0.64
m3 0.44

31
Table 1.4. continued.
direction
In the X direction
In the L direction
In the K direction
In the M direction
In the ML direction
Lattice Constant (A)
a = 2.55
C = 4.17
Optical phonon energy
~ 130
(meV)
Energy gap (eV)
4.5 5.5
Effective conduction band - 1.5 1019
density of states (cm-3)
Effective valance band
2.6 1019
density of states (cm-3)
Wurtzite
Electrical Properties:
Breakdown field (V cm-1)
Mobility (cm2 V-1 S-1)
Electrons
Holes
Diffusion coefficient (cm2 s-1)
Electrons
Holes

0.55
m1 0.36
m2 1.20

3.615

0.47
0.50
1.33
a = 2.5 2.9
c = 6.66

~130
6.1 6.4
2.1 1019

4.0 5.8

2.6 1019
Zinc Blende

Hexagonal

2 - 6 106

1 - 3 106

200
500
5
12

Optical Properties:
Refractive index

2.05

2.1

1.8

400

36.5

~ 0.6

~ 0.8

Thermal and Mechanical Properties:


Bulk Modulus (GPa)
400
Melting point (C)
Specific heat (J g-1 C-1)
~ 0.75
Thermal conductivity
(W cm-1 C-1)
Experimental
Theoretical
Parallel to c-axis
Perpendicular to c-axis
Thermal expansion, Linear
(C-1)

7.4
~ 13
1.2 10-6

0.03
6

32
Table 1.4. continued.
Parallel to c-axis
Perpendicular toc-axis

2.7 10-6
2.3 10-6

38 10-6
-2.7 10-6

33

Chapter 2

RF MAGNETRON SPUTTERING
2.1. Sputtering:
Thin films are the subject of matter for many applications and have got significant
importance in physical sciences and engineering. A number of methods and techniques
are used to deposit thin films. Sputtering is one important technique used for thin film
deposition

[39-41]

. It is simply the removal of surface atoms of a target by the interaction

of incident particles and ions of a gas. Sputtering was first observed in a dc gas discharge
tube by Grove in 1852. He discovered the cathode surface of discharge tube was
sputtered by energetic ions in the gas discharge, and cathode materials were deposited on
the inner wall of the discharge tube. At that time sputtering was regarded as an undesired
phenomena since the cathode and grid in the gas discharge tube were destroyed. Today,
however, sputtering is widely used for industrial and academic applications [39,41,42,].
The deposition of thin films by sputtering involves the acceleration of positive
ions in many cases from gaseous (inert gas) plasma to bombard a target material that is
negatively biased. There are several techniques available to generate the necessary
positive gas (inert or reactive) ions and hence glow discharge plasma but commonly radio
frequency (RF) electromagnetic radiations are used. The plasma generation starts with the
acceleration of free electrons in an applied electric field. Due to the voltage between the
target and the plasma, generated by the biased target, ions are accelerated out of the
plasma into the target. The surface atoms of the target material are ejected by momentum
transfer if the energy of the bombarding ions is higher than the sputtering threshold
energy, which for most materials is in the order of 20 eV.

34
Besides the gas ions, other energetic particles like free radicals and excited state
molecules and secondary ions are also generated, if the kinetic energy of the electrons is
high enough for electron impact reactions with gas molecules, which diffuse out of the
plasma towards the containment walls and film surface. Statistically an ionization event
will occur in such an environment producing an ion and an electron. The electron will
then be accelerated towards the anode and on its way cause further impact ionization
events called secondary ionization, and also dissociation and excitation of the present
atoms and molecules. In order to maintain stable plasma or glow discharge, additional
electrons must be introduced. This is done by the ion surface interactions at the negative
electrode causing generation of secondary electrons. These electrons are then accelerated
by the electrode potential into plasma and cause a sufficient number of ionization events
to maintain the discharge.
The plasma can be divided in two areas; the bulk center plasma and the
boundary region, called sheath region. The plasma behavior in the sheath region is crucial
both for sustaining the plasma and to control the film deposition. Therefore, the plasma
plays an important role in the deposition process and has a large influence on the deposit
obtained. The positive gas ions are the bombarding particles and the cathode material of
which film is formed is the target. Sputtering has almost no limitations regarding target
materials and both pure metals and insulators can be grown. In the latter case an rf power
supply is needed in order to avoid charge build-up on the target surface [40,42].
Sputtering is usually characterized by the sputter yield S, which is defined as the
mean number of atoms removed from the surface of a solid per incident ion and is given
by [39-42]

35

S = atoms removed / incident ions


The sputter yield is influenced by the following factors

2.1.1 RF Power.
The ion bombardment energy and the discharge current strongly depend on rf
power. As a result, the sputter rate also strongly depends on the rf power. The sputtering
yield is proportional to the ion bombardment energy. Therefore, the sputter rate increases
with increasing rf power.

2.1.2. Gas Pressure.


The collisions of target atoms with the gas particles results in deflection of the
sputtered atoms, and sometimes cause the atoms to redeposit on the target, so that the
redeposition rate increases as the gas pressure increases. On the other hand, increasing
gas pressure leads an increasing discharge current for constant rf power due to increasing
plasma potential. As a combine result, the sputtering rate may have a maximum value
somewhere as the pressure is increased.

2.1.3. Target Material.


The nature of target material significantly affects the sputtering yield because
different materials have different binding. The greater the binding energy of a target
material the lower will be its sputtering rate and yield. The reason being, more energetic
ions are needed to remove atoms from the surface of such material. Therefore those ions
having energy smaller than the binding energy of the target material can not detach atoms

36
from the target and hence the yield will also be small. The opposite is true for materials
with lower binding energies.

2.1.4. Target Structure and Topography.


Since the sputtered atoms can come only from the surface layers of the target, the
sputtering yield should be proportional to the energy deposited in a thin layer near the
surface. For not-too-oblique incidence, the density of this energy near the surface
increases with increasing angle of incidence. For a new target, most ions impact the
surface with small angles of incidence. Therefore the sputter rate is relatively small for
new target. Since the most effective sputter area of the target is the area under magnetron
trap and hence for old target, deep erosion under the magnetron trap is formed due to the
sputter. It enhances the effective area under the trap, resulting in an increasing sputter
rate. The angles of incidence of ions become larger as the erosion becomes deeper. Both
effects contribute to enhance the sputter rate.

2.2. Reactive Sputtering:


When a reactive gas is intentionally included to the sputtering system then the
sputtering process is called reactive sputtering. Nitrogen is used as reactive gas in the
present work. Reactive gas in the sputter atmosphere enables one to alter or control the
properties of the deposit. Depending on the pressure of the reactive gas, a reaction may
occur during the deposition of the film either at the cathode or at the surface. In most
cases, the composition of the film may be altered by simply varying the quantity of
reactive gas or the proportion of reactive to inert gases in the discharge.

37

2.3. RF Magnetron Sputtering:


This is an enhanced sputter method which enables a higher deposition rate at low
operating pressure together with the possibility to obtain high quality films at low as well
as high substrate temperatures. This is achieved by applying a magnetic field
perpendicular to the electric field of the target electrons. In this process surface atoms of
the target material are removed and deposited on a substrate by bombarding the target
with the ionized gas atoms. A radio frequency power source is used. The magnet, located
behind the target, enhances ionization and effectively directs the sputtered atoms towards
the substrate. Schematic diagram of the entire process is shown in the figure 2.1.

38

Figure 2.1. Diagrammatical set-up of RF Magnetron sputtering deposition.

39
In 1935 Penning first studied low pressure sputtering in which a transverse magnetic field
was superposed on a dc glow discharge tube. He found that superimposition of the
magnetic field of 300 G lowered the sputtering gas pressure by a factor of ten and
increased the deposition rate of sputtered films [42].
The electrons are forced to follow a closed drift path caused by the crossed fields
i.e. they are trapped in a channel. The effect of this is two fold: the number of ionized gas
atoms will increase significantly allowing a lower gas pressure and an increase of the
mean free path, which is inversely proportional to the pressure. The outcome is a higher
field of particles on the substrate at a lower pressure, which has a positive influence on
the quality of the deposit.
The properties of the deposit can be seriously affected by the choice of the
particular condition used for sputtering. Besides the above-mentioned magnetic field
effects and the plasma there are numerous other parameters that control the deposit.
Examples are the quantity of reactive gas, the ratio of reactive gas to the inert gas, the
pressure in the deposition chamber, geometry of the chamber and the sputtering current
and voltage.

40

Chapter 3

EXPERIMENTAL SETUP AND TECHNIQUES


3.1. Film Growth and Treatment:
3.1.1 Growth Conditions and Sample Preparation.
The experimental setup in this study consists of different parts but the initial and
fundamental work is the film growth. All films were grown by reactive magnetron
sputtering in an ultra high vacuum (UHV) deposition chamber. The sputtering gas used
was nitrogen. Substrates used in the entire work were glass, Si (111), Si (100) and optical
fiber made of glass. The flat substrates were cleaned using methanol and air cleaner.
Optical fibers were first put in the dichloromethane to remove the protective cladding
layer and then pulled in a flame to change the size according to the requirements.
Different substrate holders were used to deposit films at different temperatures. A special
holder was also used having a long cylindrical hole about 25 cm long and 1 cm in
diameter to keep liquid nitrogen contact with the substrate or optical fiber to achieve low
temperature deposition. The growth temperature ranged from -77 K to 1000 K. Pumping
of the chamber was done by a turbo pump and cryogenic pump during film growth. RF
guns were used to deposit films. Crystal thickness monitors were also used inside the
chamber to monitor film thickness during growth. Total thickness of the films on flat
substrates ranged between 50 nm to 1200 m while on optical fibers it ranged from 300
nm to 10 m. Water pump was used to keep the temperature of the cryogenic pump and
that of the growth chamber down after film deposition. AlN, GaN, InN and BN hosts

41
were used in the entire films growth. Figure 3.1 is showing picture of the growth chamber
and related instruments used for the film deposition during the entire work.

3.1.2 Nature of Film Growth and Substrate Temperature.


All of the films investigated in this work were doped or alloyed films. These films
were grown by co-sputtering two targets simultaneously. Usually a small piece of rare
earth or transition element was either put inside the main host (Al) or put on the top of the
host (Boron) or placed in the liquid host (Gallium). The growth rate for these films was
controlled by the rf power. Multilayers of such films were also deposited by using two
targets and using them alternatively by flipping the substrate towards send target after
deposition from first one.
The temperature of substrate plays an important role in the quality of films. Good
crystalline films are usually obtained at higher temperatures, while low temperature often
produces an amorphous film. In the present work, nitride films were prepared at room
temperature, high temperature and liquid nitrogen temperature. Different sample holders
were used to deposit films at different temperatures. Liquid nitrogen was used to keep the
temperature of substrate at 77 K. However we do not know about the exact temperature
of the tip of optical fiber used as substrate in some films.

42

Figure 3.1. Picture of the RF Magnetron Sputtering system used for film deposition.

43

3.1.3 Thermal Annealing of the Films.


Thermal annealing of some of the films was performed in a nitrogen atmosphere
using a Thermolyne type tube furnace Model 21100. This furnace has a transfer arm on
one side and a hole at the end of other side arm for nitrogen entry and flow. The Optical
fibers were first put in a narrow ceramic tube and then annealed in the same furnace.
Annealing temperature ranged from 600 C to 900 C for optical fiber because the fibers
melted near 1000 C. However some films on Si substrates were also annealed upto 1050
C. Annealing time was 30-60 minutes. The rapid thermal annealing allows the samples
to be activated, quickly cooled and then removes from furnace for analysis. Figure 3.2
shows picture of the tube furnace 21100.

44

Figure 3.2. Picture of the tube furnace 21100 used for thermal activation of films

45

3.2 Characterization Techniques:


3.2.1 Scanning Electron Microscopy (SEM).
The Scanning Electron Microscope (SEM) is one of the most versatile and widely
used tools allowing the study of structure, morphology and composition of biological and
physical materials. By scanning an electron probe across a specimen, high-resolution
images of the morphology or topography of a specimen, with great depth of field, at very
low or very high magnifications can be obtained.
The morphology and the structure of films particularly those deposited on optical
fibers were examined using scanning electron microscopy. Cross section images of the
fiber surrounded by the deposited film were taken. Figure 3.3 gives picture of the SEM
used in the present work.

3.2.2 X-ray Diffraction (XRD).


X-rays are electromagnetic radiations having wavelength of the order of a few
angstroms, the same as typical interatomic distances in crystalline solids, and hence they
can be diffracted from the solids according to Braggs law. Due to this property X-rays
can be used to study the internal structure of solid materials.
The structure of all films was studied by X-ray diffraction (XRD). Theta-2theta
(-2) scans were taken using a Rigaku diffractometer equipped with a two axis sample
stage goniometer. K x-rays from copper crystal were used for illumination purpose. The
existence of Bragg diffraction peaks in the XRD spectrum was used to confirm the crystal

46
structure. The amorphous films showed peaks only due to the substrate material. The
scanning was done from 20 - 80 for one hour in each case.

3.2.3 Cathodoluminescence.
Cathodoluminescence (CL) is a method conventionally used to study the emission
of light from a specimen when bombarded with an electron beam. The spectral content of
the light emitted from the specimen contains information on the impurities in
semiconductors. Spectral CL imaging allows for imaging of lateral distribution of
impurities. Optically inactive centers may inhibit the CL emission and affect the intensity
of the detected CL signal. CL measured as a function of electron beam energy can
provide insight into the depth distribution of the relevant centers. In case of a
semiconductor specimen, the CL energy is equivalent to the energy gap between the
conduction band and the valence band. Cathodoluminescence offers a very powerful
capability of studying the evolution of the CL spectra as a function of the excitation
current density over a very wide range of currents. The following are some feature of CL.

It is possible to analyze the distribution of trace impurities and also the

distribution of defects. In the case of a semiconductor specimen, donor and acceptor


elements are doped in order to form the PN junction.
It is possible to perform measurements over a minute area as small as that in
characteristic X-ray analysis. When a CL device is installed in a TEM and the CL
distribution of a thin-section specimen is observed, as the area of scattered electrons in
the thin film is small, spatial resolution of the order of one nanometer can be obtained.

47

Figure 3.3. Scanning Electron Microscope used in the present work.

48

Considering various conceivable factors involved in the CL generation process,


it is extremely difficult to analyze the data and prepare the specimen. For example, when
performing CL observation of the cross-section of a light emitting diode (LED) specimen
(we will describe later), it has been found that CL did not appear along the streak left by
the polishing.
Conventionally, the Cathodoluminescence method has been used widely for
research of mineral specimens. Recently, it has especially come to receive attention for
developing optical devices such as the commercialization of blue LEDs and laser diodes.
It is also used in various fields such as large capacity memory devices, devices for
optical communication network and materials for various kinds of displays. Table 3.1
shows the material fields in which the CL method is mainly used, and also examples of
its use.

Table 3.1. Application scope and use of CL in various fields.

Material field

Example of application of CL method

Semiconductor
devices

Device characteristics
Wavelength of emitted light
Distribution of trace defects and impurities in the
specimen

Optical fiber

Change in refractive index


Distribution of trace defects

Fluorescent materials

Wavelength of emitted light


Identification of emitted area

49

Distribution of each light- emitting particle

Ceramic materials

Distribution of grain boundaries and defects in sintering

Minerals, rocks

Distribution of trace impurities


Structural non-uniformity
Stress distribution in minerals

Steel materials

Analysis of oxide inclusion

Biological specimens

Observation using luminescent dye

In the present work, CL studies of the films were performed at room temperature
in a vacuum chamber at a pressure of about 3x10-6 Torr, which was maintained with an
Alcatel CFF 450 turbo pump. Films were excited with electron beam energy of 2.85 kV
and beam current of 100 A. Luminescence from the films was focused onto the entrance
slit of a SPEX Industries double monochromator with gratings blazed at 500 nm and
detected at a Thorn EMI fast high gain photomultiplier tube with a range of 200 900
nm.

50

Chapter 4

HOLMIUM AND GADOLINIUM


Thin films of Holmium doped AlN and Gadolinium doped AlN are deposited on flat
Silicon substrates and optical fibers by RF Magnetron sputtering at liquid nitrogen and
room temperatures. The deposited films are characterized for their structure and
luminescence

using

Cathodoluminescence

(CL),

Thermal

Annealing,

Optical

Microscopy, X-Rays Diffraction (XRD), Scanning Electron Microscopy (SEM) and


Energy Dispersive X-rays Spectroscopy (EDX). XRD and SEM results show that all the
films deposited at liquid nitrogen temperature are amorphous, except those films on
optical fibers having thickness greater than 1 micron. Luminescence from Holmium
occurs in wide range from ultraviolet to infrared region with dominant green emission at
549 nm, with other peaks at 362 nm, 394 nm, 459 nm, 662 nm and 752 nm. Gadolinium
gives a very sharp ultraviolet peak at 314 nm. Thermal annealing enhances the
luminescence of both materials. The co-doping of Gadolinium with Holmium in AlN host
enhances the Holmium green emission with thermal annealing. The enhancement
saturates at a Gadolinium concentration of about four times that of the Holmium.

51

4.1

Growth.
Thin films of AlN:Ho AND AlN:Gd were prepared at room temperature and 77 K

by RF magnetron sputtering of holmium-aluminum and gadolinium-aluminum bi-metal


targets separately in a pure nitrogen atmosphere in a cryopumped vacuum system. The
RF power was varied between 100 and 200 watts. The bimetal targets were made by
drilling one hole (6.5 mm diameter) in two different aluminum targets (42 mm diameter)
and placing circular discs of 99.999% pure holmium and gadolinium in the holes.
Holmium and gadolinium were then co-sputtered with aluminum. The background
vacuum in the chamber was < 3 X 10-5 Torr. All the holmium films were deposited on
flat Si(100) or Si(111) substrates and optical fibers with diameters ranging from 10 m to
80 m. Gadolinium films were deposited only on flat Si(100) or Si(111) substrates.
Nitrogen pressure between 5 to 8 m-Torr was used in the entire work. Liquid nitrogen
was used for cooling the substrates during low temperature deposition of films.
Hot films were also grown keeping the substrate temperature at 700 C by connecting an
electric heater to the substrate. Films of different thickness were grown on both types of
substrates ranging from 50 nm up to 10 m on fibers and up to 1.2 m on flat silicon
substrates. These thicknesses were measured with a quartz crystal thickness monitor in
the growth chamber, a transmission IR microscope and Scanning Electron Microscope
(SEM). X-rays diffraction (XRD) and SEM were used to determine the structure of the
films. The fibers of approximately 50 mm length were clamped to the liquid nitrogen
reservoir along their length, with the last 10 mm extending beyond the cooling clamp. No
measurement of the fiber tip temperature during deposition was made. The structure of
films deposited on flat silicon substrate was also analyzed by XRD.

52

Liquid Nitrogen

Fiber Holder

Fiber
Holmium
Aluminum

Figure 4.1. Diagrammatical arrangement of film deposition on optical fiber


thermally in contact with the holder containing liquid nitrogen.

53

4.2. Characterization.
4.2.1. X-Ray Diffraction.
The structure of all films was studied by X-ray diffraction (XRD). Theta-2theta
(-2) scans were taken using a Rigaku diffractometer equipped with a two axis sample
stage goniometer. K x-rays with wavelength 0.154 nm from copper crystal were used
for illumination purpose. The existence of Bragg diffraction peaks in the XRD spectrum
was used to confirm the crystal structure. Figures 4.2a and 4.2b show XRD spectra of
amorphous and crystalline AlN:Ho films. It clear from these figures that those films
deposited at 700 C have additional peaks at 33.03 and 43.5 along with Si(111) and
(311) peaks. These peaks appear from AlN:Ho indicating that films deposited at 700 C
growth temperature are crystalline. However those films deposited at liquid nitrogen
temperature do not have those additional peaks, confirming that they are amorphous. The
scanning was done from 20 - 80 for one hour in each case.
Figure 4.3 shows XRD spectrum of AlN:Gd deposited at low temperature. The
spectrum shows only one peak at 69.1 indicative of the Si(100) substrate. No additional
peaks were observed and hence the deposited films were amorphous.

54

1000000
Si (111) peak
at 28.4
degrees
100000

Log Intensity (a.u)

10000

Si(311) peak
at 58.6
degrees

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

76

2 theta (deg)

Figure 4.2a. XRD spectrum of low temperature deposited AlN:Ho films.

80

55

1000000
Si (111) peak
at 28.4 degrees

100000

Si (311) peak
at 58.6 degrees

1000
AlN:Ho (101) peak
at 33.03 degrees
100

AlN:Ho (102) peak


at 43.5 degrees

AlN:Ho (103) peak


at 65.8 degrees

10

1
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
50
52
54
56
58
60
62
64
66
68
70
72
74
76
78
80

Log Intensity (a.u)

10000

Angel (theta-2theta)

Figure 4.2b. XRD spectrum of AlN:Ho films deposited at 700 C.

56

1000000

Si (100) peak
at 69.1

100000

Log Intensity (a.u)

10000

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

2theta (deg)

Figure 4.3.

XRD spectrum of low temperature deposited AlN:Gd.

76

80

57

4.2.2. Cathodoluminescence.
CL studies of films deposited on flat silicon substrate and the optical fiber were
performed at room temperature in a vacuum chamber at a pressure of about 3x10-6 Torr,
which was maintained with an Alcatel CFF 450 turbo pump. Films were excited with an
electron beam energy of 2.85 kV and beam current of 100 A . Luminescence from the
films was focused onto the entrance slit of a SPEX Industries double monochromator
with gratings blazed at 500 nm and detected at a Thorn EMI fast high gain
photomultiplier tube with a range of 300 900 nm. Figure 4.7 shows a picture of the
green light emitting optical fiber in the CL. The fiber is surrounded by AlN:Ho which
emits the green light. Figures 4.5 and 4.6 show room temperature CL emission from
AlN:Ho and AlN:Gd films deposited at liquid nitrogen temperature. AlN:Ho gives peaks
at 362nm, 394 nm, 461nm , 549 nm, 659 nm and 758 nm . These peaks correspond to 5G5
5I8, 5G4 5I8, 5F3 5I8, 5S2 5I8, 5F5 5I8, and 5I4 5I8 transitions [12]. A peak was
also observed at 704 nm indicating the presence of chromium oxide (Ruby) impurity in
the target. The 549 nm peak is dominant one while the Cr

3+

(ruby) peak at 704 nm is

even smaller than the 659 nm and 362 nm peaks. However this situation is different when
the film was thermally activated for 1 hour at 900 C.
CL spectrum of Gadolinium gives a sharp peak at 314 nm as a result of 6P7/2
8

S7/2 transition. An extra peak at 628 nm also appears in Gd spectrum. Since diffraction

grating is used in the CL detection system and hence this peak is in fact the second order
diffraction peak due to the sharp peak at 314 nm. An ultraviolet filter was used to confirm
our result and it was found that after the absorption of UV emission from Gd by the filter,
the peak at 628 nm also disappeared. It clearly indicated that 628 nm peak was just a

58
second order diffraction peak. Table 4.1 provides information about all emission lines
and transitions of Ho and Gd deposited in AlN host. Figure 4.4a and 4.4b are giving
schematic diagrams of energy levels in Ho3+ and Gd3+ ions

[43-45]

. Figure 4.7 gives a

direct picture of luminescence from holmium in CL. The optical fiber deposited with
AlN:Ho is directly exposed to electron beam in CL and the photograph is taken under
that exposed condition. The bright green emission is coming from the AlN:Ho film
around the fiber, indicating that green emission from holmium can be easily seen with
naked eye.
Table 4.1. Summary of emissions from Ho3+ and Gd+3 ions.

Material

AlN:Ho

AlN:Gd

Transition
Number

1
2
3
4
5
6
1

Transition

Wavelength
(nm)

Relative Intensity

G5 5I8
G4 5I8
5
F1 5 I 8
5
S2 5I8
5
F3 5I8
5
S2 5I7

362
394
461
549
659
758

0.355
0.530
0.300
1.000
0.277

314

P7/2 8S7/2

59

Figure 4.4a.

Energy levels diagram of Ho3+ in amorphous AlN host.

60

Figure 4.4b.

Energy levels diagram of Gd3+ in amorphous AlN host.

61

600
549 nm
Cr+3 at 702
nm

500

400

300

362 nm
461 nm
200

659
nm

100

Wavelength (nm)

Figure 4.5. CL emission from amorphous AlN:Ho in 300-750 nm range.

738

720

703

685

668

650

633

615

598

580

563

545

528

510

493

475

458

440

423

405

388

370

353

335

318

300

Intensity (a.u)

394 nm

62

6000
314 nm

5000
Second order
diffraction peak
at 628 nm

3000

2000

1000

0
300
317
333
350
367
384
400
417
434
450
467
484
501
517
534
551
567
584
601
618
634
651
668
684
701
718
735
751
768
785

Intensity (a.u)

4000

Wavelength (nm)

Figure 4.6. CL emission from amorphous AlN:Gd in 300-800 nm range.

63

64

4.2.3 Thermal Activation.


Thermal activation was accomplished by placing the flat films in the tube furnace
type 21100 at 900 C in a nitrogen atmosphere for 1 hour. The films on fibers were first
placed in a narrow ceramic tube and then in the furnace and thermally annealed at the
same temperature for 1 hour. Activation at 1000 C melted the optical fibers and hence
900 C was considered to be a suitable activation temperature for optical fibers. All
samples were placed in the furnace through the transfer arm of the furnace. Nitrogen gas
was allowed to pass through the hole at the other side arm of the furnace.
Figure 4.8 and 4.9 represent the luminescence from thermally activated AlN:Ho
films and comparison of CL spectra from thermally activated and un-activated AlN:Ho
films. It was observed that thermal excitation enhances the intensity of the emission.
Thermal activation has affected the Cr 3+ peak more than other peaks and it is now more
intense than all of the holmium peaks. Thermal activation has also enhanced the intensity
of 549 nm peak more than the rest of the holmium peaks. This fact is clear from the
increased ratio of the 549 nm peak and rest of holmium peaks after thermal activation.
We believe that thermal activation reduces defects in the structure and thermal stress,
causing an increase in the luminescence. CL emission from both flat silicon substrate and
optical fibers were performed, giving the same peak.

65

3500

Ruby
3000

Intensity (a.u)

2500

549 nm

2000

1500

1000

362
nm

461
nm

394
nm

659 nm

500

Wavelength (nm)

Figure 4.8. Cathodoluminescence spectrum of thermally activated AlN:Ho film.

733

713

692

672

651

631

610

590

569

549

528

508

487

467

446

426

405

385

364

344

323

303

282

262

241

221

200

66

3500
Ruby

AlN:Ho
AlN:Ho thermally annealed

3000

549 nm

2000

1500

1000

362 nm

461 nm

659
nm

394 nm

500

729

706

683

660

637

614

591

568

545

522

499

476

453

430

407

384

361

338

315

292

269

246

223

0
200

Intensity (a.u)

2500

Wavelength (nm)

Figure 4.9. Comparison of thermally activated AlN:Ho film with inactivated film.

67

4.2.4. XRD analysis of AlN:Ho films after Thermal Activation.


In order to investigate whether the AlN:Ho films are still amorphous after thermal
activation at 900 C for 1 hour or not, XRD was performed after thermal activation in the
same way as did before activation.
Figure 4.10 shows the results of x-rays diffraction analysis of the films. The
figure gives two big peaks at 28.4 and 58.6. These are Si(111) and Si(311) peaks
already present in the XRD spectrum of amorphous AlN:Ho. However a small additional
peak is also observable at 43.9. This is AlN:Ho(102) peak, indicating that some
crystalline regions have developed in AlN:Ho films when thermally activated.

4.3. Co-Sputtered Ho and Gd Films in AlN host:


4.3.1. Deposition.
Thin films of Holmium and Gadolinium co-doped in AlN were prepared at room
temperature and 77 K by RF magnetron sputtering of holmium and gadolinium together
in an aluminum target in a pure nitrogen atmosphere in a cryopumped vacuum system.
The same conditions were applied as mentioned in section 4.1, for deposition. The
gadolinium concentration was varied from 0 up to 6 times with a fixed holmium
concentration to see the effect of gadolinium on the holmium luminescence. The target
was made by drilling 6 holes (6.5 mm diameter) in aluminum target (42 mm diameter)
and placed circular discs of 99.999% pure holmium in one hole and gadolinium in the
remaining six holes with increasing number of pieces each time. All the films were
deposited on flat Si(100) substrates. Films thickness was ranging from 60 nm to 200 nm.
Liquid nitrogen was used for cooling the substrates during low temperature deposition of

68
films. Stripes of AlN:Ho and AlN:1Ho4Gd were also made using a mask to see directly
the effect of gadolinium on the luminescence enhancement in holmium. Thicknesses
were measured with a quartz crystal thickness monitor in the growth chamber. X-ray
diffraction was used to determine the structure of the films. Cathodoluminescence was
used to study the effect of Gd on Ho luminescence intensity. Direct pictures of the stripes
were obtained using a digital camera when the stripes were exposed to electron beam
inside the CL apparatus. Energy Dispersive X-rays Spectroscopy (EDX) was also
performed to confirm the increasing amount of Gd in the brighter films.

4.3.2. Energy Dispersive X-rays Spectroscopy (EDX).


EDX was performed using SEM and EDX analyzer. AlN:1Ho, AlN:1Ho1Gd and
AlN:1Ho4Gd films were analyzed separately. These films were focused in the SEM and
then EDX spectra were obtained. Figures 4.11, 4.12a, 4.12b, 4.12c and 4.12d show the
EDX analysis of

AlN:1Ho, AlN:1Ho1Gd, AlN:1Ho2Gd, AlN:1Ho3Gd and

AlN:1Ho4Gd films. The Ho peak is observable in figure 4.11 at 6722 KeV energy. No
Gd peak is obtained in this figure, confirming that this is the AlN film doped with Ho
without any Gd addition. In figure 4.12a an additional peak at 6046 KeV appears along
with the Ho peak. This additional peak shows the existence of Gd in the AlN:1Ho1Gd
films. This peak increases in intensity and the area under the curve as the number of
gadolinium pieces is increased in the film, confirming that our films get higher
gadolinium concentration when numbers of Gd pieces are increased in the target. Thus
we see an increasing order of the peak intensity at 6046 KeV for AlN:1Ho1Gd,
AlN:1Ho2Gd, AlN:1Ho3Gd and AlN:1Ho4Gd.

69

1000000
Si (111) peak
at 28.4
degrees
100000

Si (311) peak
at 58.6 degrees

1000

AlN:Ho (102) peak


at 43.5 degrees

100

10

Angel (theta-2theta)

Figure 4.10.

XRD analysis of AlN:Ho films after thermal activation.

80

76

72

68

64

60

56

52

48

44

40

36

32

28

24

1
20

Log Intensity (a.u)

10000

70

4.3.3 Cathodoluminescence and Luminescence Enhancement.


Cathodoluminescence spectra of the films were obtained the way described in
section 4.2.2 and strong enhancement in the intensity of Ho peak at 549 nm was observed
with the addition of Gd to AlN:Ho films.
Figure 4.13 shows the effect of increasing Gd concentration on the Ho
luminescence. Curves for AlN:Ho with zero to six times Gd are shown here. This figure
indicates that the increasing concentration of Gd strongly enhances the luminescence
intensity of Ho peak due to 5S2 5I8 transition at 549 nm. Table 4.2 shows the increase
in luminescence of Ho peak at 549 nm due to different dopant concentration of Gd. It is
clear from the table that luminescence enhancement is more than 100%.

Figure 4.11.
Energy (KeV)

EDX spectrum of AlN:Ho


20345

19755

19165

18575

17984

17394

16804

16214

15624

15034

14444

13853

13263

12673

12083

11493

10903

10313

9722

9132

8542

800

7952

1200

7362

6772

6182

5591

5001

Intensity (a.u)

71

1400

Ho peak at
6732 KeV

1000

Ho peak at 7497 KeV

600

400

200

9522

Energy (KeV)

Figure 4.12a. EDX spectrum of AlN:1Ho1Gd


20260

19695

19130

18565

17999

17434

16869

16304

15739

15174

14609

14043

13478

12913

12348

11783

11218

10653

10087

300

8957

KeV

8392

150

7827

400

7262

6697

6131

5566

5001

Intensity (a.u)

72

450

Ho Peak at 6732

KeV

350

Gd peak at 6046 KeV

250

200

Ho peak at 7497
KeV

100

50

73

2000
Gd peak at 6046
KeV

1800

Ho peak at 6732
KeV

1600

1400

Ho peak at 7497
KeV
1000

800

600

400

200

Energy (KeV)

Figure 4.12b. EDX spectrum of AlN:1Ho2Gd

19955

19380

18805

18229

17654

17079

16504

15929

15354

14779

14203

13628

13053

12478

11903

11328

10753

10177

9602

9027

8452

7877

7302

6727

6152

5576

0
5001

Intensity (a.u)

1200

Energy (KeV)

Figure 4.12c. EDX spectrum of AlN:1Ho3Gd.


20215

19630

19045

18460

17874

17289

16704

16119

15534

14949

14364

13778

13193

12608

12023

11438

10853

10268

9682

9097

8512

7927

7342

6757

6172

5586

5001

Intensity (a.u)

74

1400

1200
Gd peak at 6046 KeV

1000
Ho peak at 6732
KeV

800

600
Ho peak at 7497
KeV

400

200

Energy (KeV)

Figure 4.12d. EDX spectrum of AlN:1Ho4Gd


20395

19825

19255

18685

18114

17544

16974

16404

15834

15264

14694

14123

13553

12983

12413

11843

11273

10703

10132

9562

8992

8422

7852

7282

6712

6141

5571

5001

Intensity (a.u)

75

1600

1400
Gd peak at 6046

1200

1000

800
Ho peak at 6732 KeV

600

400
Ho peak at 7497 KeV

200

76

8000

549 nm Ho
7000

AlN:Ho

314 nm Gd
peak

AlN:1Ho1Gd
AlN:1Ho2Gd

6000

AlN:1Ho3Gd
AlN:1Ho4Gd
AlN:1Ho5Gd

AlN:1Ho6Gd
4000

3000

661 nm
2000

758 nm
1000

Wavelength (nm)

Figure 4.13. Effect of Gd concentration on Ho luminescence.

790

768

746

723

701

679

657

634

612

590

567

545

523

501

478

456

434

411

389

367

345

322

300

Intensity (a.u)

5000

77
Table 4.2. Effect of Gd concentration on the luminescence of Ho 5S2 5I8
transition at 549 nm.

Ho relative
concentration.

Gd relative
Concentration.

Relative Intensity
of 549 nm peak.

% increase
in intensity.

1
1
1
1
1

0
1
2
3
4

1
1.275
1.564
1.747
2.074

0 (reference)
27.5
56.4
74.7
107.4

1
1

5
6

2.047
2.054

104.7
105.4

From the figure and table it is clear that the Ho intensity enhances with increasing
concentration of Gd until it reaches a Gd concentration where the enhancement saturates.
Beyond this saturation concentration the addition of Gd doesnt enhance luminescence
further. We believe that during the growth additional Gd concentration makes more Gd
ions to cluster around Ho until the vicinity of Ho is saturated. Beyond that saturation,
addition of Gd doesnt make nay difference to Ho. On the other hand we could not get
any sequential increase or decrease in Gd luminescence due to changing concentration of
Gd.
Figure 4.14 shows the increasing pattern of Ho 5S2 5I8 peak intensity. It is clear
from the figure that the luminescence of 5S2 5I8 peak is enhanced and the intensity
initially increases linearly up to 4 times the initial concentration of Gd. At this
concentration of Gd, the AlN:Ho luminescence saturates and hence beyond this
concentration of Gd the intensity is constant

78

8000

Maximum Intensity of AlN:Ho (a.u)

7000

6000

5000

4000

3000

2000

1000

0
0

Relative Concentration of Gd (a.u)

Figure 4.14. Increasing pattern of Ho luminescence with Gd concentration.

79
4.3.4. Picture of Enhanced Luminescence on Stripes.
To observe the enhancement of green luminescence from Ho due to Gd, a mask
was introduced and new films were deposited.
Figure 4.15 shows the picture of the stripes, of AlN:1Ho6Gd/AlN:Ho deposited
on Si(111) substrate using a mask, under electron illumination at 2.85 KeV, with a
nominal current density of 15 A /mm2 . First 100 nm thick films of AlN co-doped with
1 piece of Ho (circular disk shaped with 6 mm diameter) and 6 pieces of Gd (each of the
size of Ho piece) in an Al target of 42 mm diameter were deposited on Si (111)
substrates. Then a mask having alternative stripes of metal and empty space was placed
on the top of one of these films and 80 nm thick films of AlN:Ho with no Gd were
deposited on the top of the mask. The spacing between two consecutive stripes on the
mask was 1mm. In this way the areas of the already deposited AlN:1Ho6Gd films hidden
by the metal of the mask could not get AlN:Ho films. However in those regions where
the mask was having empty spaces the AlN:Ho films got deposited on the top of
AlN:1Ho6Gd films. Thus we got stripes of alternative AlN:1Ho6Gd and AlN:Ho films.
These stripes were then thermally activated for 40 minutes at 900C. After thermal
activation CL was performed in the same way and under the same conditions as done for
thermally activated AlN:Ho films. Picture was obtained using a digital camera during the
operation of CL. The wide stripes in the figure show AlN:1Ho6Gd and the narrow stripes
represent AlN:Ho films. Luminescence enhancement is very clear from the picture. The
wide stripes of AlN:1Ho6Gd are much brighter than the narrow stripes of AlN:Ho.
Obviously there is a huge enhancement of green luminescence from Ho due to the
addition of Gd. More than 100% enhancement in table 4.2 verifies the picture.

80

Figure 4.15. CL picture of the enhanced luminescence in Ho by Gd.

81

Chapter 5

PRASEODYMIUM
Thin films of Praseodymium (Pr) doped AlN , GaN and BN are deposited on flat Silicon
substrates by RF Magnetron sputtering at liquid nitrogen and room temperatures. The
deposited films are characterized for their structure and luminescence using
Cathodoluminescence (CL), Thermal Activation and X-Rays Diffraction (XRD). XRD
results show that all the films deposited at liquid nitrogen temperature are amorphous.
Luminescence from praseodymium occurs in vide range from ultraviolet to infrared
region in all hosts with intense emission near 490 nm, 526 nm, 621 nm and 651 nm as a
result from 3P0 3H4, 3P1 3H5, 3P0 3H6 and 3P0 3F2 transitions respectively. The
films are thermally activated for half an hour at a temperature of 1025 C. Thermal
activation enhances the luminescence for the films. The co-doping of Gadolinium (Gd)
with Pr in AlN host also enhances the Pr emission at 490nm and 650 nm with thermal
activation.

82

5.1

Growth.
Thin films of Pr doped AlN, GaN and BN were prepared between 77 K and 950 K

temperature by RF magnetron sputtering of solid aluminum and boron targets of 99.999%


purity and liquid gallium target in a cryopumped vacuum system and a pure nitrogen
atmosphere. The RF power was varied between 100 and 200 watts. Doping of AlN and
BN thin films with Pr was accomplished by drilling a small hole (6.5 mm diameter) in the
aluminum and boron targets and placing slugs of Pr in the holes. Pr was then cosputtered
with the aluminum and boron. GaN thin film was doped with Pr by putting solid Pr (
<6.5 mm diameter) in liquid Ga and then cosputtered in the same way. The background
vacuum in the chamber was < 3 X 10-5 Torr. All films were deposited on flat Si(111) or
Si(100) substrates. AlN:Pr and GaN:Pr films were deposited at 77 K, room temperature
and 950 K respectively while BN:Pr films were deposited at 750 K. Nitrogen pressure
between 5 to 8 m-Torr was used in the entire work. The metallic substrate holder was
designed such that it was having a half-inch diameter cylindrical hole from the top. The
substrate was pasted on the metal base of the holder below the liquid nitrogen. Liquid
nitrogen was constantly poured in the holder to provide a constant low temperature to the
substrate and the deposited AlN:Pr films. Different holder, connected to power supply,
was used to deposit hot films at 950 K. the growth condition for Pr doped in the
mentioned three hosts are given in table 5.1.
Films of different thickness were grown with thickness between 50 nm and 500 nm.
These thicknesses were measured with a quartz crystal thickness monitor in the growth.
X-rays diffraction (XRD) was used to determine the structure of the films.

83

Table 5.1.

Growth conditions for AlN:Pr, GaN:Pr and BN:Pr films.


Type of Substrates

Si (111)

Substrate temperature for AlN:Pr

77 K - 950 K

Substrate temperature for GaN:Pr

77 K

Substrate temperature for BN:Pr

750 K

Base Pressure

4-6 x 10-7 Torr

Sputtering Pressure

5-8 x 10-3 Torr

RF Power

100 140 W

Average Back Power

2 10 W

Nominal Thickness of AlN:Pr films

60 1200 nm

Nominal Thickness of GaN:Pr films

60 500 nm

Nominal Thickness of BN:Pr films

100 nm

84

5.2. Characterization.
5.2.1. X-Rays Diffraction.
The structure of all films was studied by X-ray diffraction (XRD). Theta-2theta
(-2) scans were taken using a Rigaku diffractometer equipped with a double axis
sample stage goniometer. K x-rays with wavelength 1.54 A from copper crystal were
used for illumination purpose. The existence of Bragg diffraction peaks in the XRD
spectrum was used to confirm the crystal structure.
Figure 5.1 shows that films deposited at liquid nitrogen temperature are
amorphous and showing no peaks from Praseodymium. Only two peaks were observed at
28.4 and 58.1. These are Silicon peaks arose from Si(111) and Si(311). These results
indicate that the films are amorphous. However those films deposited nearly 700 C
(approximately 950 K) were having additional peaks at 34.8 and 49.1.5 along with
Si(111) and (311) peaks. These peaks appear from AlN:Pr indicating that films deposited
at 700 C are crystalline. The scanning was done from 20 - 80 for one hour in each case.
Figure 5.2 explains these results from crystalline films.
GaN:Pr and BN:Pr films were also characterized for their structures in the same
way and found to be amorphous as clear from figure 5.3 and figure 5.4.

85

10000

Si (111) at 28.4

Log Intensity (a.u)

1000

Si(311) at 58.6
100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

76

80

Angel (theta-2theta)

Figure 5.1. XRD spectrum of AlN:Pr films deposited at liquid nitrogen temperature.

86

1000000
Si peak at
28.4 degree

100000

10000

Log Intensity (a.u)

Si peak at 58.6
degree

AlN:Pr peak
at 34.8 degree

1000

AlN peak at 65.8


de g ree

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

76

Angle (theta - 2 theta)

Figure 5.2. XRD spectrum of AlN:Pr films deposited at 700 C .

80

87

5.2.2. Cathodoluminescence of Amorphous Praseodymium.


CL studies of Pr doped AlN, GaN and BN films deposited on flat silicon
substrates were performed at room temperature in a vacuum chamber at a pressure of
about 3x10-6 Torr, which was maintained with an Alcatel CFF 450 turbo pump. Films
were excited with an electron beam energy of 2.85 kV and beam current of 100 A .
Luminescence from the films was focused onto the entrance slit of a SPEX Industries
double monochromator with gratings blazed at 500 nm and detected at a Thorn EMI fast
high gain photomultiplier tube with a range of 300 900 nm.
A number of transitions were obtained from Pr in all hosts. Table 5.2

presents a

summary of Pr3+ ion emission from Pr doped AlN, GaN and BN. Each transition is
represented by its spectroscopic notation and the corresponding wavelength and intensity
of the peak is given for all materials. The intensity of each transition shown in the table
represents the maximum value of the intensity of each peak.

Table 5.2. Summary of Pr3+ ion emissions from Pr doped AlN, GaN and BN.
Material

Transition
Number

AlN:Pr

1
2
3
4
5
6
7
8
9
10
11

Transition
Wavelength
Assignment
(nm)
1

S0 1D2
S0 1I6
3
P2 3H4
3
P0 3H4
3
P0 3H4
3
P1 3H5
3
P0 3H5
3
P0 3H6
3
P0 3F2
3
P0 3F3
3
P0 3F4
1

335
385
439
488
504
526
573
618
652
710
738

Intensity
(a.u)

162
438
362
1002
1024
4577
767
643
1227
410
520

88
Table 5.2. continued.
GaN:Pr
3
4
6
8
9
10
11

P2 3H4
3
P0 3H4
3
P1 3H5
3
P0 3H6
3
P0 3F2
3
P0 3 F3
3
P0 3F4

418
493
532
621
650
713
736

559
5687
1895
4015
4227
1508
1627

415
492
544
628
651
713
736

372
973
752
842
748
257
289

BN:Pr
3
4
6
8
9
10
11

BN transition
P0 3H4
3
P1 3H5
3
P0 3H6
3
P0 3F2
3
P0 3F3
3
P0 3F4
3

Figure 5.3 shows a schematic diagram of the energy levels of Pr responsible for
the observed transitions

[45,46,47]

. Each energy level is represented by its spectroscopic

notation and each transition is represented by a line between the involved energy levels.
The wavelength of the emitted light is also mentioned with each line. These wavelengths
corresponds to the luminescence from Pr in amorphous AlN host.
Figure 5.4 gives cathodoluminescence spectra obtained from praseodymium
doped in the three hosts mentioned. Spectra of GaN:Pr, AlN:Pr and BN:Pr are shown
together. All spectra are taken under the same conditions. A constant offset of 500 is
added to the AlN:Pr data while 1000 is added to the GaN:Pr data to make them clearly
distinguishable. It is clear from the figure that the transitions 3P0 3H4 at 488 nm, 3P0
3H6 at 621 nm and 3P0 3F2 at 650 nm are very intense in GaN:Pr as compared to the
same peaks in AlN:Pr and BN:Pr and the transition 3P1 3H5 in AlN:Pr
corresponding to a sharp green emission in AlN:Pr compared to the other two hosts . The

89

Figure 5.3. Energy levels diagram of Praseodymium in amorphous AlN host.

90
reason for this difference is the effect of host materials on the electronic transitions
[30,31,32]

.
Figure 5.4 also confirms the presence of two additional peaks in AlN:Pr films.

These ultraviolet peaks are observed at 335 nm and 385 nm as a result of 1S0 1D2 , and
1

S0 1I6 transitions. The GaN:Pr and BN:Pr films do not show these peaks. A possible

explanation for this result is the band gap of GaN in the near ultraviolet region (3.39
eV)[13,22] and that of BN in the ultraviolet (4.0 - 5.2 eV)[26,27]. GaN absorbs light in the
region where these peaks would be expected. In BN:Pr, broad emission from the host BN
centered at 415 nm obscures the two UV peaks [28]. The peaks at 335 nm and 385 nm lie
in this broad band and hence we cannot say about these peaks. On the other hand the
large band gap of AlN (6.2 eV) [13,22] will absorb only those photons having wavelengths
smaller than 200 nm and thats the reason we observe the ultraviolet peaks in the AlN:Pr
films. Thus our results match with the band gap explanation of these materials.
From the last result we can see the advantage of wide band gap of AlN over GaN
and BN. Thus we can say that AlN is the favorable host for ultraviolet emitting phosphors
or other elements and a better material for use in ultraviolet devices.

5.2.3. Thermal Activation.


Like other lanthanides Pr has also potential for luminescence enhancement with
thermal activation. Thermal annealing was accomplished by placing the films in the tube
furnace 21100 at 1025 C in a nitrogen atmosphere for half an hour. All the samples were
placed in the furnace through the transfer arm of the furnace. Nitrogen gas was allowed
to pass through the hole at the other side arm of the furnace.

91

8000
7000

AlN:Pr
GaN:Pr

6000
9

BN:Pr

Intensity (a.u)

5000

4000
6

3000

10

2000
1000

11

5
8

10

11

GaN:P
r
AlN:Pr

11

BN:Pr

Wavelength (nm)

Figure 5.4.

CL spectra of praseodymium doped amorphous AlN, GaN and BN.

838

815

793

770

748

725

703

680

658

635

613

590

568

545

523

500

478

455

433

410

388

365

343

320

92
Figure 5.5 compares the luminescence from thermally annealed AlN:Pr films
with an un-annealed film. It is observed that thermal annealing enhances the intensity of
the emission. Defects and impurities causes non-radiative emission and we believe that
thermal annealing reduces defects [48-52] in the structure, causes an increase in the
luminescence.

5.2.4. X-rays analysis of amorphous AlN:Pr after thermal activation.


To investigate the effect of thermal activation on the structure of amorphous
AlN:Pr films, XRD was performed after thermal activation for 40 minutes at 1025 C in
the same way as did before activation.
Figure 5.6 shows the results of x-rays diffraction analysis of the films. The figure
gives two big peaks at 28.4 and 58.6. These are Si(111) and Si(311) peaks already
present in the XRD spectrum of amorphous AlN:Pr. No additional peaks were found.
Thus AlN:Pr films are still amorphous even after thermal activation for 40 minutes at
1025 C.

5.2.5. Cathodoluminescence of Crystalline Praseodymium.


Crystalline AlN:Pr was also studied for its luminescence properties in
Cathodoluminescence apparatus. These films were actually deposited at 700 C and XRD
results indicated that films deposited at this temperature are crystalline. CL was
performed in the same way as performed for amorphous films.
Figure 5.7 gives results from CL studies of crystalline Pr doped AlN films. The figure
gives a dominant red peak at 648 nm due to 3P0 3F2 transition. Other peaks were also

93
observed at 432 nm, 500 nm, 526 nm, 618 nm and 732 nm as a result of 3P2 3H4 , 3P0
3H4 , 3P1 3H5 , 3P0 3H6 , 3P0 3F2 and 3P0 3F4 transitions.
If we make comparison of CL emission from amorphous AlN:Pr as shown in
figure 5.4 or figure 5.5 and CL emission from crystalline AlN:Pr shown in figure 5.6, we
find some differences between the two. First of all the UV peaks at 335 nm and 385 nm
are not clearly visible in the crystalline spectrum. This means amorphous films are
relatively better than crystalline to see UV peaks in Pr. Second, in amorphous AlN:Pr we
obtained two close blue peaks at 488 nm and 504 nm while in crystalline AlN:Pr only one
peak is observable in this region at 500 nm. Thus amorphous films are better to study
more resolved blue peaks in praseodymium as compared to crystalline one. Third blue
peak at 500 nm is more intense as compared to green peak at 527 nm. However this case
was opposite in amorphous films and hence we can say that crystalline AlN:Pr films are
good emitters in blue rather than green as compared to amorphous films, but in the mean
while we can not ignore the nice sharp green emission from amorphous films. Fourth and
the most prominent result is that the intensity of red emission at 648 nm is considerably
higher than the intensity of the same peak in amorphous films. This red peak in
crystalline materials is dominant on all other peaks and hence crystalline films get
dominancy if one wants red luminescence from praseodymium. Fifth, the 711 nm peak
does not appear in crystalline films. Thus we can conclude that amorphous and crystalline
films and materials are equally important and we cannot decline one kind of materials as
compare to other.

94

12000
Thermally inactivated film

Thermally activated film

10000

Intensity (a.u)

8000

6000

4000

6
7
4
2

2000

9
10

11
0
250 290 330 370 410 450 490 530

570 610 650 690 730 770 810 850

Wavelength (nm)

Figure 5.5. Comparison of CL spectra from thermally activated and inactivated


amorphous AlN:Pr.

95

100000
Si (111) peak
at 28.4
degrees

1000
Si (311) peak
at 58.6
degrees
100

10

1
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
50
52
54
56
58
60
62
64
66
68
70
72
74
76
78
80

LogIntensity (a.u)

10000

2Theta (degrees)

Figure 5.6.

XRD spectrum of amorphous AlN:Pr after thermal activation.

96

4500

648 nm
4000

3500

Intensity (a.u)

3000
500 nm
2500
618 nm
526 nm
2000
732
nm

1500

1000
432 nm
500

0
304.25 372.75 412.75 452.75 492.75 532.75 572.75 612.75 652.75 692.75 732.75 772.75 812.75

Wavelength (nm)
Figure 5.7. Room temperature Cathodoluminescence of crystalline AlN:Pr.

97

5.3. Co-Sputtered Pr and Gd Films in AlN host.


5.3.1 Sputtering Deposition.
Thin films of Praseodymium and Gadolinium co-doped in AlN were prepared at
77 K by RF magnetron sputtering of praseodymium and gadolinium together in
aluminum target in a pure nitrogen atmosphere in a cryopumped vacuum system. Same
conditions were applied as mentioned in section 5.1, for deposition. The gadolinium
concentration was varied from 0 up to 4 times with a fixed praseodymium concentration
to see the effect of gadolinium on the praseodymium luminescence. The target was made
by drilling 6 holes (6.5 mm diameter) in aluminum target (42 mm diameter) and placed
circular discs of 99.999% pure praseodymium in one hole and gadolinium in four of the
remaining holes with increasing number of pieces each time. All the films were deposited
on flat Si(100). Films thickness was ranging from 60 nm to 200 nm. Liquid nitrogen was
used for cooling the substrates during low temperature deposition of films. Stripes of
AlN:Pr and AlN:1Pr4Gd were also made using a mask to see directly the effect of
gadolinium on the luminescence enhancement in praseodymium. Thicknesses were
measured with a quartz crystal thickness monitor in the growth chamber. X-rays
diffraction was used to determine the structure of the films. Cathodoluminescence was
used to study the effect of Gd on Pr luminescence intensity after thermally annealing the
films.

5.3.2. Cathodoluminescence and Luminescence Enhancement in Pr.


Cathodoluminescence spectra of the films were obtained the way described in

98
section 5.2.1 and strong enhancement in the intensity of Pr peaks at 490 nm, 618 nm and
652 nm was observed with the addition of Gd to AlN:Pr films.
Figure 5.8 shows the effect of increasing Gd concentration on the Pr
luminescence. Curves for AlN:Pr with zero, one and four times Gd are shown here. The
figure indicates that the increasing concentration of Gd strongly enhances the
luminescence intensity of Pr peaks due to 3P0 3H4 transition at 490 nm, 3P0 3H6
transition at 618 nm and 3P0 3F2 transition at 652 nm. For all these three transitions the
enhancement in luminescence is more than 200 %. The figure also shows that
luminescence due to addition of 4Gd is much more than enhancement due to the addition
of 1Gd. Table 5.3 shows the increase in luminescence of Pr peaks due to different dopant
concentration of Gd.

Table 5.3. Effect of Gd concentration on the luminescence of Pr transitions.


Pr Transition

Wavelength
(nm)

Gd Relative
Concentration

Relative
Intensity

% Increase in
Intensity

P0 3H4

490

0
1
4

1
1.58
4.37

0 (reference)
58
337

P0 3H6

618

0
1
4

1
1.73
4.14

0
73
314

P0 3F2

649

0
1
4

1
1.66
3.4

0
66
240

99

7000

490 nm

AlN:Pr

AlN:1Pr1Gd

6000

AlN:1Pr4Gd
314 nm
Gd pe ak
5000

4000

618
nm

3000

542
nm
2000

738
nm
439
nm
1000

784

759

733

708

682

657

631

606

580

555

529

504

478

453

427

402

376

351

325

0
300

Intensity (a.u)

652 nm

Wavelength (nm)

Figure 5.8. Effect of Gd concentration on the luminescence of AlN:Pr films.

100

5.3.3 Cartoon of Layers, Thicknesses and Stripes.


The enhancement of luminescence in Pr due to Gd doping is also photographed
just like the one for Ho. A mask was used during the deposition of the films in stripes and
new films were deposited.
Figure 5.9 shows the cartoon of the AlN:Ho and AlN:1Pr4Gd films deposited on
Si(111) substrate using a mask. AlN film co-doped with 1 piece of Pr and 4 pieces of Gd
was deposited on Si (111) substrates. The thickness of this film was 75 nm. Then a mask
having alternative stripes of metal and empty space was placed on the top of two of these
films and 70 nm thick films of AlN:Pr with no Gd were deposited on the top of the mask.
The spacing between two consecutive stripes on the mask was 1mm. In this way areas of
the already deposited AlN:1Pr4Gd films hidden by the metal of the mask could not get
AlN:Pr films. However in those areas where the mask was having empty spaces the
AlN:Pr films got deposited on the top of AlN:1Pr4Gd films. Thus in this way we got
stripes of alternative AlN:1Pr4Gd and AlN:Pr films. These stripes were then thermally
annealed for 30 minutes at 1025C. After thermal annealing CL was performed in the
same way and under the same conditions as done for thermally annealed AlN:Pr and
AlN:Gd films. Picture was obtained using a digital camera during the operation of CL.
The wide stripes in the figure show AlN:1Pr4Gd and the narrow stripes represent AlN:Pr
films. Luminescence enhancement is very clear from the picture. The wide stripes of
AlN:1Pr4Gd are much more brighter than the narrow stripes of AlN:Pr. Also the color of
the stripes seem to be yellow . However luminescence enhancement is a combination of
blue (490 nm) and red (618 nm and 652 nm). The reason for this yellowish color is that

101
no doubt enhancement occurs in blue and red regions but the combination of blue and red
colors are giving the sensation of yellowish color to the eye and camera. The
enhancement factors are already given in table 5.3.

Figure 5.9.

Geometry of the AlN:1Pr4Gd/AlN:1Pr films deposited using a mask.

102

Figure 5.10. CL picture of the enhanced luminescence in Pr by Gd.

103

Chapter 6

THULIUM AND SAMARIUM


Thin films of Thulium (Tm) doped AlN and Samarium (Sm) doped AlN are
deposited on flat Silicon substrates by RF Magnetron sputtering at liquid nitrogen
temperature. The deposited films are characterized for their structure and luminescence
using Cathodoluminescence (CL), Thermal Activation and X-Rays Diffraction (XRD).
XRD results show that all the films deposited at liquid nitrogen temperature are
amorphous. Luminescence from Tm occurs in vide range from ultraviolet to infrared
region with dominant blue emission at 467 nm, and other peaks at 371 nm, 480 nm, 528
nm, 650 nm, 685 nm, 792 nm, 802 nm and 808 nm. Sm gives three main transitions at
564 nm, 600 nm, 648 nm and 711 nm as a result of 4G5/2 6H5/2 , 4G5/2 6H7/2 , 4G5/2
6

H9/2 and 4G5/2 6H11/2

transitions. Thermal activation enhances the luminescence of

both materials. The co-doping of Gadolinium with Sm in AlN host enhances the Sm
luminescence with thermal activation.

104

6.1. Growth.
Thin films of Aluminum Nitride (AlN) doped with Thulium (Tm) and Samarium
(Sm), were prepared at 77 K by RF magnetron sputtering of bi-metal targets of Al with
Tm and with Sm in a pure nitrogen atmosphere in a cryo- pumped vacuum system. The
RF power was varied between 100 and 200 watts. Both kind of bimetal targets were made
by drilling one hole (0.55 cm diameter) each in two aluminum targets (4.2 cm diameter)
and placing a circular disc of 99.999% pure Tm and pure Sm in each hole. Elements were
then sputtered separately with aluminum. The background vacuum in the chamber was <
3 X 10-5 Torr. Films were deposited on Si (100) substrates. Nitrogen pressure between 5
to 8 m-Torr was used in the entire work. Liquid nitrogen was used for cooling the
substrates during low temperature deposition of films. Thickness of the deposited films
was between 60 nm and 1.2 m measured with a quartz crystal thickness monitor in the
growth chamber. X-rays diffraction (XRD) was used to determine the structure of the
films

6.2. Characterization.
6.2.1 X-Rays Diffraction (XRD) of Tm and Sm.
The structure of all films was studied by X-ray diffraction (XRD). Theta-2theta
(-2) scans were taken using a Rigaku diffractometer equipped with a double axis
sample stage goniometer. K x-rays with wavelength 1.54 A from copper crystal were
used for illumination purpose. The existence of Bragg diffraction peaks in the XRD
spectrum was used to confirm the crystal structure.

105
Figure 6.1 shows that films deposited at liquid nitrogen temperature are
amorphous and showing no peaks from AlN:Tm. Only one peak was observed at 69.1.
This is Si (100) peak. No other peak was found in the spectrum, showing that AlN:Tm
films deposited at liquid nitrogen temperature are amorphous. The scanning was done
from 20 - 80 for one hour in each case.
Figure 6.2 represents the XRD spectrum of AlN:Sm deposited on Si(100)
substrate at liquid nitrogen temperature. The only observable peak is from Si (100) at
69.1. There is no peak from Sm or AlN, showing that the deposited AlN:Sm film is
amorphous.

6.2.2 Cathodoluminescence of Tm.


CL studies of the as deposited films were performed at room temperature in a
vacuum chamber at a pressure of about 3x10-6 Torr, which was maintained with an
Alcatel CFF 450 turbo pump. Films were excited with an electron beam energy of 2.85
kV and beam current of 100 A measured with a multimeter. Luminescence from the
films was focused onto the entrance slit of a SPEX Industries double monochromator
with gratings blazed at 500 nm and detected at a Thorn EMI fast high gain
photomultiplier tube with a range of 300 900 nm.
Figure 6.3 shows CL emission from Tm doped amorphous AlN thin films. A
number of transitions from Tm in a wide range from ultraviolet (UV) to infrared (IR)
were found, with the most dominant peak in blue at 467 nm as a result of 1D2 3F4

106

100000
Si (100) peak
at 69.1
degrees

Log Intensity (a.u)

10000

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

Theta-2 Theta (deg)

Figure 6.1. XRD spectrum of AlN:Tm deposited at low temperature.

76

80

107

100000

Si(100)p e ak
at
69.1 d e g r e e s
10000

Log I (a.u)

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

76

The ta -2 the ta (de gre e s )

Figure 6.2. XRD spectrum of AlN:Sm deposited at low temperature.

80

108
transition. Other peaks were also found at 371 nm, 480 nm, 528 nm, 650 nm, and 802
nm. This makes Tm a very promising material in photonic and optoelectronic
applications and to be used in UV, visible and IR devices. A complete description of all
transitions from Tm [53-55] is given in table 6.1.
Figure 6.4 gives a schematic diagram of the energy levels of Tm responsible for
the observed transitions. Each energy level is represented by its spectroscopic notation
and each transition is represented by a line between the involved energy levels. The
wavelength of the emitted light is also mentioned with each line.

Table 6.1. Summary of Tm3+ ion emissions from Tm doped AlN.


Material

AlN:Tm

Transition

Transition

Number

Assignment

1
2
3
4
5
6
7

D2 3H6
1
D2 3F4
1
D2 3H6
1
D2 3H5
1
G4 3F4
1
D2 3H4
3
H4 3H6

Wavelength

Relative Intensity

(nm)

(a.u)

371
467
480
528
650
685
802

0.126
1.000
0.648
0.092
0.126
0.061
0.563

109

4000

467
nm

3500

3000
480 nm
2500

Intensity (a.u)

802 nm

2000

1500

1000
371
nm
500

528 nm
650
nm

685
nm

0
250 287 325 362 399 437 474 511 549 586 623 661 698 735 773 810 847 885

Wavelength (nm)

Figure 6.3.

Cathodoluminescence of amorphous AlN:Tm.

110

Figure 6.4. Energy levels diagram of Tm in amorphous AlN host.

111

6.2.3 Thermal Activation of Tm.


Thermal activation of amorphous AlN:Tm films was accomplished in the same
way by placing the films in the tube furnace 21100 at 900 C in a nitrogen atmosphere for
one hour. All the samples were placed in the furnace through the transfer arm of the
furnace. Nitrogen gas was allowed to pass through the hole at the other side arm of the
furnace.
Figure 6.5 shows luminescence from thermally annealed AlN:Tm films. Thermal
activation has enhanced the intensity of the emission. We believe that thermal activation
reduces defects in the structure and thermal stress, causes an increase in the
luminescence.
Figure 6.6 shows comparison of CL emission from AlN:Tm film and thermally
annealed AlN:Tm film. It is clear from the figure that thermal activation tremendously
enhances the luminescence of 467 nm peak, followed by luminescence enhancement in
371 nm and 480 nm peaks. However the rest of the transitions seem unaffected from
thermal activation.

112

14000

467 nm
12000

Intensity (a.u)

10000

8000

6000
480 nm

4000
802 nm

2000

371
nm

528
nm

650
nm

685 nm

0
248 286 323 361 398 436 473 511 548 586 623 661 698 736 773 811 848 886

Wavelength (nm)

Figure 6.5. CL spectra from thermally activated AlN:Tm films.

113

14000

AlN:Tm

AlN:Tm Thermally Annealed


467 nm

12000

8000

6000

480 nm

4000

371 nm

802 nm

2000

528
nm

650 nm

685 nm

880

852

825

797

770

743

715

688

661

633

606

578

551

524

496

469

442

414

387

359

332

305

277

0
250

Intensity (a.u)

10000

Wavelength (nm)

Figure 6.6. Comparison of CL spectra from thermally activated AlN:Tm films.

114

6.2.4 Cathodoluminescence of Sm.


Like other rare-earth elements Sm also gives luminescence when excited by
electron beam. CL studies of the as deposited and thermally annealed films of Sm were
performed at room temperature in a vacuum chamber at a pressure of about 3x10-6 Torr,
which was maintained with an Alcatel CFF 450 turbo pump. Films were excited with an
electron beam energy of 2.85 kV and beam current of 100 A. Luminescence from the
films was focused onto the entrance slit of a SPEX Industries double monochromator
with gratings blazed at 500 nm and detected at a Thorn EMI fast high gain
photomultiplier tube with a range of 300 900 nm.
Figure 6.7 shows luminescence from amorphous AlN:Sm excited by 100 A
beam current of 2.85 KeV electrons. The CL spectrum was taken at room temperature.
No emission was observed from Sm in UV or IR . The entire luminescence comes from
the red, orange and yellow parts of the visible range. Three main transitions were
observed in Sm. These transitions were recorded at 564 nm, 600 nm, 648 nm and 711 nm
as a result of 4G5/2 6H5/2 , 4G5/2 6H7/2 , 4G5/2 6H9/2 and 4G5/2 6H11/2 transitions.
This figure also shows that the orange emission from 4G5/2 6H7/2 at 600 nm is dominant
than the others followed by red emission from 4G5/2 6H9/2 at 648 nm and yellow from
4

G5/2 6H5/2 at 564 nm. The overall effect of all colors and intensities gives us an orange

red emission when observed with naked eye. One extra peak at 314 nm is due to
Gadolinium impurity. Current from filament in the CL process was 0.280 mA. Spectra of
AlN:Sm was also taken at 0.320 mA filament current.

115

1600

600 nm

1400

1200

648
nm

800

564 nm

600

711 nm
Gd
im purity
400

200

Wavelength (nm)

Figure 6.7. CL spectra from amorphous AlN:Sm films.

763
782

745

708
726

689

671

634
652

615

597

559
578

541

522

485
504

467

448

411
430

393

374

337
356

319

300

Intensity (a.u)

1000

116

Figure 6.8 gives a comparison of the same amorphous AlN:Sm films exposed to
0.280 mA and 0.320 mA filament current. It is clear that the luminescence from the films
exposed to o.320 mA current is pretty higher than the one exposed to 0.280 mA. Thus
exposure to higher filament current in CL will give intense and bright emission.

6.2.5. Thermal Activation of Sm.


Just like other rare-earth elements we also found that thermal activation enhances
the luminescence of Sm. Thermal activation of amorphous AlN:Sm films was
accomplished in the same way by placing the films in the tube furnace 21100 at 900 C in
600 nm is doubled after thermal activation, while the effect on 4G5/2 6H5/2 is even more
than that. Careful observation indicates that thermal activation has enhanced the
luminescence of 4G5/2 6H5/2 transition at 564 nm is more than triple after thermal
activation. Further a peak has appeared at 692 nm after thermal activation. This is the
chromium (Cr) impurity. The amount of impurity Cr impurity is so small that its not
observable before thermal activation of the films. CL spectrum of thermally activated
AlN:Sm films is compared with inactivated films in figure 6.9.

117

3000

Emission at 0.28 mA

Emission at 0.32 mA

600
nm
2500

648 nm

Intensity (a.u)

2000

1500

564
nm

1000

711 nm

500

Wavelength (nm)

Figure 6.8. Comparison of CL spectra from amorphous AlN:Sm films exposed to


0.280 mA and 0.320 mA currents.

799

777

754

731

709

686

663

641

618

595

572

550

527

504

482

459

436

414

391

368

345

323

300

118

3500

Inactivated Film

Thermally Activated Film

600
nm

3000

Intensity (a.u)

2500

2000
564 nm

1500

648
nm

+3

Cr
impurity

1000
711
nm

+3

Gd
Impurity

500

0
300

346.33 392.66 438.99 485.32 531.65 577.98 624.31 670.64 716.97 763.3
Wavlength (nm)

Figure 6.9. CL spectra from amorphous AlN:Sm films before and after thermal
activation.

119

6.2.6. Addition of Gd to Sm in AlN host.


Thin films of Gd and Sm co-doped to amorphous AlN were deposited by RF
magnetron sputtering in the same way as mentioned in the previous chapter. Thickness of
the films was 75 nm having equal concentration of Sm and Gd. The goal was to study
luminescence enhancement in Sm due to Gd. After deposition, the films were thermally
annealed for 40 minutes at 900 C. After activation Cathodoluminescence was performed
at room temperature and the results were compared to the CL spectrum of AlN:Sm films
annealed under the same conditions.
Figure 6.10 shows results from the thermally activated AlN:1Sm1Gd films
compared with the AlN:Sm films. It can be observed from the figure that Gd doping has
enhanced luminescence from 4G5/2 6H7/2 transition at 600 nm and 4G5/2 6H9/2 at 648
nm. However the 4G5/2 6H5/2 transition at 564 and 4G5/2 6H11/2 transition at 711 nm
are almost unaffected. Gd addition has also not affected the Cr impurity. All transitions
[56,57]

and corresponding luminescence peaks are given in table 5.3.

Table 6.2. Summary of Sm3+ ion emissions from Sm doped AlN.

Material

AlN:Sm

Transition

Transition

Number

Assignment

1
2
3
4
5
6

G5/2 6H5/2
G5/2 6H7/2
4
G5/2 6H9/2
4
G5/2 6H11/2
Gd impurity
Cr impurity
4

Wavelength

Relative Intensity

(nm)

(a.u)

564
600
648
711
314
692

0.425
1.000
0.686
0.312
0.276
0.220

120

4500

AlN:Sm

600 nm

4000

AlN:SmGd

3500

648 nm
2500

2000

564
nm

1500

Cr impurity
1000

Gd peak at
314 nm

711 nm

500

782

763

726
745

708

689

671

652

634

615

597

578

559

541

522

504

467
485

448

430

411

393

374

356

337

319

300

Intensity (a.u)

3000

Wavelength (nm )

Figure.6.10. Effect of Gd concentration on the luminescence of AlN:Sm films.

121

Chapter 7

TUNGSTEN AND YTTRIUM


Thin films of Tungsten doped AlN and Yttrium doped AlN are deposited on flat
Silicon substrates by RF Magnetron sputtering at liquid nitrogen temperature. The
deposited films are characterized for their structure and luminescence using
Cathodoluminescence (CL), Thermal Annealing, and X-Rays Diffraction (XRD. XRD
results show that all the films deposited at liquid nitrogen temperature are amorphous. A
very broad luminescence from tungsten is observed with the maximum peak at 491 nm
and a second peak at 429 nm. Yttrium gives three peaks in ultraviolet, blue and bluish
green regions at 360 nm, 421 nm and 518 nm respectively. Thermal annealing does not
bring any change in the luminescence from yttrium. Tungsten is co-sputtered with
holmium and gadolinium to study its effect on holmium luminescence and the effect of
gadolinium on tungsten luminescence. The holmium luminescence is found to be
unaffected by tungsten. However gadolinium enhances the luminescence of tungsten.

122

7.1. Growth.
Thin films of Aluminum Nitride (AlN) doped with Tungsten (W), Yttrium (Y)
and Titanium (Ti), were prepared at 77 K by RF magnetron sputtering of bi-metal targets
of Al with Tm and with Sm in a pure nitrogen atmosphere in a cryo- pumped vacuum
system. The RF power was varied between 100 and 200 watts. All bimetal targets were
made by drilling one hole (0.55 cm diameter) each in three aluminum targets (4.2 cm
diameter) and placing a circular disc of 99.999% pure W, Y and Ti in each hole.
Elements were then sputtered separately with aluminum. The background vacuum in the
chamber was < 3 X 10-5 Torr. Nitrogen pressure between 5 to 8 m-Torr was used in the
entire work. All films were deposited either on Si (111) or Si (100) or glass substrates.
Deposition rate was between 0.1 and 0.4 A / second. Liquid nitrogen was used for
cooling the substrates during low temperature deposition of films. Thickness of the
deposited films was between 60 nm and 400 nm measured with a quartz crystal thickness
monitor in the growth chamber. X-rays diffraction (XRD) was used to determine the
structure

of

the

films.

Luminescence

characterization

was

performed

in

Cathodoluminescence apparatus.

7.2. Characterization.
7.2.1. X-Rays Diffraction (XRD) of AlN:W films.
Thin films of AlN doped with tungsten (W) deposited at liquid nitrogen
temperature were characterized for their structures using X-ray diffraction (XRD). Theta2theta (-2) scans were taken using a Rigaku diffractometer equipped with a double axis

123
sample stage goniometer. K x-rays with wavelength 1.54 A from copper crystal were
used for illumination purpose. The existence of Bragg diffraction peaks in the XRD
spectrum was used to confirm the crystal structure.
Figure 7.1 shows that AlN:W films deposited at liquid nitrogen temperature are
amorphous and showing no peaks from Thulium. Only one peak was observed at 69.1.
This is Si (100) peak. No other peak was found in the spectrum, showing that AlN:W
films deposited at liquid nitrogen temperature are amorphous. The scanning was done
from 20 - 80 for one hour in each case.

7.2.2. Cathodoluminescence of AlN:W.


Luminescence spectrum of amorphous AlN:W was obtained using CL
instruments in the same way as did for all other discussed films. The spectrum given in
figure 7.2 gives the results of CL for these films. This figure represents a very broad blue
emission from amorphous AlN:W film. The luminescence is so broad that it covers a
wide range from 360 nm UV and reaches up to 650 nm red region. However the
dominant portion of the spectrum comes in blue region with a peak at 491 nm. Another
peak in blue is also observed at 429 nm. Because of the huge portion of the spectrum and
its peaks lying in blue the films also looked blue in appearance when directly exposed to
electron beam in CL and watched unaided. Because of this broad blue emission by
tungsten it was also co-doped with holmium and gadolinium separately to test for

124

1000000

Si (100) peak
at 69.1
100000

Log Intensity (a.u)

10000

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

2theta (deg)

Figure 7.1. XRD analysis of W doped AlN films.

68

72

76

80

125

600
491 nm

500

429 nm

300

200

100

Gd
impurity
at 314 nm

Wavelength (nm)

Figure 7.2.

CL spectrum of amorphous AlN:W

788

768

749

729

710

690

671

651

632

612

593

573

554

534

515

495

476

456

437

417

398

378

359

339

320

0
300

Intensity (a.u)

400

126
luminescence enhancement in the green emission from holmium ions or luminescence
enhancement in the AlN:W by gadolinium. It was observed that no change has occurred
in Ho luminescence by the addition of W. However the luminescence from Gd has
enhanced W.
Figure 7.3 shows luminescence from amorphous AlN films co-doped with 1 piece
of W and 1-3 pieces of Gd. It is clear from the figure that the addition of Gd has a
dominant effect on the luminescence from W. Luminescence from AlN:1W2Gd is 3
times the luminescence from AlN:1W1Gd and luminescence from AlN:1W3Gd is almost
6 times the luminescence from AlN:1W1Gd. This fact shows a huge increase in
intensities in the W luminescence by Gd. The exact mechanism for this luminescence
enhancement is still not known. Moreover the films are not tested for luminescence after
thermal annealing.
Figure 7.3 also reveals some other information. It can be observed that the main
peak in blue shifts with the relative increase in Gd concentration. With one piece of Gd
the main peak of W appears at 506 nm. With two pieces of Gd added to AlN:W the peak
gets bigger but shifts to 534 nm. However with the addition of another piece of Gd the
peak gets bigger but shifts back to 523 nm. A possible explanation for this shift in the
peaks may be the activation of new radiative energy levels in W by the Gd. Further the
intensity of Gd peak is reducing with higher concentration of Gd. Table 7.1 is giving the
increase in the W intensity with the concentration of Gd and also the intensity of Gd peak
with the concentration of Gd.

127

2500

2nd order diffraction


at 628 nm

AlN:1W1Gd

AlN:1W2Gd

Gd peak
at 314 nm

AlN:1W3Gd

2000

Intensity (a.u)

1500

523
nm

1000

534
nm

426
nm

500

506 nm

780

759

737

715

693

671

649

628

606

584

562

540

518

497

475

453

431

409

387

366

344

322

300

Wavelength (nm)

Figure 7.3. Effect of Gd concentration on the luminescence of AlN:W films.

128
Table 7.1.

Gd Concentration

Effect on Gd concentration on the W and Gd luminescence.

Intensity of Gd
Luminescence(a.u)

Intensity of W
Luminescence (a.u)

% increase in
W intensity

2207

208

Reference

1318

705

339 %

538

1159

557 %

As mentioned earlier W is added to Ho in order to observe any enhancement in the green


luminescence from Ho. Figure 7.4 shows the effect of varying W concentration on Ho. A
comparison between amorphous AlN:1Ho1W, AlN:1Ho3W and thermally activated
AlN:1Ho1W is given in this figure. It is clear from the figure that there is no effect of
increasing W concentration on the Ho luminescence. The W concentration is increased 3
times which made the W peak broad but no enhancement in Ho. Further, if we compare
this figure with the pure W spectrum then a shift can be seen in the W emission
wavelength just like that happened due to Gd. And we see from the figure that in the
presence of Ho, tungsten emits a single blue light with a wavelength of 461 nm rather
than emissions at 429nm and 491 nm in pure tungsten. A possible explanation may be the
holmium has affected the energy levels distribution in tungsten. Further it can also be
given in explanation that this shift of wavelength did not occur in rare-earth elements
discussed in previous chapters but the transition element W has suffered a lot from this
shift. Since rare-earth elements are favorable for their internal f-f transitions. These flevels lying inside the other shells and hence any external change barely affect these

129
transitions. However it is not true for transition metals and they can be significantly
affected by other dopant impurities. Moreover this figure also gives us the effect of
thermal annealing on the luminescence of W and Ho. It can be easily deduced from the
figure that thermal annealing has activated Ho luminescence more than W. There is also
existence of a ruby impurity, which is considerably enhanced in luminescence by thermal
activation.

130
700
AlN:1Ho1W
AlN:1Ho3W

549 nm Ho

+3

AlN:1Ho1Wthermally annealed
600

500

461 nm

Intensity (a.u)

+3

400

702 nm
Cr

300

+3

661 nm
Ho

+3

200

100

0
198

248

298

348

398

448

498

548

598

648

698

748

Wavelength (nm)

Figure 7.4.

The effect of thermal activation and W concentration on Ho luminescence.

131

7.2.3. X-Rays Analysis of AlN:Y


The structure of AlN:Y films deposited at low temperature was also determined
by x-rays diffraction in the same way as described earlier. Figure 7.5 shows the result
from x-rays analysis of AlN:Y films. The films were deposited on Si(111) substrates and
the XRD spectrum shows just one peak at 28.4 which corresponds to Si(111). No other
peak is observable in the spectrum. Thus the AlN:Y films deposited at low temperature
are also amorphous just like the others discussed earlier.

7.2.4. Cathodoluminescence and Thermal Annealing of AlN:Y


Cathodoluminescence of cold deposited Yttrium doped AlN films were
characterized for CL in the same way described earlier for other materials. Figure 7.6
represents the luminescence from AlN:Y films when exposed to cathode rays in CL
assembly. This figure confirms emission from amorphous AlN:Y films in UV , blue and
in bluish green regions. There is also a peak in IR which may correspond to Yttrium or
possibly a ruby impurity. The wavelengths correspond to these emissions are 360 nm,
421 nm, 518 nm and 705 nm respectively. The emission 705 nm could be from alN:Y or
from Cr+3 The bluish green and blue emissions are the dominant in intensity and that is
why the films appear blue to naked eye when exposed to electron beam in CL. However
we already know from the previous materials investigations that thermal annealing
tremendously enhance the luminescence from Cr+3 and hence we can analyze this peak
obtaining the CL after performing thermal annealing.

132

10000
Si (111)
peak
at 28.4

Log Intensity (a.u)

1000

100

10

1
20

24

28

32

36

40

44

48

52

56

60

64

68

72

76

80

Theta-2 Theta (deg)

Figure 7.5. X-rays analysis of the structure of low temperature deposited AlN:Y films.

133

800
518 nm
700
421 nm
600

Intensity (a.u)

500

400
705 nm

300

360 nm

200

100

0
350

390

430

470

510

550

590

630

670

710

Wavelength (nm)

Figure 7.6.

CL spectrum of amorphous AlN:Y

750

790

134
Figure 7.7 gives a comparison of CL spectra from amorphous AlN:Y before and
after thermal annealing. Films were thermally annealed at 900 C for one hour and then
characterized for CL. From the figure it is clear that thermally annealing has almost not
affected the Yttrium luminescence at 360 nm, 421 nm and 518 nm. However the peak at
705 nm is strongly enhanced by thermal annealing. This enhancement in the peak at 705
nm is more than three and half times the luminescence from the same peak in the film,
which is not thermally annealed. Further, this peak was less than half the intensity of
peak at 518 nm before thermal annealing but after annealing its about one and half time
more intense than the peak at 518 nm. One can guess at this stage that the peak at 705 is
most favorably due to chromium oxide (Ruby) rather than emission from yttrium ion
itself but on the other hand.

135

1200
AlN:Y
AlN:Y Annealed

705 nm

1000

518 nm

421 nm
600

400
360 nm

200

Wavelength (nm)

Figure 7.7. CL spectrum of thermally activated amorphous AlN:Y films.

782

764

746

728

710

692

674

656

638

620

602

584

566

548

530

512

494

476

458

440

422

404

386

368

0
350

Intensity (a.u)

800

136

Chapter 8

STOPPING POWER OF AlN AND


ELECTRON PENETRATION DEPTH
Stopping power of AlN is determined for electrons using Cathodoluminescence. Thin
film bilayers of AlN:Ho and AlN:Tm are deposited on flat Silicon substrates by RF
Magnetron sputtering at liquid nitrogen temperatures. 15.3 nm thick AlN:Ho film on a
Si(111) substrate. On the top of this AlN:Ho film 37.8 nm thick AlN:Tm film is
deposited to make it a bilayer. Electron beam of different energies, obtained from
electron gun of the CL apparatus, is allowed to penetrate in the AlN:Tm/AlN:Ho bilayer
film. The thickness of the film and the energies of electron beam are enough to get the
stopping power of AlN. Energy of the beam just crossing 37.8 nm AlN:Tm film is used to
obtain the stopping power experimentally. Experimental results are compared to the
theoretical value using the established mathematical equations for stopping power. A
percent deviation of 6.6 % is found in the experimental and theoretical results.

137

8.1

Bilayer Films Growth.


Thin film bilayers of AlN:Ho and AlN:Tm were prepared at room temperature

and 77 K by RF magnetron sputtering of holmium-aluminum and gadolinium-aluminum


bi-metal targets separately in a pure nitrogen atmosphere in a cryopumped vacuum
system. The RF power was varied between 100 and 200 watts. The bimetal targets were
made by drilling one hole (6.5 mm diameter) in two different aluminum targets (42 mm
diameter) and placing circular discs of 99.999% pure holmium and thulium in the holes.
Holmium and thulium were then co-sputtered with aluminum. The background vacuum
in the chamber was < 3 X 10-5 Torr. All the holmium films were deposited on flat Si(100)
or Si(111) substrates and optical fibers with diameters ranging from 10 m to 80 m.
Gadolinium films were deposited only on flat Si(100) or Si(111) substrates. Nitrogen
pressure between 5 to 8 m-Torr was used in the entire work. Liquid nitrogen was used for
cooling the substrates during low temperature deposition of films. AlN:Ho film was first
deposited on Si(111) substrate for 30 minutes. AlN:Tm film was deposited on the top of
the deposited AlN:Ho film for 30 minutes under the same conditions. The thickness of
AlN:Ho film measured with a quartz crystal thickness monitor in the growth chamber,
was 15.3 nm and that of AlN:Tm was monitored to be 37.8 nm. X-rays diffraction (XRD)
was used to analyze the structure of films. No diffraction peaks were obtained and hence
the films were amorphous.

8.2

Basic Idea Behind the experiment.


The average energy loss per unit length by a charged particle in passing through a

material is called the Stopping Power of that material. Mathematically it is usually

138
represented by -dE/dx where negative sign shows the loss of charged particle energy
with penetration. Materials have different stopping powers for different charged particles.
Stopping power gets special attention when charged particles impart energy to the human
body while passing through the body. People have tried to determine stopping power of
different media for different charged particles. We have used Cathodoluminescence to
determine the stopping power of AlN for electron beam and the penetration depth of
electron beam. Since by definition the stopping power is the energy loss per unit length
and hence if we know the thickness of the material and the energy of charged particles
passing through that thickness of material then we can easily obtain stopping power and
the depth penetration of the charged particles.
In the present work a bilayer of AlN:Ho and AlN:Tm of known thicknesses on Si
(111) substrate is used as media or target material and electron beam obtained from the
electron gun in CL instrument is allowed to penetrate in this bilayer target. The aim is to
determine the stopping power of AlN for electron beam. From the discussion in our
previous chapters we know that AlN:Ho emits in broad range from UV to infrared region
with the most intense emission in green at 549 nm as a result from 5S2 5I8 transition.
Emission from AlN:Tm also extends over a wide range from UV to IR with dominant
peaks in blue at 467 nm and 480 nm caused by 1D2 3F4 and 1D2 3H6 transitions
respectively. The idea is when electron beam is allowed to fall on this bilayer target, it
will penetrate in the target. The penetration depth depends upon the energy of the beam.
When the energy of the electron beam is low, the electrons wouldnt make it possible to
cross the thick AlN:Tm film and as a result only the CL spectrum of AlN:Tm will be
obtained. The beam energy will be increased slowly and for different energetic beams CL

139
spectra will be recorded. Increasing the energy of the beam the penetration of electrons
will increase and a stage will reach that at a specific amount of energy the electrons will
just cross the AlN:Tm film and will hit the AlN:Ho film deposited below the AlN:Tm
film. At this stage spectrum of AlN:Ho will also appear in CL along with the AlN:Tm
spectrum. Since the thickness of AlN:Tm film is known and hence knowing the energy of
electron beam just crossing this thickness we can calculate the stopping power of AlN
and the penetration depth of electrons in AlN. Figure 8.1 gives pictorial explanation of
the electron beam penetration in AlN:Tm/AlN:Ho bilayer target. The addition of thulium
to AlN wouldnt affect the stopping power value of AlN because thulium is less than 0.5
%. This small concentration of thulium is not enough to change the stopping power of
AlN.

8.3.

Cathodoluminescence.
CL studies of the bilayer films deposited on flat silicon substrate was performed

at room temperature in a vacuum chamber at a pressure of about 3x10-6 Torr, which was
maintained with an Alcatel CFF 450 turbo pump. Films were excited with an electron
beam of maximum energy 2.853 keV and beam current of 100 A. Luminescence from
the films was focused onto the entrance slit of a SPEX Industries double monochromator
with gratings blazed at 500 nm and detected at a Thorn EMI fast high gain
photomultiplier tube with a range of 300 900 nm. CL spectra were obtained for incident
electron beam of energies 2.4 KeV, 2.5 KeV, 2.6 KeV, 2.7 KeV, 2.8 KeV and 2.853
KeV.

140

Figure 8.1.

A. Penetration of low energy electron beam in the AlN:Tm/AlN:Ho


bimetal target.
C. Penetration of high energy electron beam in the AlN:Tm/AlN:Ho
bimetal target.

141
Figure 8.2 shows the CL spectra of the bilayer film in 300 nm to 750 nm range,
excited by the electron beams with those different energies. For a clear observation
constant numbers 200, 400, 600, 800 and 1000 are added to the spectra taken with 2.5
KeV, 2.6 KeV, 2.7 KeV, 2.8 KeV and 2.853 KeV beam energy respectively. It is clear
from the figure that the spectrum taken with 2.4 KeV beam does not show any
appearance of AlN:Ho peak at 549 nm. CL spectrum with 2.5 KeV beam shows a slight
rise in the curve. This small rise indicate that the electron beam has just crossed the 37.8
nm thick AlN:Tm layer and reached AlN:Ho film. Beyond this energy the intensity of
AlN:Ho peak at 549 nm increases with beam energy. The increase in the intensity of the
luminescence is given in figure 8.3. This information about the 2.5 KeV electron beam
just crossing the 37.8 nm layer of AlN:Tm can be exploited to determine the stopping
power of AlN.

8.4

Stopping Power of AlN.


The information obtained from CL spectra is used to calculate the stopping power

of AlN for electrons. The value of stopping power obtained from the experimental
information is compared to its theoretical values using established mathematical
expressions for stopping power determination.
As discussed earlier the stopping power is the average energy loss per unit length
and is denoted by -dE/dx
The energy of the electron beam lost when passing through 37.8 nm thick AlN:Tm layer
is 2.5 KeV and the thickness of the film in which 2.5 KeV energy is lost is 37.8 nm. Thus
(-dE/dx)exp = 2.5 KeV / 37.8 nm

142

8000
Spectrum at 2.4 KeV
Spectrum at 2.5 KeV

467 nm

7000

480 nm

Spectrum at 2.6 KeV


Spectrum at 2.7 KeV

6000

Spectrum at 2.8 KeV


Spectrum at 2.853 KeV

Intensity (a.u)

5000
371 nm
4000

549 nm

650 nm

3000

2000

1000

0
300 325 350 375 400 425 450 475 500 525 550 575 600 625 650 675 700 725 750

Wavelength (nm)

Figure 8.2.

CL spectra of AlN:Tm/AlN:Ho bilayer at different electron beam energies.

143

500

450

400

Intensity (a.u)

350

300

250

200

150

100

50

0
2.4

2.5

2.6

2.7

2.8

2.853

Beam Energy (KeV)

Figure 8.3.

Variation in the intensity of 549 nm AlN:Ho peak with beam energy.

144

(-dE/dx)exp = 1.058 E-8 J/m

(8.1)

Equation (8.1) gives experimental value for the stopping power of AlN for electrons.
The stopping power of AlN for electrons is obtained theoretically using the equation [58,59]
-dE/dx = (4k02e4n / m0c2 2) [ ln {(m0c2 +2 ) / I 2} + F () ]
Where F () = (1-2)/2 [1+2/8 (2+1)ln2]
And

(8.2)
(8.3)

k0 = 8.997 E9 Nm2/C2
e = Charge of electron.
n = Electron density of the material.
c = Speed of light.
m0c2 = Rest mass energy of electron.
= v / c = 0.011 for 2.5 KeV electron.
= Ratio of the kinetic energy and rest mass energy of electron.
I = 11.2 + 11.7 Z eV

Using the energy of our electron beam to be 2.5 KeV in equation (8.3) we obtain the
value of F () to be 0.151. Using these values of F () , energy of electron beam and
other parameters in equation (8.2) we obtain the theoretical value of AlN stopping power,
which is given below
(-dE/dx)th = 9.88 E-9 J/m

(8.4)

Comparing equation (8.1) and (8.4) we get a percent difference of 6.6% in experimental
value and theoretical value of the stopping power of AlN.

145

Chapter 9

DISCUSSION AND CONCLUSION


9.1. Discussion.
One of the important sections of this work is the luminescence enhancement in Holmium,
Praseodymium and Samarium ions by co-doping Gadolinium to these materials doped in
AlN host, with the exceptions of Thulium where we do not see any enhancement of
luminescence.
Gadolinium was co-doped in different concentrations with AlN:Ho. The hypothesis is
that whether the addition of gadolinium can enhance luminescence in AlN:Ho or not ?
Different experiments were performed to testify our hypothesis and to study the physics
behind the results we obtained from experiments. AlN:1Ho6Gd and AlN:Ho films were
deposited on Si (111) substrates using a mask such that we get alternative stripes of these
films. The stripe of AlN:1Ho6Gd was more intense as compared to that of AlN:Ho,
showing the luminescence enhancement in Ho by Gd. First 100 nm thick films of AlN
co-doped with 1 piece of Ho (circular disk shaped with 6 mm diameter) and 6 pieces of
Gd (each of the size of Ho piece) in an Al target of 42 mm diameter were deposited on Si
(111) substrates. Then a mask with alternative stripes of metal and empty spaces was
placed on the top of one of these films and 80 nm thick films of AlN:Ho with no Gd were
deposited on the top of the mask. The spacing between two consecutive stripes on the
mask was 1mm. In this way AlN:Ho films were not deposited on the top of the already
deposited AlN:1Ho6Gd films in those areas where AlN:1Ho6Gd films were under the

146
metal of the mask. However in those areas where the mask had empty spaces the AlN:Ho
films got deposited on the top of AlN:1Ho6Gd films. In this way we got stripes of
alternative AlN:1Ho6Gd and AlN:Ho films. These stripes were then thermally activated
for 40 minutes at 900C. After thermal activation CL was performed under electron
illumination at 2.85 KeV with a nominal current density of 15 A /mm2.
Figure 9.1 gives the picture of those alternative stripes of AlN:1Ho6Gd and
AlN:Ho films obtained using a digital camera during the operation of CL. The wide
stripes in the figure show AlN:1Ho6Gd and the narrow stripes represent AlN:Ho films.
Luminescence enhancement is very clear from the picture. The wide stripes of
AlN:1Ho6Gd are much brighter than the narrow stripes of AlN:Ho. Obviously there is an
enhancement of green luminescence from Ho due to the addition of Gd.
Figure 9.2 represents picture of the stripes of pure holmium without adding any
gadolinium. Alternative three layers of AlN:Ho films were deposited upon each other
using the same mask. No Gd was added to the films this time. But still we see that some
areas are brighter than others. It means luminescence enhancement has occurred even
without adding any Gd to the films. If we analyze figure 9.1 and figure 9.2 then it is
obvious that the enhancement in luminescence of holmium is not only due to the effect
from gadolinium but also the stripes itself enhance the luminescence. We can say that the
enhancement in Ho+3 apparent in figure 9.1, is not solely from Gd+3 but some of the
spatial intensity increase is from geometrical consideration of the stripes. One possible
explanation could be the interference phenomenon and the bright stripes or spots may be
fulfilling the conditions for constructive interference. This possibility was tested using a
532 nm Neodymium diode

147

Figure 9.1. CL picture of the enhanced luminescence in Ho by Gd.

148

Figure 9.2. CL picture of three layers of AlN:Ho deposited upon each other using a
mask.

149
laser. The laser beam was allowed to fall on the film at different angles and each time the
intensity of luminescence was recorded.
Figure 9.3.shows the intensity of green luminescence of 5S2 5I8 transition in
Ho+3 ion for different angles of laser beam making with the surface of the film. It is clear
that the same peak is more intense when the laser beam makes an angle of 23 and 48
with the film. For the rest of the angles the intensity of luminescence is almost the same
and lowers as compared to the intensities at 23 and 48. This situation can be explained
with the help of interference conditions. The condition for constructive interference is
d = n (9.1)
Where d is the path difference between two coherent waves, is the wavelength and
n is an integer. The values of n can be 0,1,2,3,4.
Figure 9.4 gives the experimental setup followed by the laser beam. In this figure
Path difference = d = AB + BC
Film Thickness = BD
If the laser beam makes an angle of 48 with the surface of the film then mathematically
AB = BC = BD / sin(48)
But the total thickness of all layers BD is 211 nm so the calculation gives
AB = BC = 283 nm
And hence the path difference d will be
d = 283.5 + 283.5 = 567 nm
The index of refraction of AlN is 2.00. If we also take this index of refraction in account
then the total path difference d will be
d = 1134 nm

150
The path difference is almost twice of the wavelength of the green light emitted from
AlN:Ho and hence we can say that the constructive interference conditions are satisfied
with integer n = 2 when the laser beam makes an angle of 48 with the film.
If the laser beam makes an angle of 23 with the surface of the film then
AB = BC = BD / sin (23)
Which gives AB = BC = 541 nm
And hence the path difference d = 1082 nm
Including the effect of the index of refraction of AlN, the total path difference will be
2164 nm. This path difference is almost four times the wavelength of the emitted green
light and hence again it satisfies the conditions of constructive interference with the
integer n = 4
Thus the experimental results show that a part of the luminescence is due to the
interference phenomenon in the film.
The gadolinium concentration in the film was also tested experimentally. Figure
9.5 confirms the increasing concentration of Gd in the films where the luminescence is
enhanced. EDX was performed using SEM and EDX analyzer. AlN:1Ho, AlN:1Ho1Gd
and AlN:1Ho4Gd films were analyzed separately. The figure shows the existing of Ho
peak at 6722 KeV in the AlN:Ho sample and an additional peak at 6046 KeV in the
AlN:1Ho1Gd films. This peak corresponds to Gd and hence indicates the presence of Gd
in the films. The intensity of the Gd peak has increased in the AlN:1Ho4Gd films,
confirming the presence of higher Gd concentration in the AlN:1Ho4Gd films as
compared to AlN:1Ho1Gd films.

151
To further analyze the part of luminescence enhanced by Gd let us concentrate on
figures 9.2 and 9.6 (the former figure 4.16). Figure 9.6 shows that the Ho luminescence
increases linearly with the increasing Gd concentration. This enhancement saturates when
Gd concentration gets four times its initial concentration. After the saturation point no
further enhancement is seen. This saturation of concentration at a particular Gd
concentration indicates that Gd ions affect the luminescence centers of Ho ions when

152

10000000

1000000

Log Intensity (a.u)

100000

10000

1000

100

10

1
10

15

23

40

48

58

70

Angle with Film (degrees)

Figure 9.3.

Intensity of luminescence from AlN:Ho film with the angle of laser beam.

153

Figure 9.4.

Ray-diagram of the path followed by laser beam giving interference.

154
they are in the vicinity of these luminescence centers. This enhancement in Ho by Gd
might be due to any of the following processes.

Direct UV excitation of Ho ions by the sharp and intense UV emission


from Gd.

Energy transfer from Gd to Ho.

As far as the direct pumping is concerned it is photoluminescence in which the


light emitted from one kind of ions directly excites the other kind of ions doped together.
P. A. Budni et al [60] has already obtained direct excitation of Holmium ions by Thulium
ions and have produced 2.09 m laser in Ho:YAG from 5I7 5I8 transition, pumped
directly by

F43H6 transition at 1.9 m from Tm+3 ions. Tm+3 ion was resonantly

pumped by diode thulium laser at 1.9 m. Tm+3 ion is a strong emitter at 1.9 m which is
responsible for the direct pumping of Ho+3 ion. In our work we were expecting the same
kind of direct excitation of 5S2 5I8 transition in Ho+3 at 549 nm by the 6P7/2 8S7/2 Gd
transition at 314 nm. We measured the excitation spectrum of AlN:Ho in UV regions but
could not get any excitation peak in UV region. Figure 9.7 shows the excitation spectrum
of AlN:Ho from 200 nm to 350 nm using a Xenon lamp. The wavelength of emission set
for this excitation range was 549 nm. No peak is found in the 200 350 nm range,
showing that no excitation levels of AlN:Ho exist in the scanned UV region. Since Gd is
a UV emitter in 314 nm and hence its not possible for the Gd+3 ion to excite Ho+3 ion
directly.
The energy transfer mechanism is studied by measuring the life time of Ho before
and after the Gd doping to the pure AlN:Ho films. The basic idea in this process is to
transfer electrons from one kind of ions to the excited states of other kind of ions to

155
enhance the luminescence of the later ion. X. J. Wang et al

[61]

have observed such kind

of energy transfer phenomena in Pr+3- and Er+3- co-doped in CaAl12O19 crystal. They
found that the luminescence from 2H11/2 4I15/2 transition of Er+3 at 537 nm is enhanced
by the electrons transfer from 3P0 and 1S0 levels of Pr+3 ion to the 2H11/2 excited level of
Er+3 ion. 3P10 and 1S0 levels of Pr+3 ion are responsible for the 3P0 3H4 transition at 488
nm and 1S0 3F4 transition at 253 nm. They measured the life time of Er emission at 537
nm with and without the doping of Pr ions and found a huge difference in its life time
with and without energy transfer from Pr. In our work we were expecting such kind of
energy transfer from the 6P7/2 level of Gd+3 ion to the excited 5S2 level of Ho+3 ion to
enhance the green luminescence from 5S2 5I8 transition of Ho+3 ion at 549 nm but we
could not get any conclusive information.
Thus we do not know any mechanism exactly responsible for the luminescence
enhancement.
Figure 9.6 is the spectrum of AlN:Ho doped with different concentrations of Gd.
The spectrum directly reveals that Gd does enhance Ho luminescence. On the other hand
figure 9.2 is the direct picture of AlN:Ho stripes without any Gd added. Three layers of
AlN:Ho stripes are deposited upon each other with no Gd concentration. Luminescence
enhancement is still present even without Gd and just with the stripes. It clearly indicates
that the geometry of stripes and the thickness of films also play a role in the enhancement
of luminescence. Further since we obtained luminescence in the pure Ho stripes in the
same manner as we obtained in the presence of Gd. This pushes us to think in another
direction that its the rare-earth itself which changes the local environment of the
luminescence center. As discussed earlier the size of a rare-earth ion is even bigger than

156
the size of the host material and we believe that this bigger size plays significant role in
the vicinity of luminescence centers. But at the same time we cannot ignore the effect of
stripes geometry and thickness of the films itself.
Luminescence enhancement is also observed in AlN:Pr doped with Gd+3. Figure
9.8 is the direct spectrum and figure 9.9 is the picture of the Gd doped AlN:Pr films.
Luminescence enhancement in Praseodymium is clear from both figures. Figure 9.9 also
shows that after getting enhancement the bright regions are neither blue nor red, but they
look yellowish in appearance. The reason may be the combination of blue and red. The
wavelength of yellow color occurs in between blue and red and hence the combination of
blue at 490 nm and red at 618 nm and 652 nm give an overall effect of yellow color. The
spectrum shows that luminescence enhancement is more than 100%.
Figure 9.10 shows the luminescence enhancement in Sm due to Gd doping. This
figure shows that the luminescence enhancement is dominant in the two peaks at 600 nm
and 648 nm. Although the peak at 564 nm is also one of the Sm dominant peaks but we
do not see any luminescence enhancement in this transition. May be the addition of Gd
enhances luminescence in some transitions and not in some others. This could also be one
of the reasons we could not see any luminescence enhancement in the blue emitting Tm.
Figure 9.11 reveals information about luminescence enhancement in W due to the
addition of Gd to AlN:W films. The spectrum shows that luminescence enhancement is
about 5 times when the Gd concentration is increased to three times of its initial
concentration in AlN:W films. A shift in the W transitions is also observed which will be
discussed latter.

157
Further it was also observed that no luminescence enhancement occurs if the
stripes are not thermally activated before CL is performed. Thermal stresses and defects
are released during thermal activation, which causes enhancement in luminescence.
Similarly oxidation of the films also occurs during thermal annealing and oxygen effect
enhances luminescence in some rare-earth materials like Europium

[62]

and decreases in

some other rare-earth elements like Pr [63]. Perhaps Gd and Ho respond in different way to
oxygen effect and thermal annealing which could have brought such a big changes in the
luminescence of thermal activation.

158
6000

Gd pe ak at 6046
Ke V

AlN:1Ho
AlN:1Ho2Gd

5000

AlN:1Ho3Gd
AlN:1Ho4Gd
Ho pe ak at 6732 Ke V

3000

2000

1000

19920

19135

18349

17564

16779

15994

15209

14424

13638

12853

12068

11283

10498

9712

8927

8142

7357

6572

5786

0
5001

Intensity (a.u)

4000

Energy (KeV)

Figure 9.5. Comparison of EDX spectra of AlN:Ho and AlN:1Ho.xGd.

159

8000

549 nm Ho
7000

314 nm Gd
peak

AlN:Ho
AlN:1Ho1Gd

6000

AlN:1Ho2Gd
AlN:1Ho3Gd
AlN:1Ho4Gd

AlN:1Ho5Gd

AlN:1Ho6Gd
4000

3000

661 nm
2000

758 nm

1000

Wavelength (nm )

Figure 9.6. Effect of Gd concentration on Ho luminescence.

785

760

734

709

683

658

632

606

581

555

530

504

479

453

428

402

377

351

326

0
300

Intensity (a.u)

5000

160

350

300

200

150

100

50

335

320

305

290

275

260

245

230

215

0
200

Intensity (a.u)

250

Wavelength (nm)

Figure 9.7.

Excitation spectrum of AlN:Ho, using Xenon lamp.

161

7000
AlN:Pr

490 nm

6000

AlN:1Pr1Gd
AlN:1Pr4Gd

314 nm
Gd peak

652 nm

4000
618

3000
542 nm

2000
738
439 nm

1000

Wavelength (nm)

Figure 9.8. Effect of Gd dopant on the luminescence of Pr.

779

757

736

714

692

670

649

627

605

583

561

540

518

496

474

452

431

409

387

365

344

322

300

Intensity (a.u)

5000

162

Figure 9.9. Picture of luminescence enhancement in Pr due to Gd dopant.

163

4500

AlN:Sm

600 nm

AlN:SmGd

4000

3500

648 nm
2500

2000

564
nm

1500

Cr impurity
1000

Gd peak at
314 nm

711 nm

500

Wavelength (nm )

Figure 9.10.

Spectrum of the enhanced luminescence in Sm.

782

763

726
745

708

689

671

652

634

615

597

578

559

541

522

504

467
485

448

430

411

393

374

356

337

319

300

Intensity (a.u)

3000

164

2500
2nd order diffraction
at 628 nm

AlN:1W1Gd

AlN:1W2Gd

Gd peak
at 314 nm

AlN:1W3Gd

2000

Intensity (a.u)

1500

523
nm

1000
534
nm

426
nm

500

506 nm

780

759

737

715

693

671

649

628

606

584

562

540

518

497

475

453

431

409

387

366

344

322

300

Wavelength (nm)

Figure 9.11.

Effect of Gd concentration on the luminescence of Tungsten.

165
Another interesting phenomenon observed is the shift in W peaks when Gd or Ho
are added to pure AlN:W films. Pure tungsten gives two transitions in blue at 491 nm and
429 nm. It is observed that after the addition of Gd or Ho to AlN:W, the transitions shifts
from their original place. Look back at figure 9.11 which gives the effect of Gd addition
to AlN:W films. The figure shows that by the addition of Gd to pure AlN:W films not
only the transition wavelengths change but also if we vary the Gd concentration then the
transitions further shifts with every different concentration of Gd. When the
concentration of Gd is equal to W concentration then the W emission occurs at 503 nm.
When it is further increased to two times the W concentration then the emission in W
occurs at 534 nm while a three times bigger concentration of Gd shifts the W peak to 523
nm. A slight shift in the peak at 429 nm also occurs with an emission at 426 nm with
three times Gd concentration. A possible explanation for this shift in the peaks may be
either the activation of new radiative energy levels in W by the Gd or unlike the inner
lying f-levels of rare-earth elements, transition metals do not have any unfilled level
existing inside the core. And hence it is very easy to bring changes in the energy levels of
a transition metal. Since W is also a transition metal so we also believe that the addition
of Gd can significantly affect the energy levels of W, giving rise to a shift in the emitting
wavelength. Different concentrations of Gd can affect in different ways and hence we see
shifts in wavelength for every different Gd concentration added to pure W.
Figure 9.12 shows the effect of Ho on the energy levels and luminescence of pure
AlN:W. It is clear from the figure that with the addition of Ho no change occurs in the
luminescence intensity of W. Also there is no change in the luminescence of Ho. This
fact confirms that W and Ho do not have any effect on the luminescence intensities of

166
each other. However it is very obvious that the wavelength of emission in W has changed
after the addition of Ho. Pure W emits at 491 nm but the addition of Ho changes this
emission to 461 nm. The explanation already given for the shift in W wavelength due to
Gd addition can be a possible explanation for this shift as well. Another possible
explanation for this may be the 459 nm Ho peak. It is possible that because of
luminescence from Ho at 459 nm new radiative energy levels might be activated in W,
which made it possible, that W may be emitting more favorably at 461 nm rather than
491 nm.
Another exciting result is the confirmation of two additional levels in
praseodymium at 335 nm and 385 nm when deposited in AlN host. These levels are
already reported theoretically by E.van der Kolk et al [64]. The direct observations of these
levels show the limitations of GaN and BN as host for UV applications of rare-earth
elements and importance of AlN host in such applications. Possible explanation for the
observation of these transitions is the transparency of AlN to most of the UV region. The
wide bandgap of 6.2 eV (190 nm) of AlN allow all transitions with wavelength higher
than 190 nm. In case of GaN and BN as hosts used for Pr we do not see these transitions
because the bandgap of these materials occur either in the vicinity of these wavelengths
or at higher wavelength than the wavelength of the emitted transitions. The big gap
between 190 nm of AlN and 335 nm or 385 nm has got an edge over the other two nitride
semiconductors in this case. Figure 9.15 is giving a comparison of Pr luminescence at
these wavelengths in AlN host with GaN and BN hosts.

167
700

AlN:1Ho1W
AlN:1Ho3W

549 nm Ho

+3

AlN:1Ho1Wthermally annealed
600

500

461 nm

Intensity (a.u)

+3

400

702 nm
Cr

300

+3

661 nm
Ho

+3

200

100

0
198

248

298

348

398

448

498

548

598

648

Wavelength (nm)

Figure 9.12.

Energy levels shift in W due to Ho addition.

698

748

168

8000

7000

4
AlN:Pr

GaN:Pr

6000

BN:Pr

Intensity (a.u)

5000

4000

3000

10 11

2000

5
7

1000

10
8

GaN:P
r
AlN:P
r
BN:Pr

11

11

830

805

779

754

728

703

677

652

626

601

575

550

524

499

473

448

422

397

371

346

320

Wavelength (nm)

Figure 9.13.

Comparison of AlN, GaN and BN hosts for the observation of 335 nm and
385 nm peaks in Pr.

169

The determination of stopping power of AlN for electrons using a bilayer of


AlN:Tm/AlN:Ho and Cathodoluminescence is another exciting result of this work. A
small percent difference of 6.6 % in the theoretical value and our experimental results
gives the accuracy of our experiment and method used for the determination of stopping
power of AlN. Other materials can also be investigated for their stopping powers using
this method.

9.2. Conclusion.
In the present work we have studied the optical and luminescence properties and
surface characterization of wide bandgap III-Nitride semiconductors, AlN, GaN and BN,
doped with rare-earth rare-earth elements Ho, Gd, Pr, Tm and Sm and transition metals
W and Y. RF Magnetron Sputtering, Cathodoluminescence, Thermal Activation,
Scanning Electron Microscopy and X-rays diffraction are used for the growth and
characterization. Films were deposited using rf magnetron sputtering on silicon
substrates. Cryogenically grown amorphous films were the principal focus of this
research. The substrates were cooled using liquid nitrogen during the growth and pure
amorphous films were obtained. Crystalline films were also obtained using an electric
heater to keep substrates at high temperature. Holmium films were also deposited on
optical fiber of thickness varying between 5 and 80 microns. Stripes of Ho, Pr, and W codoped with Gd were also deposited to study luminescence enhancement in these
materials. Scanning electron microscopy and X-rays diffraction analysis gave information
about the structure of the films.

170
Thermal activation of the films was performed in the flow of nitrogen at different
temperatures. The most suitable temperature for the activation of the investigated
materials was 900 C except for praseodymium. The suitable temperature for the
activation of praseodymium was found to be 1025 C. The holmium films deposited on
optical fiber were also annealed at 1000 C but the fibers melted at that temperature and
hence like holmium films deposited on flat silicon substrate, the films on fibers were also
activated at 900 C.
All samples were investigated for their luminescence and optical properties using
electron impact excitation. Emission spectra were obtained for all materials under
investigation and on the basis of recorded luminescence spectra emission lines for all
investigated RE ions were tabulated and assigned them to the most probable and well
know RE+3 ions 4f-4f transitions. Holmium, Thulium and Praseodymium were emitting in
a wide range from ultraviolet to infrared regions. The intense peak of holmium appears at
549 nm with other relatively low intense peaks at 362 nm, 394 nm, 459 nm, and 661 nm.
The dominant luminescence in praseodymium appears around 490 nm and 650 nm.
Thulium is found to be a good blue and IR emitter. Gadolinium was giving a sharp and
intense peak in ultraviolet. No peaks in visible or infrared were obtained for Gd.
Samarium was giving transitions in the low energetic portion of the visible spectrum with
a dominant emission in 602 nm. Both the transition metals were emitting either in blue,
violet or ultraviolet. Pr is deposited with different hosts and the luminescence spectra are
compared to see the effect of host material on its emission.
The most interesting part of this work is the luminescence enhancement in Ho, Pr,
Sm and W by the addition of Gd to these elements and by the geometry of the stripes.

171
Luminescence enhancement in all these materials is more than 100%. When Gd was
doped to Ho, the luminescence of Ho increased with the increasing concentration of Gd
until a saturation point reached and beyond that saturation, no more enhancement was
found with further increase of Gd concentration. Further it was not only the Gd ion taking
part in the luminescence enhancement but the geometry of the stripes was also playing its
role in this phenomenon.
The addition of Gd also enhances luminescence from W. Despite of the fact that
the exact transitions, responsible for the luminescence from W is unknown because of its
complicated energy levels distribution but still one can improve the intensity of light
coming out of W with the help of Gd. Further work needs to study this luminescence
enhancement in the investigated materials. People can exploit these ideas and results for
the construction of Solid State Laser in green from Ho, in blue and red from Pr, in red
from Sm and in blue from W.
Another exciting result is the confirmation of two additional levels in
praseodymium at 335 nm and 385 nm when deposited in AlN host. These levels are
already reported theoretically by E.van der Kolk et al

[64]

. The direct observations of

these levels show the limitations of GaN and BN as host for UV applications of rare-earth
elements and importance of AlN host in such applications.
Stopping power of AlN for electrons was determined experimentally. The results
were compared to the theoretical values obtained using the already established
mathematical equations. The beauty of the work is that the experimental results are very
precise and accurate with a small deviation of 6.6 % from theoretical results.

172

REFERENCES
1.

John B. Gruber et al, J. Appl. Phys., Vol. 91, No. 5, 2929-2935 (2002).

2.

H. J. Lozykowski, Phys. Rev. B 48, 17758 (1993).

3.

S. Schmitt-Rink et al, Phys. Rev. Lett. 66, 2782 (1991).

4.

M. A. J. Klik, I. Izeddin, J. Phillips and T. Gregorkiewicz, Materials Science and


Engineering, B105, 141-145 (2003).

5.

E. D. Rosa-Cruz, L. A. Diaz-Torres, P. Salas, R. A. Rodriguez, G. A. Kumar, M.


A. Meneses, J. F. Mosino, J. M. Hernandez and O. Barbosa-Garcia.

6.

M. L. Caldwell, A. L. Martin, V. I. Dimitrova, P. G. Van Patten, M. E.


Kordesch, and H. H. Richardson, Appl. Phys. Lett., Vol. 78, No. 9, 1246-1248
(2001).

7.

G. Ajithkumar, P. K. Gupta, G. Jose and N. V. Unnikrishnan, Journal of NonCrystalline Solids, 275, 93-106 (2000).

8.

D. S. Lee and A. J. Steckl, Appl. Phys. Lett., Vol. 80, No. 11, 1888-1890
(2002).

9.

D. S. Lee and A. J. Steckl, Appl. Phys. Lett., Vol. 80, No. 5, 728-730 (2002).
J. T. Torvik et al, J. Appl. Phys., 81 (9), 6343-6350 (1997).

10.

Hadis Morkoc, Nitride Semiconductors and Devices, Springer, Berlin Germany,


1999.

11.

Muhammad Maqbool, H. H. Richardson and M. E. Kordesch, Mater. Res. Soc.


Symp. Proc. Vol. 831 Article E8.12.1, @2005 Materials Research Society.

12

M. Maqbool, H.H. Richardson, P.G. Van Patten and M.E. Kordesch, Mat. Res.
Soc. Symp. Proc. Vol. 798, pp 8.5.1-8.5.5, 2004 Materials Research Society.

173
13.

B. Monemar, , Luminescence in III-nitrides, Material Sciences and Engineering,


B59, 122 (1999).

14.

J. F. Philipps, T. Topfer, H. E. Heidepriem, D. Ehrt and R. Sauerbrey, Appl.


Phys. B 72, 399-405 (2001).

15.

T. Forster, Z. Naturforsch. 4a, 321 (1949).

16.

D. Dexter, J. Chem. Phys. 21, 836 (1953).

17.

A. Burshtein, Sov. Phys., JETP 35, 882 (1972).

18.

J. Caird, A. Ramponi, P. Staver, J. Opt. Soc. Am. B 8, 1391 (1991).

19.

T. Forster, Ann. Phys. 2, 55 (1948).

20.

Y. Mita, H. Yamamoto, K. Katayanagi and S. Shionoya, J. Appl. Phys. 78 (2),


1219-1223 (1995).

21.

J. M. Zavada, R. A. Mair, R. G. Wilson and P. A. Grudowski, Appl. Phys. Lett.


75, 790 (1999).

22.

Michael E. Levinshtein, S. L. Rumyantsev and M. S. Shur, Properties of


Advanced Semiconductor Materials, John Wiley & Sons, Inc. 2001.

23.

M. L. Caldwell, A. L. Martin, V. I. Dimitrova, P.G. Van Patten, M. E. Kordesch


and H. H. Richardson, Appl. Phys. Lett. 78, 1246 (2001).

24.

S. C. Jain, M. Willander, J. Narayan, R. Van Overstraeten, J. Appl. Phys. 87, 965


(2000).

25.

W. J. Fan, M. F. Li, T. C. Chong, and J. B. Xia, J. Appl. Phys. 79, 188 (1996).

26.

H. Chen, K. Chen, D. A. Drabold and M. E. Kordesch, App. Phys. Lett. 77, 1117
(2000).

27.

A. A. Oliner: Acoustic Surface Waves, Topics Appl. Phys., Vol. 24 (Springer,


Berlin, ZHeidelberg 1978).

174

28.

J. Edwards, K. Kawabe, G. Stevans, R. H. Tredgold: Solid State Commun. 3, 99


(1965).

29.

W. M. Yim, E.J. Stokfo, P.J. Zanzucchi, J.I. Pankove, M.Ettenberg, and


S.L.Gilbert: J. Appl. Phys., 44, 292 (1973).

30.

S.Yoshida, S.Misawa, Y.Fujii, S.Takada, H.Hayakawa, S.Gonda, and A.Itoh: J.


Vac. Sci. Technol. 16, 990 (1979).

31.

K.M.Taylor, C.Lenie: J. Electrochem. Soc. 107, 308 (1960).

32.

Brozek, T., J. Szmidt, A. Jakubowski, A. Olszyna, Diamond Relat. Mater. 3, 720-724


(1994).

33.

Lopatin, V.V. Properties of Group III Nitrides. Ed. by Edgar, J.H., EMIS
Datareview Series N11INSPEC 1994.

34.

Litvinov, D., A. Charles, C.A. Taylor II, R. Clarke, Diamond Relat. Mater. 7,
360-364 (1998).

35.

T. Inushima et al., J. Crystal Growth, 227-228, 481 (2001).

36.

V. Yu, Davydov et al., Phys. Stat. Sol. B 229, R1 (2002).

37.

J. Wu et al., Appl. Phys. Lett. 80, 3967 (2002).

38.

Green J. F., T. K. Bolland, and J. W. Bolland, J. Chem. Phys. 64, 2(1976), 656662.

39.

Wasa Kiyotaka and Shigeru Hayakawa, Handbook of Sputtering Deposition


Technology: principles, technology and applications, Noyes Publications, Park
Ridge, New jersey, U.S.A. 1992.

40.

R. W. Berry, P. M. Hall, M. T. Harris, Thin film technology, D. van Nostrand


Inc., Princeton, 1968.

175
41.

D. L. Smith, Thin Film Deposition: Principle & Practice, McGraw Hill, New
York, 1995.

42.

B. N. Chapman, Glow Discharge Process, Wiley, New York, 1980.

43.

A. V. Kiryanov, V. Aboites, A. M. Belovolov, M. J. Damzen, A. Minassian, M. I.


Timoshechkin and M. I. Belovolov, Journal of Luminescence, 102-103, 715-721
(2003).

44.

Ki-See Lim, P. Babu, Sun-Kyun Lee, Van-Thai Pham and D. S. Hamilton,


Journal of Luminescence, 102-103, 737-743(2003).

45.

J. B. Gruber, U. Vetter, H. Hofsass, B. Zandi and M. F. Reid, Physical Review B,


Vol. 69, 195202 (2004).

46.

U. Vetter, J. Zenneck and H. Hofsass, Appl. Phys. Lett. Vol. 83, No. 11, 21452147 (2003).

47.

R. Birkhahn, M. Garter and A. J. Steckl, Appl. Phys. Lett. Vol. 74, No. 15, 21612163 (1999).

48.

V. K. Tikhomirov, K. Lakoubovskii, P. W. Hertogen and G. J. Adriaenssens,


Appl. Phys. Lett. 71 (19), 2740-2742 (1997).

49.

H. Chen, R. Lian, M. Yin, L. Lou, W. Zhang, S. Xia and J. C. Krupa, J. Phys.:


Condens. Matter. Vol. 13, 1151-1158 (2001).

50.

M. Fleuster, C. Buchal, E. Snoeks and A. Polman, Appl. Phys. Let. Vol. 65 (2),
225-227 (1994).

51.

A. A. Andreev, Physics of the Solid State, Vol. 44, No. 2, 248-254 (2002).

52.

S. Kim, S. J. Rhee, D. A. Rurnbull, X. Li, J. J. Coleman, S. G. Bishop and P. B.


Klein, Appl. Phys. Lett. 71 (18), 2662-2664 (1997).

176
53.

B. Karunagaran, R. T. Rajendra Kumar, D. Mangalataj, S. K. Narayandass and G.


M. Rao, Cryst. Res. Technol. Vol. 37 (12), 1285-1292 (2002).

54.

D. S. Lee and A. J. Steckl, Appl. Phys. Lett. Vol. 83, No. 11, 2094-2096 (2003).

55.

B. M. Walsh, N. P. Barnes and B. D. Bartolo, J. Appl. Phys., Vol. 83, No. 5,


2772-2787 (1998).

56.

H. J. Lozykowski, W. M. Jadwisienczak and I. Brown, Solid State


Communications, 110 (5), 253 (1999).

57.

E. D. Rosa-Cruz, L. A. Diaz-Torres, P. Salas, R. A. Rodriguez, G. A. Kumar, M.


A. Meneses, J. F. Mosino, J. M. Hernandez and O. B. Barbosa-Garcia, J. Appl.
Phys. Vol. 94. No. 5, 3509-3515 (2003).

58.

James E. Turner, Atoms, Radiation and Radiation Protection Second Edition, pp


122-138, 1995, John Wiley & Sons, New York.

59.

P. Marmier and E. Sheldon, Physics of Nuclei and Particles Vol. 1, pp 156-171,


1969, Academic Press Inc. New York.

60.

P. A. Budni, C. R. Ibach, S. D. Setzler, E. J. Gustafson, R.T. Castro and E. P.


Chicklis, Optics Letters, Vol. 28, No. 12, 1016-1018 (2003).

61.

W. J. Wang, S. Huang, L. Lu, W. M. Yen, A. M. Srivastava and A. A. Setlur,


Optics Communication, Vol. 195, 405-410 (2001).

62.

Caldwell, M. L.; Van Patten, P. G.; Kordesch, M.E.; Richardson, H. H., Visible
Luminescent Activation of Amorphous AlN:Eu Thin-Film Phosphors with
Oxygen, Internet Journal of Nitride Semiconductors, Vol. 6, Article 13, 2001.

63.

J. M. Zavada, R. A. Mair, C. J. Ellis, J. Y. Lin, H. X. Jiang, R. G. Wilson, P. A.


Grudowski and R. D. Dupuis, Appl. Phys. Lett., Vol. 75, No. 6, 790-792 (1999).

177
64.

E. van der Kolk, P. Dorenbos, C. W. E. van Eijk, Optical Communication, 197,


317-326 (2001).

65.

S.V. Frolov, A. Fujii, D. Chinn, Z.V. Vardeny, K. Yoshino and R.V. Gregory,
Appl. Phys. Lett. Vol. 72, 2811-2813 (1998).

66.

S.V. Frolov, M. Shkunov, Z.V. Vardeny, and K. Yoshino, Phys. Review B, Vol.
56, No. 8, R4363-R4366 (1997).

178

APPENDIX
The structure and morphology of films deposited around the optical fibers were examined
using scanning electron microscopy. Cross section images of the fiber surrounded by the
deposited film are taken. These pictures are shown in figure 4.1 and 4.2. The figures
show the existence of grain boundaries on the deposited films around fibers and hence
indicates that films with thickness greater than ~1 m on fiber were crystalline even
when deposited in the presence of liquid nitrogen. We are still not sure about the exact
temperature of the tip of the fiber during the film growth. The radial lines on the
deposited film around fiber in figure 4.2 demonstrate the crystalline structure of the film.
The crystalline structure of the film suggests that the liquid nitrogen cooling and clamp
geometry is insufficient for cooling the fiber during deposition, since the crystalline
structure was present immediately when the film was deposited without cooling. We did
not test films of thickness smaller than 1 m deposited on optical fiber for amorphous or
crystalline structure because we were trying to construct cylindrical and ring micro-laser
in whispering gallery mode and the film thickness to construct this kind of laser should
be greater than 2 m

[43,44]

. We could not deposit amorphous film of thickness greater

than 2 m around the optical fiber to construct cylindrical micro-laser.

179

Figure A1. Top.


SEM picture of the cross-section of AlN:Ho film around fiber.
Bottom. Picture of the same fiber from a side.

180

Figure A2.

Cross-sectional images of AlN:Ho films deposited around fiber.