Title: Growth and characterization of green electroluminescent thin films
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100829/00001
 Material Information
Title: Growth and characterization of green electroluminescent thin films
Physical Description: Book
Language: English
Creator: Feng, Tao
Publisher: University of Florida
Place of Publication: Gainesville Fla
Gainesville, Fla
Publication Date: 2001
Copyright Date: 2001
 Subjects
Subject: Materials Science and Engineering thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Materials Science and Engineering -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Summary: ABSTRACT: Growth and characterization of Pulsed laser deposited and sputter deposited green Zn2GeO4:Mn and electron beam evaporated Zn1-xMgxS:Mn thin films were investigated. Experimental results show that electroluminescent brightness of Zn2GeO4:Mn grown by sputter deposition was independent of working pressure and Ar/O2 ratio. Green photo- and electroluminescence were obtained from Zn2GeO4:Mn films deposited on Si (for photoluminescence only), glass/ITO/ATO and Al2O3/Au/Pb(ZrTi)O3 substrates with the emission wavelength at 540 nm and CIE color coordinate at x = 0.263 and y = 0.683. The Zn2GeO4:Mn electroluminescent film grown on a dielectric layer of Pb(ZrTi)O3 by pulsed laser deposition showed much higher brightness than that grown by sputter deposition (450 cd/m2 versus 120 cd/m2 at 2.5 kHz) primarily due to better film crystallinity resulting from higher substrate temperature (250 degrees C versus R.T.) and from surface damage resulting from high energy Ar plasma. Zn2GeO4:Mn deposited on Al2O3/Au/Pb(ZrTi)O3 substrate showed higher EL brightness (450 cd/m2 versus 45 cd/m2 at 2.5kHz) than that deposited on glass/ITO/ATO due to better crystallinity resulting from a longer annealing time, and higher breakdown electric field for the thicker Pb(ZrTi)O3 layer. The Zn2GeO4:Mn grown by pulsed laser deposition was Zn-deficient (Zn/Ge atomic ratio = 0.83 and 0.77 at a substrate temperature of 250 degrees C and 800 degrees C). A reaction of Zn2GeO4:Mn and Pb(ZrTi)O3 was observed upon annealing at temperature over 800 degrees C. Optimum annealing for Zn2GeO4:Mn on Pb(ZrTi)O3 was obtained using a temperature of 700 degrees C for 5 hours.
Summary: ABSTRACT (cont.): The Zn2GeO4:Mn grown on Pb(ZrTi)O3 substrate showed extremely poor EL brightness when the film thickness was smaller than 5000 Angstroms. This was hypothesized to be due to a reduced voltage drop across the phosphor layer because Zn2GeO4:Mn was not continuous. In addition, a large leakage current resulted from the rough Zn2GeO4:Mn/Pb(ZrTi)O3 interface. By partially substituting Mg for Zn, Zn1-xMgxS:Mn electroluminescent emission showed a green shift compared with the yellow/orange of ZnS:Mn due to crystal field reduction. Green photo- and electroluminescence were obtained from Zn2GeO4:Mn film deposited on glass/ITO/ATO substrate with the emission wavelength at 578 nm and CIE color coordinates at x = 0.5049, and y = 0.4900 versus ZnS:Mn emission at 592 nm, and the CIE coordinates at x = 0.5451, y = 0.4529 ). Film doped from a MnS source-doping exhibited EL brightness of 40 cd/m2 for 6700 Angstroms while those doped from metallic Mn source exhibit 90 cd/m2 for 13500 Angstroms both at 60Hz. Optimum Mn doping of 0.12% in the ZnMgS:Mn film was determined using SIMS. Using cross section TEM, columnar ZnMgS crystallites were observed in Zn0.8Mg0.2S:Mn.
Summary: KEYWORDS: electroluminescence, oxide phosphors, sulfide phosphors, thin film, zinc germanate, zinc sulfide
Thesis: Thesis (Ph. D.)--University of Florida, 2001.
Bibliography: Includes bibliographical references (p. 187-192).
System Details: System requirements: World Wide Web browser and PDF reader.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Tao Feng.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains ix, 193 p.; also contains graphics.
General Note: Vita.
 Record Information
Bibliographic ID: UF00100829
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 48522826
alephbibnum - 002763576
notis - ANP1598

Downloads

This item has the following downloads:

master5 ( PDF )


Full Text













GROWTH AND CHARACTERIZATION OF
GREEN ELECTROLUMINESCENT THIN FILMS

















By

TAO FENG


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2001































Copyright 2001

by

Tao Feng















ACKNOWLEDGMENTS

First of all, I am greatly indebted to my adviser and my committee chairman Dr.

P. H. Holloway, who gave me guidance and support. I feel fortunate to be guided by Dr.

Holloway. Working for Dr. Holloway was the richest three years in my life. I learned

immensely from him.

It is also my great honor to have four other distinguished professors ( Dr. Robert

Dehoff, Dr. Kevin Jones, Dr. Rolf Hummel and Dr. Fan Ren ) as my committee

members.

In addition, I thank Dr. Mark Davidson, who offered me help and advice with

every detail of my project. Regarding pulsed laser deposition, I am truly thankful to Dr.

Rajiv Singh and Dr. D. Kumar at the University of Florida, and Dr. Jim Fitz-Gerald and

Dr. Heungsoo Kim at Naval Research Lab.

For sample analyses, I am especially grateful to Dr. Maggie Puga-Lambers on

SIMS, Wayne Acree on EDX/SEM, Dr. Michael Kaufmann on XTEM, and Eric Lambers

on XPS.

I also appreciate the help and input my colleagues in Dr. Holloway's group

offered to me. I especially thank Joohan Kim for sputtering deposition, Jay Lewis and

Jong-pyo Kim for electrical measurement and Bill Glass for optical characterization.

Finally, I am indebted immeasurably to my parents who gave me love and support

all the time.
















TABLE OF CONTENTS

CHAPTERS page

A C K N O W L E D G M E N T S ............................................................................ ...............iii

ABSTRACT ..................... ................................... .............. ..... .......... viii

1 IN TRODU CTION .................................................. .. ................. 1

2 LITER A TU R E R EV IEW ................................................................ ....................... 5

2 .1 In tro du action ............................................... .............. ....................... 5
2.2 ACTFEL Device Structures............................. ..................... .............. 6
2.3 Optical and Electrical Characterization of ACTFEL Devices............................. 7
2 .3 .1 B rig h tn e ss ................................ ........................................ 7
2.3.2 Chromaticity .......................................... ....... ........... 9
2.3.3 Electrical Behavior Ideal Circuit Model .............................................. 11
2.4 Device Physics......................................... ............ 14
2.4.1 E lectron Injection....................................... ... ................ ...... ...... .... .. 15
2.4.2 Charge Transport....................................................... 16
2.4.3 Impact Excitation and Ionization....................... ..................................... 17
2.4.4 R adiative D ecay ............. .......... ........................ ..... ........................ ...... 18
2.4.5 O ptical O utcoupling .................................................. ......................... 19
2.5 A C TFE L M materials ................................................... ................................. 20
2.5.1 Substrates ............................ ........... .................... 20
2 .5 .2 E lectrodes ......... ..... ...... ......................................................... ......... ...... 2 1
2 .5.2 .1 T ransparent E lectrodes................................................... ... ................. 2 1
2.5.2.2 O paque Electrodes .............. ............................................ .............. 22
2.5.3 Insulators ..................................... ........... .............. 22
2.5.4 Phosphors ................................................ ............ 25
2.5.4.1 H osts ........................ ........................ .................. 25
2 .5.4 .2 L um inescent C enters...................................................... .... .. .............. 26
2.5.4.3 H ost-C enter system s ........................................... ........... .............. 28
2.6 Zn2G eO4:M n A CTFEL D evices ........................................ ......... .............. 29
2.6.1 General Consideration of Oxide Phosphors ........ .................................. 29
2.6.2 Green Oxide ACTFEL Devices ...................... ..................................... 30
2.6.3 Zn2GeO4 Host Crystal Structure .............. ........................... ....... ..... 31
2.6.4 Pulsed Laser Deposition of Oxide Thin Films ........................................ 32
2.7 ZnM gS:M n A C TFEL D evices..................................................... ... ................. 34
2.7.1 ZnlixMgxS-based Thin Film Growth............ ........ .......... ......... 34









2.7.2 ZnS-M gS Crystal Structures ........................................ ........ .............. 35
2.7.3 ZnS-M gS Solubility ..................... ................................... 36
2.7.4 Crystal Field M odification of M n2 ..................................................... 37

3 EXPERIM ENTAL PROCEDURES ........................................ ......................... 40

3.1. Source Preparation ................................................................ .. 40
3.1.1 Preparation of Zn2GeO4:Mn Target................... .................... 40
3.1.2 Preparation of ZnilxMgxS:Mny Pellets .................................... ................. 44
3.2 Substrate Preparation .............. .................................................................... 45
3.2.1 G lass/ITO /A TO Substrates ........................................ ......... .............. 45
3.2.2 A1203/Au/Pb(ZrTi)03 Substrates ............................... ....................... 46
3 .3 T h in F ilm G row th ........................................................................ ..................... 4 7
3 .3 .1 V acuu m Sy stem s ............................................................................ .............. 4 7
3.3.2 Pulsed Laser Deposition of Zn2GeO4:Mn............................................. 48
3.3.3 RF Sputtering Deposition of Zn2GeO4:Mn................................................. 50
3.3.4 Electron Beam Evaporation of ZnMgS:Mn......................... ........... ... 53
3.3.5 Deposition of Conducting Layers ........ ............. ........... ............. 55
3.4 Thin Film H eat Treatm ent.................................................. ....................... 57
3.4.1 Lamp-based Rapid Thermal Annealing............................... .................... 58
3.4.2 B ox Furnace A nnealing ........................................... .......................... 58
3.4.3 In-situ V acuum A nnealing ........................................ .................. ...... 58
3.5 C characterizations ........................ ................ .......................... .. .......... .. 59
3.5.1 Optical and Electrical Characterizations ......... ......................... ... 59
3.5.1.1 C hrom aticity ..................................... ............ ........ .............. 59
3.5.1.2 Photoluminescence................ ..... .. ..... .......................... 60
3.5.1.3 Electroluminescence (Luminance vs. Voltage) ..................................... 62
3.5.1.4 Q -V and C-V C urves............... .......................... .............. ... 63
3.5.2 Microstructural and Chemical Characterizations................ .................... 67
3.5.2.1 X-ray Diffraction.................. ....... ..... ...................... 68
3.5.2.2 Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy 70
3.5.2.3 X-ray Photo-electron Spectroscopy ............................... ............. 72
3.5.2.4 Secondary Ion Mass Spectroscopy ................................. .............. 73
3.5.2.5 Transmission Electron Microscopy ............................................... 74

4 PULSED LASER DEPOSITED Zn2GeO4:Mn...............................................76

4.1 Introduction............................ ............... ..... 76
4.2 E xperim mental R results .................................................. ......................... ...... 77
4.2.1 Zn2G eO4:M n Target.................... .................................. .......................... 77
4.2.1.1 X -ray D iffraction................................................. ........... .............. 77
4.2.1.2 PL and PLE Spectra ..................................................... ................. 78
4.2.1.3 PL of Ablated Zn2GeO4:Mn Target ..................................... ............... 80
4 .2 .2 P Z T S u b state ......... .. ........................................................................... ...... 8 0
4.2.2.1 SEM ......................................... 80
4.2.2.2 X -ray D iffraction................................................. ........... .............. 81
4.2.2.3 C om position ... .... ..... ........ .... .... ............. ... .... ........... 82


v









4.2.3 Zn2GeO4:Mn Film Track-table .............................. .. .............. 84
4.2.4 Zn2GeO4:Mn on Si Substrate Temperature Issue ....................................... 84
4.2.4.1 Photoluminescence.................. ......... ........................... 86
4.2.4.2 X -ray D iffraction................................................. ........... .............. 89
4 .2 .4 .4 Surface T opography ....................................................... ... ................. 9 1
4.2.4.5 C om position .............................. ..................... ........ .. .. .......... .... 92
4.2.5 Zn2GeO4:Mn on Glass/ITO/ATO Substrate Issue .................................... 94
4.2.5.1 SEM ............ ............................ ............... 95
4.2.5.2 Com position ..................... .............. .... ......... .... .... ...... 95
4.2.6 Zn2GeO4 on glass/ITO/ATO Laser Energy Density Issue.......................... 96
4.2.6.1 Photoluminescence................ ..... .. ..... .......................... 98
4.2.6.2 Electroluminescence (Luminance vs. Voltage) ..................................... 98
4.2.6.3 X -ray D iffraction ................................................. ........... .............. 99
4.2.6.4 SEM ............................................ ......... 101
4.2.7 Zn2GeO4:Mn onPZT- Annealing Issue .............................................. 101
4.2.7.1 Annealing Tem perature vs. Tim e.......................................................... 102
4.2.7.2 Photoluminescence.................. ........ ........................... 103
4.2.7.3 Electroluminescence .................................. 104
4.2.7.4 X -ray D iffraction............................................... 106
4 .2 .7 .5 SE M ................... .................................. ................. 107
4.2.7.6 Charge vs. V oltage Curve .............. ...... ......................................... 108
4.2.7.7 Capacitance vs. V oltage Curve ........................................................ 110
4.2.8 Zn2GeO4:Mn on PZT Substrate Temperature Issue............................... 111
4.2.8.1 Photoluminescence................ ..... .. .... .......................... 112
4.2.8.2 Electroluminescence (Luminance vs. Voltage) .................................... 112
4.2.8.3 X -ray D iffraction............................................... ........... .............. 113
4.3 Sum m ary and D discussions .............. .......................................................... 114
4.3.1 Z n2G eO 4:M n T arget........................................................... .............. 114
4.3.2 PZ T Substrate ................... ...... .... .............. ........... .............. 117
4.3.3 Photoluminescence and Electroluminescence of Zn2GeO4:Mn Films......... 117
4.3.4 Zn2GeO4:Mn on Si Substrate Temperature Issue .................................... 117
4.3.5 Zn2GeO4:Mn on PZT Ambient 02 Pressure Issue ................................... 120
4.3.6 Zn2GeO4:Mn on Glass/ITO/ATO Laser Energy Density Issue ............... 121
4.3.7 Zn2GeO4:Mn on PZT Annealing Issue .............................................. 122
4.3.8 Zn2GeO4:Mn onPZT Substrate Temperature Issue................................ 123
4.3.9 Comparison of Zn2GeO4:Mn Deposition on PZT and Glass/ITO/ATO ...... 124
4.3.9.1 Electroluminescence (Luminance vs. Voltage) .................................... 124
4.3.9.2 X -ray D iffraction............................................... 124
4.3.9.3 SEM .......................................... 125
4.3.9.4 Capacitance ...... ..... .................................... ...... ............ .. 126
4.3.9.5 Sum m ary ................................. .... ..................... ......... 128
4.3.10 Zinc Deficiency of Zn2GeO4:M n Film s ..................................................... 128

5 SPUTTER DEPOSITED Zn2GeO4:M n ............................................ ............... 130

5.1 Parametrics of Zn2GeO4:Mn Sputtering Deposition.................................. 130
5.1.1 Deposition Repeatability ............................... .............. 130









5.2.2 A nnealing M mechanism s ........................................ .......................... 143
5.3 Sum m ary and Discussions ..................................... .................... 148
5.3.1 Annealing of Zn2GeO4:Mn on PZT .............. ..... .. ..................... 148
5.3.2 Comparison of PLD and SD Zn2GeO4:Mn Films ....................................... 150
5.3.3 Zinc Deficiency of Zn2GeO4:Mn Films ............................................... 156
5.3.4 Effect of PZT Roughness................................................ 158

6 Zn.ixMgxS:Mn ...................................... 164

6.1 Introduction ..... ........... .... ......... ...... ......... 165
6.2 Experim mental R results ................................................ ............................... 165
6.2.1 ZnilxM gxS:M n Source Pellet.................................................................... 167
6 .2 .2 M n2+ D o p in g ............................................................ .......... .................... 16 8
6.2.3 Optimization of M n in Evaporated ZnS:M n............................................... 170
6.2.4 Rapid Therm al Annealing of ZnS:M n....................................................... 173
6.2.5 Lattice Expansion of ZnilxMgxS:Mn..................................................... 174
6.2.6 Green Shift of Zn M g S:M n.................................................. .............. 175
6.2.7 Cross Section TEM......................................................... 178
6.3 Sum m ary and D discussions .............. .......................................................... 178
6.3.1 Z n M g S :M n Source......................................................................... ... 179
6.3.2 Optimum Mn Doping................................. .............. 180
6.3.3 Rapid Thermal Annealing.............................. .............. 181
6.3.4 Characteristics of ZnilMgxS:Mn Films.............................................. 181
6.3.5 M gS phase Segregation ....................................................... .......... ... 182

7 C O N C LU SIO N S ................... ........................................ 182

7.1 Pulsed Laser Deposited (PLD) Zn2GeO4:Mn ............................................... 182
7.2 RF Sputter Deposited (SD) Zn2GeO4:M n................................. .............. 184
7.3 Comparison of PLD and SD Zn2Ge4:Mn............................................... 185
7.4 Electron-beam Evaporated Zni-xMgxS:Mn...... .................................... 185

L IST O F R E FE R E N C E S ........................................................................ ...................187

BIOGRAPHICAL SKETCH .............. ............. .......... ..................... ............... 193















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

GROWTH AND CHARACTERIZATION OF
GREEN ELECTROLUMINESCENT THIN FILMS

By

Tao Feng

August 2001


Chairman: Dr. P. H. Holloway
Major Department: Materials Science and Engineering

Growth and characterization of Pulsed laser deposited and sputter deposited green

Zn2GeO4:Mn and electron beam evaporated Zn,. Mlu S:Mn thin films were investigated.

Experimental results show that electroluminescent brightness of Zn2GeO4:Mn

grown by sputter deposition was independent of working pressure and Ar/02 ratio. Green

photo- and electroluminescence were obtained from Zn2GeO4:Mn films deposited on Si

(for photoluminescence only), glass/ITO/ATO and A1203/Au/Pb(ZrTi)03 substrates with

the emission wavelength at 540 nm and CIE color coordinate at x = 0.263 and y = 0.683.

The Zn2GeO4:Mn electroluminescent film grown on a dielectric layer of Pb(ZrTi)O3 by

pulsed laser deposition showed much higher brightness than that grown by sputter

deposition (450 cd/m2 versus 120 cd/m2 at 2.5 kHz) primarily due to better film

crystallinity resulting from higher substrate temperature (2500C versus R.T.) and from

surface damage resulting from high energy Ar plasma. Zn2GeO4:Mn deposited on









A1203/Au/Pb(ZrTi)O3 substrate showed higher EL brightness (450 cd/m2 versus 45 cd/m2

at 2.5kHz) than that deposited on glass/ITO/ATO due to better crystallinity resulting

from a longer annealing time, and higher breakdown electric field for the thicker

Pb(ZrTi)O3 layer. The Zn2GeO4:Mn grown by pulsed laser deposition was Zn-deficient

(Zn/Ge atomic ratio = 0.83 and 0.77 at a substrate temperature of 2500C and 8000C). A

reaction of Zn2GeO4:Mn and Pb(ZrTi)O3 was observed upon annealing at temperature

over 8000C. Optimum annealing for Zn2GeO4:Mn on Pb(ZrTi)O3 was obtained using a

temperature of 7000C for 5 hours. The Zn2GeO4:Mn grown on Pb(ZrTi)O3 substrate

showed extremely poor EL brightness when the film thickness was smaller than 5000A.

This was hypothesized to be due to a reduced voltage drop across the phosphor layer

because Zn2GeO4:Mn was not continuous. In addition, a large leakage current resulted

from the rough Zn2GeO4:Mn/Pb(ZrTi)O3 interface.

By partially substituting Mg for Zn, Zni-xMgxS:Mn electroluminescent emission

showed a green shift compared with the yellow/orange of ZnS:Mn due to crystal field

reduction. Green photo- and electroluminescence were obtained from Zn2GeO4:Mn film

deposited on glass/ITO/ATO substrate with the emission wavelength at 578 nm and CIE

color coordinates at x = 0.5049, and y = 0.4900 versus ZnS:Mn emission at 592 nm, and

the CIE coordinates at x = 0.5451, y = 0.4529 ). Film doped from a MnS source-doping

exhibited EL brightness of 40 cd/m2 for 6700A while those doped from metallic Mn

source exhibit 90 cd/m2 for 13500A both at 60Hz. Optimum Mn doping of 0.12% in the

ZnMgS:Mn film was determined using SIMS. Using cross section TEM, columnar

ZnMgS crystallites were observed in Zno.sMgo.2S:Mn.














CHAPTER 1
INTRODUCTION

The history of phosphors has spanned nearly 400 years. Phosphor was first

observed in 1603.1 One of the first large-scale commercialization of a device using

phosphors was television tubes, which are also known as cathode ray tube (CRT).2 The

first practical electroluminescent display was reported by Sharp corporation in 1974.3

Use of electronic displays began with the development of the CRT. However, the CRT

had inherent disadvantages in its thick glass tube and large size (depth beyond 35"),

required for scanning of the electron beam across the phosphor screen. Driven by

portable electronics applications, flat panel display technologies are now under active

development. The flat panel display industry accounts for almost 50% of the total display

market and is expected to increase to 63% of a more than $100 billion market by the year

2005.4 Included among flat panel technologies are liquid crystal display (LCD),5 plasma

display panel (PDP),6 inorganic light emitting diode (LED),7 organic LED (OLED),8 also

known as organic electroluminescent display (OEL) and field emission display (FED),9

laser-based diode display (LDD) and electroluminescent (EL) display. 10 Currently, the

flat panel display market is dominated by LCDs and PDPs are just reaching the market as

mass produced product, even though it is clear that neither currently meets the demanding

economic and performance requirements for this application, leaving the market open to

still-developing technologies. Only EL displays are discussed in detail. The reader is

referred to several other reviews that discuss other displays for more information.11, 12,

13 14









Electroluminescence in ZnS:Cu was first reported by Gudden and Pohl in 1920. 3

Since then, electroluminescent devices were developed from powder to thin film, from

DC to AC, from inorganic to organic, and from monochrome to multicolor. The first

electroluminescent display using ZnS:Mn phosphor was developed by Sharp in 1974.15'16

Currently, the leading companies to develop EL displays are Sharp, Planar Systems and

Tektronix. The alternating-current thin-film electroluminescence (ACTFEL) display has

found its place in the market region that relies on its wide operation temperature range

from -400 to 1500C, as well as its wide view angle and large contrast ratio and

ruggedness.1 Its applications include medical, automotive, industrial and military.4 The

advantage of the wide operation temperature range is especially attractive for military

applications. Power consumption and operation voltage of these displays is higher than

the available portable electronics. The largest limitations for the technology are the lack

of a full-color, red-green-blue (RGB) product and the low brightness. The sensitivity of

human eyes to green is much lower than red and the relative brightness of R, G, B for a

good white is 1:4:1. 10 17 So far, the best red brightness is obtained from filtered yellow

ZnS:Mn, which is 65 cd/m2 at 60 Hz drive frequency.10 In order to match this best red

brightness, 260 cd/m2 of the green brightness is required for a white. However, the best

green phosphor available is ZnS:Tb, which gives 100 cd/m2 at 60 Hz drive frequency.

Therefore, further developments in green-emitting ACTFEL phosphors are desired. There

is a need to seek new green candidates to improve brightness and stability. In this

dissertation, two phosphors used for green ACTFEL devices were studied: one oxide

(Zn2GeO4:Mn) and one sulfide (Zni-xMgxS:Mn).









Chapter 2 presents a background of ACTFEL devices and a general understanding

of the phosphor thin film growth techniques and the device operation. In Chapter 3,

experimental procedures are presented that cover thin film growth, heat treatment and

characterization. In Chapter 4, characterization results are shown for the first

Zn2GeO4:Mn ACTFEL devices fabricated using pulsed laser deposition. Results show a

higher brightness in the Zn2GeO4:Mn ACTFEL devices deposited on Pb(ZrTi)03

compared with those deposited on glass/ITO/ATO, resulting from improved phosphor

film crystallinity due to much longer allowed annealing times. In Chapter 5, the film

growth optimization and characterization results of Zn2GeO4:Mn ACTFEL devices

produced using RF sputtering deposition are presented, which shows that the

performance of Zn2GeO4:Mn ACTFEL devices is insensitive to the growth parameters of

base pressure, working pressure and Ar/02 ratio over a large range. The Zn2GeO4:Mn

ACTFEL devices fabricated using the two previously mentioned deposition techniques

are shown to exhibit green emission at a broad peak around 540 nm. In Chapter 6, the

important results are discussed showing a dramatically better performance in the pulsed

laser deposited Zn2GeO4:Mn ACTFEL device compared with the sputter deposited

device. This improved performance is due to the less surface damage and a higher

substrate temperature used in PLD processing, both of which lead to improved

crystallinity. I also discuss the substantially reduced performance of Zn2GeO4:Mn

ACTFEL devices on annealing at 8000C. Degradation of device performance in this case

results from the reaction of Zn2GeO4:Mn with its underlying layer Pb(ZrTi)03. In

Chapter 7, an investigation into electron-beam (e-beam) evaporated ZnMgS:Mn

ACTFEL devices is presented showing that use of a MnS doping source gave better






4


performance than a metal Mn doping source. Further, the best performance was found at

optimum Mn2+ 0.12% in film and MgS phase segregation in the film was observed using

XTEM. Chapter 8 gives conclusions.














CHAPTER 2
LITERATURE REVIEW


2.1 Introduction

The definition of electroluminescence (EL) is the non-thermal generation of light

resulting from the application of an electric field to a substance,16 usually a

semiconductor. The electroluminescence discussed in this dissertation is different from

that of inorganic semiconductor-based light emitting diodes (LEDs), diode lasers or

organic electroluminescence (OEL). Excitation of a luminescent center in the present

type of EL involves ballistic electron transport in an inorganic host rather than the

recombination of electron-hole pairs in a p-n junction. To be more specific, the

electroluminescence discussed here is thin film alternating electroluminescence, rather

than powder electroluminescence or DC electroluminescence, each of which works by a

different mechanism.16

This chapter covers information from the fundamentals of the EL device to the

literature review of the two green phosphors Zn2GeO4:Mn and ZnMgS:Mn studied in this

dissertation. The fundamentals of the EL device include the device structures, device

operation physics and device materials. The literature review of Zn2GeO4:Mn ACTFEL

devices focuses on the pulsed laser deposition and the sputtering deposition of available

oxide phosphors. The literature review of ZnMgS:Mn concentrates on the results of the

electron-beam (e-beam) evaporated ZnMgS:Mn, the issue of ZnS-MgS solubility, and

the mechanism of crystal field modification due to Mg substitution for Zn.









2.2 ACTFEL Device Structures

Depending on whether a device emits a single color or multiple colors, ACTFEL

devices can be categorized into two types: monochrome devices and multi-color devices.

In this dissertation, only monochrome green devices were involved. A complete

ACTFEL device generally has a multi-layer structure of substrate/bottom conducting

layer/bottom dielectric layer/phosphor layer/top dielectric layer /top conducting layer.

Sometimes, this ACTFEL device structure is also referred to as MISIM (metal-insulator-

semiconductor-insulator-metal). The most important layer of all these layers is the

phosphor layer, which is sandwiched by two (isolated) conducting layers at each side to

allow a voltage to be applied across it. For the alternating current thin film

electroluminescent devices, the dielectric layer(s) is required to limit the current passing

through the phosphor layer. Each of them is inserted between the phosphor layer and the

conducting layer. Depending on which side is transparent, an EL device structure can be

categorized into two: the standard device structure and the inverted device structure as

shown in Figure 2-1. When the substrate and the two bottom layers are transparent, the

light emitted from the phosphor comes out from the substrate side and the viewing

direction is from the phosphor layer to the substrate side. In this research, the standard

device structure is glass/ITO/ATO/phosphor layer/BTO/Al. Glass is used as a transparent

substrate, ITO is a transparent bottom conducting layer, ATO is a transparent bottom

dielectric layer, BTO is a transparent top dielectric layer, and opaque metal Al is bottom

conducting layer, which also acts as an optical reflector. If any of the bottom layers are

not transparent, the top layers are required to be transparent. This latter structure is called

the inverted structure and the emitted light is viewed from the top layers. In this research,

the inverted device structure is A1203/Au/Pb(ZrTi)03/phosphor layer/BTO/ITO. Since










the opaque A1203 substrate together with Au conducting layer and Pb(ZrTi)O3 dielectric

layer is not transparent, the transparent dielectric layer BTO and transparent conducting

layer ITO are used for the top layers. In either of the two structures, the layers at one side

of the phosphor layer are transparent and the layers at the other side are opaque so that

the light can be viewed at the transparent side. So-called "half-cell" structures,

containing only one dielectric layer, are typically used for research. The general rule is

that brightness in the "half-cell" structure device is half of that in the "full-cell" complete

structure device.





Transparent conducting Layer
TranL,-arent ii- "triic Laver
La%,erF hocsrhor L ver
Phosphor Laver
~ Lielec tric Lacer
Transparent Di ectric Layer
Transparent Con ting Layer
Substrate
Transparent Glass bstrate


a) b)



Figure 2-1 ACTFEL device structures: a) Standard and b) Inverted16




2.3 Optical and Electrical Characterization of ACTFEL Devices



2.3.1 Brightness

An ACTFEL device is operated by applying high voltage pulses of alternating

polarity. Any waveform of alternating polarity may be used, including sine waves,

triangle waves, square pulses, and trapezoidal pulses, with the applied voltage specified









as the maximum voltage per pulse, rather than the RMS value.18 Frequency for the

alternating polarity may vary from 60 Hz to from 1 to 2.5 kHz. Generally, there is a

voltage threshold above which a significant amount of light is produced. The plot of

brightness as a function of voltage is known as a B-V curve. Devices often are compared

by the value of brightness at 40 V above this threshold, and this value is called the B40

value. These data are generally plotted on a linear scale. Typical data from an ACTFEL

device is shown in Figure 2-2. There are several common methods for determining the

threshold voltage. The most frequently used method involves finding the highest slope of

the B-V curve, then extrapolating this slope back to the voltage axis. This has the

advantage of defining the portion of the B-V curve relative to the voltage at which rapid

turn-on occurs.



Luminance











Vth Voltage
Vth


Figure 2-2 Typical B-V curve of ACTFEL devices showing the extrapolation method
for determining Vth


Finally, electrical characteristics are often used to define threshold. This is done most

often with Q-V and C-V curves, which are discussed in Section 3.6.1.4. The electrical










turn-on obtained from Q-V often corresponds closely with the optical turn-on obtained

from B-V.



2.3.2 Chromaticity

Quantitative methods for evaluating chromaticity and colorimetrics were

developed that are standard to the entire display community. These methods involve

relating the spectral distribution of the phosphor to values that correspond to the color

seen by an observer. The most commonly used standard is that established by

Commission Internationale de l'Eclairage (CIE), which defines chromaticity using three

unitless values (x, y and z CIE coordinates) that can be plotted on a two-dimensional

graph (since z = 1-x-y), known as the CIE diagram. These plots offer a convenient

quantitative comparison of the color of different phosphors.19 Under the additive mixing

rule, any color can be reproduced by adding the three primary colors (tristimulus values.)

The CIE system uses monochromatic sources as the primary colors for blue (k=

435.8cm), green (k= 546.1nm) and red (k= 700nm). The R, G and B tristimulus cuves are

denoted as x, y and z, and are shown in Figure 2-3. Color is determined by calculating X,

Y and Z values, such that

780
X =Km fe,,x()ddA
380

780
Y = Km (AY )dA
380

780

380










where Km is the luminous efficiency coefficient which is 680 Im/W, DeX is the spectral

output of the source, X is the wavelength and x(l), y(l) and z(l) are the spectral

response of the standard observer for each of the three primary colors: red, green and

blue.


I .-, .. __ _.. _





1.

0.4-



1


', ,.. I,. .i,-, (nm )

Figure 2-3 Red (x), green (y), and blue (z) tristimulus curves19



The chromatic coordinates are then found from

X
x -
X+Y+Z

Y
y-
y X+Y+Z


z-
X+Y+Z

Two chromaticity values are traditionally expressed as x and y. The x and y values are

plotted on the CIE diagram, shown in Figure 2-4. The curved perimeter of the CIE

diagram, which spans from blue to red, is defined by single wavelengths and represents

completely saturated colors. Any broadening of the emission spectrum results in








coordinates inside these boundaries and toward the center or "white" region of the
diagram.


I-,


.L
YELLOWIS
GREEN


GREEN


.900


.800

.700

.600

y .500oo

.400


.300

.200

.100


000


-i =


G EENIS





.100 .200 .300 .400 .500 .600 .700 .800

X


Figure 2-4 1931 Commission de L' Eclariage chromaticity diagram16


2.3.3 Electrical Behavior Ideal Circuit Model
Interpretation of the electrical characteristics of an ACTFEL device is aided by
considering an equivalent circuit model. Below threshold, the device can be modeled


i "---




' tZ


UII









accurately by considering the insulators and the phosphor layer as series capacitors. The

two insulators can be combined numerically by treating them as series capacitors and can

be considered as a single ideal capacitor. At a certain threshold voltages, rapid turn-on

occurs where a real current flow through the phosphor layer, sometimes called

conduction or dissipative current (as opposed to the charging current which charges the

capacitors and is discharged on removal of the applied voltage). In an ideal device, the

dissipative current is stored until an opposing pulse is applied and the current flows in the

opposite direction. Using the procedure described by Ono,16 the device can be modeled

as a circuit where the phosphor layer is modeled as a parallel capacitor and nonlinear

resistor. The nonlinear resistor has the current-voltage characteristics shown in Figure 2-

5. The nonlinear resistor often is modeled as a set of back-to-back Zener diodes. The

derivation of the following characteristics is given in more detail by Ono,16 but results of

the circuit analysis are presented here. Below threshold, the applied voltage is distributed

capacitively by the phosphor and the insulators giving


VP- C V (2-1)
C1 + CP


V= C V (2-2)
C1 + CP

where Va is the applied voltage, VI and Vp are the voltages across the insulator and the

phosphor layer, respectively. CI and Cp are the insulator and phosphor capacitance,

respectively. CI is the combination of Cil and Ci2, given by CilCi2/(Cil+Ci2) since they are

connected in series. When Cil = Ci2, which is the symmetric device case, CI = Ci/2.

Above the threshold, the resistive portion of the phosphor circuit allows current to flow

(conduction current) until the voltage across the phosphor returns to threshold. At the









same time, the conduction current must be balanced by capacitive current which will

charge the insulating capacitors to maintain the overall voltage across the device.

Therefore, the final voltages above threshold are

C
Vpf = VP,th = C th (2-3)
CI +CP

VIf = Va VP,f (2-4)

Here the concept of field clamping is apparent. Voltage across the phosphor layer in this

simple equivalent circuit model is maintained at the threshold voltage regardless of the

applied voltage. The amount of dissipative charge transferred through the phosphor, as

required to maintain the applied external voltage, is

Qex = 2C Vy,f = 2C (V, Vth)

and the interface charge density at the end of a pulse is half this value. (The circuit model

represents a perfectly symmetric device. Therefore, the other half of the transferred

charge neutralized the previous interface charge.) Using this value for transferred charge

along with the voltage transferred from the phosphor to the insulators, the power

consumption density is found to be

n = 2Qe"Vp,th = 4fC( (V, Vh)Vp,th (2-5)

wherefis the drive frequency. The luminous efficiency ris the brightness, L, divided by

the power consumption, given by

L L
P,, 4fC, (V -Vth)Vp,th



















Va Vp, = R



Ci2










V
-Vth Vth



Figure 2-5 Equivalent circuit for an ACTFEL device showing the current-voltage
characteristics of the non-linear resistor16




2.4 Device Physics

The physical processes of electroluminescence can be broken down into several

steps: (1) injection of electrons into the conduction band, (2) acceleration of a fraction of

the electrons to ballistic energies due to the applied field, (3) impact excitation of the

luminescent centers by electrons with sufficient energy, (4) radiative relaxation of the

centers (in competition with non-radiative relaxation), and (5) outcoupling of the

generated photons. These processes are shown schematically in Figure 2-6. Electrons are









generated at the interface between the phosphor layer and dielectric layer. The vertical

axis stands for the energy E. When a voltage drop is applied through the phosphor layer,

the band-gap is tilted from the cathode side of the phosphor layer to the anode side.

Electrons from the interface are accelerated under this electric potential, subsequently

exciting luminescent centers in the phosphor. An additional process which can occur is

(3b) impact ionization of the lattice or of defects by electrons, which is also shown in

Figure 2-6. In the following paragraphs, these processes will be discussed in more detail.



Cathode Dielectric Phosphor Layer Dielectric Anode
E C



1.4.









Figure 2-6 Electroluminescence process of ACTFEL device



2.4.1 Electron Injection

The source of charge carriers under steady state operation is generally accepted as

being primarily from interface states for ZnS:Mn devices.20 The depth and concentration

of available interface states has been shown to change significantly when the dielectric

was changed. When the local electric field at the cathode reaches threshold level, charge

injection takes place. Among the possible mechanisms responsible for injection of

electrons into the conduction band are thermionic emission, pure tunneling, and phonon-









assisted tunneling.21 At room temperature and for trap depths of 1.5eV, pure tunneling is

the dominant injection mechanism among the three. The expression for the emission rate

from pure tunneling is:

1 3
PT qf exp 44(2m ) E,2 A
e n I 3 qhfp ) ( )3]
4(2m E,,3 )2

where q is the electronic charge, fp is the electric field in the phosphor, m* refers to the

electron effective mass, E,1 is the interface trap depth, AE,a is the coulombic barrier

lowering and h is Planck's constant. If the electron injection is a result of the pure

tunneling effect, its emission rate should be independent of temperature. However, a

weak temperature dependence is observed.22 Therefore, the more accurate way to

explain the electron injection is the phonon-assistance tunneling which is a combination

of the first two mechanisms, where phonon interactions allow efficient tunneling.



2.4.2 Charge Transport

Charge transport in a high electric field, as is the case in ACTFEL devices, has

been studied extensively. A critical issue is the efficiency with which charge carriers are

able to gain sufficient energy to impact excite the luminescent centers without loss. Two

methods have been used to calculate the characteristics of high field electronic transport

in ZnS. The goal of the calculations is to determine the fraction of electrons that are able

to excite luminescent centers. One involves the use of band structure calculations and

Monte Carlo simulations, in which independent probabilities for scatting events are used

to determine the fate of a large number of carriers.3' 23, 24' 25 The second technique is the









lucky-drift transport model.26 This is an analytical model in which the mean free path for

electrons is analyzed with respect to scattering events.

In the first theory, the assumption is made that "all electrons are hot" at a high

field of-1MV/cm, and ballisitic or nearly loss-free acceleration was predicted.26 This

theory explained the avalanche breakdown observed in ZnS ACTFEL devices. In the

second theory, two transport modes are considered. One is ballistic transport, which is

collision-free. The other is the drift mode, which occurs after the electron has had one

collision, which increases the probability of subsequent collisions. This theory explained

the ACTFEL devices with poor crystallinity.26



2.4.3 Impact Excitation and Ionization

Impact excitation and ionization are two parallel processes that occur in the

phosphor device. The primary process involved in excitation in ZnS:Mn devices is direct

impact excitation,26 in which the energy threshold for the process is essentially the

transition energy for the luminescent center.26 The probability for this process is

determined by the impact cross section of the luminescent center, and the average impact

length for an electron is found from 11 = (oN)-1 where c is the impact cross section in

units of cm2 and N is the concentration of centers in units of cm-3. This cross-section has

been estimated experimentally by Mach and Mueller to be 4.0 x 10-16 cm2 (+/-20%) for

Mn2+ in ZnS, which is near the value of the ionic cross section.3 Another possible

excitation process is impact ionization followed by electron recapture and radiative

relaxation. This is believed to be the dominant process involved in excitation of Eu3+ in

ZnS. There may be other excitation mechanisms, such as energy transfer from electron-









hole pairs to luminescent centers,20 which has been suggested to account for excitation of

Tb3+. Direct impact ionization of the lattice can occur if electrons reach sufficient

energy. The energy threshold is normally thought to be either the bandgap of the host

material23 14 or 1.5x this value.27



2.4.4 Radiative Decay

When an impurity center is in an excited state, that energy will eventually be

dissipated. The relaxation process can occur by either 1) emission of a photon, b) non-

radiative relaxation, which can be in the form of emission of one or more phonons to the

lattice etc, or c) energy transfer to another center. For display applications, it is desirable

to maximize the first process and to minimize the second. The third case can be useful or

harmful in different situations.28 For a simple two-level system, the rate of return from

an excited state to the ground state is given by

dN =
dt NePg

where Ne is the number of luminescent ions in the excited state, t is the time and Peg is the

probability of spontaneous emission from the excited to the ground state. Integration

yields


N (t) = Ne (0) exp(- )
TR

where ZR (=Peg-') is the radiative decay time, and the exponential decay of intensity

results.28 An excited ion can relax through the emission of phonons by coupling to an

electromagnetic vibration or the lattice vibration. Because the probability of this process









is an Arrhenius dependence on temperature, the decrease in probability of photon

emission with temperature is known as thermal quenching.20



2.4.5 Optical Outcoupling

Optical outcoupling is dependent on the internal reflection angle. For a useful

device, generated photons must escape the device in the direction of the viewer, so the

concept of internal reflection angle must be considered. When light with incidence angle

01 passes from one media nl to the less dense media n2, the deflection angle 02 is

determined by the following equation:

n1 sin 08 = n2 sin 02

When the deflection angle 02 reaches its maximum 900, no deflection will occur and all

the light will be reflected and confined in the media 1. Since in most cases the media 2 is

normally air or vacuum, the refraction index of n2 is assumed as 1.00. According to the

above equation, the critical incidence angle 0o is given by:


0, = arcsin( ) (2-6)
nli

When all the possible incidence angles lower than the critical angle are considered, the

light that can be deflected out of the phosphor layer is given by:

arsm( )

0
24



The optical outcoupling efficiency can be improved by increasing the surface roughness

and thereby reducing internal reflection. It has been shown that a rough surface can









increase outcoupling by more than on order of magnitude.29 However, too rough a

surface can reduce contrast because of increased diffuse scattering.16




2.5 ACTFEL Materials

As mentioned earlier, three different materials are involved in an ACTFEL

device: the phosphor materials, the dielectric materials (also called insulators), and the

conducting materials (also called electrodes). These three different materials, as well as

substrate, are discussed separately here according to their individual function.



2.5.1 Substrates

For a standard device, the primary requirements for the substrate are transparency,

smooth surface and the ability to withstand thermal treatments required by subsequent

layers. It is also desirable that the material be inexpensive. For these reasons, the most

common substrate material is Corning 7059 soda-lime-silicate glass. This is an alkali-

free, barium bororsolicate glass often used for LCD substrates (J-47). It has a softening

temperature of 6000C, and large panels can undergo rapid thermal anneals (RTAs) of up

to 6500C without significant damage or warping.30 For research purposes, smaller

samples of 2"x 2" in size may be subjected to RTAs up to 8500C without significant

damage. In addition Coming 7059 is alkali-free to prevent migration of such species into

the device. For higher temperatures, glass-ceramics are available which meet the

substrate requirements, but these materials are too expensive to be economically viable

substrate materials for commercial applications. Glass substrates were used for the

standard device structures in this research.









The substrate requirements for the inverted structure devices also include a

smooth surface and the ability to meet the thermal considerations for processing of

subsequent layers. For the inverted structure, the substrate could act as part of the device,

comprising the bottom insulator, and electrical properties must be considered. These are

discussed in Section 2.5.3. For devices with the structure in which the bottom electrode is

deposited on the top of the substrate, alumina substrates have also been used.31 For active

matrix applications, the on/off state of the pixels are controlled by transistors located

underneath each pixel, and the silicon-based IC array is used as the substrate.16 In this

research, opaque polycrystalline A1203 substrate was used for the inverted device

structure since it can withstand much higher processing temperature than glass.



2.5.2 Electrodes

Two kinds of electrode materials are used for ACTFEL devices: transparent

electrodes and opaque electrodes.



2.5.2.1 Transparent Electrodes

The transparent electrode layer is used to allow emitted light to be viewed. Three

critical materials requirements for transparent electrode are: 1) it must be sufficiently

conductive; 2) it much be as transparent as possible in the spectral region of the emission

from the device to minimize absorption and/or color perturbation; 3) it must withstand

the thermal processing of subsequent layers. The material most commonly used for this

application is indium-tin-oxide (ITO).16 This is normally an alloy of 90 wt% In203 and

5-10% wt% SnO2, sometimes doped with F. An ITO thickness of 200nm is typical.

Good transparency (90%) can be obtained at this thickness with a resistivity of lx 10-4









Q-cm, providing a resultant sheet resistance of -5/square. The primary cause for high

conductivity in ITO is oxygen deficiencies, since oxygen vacancies act as shallow donors

which are efficiently ionized at room temperature.32 Oxygen vacancy concentrations can

be affected by annealing at high temperature. Another transparent conductor commonly

used in the inverted structure is ZnO:Al.33



2.5.2.2 Opaque Electrodes

For the standard structure, the requirements for the opaque electrode are not

stringent because it is the last layer deposited. The critical materials requirements for

opaque electrodes are: 1) it must have sufficient conductivity, which for most metals is

far better than that of the transparent electrode, and 2) good adhesion to the top insulator

is also critical. Aluminum is the most common electrode material for several reasons.

First, it is inexpensive. Second, it is easily deposited by evaporation or sputtering. Third,

it has a low melting temperature and exhibits desirable wetting characteristics to many

common insulator materials, providing the possibility for self-healing breakdown by

fusing of the metal surrounding a short. The typical thickness for this layer is 100-200

nm. One tradeoff to be considered for Al as an electrode is its high reflectivity, which

enhances brightness but reduces the contrast in a display. For inverted structures in which

the electrode is deposited on the bottom of the phosphor layer, other metals such as Au,

Mo, Ta and W are candidates besides aluminum.



2.5.3 Insulators

The insulating layers of an ACTFEL device are an integral part of the capacitive

nature of the device, as seen in the equivalent circuit model in Section 2.3.3.. The bottom









insulator undergoes the same thermal processing as the phosphor layer and therefore must

have thermal stability and chemical compatibility with the bottom electrode and phosphor

layer in addition to good adhesion. The most important electrical requirements are that

this insulator prevents current flow through the device and withstands electrical

breakdown under the potential fields used in the device, usually up to 2MV/cm. In

addition, the characteristics of the phosphor-insulator interface affect the depth and

concentration of interface states34 and therefore significantly influence device

performance. Critical properties for reliable, efficient performance are as follows16:

1. High dielectric constant soFr;

2. High dielectric breakdown electric field, FBD

3. Small number of pinholes and defects

4. Good adhesion

5. Small loss factor, tan6.

So and ;r are the dielectric constant of vacuum and the relative dielectric constant of the

insulating layer, respectively. tan6 is a measure of the dissipative characteristics of the

capacitor and is equal to 1/(2ifRC), wherefis the frequency, R is the resistance and C is

the capacitance of the insulating layer.

In order to optimize the insulating layers for device efficiency, it is useful to go

back to the electrical behavior using the relationships presented in Section 2.3.3..

Equation 2-1 and 2-2 show that the proportion of the voltage which is across the

phosphor is governed by the relative capacitance of the phosphor and insulating layers.

To maximize the portion of the applied voltage dropped across the phosphor layer, the









capacitance of the insulators should be maximized while that of the phosphor is

minimized. The insulator and phosphor capacitances are determined by the equation

A
C = e0,r (2-7)


where t is the layer thickness. It is apparent that to optimize the insulator capacitance,

either the dielectric constant should be maximized as noted in the above list of attributes,

or the layer thickness should decrease. Indeed it is useful to minimize the insulator

thickness in a device. However, a lower limit on thickness is set by two factors. The first

is the third attribute in the list of characteristics, namely a small number of defects and

pinholes. The thinner the insulating layer the larger the probability of having defects and

pinholes in the film. This is primarily a processing issue and careful process control can

minimize the problem. The second limitation is the breakdown electric field strength, also

noted in the above list. The second limitation is the electric field strength, E = V/t. For a

fixed voltage, the thinner the dielectric film is, the higher the electric field strength and

the greater the chance of dielectric breakdown.

Generally speaking, it has been observed that materials with a higher dielectric

constant have a small breakdown strength, and vice versa. In addition, the high dielectric

constant materials tend to exhibit propagating breaking, meaning than small defects lead

to catastrophic failure The typical dielectric materials used are BaTa206 (BTO),

PbTiO3, BaTiO3, SiNx, SiON. Two insulating materials used for this research were ATO

and PZT. ATO is made of (Al203/TiO2)n multilayers grown by atomic layer epitaxy.35 By

replacing Zr for Ti in PbTiO3, Pb(Zro.5Tio.5)03 (PZT) was used as an insulating layer. The

dielectric properties of ATO and PZT are listed Table 2-1 for comparison.36









2.5.4 Phosphors



2.5.4.1 Hosts

The host is, by its definition, the host material for the luminescent centers. A critical

consideration for the host material is its bandgap characteristics.




Table 2-1 Parameters of dielectric materials ATO and PZT


Dielectric parameters ATO ((A1203/TiO2)n) PZT (Pb(Zro.5Tio.5)03)
Dielectric constant Sr A1203 8 PbTiO3 150-1700
TiO2 60
Electric breakdown strength A1203 5-8 MV/cm PbTiO3 0.5 MV/cm
TiO2 0.2 MV/cm


It should satisfy the following several criteria:

1. The host bandgap must be large enough not to absorb any emission from the

luminescent center. For complete visible transmission, this requires a band gap of at

least 3.1eV.

2. The host must be a good insulator below threshold. This is required to maintain a

voltage drop and subsequent electric field across the phosphor layer, leading to the

sub-threshold capacitive nature of the phosphor.

3. The host must have a high breakdown strength to allow for efficient acceleration

of electrons. This requires that the breakdown field of the phosphor must be at least 1

MV/cm.

4. The host must have good crystallinity and a low phonon-coupling coefficient to

minimize electron scattering.









5. The host must provide a suitable substitutional lattice site for the luminescent

center.

These requirements have historically been met best by sulfur- and oxide-based

compounds. The typical sulfur-based hosts are ZnS, SrS, CaS, SrGa2S4 and CaGa2S4.1 10,

37' 38 The typical oxide hosts are Zn2SiO4, ZnGa204, Zn2GeO4, Ga203.

Since the band gap of host materials for ACTFEL devices is generally wider than

that of Si or III-V semiconductors, EL phosphor materials are also called wide-gap

semiconductor materials. The band-gap of ZnS and Zn2GeO4 together with that of Si39 is

listed in Table 2-2.




Table 2-2 Typical wide band-gap semiconductor host materials16' 40


Semiconductor Materials Si ZnS Zn2GeO4
Bandgap 1.12eV 3.4eV (Zincblende) 4.4eV
3.6eV (Wurtzite) (Rhombohedral)




2.5.4.2 Luminescent Centers

The spectral properties, and to a large extent the temporal properties, of optical

emission are primarily influenced by the activator impurity in the phosphor. There are

two types of luminescent centers commonly used in phosphor applications. In many lamp

and CRT phosphors, luminescence results from recombination of electrons and holes

trapped in deep donor and deep acceptor levels, respectively. This class of luminescent

center is not used in ACTFEL devices because the electron-hole pairs are unstable under

high fields and the luminescence is effectively quenched.16 41 The second type of center









uses localized transition between electronic states of an isolated dopant ion. This dopant

ion is typically either a transition metal such as Mn2+, Cr2+, Ti4+, Cu Ag+, a rare earth

ion such as Eu3+, Eu2, Ce3+, Tb3+, or Tm3+, or an S2 configuration ion such as Pb2 or

Bi3+ .16 28 One characteristic of importance in terms of both understanding the

luminescent process and for application to displays is the radiative decay time. The decay

time is determined primarily by two selection rules related to the electronic configuration

of the ground and excited states, as well as the surroundings of the ion.28 One of these

selection rules is referred to as the parity selection rule, which forbids transitions between

levels of the same parity, i.e. transitions within the d shell, within the f shell, or between

the d shell and s shell are forbidden. The other selection rule is the spin selection rule,

which forbids transition between configurations with different spin states. Forbidden

transitions have longer decay time.28

The transition metal ions mentioned above all have a dn valence configuration and

their emission spectra are all characteristic of intra-shell d-d transitions. They are

therefore all forbidden by the parity selection rule,28 and their emission is usually broad

band. Therefore, decay times for the transition metal can be in the range of 100[ts to

several ms, depending on the spin selection rule and symmetry. Another general

characteristic of the transition metal activators is that because the transitions originate in

the valence d-shell, they are strongly influenced by the crystal field and ligand fields of

the host. The crystal field modification of Mn2+ by the host will be discussed in more

detail in Section 2.7.4..

There are two very different types of transitions observed for rare earth ions. The

first is observed in Ce3+ and Eu2+ in which the ground state configurations have a 4f









valence. The excited states have one or more electrons promoted to the 5d orbital, which

is strongly affected by the surroundings and symmetry of the ions. The second type of

rare earth transition is an intra-shell 4f-4f transition. This type of transition is seen in

Eu3 Tb3 and most other trivalent rare earths. These transitions are all parity forbidden

and have decay times typically of several ms. The spectra are characterized by line

transitions, and are insensitive to the surroundings and symmetry of the ion since the 4f

electrons are shielded by the 6s orbital.



2.5.4.3 Host-Center systems

In addition to considering the electrical properties and processing conditions of

the host and the luminescent properties of the center, one must consider the compatibility

and interrelationships between the two. Size mismatching of the activator on the cation

site is important, as is the valence match (or mismatch) with the host. For the case of

ZnS:Mn, the ion size (Zn2+ is 0.60A, Mn2+ is 0.66A), the valence (both are 2+) match

well. For the case of ZnS:Tb, the ions size of Tb3+ 0.92 A is much larger than that of

Zn2+. Besides, Tb3 leaves one positive charge after it substitutes for Zn2+. In this case,

Ag+ or F- ions are used to compensate the charge.

In terms of the host material, there are two categories of phosphors: sulfides and

oxides. Green phosphors include ZnS:Mn (optically filtered), ZnS:Tb, CaS:Ce, SrS:Ce

and CaS:Ce in the sulfide based materials,17 and Zn2SiO4:Mn, Zn2GeO4:Mn,

ZnGa204:Mn and Ga203:Mn in the oxide based materials.42 Among the sulphide

phosphors, the highest brightness and efficiency is obtained in ZnS:Tb. Among the oxide

phosphors, Zn2SiO4:Mn demonstrates the highest brightness. More details about









Zn2GeO4:Mn and ZnMgS:Mn will be discussed in Section 2.6. and Section 2.7

respectively.




Table 2-3 Performance of green ACTFEL devices using sulfide and oxide phosphor
materials


Sulfide-based
ZnS:Mn filter
ZnS:Tb,F
CaS:Ce
SrS:Ce
ZnMgS:Mn43
Oxide-based42
Zn2SiO4:Mn
Zn2GeO4:Mn
ZnGa204:Mn
Ga203:Mn


L (cd/m2)
60Hz (1kHz)
L40
160
125 (2100)
10 (150)
65 (900)
(325)
Lmax
230 (3020)
39 (341)
235 (758)
227 (1018)


r1 (lm/W)
(1kHz)
r40
1.0
05-1.0
0.1
0.44


lmax
0.78
0.25
1.2
1.7


CIE(x,y)


0.65,0.35
0.30,0.60
0.27,0.52
0.19,0.38
0.35,060

0.251,0.697
0.263,0.683
0.082,0.676
0.198,0.654


2.6 Zn2GeO4:Mn ACTFEL Devices


2.6.1 General Consideration of Oxide Phosphors

The most widely used phosphor host materials in ACTFEL devices are sulfide

compounds.16'1 Requirements for the phosphor layer in ACTFEL devices were discussed

in Section 2.5.4.1. Among these requirements, the need for efficient transport of high

energy electrons and the need for good crystallinity have limited the interest in oxides as

potential EL phosphors.16 Empirically, oxides have larger band gaps than sulfides and are

not as capable of transporting significant current densities of hot electrons.16 Efficient

host materials for ACTFEL phosphors are generally limited to those materials with

bandgap energies in the range of 3.5-4.5 eV. Oxides tend to be more refractory than









sulfides and achieving good crystallinity at processing temperatures compatible with

glass substrates (-550 C) is difficult or impossible with most oxide phosphors.42

Renewed interest in oxide ACTFEL phosphors has stemmed from progress made

in alleviating both of these limitations. In terms of the current transport, much recent

work in oxide phosphors has been focused on materials with moderate band gaps such as

Ga203 (4.52-4.84 eV),44 ZnGa204 (4.3eV),45 and Zn2GeO4:Mn (4.68eV).46)



2.6.2 Green Oxide ACTFEL Devices

Typical green phosphors used for ACTFEL devices are ZnGa204:Mn,47

Zn2SiO4:Mn, Zn2GeO4:Mn, Zn2GexSil-x04:Mn, and GaO3:Mn. The color of ZnGa204:Mn

EL emission is the closest to the standard green color (546.1 nm). The highest brightness

was obtained in Zn2SiO4:Mn, with a value of 5000 cd/m2 achieved at a drive frequency of

1 kHz.42 Ga2O3:Mn EL devices demonstrated the lowest threshold voltage of 110V at a

thickness of ltm.42 The lowest processing temperature for a green oxide phosphor is

observed for Zn2GeO4:Mn, with a value of 650C. Limited by processing constraints,

Zn2SiO4:Mn crystallinity is poor due to its high crystallization temperature. By

substituting Ge for Si, Zn2GexSil-O04:Mn demonstrated the best performance of all green

oxide phosphors due to improved crystallinity.42

RF sputtering deposition has been frequently used to produce oxide ACTFEL

devices. RF sputtering deposition of a variety of oxide ACTFEL devices was reported by

Minami, et al.42 These films were deposited at 2500C on a 0.2mm BaTiO3 sheet and

annealed at 700-1100 C for 5 hrs. in Ar. The unique characteristic of this deposition is

the use of the BaTiO3 substrate which acted as both substrate and insulating layer.









BaTiO3 can be annealed at much higher temperature than typical glass substrates.

However, this large 0.2mm thickness reduced the capacitance of the BaTiO3 sheet, in

accordance with the equation C = sogrA/t. Smaller capacitance in the insulating layer

lowers the voltage drop across the phosphor layer, according to the equation Vp = Cin /(Cp

+ Cin) Va. Fortunately, BaTiO3 has a dielectric constant of over 1500, which can

compensate for the large thickness required for a substrate.

Other growth techniques for oxide phosphors include dip-coating, sol-gel and

pulsed laser deposition. Dip-coating was reported to show a better performance than

sputtering deposition for the oxide ACTFEL devices. Sol-gel processing has been

reported only for powder ZnGa204:Mn EL devices. Pulsed laser deposition of

ZnGa204:Mn deposited on a single-crystal substrate was report by Norton et al.48 Single-

crystal substrate MgO2 was used for high temperature processing in order to obtain better

crystallization of the ZnGa204:Mn layer. The films were deposited at substrate

temperatures from 2500C to 8000C. Photoluminescence was obtained from the film at

540nm. It was found that the PL intensity was more strongly dependent on crystallinity

rather than surface roughness. Zn deficiency in the phosphor gwas found to be dependent

on the substrate temperature. However, use of a ZnGa204 target containing extra Zn

showed worse PL performance.



2.6.3 Zn2GeO4 Host Crystal Structure

This research particularly focused on the green oxide phosphor Zn2GeO4:Mn. As

presented in Section 2.5.4.3., host materials accommodate the luminescent centers. For

the case of Zn2GeO4:Mn, Mn2+ should replace Zn2+ to incorporate into the Zn2GeO4 host.









There are two types of Zn2+ positions available in the Zn2GeO4 reverse spinel structure,

tetrahedral and octahedral. In the regular spinel structure, the two Zn2+ ions in the

primary Zn2GeO4 unit are located in octahedral positions and one Ge4+ ion is in a

tetrahedral position. In the reverse spinel, one Zn2+ is in an octahedral position, the other

one in a tetrahedral site, and the Ge4+ is in an octahedral position. The tetrahedral and

octahedral position of Zn2+ are shown in the Figure 2-7. The octahedral position is only

shown with a quarter of its volume, while the other three quarters are in other primary

unit cells. Transition of Mn2+ is allowed only when it is in an octahedral symmetric

position. No transition of Mn2+ is allowed when it is in a non-symmetric tetrahedral

position.16



2.6.4 Pulsed Laser Deposition of Oxide Thin Films

Pulsed laser deposition has been used for oxide materials for some time, especially for

oxide superconducting materials such as Yba2Cu307-x .49 50 51 Pulsed laser deposition has

unique advantages, including film stoichiometry close to that of the target, low

contamination level, high deposition rate and non-equilibrium processing. Pulsed laser

deposition of phosphor materials have been reported such as ZnGa204:Mn presented

above and Y203:Eu. 52 53 The interaction of the laser with the target and plasma during

pulsed laser ablation can be broken down into four regimes: 1) Interaction of incident

laser with the target resulting in ablation of target surface; 2) Interaction of the ablated

materials with laser resulting in formation of plasma. 3). Expansion of the plasma. and 4)

Interaction of the plasma with background gas resulting in formation of interaction

plume right above the target. The first regime could be either of two situations, depending

on which process is dominant, thermal diffusion or optical absorption.












OZ"

SZn 2 tetahedral


O Zn2+ octahedral

O Ge2z octahedral










Figure 2-7 Reverse spinel structure of Zn2GeO4


This is determined by three factors: 1) the laser pulse parameters (temporal power density

I(t), pulsed duration tp, and wavelength A), 2) target optical properties (reflectivity R, and

absorption coefficient a ), and 3) the thermal properties of the target materials (thermal

conductivity K, latent heat per unit volume L,, specific heat per unit volume Cv, and

ablation temperature T,, etc). Thermal diffusion length Ld is given by the equation:

L = 2*D*t
D=K/

When the thermal diffusion length is larger than the optical absorption length, the laser

penetrates deeper into the target resulting in ablation of particles from the target. When

the thermal diffusion length is smaller than the optical absorption length, only the near-









surface region of the target is heated up by laser. More details about the pulsed laser

deposition mechanism is referred to 54

Formation of particles on the phosphor film surface is an issue. There are several

ways to reduce surface particle density. One method is to increase ambient gas pressure

to prevent ablated target clusters from forming a particle before they reaches substrate.

A second option is to place the substrate at an angle to the target plane, instead of placing

it right above the target, to keep the substrate away from the particle landing area.




2.7 ZnMgS:Mn ACTFEL Devices



2.7.1 ZnIlxMgS-based Thin Film Growth

ZnMgS:Mn electroluminescent thin film devices were only reported by Noma and

Minami, et al. in 1998.43 In this report, ZnMgS:Mn films were grown on

glass/ITO/SiO2/SiNx by e-beam deposition. Mn doping was provided from a MnS source.

It was grown at a substrate temperature of 1500C and in-situ annealed at 6300C for 1

hour. The best electroluminescence was 333cd/m2 at a drive frequency 60Hz. The EL

emission wavelength of the film deposited from a Zno.7Mgo.3S:Mn source was 552nm,

which is a greenish color.43

ZnMgS is also used for quantum well devices. It was reported that ZnS/ZnMgS

films were grown on III-V compound semiconductor single-crystal substrates to form

quantum wells.55' 56 The deposition techniques used were MOCVD (metal oxide

chemical vapor deposition)55 and MBE (molecule beam epitaxy).56 A strain effect due to

lattice constant mismatch was observed in the ZnS/ZnMgS quantum wells.









The stoichiometry of ZnMgS film is strongly affected by ambient sulfur pressure

during processing. ZnMgS grown by MBE under excess sulfur pressure has been

reported.57 Extra sulfur pressure was observed to increase the growth rate. Without the

use of excess sulfur, ZnMgS film deposition on GaP was not repeatable.57



2.7.2 ZnS-MgS Crystal Structures

ZnMgS is formed from the reaction of ZnS and MgS. The typical ZnS crystal

structures are cubic zinc-blende or hexagonal wurtzite. MgS forms in the cubic rocksalt

crystal structure. The two ZnS structures are illustrated in Figure 2-8. The zinc-blende

structure is similar to the diamond structure, which is two interpenetrating face centered

cubic (f.c.c.) primary units located at positions 14 of the body diagonal from each other in

the [111] diagonal direction. The only difference between the zinc-blende structure and

diamond structure is that the two f.c.c cells in zinc-blende are occupied by two different

types of atoms (ions), such as Zn2+ and S2- for the case of ZnS. The wurtzite form of ZnS

is two hexagonal primary units with one lattice shifted in [1220] direction, as shown in

Figure 2-8, with one lattice occupied by Zn2+, the other by S2-. MgS occurs only in the

cubic rock-salt structure, which has two fc.c.c primary units with one lattice shifted /2 in

the [100] direction.



2.7.3 ZnS-MgS Solubility

For ZnMgS, the amount of MgS replacing ZnS is restricted by its solubility in ZnS. The

equilibrium phase diagram of ZnS and MgS is shown as Figure 2-9. As shown in this

diagram, from pure ZnS to ZnMgS with 7% MgS incorporation exhibits the ZnS-P phase









in the wurtzite hexagonal structure, while ZnMgS with 7% to 24% MgS replacement

produces the ZnS-ac phase, which is zinc-blende cubic structure. In ZnMgS with over

24% MgS replacement, a separate MgS-y phase will be formed which is rocksalt cubic

structure.














a) b)

Figure 2-8 ZnS crystal structures: a) zinc-blende and b) wurtzite



Therefore, ZnS can be replaced by only 24% MgS before the MgS phase is segregated

out. Slightly different solubility for ZnS and MgS was reported in other literature.58 Zn

can replace Mg at atomic percentages up to 0.87 percent, resulting in a compound

formula of Zno.s7Mg0.13S, compared with pure ZnS. It was reported that Zn-rich ZnxMgl_

xS at l
structure.59



2.7.4 Crystal Field Modification of Mn2+

The outer shell electron arrangement of a Mn atom is 3d52s2. After two electrons are lost

from the outer shell, the Mn2+ ion is formed with an outer shell electron arrangement of

3d5. The quantum level of the five electrons is 6S, 4G, 4D, 4, 4F, respectively, in terms of










quantum level definition: 2s +L.16 These five quantum levels are the case for an isolated

Mn2+



\ ca-ZnS: wurtzite hexagonal
\. pB-ZnS: zinc-blende cubic
y-MgS: rocksalt cubic -, 10



L- -1 1 ---0

1t-e
SI,
i -. ,




,_ .. ,: ,




Figure 2-9 Equilibrium phase diagram ofZnS-MgS



The 6S is the ground state of an isolated Mn2+. These levels are degenerated under the

crystal field, but the extent depends on the environment of the Mn2+ ion. The five levels

are split in the way as shown in Figure 2-10. For instance, 4G is split into three 4T1, 4T2,

and 6A. The only allowed transition in Mn2+ is the transition from 4T1 of 4G to 6A of 6S.

The level of 4T1 of the Mn2+ ion is changed depending on the strength of the crystal field

formed around it. How the crystal field strength effects the energy states is governed by

Schrodinger's Equation:


2E
( +V) = 0 (2-8)


in which h is Plank's constant, E is the kinetic energy, Vis the potential and is the

wave function of an electron. Vis the crystal field strength mentioned above, which is a










vector quantity. According to the crystal field theory, the transition of dopant ions can be

simulated.60 According to the parity rule discussed in Section 2.7.4., the 4T1 to 6Al

transition of Mn2+ in a symmetric position is allowed, while 4T1 to 6A1 in a non-

symmetric position is not allowed. An octahedral position is formed by eight facets,

which are constructed by six atoms, as shown below in Figure 2-11 (a). In contrast, the

tetrahedral position is formed by four facets and constructed by four atoms. The

tetrahedral position of Mn2+ shown in the Figure 2-11 (b).


5'

4F
41


t 4
-o 4D
S4
21


0 -T





"- : 4E
4T2



0 -
I I I I I
Tj



L II Al


5 10
A(03 cm) --


Figure 2-10 Degneration of Mn2+ energy states under crystal Field


6S 0


















r------------------------7











I I
I I I i


--I I- - -I-- - -
- - -i- - - ----










II -
- --


a) b)





Figure 2-11 Positions of Mn2 luminescent center in host a) at octahedral and b) at
tetrahedral














CHAPTER 3
EXPERIMENTAL PROCEDURES

Preparation of Zn2GeO4:Mn and ZnMgS:Mn green ACTFEL devices includes

source preparation, substrate choice, thin film deposition and heat treatment.

Characterization of the devices includes the optical, electrical, microstructural and

chemical characterization.


3.1. Source Preparation

Two different phosphor layers were used for green ACTFEL devices in this

research: Zn2GeO4:Mn and ZnMgS:Mn. Targets (or pellets) of source material were

required for the thin film deposition techniques used. These source materials were not

commercially available in target form, but required careful preparation before any

deposition processes could be utilized. The compound Zn2GeO4:Mn target was prepared

both for pulsed laser deposition and RF sputtering, while Zn,. Mlu S:Mny pellets were

used for e-beam evaporation. Generally the target can be used for many runs of sputter

deposition, while the pellet can be used only once for evaporation and should be

completely evaporated during the evaporation process.



3.1.1 Preparation of Zn2GeO4:Mn Target

Two Zn2GeO4:Mn targets at one-inch diameter and prepared exactly in the same

way were used respectively for pulsed laser deposition and sputtering deposition. The Mn

concentration ([Mn]) was fixed at 1.5% atomic relative to Zn concentration ([Zn]) in the









target. According to the literature, there is an optimum [Mn] for best optical performance

in Zn2GeO4:Mn of approximately 1%-2%.61 62

The Zn2GeO4:Mn target was obtained by firing a powder mixture of ZnO, GeO2

and MnO2 at a temperature of 1100C. There were two firing processes involved: firing

before and after the target was pressed. The goal of the pre-firing process was to obtain

Zn2GeO4:Mn resulting from the reaction of the powder mixture. The goal of the post-

firing process was to condense the target for better thin film growth. Details of the

Zn2Geo4:Mn target preparation are shown in the following chart:

Mixing powder
ZnO, GeO2 and MnO2at
formula Zn2GeO4:Mn l5%


Ball-milling powder
mixture


Pre-firing powder mixture
at 11500C in air for 24 hrs.


Pressing powder
into 1" target


Post-firing target
at 10500C in air for 24hrs.


Figure 3-1 Flow chart of Zn2GeO4:Mn target preparation procedure



First, ZnO, GeO2 and MnO2 powders were mixed at the formula ratio

Zn2GeO4:Mn 1.5%. All the source powders were purchased from Alfa Aesar. Even

though Mn2+ is the luminescent center in Zn2GeO4:Mn, the source used for Mn was









MnO2, with an ionic valence of Mn4+. This was done because Mn4 was expected to

substitute one of the two Zn2+ positions to become Mn2+ after the powder mixture was

fired. The physical properties of the three powder sources are listed in the table:




Table 3-1 Physical properties of Zn2GeO4;Mn target source materials63


Zn2GeO4:Mn ZnO GeO2 MnO2
Atomic ratio 2 1 0.15%*2
F.W. (g/mol) 81.37 104.59 86.94
Melting point(C) 19750C 10860C 535C (decompose.)
Purity (Puratronic) 99.9995% 99.999% 99.9995%



Next, in order to obtain an intimate physical mixing, the powder mixture was ball-

milled with ceramic media for 24 hrs. During ball milling, the powder mixture was

immersed in acetone. The color of the pure MnO2 source was dark gray, and the ZnO and

GeO2 were white. It was evident that the powders were well mixed after ball milling

since the powder mixture was changed from white to gray even though it contained only

a very small amount of MnO2. Mixing and grinding by hand would not be capable of

changing the mixture color from white to gray. After ball-milling, the mixture was dried

on a hot plate to remove the acetone.

Third, the mixture powder was fired in a furnace at 11500C in N2 for 24 hours to

form Zn2GeO4:Mn. The Zn2GeO4 crystal structure was expected to be formed upon

firing. Firing is a chemical process since ZnO, GeO2 and MnO2 reacted with each other

during firing. During firing, Mn4+ in MnO2 was expected to convert into Mn2+ and

replace the Zn2+ in the octahedral position of the spinel structure, but not the Zn2+ in the

tetrahedral position since the transition of Mn2+ is only allowed when it is at the









symmetric octahedral position rather than non-symmetric tetrahedral position. The Mn2+

transition is referred to in Section 2.7.4..

Fourth, the fired powder mixture was pressed into an one inch diameter target

using an hydraulic press. After placing the powder into a 1" diameter steel die, a 50001b

force was slowly loaded onto it to form the target. Porosity was a concern for the target.

A solid and highly condensed target was desired for both pulsed laser deposition and

sputtering deposition, especially for the former since it is less likely to be cracked during

deposition by high energy laser because of higher thermal conductivity. A good quality

film starts with a good target. A large press loading can reduce porosity in the target.

However, overloading during pressing would result in the cracked target. The way to

measure porosity of a target is by the pore density given by the following equations,

Pore density = Actual mass density / Theoretical mass density (3-1)

Real mass density = Target weight/ Target Volume (3-2)

Theoretical mass density
(3-3)
Weight of all atoms in the primary cell / Volume of the primary cell

Since the size of Zn2GeO4:Mn target was 1" in diameter and 0.5cm high and the weight

was 3.74g, the actual mass density of Zn2GeO4:Mn was given by

Actual mass density ofZn2GeO4 :Mn target =
(3-4)
Weight of the target/ (r* (diameter of the target/2)2 height of the target

The volume of a primary unit Zn2GeO4 is a*b*c where a,b,c are lattice constants. A

Zn2GeO4 primary unit contains two Zn2GeO4 molecules and each Zn2GeO4 molecule

contains two Zn atoms, one Ge atom and four O atoms. Therefore, the theoretical mass

density of Zn2GeO4:Mn is given below,

Theoretical mass density ofZn2Ge04 :Mn Target = (3-5)

2* (2*F. W(Zn) + F.W.(Ge) + 4*F. W(O)) /(a*b *c)









It should be noted that since Mn concentration is low, its contribution to the theoretical

mass density of Zn2GeO4:Mn was ignored. The pore density of the Zn2GeO4:Mn target

was 0.45 according to the above equations, assuming Zn2GeO4:Mn is in the

rhombohedral crystal structure. It should be noted that there are three different

Zn2GeO4:Mn forms, which have different lattice constant a, b, and c. Here the

rhombohedral Zn2GeO4:Mn was chosen for the pore density calculation.

Finally, the target was post-fired in a furnace at 10500C in N2 for 24 hours to

consolidate the target. The temperature was increased slowly to 1500C in 4 hours and

remained at 1500C for 4 hours in order to completely remove moisture from the target

before it was heated up quickly to 10500C in 2 hours. Normally, the outer part of a target

heats up faster than the inner part. If the temperature were increased too quickly from

room temperature to 10500C, the outer part of the target would become dense before the

moisture in the inner part came out. The moisture would then be trapped inside the target

and the target tends to be cracking during firing.



3.1.2 Preparation of ZnjxMgxS_:Mn Pellets


The Zn,. l.M S:Mny pellets were obtained by reacting ZnS, MgS and MnS (or

metal Mn) powders which were mixed according to the formula ratio of Zn, Mg, and Mn.

As shown in Figure 3-2, the pellet preparation procedure basically can be broken into

three steps: 1) grinding the power mixture, 2) sintering in Ar/H2S at 650-8500C and 3)

shaping it into 0.25" diameter pellets which can fit into the 0.5" diameter graphite

crucible. The most important part for the pellet preparation was sintering, in which step

the ZnMgS crystal structure was expected to form.









3.2 Substrate Preparation


3.2.1 Glass/ITO/ATO Substrates


The substrate types used for this research were glass and Al203 ceramic. Both

types were deposited with a bottom conducting layer and a bottom dielectric layer before

they were used for the phosphor layer growth. Therefore, the substrates, along with the

two bottom layers, were glass/ITO/ATO and A1203/Au/Pb(ZrTi)03, which are used for

standard EL devices and inverted EL devices respectively.



Grinding

Powder ZnS, MgS and MnS or metal
Mn weighed according to the formula


Sintering

Powder mixture placed in quartz
cnicible in turbe fiurnace- Furnace


Figure 3-2 Flow chart of Zni. .My S:Mny pellet preparation procedure



Transparent glass/ITO/ATO substrates were used for the standard ACTFEL devices in

which the emitted light comes out of the substrate side. The glass substrate used here can

resist temperatures up to 8500C. ITO is a 2000A transparent conducting layer and ATO is

a 2500A transparent dielectric layer. The ATO layer was formed from A1203/TiO2 layers

deposited using atomic layer deposition by Planar Systems, Inc. It was used for both









standard ZnMgS:Mn and Zn2GeO4:Mn ACTEFL devices. Since the annealing

temperature of this substrate was restricted by the glass substrate transition temperature

and ITO degradation temperature, the highest annealing temperature used for this

substrate was 8000C for 2 minutes.46 Before the substrate was loaded into the vacuum

chamber, it was cleaned in an ozone cleaner for 6 minutes and blow dried. The substrate

was cut from 2"x2" into 0.67"xl" for e-beam evaporation of ZnMgS:Mn, into 0.5" x 0.5"

for pulsed laser deposition of Zn2GeO4:Mn and into l"xl" for sputtering deposition of

Zn2GeO4:Mn depending on the available thickness uniformity for each of the deposition

techniques.



3.2.2 A1203/Au/Pb(ZrTi)03 Substrates


The opaque Al203/Au/Pb(ZrTi)03 substrate was used for the inverted ACTFEL

devices. In this case, the emitted light comes out of the top transparent layer(s) instead of

the substrate side. The A1203 substrate is a 0.5mm polycrystalline ceramic sheet. A

2000A gold layer was deposited on the ceramic sheet and patterned as electric contacts.

A 20[tm Pb(ZrTi)03 (PZT) layer was screen printed on the gold layer. The 2"x2"

substrates with Au and PZT were provided by Westaim. The A1203 substrate, including

the Au layer and the PZT layer, is referred to as the PZT substrate in this dissertation.

Pb(ZrTi)03 has a much higher processing temperature and much higher dielectric

constant compared with ATO. The processing temperature of Pb(ZrTi)03 is limited by its

melting temperature of 950C 63 and its dielectric constant ranges from 300-1500.36 This

substrate was used for Zn2GeO4:Mn ACTFEL devices since high temperature was

required for crystallization of Zn2GeO4:Mn. The crystallization temperature of









Zn2GeO4:Mn is 6500C.46 The patterned 2"x2" substrate could be used for four individual

EL device units. One 2"x2" substrate was cut into four 0.25" x 0.25" substrates for pulsed

laser deposition of Zn2GeO4:Mn or four l"xl" for sputtering deposition of Zn2GeO4:Mn.

The poor thickness uniformity of pulsed laser deposition restricted the substrate size

down to 0.25" x 0.25". The size of the substrate used for sputtering deposition of

Zn2GeO4:Mn was restricted by the size of the one inch diameter Zn2GeO4:Mn target and

the substrate-target distance of 3.1cm.




3.3 Thin Film Growth

Growth of phosphor layers was one of the most important parts of this research.

The ZnMgS:Mn phosphor layer was grown by e-beam evaporation and the Zn2GeO4:Mn

phosphor layer was grown by pulsed laser deposition and RF sputtering deposition. All

phosphor layers were grown in vacuum systems at base pressures varying from 10-5 to

10-6 Torr.



3.3.1 Vacuum Systems


The deposition techniques for the top opaque Al conducting layer, the top

transparent conducting layer and the top transparent dielectric layer are not included here.

The vacuum system used for each of the three phosphor layer deposition techniques was

different. However, in all these cases the roughing vacuum was provided by an oil-sealed

mechanical pump. When the roughing pressure reaches the crossover point of 100 mTorr,

the roughing valve is closed, backing valve is opened and the gate valve for high vacuum

is opened. The high vacuum pump is typically backed up by the same mechanical pump









used for roughing. Several interlocks are used in the vacuum systems. Interlock between

vacuum gauge and high vacuum valve is designed to prevent damaging the high vacuum

pump when the chamber pressure is too high. Vacuum gauges varied from low vacuum

thermo-couple and capacitance gauges to high vacuum ion gauges. More details about

vacuum system design and maintenance is referred to.64' 65


Figure 3-3 Schematic diagram of vacuum systems


3.3.2 Pulsed Laser Deposition of Zn2GeO4:Mn


Pulsed laser deposition of Zn2GeO4:Mn was prepared by two different systems,

one at University of Florida and one at the Naval Research Laboratory (NRL) in

Washington D.C. The Zn2GeO4:Mn samples prepared by University of Florida focused

on photoluminescence of Zn2GeO4:Mn on silicon. The samples prepared by NRL were









aimed at obtaining electroluminescence of Zn2GeO4:Mn on glass/ITO/ATO and on

Al203/Au/Pb(Zro.5Tio.5)03.

The pulsed laser deposition parameters generally include laser power density,

laser repetition rate, background ambient gas pressure, deposition time and substrate

temperature. The deposition condition for pulsed laser deposition of Zn2GeO4:Mn at UF

is summarized in Table 3-2.




Table 3-2 Pulsed laser deposition parameters of Zn2GeO4:Mn at UF

PLD parameters
Target material Zn2GeO4:Mn
Target rotation 1 rotation/sec
Distance of substrate-target 10cm
Laser source KrF 248nm
Laser energy 400 mJ
800 mJ
Laser focal size 0.25cm2
Laser repetition rate 10 shots/sec.
Base pressure 5*10-' Torr
Ambient gas 02
Ambient gas pressure 200m Torr
Substrate material Si
glass/ITO/ATO
Substrate temperature 2500C and 800C
Deposition time 30 min.


The pulsed laser deposition parameters for the samples prepared by NRL is listed

in Table 3-3.

The laser deposition parameters shown in the two tables are slightly different.

However, the following two formulas can be used for the conversion:

Laser energy density (J/cm2) = Laser energy (J) /laser focal size (cm2)

Total laser shots = laser repetition (shots/sec) deposition time (s)









The film quality and thickness was dependent on the extrinsic parameter laser energy

density rather than the intrinsic parameter laser energy. The film quality and thickness

was dependent on the total laser shots rather than deposition time.




Table 3-3 Pulsed laser deposition parameters of Zn2GeO4:Mn at NRL


PLD parameters
Target material Zn2GeO4:Mn
Laser source KrF 248nm
Laser energy density 0.8J/cm
1.6J/cm2
Laser shots 10,000 shots
Base pressure 5*10-' Torr
Ambient gas 02
Ambient gas pressure 150 mTorr
200 mTorr
Substrate material Si
glass/ITO/ATO,
Al203/Au/PZT
Substrate temperature R.T.
2500C
8000C


The samples prepared respectively by the two different facilities were not intended to use

for comparison because the systems were different, despite the use of similar laser

parameters.



3.3.3 RF Sputtering Deposition of Zn2GeO4:Mn


A schematic of the RF sputtering deposition of Zn2GeO4:Mn used for this

research is shown in Figure 3-4. An RF power source was used to sputter the insulating

Zn2GeO4:Mn target, which was placed 3.1 cm above the substrate. A gas mixture of Ar








and 02 was supplied to the chamber during sputtering. Oxygen was added to reduce

possible O deficiency in the Zn2GeO4 film growth.


Zn2GeO4:Mn Target v

Ar


Ar+

00 0 0

/ Ar
A-- Ar

GND

Figure 3-4 Schematic diagram of Zn2GeO4:Mn RF sputtering deposition


Argon atoms were ionized by the RF electric field into Ar+ ions. Electrons from Ar

ionization were accelerated and impacted other Ar atoms to form a self-sustaining Ar

plasma in the chamber. Ar+ ions bombarded the Zn2GeO4:Mn target to knock off atoms

of Zn, Ge and O from the target by momentum transfer. Since no heating element was

used during the sputtering deposition of Zn2GeO4:Mn, all the sputter deposited

Zn2GeO4:Mn films used for this research were grown at room temperature. The RF

sputtering deposition parameters used for Zn2GeO4:Mn growth are listed in Table 3-4.

The target used for Zn2GeO4:Mn sputtering deposition was prepared in the same way as

the one used pulsed laser deposition. Further details about the target preparation are

referred to in Section 3.1.1.. This sputtering system was originally made at Materials

Research Inc. It accommodated three large targets, one 6" and two 8". One of the 8"









target sites was modified into three 1" target sites. The distance between the 1"

Zn2GeO4:Mn target and the substrate plane was 3.1cm.




Table 3-4 RF sputtering deposition parameters of Zn2GeO4:Mn


Target Sintered 1" Zn2GeO4:Mn
Substrates A1203/Au/PZT
NEG/ITO/ATO
Si
Base pressure 5*10-5 Torr
Ambient gas pure Ar
Ar 70 sccm /02 30 sccm
Ar 50 sccm/02 50 sccm
Working pressure 7.2mT
15mT
25 mT
Target power RF AC 25w (reflected Ow)
Substrate temp. R.T.


Due to limitations of the size of the target and the target-substrate distance, l"xl"

substrates were used to ensure better film thickness uniformity. The three different

substrates used for sputtering deposition of Zn2GeO4:Mn were A1203/Au/PZT,

glass/ITO/ATO and Si. The PZT substrates and the glass/ITO/ATO substrates were used

for ACTFEL Zn2GeO4:Mn devices. The Zn2GeO4:Mn films deposited on Si were used

for microstructural characterization since the background interference from substrate Si is

much less than that from substrate PZT.









3.3.4 Electron Beam Evaporation of ZnMgS:Mn


A schematic of the electron beam evaporation used for ZnMgS:Mn growth is

shown in Figure 3-5. Electrons were emitted from a filament and accelerated by a 10KV

supply. The current for accelerated electrons ranged from 0 -10A, and the electrons were

directed and focused by a pair of permanent magnets. A ZnMgS:Mn pellet was placed in

a carbon crucible. When the energetic electrons bombarded the pellet, decomposition or

evaporation or sublimation of the pellet occurred. The ZnMgS:Mn pellet sublimed and

decomposed into atomic Zn, Mg, Mn and S. Upon sublimation S atoms could be formed

as clusters of S2 or S4 before reaching the substrate. The substrate was heated up from

room temperature to 2000C. A ZnMgS:Mn film grown at elevated temperatures was

expected to have better film crystallization than one grown at room temperature.

Deposition rate was controlled by the filament current. Film thickness was monitored by

a quartz crystal oscillator, and controlled by a shutter located over the substrate, which

was operated manually. The distance between the pellet source and the substrate in the

system was 15 cm. The E-beam evaporation parameters for ZnMgS:Mn growth are listed

in Table 3-5. The glass/ITO/ATO substrates were used for the ZnMgS:Mn standard EL

device structure. A base pressure of 5*10-6 Torr provided a mean free path of 1 cm for the

background gas atoms. The mean free path is inversely proportional to the base pressure,

therefore higher vacuum establishes a longer mean free path. Also, a low base pressure

provides less contamination from background gas atoms during the ZnMgS:Mn growth.

The ultimate base pressure of the high vacuum diffusion pump used was 1* 10-7 after 6

hours of pumping. Elevated substrate temperatures resulted in two things: a better

ZnMgS:Mn crystalline growth and variation in the Mn doping in the film. The ZnS (or









ZnMgS) sticking coefficient is a strong function of substrate temperature. When the

substrate temperature is over 2500C, there is no ZnS growth on the substrate since the

desorption rate is greater than the deposition rate.66 The reduction of ZnS growth rate

with substrate temperature indirectly changed the Mn concentration relative to the host

ZnS (or ZnMgS).


S\ate










Mn



Figure 3-5 Schematic diagram of Zn Mi S:Mn E-beam evaporation


There is an optimal luminescent center concentration in the host. As explained in Section

2.5.4.2., either too low and too high luminescent center concentration will reduce the

brightness. Therefore, an optimal deposition rate is preferred. If it is too high, it is

obvious that the growth quality of ZnMgS:Mn is poor, since more defects and

dislocations tend to be formed at high deposition rate. If the rate is too low, the

incorporation of the background gas atoms into the ZnMgS:Mn films is more likely to

occur. Generally, films of the same thickness are used for comparison since the

brightness is a strong function of the thickness. One micron was typically used for the









standard thickness. However, the e-beam evaporation rate of ZnMgS:Mn was difficult to

control since it was dependent on several parameters, including Mg replacing percentage,

the electron filament current, the substrate temperature, etc. The e-beam evaporation of

ZnMgS:Mn in this research was more focused on the quality of the film rather than the

thickness of the film.




Table 3-5 E-beam evaporation parameters of Zn21-xMgxS:Mn


Evaporation Parameters
Substrate glass /ITO/ATO
Base pressure 5-8x10- Torr
Substrate temperature R.T. -2000C
Deposition rate 1-20nm/s
Thickness 0.3-1.5um


3.3.5 Deposition of Conducting Layers

After the phosphor layer was finished, a top conducting layer was required to

form an EL device. The majority of the EL devices used for this research were half cell

structures, which do not require a top dielectric layer. In this case, the top conducting

layer was deposited directly on the phosphor layer. For the standard EL devices in which

the phosphor layer was deposited on transparent glass/ITO/ATO substrates, an opaque

metal conducting layer was required to reflect back the emitted light. In this research, a

2000A Al film was thermally evaporated onto the phosphor layer. The base pressure for

Al coating was 5*10-4 Torr and the deposition rate was approximately lA/s, monitored by

a quartz oscillator. The Al dots used for optical testing of the EL device were formed

using a mechanical mask. For the inverted EL devices in which the phosphor layer was

deposited on opaque Al203/Au/PZT substrates, a transparent conducting layer was









required on top of the phosphor layer. In this case, a 2000A ITO film was grown by RF

sputtering. It was deposited at a working pressure of 3.1 mTorr (base pressure 1*10-6

Torr), with an ambient of Ar 100sccm/ 02 3 sccm and RF power 700 W (reflected power

0 W). The target was 8" in diameter so that a large quantity of samples could be

deposited with ITO in the same run. The substrate temperature was room temperature. As

measured by multimeter with two probes separated at a distance of 5 mm, the sheet

resistance of 2000A ITO ranged from 1-10 6 Q, depending on the degree of addition of

O2. The general rule is that as more 02 is added, the ITO is less conductive and more

transparent, since the mechanism of the ITO electrical conductivity is involved with O

vacancies in ITO.67 68

Since Zn2GeO4:Mn was deposited on opaque Al203/Au/PZT substrate. In order to

form an ACTFEL device, the top conducting layer should be transparent. In this research,

ITO transparent conducting layer was used. ITO sputtering deposition was done in

sputtering system. The ITO target was 8" in diameter. To make ITO electrically

conductive and optically transparent is of contradiction according to its solid state nature.

More details about the electrical conductivity of ITO is referred to Section 2.5.2.1. In our

RF sputtering system, the lowest 02 flow rate can be reaches is 1.5% of full 200 s.c.c.m

which is still too high for ITO deposition. Another way to further reduce 02 in chamber is

to reduce working pressure. After adjusted the open position of the throttle valve, the

lowest working pressure which can be reached was 7.5 mT. Deposited at the working

pressure, the ITO sheet resistance was still up to MQ. In order to further reduce chamber

pressure, the throttle valve was left completely open. The lowest pressure dropped to 3.1

mT. The sheet resistance was also reduced enormously. The resistance comparison at the









two different working pressures is listed in Table 3-6. The resistance was measured by

multi-meter with two terminals at a distance of 1 cm. The ITO thickness is 2000 A. The

thickness 2000 A was obtained at deposition conditions: base pressure 1*10-5 Torr, Ar

100 s.c.c.m. /02 3 s.c.c.m, working pressure 3.1 Torr, power 700 W for 10 minutes. The

high deposition rate of 200 A/min was reached by applying this large power. The

reflected power for the 700 W supply was zero. Since oxygen pressure in the chamber

should be minimized in order to obtain high ITO conductivity, the base pressure is

preferred as low as possible accordingly to reduce oxygen partial pressure in the

background.




Table 3-6 ITO sheet resistance at two different working pressures


Working Pressure ITO sheet resistance
(Ar 100sccm. / 02 (t = 2000A and s=lcm)
2%,*200sccm)
8.1mT 105- 10 i7
3.1mT 10-102hM


3.4 Thin Film Heat Treatment

Heat treatment is required for re-crystallization of phosphor thin films in most

cases. Phosphor thin films are normally grown at low temperatures, especially sulfide

phosphor thin films since the substrate temperature is restricted by the high sulfur vapor

pressure. Re-crystallization can be achieved upon annealing at high temperature.









Luminous efficiency and luminance is strongly dependent on phosphor film crystallinity

16' 1, 12 since these defects and grain boundaries are responsible for irradiative emission.



3.4.1 Lamp-based Rapid Thermal Annealing


In this research, phosphor thin films deposited on glass/ITO/ATO were annealed

by a lamp-based furnace in Ar. High ramping rate is a big advantage of lamp-based

furnaces. In this system, ramp up time was 1.5 min from R.T. to 7000C and 2.5 min. from

R.T. to 8000C. We called this system rapid thermal annealing (RTA), even though the

annealing time was much longer than that defined by the semiconductor industry.

Annealing times from 1 min. to 30 min. were typically applied, using programmed

recipes. The cooling rate was slow, normally taking 30 minutes to reduce temperature

from 700-8000C to 1000C by water and N2 cooling the lamps.



3.4.2 Box Furnace Annealing

Zn2GeO4:Mn films deposited on PZT substrate can be annealed at 8000C for

much longer times in air. A box furnace was used to anneal all Zn2GeO4:Mn films on

PZT substrates. The ramping rate for the box furnace was low, taking 1.5 hrs to reach

7000C and about 2 hrs to reach 8000C.



3.4.3 In-situ Vacuum Annealing


Some of the ZnMgS:Mn films deposited on glass/ITO/ATO were in-situ annealed

in vacuum after the evaporation was finished. A tungsten wire heater was placed under









the substrate. Maximum temperatures up to 5000C could be reached for this in-situ

annealing.


3.5 Characterizations


In this research, the information about the photoluminescence (PL),

electroluminescence (EL) and its charge transfer is given in "Optical and Electrical

Characterization". The information about thin film structure and composition is given in

"Microstructural and Chemical characterization", which includes the following analytical

techniques: XRD, SEM/EDX, XPS, SIMS and TEM.



3.5.1 Optical and Electrical Characterizations


3.5.1.1 Chromaticity


Chromaticity as defined in Section 2.3.2., is determined by the CIE coordinate nos

x and y. In this research, the CIE coordinated x and y for both photoluminescence and

electroluminescence were measured by a spectrometer. At the same time, the brightness

and emission spectrum were measured. The model of the spectrometer was PR-550

Spectra Colorimeter. Before the optical data was collected, both the image of the

interested sample area and the image of the dark dot in the camera were ensured clear by

adjusting focal length of the lenses.









3.5.1.2 Photoluminescence


When a phosphor material is struck by a light source at a certain wavelength, an

electron of a luminescent center in the phosphor material is excited to a higher energy

state before it relaxes back to the ground state, emitting a photon. This process is called

photoluminescence (PL) because the excitation source here is light. PL can be used to

characterize both the transitions leading to excitation as well as emission.



3.5.1.2.1 PL/PLE Spectra Excited by Xe Lamp

PL is photoluminescence emission and PLE is photoluminescence excitation. The

PL emission spectrum is defined as the emission intensity as a function of wavelength

emitted for a constant excitation wavelength. The excitation spectrum is the emitted

intensity at a fixed wavelength versus the excitation wavelength. For instance, in order to

obtain an emission spectrum of Zn2GeO4:Mn, the excitation source was typically set at

the wavelength 325 nm, which is close to the bandgap wavelength of Zn2GeO4 (4.68

eV).46 The relationship between wavelength and photon energy is:

1240
(nm) E(eV)= (3-6)
E(eV)

The emitted intensity versus wavelength was measured from 300 nm to 800 nm, which

covers the main peak of Zn2GeO4:Mn at around 540 nm. The maximum intensity was at

540 nm. To measure the excitation spectrum, the intensity of emission at 540 nm was

measured as the excitation wavelength was scanned from 300 nm to 540 nm. The

maximum intensity of the 540 nm emission was identified at an excitation wavelength of

340 nm. The emission spectrum was maximized by 340 nm excitation, which was used









for this study. Photoluminescence emission spectra and photoluminescence excitation

spectra were collected using a Photon Technology International integrated PL system that

used a 150 W Xe lamp source for excitation and double 14 cm monochromators for both

excitation and detection. This system was especially used for PL emission and excitation

spectra of the Zn2GeO4:Mn target. The PL intensity from thin film samples was too low

to be detected by this system, as the background noise intensity was higher than the

sample peak intensity.



3.5.1.2.2 PL Spectra Excited by UV Lamp

A UV lamp was used to measure PL spectra of Zn2GeO4:Mn thin films. ZnS:Mn

films do not demonstrate any PL at room temperature.12 An ultra-violet lamp with broad

wavelength range from 254 nm to 600 nm was used as the excitation source. The UV

lamp incidence plane was placed at an angle of 450 to the sample plane normal as shown

in Figure 3-6.




1 Photon detecting direction
ITV inrcidpnrp






Position 1: Normal to sample plane
Position 2: 450 to sample plane

Figure 3-6 Schematic diagram of photoluminescence spectrum measurement
under excitation of ultraviolet lamp









UV lamp light is not collimated, but comes out in an angular distribution. The illustrated

UV incidence direction is simply the UV lamp plane normal. Emitted photons from the

sample are detected at an angle either normal to the sample plane or at an angle 450 to the

sample plane. Emitted light was detected by a Photoresearch 650 spectrometer and the

detected sample area had a spot size of 0.675cm2. The UV source light reflected by the

sample was collected together with the light emitted from the sample at the detecting

angle 45. By placing the detector normal to the sample plane, the UV source reflection

could be minimized. The PL spectrum peak shape was broader for the case of the

detecting angle at 450 than that of the angle normal to the sample plane.



3.5.1.2.3 PL Spectra Excited by Laser Source

A laser excitation source of 325nm wavelength (provided by Professor Hummel's

group) was used to measure ZnMgS:Mn photoluminescence spectra. No detectable PL

could be obtained from ZnMgS:Mn when excited by the above UV lamp. ZnS:Mn thin

films do not give any photoluminescence at room temperature.12 However, after Mg

replaced Zn in ZnS:Mn, photoluminescence could be detected in ZnMgS:Mn when using

high energy density laser excitation.



3.5.1.3 Electroluminescence (Luminance vs. Voltage)

Electroluminescent brightness is typically determined using a luminance vs.

voltage curve (L-V curve). A typical L-V curve has been shown in Section 2.3.1.. Two

important parameters obtained from the L-V curve are the threshold voltage (Vth) and the

brightness 40V above the threshold (B40). In most cases, the terms luminance and









brightness mean the same thing. In the literature, either L or B were used for the

brightness. The method for determining Vth and B40 from L-V curves has been given in

Section 2.3.1.. The L-V behavior for an ideal ACTFEL device shows zero brightness

below the threshold voltage, since no light is emitted, and above the threshold voltage,

the brightness increase dramatically with applied voltage. A real ACTFEL device has a

"leaky" voltage process, which means below the threshold voltage the brightness

increases slowly and at or above the threshold voltage the brightness increases rapidly.

The voltage driver used for EL measurement was provided by Planar Systems,

Inc. and electroluminescence information including emission spectra, chromaticity and

luminance was collected by a Photoresearch PR-650 spectrometer. The light emitting

sample area was focused and the detected area was a 0.675cm2 dot. The AC voltage

driver was powered by a DC power supply at 300V and 0.025A. The voltage waveform

was trapezoidal with a rise time of 5[ts, stay time of 30[ts, and a drop time of 5|ts. The

stay time could also be set for 10ts or 100[ts. Generally, when longer stay time is used,

higher brightness is obtained from the device. The frequencies available for the driver

were 60Hz, 1kHz and 2.5kHz. Generally, higher frequencies produce higher brightness.

All the Zn2GeO4:Mn ACTFEL devices were tested at 2.5kHz and all the ZnMgS:Mn

ACTFEL devices were tested at 60Hz. Voltage waveforms seen in the literature also

include sinusoidal and triangular.69



3.5.1.4 Q-V and C-V Curves

When measuring the L-V curve, information about charge transfer and the

capacitance change can be obtained by collecting the device current. When a voltage V(t)









is applied to a TFEL device, the charge Q(t) accumulates in both sides of an EL device,

one side negative with electrons, the other side positive with holes. Dynamic capacitance

C(t) is defined as differentiation of charge by voltage, dQ(t)/dV(t). Dynamic capacitance

is different from static capacitance, which is defined as charge over voltage. If the charge

is linearly changed with the voltage, the dynamic capacitance becomes the same as the

static capacitance. The test configuration for Q-V data and C-V data measurement is

shown schematically in Figure 3-7. An AC voltage at a certain frequency is applied at

terminal Vi and V4 as shown in Figure 3-7.


/ /




i GND


Vi V2 V3 V4



Figure 3-7 Schematic diagram of Q-V measurement for ACTFEL device



Voltage dropped on the TFEL device is given by the difference of terminal V2 and V3.

Since the two resistors are connected to the EL device in series, current through each of

the resistors should be the same as that through the TFEL device due to the charge

conservation rule. The current flow through either one of the resistors, I(t), is its voltage

drop over its resistance, V(t)/R.. The resistors were 1000 each. The voltage drop across

the resistor, for instance, the one on the left side is therefore the voltage difference

between terminal V, and V2. The current flow through the resistor is (Vi-V2)/R. The

voltage and the current across the phosphor layer are given by the following equations:

(3-7)










VEL-device (t) = V3 (t) V2 (t)


IEL-devce(t)V2(t) (t)

The V(t) and I(t) is changed with time periodically. They are shown in Figure 3-8.

Voltage waveform used for the Q-V curve was the same trapezoidal waveform described

in the L-V curve, using 5[ts rise time, 30[ts stay time and 5[ts drop time at each of

positive and negative polarization cycle. The current, I(t), reaches its maximum when the

voltage rises and drops since I(t) oc dV(t)/dt.


C D

BB

SF J

I I I I
I I I I
5ts 30[ts 5ts H I



Figure 3-8 Trapezoidal voltage waveform shape of ACTFEL device



A frequency of 2.5KHz was used for the Q-V curve of all Zn2GeO4:Mn TFEL devices.

The voltage at each of the four terminals was monitored using a Tektronix TDS 510A

digitizing oscilloscope. The charge Q(t) is an integration of the current I(t) over time t

and it is given by the following equation,


F-device (t) = I dece (t)* dt (3-9)









Assuming all the charge is relaxed at the end of one polarization, before the next cycle

starts, Q(t) is an integration of I(t) covering one polarization period. Dynamic

capacitance C(t) is therefore given by the following equation,70

d EL-device
CEL-de... (t)- dVEL -dce (3-10)
dV L-dence

Since V(t), Q(t) and C(t) are determined, Q-V and C-V curves are obtained. A typical Q-

V curve and C-V curve for Zn2GeO4:Mn device are shown in Figure 3-9. The Q-V

behavior of the simple equivalent circuit was discussed in Section 2.3.3.. "Equivalent

circuit" is shown in Figure 3-7 as the broken lines. The conduction charge and

polarization charge discussed for an ideal device are also seen in the Figure 3-7 (broken

lines). Two new features seen are the relaxation charge and leakage charge.

SVQ(t)

E. relax

Qleak -T-- SlopeCBD Qcond
Qpol lopeAB Cp/ -----------
rM V(t)
G A Vth Vmax




I



Figure 3-9 Ideal (broken line) and real (solid line) Q-V Curve
of Zn2GeO4:Mn ACTFEL device



Relaxation charge occurs when the applied voltage reaches its maximum value (or near

maximum value in the real case) and reflects current which flows during the dwell, or C-









D portion of the waveform. The name refers to the fact that at a constant applied voltage,

conduction charge creates an opposing electric field that relaxes the field across the

phosphor layer. The leakage charge occurs once the applied voltage reaches zero and

refers to current that flows in the opposite direction as the previous pulse due to the

polarization field present at the end of the pulse. For one pulse cycle (positive and

negative), before the threshold voltage is reached, the capacitance of the device is the

capacitance of the dielectric layer and the phosphor layer in series, which is

CpCi/(Cp+Ci). After the threshold is reached, the phosphor layer is broken down and

there is no capacitance contribution from the phosphor layer. At this point, the

capacitance of the device is the capacitance of the dielectric layer, Ci, which is higher

than the value measured before the phosphor layer is broken down. The capacitance is the

slope of the Q-V curve. Wager, et al., did an extensive study on electrical characterization

of ACTFEL devices using Q-V and C-V. More details about electric characterization is

referred to. 70' 71' 72' 73' 74



3.5.2 Microstructural and Chemical Characterizations

The microstructural and chemical characterizations will be discussed. Their

fundamental mechanism and the experimental procedures will be are focused. More

details about the microstructural and chemical characterizations can be referred to the

books "Materials Characterization Encyclopedia75 and the TEM technique can be

referred to 76











3.5.2.1 X-ray Diffraction

XRD is acronym of x-ray diffraction. X-ray diffraction is the result of the

interference of the incidence x-ray and the reflected x-ray. When the phase of the two x-

rays is synchronized, the intensity is added up, and when the phase of the two is 1800

shifted, the intensity is subtracted. A single crystal contains many crystal planes at a

certain orientation that can be expressed by Miller indices, such as plane (001) and (111).

When incident x-rays strike a single crystal, they are reflected from different crystal

planes. Whether the reflected light from the crystal plane is added up or subtracted from

the incident light depends on the light travelling distance between the same stacking

planes. The added up intensity appears as a peak in the X-ray diffraction pattern. The

diffraction occurs according to Bragg's Law, shown below,



2d sin 0 = ni (3-11)

ao

hk h 2 + k2 2 (3-12)

where X is wavelength of the incident x-ray, 0 is the angle between the crystal plane

normal and both incident and diffraction rays, h, k and 1 are the Miller indices of the

crystal plane and ao is lattice constant. In the above equation, the cubic crystal structure is

assumed otherwise a, b, and c should be used instead of ao. Information that can be

obtained from x-ray diffraction includes crystallite size, crystallinity, strain and lattice

constant. If the sample is polycrystalline and the crystallites in the sample are randomly

oriented, the numerous diffraction peaks are characteristic of the crystal structure. The

FWHM method is used to compare the crystallinity of crystalline samples. FWHM means









the full width at half maximum intensity, A(20). The A(20) for both a sharp and broad

peak are shown in Figure 3-10. Better crystallinity is associated with a smaller A(20). The

crystallite size L is generally inversely proportional to A(20) and the relationship of the
(3-13)
two is given by,

L
A(20)*cos 0

where X is the wavelength and 0 is the angle between the incident x-ray and the crystal

plane normal. The X-ray diffraction system used for this research was a Philips APD

3720. The x-ray source was Cu-Kac, which was generated from a Cu tube operated at

40kV and 20mA, with a wavelength of 1.54A. The scanning angle 2 0 used for the

research was from 100 to 800, with a scanning step of 0.050. More details about x-ray

diffraction are referred to 75


Intensity











I----

S2>20

A(201)

Figure 3-10 Schematic diagram of crystallity comparison using FWHM of x-ray
diffraction patterns









3.5.2.2 Scanning Electron Microscopy/Energy Dispersive X-ray Spectroscopy


SEM is an acronym for Scanning Electron Microscopy and EDX is an acronym of

Energy Dispersive X-ray Spectroscopy. SEM was used to obtain the film surface image

at magnifications up to 20,000x. EDX was used for quantitative analysis of elemental

composition in films. EDX is generally more sensitive to heavy chemical elements since

the characteristic x-ray yield is increased with the atomic number. Theoretically,

chemical elements with atomic number lower than 3 are undetectable by EDX.

Practically, the chemical elements with atomic number lower than 5 are almost

undetectable by EDX. The SEM functions using electrons emitted from a heated

filament. These electrons are then accelerated by a voltage of 15KV before the electron

beam size is reduced by a magnetic condenser lens. Magnetic lenses are used to reduce

the spot size further to provide the appropriate magnification for secondary electron

image. Secondary electron image provides information about sample surface

morphology. The secondary electron collector is placed at a low angle close to the

sample. A collector for the back-scattered electrons, which are strongly dependent on

atomic number, is generally placed at a high angle almost normal to the sample plane. An

EDX detector is also placed close to the sample to collect the characteristic x-rays. More

details about SEM/EDX is referred to 75. The mechanism of generation of characteristic

x-ray by incident electrons is shown in Figure 3-11. Accelerated electrons from the

scanning beam strike atoms in the sample. Generally, electrons in the K shell tend to be

excited by these primary electrons into higher shells, such as the L shell. When a primary

electron strikes an electron at K shell, an electron at K shell is excited to L shell in the

same time.










Primary
Auger electron


L Shell

Nuclei




Figure 3-11 Schematic diagram of energy dispersive x-ray spectrum



When the excited electron in the L shell relaxes back to the K shell, either a photon or an

Auger electron will be created. The photon, which has an energy equal to the difference

between the K shell and L shell, is called a characteristic x-ray, given by the following

equation:

h = E(L-shell) E(K-shell) (3-14)

Each of the elements has its own set of characteristic x-rays that are used to identify

chemical elements in a given sample. When a multilayered thin film sample is being

quantitatively analyzed, penetration depth of the primary electrons becomes an issue. The

penetration depth is determined by the primary electrons' energy, E, and is given by the

following equation:

4120
R(pm) = 412/ E(Ke V)( 265- 0954nE(KeV)) (3-15)
p(g/cm3) (3-15)

Assuming the density is 5g/cm3 and the applied voltage is 15KV, the penetration depth is

1.23[tm. The interaction volume of electrons in the sample is roughly pear shaped75 and

would not give accurate quantitative information of the top layer if the primary electrons

penetrate into the layer adjacent to the top layer.









Insulating samples were coated with a thin layer of carbon or Au to make them

conductive prior to SEM/EDX. Carbon coating is more often used for EDX analysis

because it's characteristic x-ray peaks are less likely to interfere with those from the

elements of interest. Au coating is better for SEM imaging. The SEM/EDX system used

for this work was the JEOL 6400 and the SEM image and EDX quantitative analysis was

processed by the software LinkISIS. SEM image resolution was improved by using a

higher condenser lens current and a smaller aperture size. The condenser lens current at

scale 11-12 was typically used for a fixed working distance of 15mm. The second

smallest aperture was used of the four aperture sizes available for SEM imaging. For

EDX analysis, a smaller condenser lens current was used to allow more x-rays to be

detected. The typical condenser lens current setting used for EDX analysis was at scale

6-7.



3.5.2.3 X-ray Photo-electron Spectroscopy


XPS is an acronym for x-ray photo-electron spectroscopy. Details about XPS are

referred to 75. In summary, when the Mg source x-ray strikes a given sample, they may

knock off inner shell electrons such as those in the K shell. Information about the given

chemical element is provided by the energy of the secondary electron, according to the

following equation:

E = Ebondi (3-16)

Inner shell electrons could also be excited to a higher shell such as the L shell and relax

back from the L shell to the K shell to knock out another L shell electron, which is an

Auger electron. Its energy is given by the following equation:


(3-17)









E -E -E -E
Auger (K-shell) (L-shell) L-shell

The generation of the photoelectron and Auger electron is shown in Figure 3-12. XPS

was used to analyze the chemical bond energy of Ge4+ in Zn2GeO4:Mn films. If the

Zn2GeO4:Mn film contains ZnGeO3 or GeO2, the Ge4+ peaks in XPS spectrum should

be different since the environment of Ge4+ in each of the three crystal structures

(Zn2GeO4, ZnGeO3 and GeO2) is different.



3.5.2.4 Secondary Ion Mass Spectroscopy


SIMS is an acronym for secondary ion mass spectroscopy. Primary ions such as O- or

Cs+ bombard the specimen to knock off atoms and ions from the specimen from the

surface. The sputtered secondary ions are then separated out according to their mass over

charge ratio (m/q) using a magnetic field.


/ Auger

L Shell
*^ L Shell


Figure 3-12 Schematic diagram of x-ray photo-electron spectrum









SIMS is especially used for doping profile analysis. The secondary ion yield of a dopant

is dependent on yield sensitivity of doping in a given matrix material and it is given by

the following equation:

Imatnx =RSF Idopant (3-18)
C Cdopant-to-matnx
matrix dopant


Cdopant (RSF Cmatix) dopan (3-19)
matrx

where Cdopant is the dopant concentration, Cmatrix is the matrix element concentration

(which can be normalized as 1), RSF is the relative sensitivity factor of the dopant

relative to the matrix element, and Idopant is the dopant secondary ion yield and Imatrix

is the matrix element secondary ion yield. In this case, a standard is required for absolute

doping concentration analysis. SIMS was used to obtain ZnMgS:Mn film profile and to

identify Mn2+ absolute doping concentration in ZnMgS.



3.5.2.5 Transmission Electron Microscopy


TEM is an acronym for transmission electron microscopy. This technique was

used to analyze the microstructure of ZnMgS:Mn films in this research. A specimen for

TEM should be at least 200A thin so that electrons can effectively pass through.

Specimens are normally ion milled to form a wedge where a sufficiently thin region can

be used. Before the specimen is ion milled, the sample must first be cut into a square in

the size of 0.3mm and glued face to face (in the case of a cross-section sample) with the

film part in the middle, hand polished down to 1mm, and dimpled thinner to 20[tm. The

gluing step is critical. If the initial gluing fails in the later steps, the two pieces are









separated and the sample preparation must start all over again. The glue used for

ZnMgS:Mn was G-bond epoxy. The glued sample was heated up to 1500C and remained

at that temperature for 30 minutes to ensure a good curing. Since ZnMgS:Mn was

deposited on glass substrates, the ion milling time for thinning the glass specimens was

much longer than that for silicon specimens. It took 10 hours to finish a 20[tm

ZnMgS:Mn specimen.

In the TEM, electrons are accelerated under a high voltage up to 400KV, which

corresponds to a wavelength of 0.0016nm. Wavelength becomes shorter as higher

voltage is applied. The TEM resolution r is restricted by the wavelength X in accordance

with the relationship below,

r oc 13/4 (3-20)

The wavelength can be deduced from the applied voltage, after the theory of relativity is

considered for determining the effective electron mass. In this research, the

microstructure of ZnMgS:Mn ACTFEL devices and the MgS segregation was identified

using TEM cross section. The TEM system used was a JOEL 200CX, operated at

200KV. Two modes were used in the analysis: image mode and diffraction mode. The

image mode included bright field mode and dark field mode. The bright field mode

generally gives a lower resolution than the dark field mode since the chromatic and

spherical aberration are minimized for the dark field mode. The diffraction mode

included selected area mode (SAD) and microdiffraction mode. The beam used for

diffraction should be parallel. However, when the interested sample area is larger than

the beam size, the beam size has to be reduced by compromising that the beam is not

parallel. This situation is called microdiffraction mode.














CHAPTER 4
PULSED LASER DEPOSITED Zn2GeO4:Mn


4.1 Introduction

Full-color ACTFEL displays require higher brightness of the green phosphors to

match the high brightness of the available red phosphors.17 So far, the best green

phosphor for ACTFEL displays is ZnS:Tb3 Oxide phosphors doped with Mn2 are

among the promising green phosphor candidates for ACTFEL devices.42 Oxide

phosphors have the features: 1) larger band-gap than that of sulfide phosphors;47 2)

chemically more stable than sulfide phosphors;61 3) less vacuum-restricted thin film

growth;69 and 4) high temperature processing required.42 These features could be good

and bad for the performance of ACTFEL devices. So far, the best green oxide phosphor

is Zn2Geo.5Sio.504:Mn.42 In this dissertation, Zn2GeO4:Mn was investigated. Its

processing temperature is the lowest among the available green oxide phosphors.42 So

far, the deposition techniques used for Zn2GeO4:Mn thin film growth include sputtering

deposition42 and sol-gel coating.69 There have been no reports of pulsed laser deposition

used for oxide ACTFEL devices. Pulsed laser deposition was used in an investigation of

photoluminescence of ZnGa204:Mn on single crystal substrates such as MgAl204.77

Pulsed laser deposition has been commonly used for oxide materials, especially for oxide

superconducting materials for a while.49 Pulsed laser deposition features the good

stoichiometry of the film compared to the target, short down time and low cost compared









with MBE (molecule beam epitaxy), and non-equilibrium growth due to the interaction of

the target with high laser energy.54

Experimental procedure of Zn2GeO4:Mn pulsed laser deposition has been given in

Section 3.3.2.. The experimental results of Zn2GeO4:Mn pulsed laser deposition include

the results from the Zn2GeO4:Mn target and from the Zn2GeO4:Mn films deposited at

different deposition conditions including laser energy density, ambient pressure, substrate

temperature and substrate material, and different annealing temperature and time. The

pulsed laser deposited Zn2GeO4:Mn were characterized by optical, electrical, structural

and chemical techniques.


4.2 Experimental Results



4.2.1 Zn2GeO4:Mn Target



4.2.1.1 X-ray Diffraction

The Zn2GeO4:Mn powder was pre-fired at 11500C for 24hrs. before hydraulically

pressed into a target and post-fired at 10500C for 24 hrs. after pressing. The crystal

structure of the Zn2GeO4:Mn target after firing was identified by x-ray diffraction. The x-

ray diffraction pattern of the Zn2GeO4:Mn target is shown in Figure 4-1 together with one

of the two reference Zn2GeO4 rhombohedral patterns. The x-ray diffraction pattern of the

Zn2GeO4:Mn target indicates that the Zn2GeO4 crystal structure is formed since all the

peaks from the target match well those of the Zn2GeO4 rhombohedral reference.

However, some of the peaks from target are shifted compared with the Zn2GeO4

rhombohedral reference peaks. Especially, these peaks at high 20 have a even larger shift







78


which is up to 20. The Zn2GeO4 is well crystallized after the two firing processes at

11500C and 1050C before and after pressing since the intensity for all the peaks is

relatively strong.



Zn2GeO4:Mn Target







Zn2eO4 Rohme
20 30 40 50 60 70
4 -
8 5














Figure 4-1 X-ray diffraction pattern of Zn2GeO4:Mn target compared with the pattern of
Zn2GeO4 rhombohedral reference36



4.2.1.2 PL and PLE Spectra

Photoluminescence emission spectrum (PL) and photoluminescence excitation

spectrum (PLE) of the Zn2GeO4:Mn target were measured. A monochromatized light

source was available for the PL and PLE measurement which were completely

automated. The PL spectrum is the emission spectrum at a constant excitation

wavelength. The PLE spectrum is the excitation spectrum at a constant emission

wavelength. Before an optimum excitation wavelength is identified, an excitation

wavelength 325nm was applied initially. Excited by 325nm excitation wavelength, the

Zn2GeO4:Mn target emits the green light and the PL emission spectrum was obtained










with a strong and broad peak at 540 nm. When the excitation wavelength is changed

from 325nm to 540nm, the emission peak intensity at 540nm is changed accordingly. The

excitation spectrum was therefore for the emission peak at 540nm. The excitation peak

was seen at 340nm The final PL emission spectrum was obtained at this optimum

excitation wavelength 340nm rather than at 325nm. The PL and PLE spectrum of the

Zn2GeO4:Mn target is shown Figure 4-2. One broad peak at 540nm is seen in the

photoluminescence emission spectrum of the Zn2GeO4:Mn target and one broad peak at

340nm is seen in the photoluminescence excitation spectrum of the target. The emission

peak wavelength and excitation peak wavelength is associated with the transition of the

luminescent center Mn2+ in host Zn2GeO4.



6.00E-02
Excitation spectrum Emission spectrum
5.00E-02


4.00E-02


3.00E-02
I,

2.00E-02
340r m 540nm
1.00E-02


O.OOE+00 --
200 250 300 350 400 450 500 550 600 650 700
X (nm)


Figure 4-2 Photoluminescence emission and excitation spectrum of Zn2GeO4:Mn target









4.2.1.3 PL of Ablated Zn2GeO4:Mn Target

After ablation by the pulsed laser, the originally white Zn2GeO4:Mn target

became dark. Oxygen loss at the target surface could be the reason for the color change.

Excited by a UV lamp, the photoluminescence of the unablated (fresh) Zn2GeO4:Mn

target was bright green color, while the PL of the ablated target was a dim green color.

The PL spectra for the two cases are shown in Figure 4-3. Both the unablated and the

ablated target show a broad peak at 540nm, which is in the green region. The peak from

the ablated Zn2GeO4:Mn target is much lower than that from the unablated target. The

small peak at 390nm is from the UV lamp background. The reduced PL intensity

probably result from the oxygen loss on the target surface after ablation.49




Unablated Zn2GeO4:Mn target








380 430 480 530 580 630 680 730 780
S(nm)

Figure 4-3 Photoluminescence emission spectrum of unablated and ablated
Zn2GeO4:Mn target under excitation of ultra-violet lamp



4.2.2 PZT Substrate

4.2.2.1 SEM

A PZT substrate was used for the Zn2GeO4:Mn ACTFEL inverted device

structure. This substrate is a 0.2mm A1203 ceramic sheet with a 2000A Au layer and a









20tlm Pb(Zro.5Tio.5)03 (PZT) layer. The yellowish PZT layer is comprised of two layers:

a thin bottom PZT layer prepared by spin-coating and a thick top PZT layer prepared by

screen printing for a total thickness of 20[tm. The screen printing PZT layer is expected

to result in a rough surface. The plain view and cross section of the PZT substrate were

obtained using SEM and they are shown in Figure 4-4. The thickness of PZT was first

measured by profilometry and 20[tm was measured from PZT layer step on Au layer. The

cross section SEM image of the PZT substrate further confirmed that the thickness of

PZT layer was 20[tm. The cross section SEM image indicates that the PZT layer is

porous. The A1203 substrate is even more porous. However, the A1203 substrate's

physical properties except its refractory nature are almost irrelevant to the optical

performance of Zn2GeO4:Mn ACTFEL devices. The primary reason to use A1203 as

substrate for Zn2GeO4:Mn ACTFEL devices was that it could be heated up to high

temperatures (up to 11500C). The top plain view SEM image of the PZT substrate

indicates that the PZT substrate surface is not smooth at all. Many bumps and valleys are

seen on surface of the PZT layer. Their average feature size is 5[tm Their depth however

cannot be is identified from the plain view image. The rough surface is likely the result of

the screen printing. The Au layer is not seen in the cross section of the PZT substrate

probably because it is too thin to be seen or the patterned Au area did not cross this

section of the cross section.



4.2.2.2 X-ray Diffraction

X-ray diffraction from the PZT substrate was obtained. Only part of the surface area of

the 2"x 2" A1203 substrate was covered by the Au layer and the PZT layer. The x-ray









diffraction pattern of the PZT substrate together with the Pb(Zr0.52Ti0.48)03 reference

pattern36 is shown in Figure 4-5.









S. ,:: :: ....* ..




Figure 4-4 Plain view and cross-section of substrate Al203/Au/Pb(Zro.5Tio.5)03



The x-ray diffraction pattern of the PZT substrate shows that the PZT layer is well

crystallized since all the peaks from the PZT layer are strong; the PZT layer peaks match

those from a Pb(Zr0.52Ti0.48)03 crystalline reference. It should be noted that all the peaks

from the PZT substrate come from the top 20jtm thick PZT layer since its thickness is

larger than the typical penetration depth of x-ray diffraction (10tm).75 This means that x-

ray don't penetrate into the PZT layer and no x-ray diffraction from the underlying Au or

the A1203 substrate were detected.



4.2.2.3 Composition

The energy dispersive x-ray spectrum was used to confirm the composition of the PZT

layer which was carbon coated before analysis. The EDX spectrum was obtained at

electron acceleration voltage 15kV and is shown in Figure 4-6. This data indicate that the

PZT layer contains Pb, Zr, Ti O and Si. The C is from the carbon coating. A Si peak is

observed at around 2.5keV.

































Figure 4-5 X-ray diffraction pattern of Pb(ZrTi)03, compared with Pb(Zr0.62Ti0.48)03
reference crystal structure



The small amount of Si in the PZT layer could be from the Si-based organic binder which

was typically used during screen printing.


Counts


Energy (keV)


0 5


Figure 4-6 Energy dispersive x-ray spectrum of the PZT layer from substrate
A1203/Au/PZT


o 0 0 0 0 -
o C C

S PZT i bstrate 60







20 30 40 2*0 50 60 70









4.2.3 Zn2GeO4:Mn Film Track-table

Zn2GeO4:Mn thin films were PLDed in two different systems, one at the

University of Florida and one at the Naval National Lab. Pulsed laser deposited films

from the University of Florida were primarily characterized by photoluminescence of

Zn2GeO4:Mn on silicon. Pulsed laser deposition at NRL were primarily characterized by

electroluminescence of Zn2GeO4:Mn on glass/ITO/ATO and on

Al203/Au/Pb(Zro.5Tio.5)03. The use of the PZT substrate was to withstand higher

annealing temperatures up to 9500C.(63) All pulsed laser deposited films used in the

chapter are listed in Table 4-1 for easy tracking.



4.2.4 Zn2GeO4:Mn on Si Substrate Temperature Issue

Processing of Zn2GeO4:Mn at >= 8000C is required for crystallization.42 A single

crystal (100) silicon wafer was initially used as substrate for Zn2GeO4:Mn thin film

growth. Si can be heated up to 11000C without substantial degradation or diffusion. 78

Another advantage of using silicon substrate is that many x-ray diffraction peaks from the

polycrystalline Zn2GeO4:Mn film can be easily isolated from the single crystal silicon

background peak(s) rather than from the complicated and possibly overlapped

polycrystalline PZT background peaks. It was expected that the film quality of

Zn2GeO4:Mn deposited on silicon would be much better than if deposited on

glass/ITO/ATO since the Zn2GeO4:Mn grows on the single crystal surface rather than on

the ATO amorphous surface. Two different substrate temperatures were used for

Zn2GeO4:Mn film growth, 2500C (Sample UF-Si-1) and 8000C (Sample UF-Si-2). All

other deposition parameters and the detailed deposition parameters are referred to Table

4-1.









Table 4-1 Tracking table of pulsed laser deposited Zn2Ge04:Mn at UF and NRL

Zn2GeO4 deposition Annealing Characterization
UF-Si-1 Si sub., 2500C 8000C, air, PL, XRD, SEM, EDX
5*10-5T, 200mT 02 24hrs
1.6J/cm2, 36,000 pulses
UF-Si-2 Si sub., 8000C No PL, XRD, SEM, EDX
5*10-5T, 200mT 02
1.6J/cm2, 36,000 pulses
UF-ATO-2 Si sub., 8000C No SEM, EDX
5*10-5T, 200mT 02
1.6J/cm2, 36,000 pulses
NRL-PZT-1 PZT sub., 2500C 7500C, 2.5hrs PL, EL*, XRD, SEM
5*105T, 200mT 02 N2 and air
0.8J/cm2, 10,000 pulses
NRL-PZT-2 PZT sub., 2500C 7700C, air, PL, EL*, XRD, SEM
5*105T, 150mT02 10 min.
0.8J/cm2, 10,000 pulses
NRL-PZT-3 PZT sub., RT 7500C, air, PL, EL*, XRD, SEM
5*105T, 150mT 02 2.5hrs.
0.8J/cm2, 10,000 pulses
NRL-PZT-4 PZT sub., 8000C No PL, EL*, XRD, SEM
5*10-5T, 150mT02
0.8J/cm2, 10,000 pulses
NRL-ATO-1 glass/ITO/ATO sub., RTA, Ar, PL, EL*, XRD, SEM
2500C Imin.
5*10-5T, 200mT 02
0.8J/cm2, 10,000 pulses
NRL-ATO-2 glass/ITO/ATO sub, RTA, Ar, PL, EL*, XRD, SEM
2500C Imin.
5*10-5T, 200mT 02
1.6J/cm2, 10,000 pulses
NRL-Si-1 Si sub., RT 8000C, air, Thickness, XRD
5*10-5T, 150mT02 21hrs.
0.8J/cm2, 10,000 pulses
NRL-Si-2 Si sub., 8000C No Thickness
5*10-5T, 150mT02
0.8J/cm2, 10,000 pulses
*EL device is half-cell structure and the top conducting layer is either Al dot for
glass/ITO/ATO substrate and or ITO for A1203/Au/PZT substrate.


Typically the number of the laser shots was given constant instead of laser repetition rate

and deposition time. The film thickness is a function of laser energy density and the









number of laser pulses. The film crystallization is generally dependent on the substrate

temperature. High substrate temperature dramatically enhances film crystallization during

growth.48 However, the film generally becomes less stoichiometric at high substrate

temperatures, which is especially true for pulsed laser deposition.48' 79,80 Zn2GeO4:Mn

films deposited at low temperatures tend to be amorphous. Re-crystallization is expected

upon an annealing at high temperatures. The films deposited at 2500C and then annealed

at 800C is compared with the film deposited at 8000C.



4.2.4.1 Photoluminescence

Photoluminescence of the Zn2GeO4:Mn films deposited at 2500C (UF-Si-1) and 800C

(UF-Si-2) was obtained under excitation of UV lamp and their PL spectra are shown

Figure 4-7. The film deposited at 2500C was annealed at 8000C in two different ambient

gases N2 and air. The PL spectrum of the Zn2GeO4:Mn target is plotted here for

comparison. Since the intensity of the target is so much stronger that of the films, the

insert shows intensity plotted on a larger scale to accommodate the whole PL spectrum of

the target. The first observation from Figure 4-7 is that both the films, annealed and

deposited at 8000C, demonstrate a broad PL emission peak at 540nm. Another

observation is that both films demonstrate a much lower PL intensity than that of the

target. These observations can be explained by that 1) the target surface is much rougher

than the films, which enhances light outcoupling; and 2) The amount of Zn2GeO4:Mn

contained in the target is much greater than that in the films and 3) The target is much

better crystallized than the films since it was fired at higher temperatures for a much

longer time. The interesting evidence here is that the film annealed at 8000C shows a

stronger photoluminescence than that deposited at 8000C. Crystallization of Zn2GeO4:Mn










deposited at 8000C should be definitely better than that deposited at 2500C. In this case,

the better photoluminescence for the annealed film should be contributed by the post-

annealing at 8000C for 24hrs which might have resulted in a good re-crystallization.

Cystallinity is the critical structural parameter for PL and EL materials. If the film is

poorly crystallized, the luminescent center is more likely to relax back to the ground state

by giving out the energy non-radiatively through grain boundaries or defects.


1.00E-03
9.00E-04 -Zn2GeO4:Mn Target
8.00E-04 -
7.00E-04 ilm depo ited at 2500C and
6.00E-04 nealed a 8000C in N2

= 4.00E-04 -
E 3.00E-04
2.00E-04
1.00E-04
0.OOE+00 -
400 450 500 550 600 650 700
? (nm)


Figure 4-7 Photoluminescence emission spectrum ofZn2GeO4:Mn film deposited at 1)
2500C (followed by an annealing at 8000C) (UF-Si-1) and 2) deposited at
800C without any annealing I(UF-Si-2)



The film annealed at 8000C in N2 was annealed a second time in air at 8000C. It has been

reported that oxygen vacancy played the shallow donor level to enhance excitation

efficiency of the luminescent center,73 which suggests that more oxygen in Zn2GeO4:Mn

is not necessarily good for optical performance. However, there was no significant

change in PL intensity after the second anneal in air.

There are several reasons that films annealed at 8000C demonstrate a much better

PL intensity than those deposited at 8000C. The hypotheses include 1) possible thickness









difference; 2) possible crystallinity difference 3) and possible surface roughness

difference. The x-ray diffraction can give a better picture how well a film is crystallized.

Regarding surface roughness difference, the substrate temperature difference could make

film surface roughness different. Re-crystallization during annealing would make the film

surface much rougher, compared with the as-deposited film.

The first hypothesis is thickness difference. No direct information is available to

compare the thickness of the two films Sample UF-Si-1 and UF-Si-2. The film deposited

at 250C was 10,000A given by profilometry using a step formed during deposition. No

thickness data was available for the film deposited at 8000C.

Zn2GeO4:Mn pulsed laser deposition on Si was also done at NRL. The substrate

temperatures were room temperature (NRL-Si-1) and 8000C(NRL-Si-2). In order to do

comparison with sputter deposited Zn2GeO4:Mn films which were all grown at room

temperature, the pulsed laser deposited Zn2GeO4:Mn film was especially grown at room

temperature. The deposition condition was slightly different from those from UF. The

system is different. The number of the laser shots here were 10,000 shots. While, UF

laser pulses in term of the laser repetition (10Hz) and the deposition time (30 min.), the

total laser shots should be 36,000 shots. This means even if the other parameters (the

position and the distance of substrate relative to the target) are exactly the same, the film

thickness should be only less than 1/3 of the UF films.

The photoluminescence was obtained from the films by using UV lamp excitation

source. The Zn2GeO4:Mn film deposited at R.T. didn't show any PL. This result is not

surprising. Since the x-ray diffraction of this film shown above is amorphous no

crystallization of Zn2GeO4:Mn should have been obtained during film growth. The










photoluminescence was obtained after the film was annealed at 8000C for 21 hours and

its PL spectra is shown in Figure 4-8. The broad peak at 540nm is seen in this film. The

background UV lamp reflection is plotted as a reference.


7.00E-04

6.00E-04

5.00E-04
E
S4.00E-04

S3.00E-04

S2.00E-04

1.00E-04

0.00E+-00
400 450 500 550 600 650 700
X(nm)

Figure 4-8 Photoluminescence spectrum of the Zn2GeO4:Mn film deposited on Si at
R.T.(NRL-Si-1) and annealed at 8000C for 21hrs.



4.2.4.2 X-ray Diffraction

X-ray diffraction from the two films and the Si substrate is shown in Figure 4-9.

The x-ray diffraction pattern of the Zn2GeO4:Mn target is shown for comparison. The

film deposited at 250C doesn't show any peaks except a huge peak at 690, which is the

Si (400) peak. This observation implies that Zn2GeO4:Mn deposited at 2500C is

amorphous. After being annealed at 8000C for 24hrs, the 250C deposited film exhibits

all the rhombohedral Zn2GeO4 peaks when compared with the rhombohedral

Zn2GeO4:Mn target. It should be noted that the x-ray diffraction peaks from the film are

slightly shifted to the right compared with those from the target. The 20 shift occurs

probably because the sample height was when x-ray diffraction was being detected. It









should be noted as well that Si (400) peak disappears in the 8000C annealed film, which

implies that the Zn2GeO4:Mn layer might have reacted with part of its underlying Si

substrate to form Zn2Gel-xSiO4:Mn which should demonstrate a much higher

photoluminescence or electroluminescence.42 The peaks from the film deposited at 8000C

are strong and match well the peaks from the target as well. This implies that

Zn2GeO4:Mn was well crystallized at substrate temperature 8000C. It is observed in this

film that the peak at 340 which is corresponding the Zn2GeO4 rhombehedral (321) crystal

plane is much stronger than any other peaks. This implies that even if Zn2GeO4 grown at

8000C is polycrystalline it is textured at (321) which is parallel to the wafer surface. It is

clear that rhombohedral Zn2GeO4 crystal structure is formed in both the 8000C annealed

and the 800C deposited films. The FWHM defines of the 20 = 340 (321) peak for

crystallinity comparison. The A20 for the film deposited at 8000C is smaller than that for

the film annealed at 8000C. This implies that the film deposited at 8000C is slightly better

crystallized than that annealed at 8000C. Therefore, in term of crystallinity, the film

deposited at 8000C should demonstrate a stronger PL than that annealed at 8000C.

However, the result is reversed. The x-ray diffraction pattern of the films on Si (Sample

NRL-1) is shown in Figure 4-10. One large peak appears in the x-ray diffraction pattern

of the film deposited at room temperature. The Zn2GeO4:Mn film deposited at 2500C

given in Figure 4-9 was amorphous. It is obvious that the film deposited at R.T. is

amorphous too. This large peak should be from the single crystal Si substrate. Upon an

annealing at 8000C for 21hrs., the Si peak height is reduced and the new peaks appear.

Compared with the peaks from the Zn2GeO4:Mn target which is in the Zn2GeO4










rhombehedral structure, the major Zn2GeO4 rhombohedral peaks appear in the annealed

film.



Zn2GeO4:Mn target



Si substrate

0
2 Film deposited at 2500C

Film deposited at 2500C and annealed at 8000C


Film deposited at 800C

20 25 30 35 40 45 50 55 60 65 70
20



Figure 4-9 X-ray diffraction pattern of pulsed laser deposited Zn2GeO4:Mn at substrate
temperature 1) 2500C (UF-Si-1) and 2) then annealed at 8000C, and 3) deposited at 8000C
(UF-Si-2)



The reduced Si peak intensity implies that the underlying Si might have reacted with

Zn2GeO4 during annealing. Re-crystallization during annealing definitely occurred since

several new peaks appear. However, the re-crystallization is not thoroughgoing since the

intensity of the new peaks appears not strong.




4.2.4.4 Surface Topography

Another possible reason for difference in PL intensity is difference of film surface

roughness which would result in the difference of light out-coupling.46 The surface

morphology was imaged by SEM as shown in Figure 4-11. The film annealed at 8000C




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs