<%BANNER%>

In Situ X-Ray Diffraction Based Investigation of Crystallization in Solution Deposited PZT Thin Films

Permanent Link: http://ufdc.ufl.edu/UFE0044015/00001

Material Information

Title: In Situ X-Ray Diffraction Based Investigation of Crystallization in Solution Deposited PZT Thin Films
Physical Description: 1 online resource (272 p.)
Language: english
Creator: Nittala, Krishna
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: diffraction -- films -- pzt -- thin
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Solution deposited PZT based thin films have potential applications in embedded decoupling capacitors and pulse discharge capacitors. During solution deposition, precursor solution is deposited onto a substrate to obtain an amorphous film. The film is then crystallized by heating it at a high temperature (~600 - 700°C). Conditions during the crystallization anneal such as precursor stoichiometry in solution, heating rate and adhesion layer in the substrate are known to influence phase and texture evolution in these films. However, a mechanistic understanding of the changes taking place in these thin films during crystallization is lacking. A better understanding of the crystallization processes in these thin films could enable tailoring the properties of thin films to suit specific applications. To explore the crystallization process in solution deposited PZT thin films, high temperature in situ laboratory and synchrotron X-ray diffraction based techniques were developed. Taking advantage of the high X-ray flux available at synchrotron facilities such as beamline 6-ID-B, Advanced Photon Source, Argonne National Laboratory, crystalline phases formed in the thin films during crystallization at the high heating rates (0.5 – 60°C/s)typically used during film processing could be measured. Using a 2-D detector for these measurements allowed the simultaneous measurement of both phase and texture information during crystallization. Analytical treatment of the unconventional diffraction geometry used during the synchrotron based measurements was performed to develop methodologies for quantitative estimation of texture components. The nominal lead content in the starting solutions and the heating rate used during crystallization was observed to influence the sequence of phases formed during crystallization of the films. In films crystallized at fast heating rates,titanium segregation, probably due to diffusion of titanium from the adhesion layer, was observed. To further investigate the effect of adhesion layer on the crystallization behavior of these films, PZT films were solution deposited onto substrates with titanium (Ti), titanium oxide (TiOx) and zinc oxide (ZnO) adhesion layers. Phase evolution was observed to be unaffected by the adhesion layer. (111) oriented PZT films were obtained on all substrates and hence Ti diffusion from the adhesion layer appeared to have limited influence on texture control in these films. It is suggested that the (111) orientation of the perovskite phase is directly seeded by the Pt(111) texture.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Krishna Nittala.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Jones, Jacob L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044015:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044015/00001

Material Information

Title: In Situ X-Ray Diffraction Based Investigation of Crystallization in Solution Deposited PZT Thin Films
Physical Description: 1 online resource (272 p.)
Language: english
Creator: Nittala, Krishna
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: diffraction -- films -- pzt -- thin
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Solution deposited PZT based thin films have potential applications in embedded decoupling capacitors and pulse discharge capacitors. During solution deposition, precursor solution is deposited onto a substrate to obtain an amorphous film. The film is then crystallized by heating it at a high temperature (~600 - 700°C). Conditions during the crystallization anneal such as precursor stoichiometry in solution, heating rate and adhesion layer in the substrate are known to influence phase and texture evolution in these films. However, a mechanistic understanding of the changes taking place in these thin films during crystallization is lacking. A better understanding of the crystallization processes in these thin films could enable tailoring the properties of thin films to suit specific applications. To explore the crystallization process in solution deposited PZT thin films, high temperature in situ laboratory and synchrotron X-ray diffraction based techniques were developed. Taking advantage of the high X-ray flux available at synchrotron facilities such as beamline 6-ID-B, Advanced Photon Source, Argonne National Laboratory, crystalline phases formed in the thin films during crystallization at the high heating rates (0.5 – 60°C/s)typically used during film processing could be measured. Using a 2-D detector for these measurements allowed the simultaneous measurement of both phase and texture information during crystallization. Analytical treatment of the unconventional diffraction geometry used during the synchrotron based measurements was performed to develop methodologies for quantitative estimation of texture components. The nominal lead content in the starting solutions and the heating rate used during crystallization was observed to influence the sequence of phases formed during crystallization of the films. In films crystallized at fast heating rates,titanium segregation, probably due to diffusion of titanium from the adhesion layer, was observed. To further investigate the effect of adhesion layer on the crystallization behavior of these films, PZT films were solution deposited onto substrates with titanium (Ti), titanium oxide (TiOx) and zinc oxide (ZnO) adhesion layers. Phase evolution was observed to be unaffected by the adhesion layer. (111) oriented PZT films were obtained on all substrates and hence Ti diffusion from the adhesion layer appeared to have limited influence on texture control in these films. It is suggested that the (111) orientation of the perovskite phase is directly seeded by the Pt(111) texture.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Krishna Nittala.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Jones, Jacob L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044015:00001


This item has the following downloads:


Full Text

PAGE 1

1 IN SITU X RAY DIFFRACTION BASED INVESTIGATION OF CRYSTALLIZATION IN SOLUTION DEPOSITED PZT THIN FILMS By KRISHNA NITTALA 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 2012

PAGE 2

2 2012 K rishna Nittala

PAGE 3

3 To my grandparents

PAGE 4

4 ACKNOWLEDGMENTS It is said that it takes a village to make a man. C ompletion of my PhD required resources from a major university and two national laboratories. I am grateful for the kindness showed by a great number of people during this time and would lik e to recognize their generosity H owever, I am limited by the cons traints imposed by space and time. As a humble attempt to remain in the kind graces of the people who ha ve helped me during this stint and whose kindness has not been personally acknowledged I offer the following compromise : I will gladly buy you a coffee (or two). I am grateful to m y committee chair and advisor, Dr. Jacob L. Jones, for providing me the opportunity to work in his group and giving me a free reign in choosing the direction for research. so friendly ping pong tournaments during the socials provided much needed breaks from science. For these fun times and help with research and presentations, I would like to thank Dr. Jones and the Dr. Jones research group. This work hugely benefitted from the collaboration with Sandia National Laboratories. In particular Drs. Geoff L. Brennecka, Jon F. Ihlefeld and Bruce A. Tuttle from Sandia National Laboratories provided valuable scientific input and many thin film samples for measurement s I would also like to thank them for hosting me at Sandia National Laboratories, teaching me the solution deposition techniques and for kindly putting up with my mistakes during my visits I am also grateful to the NINE consortium (National Insititute for Nanotechnolog ies) for providing funding for this research. I would also like to acknowledge Dr. Douglas S. Robinson at beamline 6 ID B, Advanced Photon Source, Argonne National Laboratories for helping with the in situ diffraction experiments.

PAGE 5

5 Training provided by Dr. Valentin Craciun on various X ray diffractometers helped immensely in improving my understanding of X ray diffraction. I am also grateful to m y other committee members Drs. Amlan Biswas, Simon Phillpot and Scott Perry for their trust in my ability to carry out this research. I was under the supervision of Dr. Henry Hess (now at Columbia University) for the first 18 months of my doctoral studies. The many discussions during group meeting s and journal clubs in the Dr. Hess research group helped me in sha rpen ing my acumen for research. For this and the personal support, I am indebted to Dr. Hess and the former Hess research group. I was lucky to have met many wonderful people in Gainesville and Albuquerque. In particular, I am grateful to Bea Doyle for let ting me stay at her place during my visits to A lbuquerque. Her gentl y goading me to practice yoga has ha d a profound effect on my life and has helped me manage stress levels during the frenzied thesis writing stage. The inspiration, love and support provid ed by my family helped me maintain vitality both towards science and life during my doctoral studies In particular, I would like to thank my grandparents who made great personal sacrifice s for the educat ion of their children. T he high value placed by my g randparents on education is the inspir ation for this academic endeavor.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDG MENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 18 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 20 1.1 Chemical Solution Deposition of PZT Thin Films ................................ ........... 22 1.1.1 Solution Processing Routes ................................ ................................ 23 1.1.1.1 Sol gel based routes ................................ ............................... 23 1.1.1.2 Metallo organic decomposition ................................ ............... 24 1.1.1.3 Hybrid chelate based routes ................................ ................... 24 1.1.2 Inhomogeneity in Solutions ................................ ................................ .. 26 1.2 Deposition ................................ ................................ ................................ ...... 27 1.2.1 Spin Coa ting ................................ ................................ ........................ 27 1.2.2 Platinized Silicon Substrates ................................ ............................... 27 1.3 Phase Evolution ................................ ................................ ............................. 30 1.4 Texture of Crystallized Films ................................ ................................ .......... 33 1.4.1 Homogenous Vs Heterogenous Nucleation ................................ ......... 34 1.4.2 Seeding of (111) Orientation by Pt x pb ................................ ................. 35 1.4.3 Transformation of Fluorite Phase ................................ ........................ 36 1.4.4 Ti Diffusion from Adhesion Layer ................................ ......................... 36 1.5 Motivation and Scope of the Present Work ................................ .................... 37 1.6 Outline ................................ ................................ ................................ ............ 39 1.7 Conclusion ................................ ................................ ................................ ..... 39 2 DETAILS OF EXPERIMENTAL METHODS ................................ ........................... 45 2.1 IMO Solution Preparation and Deposition ................................ ...................... 45 2.1.1 Standard Processing ................................ ................................ ........... 45 2.1.2 Processing for Ceria Dispersed PZT Thin Films ................................ .. 46 2.2 Preparation of Platinized Silicon Substrates ................................ ................... 46 2.2.1 Pt/TiO x /SiO 2 /Si ................................ ................................ ..................... 47 2.2.2 Pt/ZnO/SiO2/Si ................................ ................................ .................... 47 2.3 In Situ Synchrotron X Ray Diffraction Setup ................................ .................. 47 3 IN SITU X RAY DIFFRACTION SETUPS: DIFFRACTION GEOMETRY AND DATA EXTRACTION METHODOLOGIES ................................ .............................. 52

PAGE 7

7 3.1 Texture Measurement using X Ray Diffraction ................................ ............... 52 3.2 Laboratory based In Situ X Ray Diffraction Setup ................................ .......... 54 3.3 In Situ Synchrotron X Ray based Setup ................................ ......................... 54 3.4 Data Reduction and Representation ................................ .............................. 56 3.5 Theoretical Absorption Correction ................................ ................................ .. 57 3.6 Correction for Tilting of the Sample ................................ ................................ 59 3.7 Section of Pole Figure Sampled ................................ ................................ ..... 61 3.7.1 Determination of hkl Texture Components using non (hkl) Diffraction Intens ities ................................ ................................ ........... 64 3.7.2 Plane Multiplicities and Angles between Planes ................................ .. 65 3.8 Calculation of Texture Component Volume and Volume Fraction .................. 68 3.9 Temperature Calibration ................................ ................................ ................. 69 4 PHASE EVOLUTION IN P b O DEFICIENT SOLUTION DEPOSITED PLZT THIN FILMS ................................ ................................ ................................ ..................... 84 4.1 Introduction ................................ ................................ ................................ .... 84 4.2 Experimental Methodologies ................................ ................................ .......... 86 4.2.1 Film Deposition ................................ ................................ .................... 86 4.2.2 Laboratory based In Situ Diffraction Setup ................................ .......... 87 4.3 Results ................................ ................................ ................................ ........... 90 4.4 Discussion ................................ ................................ ................................ ...... 92 4.5 Conclusion s ................................ ................................ ................................ .... 96 5 AGING IN PRECRYSTALLIZED SOLUTION DERIVED PZT THIN FILMS .......... 106 5.1 Experimental Methods ................................ ................................ .................. 106 5.2 Results ................................ ................................ ................................ ......... 108 5.2.1 Electrical Characterization ................................ ................................ 108 5.2.2 FTIR Measurements ................................ ................................ .......... 110 5.2. 3 Microstructural Characterization using SEM ................................ ...... 111 5.2.4 XRD Measurements ................................ ................................ .......... 113 5.3 Discussion ................................ ................................ ................................ .... 114 5.3.1 Aging of Thin Films ................................ ................................ ............ 114 5. 3.2 Solution Derived Powders ................................ ................................ 118 5.3.2.1 FTIR ................................ ................................ ...................... 118 5.3.2.2 Thermogravimetric analysis ................................ .................. 119 5.4 Conclusions ................................ ................................ ................................ .. 120 6 EFFECT OF HEATING RATE ON PHASE AND TEXTURE EVOLUTION ........... 145 6.1 Experimental Methods ................................ ................................ .................. 145 6.2 Results ................................ ................................ ................................ ......... 146 6.2.1 Phase Evol ution ................................ ................................ ................. 146 6.2.2 Texture Evolution ................................ ................................ ............... 147 6.2.3 SEM and STEM EDS ................................ ................................ ........ 148 6.3 Discussion ................................ ................................ ................................ .... 149

PAGE 8

8 6.4 Conclusion ................................ ................................ ................................ ... 154 6.5 Acknowledgements ................................ ................................ ...................... 154 7 FORMATION OF PT X PB ................................ ................................ ...................... 164 7.1 In troduction ................................ ................................ ................................ .. 164 7.2 Experimental ................................ ................................ ................................ 165 7.3 Results and Discussion ................................ ................................ ................ 166 7.4 Conclusion ................................ ................................ ................................ ... 170 8 EFFECT OF ADHESION LAYER ON PHASE AND TEXTURE EVOLUTION ...... 181 8.1 Experimental Methods ................................ ................................ .................. 181 8.1.1 Preparation of Substrates ................................ ................................ .. 181 8.1.2 APS Setup ................................ ................................ ......................... 182 8.2 Results ................................ ................................ ................................ ......... 183 8.3 Discussion ................................ ................................ ................................ .... 186 8.4 Conclusion ................................ ................................ ................................ ... 190 9 SUMMARY ................................ ................................ ................................ ........... 211 APPENDIX A MATLAB CODES FOR DATA REDUCTION ................................ ........................ 212 A.1 Codes for Integration over ................................ ................................ ......... 212 A.2 Matlab Codes for Integrating over 2 ................................ ........................... 224 B ANALYTICAL CORRECTION FOR BRAGG PLANE DISPLACEMENT IN SYNCHROTRON IN SITU DIFFRACTION SETUP ................................ .............. 235 C MARCH 2010 EXPERIMENTAL RUN ................................ ................................ .. 241 C.1 Phase Evolution Plots ................................ ................................ .................. 241 C.2 (100) Azimuthal Time Plots ................................ ................................ .......... 242 D OCTOBER 2011 EXPERIMENTAL RUN ................................ .............................. 243 D.1 PZT/Pt/Ti/SiO2/Si (SQI) ................................ ................................ ............... 243 D.1.1 Phase Evolution Plots ................................ ................................ ........ 243 D.1.2 (100) Azimuthal Time Plots ................................ ................................ 245 D.2 PZT/Pt/TiOx/SiO2/Si (Sandia) ................................ ................................ ...... 247 D.2.1 Phase Evolution Plots ................................ ................................ ........ 247 D.2.2 (100) Azimuthal Time Plots ................................ ................................ 249 D.3 PZT/Pt/ZnO/SiO2/Si (Sandia) ................................ ................................ ...... 251 D.3.1 Phase Evolution Plots ................................ ................................ ........ 251 D.3.2 (100) Azimuthal Time Plots ................................ ................................ 253 D.4 PZT/Pt/TiOx/SiO2/Si (ARL) ................................ ................................ .......... 255

PAGE 9

9 D.4.1 Phase Evolution Plots ................................ ................................ ........ 255 D.4.2 (100) Azimuthal Time Plots ................................ ................................ 257 E (110) AZIMUTHAL SCAN PEAK FITTING RESULTS ................................ .......... 259 E.1 PZT/Pt/Ti/SiO2/Si (SQI) ................................ ................................ ............... 259 E.2 Pt/TiO x /SiO 2 /Si ................................ ................................ ............................. 260 E.3 PZT/Pt/ZnO/SiO2/Si ................................ ................................ ..................... 261 E.4 PZT/Pt/TiO 2 /SiO 2 /Si (ARL) ................................ ................................ ........... 262 LIST OF REFERENCES ................................ ................................ ............................. 263 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 272

PAGE 10

10 LIST OF TABLES Table page 3 1 Mass absorption coefficients of the different elements at 20 keV [135] ............. 73 3 2 Angles between the (110) type planes and (100) and (111) planes. Planes with angles greater than 90 are not listed as they are not detected in the diffraction geometry used at APS ................................ ................................ ....... 73 3 3 Temperature of the film measured using a thermocouple at different voltage settings on the lamp ................................ ................................ ........................... 74 4 1 Lattice parameters of the different perovskite phases formed in the thin films investigated. Perovskite 1 and perovskite 2 refer to the perovskite phases formed in Pb deficient thin films.. ................................ ................................ ....... 98 7 1 Calculated powder diffraction reflection positions for PtPb at = 0.5462 angstroms. The space group of the material is P6 3 /mmc, with a = 3.358, b = 3.358 and c = 4.0580 angstroms. ................................ ................................ ..... 171 7 2 Plane indices and 2 values expected in the powder pattern for Pt 3 Pb at = 0.5462 angstroms. The 2 values were calculated using Crystal Diffract .. ... 171 8 1 Average heating rates corresponding to different voltage rates used during the in situ diffraction experiments. ................................ ................................ .... 191 B 1 z height and eta matrix used for validating the derived analytical model .......... 238

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Typical hysteresis loop obtained for a ferroelectric material ............................... 41 1 2 Equilibrium phase diagram of PZT. Reproduced from [10] ................................ 41 1 3 Typical steps involved in the solution deposition of thin films ............................. 42 1 4 Different reactions taki ng place during sol gel preparation of PZT solutions ...... 42 1 5 Steps in the IMO solution preparation route ................................ ....................... 43 1 6 Additional reactions taking place during hybrid chelate solution preparation route ................................ ................................ ................................ ................... 43 1 7 Phases present in solution deposited thin films during various stages of thermal processing. Adapted from [84] ................................ .............................. 44 1 8 Heterogeneous and homogenous nucleation of a crystalline region in an amorphous film. ................................ ................................ ................................ .. 44 2 2 Experimental setup used at beamline 6 ID B, APS during beamtime access in March, 2010. ................................ ................................ ................................ ... 51 2 3 Experimental setup used at beamline 6 ID B, APS during beamtime access in November, 2010. ................................ ................................ ............................ 51 3 1 Euler cradle used for measurement of texture in materials. ............................... 75 3 2 Variation in the direction of with for the Inel diffractometer diffraction geometry. Radial direction represents absolute value of the momentum transfer vector. ................................ ................................ ................................ .... 75 3 3 Experimental setup used for the in situ synchrotron X ray measurements taken at beamline 6 ID B, Advanced Photon Source, Argonne National Laboratories. The blue line indicates the X ray beam path. ................................ 76 3 4 Definition of detector angles and schematics for data extraction.. ...................... 76 3 5 Data representations of the in situ diffraction data. ................................ ............. 76 3 6 Definition of terms for calculation of Absorption correction. ................................ 77 3 7 Absorption correction for different film thicknesses. ................................ ........... 77 3 8 Rotation of diffraction cones due to tilting of the sample in the YZ plane. The position of the segregated intensities changes due to tilting. .............................. 78

PAGE 12

12 3 9 Schematic showing the diffraction geometry for the APS exper iments. .............. 78 3 10 Definition of the pole figure angles and based on the sample coordinates. .. 79 3 11 Section of (100) pole figure measured in the APS in situ diffraction experiments. ................................ ................................ ................................ ....... 79 3 12 Pole figures corresponding to different types of texture. For the sake of illustration, only (100), (110) and (111) pole figures of (100), (110) and (111 ) type textures are represented.. ................................ ................................ ........... 80 3 13 (110) simulated pole figure for mixed (100) and (111) texture mode. ................. 80 3 14 Representative result from the peak fitting procedure. ................................ ....... 81 3 15 Variation of apparent lattice parameter with change in temperature measured on two different samples.. ................................ ................................ ................... 81 3 16 Methodology used for the determination of the temperature of the thin film at different times during the crystallization heat treatment. ................................ ..... 82 3 17 Contour plot showing phase evolution during in situ crystallization of PZT(52/48) film on Pt/Ti substrate. The change in the temperature of the film is sh own in the plot in the right. ................................ ................................ .......... 83 4 1 Laboratory based setup used for in situ crystallization of PZT thin films ............ 99 4 2 Cross section of the thin film stack and interaction of the X rays with the thin films ................................ ................................ ................................ .................. 100 4 3 Attenuation of the X ray intensity due to absorption by the different layers ...... 100 4 4 Phase evolution in PLZT + 20% Pb excess film ................................ ............... 101 4 5 Limited 2 surface plot showing the evolution of the fluorite and perovskite phases ................................ ................................ ................................ .............. 101 4 6 Phase evolution in Pb deficient thin film ................................ ........................... 102 4 7 Limited 2 theta plot showing the evolution of the fluorite and perovskite phases. A secondary perovskite phase is observed to fo rm. ............................ 102 4 8 Phase evolution during crystallization of PLZT + 20% excess PbO thin film .... 103 4 9 Phase evolution during crystallization of PLZT + 20% PbO deficient film ......... 104 4 10 GIXRD plots for the Pb deficient thin films at different angles of incidence ...... 104

PAGE 13

13 4 11 Normalized integrated intensities obtained at increasing x ray incidence angles ................................ ................................ ................................ ............... 105 5 1 Polarization vs electric field measurements for films pyrolyzed at 300 C and stored in different RH conditions.. ................................ ................................ ..... 122 5 2 Polarization electric field measurements for films pyrolyzed at 400 C and aged in different RH conditions.. ................................ ................................ ....... 122 5 3 C vs F measurements for films pyrolyzed at 300 C and aged in different RH conditions.. ................................ ................................ ................................ ....... 123 5 4 C vs F measurements for films pyrolyzed at 400 C and aged in different RH conditions.. ................................ ................................ ................................ ....... 124 5 5 Variation of relative dielectric constant and tan for films pyrolyzed at 300 C with aging time. Measurements were performed at 100 Hz. ............................. 125 5 6 Films pyrolyzed at 400 C: variation of relative dielectric constant and tan with aging time ................................ ................................ ................................ .. 126 5 7 FTIR spectrum of films pyrolyzed at 300 C and aged in different RH conditions. A) 20%, B) 44%, C) 55%, and D) 98% relative humidity. ............... 127 5 8 FTIR spectrum of films pyrolyzed at 400 C and aged in different RH conditions. ................................ ................................ ................................ ........ 128 5 9 Films pyrolyzed at 300 C and aged in 20% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times.. ................ 129 5 10 Films pyrolyzed at 300 C and aged in 44% RH: SEM plan view and cross section micrographs of crystallized films after specific aging tim es.. ................ 130 5 11 Films pyrolyzed at 300 C and aged in 55% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times.. ................ 131 5 12 Films pyrolyzed at 300 C and aged in 98% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times.. ................ 132 5 13 Films pyrolyzed at 400 C and aged in 20% RH: SEM plan view and cross se ction micrographs of crystallized films after specific aging times.. ................ 133 5 14 Films pyrolyzed at 400 C and aged in 44% RH: SEM pl an view and cross section micrographs of crystallized films after specific aging times.. ................ 134 5 15 Films pyrolyzed at 400 C and aged in 55% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times.. ................ 135

PAGE 14

14 5 16 Films pyr olyzed at 400 C and aged in 98% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times.. ................ 136 5 17 XRD measurements of films pyrolyzed at 300 C and crystallized after aging for specific times. ................................ ................................ .............................. 137 5 18 Lotgering f 111 factor values for films pyrolyzed at 300 C and crystallized after different aging times ................................ ................................ ......................... 138 5 19 XRD measurements of films pyrol yzed at 400 C and crystallized after aging for specific times. ................................ ................................ .............................. 139 5 20 Lotgering f 111 factor values for films pyrolyzed at 400 C and crystallized after different aging times ................................ ................................ ......................... 140 5 21 Different modes of coordination between the metal ion and the acetate ions. Adapted from [144] ................................ ................................ .......................... 140 5 22 Variation in the full wid th at half maximum (FWHM) of s (C=O) with aging time in films pyrolyzed at 300 C. Error bars correspond to 95% confidence interval. ................................ ................................ ................................ ............. 141 5 23 Variation in the FWHM of s (C=O) of films pyrolyzed at 400 C and aged in 98% relative humidity. Error bars indicate 95% confidence intervals. ............... 142 5 24 FTIR spectra of powders aged in aged for different times in 98% RH.. ............ 142 5 25 Variation of FW HM with aging time of the a (CO 2 ) resonance peak in solution derived powders pyrolyzed 300 C and 400 C. ................................ .... 143 5 26 Thermogravimetry plots for powders. The masses were normalized with respect to the final mass of the powders.. ................................ ........................ 143 5 27 Limited t emperature thermogravimetry plots for solution derived powders.. ..... 144 6 1 Phase evolution during crystallization of solution depos ited PZT thin films. ..... 155 6 2 Variation of integrated intensity of the crystalline phases formed in solution deposited PZT thin fi lms.. ................................ ................................ ................. 156 6 3 (110) AT plots for films crystallized at different heating rates. indicate intensities from (111) type texture and arrows indicate (100) type texture.. ..... 157 6 4 Variation of texture in films in situ crystallized at different heating rates ........... 158 6 5 SEM micrographs of films crystallized at 20 C/s.. ................................ ............ 158 6 6 SEM micrographs of films crystallized at 3 C/s.. ................................ .............. 159

PAGE 15

15 6 7 SEM micrographs of films crystallized at 1 C/s.. ................................ .............. 159 6 8 SEM micrographs of films crystallized at 0.2 C/s.. ................................ ........... 159 6 9 STEM EDS compositional maps of films cry stallized at 20 C/s ........................ 160 6 10 STEM EDS compositional maps of films crystallized at 3 C/s .......................... 160 6 11 STEM EDS compositional maps of film crystallized at 1 C/s ........................... 161 6 12 STEM EDS compositional maps of film crystallized at 0.2 C/s ........................ 161 6 13 Change in the latti ce parameter of perovskite PZT after formation during crystallization.. ................................ ................................ ................................ .. 162 6 14 Variation of the FWHM of the fluorite (111) peak with heating rate .................. 163 6 15 Variation of pole intensity with azimuthal angle ( ) for the fluorite phase formed during crystallizati on at different heating rates ................................ ..... 163 7 1 Thermodynamic stability of the constituent metals in PZT towards oxidation and reduction ................................ ................................ ................................ .... 172 7 2 Phase evolution showing the formation of Pt x Pb metastable phase and the final perovskite phase (Pe) during in situ crystallization in the Inel diffractometer. ................................ ................................ ................................ ... 172 7 3 Results on in situ crystallization experiment performed at 64 C/s .................... 173 7 4 Results on in situ crystallization experiment performed at 68 C/s. ................... 173 7 5 Results on in situ crystallization experiment performed at 32 C/s .................... 174 7 6 Results on in situ crystallization experiment performed at 11 C/s .................... 174 7 7 Variation of maximum normalized Pt x Pb intensity with decrease in heating rate. Error bars indicate 95% confidence intervals. ................................ .......... 175 7 8 2 plot for film crystallized at 64 C/s. ................................ ........................... 175 7 9 2 plot for films crystallized at 68 C/s.. ................................ ........................ 176 7 10 2 plot for films crystallized at 32 C/s. ................................ ......................... 176 7 11 2 plot for films crystallized at 11 C/s.. ................................ ........................ 177 7 12 (111) pole figure of Pt substrate in the quenched sample ................................ 177

PAGE 16

16 7 13 (111) pole figure of Pt x Pb phase in the quenched sample ................................ 178 7 14 Pt Pb binary phase diagram [160] ................................ ................................ 178 7 15 Crystal structure of Pt 3 Pb phase (JCPDF#06 0574) ................................ ......... 179 7 16 Stacking of Pt(111) and Pt 3 Pb(111) planes showing sim ilarities in atomic arrangements. ................................ ................................ ................................ .. 179 7 17 Diffraction patterns of the quenched in Pt 3 Pb phase in solution deposited PZT thin film measured at sequentially increasing aging times. ....................... 180 7 18 Proposed sequence for the formation and disappearance of the Pt 3 Pb phas e. 180 8 1 FWHM of different Pt peaks along ................................ ................................ 192 8 2 Key for interpretation of phase evolution plots.. ................................ ................ 193 8 3 Phase evolution plots of PZT films on different substrates crystallized at 64 C/s (inst), 68 C/s and 32 C/s. ................................ ................................ ..... 194 8 4 Phase evolution plots of PZT films on different substrates crystallized at 11 C/s, 1 C/s and 0.5 C/s. ................................ ................................ ............... 195 8 5 (110) azimuthal time (AT) plots for PZT films on different substrates and crystallized at 64 C/s (inst), 68 C/s and 32 C/s.. ................................ .............. 196 8 6 (110) azimuthal time (AT) plots for PZT films on different substrates and crystallized at 11 C/s, 1 C/s and 0.5 C/s.. ................................ ........................ 197 8 7 Variation of (111), (100) and random texture components in PZT thin films deposited on Pt/Ti substrates and crystallized at different heating rates. ......... 198 8 8 Variation of (111), (100) and random texture components in PZT thin films deposited on Pt/TiO x substrates and crystallized at different heating rates. ..... 198 8 9 Variation of (111), (100) and random texture components in PZT t hin films deposited on Pt/ZnO substrates and crystallized at different heating rates. ..... 199 8 10 Variation in the FWHM of the 111 an d 100 texture components with heating rate during crystallization for films deposited on Pt/Ti substrates. .................... 199 8 11 Variation in the FWHM of the 111 and 100 texture components with heating rate during crystallization for films deposited on Pt/TiO x substrates. ................ 200 8 12 Variation in the FWHM of the 111 and 100 texture components with heating rate during crystallization for films deposited on Pt/ZnO substrates. ................ 200

PAGE 17

17 8 13 Plan view and cross section SEM images for PZT films deposited on Pt/Ti substrates.. ................................ ................................ ................................ ....... 201 8 14 Plan view and cross section SEM images for PZT films deposited on Pt/Ti substrates.. ................................ ................................ ................................ ....... 202 8 15 Plan view and cross section SEM images for PZT films deposited on Pt/TiO x substrates.. ................................ ................................ ................................ ....... 203 8 16 Plan view and cross section SEM i mages for PZT films deposited on Pt/TiO x substrates.. ................................ ................................ ................................ ....... 204 8 17 Plan view and cross section SEM images for PZT films deposited on Pt/ZnO substrates.. ................................ ................................ ................................ ....... 205 8 18 Plan view and cross section SEM images for PZT films deposited on Pt/ZnO substrates.. ................................ ................................ ................................ ....... 206 8 19 Results of peak fitting of 100 and 111 texture components in PZT thin films deposited on SQI substrates and heated at 0.5 C/s. ................................ ........ 207 8 20 Intensity maps (I(2 )) at various times during crystallization of solution deposited PZT thin film deposited on platinized silicon substrates with titanium adhesion layer (SQI).. ................................ ................................ ......... 208 8 21 Results of in situ experiment performed on PZT films deposited on Pt/TiO 2 (ARL) substrates and crystallized at 50 C/s.. ................................ ................... 209 8 22 Variation in the volume fraction of different texture components with heating rate for PZT films deposited on Pt/TiO 2 (ARL) substrates. ............................... 209 8 23 Variation in the sharpness of different texture components with heating rate for PZT films deposited on Pt/TiO 2 (ARL) substrates. Error bars indicate 95% confidence intervals ................................ ................................ .......................... 210 B 1 Definition of terms for analytical flat plate correction ................................ ........ 239 B 2 Variation of uncorrected lattice parameters of the (111) C eO 2 peak ................ 239 B 3 Calculated lattice parameters using the calculated analytical function ............. 240

PAGE 18

18 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 IN SITU X RAY DIFFRACTION BASED INVESTIGATION OF CRYSTALLIZATION IN SOLUTION DEPOSITED PZT THIN FILMS By Krishna Nittala August 2012 Chair: Jacob L. Jones Major: Materials Science and Engineering Solution deposited PZT based thin films have potential applicati on s in embedded decoupling capacitors and pulse discharge capacitors During solution deposition, precursor solution is deposited o nto a substrate to obtain an amorphous film. The film is then crysta llized by heating it at a high temperature (~600 700 C). Conditions during the crystallization anneal such as precursor stoichiometry in solution, heating rate and adhesion layer in the substrate are known to influence phase and texture evolution in these films However, a mechanistic understanding of the ch anges taking place in these thin films during crystallization is lacking. A better understanding of the crystallization processes in these thin films could enable tailoring the properties of thin films to suit specific applications. To explore the crystall ization process in solution deposited PZT thin films, high temperature in situ laboratory and synchrotron X ray diffraction based techniques were developed. Taking advantage of the high X ray flux available at synchrotron facilities such as beamline 6 ID B Advanced Photon Source, Argonne National Laboratory, crystalline phases formed in the thin films during crystallization at the high heating rates (0.5 60 C/s) typically used during film processing could be measured. Using a 2 D

PAGE 19

19 detector for these measu rements allowed the simultaneous measurement of both phase and texture information during crystallization. Analytical treatment of the unconventional diffraction geometry used during the synchrotron based measurements was performed to develop methodologies for quantitative estimation of texture components. The nominal lead content in the starting solutions and the heating rate used during crystallization was observed to influence the sequence of phases formed during crystallization of the films. In films cr ystallized at fast heating rates, titanium segregation, probably due to diffusion of titanium from the adhesion layer, was observed. To further investigate the effect of adhesion layer on the crystallization behavior of these films, PZT films were solution deposited onto substrates with titanium (Ti), titanium oxide (TiO x ) and zinc oxide (ZnO) adhesion layers. Phase evolution was observed to be unaffected by the adhesion layer. (111) oriented PZT films were obtained on all substrates and hence Ti diffusion from the adhesion layer appeared to have limited influence on texture control in these films. It is suggested that the (111) orientation of the perovskite phase is directly seeded by the Pt(111) texture.

PAGE 20

20 CHAPTER 1 INTRODUCTION Lead zirconate titanate (PZT) thin films have potential applications in embedded decoupling capacitors [1, 2] micromechanical actuators [3 5] and ferroelectric memori es [6 8] Efforts to use PZT based materials in these applications are motivated by the excellent dielectric, piezoelectric and ferroelectric properties possessed by these materials Piezoelectricity is defined a s the co upling between the electric and the elastic displacements [9, 10] and is represented by the following relationship ( 1 1 ) represents the strain, is the piezoelectric coefficient and is the electric field Since strain is a second rank ten s or and the electric field is a first rank tensor, it follows that the piezoelectric coefficient is a third rank tensor. Due t o the tensorial property of the piezoelectric coefficient, piezoelectricity can exist only in materials with a non centrosymmetric crystal structure [11] Of the 32 crystal point groups, 20 point groups are non centrosymmetric and hence, permit piezoelectricity. Among these 20 non centrosymmetric point groups, 10 point groups have a polar axis and allow for an electric dipole moment and spontaneous polarization [11] These 10 polar point groups are also referred to as pyroelectric point groups. Ferroelectric materials are a fu r ther subset of the materials in these 10 polar point groups in which the spontaneous polarization can be reoriented through the application of an electric field [10] H ence the presence of a spontaneous p olarization does not necessarily imply ferroelectricity as the spontaneous polarization is not r eorientable in some materials A t least two stable orientations for the spontaneous polarization need to be present for a polar material to be ferroelectric [12] Measurement of a ferroelectric hysteresis loop is often presented

PAGE 21

21 as evidence for ferroelectricity in materials ( Figure 1 1 ). The typical material characteristics that can be measured from the ferroelectric hysteresis loop are the coercive field (E C ), remnant polarization (P R ) and saturation polarization (P SAT ). Ferroelectric materials are also observed to have anomalously high dielectric constants. Barium titanate ( BaTiO 3 ) another popular ferroelectric material has a higher dielectric constant and is widely used in the multilayer ceramic capacitor (MLCC) industry [9, 13, 14] However, integration of BaTiO 3 in embedded applications is severely limited by the high temperatures required for processing. The applicability of BaTiO 3 ferroelectric based f ilms has been extended based on the film on foil approaches developed in recent years [15 20] However, for applications which require deposition of high dielectric constant ma terial onto integrated circuits such as embedded decoupling capacitors, PZT remains the preferred material due to the lower temperatures required for processing [1, 6] PZT is a solid solution between PbZrO 3 (PZ) and PbTiO 3 (PT) ( Figure 1 2 ) PZ has a rhombohedral crystal structure and is antiferroelectric while PT has a tetragonal crystal structure and is ferroelectric. Addition of PZ to PT decreases the tetragonal distortion until the cryst al structure changes to a rhombohedral unit cell. The phase boundary between the tetragonal and the rhombohedral phase regions (PZ/PT = 52/48) is called the morphotropic phase boundary. The dielectric constant of PZT ceramics are observed to peak at th e morphotropic phase boundary. Concurrently, high piezoelectric coefficient s are also observed. These high dielectric and piezoelectric coefficients are suggested to be due to greater number of polarization directions available [10]

PAGE 22

22 M easurements performed on thi n films indicate that the properties of thin films are different from bulk ceramics [21 23] Chen et al. [24] m easured dielectric and piezoelectric properties of thin films at various Zr/Ti ratios and found that the measured values for thin films are lower than the corresponding values measured in bulk materials. The decrease in electrical properties in thin films is suggested to be due to clamping of the film by the substrate [25] and unintentional contamination from the substrate [26] While the difference in properties of thin films and bulk materials is recognized lack of proper models limits extending the present knowledge of bulk material properties to thin films [27] 1.1. Chemical Solution Deposition o f PZT Thin Films PZT thin films have been fabricated using various thin film deposition techniques such as molecular beam epitaxy (MBE) [28, 29] chemical vapor deposition (CVD) [30, 31] sputtering [32, 33] and chemical solution deposition (CSD) [34 37] For application s such as senso r a rrays and decoupling capacitors where cost is a major driver, chemical solution deposition (CSD) routes are especially attractive [1, 27] CSD routes offer multiple advantages such as low cost, good control over stoichiometry and scalability The major disadvantage of solution deposition is lack of reliability. Reliability is difficult to achieve in solution deposition due to t he complex processes taking place during solution preparation and deposition. During solution deposition, the precursors are dissolved in a suitable solvent and then deposited onto the substrate, usually by spin coating. After spin coating, the film is ta ken through a combination of thermal processes to evaporate the solvent and crystallize the thin film ( Figure 1 3 ) The first thermal process involves holding the film at an intermediate temperature to evaporate the solution (pyrolysis). After pyrolysis the

PAGE 23

23 film is annealed at a higher temperature to crystallize the film into the desired phase ( crystallization). Depending on the process, the pyrol ysis step is sometimes skipped. 1.1.1. Solution Processing Routes The specific route used for the preparation of the solution is known to have an influence on the final properties of PZT thin films [38] Solution preparation routes used for solution deposition of PZT th in films can be classified as [39, 40] : Sol gel based Metalloorganic decomposition Chelate processes 1.1.1.1. Sol gel based routes The sol gel route is the most widely used solution preparation route for PZT thin films [39, 41] A sol is defined as a colloidal suspension of particles in a sol u tion. If the particles in the sol have a propensity to r eact with each other, in due course of time a polymer is formed. When t his polymer skeleton enclos es the liquid it is called a gel [42] For obtaining oxides, m etal alkoxides are commonly used as precursors since they readily react with water. The key steps taking place in the solution during the sol gel process are shown in Figure 1 4 Alcohol exchange of the metal alko xides with the solvent leads to a decrease in the sensitivity of the metal alkoxide to hydrolysis. For sol gel processing of PZT based thin films, zirconium and titanium propoxides are used as the precursors for the B sites and lead (II) acetate is used a s the source for lead. 2 methoxymethanol (2 MOE) is commonly used as a solvent for this process. Since the precursors used are very sensitive to water, the solutions are prepared using schlenk lines and the solutions are processed through multiple refluxin g and distillation steps to remove any trace water [43]

PAGE 24

24 Solutions prepared through this route have high controllability and reproducibility. If care is taken to ensure that no water is present in the solution s, these solutions show very little aging. Some of the disadvantages of this process are the complexity of the process and the toxicity of the solvent. 2 MOE is a known teratogen and presents a significant safety risk. 1.1.1.2. Metal lo organic decomposition In met allo organic decomposition (MOD) water insensitive precursors such as metal acetates or metal diketonates are used [35, 44 47] The metal precursors are usually in the liquid form and are mixed in a solvent. Further modifications performed on the solutions include adjusting the pH and the viscosity [47] Since the precursors used in this route have very low reactivity, these solutions show minimal aging. In the solution, the metal precursors are present as a mixture. These solutions conta in significantly highe r organic content than sol gel based solutions. The loss of this organic content during pyrolysis/crystallization could cause cracking of the films. 1.1.1.3. Hybrid chelate based routes The hybrid chelate routes use either alcohol exchange or the formation of a coo rdination compound (chelation) to decrease the sensitivity of the metal alkoxide precursors. For preparation of PZT thin films, these methods can be differentiated by the sequence of the addition of metal alkoxide precursors. In the original preparation ro ute proposed by Yi et al. [48] the A site metal alkoxides were added first, followed by the B site metal alkoxides and this route is often referred to as seque ntial precursor addition (SPA). Modifications were made to the SPA route by Assink et al. [37 ] at Sandia National Laboratories. The specific modifications made are (1) inverting the order in which the metal alkoxides are mixed into the solution (B site first and A site

PAGE 25

25 second) and (2) using lead (IV) acetate instead of lead (II) acetate. This m odified chelate based solution preparation route is called the inverted mixing order (IMO) route All PZT films used in this study were deposited using IMO based solutions. In the IMO solution preparation route, the B site precursors, zirconium butoxide an d titanium isoproxide are first mixed in the required stoichiometric amounts and vortexed. Acetic acid is then added to this mixture to chelate the alkoxides and stabilize them against reaction with water. The solution is vortexed for five minutes before m ethanol is added to quench the reactions taking place in the solution. The solution is again vortexed for five minutes before addition of lead (IV) acetate. The solution is heated on a hot plate at ~ 90 C to dissolve the lead acetate. Upon dissolution of l ead acetate, the solution is allowed to cool slightly bef ore adding methanol. Subsequently, acetic acid and methanol are added to adjust the final molarity of the solution [1, 37] During solution preparation in the IMO route, hydrolysis and condensation ( Figure 1 5 ) reactions occur However, the main reaction is the chelation of the alkoxides by acetic acid Chelation of the alkoxides with acetic acid decreases the se nsitivity of these alkoxides to air and allows for handing these in normal atmospheric conditions [1] Acetic acid does not completely replace the alkoxide groups in the metal alkoxides, possibly due to stear ic hindrance [37] During the solution preparation process, due to the esterification reaction between the acetic ac id and the alcohols in the solution, water is formed This formed water cannot be removed in situ and leads to the continued reaction between the precursors in the solution. This continued reaction between the precursors leads to a change in the properties of the solution with time and

PAGE 26

26 subsequently a change in the microstructure of the thin films [49] Similar aging was also observed for solutions prepared through SPA route [50 52] 1.1.2. I n homogeneity i n Solutions Compositional homogeneity and molecular mixing are often assumed to occur in the precursor solutions. Sengupta et al. [53] performed extended x ray absorption fine structure (EXAFS) studies on powders derived from sol gel based solutions to identify the local structure. Compositional inhomogeneity in the form of separate Ti O Ti, Zr O Zr and Pb O Pb linkages were observed. These observations are consistent with the EXAFS investigations by Feth et al. [54] The T i Ti and Zr Zr coordination and distances were observed to similar for PZT powders of different compositions. The local ordering of the cations was observed to change significantly between the amorphous phase and the perovskite phase. Wilkinson et al. [55] analyzed the local structure present in the amorphous phase using reitveld and pair distribution function analysis of neutron diffraction patterns and observed that s imilar compositional inhomogeneity was found in the amorphous phase of the powders made from the hydrolysis of PZ and PT precursors prior to mixing or after mixing. Hence, the metal (Zr/Ti) alkoxide precursors are observed to have a tendency for homocondensation. Improving the homogeneity of the solutions is found to improve the properties of the final films and decrease the temperature required for formation of the perovskite phase [56 58] Most studies on solution derived PZT powders have been performed using solutions prepared through sol gel chemistries and this information cannot be readily extrapolated to powders/films derived via IMO chemistry. However, some similarit ies are observed between sol gel and IMO chemistries. The lead precursor in the IMO solution could be selectively isolated from the solution, implying that the lead precursor

PAGE 27

27 in the IMO solutions does not homogenously interact with the titanium and zirconi um precursors [49] However, no studies have been performed to check for homogeneity in the mixing of the zirconium and titanium precursors in IMO solutions. 1.2. Deposition 1.2.1. Spin Coating Evaporation of the s olution starts during the spin coating step, however, the film contains significant amount of solution after spin coating [5 9, 60] The evaporation of the solvent from the film leads to significant contraction of the film during high temperature thermal treatments. This contraction causes th e formation of a biaxial stress state in the film. Formation of very high levels of these stresses could lead to the formation of cracks, microcracks or mud cracks, in the thin film. The critical stress required for the formation of these cracks can be est propagation. From this calculation, if follows that the thicker films are more susceptible to the formation of cracks [61] Several strategies exist for avoiding the formation of cracks in solution deposited thin films. One strategy is to deposit a thin layer of the film and crystallize the film before deposition of another layer. This sequence of steps can be repeated until the desired thickness is achieved. Anoth er strategy is to pyrolyze the film after each deposition step and then to include a crystallization step after a few pyrolysis steps. 1.2.2. Platinized Silicon Substrates Platinized silicon substrates are the most widely used substrates for fabrication of PZT ba sed thin films. Direct integration of PZT on Si is not possible due to the reaction between PZT and silicon. Hence a film of platinum is used to act both as a conductive electrode and to prevent reaction between the PZT thin film and the substrate [62]

PAGE 28

28 Platinum is preferred over other metals such as Au, Ag or Cu due to its resistance to oxidation. However, platinum does not adhere well to silicon or silicon oxide an d adhesion layer such as titanium, is commonly use d to avoid film peeling Titanium has been the most commonly used adhesion layer for integrating platinum onto silicon substrates. Sreenivas et al. [63] investigated the stability of the Pt/Ti/SiO 2 /Si stack durin g processing by varying thickness of the Ti adhesion layer in substrate stack and processing through similar thermal heat treatment. The Ti adhesion layer was found interact with the underlying SiO 2 layer by reducing the top layers of the film. This reacti on between the Ti and SiO 2 layers was suggested to lead to increased bonding between these two layers. In addition, in post processed substrates Ti was observed in the Pt layer and on the surface, suggesting that Ti migrates through the Pt film during proc essing. Elemental analysis by Rutherford back scattering (RBS) indicated that oxygen was diffusing from surface, through the Pt film, to oxidize titanium to titanium oxide. This observation of titanium oxide formation in the middle of the Pt layers is also consistent with the observations made by Tani [64] Oxidation of the underlying titanium adhesion layer, possibly due to diffusion of oxygen from the surface, was also observed by Fox et al. [65] In addition, Fox et al. [65] observed microstruc tural changes in the platinum layer during annealing. The Pt grains were observed to grow when the substrate was held at temperatures higher than 500 C. However, these changes were not observed when films were annealed after sputter deposition of a PLZ or PZT layer on the top of the substrate. Al Shareef et al. [66] sug gested that these changes in the microstructure of the Pt film were due to stress relieving through the formation of hillocks. Annealing the

PAGE 29

29 substrate after deposition of a PZT layer was suggested to constrain the Pt layer against the formation of these hi llocks [67] Al Shareef [66] also investigated the effect of the adhesion layer on the roughness of the PZT thin film after crystallization. It was found that TiO 2 and ZrO 2 adhesion layers resulted PZT thin films with lower RMS roughness than substrates with Ti adhesion layer. Howeve r, the adhesion properties of Ti adhesion layers were observed to be better than TiO 2 or ZrO 2 adhesion layers. Hence, it was suggested that Ti adhesion layers need to be used when thick layers of Pt need to be used. Thick layers of Pt need to be used when higher conductivity of the bottom electrode is required, such as for capacitor applications. Maeder et al. [68] investigated the stability of the Pt electrode when deposited on various metal adhesion layers (Ti, Zr and Ta). Oxidation of the adhesion laye rs prior to deposition of the Pt layers was observed to decrease the diffusivity of the adhesion layer metals through the Pt film. Oxidizing the adhesion layer was also found to decrease the diffusivity of PbO from the film through the Pt electrode. It was suggested that adhesion layer could play a role in determining the diffusivity of PbO from the surface through the Pt electrode. The reduction of PbO by the more reactive adhesion metal atoms to metallic Pb was suggested to be a possible mechanism. In su mmary, while Pt electrode is resistant to oxidation during processing, it allows for the diffusion of PbO and oxygen from the surface and the metal in the adhesion layer from the substrate. The formation of reaction products at the adhesion layer and in th e Pt electrode could potentially deteriorate the properties of the films and could cause problems during integration of these films onto devices. The diffusion of the metal atoms from the adhesion layer have also been suggested to affect the crystallizatio n behavior

PAGE 30

30 of the PZT thin films [69] The possible effects of adhesion layer on the crystallization of the PZT film are described in later sections. 1.3. Phase Evolution se starts during pyrolysis and is complete upon formation of the perovskite phase during crystallization [61, 70, 71] Most of the organic moieties presen t in the thin film after spin coating are removed during pyrolysis [70] The film is observed to be amorphous after the pyrolysis step. A short lived fluorite type phase has been observed to form in sol gel derived PZT thin films by Norga et al. [72] during pyrolysis. However, this fluorite ph ase has not been observed in other studies [70, 73, 74] Upon further heat treatment of the f ilms at higher temperatures, a fluorite type phase formed [75 77] In addition to th is fluorite type phase, an intermetallic Pt x Pb phase has also been reported. This transient intermetallic is suggested to form due to the reduction of Pb +2 to metallic Pb 0 and consequent alloying with the platinum bottom electrode. Processing conditions su ch as high heating rate during crystallization [78 81] annealing atmosphere [82, 83] and organic content in the solut ion [84] have been suggested to affect the stability and formation of this intermetallic phase. The temperatures experienced by the films during crystallization (600 700 C) are low on a homologous scale (T M ~ 1400 C for PbZr 0.5 Ti 0.5 O 3 [85] ). Hence, very little long range diffusion is expected to occur during these heat treatments [61] This limited long range diffusion hinders the formation of the thermodynamically stable phase s until high enough temperatures are reached. Due to the kinetic limitations for the formation of the thermodynamically stable phase, metastable fluorite/pyrochlore type phases are

PAGE 31

31 known to form in these thin films [40, 86] T his phenomenon of formation of metastable phases is common to solution based methods for synthesis of inorganic compounds [76] It has also been suggest ed that the formation of the intermediate fluorite phase could be due to the preservation of the structural elements of the initial precursors used during the synthesis [76] Several studies have focused on understanding the kinetics of the fluorite to perovskite transformation [75, 77, 79, 80, 87, 88] The results from a few key studies are summarized in the following paragraphs. Kwok et al. [89] investigated the fluorite to perovskite phase transformation in PZT thin films. The fluorite grains were observ ed to be ~10 nm and were dispersed in an amorphous matrix Very little compositional differences were observed between fluorite and the perovskite grains. Based on the grain growth and by deriving the Avrami coefficient for the grain growth, it was suggest ed that the structural transformation from the fluorite phase to the perovskite phase occurs at t he interface of the two phases. In TEM studies performed by Tuttle et al. [90] the nanoscale fluorite type phase dispersed in an amorphous matrix was confirmed Similar to results of Kwok et al. [77] no compositional differences were observed between the fluorite type phase and the perovskite phase. However, c ompositional differences were observed between the nanoscale fluorite type phase and the amorphous reg ions. The amorphous phase was observed to be Pb deficient and had a higher Zr/Ti ratio. The formation of this two phase microstructure was suggested to be due to phase separation. The perovskite grains are observed to grow into this matrix with nanoscale c rystalline fluorite regions.

PAGE 32

32 Griswold et al. [79] used rapid thermal annealing to generate snapshots of the crystallization process and the quenched sample s were used to characterize the phase evolution in PZT thin films. The pyrochlore phase could not be bypassed by using RTA in these films (PZT 40/60 + 10% Pb). The film was observed to transform completely to a perovskite phase after 15s of annealing. The final phases present in the film for RTA and conventional furnace anneal were investigated. It was observed that for films annealed in the conventional furnace, residual fluorite was always present in the films. This was attributed to the greater possibili ty for PbO volatilization from the film before the formation of the perovskite phase. The kinetics of phase transformation from the pyrochlore to the perovskite phase was evaluated for films heated at diff erent temperatures. From the A v rami coefficient, the perovskite phase seemed to grow in a two dimensional linear growth with decreasing nucleation rate regime between 550 C and 625 C. Lead volatility is a concern during processing of PZT thin films. The fluorite type phase is stabilized in nominally lea d deficient PZT compositions [91] The presence of a dielectric fluorite phase could lead to significant decrease in the dielectric properties of these thin films [74] Hence, care is taken during processing to avoid the presence of the fluorite type phase, either due to incomplete transformation of the intermediate fluorite type phase or due to lead volatilization. F a st heating during crystallization of the thin films ha ve been suggested to decrease lead volatilization [92, 93] Hu et al. [88] observed that the formation of the intermediate fluorite type phase can be avoided completely in sputter deposited PZT thin films if high enough heating rates are u sed during crystallization. However, a fluorite type phase has always been observed to form

PAGE 33

33 in solution derived PZT thin films [79, 94] Even for the very fast heating rates achievable in a rapid thermal annealer (RTA), lead volatilization cannot be completely avoided [74] H ence, PZT solutions are generally batched to include nominally excess amounts of lead precursor to account for volatilization during thermal processing. Several strategies such as post annealing with a PbO layer [84, 95] and annealing in a PbO atmosphere powder [96] have been observed to compensate for Pb volatilization during annealing. 1.4. Texture o f Crystallized Films To take advantage of the a nisotropy of PZT, it is desirable to control the texture of these films. For example, (111) texture d films are desirable for capacitor applications [97] and (100) textured films are preferred for sensor applications [98] Texture of solution deposited PZT thin films on platinum layers is observed to be dependent on the pr ocessing conditions during crystallization. While epitaxial films of lead titanate ( PT ) could be obtained through solution deposition on single crystal strontium titanate [99] and lanthanum aluminate [100] substrates coating these single crystal substrates with platinum was observed to markedly change the textur e obtained in the PZT films. Films deposited on single crystal oxide substrates with a sputtered (111) Pt layer were observed to be randomly oriented [101] For PZT thin films solution deposited on platinized silicon substrates (111) and (100) type texture are commonly observed [82, 102 104] Processing f actors such as pyrolysis temperature [1, 82, 105] heating rate [106] adhesion layer [107] in the substrate and atmosphere during crystallization [83] have been observed to control the f inal texture obtained in these thin films In general, the trends observed in film texture with change in processing conditions are consistent between different studies. A l ow pyrolysis temperature and fast heating rates have been

PAGE 34

34 observed to lead to (111) oriented films [81, 108] and a h igh pyrolysis temperature and slow heating rates are observed to lead to (100) oriented films [105] Based on these trends various mechanisms have been suggested [69, 81, 82, 104, 106, 109] However, a consensus on the mechanism for texture control in CSD PZT films on platinized silicon substrates is lacking. 1.4.1. Homogenous Vs Heterogenous Nucleation Crystallization of PZT thin films is observed to be nucleation limited [110] Hence, the final orientation obtained in the thin films could be governed by the nucleation processes taking place in the thin film [111] Due to the high surface to volume ratio in thin films, crystallization is expected to be predominantly through heterogeneous nucleation of crystalline regions at the film substrate interface. The classical nucleation theor y or the capillarity theory provides a simple mechanistic model for understanding the effect of va rious processing factors on nucleation and is briefly introduced here The details of capillarity theory for nucleation are well described in books [111 114] and hen ce, only the relevant relationships are listed. The free energy change for heterogeneous and homogenous nucleation are related by ( 1 2 ) Where, is referred to as the shape factor and is given by ( 1 3 ) is related to the surface free energies of the different phases present during nucleation ( Figure 1 8 )

PAGE 35

35 ( 1 4 ) From this formalism, it follows that heterogeneous nu cleation is always more preferred over homogeneous nucleation. Isotropy between different crystal orientations is an inherent assumption in this formalism. The situation in real materials is considerably more complex due to the anisotropy in the surface fr ee energies between different facets of the crystal. Particularly, in oxide materials the dipoles formed due to the termination of surfaces also need s to be considered. Using the framework of the capillarity theory for nucleation, Muralt [69] suggested that the stability of the nuclei is dependent on both the interface energy between the substrate and the nuclei, and the surface energy of the nuclei. The surface energy of the nuclei is suggested to be influenced primarily by the uncompensated charges on the surface (dipoles). Based on this formalism, it was suggested that for the formation of (100) oriented perovskite nuclei the surface energy is low but the interface energy between the perovskite phase and the Pt(111) is high. The situation is reversed for (111) oriented nuclei, which might have a high surface energy due to large dipoles but a low interface energy due to lattice matching. Hence, processing conditions that preferentially stabilize specific nuclei could result in the formation of textured thin films. 1.4.2. Seeding of (111) Orientation b y Pt x pb In PZT films crystallized at fast heating rates, a Pt x Pb type intermetallic phase is observed to form [79, 81, 115] This metastable intermetallic phase is observed to have an epitaxial relationship with the platinum electrode [116] Th e lattice parameter of th is metastable phase (a 0 = 4.05 angstroms) is closer to the lattice parameter of PZT (a 0 = 4.07 angstroms) than platinum ( a 0 = 3.92 angstroms). It was suggested that the

PAGE 36

36 perovskite (111) orientation is seeded due to the close lattice matching with Pt x Pb [78, 117] At slow heating rates, the P t x Pb phase was suggested to decompose to form a PbO layer on the Pt substrate. This PbO layer was suggested to seed (100) orientation [118, 119] 1.4.3. Transformation o f Fluorite Phase The fluorite type phase formed in CSD PZT thin films prior to perovskite formation has been suggested to stabilize specifically oriented perovskite nuclei. Norga et al [72] and Voigt et al [120] suggested that heterogeneous nucleation of fluorite phase can occur at the film substrate interface. These heterogeneously nucleated fluorite grains could transform to perovskite while maintaining their orientational relationship with the substrate. Bro oks et al [82] suggested that the relative stability of the fluorite type phase with respect to the perovskite phase controls the orientation selection in solution deposited thin films. The relative stability of the fluorite type phase formed during low pyrolysis temperature was suggested to be lower than the fluorite phase formed under high pyrolysis temperatures [121] The lower stability of the fluorite phase in low temperature pyrolyzed thin films was suggested to allow for the growth of (111) oriented grains seeded by the Pt(111) layer [122] In films pyrolyzed at higher temperatures, the fluorite phase formed in the films is mo re stable and hence only the lowest energy surface (100) of the perovskite phase was suggested to grow. 1.4.4. Ti Diffusion f rom Adhesion Layer Considerable titanium diffusion from the adhesion layer of platinized silicon substrates is observed to occur during pr ocessing of PZT thin films [63, 64, 68] The migration of titanium from the adhesion has been suggested to seed particular orientations in solution deposited PZT thin films. Tani [64] suggested that a Pt 3 Ti phase

PAGE 37

37 is formed on the surface of the pla tinum layer due to titanium diffusion. This Pt 3 Ti layer was suggested to seed the (111) orientation in PZT thin films. A different pathway for the effect of titanium diffusion from the adhesion layer on PZT texture was forwarded by Muralt et al [107] The titanium diffusion from the adhesion layer could cause the formation of a rutile (TiO 2 ) layer on the surface of the platinum layer. Th e rutile phase can nucleate heteroepitaxially on Pt (111) and transform into the (111) perovskite nuclei [123] Bour egba et al. [124] observed that the final orientation of PZT films seeded with rutile seed layer varied with the oxygen partial pressure during the crystallization process. Under reducing conditions, a (111) orientation was obtained and in oxidizing conditions a (100) orientation was obtained. Thus extending the scope of control of PZT texture by rutile seeds to explain both (111) and (100) textur es in solution deposited thin films. 1.5. Motivation a nd Scope o f t he Present Work Most studies performed to improve understanding of crystallization in solution deposited PZT thin films have focused on varying a particular processing condition during crystalli zation. The effect of the processing condition was then retrospectively evaluated based on the structure and properties observed in the thin film after crystallization. While these previous studies have improved the understanding of crystallization in solu tion deposited thin films in a general sense, the particular effect of change in processing conditions on nucleation during crystallization is not well understood. X ray diffraction is an attractive technique to characterize changes in the thin film during crystallization since it can measure the changes at the film substrate interface nondestructively. However, few studies have focused on measuring changes during

PAGE 38

38 crystallization of PZT thin films. These previous studies have used laboratory X ray diffracti on techniques to measure phase evolution during crystallization of PZT thin films at slow heating rates [75] In typical PZT thin film processing, the films are crystallized by heating them at fast heating rates by using a rapid thermal annealer (RTA).Hence, the conditions present during typical thin film processing co uld be significantly from the measurements made at slow heating rates in previous studies. The low X ray flux of the laboratory X ray sources limit the in situ laboratory X ray diffraction techniques used in these previous studies to slow heating rates. Th e high X ray flux available at synchrotron X ray sources such as the Advanced Photon Source could be used to overcome the limitations of the laboratory based in situ diffraction techniques. Advantage can also be taken of the availability of advanced two di mensional (2D) detectors at these synchrotron facilities. 2D detectors are capable of measuring the entire Debye Scherrer diffraction intensities thus, both phase and texture information can be simultaneously captured in a single diffraction image. Modern 2D detectors also allow for the acquisition of diffraction intensities very rapidly. For example diffraction images can be measured with an acquisition time of 0.25 s. Thus the combination of high X ray flux from synchrotron sources and 2D area detectors w ould allow for the characterization of phase and texture evolution with texture, phase and timing resolution that are appropriate for the typical heating rates used during crystallization of solution derived PZT thin films. The objective of the present wo rk is to develop in situ techniques based on laboratory and synchrotron X ray diffraction to enable the measurement of phase and texture evolution during crystallization solution deposited PZT thin films. The

PAGE 39

39 methodologies for the extraction of phase and t exture information from the measured 2D diffraction images were also developed and implemented. 1.6. Outline The details of the film deposition and X ray diffraction measurements are listed in CHAPTER 2 CHAPTER 3 provides the details of the in situ X ray diffraction techniques used in this work. The theory for the interpretation of diffraction data measured at synchrotron X ray sources using 2D detect or is also presented. CHAPTER 4 presents the effect of PbO content on phase evolution in solution deposited PLZT thin films. The effect of aging the thin films aft er pyrolysis is presented in CHAPTER 5 The effect of heating rate during crystallization was investigated using synchrotron X ray diffraction techniques and is pre sented in CHAPTER 6 Using the in situ techniques developed in this work, the structure and formation sequence of the Pt x Pb is evaluated in CHAPTER 7 The effect of adhesion layer in the substrate on the phase and texture evolution in solution deposited thin films was investigated and is presented in CHAPTER 8 The results obtained through this research are then summarized and listed in CHAPTER 9 1.7. Conclusion Solution deposition is a viable route for the preparation of PZT thin films for integrated decoupling capacitor applications During thermal processing of solution deposited PZT thin films, a metastable Pt x Pb phase and a fluorite type phase are formed prior to the formation of the perovskite phase. Final properties of solution deposited PZT thin films are observed to be dependent on processing variables such as heating rate and pyrolysis temperature. Several mechanisms have been suggested to explain the variation of texture in these films However, a consensus is still lacking. A

PAGE 40

40 better understanding of the processes during crystallization could help to improve the reliability of the solution deposit ion route for fabricating PZT thin films.

PAGE 41

41 Figure 1 1 Typical hysteresis loop obtained for a ferroelectric material Figure 1 2 Equilibrium phase diagram of PZT. Reproduced from [10]

PAGE 42

42 Figure 1 3 Typical steps involved in the solution deposition of thin films Figure 1 4 Different reactions taking place during sol gel preparation of PZT solutions

PAGE 43

43 Figure 1 5 Steps in the IMO solution preparation route Figure 1 6 Additional reactions taking place during hybrid chelate solution preparation route

PAGE 44

44 Figure 1 7 Phases present in solution deposited thin films during various stages of thermal processing. Adapted from [84] Figure 1 8 Heterogeneous and homogenous nucleation of a crystallin e region in an amorphous film.

PAGE 45

45 CHAPTER 2 DETAILS OF EXPERIMEN TAL METHODS Various thin film samples were prepared during the course of this work. These thin film samples were crystallized using laboratory and synchrotron X ray diffraction based in situ techniques. This chapter provides the details of the sample preparation and e xperimental setups used in this work. 2.1. IMO Solution Preparation a nd Deposition 2.1.1. Standard Processing Solutions used for the deposition of PZT thin films in this study were prepared through the Inverted Mixing Order (IMO) route [37] Films of different compositions were used during this work. Preparation of 20 mL of PZT (52/48) with 10% nominal excess lead precursor is desc ribed in detail below as an example to highlight the steps involved in the preparation of solutions through the IMO route. These steps can be suitably modified to synthesize IMO solutions with different Zr/Ti ratios and lead precursor excess. To prepare the solutions, 1.744g of zirconium butoxide and 0.955g of titanium isopropoxide were first mixed for 5 minutes. This mixture was diluted with 1.6 mL of acetic acid and 5 mL of methanol with a mixing time of 5 minutes between each step. 3.594g of lead (IV) acetate was added to the solution after dilution. To dissolve the lead acetate into the solution, the solution was stirred while heating at ~90C. The solution turns from a pale yellow color to clear after all the lead acetate has dissolved int o solution. The solution was allowed to stir for some additional time after the dissolution of lead acetate was complete. After complete dissolution of lead acetate into the solution, the solution was diluted to obtain a final molarity of 0.35. The dilutio n was done in 2

PAGE 46

46 steps with successive additions of 5 mL of methanol and 1.6 mL of acetic acid in each step. The solution was allowed to mix by vortexing for 2 minutes between each dilution step. All metallorganic precursors used were ac quired from Acros Or ganics Ltd. These steps listed as a flowchart in Figure 2 1 Prior to spin coating, s olution s were dispensed onto platinized silicon substrates through a syringe attac hed with a 0.45 m filter Spin coating was performed at 3000 rpm for 30 seconds using a Headway Research spin coater. The films were then pyrolyzed (300 C 400C ) for 1 minute after each spin coating step. Typically, the films were deposited in three spi n coating and pyrolysis cycles. The film thickness obtained after three spin coating and pyrolysis cycles is approximately 250 nm. 2.1.2. Processing f or Ceria Dispersed PZT Thin Films Ceria (NIST, SRM674b) was used as an internal temperature standard to measure temperature of solution deposited thin films. To prepare ceria dispersed PZT thin films, ceria was mixed into PZT solution. Prior to dispensing the solution onto the platinized s ilicon substrate, the solution was agitated to suspend the ceria powder in the solution. A syringe filter was not used during preparation of ceria dispersed PZT thin films. Multiple deposition, spin coating and pyrolysis of these solutions onto platinized silicon substrates was performed to obtain films of amorphous PZT thin films with ceria dispersed in them. 2.2. Preparation o f Platinized Silicon Substrates Platinized silicon substrates with TiO x and ZnO adhesion layers were prepared to investigate the effect of adhesion layer on phase and texture evolution ( CHAPTER 8 ).

PAGE 47

47 were used as the starting substrates for preparation of these substrates. Sputtering was performed using a Kurt J. Lesker sputtering tool at the Sandia National Laboratories. 2.2.1. Pt/TiO x /SiO 2 /Si Titanium was sputtered onto the starting substrates using RF magnetron sputte ring at 300 W and under 3 mTorr of Argon. The targets were presputtered for 4 mins prior to the start of deposition on the silicon substrate. After deposition of 40 nm of titanium, the sputtering was stopped and the atmosphere of the chamber was switched t o 15 mTorr of oxygen. The substrates were then heated to 450 C and held for 30 minutes to facilitate oxidation of the titanium adhesion layer. The oxidation of the titanium layer might not go to completion under these conditions and hence this layer is des ignated at TiO x After the oxidization heat treatment, the substrates were cooled and then 100 nm of Pt was deposited 2.2.2. Pt/ZnO/SiO2/Si 40 nm of ZnO was RF sputter deposited onto silicon substrates at 150 W under a 5 mTorr of chamber pressure with equal parti al pressures of argon and oxygen. 100 nm of platinum was deposited after deposition of the ZnO layer. After deposition of the Pt layer the Pt/ZnO substrates were directly inserted into a furnace preheated to 700 C and annealed for 10 minutes under ambient atmospheric conditions. 2.3. In Situ Synchrotron X Ray Diffraction Setup In situ experiments using synchrotron X ray diffraction were performed during three different beam time allocations: March 2010, November 2010 and October 2011. The instrumental setups use d for in situ crystallization during each of these beam time allocations were slightly different. A significant difference between these beam time accesses was the orientation of the sample with respect to the IR lamp during heating.

PAGE 48

48 During beam time acces s in March, 2010 and October, 2011 the samples were heated with the film side of the samples facing the IR lamp. In general the results obtained during the March, 2010 and October, 2011 beam time access were consistent with existing literature. However, in the November, 2010 beam time access, the films were heated with the substrate side of the samples facing the IR lamp. The phase and texture evolution observed during the November, 2010 beam time access were not consistent with the results of the beam time access in March, 2010 and October, 2011. This inconsistency in the results could be due to the differences in the heating geometry. The effect of change in heating geometry on crystallization of solution deposited PZT thin films was not investigated furth er and the results of the November, 2011 beam time access are not presented in the dissertation. The thin films were crystallized at different heating rates by heating the samples using an IR lamp similar to a rapid thermal annealer (RTA). The IR lamp (O sram Sylvania No. 64635HLX) was placed ~ 10 cms above the sample. Heating rate during experiment was controlled by gradually stepping up the voltage applied on the IR lamp. The voltage was controlled by using a digital to analog converter under program con trol which provided a 0 to 5 volt signal to a Control Concepts Inc. model 1032A solid state power controller that in turn provided phase angle modulated AC power to a step down transformer that lowered the voltage to a maximum of 15 VAC for the lamp. X ray energies of 23.99 keV and 22.7 keV were used during in situ measurement in March, 2010 and October, 2011, respectively. The incidence angle of the incoming X ray beam on the sample was controlled by changing the orientation of the sample. A low

PAGE 49

49 incidence angle was used in this work ( ~ 0.2 0.5 ) to maximize the interaction volume of the X ray in the thin film. A significant difference between the experimental setup used in March, 2010 and October, 2011 are the detectors used for the measurement of the diffracted X rays. In the beam time access in March, 2010 a MAR SX 165 CCD detector was used. This detector had a few disadvantages: the highest frame rate achievable was 1 s 1 and the readout time is ~2 s. To overcome these disadvantages, a GE amorphous silicon was used during the beam time in October, 2011. Using the GE detector frame rates as high as 4 s 1 could be achieved. The images of the setups used during the March, 2010 beam time access and the October, 2011 beam time access are shown in Figure 2 2 a nd Figure 2 3 respectively.

PAGE 50

50 Figure 2 1 Steps in the preparation of PZT solution through the IMO route.

PAGE 51

51 Figure 2 2 .Experimental setup used at beamline 6 ID B, APS during beamtime a ccess in March, 2010. Photo courtesy of Krishna Nittala Figure 2 3 .Experimental setup used at beamline 6 ID B, APS during beamtime access in November 201 0 Photo courtesy of Krishna Nittala

PAGE 52

5 2 CHAPTER 3 IN SITU X RAY DIFFRACTION SETUPS: DIFFRACTION GEOMETRY AND DATA EXTRACTION METHODOLOGIES To better understa nd phase and texture evolution in solution deposited PZT thin films, i n situ X ray diffraction studies were conducted using both laboratory and synchrotron X ray sources. The setups used for these in situ diffraction studies are considerably different from the conventional diffraction geometries used for thin film studies. This chapter provides details of the experimental setup and data treatment methods used for the analysis of these methods. 3.1. Texture Mea surement u sing X Ray Diffraction Texture can be defined as the preferential distribution of crystallite orientations in a polycrystalline material [125] Thin films are often observed to be textured. C ontrolling the texture obtained in the thin films is desirable as it allows for taking advantage of the inherent anisotropy in the properties of materia ls. For PZT based materials, a 100 or 111 type texture is observed to naturally form in the thin films. Depending on the pa rticular application, either a 100 or a 111 texture is desirable in the films. S tudies on texture in PZT based thin films have mostly used Lotgering factor to identify the components of texture and the ir relative fraction in the film. Lo tgering factor, for a particular plane is calculated by ( 3 1 ) Here, represents the relative intensity ratio as expected from a powder sample.

PAGE 53

53 While Lotgering factor can be used as a qualitative method for evaluation of texture, using L otgering factor as a quantitative evaluation for texture is not accurate [126] Approaches such as complete pole figure measurement or rocking curves are need ed to quantify volume of the different texture components and the sharpness of the t exture. In typical laboratory X ray diffraction based measurement of texture, the diffraction intensity of a plane is used to measure the orientation distributions in the sample. Different sections of the orientation space are sampled by using a goniometer coupled to an Euler cradle ( Figure 3 1 ). For planar representation of the change in diffraction intensities with angles and an equal area stereographic projection is often used. These representations are commonl y referred to as pole figures. Using a 2 D detector instead of a point detector allows for the simultaneous measurement of a section of the pole figure [127] X ray diffraction is an attractive technique to characterize changes in the thin film during crystallization since it can measure the changes at the film substrate interface nondestructively. However, f ew studies have focused on measuring changes during crystallization of PZT thin films [75, 128] These previous studies have used laboratory X ray diffraction techniques to measure phase evolution during crystallization of PZT thin films at slow heating rates. In typical PZT thin film processing, the films are crystallized by heating them at fast he ating rates by using a rapid thermal annealer (RTA). Hence, the conditions present during typical thin film processing could be significantly from the measurements made at slow heating rates in previous studies. The low X ray flux of the laboratory X ray s ources limit the in situ laboratory X ray diffraction techniques used in these previous studies to slow heating rates.

PAGE 54

54 3.2. Laboratory b ased In Situ X Ray Diffraction Setup An Inel diffractometer with a curved position sensitive detector and furnace attachment was used for in situ laboratory X ray diffraction measurements on thin films. This equipment was used to measure evolution of the crystalline phases during slow heating rates. Formalisms such as Lotgering factor have been developed for estimating the textu re present in thin film materials [125] based on diffraction patterns measured in Bragg Brentano geometry In the coupled /2 geometry, the poles of the planes contributing to the diffraction intensit ies are always perpendicular to the sample surface. However, this is not the case in asymmetric /2 geometry of the Inel diffractometer. The direction of the poles contributing to the diffraction intensity varies with the pole being measured ( Figure 3 2 ). Hence, estimation of the texture present in the thin films based on diffraction measurement on the Inel diffractometer is not straightforward and was not undertaken i n this work. More details on this in situ technique are provided in Chapter 3 3.3. In Situ Synchrotron X Ray b ased Setup The high X ray flux available at synchrotron X ray sources such as the Advanced Photon Source could be used to overcome the limitations of the laboratory based in situ diffraction techniques. Advantage can also be taken of the availability of advanced two dimensional (2D) detectors at these synchrotron facilities. 2D detectors are capable of measuring diffraction intensities from the entire Debye Scherrer ring. T hus, both phase and texture information can be simultaneously captured in a single diffraction i mage. Modern 2D detectors also allow for the acquisition of diffraction intensities very rapidly. For example diffraction images can be measured with an acquisition time of 0.25 s.

PAGE 55

55 Thus the combination of high X ray flux from synchrotron sources and 2D are a detectors would allow for the characterization of phase and texture evolution with texture, phase and timing resolution that are appropriate for the typical heating rates used during crystallization of solution derived PZT thin films. In t he setup used a t beamline 6 ID B at the APS ( Figure 3 3 ) the X rays were parallel to the horizontal The angle of incidence was controlled by changing the sample orientation with r espect to the incident beam. A s mall inciden ce angle (0.2 or 0.5 ) was used to maximize the X ray interaction volume in the thin film. The diffracted X rays were measured using a MAR SX 165 CCD detector or GE amorphous silicon 2D detector s A schematic de scribing the detector angles is presented in Figure 3 4 A. Sections of the diffraction image sampled to extract phase and texture information are indicated in Figure 3 4 B and C, respectively To measure the changes in the thin film during crystallization multiple diffraction images were collected while the films were crystallized by heating with an IR lamp ( Osram Sylvania No. 64635HLX ) placed ~ 10 cms above the sample Heating rate during the experiment was controlled by gradually increasing the voltage applied on the IR lamp. The voltage was controlled by using a digital to analog converter under program control which provided a 0 to 5 volt signal to a Control Concepts Inc. model 1032A solid state power controller that in turn provided phase angle modulated AC power to a step down transformer that lowered the voltage to a maximum o f 15 VAC for the lamp During heating of the sample, diffraction images were continuously collected using a 2D detector. Data reduction methodologies were developed to extract relevant phase and texture information from the diffraction images. Prior to dat a reduction, the sample

PAGE 56

56 to detector distance and the detector tilt were refined using d iffraction images measured on NIST Ceria dispersed on platinized silicon substrates. The D ebye Scherrer diffraction rings from NIST ceria were used to refine for detecto r distance and tilt angle in Fit2D [129] The obtained detector distance and detector tilt obtained from this refinement was used during subsequent data reduction in Fit2D. 3.4. Data Reduction a nd Representation Data reduction of the 2D diffraction for phase identification an d texture analysis involve integration of the data over different detector angles [127] For phase identification, the intensities are integrated over a limited range. This results in data similar to a typical diffraction pattern obtained in a Debye Scherrer diffractometer. The integration over can be represented as ( 3 2 ) represents the intensity measured by the detector at the pixel corresponding to and To obtain texture information diffraction intensities of a partic ular diffraction plane were extracted by integrati ng over 2 as shown in ( 3 3 ) ( 3 3 ) Integration over 2 can be performed at 2 values corresponding to different (hkl) planes of the structure to obtain the change in diffraction ntensity distributions over [130] Integra tions or data binning over and 2 were performed using Fit2D [129] Data reduction was performed for each of the diffraction images collected during the experiment. The reduced data was then plotted with respect to time of acquisition to

PAGE 57

57 generate a map of the evolutio n of phases and texture with time. These plots are designated as phase evolution plots ( Figure 3 5 A ) and azimuthal time plots ( Figure 3 5 B ). To automate data reduction and representation, routines were developed in MATLAB The codes for these routines are included in the APPENDIX A In the phase evolution plots, the crystallites that contribute to the diffraction intensities in the range for integration are oriented at different angles with respect to the surface normal. In the azimuthal time plots, intensity correction for absorption and the relationship between the diffraction angles and the pole figure angles need to be evaluat ed before analysis of the data. Details of these calculations are provided in the following sections. 3.5. Theoretical Absorption Correction Accounting for absorption of X ray intensities is important for interpretatio n of thin film diffraction data [125] In bulk materials, the change in the path length of the diffracted beam is compensated by the change in the interaction volume. For thin films, since a limited thickness is available for contribution to th e diffracted intensities, this compensation is not possible. Due to the difficulty in fabricating thin films with random orientation, an empirical intensity correction is not usually undertaken and theoretical correction factors for attenuation of the X ra y intensity due to absorption are used [131] The absorption correction is given by ( 3 4 ) is the X ray linear attenuation coefficient and is the total path length of the incident and the diffracted X rays in the sample. The X ray linear attenuation coefficient

PAGE 58

58 ( ) can be calculated from the X ray mass absorption coefficient ( ) and density ( ) by using ( 3 5 ) ( 3 5 ) For a mixture or a compound, t he mass absorption coefficient can be calculated from the constitutive elements by using ( 3 6 ) ( 3 6 ) represents the weight fraction of the constitutive elements and is the mass absorption coefficient of th e i th element These values have been calculated for PZT (52/48) for X ray wavelength of 20 keV and are listed in Table 3 1 From these values, and assuming density of PZT to be 8.605 g/cm 3 the linear attenuation coefficient ( ) is calculated to be 5.7188*10 4 m 1 The total path length is giv en by ( 3 7 ) ( 3 7 ) The projection of the path length along the sample normal is equal to the film thickness ( ) This relationship is used to calculate both and These relationships can be represented as: ( 3 8 ) a nd ( 3 9 ) W here is the unit vector parallel to the sample surface normal and is given by:

PAGE 59

59 ( 3 10 ) and are parallel to the incident and diffracted X ray beams ( Figure 3 6 ) The directional coordinates of the incoming ( ) and diffracted ( ) X ray beams are given by ( 3 11 ) and ( 3 12 ) respectively. ( 3 11 ) ( 3 12 ) ( 3 13 ) is obtained by combining ( 3 8 ) ( 3 9 ) ( 3 10 ) ( 3 11 ) and ( 3 12 ) ( 3 13 ) can be used to obtain th e magnitude of ( 3 13 ) Diffracted X ray intensities measured at different 2 and angles have travelled different distances in the thin films. Hence, the X ray attenuation at these different 2 and angles is different. ( 3 13 ) can be used to calculate the effect of X ray absorption on the measured intensity at different 2 and angles. A representative calculation for the variation in value of the absorption coeff icient for the (110) perovskite planes with is shown in ( Figure 3 7 ). It is observed that the effect of absorption is greater for thinner films. Films used in this s tudy are 250 300 nm thick. 3.6. Correction f or Tilting o f t he Sample During the experiment, only the inclination of the sample to the incident X ray beam is controlled ( Figure 3 8 ). Alignment of the tilt of the sample about the X axis is not possible during sample setup. Tilting of the sample about the X axis causes an

PAGE 60

60 apparent shift along the in plane angle of the textured intensities. To correct for this shift in textur e intensities due sample misalignment, an analytical relationship between the tilt of the sample and the measured shift in the in plane angle ( off ) is derived. The tilting of the sample about the X axis rotates the sample coordinates and t hi s rotation can be represented as ( 3 14 ) The orientation of the sample normal after th is rotation can be calculated using ( 3 15 ) ( 3 15 ) Where is the unit vector along the Z axis ( ( 3 17 ) ) and is the unit vector along the surface normal after tilting ( ( 3 17 ) ). ( 3 16 ) ( 3 17 ) Ideally, the crystallites with their momentum transfer vectors parallel to will correspond to the maximum diffracted intensity measured along However, due to the flat plate geometry used in the setup these crystallites donot contribute to diffraction intensities measured at = 90 A more elaborate derivation of the planes sampled in this geometry is given in Section 3.7 The position of the maximum intensity along corresponds to the condition when the tilted sample normal is coplanar with the incident and the diffracted beams. This can be relationship between the sample normal and the incident and diffracted X ray beam s is represented by ( 3 18 )

PAGE 61

61 ( 3 18 ) Substituti ng ( 3 11 ) and ( 3 12 ) into ( 3 18 ) we get ( 3 19 ) ( 3 19 ) The simultaneous equations in ( 3 19 ) can be solved to obtain the relationship between the sample tilt and the rotation of the diffraction image along the in plane direction. ( 3 20 ) From ( 3 20 ) it can be concluded that a rotation of the sample by angle 0 about the X axis causes a n equal rotation of the diffraction intensities along the in plane angle. Additionally, this shift in the intensities along the in plane angle ( ) is independent of 2 To correct for the shift in the intensities due to sample tilt, the tilt of the sample was estimated using the textured diffraction intensities of the platinum substrate. The sample tilt was refined by peak fitting the Pt (111) textured intensities with a Gaussian peak shape function. These tilts were estimated using the first diffraction pat tern collected for each sample. The sample tilt correction calculated from this first pattern was then applied to the subsequent pattern collected during data treatment. 3.7. Section o f Pole Figure Sampled In laboratory X ray based characterization of texture in materials, a n Euler cradle ( Figure 3 1 ) is used to measure the variation in the intensity of a particular pole of the sample with change in orientation. The orientation of the sample is varied by changing the goniometer angles and These scans are frequently referred to as scan, scan and rocking curve scans, respectively. Angles and are directly rel ated to the

PAGE 62

62 pole figure angles (= ) and (= 90 ), respectively. The measured intensities of the poles with change in orientation is frequently represented as a stereographic projection and is referred to as an intensity pole figure [125, 131] During measurements taken at beamline 6 ID B at the Advanced Photon Source (APS), the detector and the sample are held stationary during the ex periment and the X ray beam is incident at a small angle to the sample. Due to this specific geometry ( Figure 3 9 ), the pole intensities from planes parallel to the sa mple surface are not measured. For pole intensities measured at = 90 the poles are an angle away from the sample normal. Where is given by ( 3 21 ) From the above equation, it is clear that for planes at higher 2 is larger Based on the diffraction geometry, the orientation of the crystallited leading to diffraction intensities measured at differen t 2 and in the diffraction geometry can be calculated in terms of the pole figure angles and Let represent the pole orientation with respect to the laboratory coordinate system ( Figure 3 10 ) The pole orientation can be calculated from tensorial multiplication of the momentum transfer vector and the rotation matrix representing the tilt of the sample with respect to the normal ( ). This can be represented as: ( 3 22 ) w here, ( 3 23 )

PAGE 63

63 is parallel to the momentum transfer vector and can be calculate d from diffraction angles using the L aue equation for diffraction [125, 127] and is given by ( 3 24 ) Combining ( 3 22 ) ( 3 23 ) and ( 3 24 ) the directional coordinates of can be calculated. ( 3 25 ) The pole figure angles and can be calculated from the pole orientation by using ( 3 26 ) and ( 3 27 ) ( 3 26 ) ( 3 27 ) From ( 3 21 ) it follows that the center of the pole figure is measured only when the incident angle is equal to During measurements at the APS, a shallow incident angle (0.2 0.5 ) was chosen to maximize the interaction volume of the X ray beam in the thin film Hence, the poles parallel to the surface normal are not measured for hkl planes with 2 > ~1 For a particular hkl pole, corresponding to the change in from 0 180 orientations corres ponding to different angles on the pole figure are measured. The orientations of the poles contributing to the diffraction intensities with change in for the perovskite (100) pole are shown in Figure 3 11 An incident angle of 0.5 was used for this calculation. It is observed that for a particular hkl pole, the orientations from = to = 90 are measured. However, o nly a limited range of is sampled.

PAGE 64

64 The perovskite phase formed in solution derived PZT thin films deposited on platinized silicon substrates crystallizes predominantly with (100) and (111) textures [102] The texture axis is aligned parallel to the texture axis of the platinum layer and is parallel to the substrate normal. For fiber or axisymmetric textures with the fiber axis parallel to the sa mple normal, the texture components are invariant over Hence, determination of texture components based on pole intensity distributions over the in plane angle is justified. 3.7.1. Determination o f h kl Texture Components u sing n on ( h kl) Diffraction Intensiti es In traditional X ray measurements, a hkl texture component (hence forth designated at hkl ) is quantitatively characterized through the measurement of the change in hkl diffraction intensities with sample orientation. However, for the in situ measuremen ts based on synchrotron X ray diffraction developed in this work, diffraction intensities from crystallites oriented parallel to the su rface normal are not measured Hence, determination of the volume fraction of the hkl texture component based on the intensity of the ( hkl ) pole intensity at = 90 c ould lead to overestimation of the FWHM of the texture component and underestimation of the volume of the texture component. However, crystallites with their ( hkl ) planes oriented parallel to the sample surface also contribute to diffraction intensities for non ( hkl ) planes. Hence, advantage can be taken for the simultaneous measurement of the large number of Debye Scherrer diffraction rings measured by the 2D detector. For example, (110) diffraction intensities could be used to estimate both 100 and 111 texture components. To illustrate this proposed methodology, representative (100), (110) and (111) pole figures are generated for (100), (110) and (111) type textures ( Figure 3 12 ). From these figures, it

PAGE 65

65 can also be seen that the center of the pole figures are not measured in the APS geometry. However, the section measured by the 2D det ector on the (110) pole figure cuts ac ross intensities from both the 100 and 111 texture components. A simulated (110) pole figure for mixed (111) and (100) type textures is shown in Figure 3 13 to further illustrate this point Hence, the (110) diffraction intensities can be used to estimate the (111) and (100) texture components. However, for quantitative interpretation of intensities corresponding to different texture components the plane multiplicities and the solid angle over which the intensity is distributed have to be considered. 3.7.2. Plane Multiplicities a nd Angles b etween Planes For characterizing a hkl texture component using non ( hkl ) pole figure, the plan e multiplicities need to be considered. For a cubic crystal structure, there are 6 equivalent (110) type planes (only planes with positive indices are considered) The angles between the different (110) planes and the (111) and (100) planes are listed Table 3 2 For a film with (111) texture, equal numbers of (110) type planes contribute to diffracted intensities measured at = 35.2 and 90 .Similarly for a film with ( 10 0) texture, 4 (110) type planes contribute to diffraction intensities at = 45 and 2 planes contribute to diffraction intensities at = 90 The ratios of the plane multiplicities at different pole angles are expected to corresp ond to the ratios of the diffraction intensities measured at these angles, or the volume of the texture component on the pole figure It should be noted that the volume of the texture components needs to be calculated by integrating over the pole sphere angles and T he nor malization condition for pole intensity distribution is defined as

PAGE 66

66 ( 3 28 ) is an arbitrary orientation distribution function of the texture. For fiber texture with the symmetry axis parallel to the sample normal the texture is invariant in Hence ( 3 28 ) can be simplified to ( 3 29 ) If the sample contains multiple texture components ( 3 29 ) c an be expanded as ( 3 30 ) Where, corresponds to the different texture components present in the sample. In PZT thin films deposited on platinized silicon substrates, (100) and (111) are the dominant texture components [102] henc e ( 3 30 ) can be simplified to the following form ( 3 31 ) and represent the orientation distribution functions of the (110) pole for the (100) and (111) texture components, respectively. (110) pole intensities due to (100) texture component is distributed at both = 45 and = 90 The ratio of the volume of these pole distributions is constrained by the multiplicity of the (110) planes at these angles. Based on the plane multiplicities ( 3 32 ) can be derived ( 3 32 )

PAGE 67

67 Similarly for (110) pole distribution corresponding to the (111) texture component ( 3 33 ) is obtained. ( 3 33 ) Based on ( 3 32 ) and ( 3 33 ) volume of the texture fractions in a sample containing multiple components of texture can be obtained. Consider that the volume fraction of the 100 texture component is Hence, ( 3 34 ) Us ing ( 3 32 ) the above expression can be simplified to ( 3 35 ) A similar expression can be derived fo r intensities corresponding to 111 texture component. ( 3 36 ) Combining ( 3 35 ) and ( 3 36 ) ( 3 37 ) From ( 3 37 ) it follows that for a samp le containi ng equal volumes of 100 and 111 texture components, t he ratio of the measured volume of ori entations corresponding to the 100 and 111 texture components would be equal to 4/3 (~1.67).

PAGE 68

68 3.8. Calculation o f Texture Component Volume a nd Volume Fraction To quantify the volume fraction of the different texture components, the intensities corresponding to the 111 and the 100 texture components were peak fit using the curve fitting tool box in MATLAB. The background was estimated using a linear polynomial. T he background intensity corresponds to the diffracted intensity from the randomly oriented grains. Peak fitting enabled the deconvolution of the diffraction intensities from the 111 and the 100 texture components. A representative result obtained through a fter peak fitting is presented in Figure 3 14 To estimate the volume of the 111 and 100 texture component, integration of the pole intensities needs to be performed over the pole figure angle The general formalism for integrating pole intensities over the pole angle is given in ( 3 28 ) ( 3 28 ) can be discretized to obtain ( 3 38 ) [132] ( 3 38 ) The refined peak parameters from the peak fitting routine were used to calculate the contribution of different texture components to observed diffraction intensities. The diffraction intensities of the texture componen ts were then individually integrated over ( 3 38 ) was calculated from the azimuthal angle by usin g ( 3 26 ) Texture fraction of the different texture components was calculated by using the volume of the texture component and the multiplicity of the planes. The equation used for the calcul ation of the texture com ponent i s shown in ( 3 39 ) ( 3 39 )

PAGE 69

69 Here represents the fractional multiplicity of the hkl texture component. The fractional multiplicities used for the different texture components are and 3.9. Temperature Calibration Due to the fast heating rates used in these studies, the r ate of change of temperature is considerabl y greater than the time constant for the response of the thermocouple. Hence, thermocouple based temperature measurement is not possible. The response time is much greater than the time scales at which measurement needs to made. To get an estimate of the temperature at various times during the experiment, the thermal expansion of NIST ceria was used. To embed ceria in PZT thin films, solutions for deposition of PZT thin films were prepared using the IMO route and NIST ceria was mixed into this solution after the final dilution steps. This mixture was agitated before being dispersed and spin coated onto the substrate. After spin coating, the film was pyrolyzed at 300 C or 400 C. Similar to the standard procedure for obtaining PZT thin films, two more deposition and pyrolysis steps were repeated. Diffraction images for these samples were then in situ collected during heating. Thermal expansion of ceria due to temperature increase would lead to a shift in the diffract ion peak positions towards lower 2 values. This change in the position of the Bragg peaks can be used to estimate the change in lattice parameter of ceria. The temperature of the sample can be calculated through the relationship between the change in latt ice parameter and the coe fficient of thermal expansion [133] However, the apparent shift in the 2 position of the ceria diffraction peaks is also influenced by the z

PAGE 70

70 height shift caused by the thermal expansion of the sample stage [134] Change in the z height due to t hermal expansion of the stage would lead to a shift in the position of the diffraction peaks towards a lower 2 Hence, evaluating the temperature of the thin film solely based on the thermal expansion would overestimate the temperature of the thin film. A n analytical correction model the z height displacement due thermal expansion of the sample stage was first attempted. However, the derived analytical model could not satisfactorily account for the variation in the lattice parameter and hence an empirical correction is used to estimate the temperatures. Details of the analytical model are provided in APPENDIX B To better understand the contribution of the thermal expansion of the stage towards the apparent shift in the 2 positions of the ceria diffraction peaks, ceria dispersed PZT thin films were slowly heated to a particular voltage and held for five minutes to allow for thermal equilibration. Diffraction measu rements and temperature measurement by using a thermocouple were then performed. The thermocouple was placed away from the incident and diffracted X ray beam paths to avoid any possible interaction. Such temperature and diffraction measurements were perfor med at particular voltage steps by sequentially increasing the voltage. The experimentally measured thermocouple temperature on equilibrated samples provides an independent measure of temperature and allows for correlating the relative change in the appare nt lattice parameter to the temperature of the thin film. However, the following sources of error could decrease the accuracy of the temperature measurement: (1) the area of the sample measured by the thermocouple is different from the area probed by the X ray beam, (2) The thermocouple could be preferentially heated by the IR lamp, and (3) a

PAGE 71

71 reliable contact between the thermocouple and the thin films might not be made. Based on these considerations, it is expected that the temperature measured by the ther mocouple is lower than the actual temperature of the thin film. During each of the hold step diffraction images of the films were taken. The apparent change in the lattice parameter of ceria was obtained by refining the position of the ceria (111) peak. T he variation of the ceria lattice parameter with increase in temperatures is shown in Figure 3 15 The trend in the variation of the lattice parameter with change in t emperature is observed to be fairly consistent. The anomalous decrease in the lattice parameter of ceria observed for sample#3 could be due to change in the sample position during heating. While solution deposited PZT thin films are expected to crystallize at ~600 C, the film was observed to not completely crystallize upon holding at 5.6V during the in situ experiments. Hence, the final voltage applied on the lamp was increased to 6.2V. The temperature corresponding to this voltage was measured to be 760 C. The lattice parameter of ceria obtained from diffraction measurement is 5.430 angstroms. This lattice parameter value cannot be used for estimating the apparent lattice expansion with temperature as the z height between samples could be different. The me thodology followed for estimating temperature at various time during in situ crystallization is shown in Figure 3 16 During these calibration runs, ceria dispersed P ZT thin films were crystallized at various heating rates and diffraction patterns were simultaneously collected. The experimental conditions for heating used on the calibration samples and during the actual experimental runs is the same. Hence, the tempera tures in the samples at various times could be approximated to the films with ceria dispersion

PAGE 72

72 The theoretical temperature was also calculated at the different voltage steps from the relative change in the lattice parameter and the coefficient of thermal expansion. Due to the reasons stated earlier, the theoretical temperature is expected to overestimate the temperature of the thin film. Hence, the actual temperature of the thin films is bound by the experimentally measured thermocouple temperature (T TC ) a nd the theoretical temperature obtained through the relationship between the relative change in lattice parameter and the coefficient of thermal expansion (T CTE ). These temperatures were then plotted with respect to time adjacent to the phase evolution and texture evolution plots ( Figure 3 17 ).

PAGE 73

73 Table 3 1 Mass absorption coefficients of the different elements at 20 keV [135] Element Atomic Weight (cm 2 /g) Pb 207.2 86.36 0.638 Zr 91.2 72.37 0.14 Ti 47.86 15.86 0.07 O 16.0 0.86 0.148 Table 3 2 Angles between the (110) type planes and (100) and (111) planes. Planes with angles greater than 90 are not listed as they are not detected in the diffraction geometry used at APS Plane Angle Number of planes 100 45.0 4 90.0 2 111 35.2 3 90.0 3

PAGE 74

74 Table 3 3 Temperature of the film measured using a thermocouple at different voltage settings on the lamp Voltage on lamp (V) Temperatures ( o C) Sample#3 Sample#4 0.0 25 25 2.0 155 140 3.0 295 260 4.0 440 430 4.5 520 500 5.0 602 580 5.5 670 655 5.6 680 670

PAGE 75

75 Figure 3 1 Euler cradle used for measurement of texture in materials Figure 3 2 Variation in the direction of with for the Inel diffractometer diffraction geometry. Radial d irection represents absolute value of the momentum transfer vector.

PAGE 76

76 Figure 3 3 Experimental setup used for the in situ synchrotron X ray measurements taken at beamline 6 ID B, Advanced Photon Source, Argonne National Laboratories. The blue line indicates the X ray beam path. Figure 3 4 Definition of detector angles and schematics for data extraction. A ) The definition of the angles i n the 2D diffraction geometry. B ) gamma integratio n for phase identification and C ) two theta integration for texture analysis Figure 3 5 Data representations of the in situ diffraction data A ) evolution of phases and B ) texture. A B

PAGE 77

77 Figure 3 6 Definition of terms for calculation of Absorption correction Figure 3 7 Absorption correction for different film thicknesses

PAGE 78

78 Figure 3 8 Rotation of diffraction cones due to tilting of the sample in the YZ plane. The position of the segregated intensities changes due to tilting. Figure 3 9 Schematic showing the diffraction geometry for the APS experiments

PAGE 79

79 Figure 3 10 Definition of the pole figure angles and based on the sample coordinates Figure 3 11 Section of (100) pole figure measured in the APS in situ diffraction experiments

PAGE 80

80 Figure 3 12 Pole figures corresponding to different types of texture. For the sake of illustration, only (100), (110) and (111) pole figures of (100), (110) and (111) type textures are represented. Solid line across the pole figure represents the section of th e pole figure measured in the APS diffraction geometry. Figure 3 13 (110) simulated pole figure for mixed (100) and (111) texture mode

PAGE 81

81 Figure 3 14 Representative result from the peak fitting procedure. Figure 3 15 Variation of apparent lattice parameter with change in temperature measured on two different samples A) sample#3 and B) sample#4 The decrease in the lattice parameter seen in A ) at ~650 could be due to change in sample position during heating. A B

PAGE 82

82 Figure 3 16 Methodology used for the determination of the temperature of the thin film at different times during the crystallization heat treatment.

PAGE 83

83 Figure 3 17 Contour plot showing phase evolution during in situ crystallization of PZT(52/48) film on Pt/Ti substrate. The change in the temperature of the film is shown in the plot in the right.

PAGE 84

84 CHAPTER 4 PHASE EVOLUTION IN P b O DEFICIENT SOLUTION DEPOSITED PLZT THIN FILMS The effect of lead precursor content in solution on the stability of the different phases formed during crystallization is not well understood. To characterize phase evolution in nominally lead deficient thin films, a new in situ high temperature laboratory X ray diffraction based measurement approach was developed. The well characterized Pb excess PLZT composition was used to compare and validate the newly developed XRD setup. The results described in this cha pter have been published in the Journal of Materials Science [128] 4.1. Introduction Traditionally, PZT solutions are batched with nominally excess lead precursor to include for any lead volatilization during thermal processing. However, the properties of the films were observed to deteriorate when gre ater amounts of lead were included in the starting solutions [136] particularly during fabrication of ultrathin films [137] The reaction between the film and Pt substrate has been suggested to be a possible cause for this degradation in properties [84] To decrease the possible reaction between the film and bottom electrode, Brennecka et al. [84] recently developed a new route for making ultra thin PZT films using solution deposition. The uniqueness of this approach is to deliberately start with a Pb deficient solution, which is meant to decrease the film substrate interaction. The resulting multi phase (stoichiometri c perovskite + Pb deficient fluorite) films are converted to single phase perovskite through a post crystallization heat treatment under a PbO overcoat. Using this approach, high quality ultra thin films of PZT can be made.

PAGE 85

85 Few studies have focused explic itly on phase evolution of Pb deficient PZT based materials. Polli and coworkers carried out a comprehensive study of phase formation in sol gel derived PZT powders [91] They observed that the fluorite phas e is the stable phase on crystallizing the powders at 500C. A metastable perovskite phase formed upon crystallizing the powders at higher temperatures ~ 600 700C and finally, a perovskite single phase was formed on crystallizing at ~800C. While this stu dy provides insight into the crystallization behavior of PZT based materials, phase evolution is expected to be significantly different in thin films due to the presence of the film substrate interface and the unique sample geometries. For example, Pt x Pb a metastable intermetallic phase, has been observed during crystallization of Pb rich thin films deposited on Pt bottom electrodes [78, 81, 104] This transient intermetallic phase is suggested to af fect phase and texture evolution in these thin films. X ray diffraction (XRD) is a standard technique for characterizing phase evolution in solution deposited thin films. Most studies using XRD for characterization of phase evolution in solution deposited thin films have been done using quenched samples or samples that were intermittently heated [79, 117] Sample to sample variability and the dynamic nature of phase evolution may introduce significant error in these methods. Phase evolut ion in solution deposited thin films is a dynamic process affected by the heat treatment history and the processing conditions. Hence, it is desirable to characterize phase evolution in thin films in situ during crystallization. In an earlier study, Wilkinson et al. studied phase evolution in sol gel derived powders in situ by collecting the diffraction pattern over a limited 2 range on a Bragg Brentano type diffractometer during crystallization [75] Multiple diffraction patterns were

PAGE 86

86 measured to map the evolution of phases with time. The kinetics of phase evolution were studied by tracking the integrated peak intensity of the (200) peak with respect to heat treatment time. It was observed that the fluorite phase always preceded the formation of perovskite in PZT and lead zirconate powders No fluorite phase was observed in lead titanate powders. Although this method provided valuable information about phase evolution in sol gel derived powders, it cannot be directly applied to solution deposited thin films because the movement of the detec tor over the required 2 range makes this technique inherently slow. Phase evolution in solution deposited thin films is faster in comparison to powders [92] and hence, it is desirable to develop a technique which allows for rapid characterization of phases during crystallization of thin films. With the above motivation, a new diffraction based technique was developed to study phase evolution in solution deposited thin films in situ during crystallization. Briefly, a curved position sensitive (CPS) detector is used to simultaneously collect diffraction data over a wide 2 range while the sample is heated to the crystallization temperature. The use of a CPS detector allows for rapid acquisition of diffraction data and characterization of phase evolution in the thin film. 4.2. Experimental Methodologies 4.2.1. Film Deposition La doped PZT thin films (PLZT) wer e deposited from solution on single crystal Si substrates with a sputtered Pt layer (Pt(170 nm)//Ti(40 nm)//SiO 2 (400 nm)//Si). For this study the La, Zr, and Ti contents were all maintained constant with a ratio of 6/52.2/46.3 (corresponding to a stoichiom etric PLZT composition of 6/53/47 with B site

PAGE 87

87 compensation for the donor doping associated with La addition). From this base composition, 20% of the Pb content was either added to the batch calculations (PLZT 6/53/47 + 20% Pb, referred to for the remainder of this chapter simply as Pb excess) or subtracted from the batch calculations (PLZT 6/53/47 20% Pb, referred to as Pb deficient). Solutions for spin coating were prepared by the inverted mixing order (IMO) method developed by Assink and Schwartz [37] The B site precursors (zirconium butoxide and titanium isopropoxide) were first weighed and mixed in glacial acetic acid. After mixing, the A site precursors (lead acetate and lanthanum acetate) were added to the solution. The solution was mixed at ~90 C to dissolve the Pb(IV) acetate into the solution. After dissolution of lead acetate, the solution was cooled and dil uted to the required molarity using acetic acid and methanol. The solution thus prepared was spin coated onto platinized silicon wafers at 3000 rpm. Three layers were spin coated and the films were pyrolyzed between each spin coating step and were pyrolyze d at 300C after each spin deposited layer The films were crystallized during heating within a furnace mounted on a diffractometer to allow in situ characterization of the crystallization process. The thin films were ~350 nm thick after crystallization. T he layer stack sequence at the end of processing is shown in Figure 4 2 4.2.2. Laboratory b ased In Situ Diffraction Setup A CPS detector was used to measure diffraction patterns in situ while the thin films were crystallized in the furnace during heating ( Figure 4 1 ) The CPS detector is a one dimensional detector which simultaneously measures the diffraction pattern over a wide 2 range. This allows for rapid acquisition of the diffraction data. The furnace used in the experiment is equipped with a Kapton sleeve to allow for the transmis sion of the

PAGE 88

88 incident X rays to the sample and the diffracted X rays to the detector. All films were crystallized in air by continuously heating from room temperature to a final temperature in the range of 600 650C at a rate of 5C/min. A heating rate of 5 C/min was used to minimize the uncertainty in the temperature. Patterns were each measured sequentially for an acquisition time of 60 s. Each pattern therefore represents the measurement of the time averaged structure across a 5C window of temperature. The slow heating rate also allowed for collection of considerable diffraction intensities from the different phases formed during crystallization. To compensate for the possibility of greater PbO volatilization due to the slower heating rate, for Pb excess films, 20% excess Pb precursor was added to the starting solution. After crystallization, the samples were cooled to room temperature at the same rate (5C/min). Diffraction patterns were measured during both heating and cooling. To collect the diffracti on patterns, the thin films were placed at an incidence angle of ~ 10 to the incoming X ray beam. The decrease in the intensity of the X ray beam as it travels through and interacts with the sample can be calculated by using the Beer Lambert law, ( 4 1 ) Where, I(t) is the final intensity. I 0 coefficient of the material and t is the distance travelled in the material. The decrease in the intensity of the incident X rays ( Figure 4 3 ) for the particular case of a multilayer stack can be calculated by slightly modifying equation (1) to include for attenuation due to overlying layers [125] Hence, a knowled ge of the linear attenuation coefficients [135] and the thickness of the material can be used to calculate the decay in intensity as the

PAGE 89

89 X ray beam penetrates the sam ple. The calculated intensity profile is shown in Figure 4 3 The incident X rays interact with all the different layers present in the thin films. However, diffracti on peaks only corresponding to the PZT thin film and the Pt bottom electrode are observed ( Figure 4 4 and Figure 4 6 ) in the measured diffraction patterns. This can be explained as follows. The titanium adhesion layer is relatively thin (40 nm). Any weak diffraction peaks arising from the titanium adhesion layer are heavily attenuat ed by the overlying PZT and Pt layers. While SiO 2 is present in significant quantities (400 nm), SiO 2 has a small X ray cross section and is amorphous in these samples. Therefore, diffraction intensities from SiO 2 are comparatively weak. This intensity is also heavily attenuated from reaching the detector due to the overlying Pt and PZT layers. A significant amount of the X ray intensity reaches the silicon substrate and is diffracted by the silicon. However, as the silicon substrate is a single crystal, s trict conditions have to be met to observe the diffraction spots resulting from silicon using this diffraction geometry. Thus, the diffracted X rays from silicon are most often not found in the diffraction pattern measured on the Inel diffractometer The possibility of observing silicon diffraction peaks ({110} and {113}) in the Inel diffractometer geometry was further decreased by slight in plane rotation such that the (110) are not parallel to a Bragg plane in the present geometry. The <110> of the singl e crystal was estimated by using the wafer flats marked on the wafer and by tracking the specific orientation of the cut samples. Certain peaks in the measured XRD patterns were fit to profile shape functions to extract quantitative information. In peak fi tting, the background of the entire diffraction

PAGE 90

90 pattern from 2 = 20 to 2 = 80 was fit first using an eighth order polynomial function. The peaks were then each individually fit to a Pearson VII type function using the curve fitting tool box in Matlab. The Pearson VII function was modified to include diffracted intensities from both K and K of the incident X ray beam; the position between the components was constrained by their difference in wavelength and the peak shape parameters of each component were set equal. Lattice parameters were calculated after correcting for the shift in 2 posit ions due to sample displacement by thermal expansion [134] Grazing incidence X ray diffraction (GIXRD) patterns were measured on the measured at different incidence angles to probe different depths of the thin film. 4.3. Results The first crystalline phase observed during crystallization of Pb excess thin films is a Pt x Pb intermetallic phase. The evolution of the diffraction pattern of the Pb excess thin film during crystallization is shown in Figure 4 4 The changes in the phases formed during the crystallization of Pb excess thin films are shown in Figure 4 5 Based on previous reports on the Pt x Pb intermetallic, it was assumed that any intermetallic formed would be crystalline and will be observed in the diffraction pattern [78] The maximum amount of Pt x Pb is observed to be present at T=4302.5C. The intermetallic phase eventually disappears as the film is ramped to higher temperatures and is no longer seen after T=4712.5C. A fluorite phase appears at T=4402.5C and reaches maximum diffraction intensity at T=4712.5C. Finally, the perovskite phase appears at

PAGE 91

91 T=5112.5C and reaches a maximum by T=5522.5C. The appearance of the perovskite phase and the disappearance of the fluor ite phase seem to be correlated Figure 4 5 is a contour plot on a limited 2 and time region showing the disappearance of the fluorite phase and the appearance of the perovskite. In Pb deficient thin films, a perovskite and fluorite type phase appear together at T=5062.5C. The evolution of the diffraction patterns during crystallization is shown in Figure 4 6 The evolution of the phases during crystallization is shown in Figure 4 7 The amount of fluorite observed reaches a maximum at T=5532.5C. The perovskite reaches a maximum at T=5802.5C. Thereafter, on holding at T=600C, an additional distinct perovskite phase is observed. To distinguish from the perovskite phase formed in the Pb excess thin films, the first perovskite phase formed in the Pb deficient thin film is designated as perovskite 1 and the secondary perovskite phase formed later in processing time is called perovskite 2 for the remainder of this chapter There is a decrease in the observed fluorite phase content from the point where the perovskite 2 phase is observed. At the end of the heat treatment the fluorite type, perovskite 1 and perovskite 2 phases are all observed to be present. The lattice parameters of the perovskite phases formed in Pb excess and Pb deficient thin films are listed in Table 4 The lattice parameters of the perovskite phases were calculated at 500C from the diffraction patterns measured during the cooling of the sample after crystallization. Since this temperature is above the Curie temperature of the perovskite phases, the crystal structure was assumed to be pseudo cubic.

PAGE 92

92 4.4. Discussion In Pb excess thin films ( Figure 4 8 ), the temperature ran ge in which the Pt x Pb intermetallic phase is observed is consistent with that reported by Chen et al. [81] (400 700C) and Huang et al. [117] (330C). From the observed diffraction peaks of the Pt x Pb intermetallic, the structure corresponds to the Pt 3 Pb phase reported by Huang et al. [117] The temperature at which the perovskite phase is observed to form is consistent with that observed by Kwok and Desu [110] The observed sequence of phases formed during crystallization in the Pb excess thin films is found to be consistent with previous reports on crystallization of PZT thin films [70, 106, 110] Pt x Pb is not observed to form in the Pb deficient thin films ( Figure 4 6 ), indicating that this reaction between the electrode and the thin films is reduced with a decrease in the Pb content in the thin film. Several possible reasons fo r the absence of Pt x Pb in Pb deficient thin films can be hypothesized. First, due to the lower amount of total PbO present in the film, the rate at which Pt x Pb is formed may decrease. However, the rate at which oxygen diffuses from the atmosphere to the fi lm substrate interface remains unchanged. Pt x Pb is unstable in the presence of oxygen and decomposes to PbO and platinum. Due to the decreased rate of formation of Pt x Pb in Pb deficient films, the rate of oxygen diffusion from the atmosphere may be suffici ent to destabilize any Pt x Pb formed. Second, it is possible that the amorphous phase consists of a nanocrystalline fluorite phase [72] which is not detected by X ray diffraction. This nanocrystalline fluorite phase could control the availability of free PbO, leading to very little availability of PbO for reduction and subsequent formation of Pt x Pb at the interface. Polli et al. [91] observed that the formation of the fluorite and perovskite phases is delayed with a decrease in the Pb content in the powders. The trend in fluorite phase

PAGE 93

93 formation observed in this study (T~4405C for Pb exces s and T~5065C for Pb deficient) agrees with the trend reported by Polli et al. However, the perovskite phase formation observed in the present work does not agree with the trend reported for sol gel derived powders in Polli et al. The perovskite phases w ere observed to form at similar temperatures (T~5065C for Pb deficient and T~5175C for Pb excess) for the thin films. Moreover, the fluorite phase and the perovskite 1 phase form at the same time and temperature in the Pb deficient thin film. These dis crepancies with earlier results are explained as follows. Nucleation of the perovskite phase in thin films preferentially occurs heterogeneously at the film electrode interface. As crystallization of solution derived PZT is nucleation dependent [86] the temperature at which the perovskite phase is first observed is determined by the stabilization of the perovskite nuclei by the interface. Although the growth rate of the perovskite nuclei has bee n observed to be affected by the Pb content in the films, the nucleation rates were found to not significantly differ with Pb content (Pb excess = 0 20%) in PZT thin films [138] The observation of the formation of perovskite phases in both Pb excess and Pb deficient films at similar temperatures implies that the nucleation rates could be similar for both Pb excess and Pb deficient films used in this study. During cry stallization, the perovskite phase in Pb excess thin films is observed to reach its maximum intensity by T~550C (10 min), while the perovskite 1 phase in Pb deficient films reaches a maximum by around T~580C ( Figure 4 9 ). Deviation from ABO 3 stoichiometry in PZT is expected to destabilize the perovskite phase. However, in the present study a Pb excess in the solution is observed to help to crystallize the film faster than starting with a Pb deficient solution. Similar behavior was observed in

PAGE 94

94 solution derived powders [91] and sol gel deposited PZT layers on LaAlO 3 [138] substrates. Based on EXAFS studies by Sengupta et al. [53] Polli and coworkers [91] hypothesized that excess Pb acts as a network modifier that assists in rearranging the atoms for the formation of the required perovskite phase. This is a possible explanation for the faster crystallization of the perovskite phase in Pb excess thin films. It should also be noted that the starting solutions nominally varied from stoichiometry by the same magnitudes, though in different directions. Pb loss during processing may bring the initially Pb rich films closer to stoichiometry w hile pushing the Pb deficient films further from stoichiometry. Parish et al. [139] characterized the final phases obtained after crystallization of Pb deficient PLZT (12/7 0/30) thin films. Chemical segregation was observed in thin films which had been crystallized at 650C for 30 min using a 50C/min furnace ramp rate. The thin films were found to contain a perovskite and a fluorite phase. The perovskite phase was found to be Pb rich and La deficient relative to the nominal composition and was formed closer to the electrode than the fluorite phase, which was Pb deficient, La rich, and present near the top surface of the film. In the thin films in this study, apart from the f luorite and perovskite 1 phases, an additional perovskite phase (perovskite 2) is observed. Substitution of La 3+ for Pb 2+ in PZT causes a decrease in unit cell volume and lattice parameter [140] The diffraction peaks of this La substituted perovskite phase would be observed at higher 2 values relative to the perovskite 1 phase formed initially. Hence, it is proposed that the perovskite 2 formed in the Pb deficient films crystallized from the remnant Pb poor and La rich fluorite phase at lon ger times or slower ramp rates.

PAGE 95

95 Segregation in PZT thin films due to easier nucleation of lead titanate (PT) rich solid solutions at the interface has been reported to occur during crystallization [141] PT rich compositions have a smaller unit cell volume compared to lead zirconate (PZ) rich compositions [10] Early crystallization of a PT rich phase during in situ crystallization would show in the diffraction pattern as peaks at slightly higher 2 values compared to the later formed PZ rich phase. However, during crystallization of the Pb deficient thin film, diffraction peaks from the earlier formed perovskite 1 phase are observed at lower 2 values compared to the later formed perovskite 2 phase. Hence, the perovskite 1 unit cell is larger than that of perovskite 2 ( Table 4 ). Since it is unlikely that a PZ rich composition would crystallize first, therefore perovskite 1 and perovskite 2 phases are possibly not due to PT rich and PZ rich regions seg regating in the thin film. Due to the specific geometry of the X ray diffractometer used in this study, the poles of the diffracting planes are not perpendicular to the sample surface. Hence, texture relationships cannot be directly interpreted from the diffraction data. However, it can be qualitatively observed that (00l) type peaks are present in Pb excess films Figure 4 4 )and are not present in Pb deficient films ( Figure 4 6 ). This is consistent with previous reports suggesting Pb excess is a necessary condition for the formation of (00l) oriented grains [106] GIXRD was undertaken to selectively probe the structure of the thin film. By varying the angle of incidence of the incoming X ray beam, different depths of the thin film were probed. The change observed in the diffraction pattern with change in a ngle of incidence is shown in Figure 4 10 It is observed that the perovskite 1 phase is present in greater quantities closer to the interface of the film with the bottom electrode, while the perovskite 2 phase and the fluorite phase are present closer to the

PAGE 96

96 surface ( Figure 4 10 ) This observation agrees well with the results of Parish et al. [139] A possible explanation for t he proximity of both the fluorite and the perovskite 2 phase to the surface is that the perovskite 2 phase is formed from the segregated fluorite phase. Formation of a secondary perovskite phase was not observed in earlier studies on Pb deficient PLZT thin films made by similar methods [84, 139, 142] The perovskite 2 phase was observed to form in this study even though the final temperatures to which the thin f ilms were heated were comparable or lower. The longer time available for crystallization in these experiments could be the reason the perovskite 2 phase was observed. In this study, the thin films were heated using a ramp rate of 5C/min, while Brennecka et al. [84] and Parish et al. [139, 142] crystallized the thin films using a heating rate of 50C/min or by inserting the sample in a furnace pre heated to 700C. The slower ramp rate used in the present work (5C/min) may provide additional time for PbO volatility as well as cation diffusion and segregation. 4.5. Conclusions Phase evolution in solution deposited PLZT thin films was investigated by using a newly developed in situ diffraction technique. The observed phase evolution in Pb excess PLZT thin films was consistent with the results reported on the temperature of formation of Pt x Pb by other groups. For Pb deficient PLZT th in films, the film is observed to partition into two phases: (1) a Pb rich and La poor perovskite phase and (2) Pb poor/La rich fluorite phase. The fluorite phase partially transforms to a perovskite phase at higher temperature. Moreover, Pt x Pb, an interme tallic formed due to interaction between the thin film and the electrode, was not observed in the Pb deficient thin film.

PAGE 97

97 It is proposed that this in situ high temperature diffraction based technique can be used to characterize the effects of dopants, pyro lysis and solution chemistry on phase evolution. Knowledge of how different variables affect phase evolution can help to identify strategies to decrease the crystallization temperature.

PAGE 98

98 Table 4 1 Lattice parameters of the different perovskite phases formed in the thin films investigated. Perovskite 1 and perovskite 2 refer to the perovskite phases formed in Pb deficient thin films. Lattice parameters were calculated at ~ 500C assuming a pseudo cubic perovskite crystal structure. Film Composition Phase Lattice parameter () Pb excess Perovskite 4.08 Pb deficient Perovskite 1 4.08 Pb deficient Perovskite 2 4.02

PAGE 99

99 Figure 4 1 Laboratory based setup used for in situ crystallization of PZT thin films Photo courtesy of Krishna Nittala

PAGE 100

100 Figure 4 2 Cross section of the thin film stack and interaction of the X rays with the thin films Figure 4 3 Attenuation of the X ray intensity due to absorption by the different lay ers

PAGE 101

101 Figure 4 4 Phase evolution in PLZT + 20% Pb excess film Figure 4 5 Limited 2 surfac e plot showing the evolution of the fluorite and perovskite phases

PAGE 102

102 Figure 4 6 Phase evolution in Pb deficient thin film Figure 4 7 Limited 2 theta plot showing the evolution of the fluorite and perovskite phases. A secondary perovskite phase is observed to form.

PAGE 103

103 Figure 4 8 Phase evolution during crystallization of PLZT + 20% excess PbO thin film

PAGE 104

104 Figure 4 9 Phase evolution during crystallization of PLZT + 20% PbO deficient film Figure 4 10 GIXRD plots for the Pb deficient thin films at different angles of incidence

PAGE 105

105 Figure 4 11 Normalized integrated intensities obtained at increasing x ray incidence angles

PAGE 106

106 CHAPTER 5 AGING IN PRECRYSTALLIZED SOLUTION DERIVED PZT THIN FILMS Samples for the in situ measurements performed using the high temperature lab oratory and synchrotron X ray diffraction setups were fabricated at Sandia National Laboratories and then transported to University of Florida and APS, respectively. The effect of delay between the pyrolysis and crystallization treatments on the crystallization behavior of these thin films is not well understood This chapter provides the details of a systematic study performed to understand the effect of delay between pyrolysis and crystallization heat treatments on the crystallization behavior o f these thin films. 5.1. Experimental Methods The solutions used for film deposition were prepared through the Inverted Mixing Order (IMO) route developed by Assink et al. [37] Briefly, precursors of the B site cations (zirconium butoxide, Zr(OBu) 4 80% in 1 butanol and titanium isopropoxide, 9 8 % both from Sigma Aldrich, Inc., St. Louis, MO) were first vortexed together for 5 minutes. Acetic acid was then added to this mixture to chelate the zirconium and titanium precursors. After diluting the solution with methanol (MeOH, electronic grade, Fisher Scientific), lead (IV) acetate was added and dissolved at ~ 90 C. To compens ate for lead volatility during processing a 15 mole % excess of the nominal stoichiometric quantity of Pb was added. After complete dissolution of lead acetate, the solution was diluted to the required molarity by further addition of acetic acid and methan ol. All solutions used in this work had a final molarity of 0.35 M. Pt (170 nm)/ Ti (40 nm)/ SiO 2 (400 nm)/ Si) used for deposition of the films were acquired from Silicon Quest International (SQI), Inc., Santa Clara, CA. Films were deposited by spin coating the solution onto the substrate at

PAGE 107

107 3000 rpm for 30 s using a standard photoresist spin coater (Headway Research, Inc., Garland, TX). Films were pyrolyzed after each deposition step. In this work, two different pyrolysi s temperatures were used: 300C or 400C. After three deposition and pyrolysis steps, the wafer was cut into smaller pieces and the pieces were aged in different relative humidity (RH) conditions. Controlled relative humidity conditions were maintained by placing supersaturated salt solutions at the bottom of the desiccators. The specific super saturated salt solutions used are 1. LiCl (RH = 15%), 2. K 2 CO 3 (RH = 43%), 3. NaBr (RH = 58%) and CuSO 4 (RH = 98%) [133] After aging the pyro lyzed films for specific times in these controlled humidity conditions, the films were characterized using FTIR and then crystallized. The films were crystallized by directly inserting into a preheated furnace at 700 C and holding for 10 minutes. Phase cha racterization of the crystallized thin film samples was done using X ray diffraction. The X ray diffraction measurements were performed using a Philips APD 3720 diffractometer. To qualitative ly evaluate the texture of the films, t he Lotgering factor of the (111) diffraction peak was used The equation used for calculating the Lotgering factor is shown in ( 5 1 ) ( 5 1 ) For electrical testing of the films, Pt electrodes were deposited on the films by radio frequency (rf) sputtering through a shadow mask. Polarization field measurements were performed using standard semiconductor probes (Signatone, Gilroy, CA) connected to a Precision Workstation (Radiant Technologies, Al buquerque, NM) at 100 Hz and the maximum applied voltage was varied from 2.5V 20 V with a constant

PAGE 108

108 increment of 2.5V. The variation of c apacitance and loss tangent with change in frequency was measured using similar probes connected to an HP4294A precisi on impedance analyzer using a probe voltage of 100 mV. Microstructural characterization was performed using SEM. Solution derived powders were obtained by evaporating the solvent from the final solutions by continuous stirring at 150C. The powders thus ob tained were then pyrolyzed at 300C and 400C for 5 minutes by inserting them in a box furnace. These powders were then aged in desiccators with controlled humidity conditions maintained using supersaturated CuSO 4 solution (RH = 98%). After aging for speci fic times, the chemical changes in the powders were characterized using FTIR. Thermogravimetric data was then obtained on the powders using a NETZSCH STA 449 F3 Jupiter system conditions. The powders were heated from room temperature to 700C at a heating rate of 10C/min while compressed air was flown through the system The FTIR characterization of both solution derived powders and thin films was performed using a Varian Scimitar 2000 Fourier Transform Infrared Spectrometer with a Pike MIRacle ATR attachm ent. A FTIR scan of air was used for background subtraction. 5.2. Results 5.2.1. Electrical Characterization For films pyrolyzed at 300 C, variation in the polarization field measurements was observed in all relative humidity conditions. With aging time, the polariz ation field loops are observed to become more open. While the remnant polarization as observed to vary with aging, no systematic variation was observed with aging time. The measured hysteresis loops were observed to become more open for films aged in highe r relative humidity conditions ( Figure 5 1 A and B ). The hysteresis loops measured on films

PAGE 109

109 pyrolyzed at 400 C also showed variation with aging time ( Figure 5 2 A D ). However, the openness of the loops was not observed to change significantly. No systematic variation in the remnant polarization was observed for films pyrolyzed at 400 C. The dependence of the relative dielectric constant and tan with frequency for films pyrolyzed at 300 and 400 C for the different relative humidity conditions are shown in Figure 5 3 and Figure 5 4 respectively. To identify specific trends with aging time and relative humidity condition, the dielectric constant and loss values at 100 Hz for both films pyrolyzed at 300 C and 400 C were extracted and are represented in Figure 5 5 and Figure 5 6 respectively. Variation was also observed in the dielectric properties of both films pyrolyzed at 300 C and 400 C with aging time. For films pyrolyzed at 300 C, a sudden drop in the relative dielectric constant is observed for films aged for 6 days in 20% RH ( Figure 5 3 A ) and 2 days in 98% RH ( Figure 5 3 D ) at ~10 4 Hz and the tan values are observed to peak at this frequency. This relaxation seems to occur at the frequency range typical for a space charge type relaxation mechanism. However, t his relaxation is not observed for the rest of the films. The variation in the relative dielectric constant ( Figure 5 5 ) seems to be higher for films aged in lower RH conditions (20% an d 44%) than the higher RH conditions (55% and 98%). The tan values for these films were observed to increase initially upon aging and then decrease. The relative dielectric constant for films pyrolyzed at 400 C and crystallized immediately (~1090 at 100H z) is higher than for films pyrolyzed at 300 C and processed similarly (~830 at 100Hz). The relative dielectric constant of the films pyrolyzed at 400 C seems to change with aging time for all RH conditions ( Figure 5 6 )

PAGE 110

110 In general, the dielectric constant seems to decrease with increasing aging time. The tan for films aged in 98% RH seems to increase with increase in aging time. No significant change was observed in the variation of tan values with aging for 20%, 44% and 55% RH conditions. 5.2.2. FTIR Measurements The FTIR spectr a of films pyrolyzed at 300 C and 400 C and aged in different humidity conditions are shown in Figure 5 7 and Figure 5 8 respectively Resonance peaks at 1533 cm 1 and 1393 cm 1 correspond to the anti symmet ric and symmetric vibrational stretching modes of the acetate groups ( (C O 2 )) (labeled 1 and 2 in the FTIR spectra). Resonance peaks are also observed at ~1010 cm 1 and ~930 cm 1 for films pyrolyzed at 300 C (labeled 3 in Figure 5 7 ). These peaks correspond to the (C O)M stretching modes of the alkoxy groups. The two peaks centered ~1010 cm 1 and ~930 cm 1 are from titanium isopropoxide and zirconium butoxide, respe ctively [49, 143, 144] Resonance peaks corresponding to (MO) are expected to occur in 300 600 cm 1 range. However, these peaks could not be measured due to the limitation of the FTIR instrument. In addition to the peaks from the or ganic moieties in the film a broad peak centered ~3200 cm 1 is observed for all films This peak seems to correspond to the surface hydroxyl stretching mode ( ( OH) ) The resonance peak from the surface hydroxyl group seems to be present in films pyrolyze d at 300 C and 400 C for all aging times. No systematic changes in the intensity of this peak were observed with both the amount of relative humidity and aging time. For films pyrolyzed at 400 C, resonances corresponding to the acetate group or the metal a lkoxide groups are not observed immediately after pyrolyzation ( Figure 5 8 ).

PAGE 111

111 No significant changes are observed in the FTIR spectra of these samples when aged in 20%, 44% and 55% RH levels. However, for films aged in 98% RH, the acetate and the alkoxy groups are ob served ( Figure 5 8 D ) upon aging for 1 day and are present in samples with higher aging times. No significant change was observed in the peak intensities of the acetat e resonances with aging time. However, similar to the films pyrolyzed at 300 C and aged in 98% RH, the peak widths were observed to increase with increase in aging time. 5.2.3. Microstructural Characterization u sing SEM Both films pyrolyzed at 300 C and 400 C we re observed to have a fine grained microstructure from the plan view ( Figure 5 9 A and Figure 5 13 A ). The coverage of this fine grained material is observed to be higher for films pyrolyzed at 300 C than for films pyrolyzed at 400 C. From the micrographs of the cross sections ( Figure 5 9 B and Figure 5 13 B ), the films crystallized immediately after pyrolysis were observed to have a columnar grain structure. No changes in the microstructure were observed for films pyrolyzed at 300 C and aged i n 20% RH ( Figure 5 9 A F ). For films aged in 44% RH, isolated rosette type grains are observed for films aged for 9 days ( Figure 5 10 C ) in the plan view. However, no rosette type grains were observed for films aged for 30 days ( Figure 5 10 E ). No changes were observed for films aged in 44% RH in the cross section views. No significant changes in microstructure were observed for films aged in 55% RH both in the plan and cross section micrographs ( Figure 5 11 A F ). For films aged in 98% RH, no significant changes were observed between films crystallized immediately after pyrolysis and those aged for 9 days ( Figure 5 12 A D ). Upon aging for 30 days,

PAGE 112

112 considerable numbers of rosette type grains were observed to form in these films ( Figure 5 12 E ). Additionally, porosity was observed to be present in the middle of the films aged for 30 days ( Figure 5 12 F ). For films pyrolyzed at 400 C and aged in 20% RH, the coverage of the fluorite type phase seems to increase with aging time ( Figure 5 13 A C E and G ). How ever, no significant differences are observed in the cross sections of the sample ( Figure 5 13 B D F and H ). For films aged in 44% RH, isolated rosette type grains are observed to form for films aged for 8 days ( Figure 5 14 E ). However, no significant changes are observed in the cross sectional images of the films ( Figure 5 14 B D F and H ).The film microstructure was observed to resemble a columnar grain structure for the rest of aging conditions ( Figure 5 14 A C and G ). For films aged in 55% RH, an increase in the coverage of the fluorite phase was observed with increase in aging time ( Figure 5 15 A C E and G ). In addition, isolated rosette type grains are observed for films aged for 8 days ( Figure 5 15 E ). All fi lms were observed to have a columnar microstructure and this was consistent for all aging times ( Figure 5 15 B D F and H ). For films aged in 98% RH, compared to fi lms crystallized immediately after pyrolyzation ( Figure 5 16 A and B ), the film morphology is observed to change significantly with aging. For films aged for 2 and 8 days, extensive formation of rosette type grains was observed in the plan view images ( Figure 5 16 C and E ). From the cross sectional images, these rosette type grains were observed to be localized to the surface, while columnar type grains were formed t owards the interface ( Figure 5 16 D and F ). For films aged for 30 days, the surface was observed to consist entirely of rosette type grain structure ( Figure 5 16 G ). These rosette type grains were observed to be formed close to the surface, while

PAGE 113

113 different grains nucleated towards the film substrate interface ( Figure 5 16 H ). Porosity is also observed for these films in the middle, between the two different grains. 5.2.4. XRD Measurements The films were observed to be single phase perovskite from the XRD measurements. In addition to the expected diffraction peaks from the perovskite phase and the platinum substrate, additional peaks are observed at 2 ~ 32 and 2 ~34 These peaks were observed only for films pyrolyzed at 300 C. These peaks are designated to be from the under layers of the platinized silicon substrate. For films pyrolyzed at 300 C, the diffracted intensity from the (100) perovskite planes is observed to be very low. Indicating that these films are predominantly (111) oriented ( Figure 5 17 A D ). Some changes in the intensity of the (100) peak were observed with aging time. To quantify the changes in the texture of the film with aging time, th e lotgering factor for the (111) plane (f 111 ) was calculated. The lotgering factor is a crude measurement of texture [126] however, it is appropriate for qualitatively assessing the changes taking place in the texture of the film. The variation of f 111 with aging time for the different RH conditions is shown in Figure 5 18 No significant changes were observed in the thin films with aging time or relative humidity conditions. For films pyrolyzed at 400 C, the diffraction intensities of the (100) perovskite planes seems to be higher than for films pyrolyzed at 300 C ( Figure 5 19 ) This indicates that the films pyrolyzed at 400 C (f 111 = 0.70) are less (111) textured than the films pyrolyzed at 300 C (f 111 = 0.97). No significant changes are observed to take place for films aged in 20%, 44%, and 55% RH ( Figure 5 20 ). For films aged in 98% relative humidity ( Figure 5 20 ), a sudden increase in f 111 is observed between films crystallized

PAGE 114

114 immediately after pyrolysis and films aged for 1 day and then crystallized. The f 111 upon further ag ing in 98% RH of films pyrolyzed at 400 C seems to not change significantly. 5.3. Discussion 5.3.1. Aging o f Thin Films Acetic acid is used during the solution preparation process to stabilize the metal alkoxide precursors against reaction with water. Acetic ac id stabilizes the metal alkoxide by forming a coordination complex. Three modes of coordination (Unidentate ( Figure 5 21 A ), chelating ( Figure 5 21 B ) and bridging complexes ( Figure 5 21 C ) ) are possible between acetic acid and the metal alkoxide. These three modes can be differentiated based on the difference in the peak positions of the symmetric and anti symmetric resonanc es of the acetate group [144] The difference in the peak resonances for the acetate groups for the films investigated in this study was found to be ~ 140 cm 1 This value is le ss than the difference in peak resonances for ionic complexes of acetic acid (~201 cm 1 ). This indicates that the acetate group is present as a chelating complex with the metal ions in these films. During pyrolysis, the main processes that occur are: (1) e vaporation of any remnant solvents and (2) break up of metal organic compounds and subsequent removal of organic moieties [39] While the specific temperature required for the complete removal of the solvent is dependent on the solution preparati on route, it is mostly complete by ~ 200 C. However, organic moieties, particularly in the form of acetates are known to persist until higher temperatures [70] Higher pyrolysis temperatures would help to remove a greater amount of the remnant or ganics. This is consistent with the FTIR spectra wherein resonance peaks corresponding to the acetate

PAGE 115

115 groups are observed in films immediately after pyrolysis at 300 C but are not observed in films pyrolyzed at 400 C. For films pyrolyzed at 300 C no signif icant changes are observed in the peak intensities of the acetate resonances with time. However, the acetate peaks seem to broaden with time. To quantify this variation of the acetate peaks with time, the resonances corresponding to the acetate groups were peak fit with a Gaussian function. Peak fitting was performed using the curve fitting toolbox in Matlab. Full width at half maximum (FWHM) of the peaks was used as a representative metric for the width of the peaks. The variation in the FWHM of the anti symmetric resonance of the acetate group ( a (CO 2 )) is shown in Figure 5 22 The FWHM of the peak was observed to slightly decrease with aging time for films aged in 20%, 44% and 55% relative humidity. However, for films aged in 98% RH, the FWHM was observed to incre ase with aging time. Significant changes in the FTIR spectra are only observed for films pyrolyzed at 300 C and 400 C and aged in 98% RH. Lakeman et al. [70] reported that the acetate groups are the last organic moieties that escape from the films for 2 MOE based solution deposited PZT thin films. In the IMO based solution deposited PZT thin films used in this study, IR resonance signatures from the acetate and alkoxy groups are observed in the films pyrolyzed at 300 C but are not observed for film s pyrolyzed at 400 C. This trend in consistent with the results of Lakeman et al. [70] Coffman et al. reported the formation of carbonate (CO 3 2 ) groups in sol gel derived PZT powders upon prolonged exposure to air [145] The carbonate groups were suggested to form due to the absorption of carbon dioxide from the atmosphere. In the films used in this study,

PAGE 116

116 the formation of the carbonate groups ( (CO 3 2 ) ~ 900 cm 1 ) is not observed. The appearance of the acetate and alkoxy groups in films pyrolyzed at 400 C and aged in 98% RH is presumed to be due to the reformation of these groups. The exact mechanism for the reformation of these groups is not clear. However, very high (98%) RH levels seem to be required for the reformation of these organic moieties. The symmetric and anti symmetric stretching modes are normal vibrational modes and the frequency for resonance of these modes is dependent on the bond strength [146] Hence, the spread in the intensity centered about the peak value correspon ds to the variation in the bond strength for a particular bond. In films pyrolyzed at 300 C and 400 C and aged in 98% RH, the FWHM of the acetate anti symmetric stretching mode is observed to increase. This implies that the variability of the metal acetate bond strength seems to increase with aging time. Therefore, a change in the chemical in films aged at 98% RH. Changes in the chemical character of the films, eithe r due to the specific solution chemistry or due to the addition of particular solvents, could be the reason for the differences in the final microstructure of solution derived PZT thin films [38] A fine grained material is observed in the plan views of the crystallized films appears to be localized towards the surface of the thin film from the cross sectional imaging of the sample ( Figure 5 9 B and Figure 5 13 B ). This fine grained material could be a fluorite type phase formed due to incomplete conversion for the thin film. Lack of stoichiometric amou nts of the lead precursor in the thin film could stabilize the fluorite phase during processing [74] However, no fluorite type phase was detected in XRD.

PAGE 117

117 This inco nsistency could be due to the limited sensitivity of coupled /2 X ray diffraction measurements to the thin surface fluorite layer. The increased tendency for the formation of rosette type grains in films aged in 98% RH could be due to changes occurring i n the pyrolyzed films during aging. Rosette type grains are formed from homogenous nucleation while columnar grains are formed from heterogeneous nucleation [38] .The correlation observed between the formation of rosette type grains structure and the increase in the FWHM of the acetate resonance peaks in the FTIR spectra suggests that chemical changes could be taking place in the thin film during aging. The microstructure of the films pyrolyzed at 400 C seems to be more susceptible to aging in high humidity conditions. Rosette type grains are observed in films pyrolyzed at 400 C and aged in 98% RH for 2 days. No rosette type grains are observed in films pyrolyzed at 300 C and aged in 98% RH even after 9 days of aging. Since the acetate groups s eem to undergo similar changes with aging, the cause for the difference in the formation of the rosette type grains morphology is not known. The gel state of the 300 C pyrolyzed films could have a higher tolerance for chemical changes than the 400 C pyroly zed films. Dielectric fluorite phase has a lower relative dielectric constant and the presence of this fluorite phase decreases the final dielectric properties of the film. Fluorite coverage observed in the plan view SEM images seems to be lower for films pyrolyzed at 400 C and immediately crystallized in comparison to films pyrolyzed at 300 C and similarly processed. This lower fluorite coverage could be the reason for the higher dielectric constant observed in the 400 C pyrolyzed films.

PAGE 118

118 This trend of decrease in the (111) texture with increase in pyrolysis temperature is consistent with the trend observed in studies done at Sandia National Laboratories [1, 105] The increase in the f 111 ( Figure 5 20 ) seems to correspond well with formation and presence of the acetate groups in the films pyrolyzed at 400 C and aged in 98% RH ( Figure 5 23 ). The presence of greater amount of remnant organic moieties in the film after pyrolysis was suggested to cause reducing conditions to be present at the film substrate interface [106] The presence of reducing con ditions during crystallization could preferentially nucleate (111) orientation in these thin films [82, 121] 5.3.2. Solution Derived Powders Changes are observed in the microstructure and properties of the films with aging. These changes pyrolyzed film. To better understand the changes taking place in the gel state, powders were isolated from IMO solutions by drying the solutions at 150 C. To mimic the state of the fi lms after pyrolysis, the powders were pyrolyzed at 300 C and 400 C for 5 minutes. Since films aged in 98% RH seem to be most affected. Aging of solution derived powders were only performed at 98% RH. 5.3.2.1. FTIR For powders pyrolyzed at 300 C, the reson ances from the stretching modes of acetate and alkoxy groups are observed (labeled 1, 2 and 3 in Figure 5 24 A ). This is consistent with the FTIR spectra obtained for films pyrolyzed at 300 C ( Figure 5 7 ). However, some differences are observed between the FTIR spectra of powders and film. In films, the intensity of the resonance p eaks from organic moieties is not observed to change with time. While in powders, the intensity of the intensity of the resonance

PAGE 119

119 peaks of the organic moieties is observed to increase with time. In addition, the broadening observed in the acetate resonance peaks in films aged in 98% RH ( Figure 5 22 ) is not observed in powders ( Figure 5 25 ). The reasons for this difference in behavior are currently not known. In addition to the changes in peaks of the expected organic moieties, a broad peak is observed centered ~ 3250 cm 1 This peak seems to correspond to (OH ). This could be either due to absorption of water by the powders or due to reformation of alcohols. Since no resonance peaks corresponding to C H bonds or alkyl groups are observed, this peak is presumed to be due to absorption of water by the powders during aging. The intensit y of this peak is also observed to increase with aging time. Resonance intensities from a (CO 2 ) are observed to form and grow with aging time. This seems to agree with the trend observed for films pyrolyzed at 400 C and aged in 98% RH ( Figure 5 8 D ). However, no peaks corresponding to the (OH ) are observed in these powders. 5.3.2.2. Thermogravimetric a nalysis Powders pyrolyzed at 300 C seem to have considerable mass loss at ~100 C upon aging ( Figure 5 26 A ). This mass loss could be due to evaporation of water from the samples. This mass loss is also observed for powder pyrolyzed at 400 C. However, this mass loss is observed to be considerably low ( Figure 5 26 B and inset). For powders pyrolyzed at 300 C, changes are also o bserved in the weight loss trends with aging. A plateau region is observed for powders at Day 0 between ~ 300 C and 450 C. The temperature region of this plateau is observed to decrease for powders aged for 2 and 3 days to ~ 400 C. For powders aged for lon ger times, this plateau is observed to

PAGE 120

120 increase to ~ 475 C. No significant changes are observed in the powders pyrolyzed at 400 C. While thermogravimetric analysis (TGA) gives information about the inherent response of the gel state to aging, the behavior observed in thin films is affected both by the presence of the substrate and inherent behavior of the gel phase. From TGA measurements, the inherent crystallization behavior of the gel phase was observed to change considerably for powders pyrolyzed at 300 C. 5.4. Conclusions The properties of IMO based solution deposited PZT thin films are observed to be affected due to aging after pyrolysis. The electrical properties of the films pyrolyzed at lower temperature (300 C) showed higher variability due to aging tha n films pyrolyzed at higher temperature (400 C). The relative humidity during aging seems to affect the films differently. No significant differences were observed in the grain morphologies of films pyrolyzed at 300 C and aged in 20% and 400 C and 20%, 44% and 55% RH. Films pyrolyzed at 300 C and aged in 44% and 55% RH seem to form isolated rosette type grains. Films aged in 98% relative humidity seem to be the most affected for both pyrolysis conditions. Extensive rosette type grain structure is observed t o form upon aging. This change in grain morphology could be due to the change in the chemical nature of the film as observed in the FTIR spectra. Films aged in 20% RH for both pyrolysis conditions were observed to undergo minimal changes in the microstruct ure and texture and, in general, the grain morphology was not observed to change.

PAGE 121

121 This thesis aims to improve the understanding of the effect of processing conditions on the phase and texture evolution of solution deposited thin films. E quipment for film d eposition does not exist at the Advanced Photon Source, the samples had to be fabricated prior to the experiment at Sandia National Laboratories. A ging of the films is inevitable. However, since the microstructure of films 300 C or 400 C undergoes minimal aging when stored in 20% RH, films processed and stored in these conditions could be used to understand phase and texture evolution in these films during processing.

PAGE 122

122 Figure 5 1 Polarization vs electric field measurements for films pyrolyzed at 300 C and stored in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity conditions. Figure 5 2 Polarization electric field measurements for films pyrolyzed at 400 C and aged in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity.

PAGE 123

123 Figure 5 3 C vs F measurements for films pyrolyzed at 300 C and aged in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D) 98% relative humidity.

PAGE 124

124 Figure 5 4 C vs F measurements for films pyrolyzed at 400 C and aged in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D) 98% relative humidity.

PAGE 125

125 Figure 5 5 Variation of relative dielectric constant and tan for films pyrolyzed at 300 C with aging time. Measurements were performed at 100 Hz.

PAGE 126

126 Figure 5 6 Films pyrolyzed at 400 C: variatio n of relative dielectric constant and tan with aging time

PAGE 127

127 Figure 5 7 FTIR spectrum of films pyrolyzed at 300 C and aged in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity.

PAGE 128

128 Figure 5 8 FTIR spectrum of films pyrolyzed at 400 C a nd aged in different RH conditions. A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity

PAGE 129

129 Figure 5 9 Films pyrolyzed at 300 C and aged in 20% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humidity for 0 days ( A and B ), 9 days ( C and D ) and 30 days ( E and F ) are shown.

PAGE 130

130 Figure 5 10 Films pyrolyzed at 300 C and aged in 44% RH: SEM plan vie w and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humidity for 0 days ( A and B ), 9 days ( C and D ) and 30 days ( E and F ) are shown.

PAGE 131

131 Figure 5 11 Films pyrolyzed at 300 C and aged in 55% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humidity for 0 days ( A and B ), 9 days ( C and D ) an d 30 days ( E and F ) are shown.

PAGE 132

132 Figure 5 12 Films pyrolyzed at 300 C and aged in 98% RH: SEM plan view and cross section micrographs of crystallized films after specific a ging times. Films aged in 20% relative humidity for 0 days ( A and B ), 9 days ( C and D ) and 30 days ( E and F ) are shown.

PAGE 133

133 Figure 5 13 Films pyrolyzed at 400 C and aged in 20% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humidity for 0 days ( A and B ), 2 days ( C and D ), 8 days ( E and F ), and 16 days ( G and H ) are shown.

PAGE 134

134 Figure 5 14 Films pyrolyzed at 400 C and aged in 44% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humid ity for 0 days ( A and B ), 2 days ( C and D ), 8 days ( E and F ), and 16 days ( G and H ) are shown.

PAGE 135

135 Figure 5 15 Films pyrolyzed at 400 C and aged in 55% RH: SEM plan vie w and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humidity for 0 days ( A and B ), 2 days ( C and D ), 8 days ( E and F ), and 16 days ( G and H ) are shown.

PAGE 136

136 Figure 5 16 Films pyrolyzed at 400 C and aged in 98% RH: SEM plan view and cross section micrographs of crystallized films after specific aging times. Films aged in 20% relative humid ity for 0 days ( A and B ), 2 days ( C and D ), 8 days ( E and F ), and 16 days ( G and H ) are shown.

PAGE 137

137 Figure 5 17 XRD measurements of films pyrolyzed at 300 C and crystall ized after aging for spec ific times. Films were aged in A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity

PAGE 138

138 Figure 5 18 Lotgering f 111 factor values for films pyrolyzed at 300 C and crystallized after different aging times

PAGE 139

139 Figure 5 19 XRD measurements of films pyrolyzed at 400 C and crystallized after agi ng for specific times. Films were aged in A ) 20%, B ) 44%, C ) 55%, and D ) 98% relative humidity

PAGE 140

140 Figure 5 20 Lotgering f 111 factor values for films pyrolyzed at 400 C and cry stallized after different aging times Figure 5 21 Different modes of coordination between the metal ion and the acetate ions. Adapted from [144]

PAGE 141

141 Figure 5 22 Variation in the full width at half maximum (FWHM) of s (C=O) with aging time in films pyrolyzed at 300 C. Error bars correspond to 95% confidence interval.

PAGE 142

142 Figure 5 23 Variation in the FWHM of s (C=O) of films pyrolyzed at 400 C and aged in 98% relative humidity. Error bars indicate 95% confidence intervals. Figure 5 24 FTIR spectra of powders aged in aged for different times in 98% RH. Powders were pyrolyzed at A ) 300 C and B ) 400 C

PAGE 143

143 Figure 5 25 Variation of FWHM with aging time of the a (CO 2 ) resonance peak in solution derived powders pyrolyzed 300 C and 400 C. Figure 5 26 Thermogravimetry plots for powders The masses were normalized with respect to the final mass of the powders. Powders were pyrolyzed at A ) 300 C and B ) 400 C and aged for different times.

PAGE 144

144 Figure 5 27 Limited temperature t hermogravimetry plots for solution derived powders Powders were pyrolyzed at A ) 300 C and B ) 400 C

PAGE 145

145 CHAPTER 6 EFFECT OF HEATING RA TE ON PHASE AND TEX TURE EVOLUTION In this chapter, the effect of heating rate on phase and texture evolution in solution deposited PZT thin films is presented. The in situ high temperature synchrotron X ray based diffraction techniques described in CHAPTER 3 are used to measure phase and texture evolution in solution deposited PZT thin films during crystallization. 6.1. Experimental Methods PZT (52/48) + 10% Pb excess thin films were deposited using IMO based solutions. Details on the specifics of the solution preparation have been listed in Section 5.1 and are only briefly presented here. Requ ired amounts of zirconium butoxide and titanium isopropoxide were vortexed for 5 minutes before addition of acetic acid. The solution was further vortexed for 5 minutes to allow for chelation of the precursors before addition of methanol. Lead (IV) acetate was then added to the solution and solution was heated to 90 C to aid the dissolution of lead acetate. After dissolution of lead acetate, the solution was cooled and diluted with acetic acid and metha nol to obtain 0.35 M solutions. F ilms were deposited by spin coating the solution to the platinized silicon substrates and then pyrolyzed at 300 C after each deposition Multiple deposition step were performed to achieve a final thickness of ~300nm To measure the phase and texture evolution during crystalliza tion in these films, in situ high temperature diffraction experiments were performed at beamline 6 ID B Advanced Photon Source (APS) Argonne National Laboratory. In these experiments, the X ray beam was incident at an angle of ~0.2 to the sample. Diffra ction patterns of the sample were measured using a MAR SX 165 CCD detector. Each diffraction pattern was collected for an acquisition

PAGE 146

146 time of 1 s. The readout time for the detector was ~2 s. To crystallize the thin films, the samples were heated using an I R lamp (Osram XENOPHOT). The heating rate for crystallization was varied by controlling the voltage on the IR lamp. The voltage applied on the lamp was increased in constant steps during the acquisition of the diffraction pattern and kept constant during d ata readout. Different heating rates were achieved by changing the number of steps required to reach the final voltage. It was assumed that a temperature of ~ 700 C needs to be reached for the formation of the perovskite phase. Based on this assumption, th e average heating rate s for the different voltage rates w ere calculated. The heating rates were calculated to be 20 C/s, 3 C/s, 1 C/s and 0.2 C/s. SEM characterization and TEM analysis of these samples was performed at Sandia National Laboratories. 6.2. Results To check for the repeatability of results, each condition was repeated The phase evolution and azimuthal time plots for the different conditions and the repeat measurements are provided in APPENDIX C 6.2.1. Phase Evolution Phase evolution during crystallization of solution deposited PZT thin films at different heating rates is shown in Figure 6 1 Diffraction peaks corresponding to Pt x Pb(111), fluorite(111) and perovskite (100) were peak fit with P earson VII type funct ion to extract the integrated intensity of these peaks. Peak fitting was performed using th e curve fitting toolbox in MATLAB. The variation in the integrated intensity of these phases during crystallization for different heating rates is shown in Figure 6 2 A Pt x Pb type phase is observed in films crystallized at 20 C/s, 3 C/s and 1 C/s. The Pt x Pb

PAGE 147

147 and fluorite type phases are observed to form before the perovskite phase. The general sequence of phase evolution is observed to be similar to the in situ crystallization experiments performed using a laboratory diffractometer ( Figure 6 1 ). N o Pt x Pb phase is observed in films crystallized at 0.2 C/s (12 C/min) at APS while a Pt x Pb phase is observed in films in situ crystallized at 5 C/min in the laboratory diffractometer. This inconsistency in phase evolution during crystallization could be due to the following factors: The films used in the in situ laboratory based ex periments had 20% nominally excess lead precursor in them while the films used for experiments in the synchrotron based diffraction experiments had 10% excess lead precursor. The lower amount of lead precursor in the starting solutions could lead to decrea se in the formation of the Pt x Pb phase [128] This inconsistency could also be due to the difference in the types of heating used in these in situ measurements. An IR lamp is used in the APS measurements while the films are heated in a furnace in the laboratory based measurement. The fluorite phase is always observed to precede the perovskite type phase during crystallization. 6.2.2. Texture Evolution The 110 AT plots for the films crystallized using the in situ diffraction setup are shown in Figure 6 3 Films crystallized at 20 C/s are observed to have a random texture. 111 and 100 texture components are observed to be present in films crystallized at slower heating rates (3 C/s, 1 C/s and 0.2 C/s). Intensity of the 1 00 texture component is observed to increase with decrease in heating rate. To quantify the volume fraction of the different texture components the in tensities corresponding to the 111 and the 100 texture components were peak fit using the curve fitting tool box in

PAGE 148

148 MATLAB. The background was estimate d using a linear polynomial. The background intensity corresponds to the diffracted intensity from the randomly oriented grains. Peak fitting enabled the deconvolution of the diffraction intensities from the 111 and the 100 texture component s. To estimate the volume of the 111 and 100 texture component integration of the pole intensities needs to be performed over the pole figure angle The details of the methodology us ed for the calculation of the volume fraction of the texture components is presented in detail in Section 3.8 The variation of the 111 and 100 texture fractions for films crystallized in situ at the APS are shown in Figure 6 4 Films crystallized at 3 C/s are observed to be (111) oriented. The degree of (111) texture is observed to decrease with decrease in heating rate. Corresponding to the decrease in the (111) t exture, (100) texture fraction is observed to increase with decrease in heating rate. 6.2.3. SEM and STEM EDS The grain morphology of the in situ crystallized samples was characterized using SEM. In films crystallized at 20 C/s, r osette type grains are observed to form in this film ( Figure 6 5 A ). A bimodal grain structure is observed in the cross section ( Figure 6 5 B ) along with porosity in the middle of the film. This trapped porosity could form due to bimodal grain growth [61] A columnar grain structure is observed for films crystallized a t lower heating rates. The grain sizes observed in the plan view are observed to increase with decrease in heating rate ( Figure 6 5 Figure 6 8 ) Chemical mapping of the cation segregation across the thickness of the thin films was p erformed using energy dispersive spectroscopy (EDS) in a scanning transmission electron microscope (STEM). The obtained EDS spectra were processed using

PAGE 149

149 multivariate statistical analysis (MSA) to achieve nanometer scale resolution of cation segregation. De tails on this STEM EDS technique can be found elsewhere [139, 142, 147 149] A strong Ti segregation is observed at the film Pt interface in films crystallized at 20 C/s ( Figure 6 9 ) The Zr/Ti ratio in this film is observed to increase towards the center of the thin film. Ti segregation is also observed for films crystallized at 3 C/s ( Figure 6 10 ) The Zr/Ti ratio in these films is observed to increase towards the surface. For films crystallized at 1 C/s and 0.2 C/s, no Ti segregation at the film Pt electrode interface is observed ( Figure 6 11 and Figure 6 12 respectively) Variation in the Zr/Ti ratio is o bserved across the thickness of the film. However, this variation in the Zr/Ti ratio is observed to be less than the variation observed in films crystallized at 20 C/s and 3 C/s. 6.3. Discussion Crystallization of solution derived PZT thin films is nucleation l imited [110] Both heterogeneous and homogenous nucleations are observed to occur in CSD PZT thin films. H eterogeneous nucleation is the rmodynamically more favorable However, homogenous nucleation could be activated in the thin film if high temperatures are reached prior to complete crystallization of the film. Hence kinetic factors such as s mall energy difference between heterogeneous an d homogeneous nucleation free energies and fast heating rates have been suggested to cause activation of the homogenous nucleation during crystallization [38] Heterogeneous and homogeneous nucleation events can be distinguished from the final grain morphology of the films. Heterogeneous nucleation of PZT grains at the film substrate interface has been shown to lead to the

PAGE 150

150 formation of columnar grains while hom ogenous nucleation leads to the formation of rosette type grain morphology [38] In thin films in situ crystallized at 20 C/s r osette type grains are observed. These grains could be due to crystallization through homogenous nucleation. Consis tent with the observation of rosette type grains, no preferred orientation is observed in these films. The formation of porosity in the middle of the thin films can also be explained based on the nucleation mechanism during crystallization. Transformation from the fluorite to the perovskite phase involves a decrease in volume. Since homogenously nucleated grains lack directionality, the porosity formed due to this decrease in volume could be trapped in the bulk of the film [61] The variation in the Zr/Ti ratio across the film thickness observed in the films in this study is similar to that observed in previous studies [141, 150, 151] However, in this study, the Zr/Ti segregation is observed to decrease with decrease in heating rate. Hence, the Zr/Ti segregation observed in these thin films may not be thermodynamically driven. The lack of Zr/Ti segregation is films crystallized at slow heating rates in this study is consistent with the results obtained by Shelton et al. [26] on films processed through the IMO route similar to films used in this study. The films used in this study were prepared through the IMO route while earli er studies used sol gel based routes. The thermodynamic behavior of the thin films could be dependent on the route used for the preparation of the solution. To better understand the formation of Zr/Ti segregation across the thickness of the thin film, the change in the lattice parameter of perovskite PZT during crystallization are plotted ( Figure 6 13 ). Lattice parameter of perovskite PZT was obtained th rough the

PAGE 151

151 peak fitting of the perovskite (100) peak with a Gaussian profile shape function in MATLAB. No significant differences are observed in the change of perovskite PZT lattice parameter between films crystallized at different heating rates. It is obs erved that the lattice parameter of perovskite PZT decreases after formation in films crystallized at all heating rates. This decrease in lattice parameter could be due to volatilization of any excess lead present in the thin film [152] However, no correlation could be found between the decrease in the perovskite PZT lattice parameter during crystallization and the Zr/Ti segregation obs erved in the thin film. From TEM EDS maps ( Figure 6 9 Figure 6 12 ) a sharp Ti segregation layer is observed in films crystallized at 20 C/s and 3 C/s. This Ti segregation could be due to the diffusion of Ti from the adhesion layer to the surface of the Pt electrode. Diffusion of titanium from the adhesion layer was observed in se veral studies by previous investigators [63 65, 68] Recently, Sh elton et al. [26] characterized Zr/Ti segregation in IMO based solution derived PZT thin films with different adhesion layers (Ti, TiO x and ZnO). Zr/Ti segregation was observed in PZT films deposited on Pt/Ti bilayers, while less Zr/Ti segregation was observed in films deposited on Pt/TiO x bilayers and P t/ZnO bilayers. The titanium segregation observed in PZT films crystallized on Pt/Ti layers was suggested to be due to the diffusion of the titanium from the adhesion layer. Hence, the titanium segregation observed towards the film substrate interface co uld be due to the migration of the titanium for the adhesion layer. During crystallization at slow heating rates, oxygen from the atmosphere could diffuse through the film and oxidize titanium. This oxidation of titanium could decrease the mobility of tita nium and prevent it from diffusion to the surface of the platinum electrode [64]

PAGE 152

152 Th e trend in the variation of (111) and (100) texture with heating rate is consistent with observations of Chen et al. [106] However, Chen et al. o bserve a (111) texture for t he fastest heating rate used in their study The fastest heating rate during crystallization in the study by Chen et al. was achieved by directly inserting the films into a preheated furnace at 700 C, while in this study the films were crystallized using a n IR lamp similar to an RTA. The difference in the heating method could lead to this difference in crystallization behavior [153] However, the general trend in the variation of (111) and (100) texture s with change in heating rate is observed to be similar to that reported by Chen et al. [106] The metastable intermetallic Pt x Pb phase has been suggested preferential ly nucleat e the perovskite 1 11 texture component crystallization of PZT thin films The perovskite phase is suggested to be seeded by the Pt x Pb phase due to the smaller lattice mismatch between the Pt x Pb phase and the perovskite phase [78, 83, 154] However, from the phase evolution plots, no overlap is observed between the Pt x Pb ph ase and the perovskite phase. Hence the PtxPb phase might not directly seed the 111 texture component in these thin films. A more detailed investigation into the formation of the Pt x Pb phase is presented in CHAPTER 7 The fluorite phase is observed to precede the perovskite phase for all heating rates. It has been suggested that the formation of fluorite phase could be completely avoided by using the high heating rates such as achievable by r apid thermal annealing (RTA) [88, 122] fluorite phase might not be possible even for films crystallized at fast heating rates [79, 80, 155] T ransformation of the amorphous phase to the fluorite type structure is

PAGE 153

153 suggested to be preferred due to the lower atomic rearrangement required for the transformation [61, 76, 94] These results by previous investigators agree well with the phase evolution observed in films crystallized at different heating rates. Other mechanisms suggested to control texture in PZT thin films include fluorite crystallinity [72] formation of Pt 3 Ti intermetallic at the interface [64] formation of rutile seeds due to di ffusion of the Ti from the adhesion layer to the surface of the Pt electrode [69] and oxygen stoichiometry during processing [82, 83, 121] T o identify the differences in the fluorite phase present in the thin film at different heating rates, the fluorite (111) peak before the formation of the perovskite phase was peak fit to extract the FWHM of the peak. The variation of the FWHM of the fluori te (111) peak with heating rate is shown in Figure 6 14 No consistent changes with heating rate are observed in the width of the fluorite phase. To verify the presenc e of preferred orientation in the fluorite phase, the measured f luorite (111) diffraction intensities over the measured azimuthal angle range was extracted. No preferred orientation could be observed from the fluorite (111) diffraction intensities for films crystallized at different heating rates ( Figure 6 15 ). Hence, no evidence for seeding of the perovskite ori entation by the fluorite phase could be found. Diffus ion of the titanium from the adhesion layer has been suggested to seed 111 texture component during crystallization of solution deposited PZT thin films [107] The titanium diffusing from the adhesion layer was suggested control the texture of th e thin film through the formation of rutile seeds at the film substrate interface. The rutile seeds could form due to the oxidation of titanium [63, 68] This me chanism agrees well with the compositional maps of the films in situ crystallized in this study. Intense Ti

PAGE 154

154 segregation, possibly due to diffusion of Ti from the adhesion layer, is observed for films crystallized at fast heating rates. However, fast heatin g rates are also concomitant with the presence of reducing conditions at the film substrate interface. Reducing conditions have been suggested to control the orientation of these films Hence either rutile seed formation and/or reducing conditions could lead to the stabilization of the (111) orientation in these thin films. Platinized silicon substrates with TiO x and ZnO adhesion layers allow for limited diffusion of titanium to the film substrate interface. Hence, i n situ measurements were performed on CSD PZT thin films on platinized silicon substrates with different adhesion layers to better understand the effect of adhesion layer on phase and texture evolution in these thin films The results of these measurements are p resented in CHAPTER 8 6.4. Conclusion In situ diffraction measurements were performed during crystallization of PZT thin films. Films crystallized at 20 C/s were obser ved to be randomly texture, possibly due to homogenous nucleation. Intense titanium segregation is observed in films crystallized at fast heating rates (20 C/s and 3 C/s). This titanium segregation could be due to diffusion of titanium from the adhesion la yer. No evidence was found for seeding of (111) orientation in PZT thin films by the Pt x Pb intermetallic or the fluorite phase. 6.5. Acknowledgements I would like to thank Bonnie McKenzie at Sandia National Laboratories for performing SEM. I am also grateful t o Dr. Ping Lu, SNL for TEM analysis and composition mapping. Help from Dr. Bryan Gauntt in extraction of the PCA data for data representation is also appreciated.

PAGE 155

155 Figure 6 1 Phase evolution during crystallization of solution deposited PZT thin films. F ilms were crystallized by heat ing at A ) 20 C/s, B ) 3 C/s, C ) 1 C/s, and D ) 0.2 C/s

PAGE 156

156 Figure 6 2 Variation of integrated intensity of the crystalline phases formed in solution deposited PZT thin films. A ) 20 C/s, B ) 3 C/s, C ) 1 C/s, and D ) 0.2 C/s.

PAGE 157

157 Figure 6 3 (110) AT plots for films crystallized at different heating rates. indicate intensities from (111) type texture and arrows indicate (100) type texture. A ) 20 C/s, B ) 3 C/s, C ) 1 C/s, and D ) 0.2 C/s.

PAGE 158

158 Figure 6 4 Variation of texture in films in situ crystallized at different heating rates Figure 6 5 SEM micrographs of films crystallized at 20 C/s. A) Plan and B) cross section views.

PAGE 159

159 Figure 6 6 SEM micrographs of films crystallized at 3 C/s. A) Plan and B) cross section views. Figure 6 7 SEM micrographs of films crystallized at 1 C/s. A) Plan and B) cross section views. Figure 6 8 SEM micrographs of films crystallized at 0.2 C/s. A) Plan and B) cross section views.

PAGE 160

160 Figure 6 9 STEM EDS compositional maps of films crystallized at 20 C/s Figure 6 10 STEM EDS compositional maps of films crystallized at 3 C/s

PAGE 161

161 Figure 6 11 STEM EDS compositional maps of film crystallized at 1 C/s Figure 6 12 STEM EDS compositional maps of film crystalli zed at 0.2 C/s

PAGE 162

162 Figure 6 13 Change in the lattice parameter of perovskite PZT after formation during crystallization F ilms were crystallized at A) 20 C/s, B) 3 C/s, C) 1 C/s and D) 0.2 C/s.

PAGE 163

163 Figure 6 14 Variation of the FWHM of the fluorite (111) peak with heating rate Figure 6 15 Variation of pole intensity with azimuthal angle ( ) for the fluorite phase formed during crystallization at different heating rates

PAGE 164

164 CHAPTER 7 FORMATION OF PT X PB The Pt x Pb metastable intermetallic phas e formed during crystallization of solution deposited PZT thin films has been suggested to seed (111) orientation in PZT thin films [78, 81, 83] However, a consensus on the structure and stoichiometry of the Pt x Pb phase formed during crystallization is lacking. This chapter focuses on the improving the understanding of the formation of Pt x Pb phase during crystallization using the synchrotron X ray diffraction based in situ measurement technique developed in this work 7.1. Introduction The formation of Pt x Pb in solution deposited PZT thin films was suggested to be due to the highly reducing conditions present in the thin films during crystallization [81] R educing conditions in the thin film could form due to the oxidization of the remnant organic moieties. The reducing condition in the film, are suggested to lead to reduction of the lead precursor in the solution d eposited thin film from a Pb +2 to a metallic Pb 0 [81] Based on thermodynamic calculations [156] at ~400 C a p O 2 of ~ 10 2 0 atm needs to be reached for the reduction of lead(II) oxide to metallic lead ( Figure 7 1 ) [18] Ve ry fast heating rates and low pyrolysis temperatures are suggested to enable the formation of these highly reducing conditions [157] Brooks et al. [82] observed that the Pt x Pb intermetallic phase could be reformed in an uncrystallized PZT thin film by using a reducing atmos phere. Different crystal structures and stoichiometry were suggested for the Pt x Pb phase in different studies. Chen et al. [81] based on SEM EDS suggested that the Pt/Pb molar r atio in Pt x Pb is ~ 5 7 O ther researchers reported a Pt/Pb molar ratio of 3 4 [78, 82,

PAGE 165

165 116] In a recent study by Dippel et al. [158] two different Pt x Pb phases a Pt 3 Pb (cubic, space group Pm 3m) and a PtPb type (hexagonal, space group P6 3 /mmc) phase, were obs erved to form The relative stability of the Pt 3 Pb and the PtPb phase was found to be dependent on the specific Zr/Ti stoichiometry of the thin films. In general, the Pt x Pb type phases were reported to form at higher temperature s in Zr rich compositions. The PtPb phase preceded the formation of the Pt 3 Pb phase In the study by Dippel et al. atypical processing conditions such as hyperstoichiometry of lead precursor and very thick films ( t~ 1 m) were used to create favorable cond itions for the reduction of the lead precursor. Hence, the formation sequence of the Pt x Pb type phases observed in this work could have limited applicability to conventionally processed thin films. In situ high temperature diffraction techniques could help to improve the understanding of the formation and stability of Pt x Pb phases in conventionally processed solution deposited thin films. Better knowledge of the formation and stability of the Pt x Pb phase could improve understanding of texture evolution in s olution deposited PZT thin films. 7.2. Experimental PZT (52/48) s olutions with 10% and 20% nominally excess lead precursor were prepared through the IMO route with a solution molarity of 0.35 M These solutions were spin coated onto platinized silicon substrate s (Pt(170 nm)/Ti(40 nm)/SiO 2 ( 400 nm ) /Si acquired from SQI) at 3000 rpm These films were pyrolyzed at 300 C on hot plate before subsequent depositions. Two more deposition and pyrolysis steps were performed to achieve a film thickness of ~300 nm. The final crystallization step was

PAGE 166

166 performed using the in situ laboratory and synchrotron setups while diffraction intensities were continuously during crystallization. During in situ laboratory measurements, the films were crystallized by heating the films in a furnace. Films made from PZT (52/48) + 20% Pb excess solutions were used in these measurements. The furnace was ramped to 650 C at 5 C/s and held for 20 minutes upon reaching the final temperature. Diffraction pattern s were continuously acquired during heating with an acquisition time of 60s. To investigate the texture relationships between the Pt x Pb formed and the Pt electrode the film was quenched after the formation of the Pt x Pb phase(T ~ 400 C) by turning the furn ace off and letting the sample furnace cool. (111) p ole figure measurements of Pt and the Pt x Pb phase circle diffractometer. I n situ measurements using synchrotron x ray diffraction were performed during crystallization of solution deposited PZT + 10% Pb excess thin films Films were crystallized at different heating rates while diffraction images were simultaneously collected using a GE amorphous silicon detector. Data reduction methodologies were then applied to extract and represent evolution of phases and texture during crystallization. For extraction of phase information, the methodologies applied were similar to those described in CHAPTER 3 except that intensities were integrated from = 5 175 Integration over a wide range was performed to account for the presence of any textured Pt x Pb phases. 7.3. Results and Discussion Phase evolution measured using the laboratory setup is shown in Figure 7 2 A Pt x Pb phase is observed first followed by a fluorite phase and finally a perovskite phas e.

PAGE 167

167 The trend in phase evolution is observed to be similar to the PLZT films investigated in CHAPTER 4 and hence further discussion on phase evolution is not presen ted here. Diffraction intensities from the Pt x Pb phase are observed to be present at ~ 38.5 This peak position corresponds to the Pt 3 Pb phase observed by earlier investigators [116, 159] No intensities corresponding to the formation of a PtPb phase were observed. During in situ crystallization a Pt x Pb is observed to form in the thin films heated at 65 C/s, 32 C/s and 11 C/s ( Figure 7 3 Figure 7 6 ) The position of the diffrac tion intensities from the Pt x Pb phase formed in these films corresponds well with Pt 3 Pb type phase reported by Huang et al. [78] No diffraction intensities corresponding to the formation of a PtPb ( Table 7 1 ) [158] or other Pt x Pb type phases were observed. Limited 2 time phase evolution plots for the different heating rates are shown in Figure 7 3 B Figure 7 4 B Figure 7 5 B and Figure 7 6 B The structure of the Pt 3 Pb seems to correspond well to the Pt x Pb phase (JCPDF# 06 0574) however the lattice parameter of the Pt x Pb phase formed in these films seems to be larger (a = 4.08 angstroms). The observed higher lattice parameter is consistent with reports by Huang et al. and Kaewchinda et al. [78, 116] The powder diffraction spectrum expected for this structure with a higher lattice parameter was calculated using CrystalDiffract and is listed in Table 7 2 Hence, the Pt x Pb phase formed during crystallization of these films corresponds to a Pt 3 Pb phase. From the limited phase evolution plots, the maximum intensity of the Pt 3 Pb formed is observed to decrease with decrease in heat ing rate during crystallization To quantify the variation of the Pt 3 Pb phase with heating rate, the intensities corresponding to the Pt 3 Pb phase were peak fit with a Pearson VII type function using the curve fitting

PAGE 168

168 toolbox in Matlab. To account for any changes in the measured intensity due to differences in the X ray beam attenuation and acquisition time used during in situ diffraction measurements at different heating rates, the integrated intensity of the Pt 3 Pb phase was normalized using the integrated intensity of the Pt peak observed at the start of the experiment. The variation of the intensity of the Pt 3 Pb phase with heating rate is shown in Figure 7 7 The normalized Pt 3 Pb integrated intensity is observed to decrease with decrease in heating rate. This trend in the decrease in the amount of Pt 3 Pb formed with decrease in heating rate is con sistent with trend suggested by Chen et al. [81] and Huang et al [159] The 2D detector used in the A PS experiments captures both phase and texture information simultaneously. The diffraction images measured by the 2D detector at times corresponding to the maximum intensity of the Pt 3 Pb are shown in Figure 7 8 Figure 7 11 Using this representation, both the phases present in the thin film and their texture can be observed. The texture of the Pt 3 Pb phase is observed to template from Pt bottom layer. The texture relationship between Pt 3 Pb phase and platinum is observed to be similar at all heating rates. The similarity in the texture of the Pt 3 Pb phase and the Pt layer is consiste nt with the epitaxial relationship observed between the Pt 3 Pb phase and the Pt layer by Kaewchinda et al. [116] The texture relationship between the Pt 3 Pb phase and the platinum bottom electrode were further verified through pole figure measurements on samples with quenched in Pt 3 Pb phase. Solution deposited PZT + 20% Pb excess films were heated in the laboratory X ray diffraction based in situ measurement setup at 5 C/s. Upon the formation of the Pt 3 Pb phase at 400 C, the samples were q uenched by turning the

PAGE 169

169 furnace off and letting the samples furnace cool. Upon cooling, pole figures of Pt(111) and Pt 3 circle MRD diffractometer and are shown in Figure 7 12 and Figure 7 13 respectively. Consistent with the results obtained on films in situ crystallized using the syn chrotron X ray diffraction based techniques, the fiber axis of the Pt 3 Pb(111) diffraction intensities is aligned along the fiber axis of the Pt 111 texture component. The Pt 3 Pb phase has a cubic crystal structure with a lattice parameter close to platinum ( Figure 7 14 and Figure 7 15 ). In addition the Pb Pt coordination in Pt 3 Pb is similar to the Pt Pt coordination in platinum. Due to the low atomic rearrangement and similar lattice parameters, the Pt 3 Pb could form a coherent/semi coherent interface with the platinum layer and thus mimic the platinum texture. The similarity in the atomic arrangements in the Pt3Pb and Pt are shown in Figure 7 16 The absence of formation of a PtPb phase could be due to the lower amount of metallic lead formed in these thin films or due to greater amount of atomic rearrangement required for the formation of PtPb. Hence, from the synchrotron and labor atory based measurements, it can be concluded that the Pt 3 Pb inherits its texture from the Pt bottom electrode due to a cube on cube orietational relationship. The stability of the Pt 3 Pb phase during storage was also investigated by measuring diffraction p atterns of the quenched Pt 3 Pb phase at sequentially increasing aging times ( Figure 7 17 ). No significant variation in the intensity of the Pt 3 Pb phase was observed until 28 days after the Pt 3 Pb phase was quenched in the thin film. Measurements beyond 28 days of aging were not performed. Hence, the Pt 3 Pb phase is stable against decomposition for up to 28 days after formation.

PAGE 170

170 Based on the experimental ob servation of the formation of Pt 3 Pb, a mechanism for the formation of Pt 3 Pb phase in these films is suggested ( Figure 7 18 ) The metallic lead formed due to the reduc tion of lead oxide in the amorphous film could diffuse into the platinum layer. Platinum alloys with lead and forms a substitutional alloy until Pt:Pb = 3:1 ( Figure 7 14 ). Upon further formation and diffusion of metallic lead into platinum a Pt 3 Pb phase could be formed. With further heating during crystallization, the Pt 3 Pb phase decomposes and the Pb from the Pt 3 Pb phase diffuses into the film and towards the substrate Diffusion of Pb into the substrate is kinetically limited and hence most of the Pb could be absorbed by the PZT thin film. Some amount of Pb could reach the adhesion layer and beyond into the thermal oxide layer. 7.4. Conclusion The metastable Pt 3 Pb phase is observed to form in PZT thin films crystallized at fast heating rates. The amount of the Pt 3 Pb phase formed in the thin films is observed to decrease with decrease in heating rate. The Pt 3 Pb phase is observed to have a cube on cube orientational relationship with the platinum bottom electrode.

PAGE 171

171 Table 7 1 Calculated p owder diffraction reflection positions for PtPb at = 0.5462 angstroms. The space group of the material is P6 3 /mmc, with a = 3.358, b = 3.358 and c = 4.0580 angstroms. h k l 2 d 1 0 0 10.7771 2.90811 1 0 1 13.2689 2.36380 0 0 2 15.4708 2.02900 1 1 0 18.7222 1.67900 0 1 2 18.8924 1.66402 0 2 0 21.6511 1.45406 0 2 1 23.0169 1.36884 1 1 2 24.3766 1.29355 Table 7 2 Plane indices and 2 values expected in the powder pattern for Pt 3 Pb at = 0.5462 angstroms The 2 values were calculated using Crystal Diffract The crystal structure was modeled based on JCPDF#06 0574 with a Pm 3m space group and lattice parameter of 4.05 angstroms. h k l 2 d 0 0 1 7.7328 4.05000 0 1 1 10.9442 2.86378 1 1 1 13.4141 2.33827 0 0 2 15.5011 2.02500 0 1 2 17.3441 1.81122 1 1 2 19.0142 1.65341 0 2 2 21.9898 1.43189

PAGE 172

172 Figure 7 1 Thermodynamic stability of the constituent metals in PZT towards oxidation and reduction Figure 7 2 Phase evolution showing the formation of Pt x Pb metastable phase and the final perovskite phase (Pe) during in situ crystallization in the Inel diffractometer.

PAGE 173

173 Figure 7 3 Results on in situ crystallization experiment performed at 64 C/s A) Ph ase evolution and B) limited 2 plot for films. indicates Pt diffraction peaks and denotes Pt 3 Pb diffraction peaks Figure 7 4 Results on in situ crystallization experim ent performed at 68 C/s. A) Phase evolution and B) limited 2 plot films. indicates Pt diffraction peaks and denotes Pt 3 Pb diffraction peaks

PAGE 174

174 Figure 7 5 Results on in situ crystallization experiment performed at 32 C/s A) Phase evolution and B) limited 2 plot for films. indicates Pt diffraction peaks and denotes Pt 3 Pb diffraction peaks Figure 7 6 Results on in situ crystallization experiment performed at 11 C/s A) Phase evolution and B) limited 2 plot for films. indicates Pt diffraction peaks and denotes Pt 3 Pb diffraction peaks

PAGE 175

175 Figure 7 7 Variation of maximum normalized Pt x Pb intensity with decrease in heating rate. Error bars indicate 95% confidence intervals. Figure 7 8 2 plot for film crystallized at 64 C/s. indicates Pt 3 Pb phase and denotes Pt.

PAGE 176

176 Figure 7 9 2 plot for films crys tallized at 68 C/s. indicates Pt 3 Pb phase and denotes Pt. Figure 7 10 2 plot for films crystallized at 32 C/s. indicates Pt 3 Pb phase and denotes Pt.

PAGE 177

177 Figure 7 11 .2 plot for films crystallized at 11 C/s. indicates Pt 3 Pb phase and denotes Pt. Figure 7 12 (111) pole figure of Pt substrate in the quenched sample

PAGE 178

178 Figure 7 13 (111) pole figure of Pt x Pb phase in the quenched sample Figure 7 14 Pt Pb binary phase diagram [160]

PAGE 179

179 Figure 7 15 Crystal structure of Pt 3 Pb phase (JCPDF#06 0574) Figure 7 16 Stacking of Pt(111) and Pt 3 Pb(111) planes showing similarities in atomic arrangements.

PAGE 180

180 Figure 7 17 Diffraction patterns of the quenched in Pt 3 Pb phase in solution deposited PZT thin film measured at sequentially increasing aging times. Figure 7 18 Proposed sequence for the formation and disappearance of the Pt 3 Pb phase.

PAGE 181

181 CHAPTER 8 EFFECT OF ADHESION L AYER ON PHASE AND TEXTURE EVOLUTION The effect of the adhesion layer in the substrates on the phase and texture evolution in PZT thin films is not well understood. To better understand the effect of the adhesion layer on crystallization of PZT thin films solution deposited on platinized silicon substrates films were deposited on substrates with various adhesion layers. Substrates with titanium, titanium oxide (TiO x ) and zinc oxide (ZnO) adhesion layers were investgated Phase and texture evolution during crystallization of solution deposited PZT thin films on these substrates was measured in situ at beamline 6 ID B, Advanced Photon Source, Argonne National Laboratories. 8.1. Experimental Methods 8.1.1. Preparation o f Substrates Th e Pt/Ti substrates used in this study were purchased from Silicon Quest International, Inc. In these substrates, 40nm of titanium was deposited on silicon substrates with 400 nm of thermal oxide (SiO 2 ). Prior to deposition of titanium the chamber was pumpe d to 2.9*10 7 Torr and the sputtering was performed in an argon environment. The platinum layer (170 nm) was deposited after deposition of the titanium layer without breaking the vacuum The Pt/TiO x and the Pt/ZnO substrates used were made in house at San dia National Laboratories. Details of the preparation of these substrates are provided in CHAPTER 2 and the process is only briefly described here. rs with 400 nm of thermal oxide were used as the starting substrates for preparation of these substrates. For Pt/TiO x substrates, 40 nm of t itanium was sputtered onto the starting substrates using RF magnetron sputtering. After deposition, the titanium layer was

PAGE 182

182 oxidized in 15 mTorr of oxygen at 450 C for 30 minutes. T he titanium layer might not completely oxidize in these conditions and hence this layer is designated at TiO x For the preparation of Pt/ZnO substrates, 40 nm of ZnO was RF spu tter deposited onto silicon substrates. 100 nm of platinum was rf sputter deposited after deposition of the adhesion layer. After deposition of the Pt layer the Pt/ZnO substrates were directly inserted into a furnace preheated to 700 C and annealed for 10 minutes under ambient atmospheric conditions. Only substrates with ZnO adhesion layers were additionally processed through this annealing heat treatment prior to deposition of PZT thin films. The fourth set of substrate used in this investigation were obt ained from the army research lab and had a TiO x adhesion layer [16 1] During the fabrication of these substrates, t oxide. After sputter deposition of titanium, the substrates were annealed (650 C 800 C) in a tube furnace to oxidize the titanium to titanium dioxide. 100nm of platinum was sputter deposited after the oxidation of titanium to titanium oxide These substrates are designated as ARL Pt/TiO x for the reminder of the chapter. 8.1.2. APS Setup PZT (52/48) solutions with 10% nominal excess Pb prec ursor prepared using the IMO solution preparation route and spin coated onto the platinized silicon substrates Films were pyrolyzed at 300 C excep t film deposited on ARL Pt/TiOx substrates these were pyrolyzed at 350 C Three sets of deposition and pyrolysis steps were performed to achieve films with t~250 nm on each substrate. After deposition and pyrolysis the films were stored in desiccators maintained at ~20% relative humidity (RH) RH was maintained through the eq uilibrium aqueous tension of the LiCl salt in contact with its

PAGE 183

183 supersaturated solution at the bottom of the desiccator [133] The films were crystallized using the in situ setup at beamline 6 ID B Advanced Photon Source Argonne Nati onal Laboratory For these measurements a GE amorphous silicon detector was used. T he GE amorphous silicon detector has a higher frame rate capability (8 frames/s) and allows for in situ investigation at faster heating rates [162, 163] The heating rate used during in situ crystallization was varied by controlling the rate of change of voltage on the IR lamp. The temperatures at different times were estimated using ceria temperature calibration ( chapter 3.9 ). These average heating rates at the highest heating rate observed in these samples are listed in Tabl e 8 1 To ensure repeatability in measu rements two sets of samples were in situ crystallized for each of the heating rate investigated in this study. The results of the phase and texture evolution were observed to be consistent with the duplicate set of samples and the phase evolution and azim uthal time plots for both sample sets are shown in APPENDIX D 8.2. Results The texture of the platinum layer was observed to vary with the adhesion layer used. To quan tify the variation of the platinum texture, the texture intensities of the (111) and (222) Pt peaks were peak fit using a Gaussian function. Since the texture intensities present at = 90 could be artificially broadened due the diffraction geometry, the FWHM of the diffraction peaks present at = 19.48 and 160.52 were used to evaluate the sharpness of the Pt texture. Additional (111) and (222) diffraction intensities corresponding to the 111 tex ture component are present at the = 19.48 and 160.52

PAGE 184

184 due to plane multiplicities of the (111) and (222) plane s The obtained FWHM of platinum on substrates with different adhesion layers are shown in Figure 8 1 D ue to the higher frame rate achievable by the GE detector, phase and texture evolution during faster heating rates could be measured. The fastest heating rate achieved using this setup was ~68 C/s upon ra mping the voltage applied on the IR lamp at 1V/s. Ramping the voltage on the IR lamp at faster rates did not lead to any significant increase in the heating rate. Hence, heating rate of the film could be limited by its IR absorptivity The phase evolution plots for films deposited on different substrates and in situ crystallized at various heating rates are presented in Figure 8 3 and Figure 8 4 The key to the interpretation of the se phase evolution plots is provided in Figure 8 2 During fast heating rates, a Pt 3 Pb metastable intermetallic is observed to form first and then a pyrochlore structured phase is formed The perovskite phase is observed to form after the for mation of a pyrochlore structured phase. The maximum intensity of the Pt 3 Pb phase is observed to decrease with decrease in heating rate and is not observed for films crystallized at 1 C/s and 0.5 C/s. The sequence of phase formed during crystallization of the films is observed to be consistent with the results obtained on the laboratory diffractometer ( CHAPTER 4 ) and previous APS experiments ( CHAPTER 6 ). Phase evolution is observed to be also consistent between the films deposited on various substrates crystallized at similar heating rates. The additional peak observed in the Pt/ZnO substrate at ~12.5 c orresponds to the (0 002 ) ZnO diffraction peak Only the (0002) diffraction peak is observed from the ZnO adhesion layer, suggesting that this layer is textured. The Pt 3 Pb

PAGE 185

185 intermetallic is observed to form in films crystalli zed at 64 C/s (inst), 68 C/s, 32 C/s and 11 C/s. No Pt 3 Pb phase is observed in films crystallized at 1 C/s and 0.5 C/s. The 110 AT plots showing the texture evolution in these thin films during crystallization are shown in Figure 8 5 and Figure 8 6 A 111 texture component is observed in f ilms crystallized at fast heating rates (6 4 C/s, 68 C/s, 32 C/s, and 11 C/s) This trend is consistent between films deposited on different substrates. For films crystallized at 1 C/s and 0.5 C/s the texture of the PZT thin films is observed to be different for films deposited on different substrates. A mixed 111 and 100 texture components were observed for f ilms deposited on Pt/Ti substrates. However, f or films deposited on Pt/TiO x 100 and 111 te xture components are observed for 1 C/s heating rate and 111 texture component is observed for films crystallized at 0.5 C/s. For films deposited on substra tes with ZnO adhesion layer, a 111 texture component is observed for films crystallized at 1 C/s and films crystallized at 0.5 C/s are observed to be randomly oriented 110 texture components are also observed to be present in these thin films. However, the se texture intensities are only observed in films crystallized at fast heating rates To quantify the formation and growth of texture components in the thin film during crystallization peak fitting of the ( 110 ) diffraction intensities as performed. Peak fitting was performed using the curve fitting toolbox in MATLAB. A Pearson VII type fun ction was used to estimate the peak profile of the texture components. A Pearson VII function was used to account for broad shoulders of the texture components. Prior to peak fitting of the texture components, the background intensity of the (110) diffract ion intensities was estimated using a linear polynomial. For a thin film with no randomly oriented

PAGE 186

186 grains, this background intensity would be zero since a background subtraction was performed prior to data binning. Hence, t he background intensity of the (110) diffraction intensities is contributed by the random texture component of the film. Details of the methodology used for the calculation of the volume of the texture fraction are presented in Section 1.5 The variation of the different texture fractions in the films deposited on various substrates are shown in Figure 8 7 Figure 8 9 The refi ned values for the FWHM of the 111 and 100 texture components from the peak fitting routine are shown in Figure 8 10 Figure 8 12 Plan and cross section micrographs for films deposited on Ti ( Figure 8 13 and Figure 8 14 ), TiO x ( Figure 8 15 and Figure 8 16 ) and ZnO ( Figure 8 17 and Figure 8 18 ) adhesion layers were captured using SEM. A fine grained microstructure is observed to form in some of the films (eg. Films d eposited on Ti adhesion layer substrates and crystallized at 64 C/s and 32 C/s). A predominantly columnar type grain structure is observed in the rest of the films. In general, the grain size is observed to increase with decrease in heating rate. 8.3. Discussion The results obtained on the films investigated in this chapter are observed to be mostly consistent with previous results obtained during in situ crystallization of solution deposited PZT thin films presented in CHAPTER 6 Films heated at 20 C/s in CHAPTER 6 were observed to be random ly texture d while 111 texture components are observed in the thin film crystallized at fast heating rates in this chapter. Thin film samples used in CHAPTER 6 were not controlled for aging in th e precrystallized thin film during storage ( CHAPTER 5 ) while films used during later beam time were stored in

PAGE 187

187 controlled humidity conditions. This difference in st orage during transportation of the samples to APS could be the reason for th e difference in texture A consistent trend towards the formation of 100 texture components during crystallization is only observed in thin films with titanium adhesion layer (SQI substrates). In films crystallized by increasing the voltage applied on the IR lamp at 0.005V/s the 111 and 100 texture components are observed to form at the same instant ( Figure 8 19 A) The (100) planes are the low surface energy planes for the perovskite crystal structure [164] and are expected to have a faster growth rate compared to other planes such as (111) However, the growth rates of the 100 and 111 texture components are observ ed to be similar in these thin films. Hence, kinetic factor s might have a more prominent role during crystallization of solution deposited PZT thin films than the equilibrium surface free energies. FWHM obtained from the peak fitting procedure was used as a metric to quantify the sharpness of the texture components ( Figure 8 19 B). During formation of the perovskite phase, the FWHM of the 100 and the 111 texture components is observed to follow different trends. The FWHM of the 100 texture component is broad initially while the 111 texture component is sharp during initial stages of formation. Both 111 and 100 texture components are observed to converge t o similar value s at later times during crystallization. To better understand the reasons for this behavior of the FWHM of the texture components, sections of the detector image binned to generat e the 110 azimuthal intensities were plotted as intensity maps (I(2 )) Such i ntensity maps (I(2 )) at various times during the formation and growth of the perovskite phase were plotted ( Figure 8 20 ). From these intensity maps, it is observed that d uring initial stages

PAGE 188

188 of the perovskite formation the pyrochlore structured phase is still present in the thin film. Contribution of the pyrochlore structured phase to the binned 110 azimuthal diffraction intensities cannot be avo ided due to the proximity of the perovskite (110) and pyrochlore (222) diffraction peaks in 2 These diffraction intensities from the pyrochlore structured phase could lead to artificial broadening in the 100 texture component. The evolution of intensitie s and FWHM of the texture components during crystallization of PZT thin films solution deposited on various substrates are provided in APPENDIX E The fine grained microstructure observed in the cross sectional images of some films for example: PZT on Pt/Ti crystallized at 64 C/s, PZT on Pt/TiOx crystallized at 0.5 C/s, etc., is usu ally associated with the presence of pyrochlore structured phase in the solution deposited thin films. However, no diffraction peaks corresponding to the presence of a pyrochlore structured phase were observed in the in situ diffraction data The sharpnes s of the 111 and 100 texture components are observed to be similar to the sharpness of the Pt 111 texture component for films deposited on SQI substrates Since no evidence for the transfer of texture from the platinum layer to the perovskite phase through Pt x Pb or oriented pyrochlore structure phases could be observed ( CHAPTER 6 ), other extant proposed mechanisms for texture control are considered. To explain the v ariation in texture observed in films deposited on different substrates, the extant mechanisms for orientation control in solution deposited PZT thin films are considered. Muralt et al. [69, 107] and Bouregba et al. [124] suggested that the formation of a rutile layer at the films substrate controls the final texture obtained in solution deposited PZT thin films. The rutile layer was proposed to form due to the oxidation of

PAGE 189

189 t he titanium diffusing from the adhesion layer. In solution deposited films investigated in this study, 111 texture components are observed to form in films crystallized at fast heating rates irrespective of the adhesion layer. Particularly, in films depo sited on substrates with ZnO adhesion layer, the adhesion layer does not contribute any titatium to the PZT thin films [26] Hence, the 111 texture component is observed to be not dependent on the adhesion layer used in the substrate. To further investigate the dependence of the perovskite texture compo nents on Pt texture, PZT films were solution deposited on Pt/TiO 2 (ARL) substrates These films were pyrolyzed at 350 C during deposition. The phase evolution and (110) azimuthal time plot for these films crystallized at 50 C /s are shown in Figure 8 21 No Pt 3 Pb phase is observed to form in these thin films, probably due the lower expected organic content present in films pyrolyzed at higher temperatures [158] Lower organic content in the thin films could lead to the formation of less severe reducing conditions during crystallization and hence decrease the possibility for Pt 3 Pb formation. The variation of the volume fraction of the texture components with heating rate during crystallization is shown in Figure 8 22 111 100 and random texture components are observed to be present in these thin films after crystallization. The volume fraction of the 100 texture component is obser ved to be similar in films crystallized at different heating rates. However, the volume fraction of the 111 texture component is observed to decrease with decrease in heating rate. Corresponding to the decrease in the volume fraction of the 111 texture com ponent, the random component of the texture is observed to increase. From the refined FWHM of the texture components, t he sharpness of the perovskite 111 texture component is observed to be similar to the Pt(111) texture

PAGE 190

190 ( Figure 8 23 ) No correlation is observed between the sharpness of the (100) texture component and Pt texture. Due to the oxidation anneal of the titanium layer prior to platinum deposition, diffusion o f titanium through the Pt electrode during crystallization is not expected. In addition formation of Pt 3 Pb is not observed in these thin films. Hence, the (111) orientation in these thin films might nucleate directly from the platinum (111). 8.4. Conclusion Phase evolution during crystallization is observed to be consistent between PZT thin films deposited on Pt/Ti, Pt/TiOx and Pt/ZnO substrates. Films crystallized at fast heating rates are observed to be predominantly (111) textured for all substrates. Hence the (111) texture components formed in solution deposited PZT thin films might not be seeded due to the titanium diffusion from the adhesion layer The (111) orientation in these thin films templates could be seeded directly by the platinum layer

PAGE 191

191 Tabl e 8 1 Average heating rates corresponding to different voltage rates used during the in situ diffraction experiments. Voltage rate (V/s) Heating rate Inst 64 .0 1 .0 68 .0 0.4 32 .0 0.1 11 .0 0.01 1 .0 0.005 0.5

PAGE 192

192 Figure 8 1 FWHM of different Pt peaks along

PAGE 193

193 Figure 8 2 Key for interpretation of phase evolution plots denotes perovskite phase and denotes the pyrochlore structured phase. The transient Pt 3 Pb intermetallic and the platinum (111) diffraction peak from the substrate are indicated.

PAGE 194

194 Figure 8 3 Phase evolution plots of PZT films on different substrates crystallized at 64 C/s (inst), 68 C/s and 32 C/s.

PAGE 195

195 Figure 8 4 Phase evolution plots of PZT films on different substrates crystallized at 11 C/s, 1 C/s and 0.5 C/s.

PAGE 196

196 Figure 8 5 (110) azimuthal time (AT) plots for PZT films on different substrates and crystallized at 64 C/s (inst), 68 C/s and 32 C/s. ( ) indicates (111) texture component. The anomalous intensity observed at ~150 in films deposited on substrates with ZnO adhesi on layer and crystallized at 32 C/s could be from the silicon substrate.

PAGE 197

197 Figure 8 6 (110) azimuthal time (AT) plots for PZT films on different substrates and crystallized at 11 C/s, 1 C/s and 0.5 C/s. ( ) indicates (111) texture component and ( ) indicates (100) texture component.

PAGE 198

198 Figure 8 7 Variation of (111), (100) and random texture components in PZT thin films deposited on Pt/Ti substrates and crystallized at different heating rates. Figure 8 8 Variation of (111), (100) and random texture components in PZT thin films deposited on Pt/TiO x substrates and crystallized at different heating rates.

PAGE 199

199 Figure 8 9 Variatio n of (111), (100) and random texture components in PZT thin films deposited on Pt/ZnO substrates and crystallized at different heating rates. Figure 8 10 Variation in the FWHM of the 111 and 100 texture components with heating rate during crystallization for films deposited on Pt/Ti substrates

PAGE 200

200 Figure 8 11 Va riation in the FWHM of the 111 and 100 texture components with heating rate during crystallization for films deposited on Pt/TiO x substrates. Figure 8 12 Variation in the FWHM of the 111 and 100 textur e components with heating rate during crystallization for films deposited on Pt/ZnO substrates.

PAGE 201

201 Figure 8 13 Plan view and cross section SEM images for PZT films deposited on Pt/Ti substrates Films were crystallized at 64 C/s ( A and B ), 68 C/s ( C and D ), and 32 C/s ( E and F ).

PAGE 202

202 Figure 8 14 Plan view and cross section SEM images for PZT films deposited on Pt/Ti substrates Films were crystallized at 11 C/s ( A and B ), 1 C/s ( C and D ), and 0.5 C/s ( E and F ).

PAGE 203

203 Figure 8 15 Plan view and cross section SEM images for PZT films deposited on Pt/TiO x substrates Films were crystallized at 64 C/s ( A and B ), 68 C/s ( C and D ), and 32 C/s ( E and F ).

PAGE 204

204 Figure 8 16 Plan view and cross section SEM images for PZT films deposited on Pt/TiO x substrates Films were crystallized at 11 C/s ( A and B ), 1 C/s ( C and D ), and 0.5 C/s ( E and F ).

PAGE 205

205 Figure 8 17 Plan view and cross section SEM images for PZT films deposited on Pt/ZnO substrates Films were crystallized at 64 C/s ( A ), 68 C/s ( B and C ), and 32 C/s ( D and E ).

PAGE 206

206 Figure 8 18 Plan view and cross section SEM images for PZT films deposited on Pt/ZnO substrates Films were crystallized at 11 C/s ( A and B ) and 1 C/s ( C and D ).

PAGE 207

207 Figure 8 19 Result s of peak fitting of 100 and 111 texture components in PZT thin films deposited on SQI substrates and heated at 0.5 C/s. A ) Change in the integrated intensities of 100 and 111 texture component s and B) c hange in the FWHM of the 100 and 111 texture components during crystallization.

PAGE 208

208 Figure 8 20 Intensity maps (I(2 )) at various times during crystallization of solution deposited PZT thin film depos ited on platinized silicon substrates with titanium adhesion layer (SQI). Diffraction intensities at 2 = 10.3 and 2 = 10.8 correspond to the pyrochlore structured phase and perovskite PZT phases, respectively. Time

PAGE 209

209 Figure 8 21 Results of in situ experiment performed on PZT films deposited on Pt/TiO 2 (ARL) substrates and crystallized at 50 C/s. A ) Phase evolution and B ) (110) azimuthal time plots In the phase evolution plot, ( ) indicated perovskite phase and ( ) indicates pyrochlore structured phase. In azimuthal time plot, ( ) indicates (111) texture component and ( ) indicates (100) texture component. Figure 8 22 Variation in the volume fraction of different texture components with heating rate for PZT films deposited on Pt/TiO 2 (ARL) substrates.

PAGE 210

210 Figure 8 23 Variation in the sharpness of different texture components with heating rate for PZT films deposited on Pt/TiO 2 (ARL) substrates. Error bars indicate 95% confidence intervals

PAGE 211

211 CHAPTER 9 SUMMARY To measure the changes taking place in the films during crystallization, in situ synchrotron and laboratory X ray based diffraction techniques were developed. Analytical treatment was performed to develop i ntensity corrections and analytical expressions for quantita tively analyzing the diffraction intensities measured in these non standard diffraction geometries. Measurements performed using these techniques confirmed that t he Pt x Pb phase formed in these films to be a Pt 3 Pb structured phase. The Pt 3 Pb phase is observ ed to inherit its texture from the platinum bottom electrode and the orientational relationship between Pt 3 Pb and Pt was found to be [111] Pt 3 Pb || [111]Pt. Laboratory based in situ measurements revealed that d ecreasing the nominal lead precursor content i n the initial solution decrease s the formation of the Pt 3 Pb metastable reaction phase between the thin film and the platinum bottom electrode Heating rate during crystallization was found to influence the texture of the crystallized films. N o evidence was observed for seeding of the (111) perovskite orientation by the Pt 3 Pb intermetallic phase or the intermediate pyr o chlore structured phase The adhesion layers used for the Pt bottom electrode had a greater effect on the variation of texture with heating rate Particularly, the texture obtained in films crystallized at slow heating rates was significantly different between films with different adhesion layers.

PAGE 212

212 APPENDIX A MATLAB CODES FOR DAT A REDUCTION A.1. Codes f or Integration o ver % Frame integration about % script for data reduction over gamma % script to run the sequential_command_script in each of the folder % kn 11/21/2011 % list of folder to run the script in %% clearing buffer clear all; close all; %% iteratively move to each folder and run the script root_path = pwd; folder_names = {... 's79' 's80' 's81'... 's82' 's83' 's84' 's85' 's86' 's87' 's90' ... 's92' 's93'... 's94' 's95' 's96' 's97' 's98' 's99' 's100' 's101' 's102'... 's103' 's104' 's105' 's106' 's107' 's108' 's109' 's110'... 's111' ... }; %% path(path, 'I: \ APS_Oct_2011 \ test_automated_data_extractor \ script_files \ final_script_files_chi_plot_ 170_150'); for path_counter = 1:length(folder_names) path_temp = strcat(root_pa th, \ ', folder_names{path_counter}); cd(path_temp); sequential_command_script(); cd .. end rmpath('I: \ APS_Oct_2011 \ test_automated_data_extractor \ script_files \ final_script_files_c hi_plot_170_150');

PAGE 213

213 function [] = sequential_command_script() % program to run all script files in order %% % copy jscript file from detector files to present location path_current = pwd; command = ['copy', 'I: \ APS_Oct_2011 \ s5 \ correct2seq2k_gmap_dsr.js' ',path_current]; dos( command); %% % extract data from .cor files process_batch_gen(); %% % rename .cor files to remove extra commas rename_cor_files_v2; %% % generate chi files from .cor files iterative_chi_file_generator; %% % plot and save data in chi files plot_chi_data; % move all chi files, data and fig file to folder move_chi_files; function [] = process_batch_gen() %% jscript_name = 'correct2seq2k_gmap_dsr.js'; dark_filename = dir('dark*'); % taking only one dark file if there are multiple dark files dark_filename = dark_filename(1).name; all_filenames = dir('s*.*'); % searching in filenames for the sample file for i=1:length(all_filenames) test = regexp(all_filenames(i).name, 's.(.?)(.?)_', 'match'); if ~isempty(test) filename = all_filenames( i).name; break; end end %%

PAGE 214

214 % concatenating to obtain the command string command_string = [jscript_name, ', filename, ', dark_filename,' ', num2str(190), ', num2str(5)]; system(command_string); function [] = rename_cor_files_v2() % new rename cor files program based on dos rename command % the extension of the renamed .cor file is changed to .xyz % kn 11/18/2011 %% clearing buffer %{ clear all; close all; %} %% filenames = dir('*.cor'); %% generating new filenames % cell array for new file names filenames_new = cell(length(filenames),1); %% selectively renaming the .cor.#.cor files to _#.xyz expr = '.cor(.?)'; for i=1:length(filenames) name_parts = regexp(filenames(i).name, expr, 'split'); if length(name_parts)==3 filenames_new{i} = strcat(name_parts ,'_ren_', name_parts [166] ,'.cor'); % renaming the file using dos commands command = ['rename', ', filenames(i).name, ',filenames_new{i}]; system(command); end

PAGE 215

215 end function [] = iterative_chi_file_generator() % function generates a Fit2D macro to extract chi files from .cor file % the macro is then run from Matlab on Fit2D using the com mode % kn 11/17/2011 %{ %% clearing buffer clear all; close all; %} %% extracting the raw image (.xyz) filenames % extracting the names of all the .cor file s file_data = dir('*.cor'); % extracting the file path filepath = pwd; for i = 1:length(file_data) test = regexp(file_data(i).name, 'ren', 'split'); if length(test)==2 filename_prefix = regexp(file_data(i).name, 'ren', 'split'); filename_prefix = filename_prefix; break; end end %% iteratively generating the macro file % opening file to write macro text_filename = strcat(filename_prefix, '_fit2d_chi.txt'); fid = fopen(text_filename, 'w'); for i=1:lengt h(file_data) 1 input_file_full_path = strcat(filepath, \ ', filename_prefix,'ren_', num2str(i 1), '.cor'); output_chi_full_path = strcat(filepath, \ ', filename_prefix, '_',num2str(i 1), '.chi'); if i==1 [part1, part2, part3] = chi _macro_first_file();

PAGE 216

216 else [part1, part2, part3] = chi_macro_general(); end % writing fit2d commands into the macro file for k=1:length(part1) fprintf(fid, '%s \ n', part1{k}); end fprintf(fid, '%s \ n', input_file_full_path); for k=1:length(part2) fprintf(fid, '%s \ n', part2{k}); end fprintf(fid, '%s \ n', output_chi_full_path); for k=1:length(part3) fprintf(fid, '%s \ n', part3{k}); end end %% final_part ={ ... 'EXIT' 'EXIT FIT2D' 'YES' }; % writing the final part to the macro file for i=1:length(final_part) fprintf(fid, '%s \ n', final_part{i}); end % closing the file fclose(fid); %% running the fit2d macro in com mode %{ system('diary'); command = 'diary test'; system(command); dairy off; %} run_fit2d_com_mode(text_filename);

PAGE 217

217 function [part1, part2, part3] = chi_macro_first_file() part1 = {... 'I ACCEPT' 'O.K.' 'POWDER DIFFRACTION (2 D)' 'INPUT' }; part2 = {... 'O.K.' 'O.K.' 'CAKE' 'NO CHANGE' '0' '0' '0' '0' 'INTEGRATE' 'X PIXEL SIZE' '200.0000' 'Y PIXEL SIZE' '200.0000' 'DISTANCE' '296.3460' 'WAVELENGTH' '0.546186' 'X BEAM CENTRE' '1021.777' 'Y BEAM CENTRE' '1092.177' 'TILT ROTATION' '24.98881' 'ANGLE OF TILT' 0.690388' 'O.K.' 'START AZIMUTH' 9 5' 'END AZIMUTH' 85' 'INNER RADIUS' '50' 'OUTER RADIUS'

PAGE 218

218 '950' 'SCAN TYPE' '2 THETA' '1 DEGREE AZ' 'NO' 'AZIMUTH BINS' '1' 'RADIAL BINS' '900' 'CONSERVE INT.' 'NO' 'POLARISATION' 'NO' 'GEOMETRY COR.' 'NO' 'O.K.' 'EXIT' 'OUTPUT' 'CHIPLOT' 'FILE NAME' }; part3 = {... 'O.K.' 'EXCHANGE' }; function [part1, part2, part3] = chi_macro_general() part1 = {... 'INPUT' }; part2 = {... 'O.K.' 'O.K.' 'CAKE' 'INTEGRATE' 'X PIXEL SIZE' '200.0000' 'Y PIXEL SIZE' '200.0000' 'DISTANCE'

PAGE 219

219 '296.3460' 'WAVELENGTH' '0.546186' 'X BEAM CENTRE' '1021.777' 'Y BEAM CENTRE' '1092.177' 'TILT ROTATION '24.98881' 'ANGLE OF TILT' 0.690388' 'O.K.' 'START AZIMUTH' 95' 'END AZIMUTH' 85' 'INNER RADIUS' '50' 'OUTER RADIUS' '950' 'SCAN TYPE' '2 THETA' '1 DEGREE AZ' 'NO' 'AZIMUTH BINS' '1' 'RADIAL BINS' '900' 'CONSERVE INT.' 'NO' 'POLARISATION' 'NO' 'GEOMETRY COR.' 'NO' 'O.K.' 'EXIT' 'OUTPUT' 'CHIPLOT' 'FILE NAME' }; part3 = {... 'O.K.' 'EXCHANGE'

PAGE 220

220 }; function [] = plot_chi_data() % program to read all chi data files and plot the data % data from 5 30 deg is plotted %{ %% clear all; close all; %} %% defining 2theta limits for plotting min_2theta = 5; max_2theta = 25; %% reading the files file_data = dir('*.chi'); filename_prefix = regexp(file_data(1).name, '__', 'split'); filename_prefix = filename_prefix; mat_filename = strcat(filename_prefix, '_mat_file.mat'); %% preallocation of data space fid = fopen(file_data(1).name, 'r'); % first 4 lines have text data for counter = 1:1:4 fgets(fid); end % reading the data data_chi_test = fscanf(fid, '%f %f', [2 inf]); fclose(fid); [m_file, n_file] = size (data_chi_test);

PAGE 221

221 data_intensity = zeros(length(file_data) 1, n_file); %% reading all files data for run_counter = 1:length(file_data) 1 filename = strcat(filename_prefix, '__', num2str(run_counter 1), '.chi'); fid_chi = fopen(filename, 'r'); % first 4 lines have text data for counter = 1:1:4 fgets(fid_chi); end data_chi = fscanf(fid_chi, '%f %f', [2 inf]); fclose(fid_chi); if run_counter ==1 two_theta = data_chi(1,:); end data _intensity(run_counter,:) = data_chi(2,:); end % saving data as a mat file save( mat_filename, 'two_theta', 'data_intensity'); %% acquisition_number = 1:1:(length(file_data) 1); two_theta_start = find(two_theta> min_2theta, 1); two_theta_end = find(two_theta>max_2theta, 1); [twt_grid, time_grid] = meshgrid(two_theta(two_theta_start:two_theta_end), acquisition_number); %% plotting the data : surface plot figure1 = figure;

PAGE 222

222 % Create axes axes1 = axes('Parent',fi gure1,'FontSize',20,'FontName','Arial'); view(axes1,[ 10 80]); hold(axes1,'all'); surf(twt_grid, time_grid, log(data_intensity(:, two_theta_start:two_theta_end)),'Parent',axes1,... 'EdgeColor','none','LineStyle','none','FaceLighting','phong') axis([ two_theta(two_theta_start) two_theta(two_theta_end) acquisition_number(1) acquisition_number(end) min(min(log(data_intensity))) max(max(log(data_intensity)))]); colormap jet grid off; view( 10,80); xlabel('2 \ theta ( ^o )','FontSize',24,'FontName','Arial') ; % Create ylabel %ylabel('Acquisition number','Rotation',104,'FontSize',24,'FontName','Arial'); %% plotting the data: contour plot figure2 = figure; % Create axes axes1 = axes('Parent',figure2,'FontSize',20,'FontName','Arial'); hold(axes1,'all'); contourf(twt_grid, time_grid, log(data_intensity(:, two_theta_start:two_theta_end)),... 'Parent',axes1) axis([two_theta(two_theta_start) two_theta(two_theta_end) acquisition_number(1) acquisition_number(end) min(min(log(data_intensity))) max(max(log(d ata_intensity)))]); colormap jet grid off; xlabel('2 \ theta ( ^o )','FontSize',24,'FontName','Arial'); % Create ylabel ylabel('Acquisition number','FontSize',24,'FontName','Arial');

PAGE 223

223 %% % surface plot filenames surface_fig_name = strcat( filename_prefix, 'surface_plot.fig'); surface_png_name = strcat(filename_prefix, 'surface_plot.png'); % contour figure filenames contour_fig_name = strcat(filename_prefix, 'contour_plot.fig'); contour_png_name = strcat(filename_prefix, 'contour_plot.png') ; %% % saving the surface plots saveas(figure1,surface_fig_name, 'fig' ); print(figure1, r300', dpng', surface_png_name); % saving the contour plot saveas(figure2,contour_fig_name, 'fig' ); print(figure2, r300', dpng', contour_png_name); %% closing all figures close all; function [] = move_chi_files() % function for moving all chi files and plots % into a separate folder %% % making new directory dos('mkdir chi_files'); % moving chi files dos('move *.chi chi_files \ '); % moving mat files dos('move *.mat chi_files \ '); % moving png and .fig files

PAGE 224

224 dos('move *.fig chi_files \ '); dos('move *.png chi_files \ '); A.2. Matlab Codes f or Integrating o ver 2 function [] = sequential_command_azi_plot() iterative_azi_file_generator(); %% projecti on of data onto the gamma range azi_data_projection(); %% plot and save azi_data_plotter(); %% move all azi related files to a specific folder move_azi_files(); function [] = iterative_azi_file_generator() % function generates a Fit2D macro to extract chi files from .cor file % the macro is then run from Matlab on Fit2D using the com mode % kn 11/17/2011 %{ %% clearing buffer clear all; close all; %} %% extracting the raw image (.xyz) filenames % extracting the names of all the .cor files file_data = dir('*.cor'); % extracting the file path filepath = pwd; % determining the filename prefix for i = 1:length(file_data) test = regexp(file_data(i).name, 'ren', 'split'); if length(test)==2 filename_prefix = regexp(file_data(i).name, 'ren ', 'split'); filename_prefix = filename_prefix; break; end end %% iteratively generating the macro file

PAGE 225

225 % opening file to write macro text_filename = strcat(filename_prefix, '_fit2d_azi.txt'); fid = fopen(text_filename, 'w'); % an filedata has one extra .cor file... this is due the extraction % procedure followed by the jscript for i=1:length(file_data) 1 input_file_full_path = strcat(filepath, \ ', filename_prefix,'ren_', num2str(i 1), '.cor'); output_chi_full_path = s trcat(filepath, \ ', filename_prefix, '_',num2str(i 1), '.spr'); if i==1 [part1, part2, part3] = azi_macro_first_file(); else [part1, part2, part3] = azi_macro_general(); end % writing fit2d commands into the macro file for k=1:length(part1) fprintf(fid, '%s \ n', part1{k}); end fprintf(fid, '%s \ n', input_file_full_path); for k=1:length(part2) fprintf(fid, '%s \ n', part2{k}); end fprintf(fid, '%s \ n', output_chi_full_path); for k=1:length(part3) fprintf(fid, '%s \ n', part3{k}); end end %% final_part ={ ... 'EXIT' 'EXIT FIT2D' 'YES' }; % writing the final part to the macro file for i=1:length(final_part) fpri ntf(fid, '%s \ n', final_part{i});

PAGE 226

226 end % closing the file fclose(fid); %% running the fit2d macro in com mode %{ system('diary'); command = 'diary test'; system(command); dairy off; %} run_fit2d_com_mode( text_filename); function [part1, part2, part3] = azi_macro_first_file() part1 = {... 'I ACCEPT' 'O.K.' 'POWDER DIFFRACTION (2 D)' 'INPUT' }; part2 = {... 'O.K.' 'O.K.' 'CAKE' 'NO CHANGE' '0' '0' '0' 0' 'INTEGRATE' 'X PIXEL SIZE' '200.0000' 'Y PIXEL SIZE' '200.0000' 'DISTANCE' '296.3460' 'WAVELENGTH' '0.546186' 'X BEAM CENTRE' '1021.777' 'Y BEAM CENTRE'

PAGE 227

227 '1092.177' 'TILT ROTATION' '24.98881' 'ANGLE OF TILT' 0.690388' 'O.K.' 'START AZIMUTH' 175.00000' 'END AZIMUTH' 5.00000' 'INNER RADIUS' '270.0000' 'OUTER RADIUS' '290.0000' 'SCAN TYPE' '2 THETA' '1 DEGREE AZ' 'NO' 'AZIMUTH B INS' '170' 'RADIAL BINS' '20' 'CONSERVE INT.' 'NO' 'POLARISATION' 'NO' 'GEOMETRY COR.' 'NO' 'O.K.' 'EXIT' 'OUTPUT' 'SPREAD SHEET' 'YES' }; part3 = {... 'O.K.' 'EXCHANGE' }; function [part1, part2, part3] = azi_macro_general() part1 = {...

PAGE 228

228 'INPUT' }; part2 = {... 'O.K.' 'O.K.' 'CAKE' 'INTEGRATE' 'X PIXEL SIZE' '200.0000' 'Y PIXEL SIZE' '200.0000' 'DISTANCE' '296.3460' 'WAVELENGTH' '0.546186' 'X BEAM CENTRE' '1021.777' 'Y BEAM CENTRE' '1092.177' 'TILT ROTATION' '24.98881' 'ANGLE OF TILT' 0.690388' 'O.K.' 'START AZIMUTH' 175.00000' 'END AZIMUTH' 5.00000' 'INNER RADIUS' '270.0000' 'OUTER RADIUS' '290.0000' 'SCAN TYPE' '2 THETA' '1 DEGREE AZ' 'NO' 'AZIMUTH BINS' '170' 'RADIAL BINS' '20' 'CONSERVE INT.' 'NO' 'POLARISATION' 'NO' 'GEOMETRY COR.'

PAGE 229

229 'NO' 'O.K.' 'EXIT' 'OUTPUT' 'SPREAD SHEET' 'YES' }; part3 = {... 'O.K.' 'EXCHANGE' }; function [] = run_fit2d_com_mode(text_filename) % path of the fit2d executable fit2d_path = 'C: \ Users \ Krishna \ Desktop \ '; fit2d_exe = 'fit2d_12_077_i686_WXP.exe'; options = dim2048x2048 com'; command_fit2d = [fit2d_path, fit2d_exe, options]; command = [command_fit2d, ','<', ', text_filename]; system(command); function [] = azi_data_projection() %% reading names of files a nd extracting filename prefix file_data = dir('*.spr'); filename_prefix = regexp(file_data(1).name, '__', 'split'); filename_prefix = filename_prefix; %%

PAGE 230

230 %% % defining azimuthal angles and arrays start_azimuth = 175; end_azimuth = 5; step_no = 170; azi_step = (end_azimuth start_azimuth)/(step_no 1); azimuth_array = start_azimuth:azi_step:end_azimuth; %% reading all files data for run_counter = 1:length(file_data) filename = strcat(filename_prefix, '__', num2str(run_co unter 1), '.spr'); data = importdata(filename); data_intensity = data.data; clear data; [m,n] = size(data_intensity); % resetting data_azi_projection to zeros data_azi_projection = zeros(1,m); for counter = 1:1:m data_azi_projection(counter) = sum(data_intensity(counter,:)); end azi_projection_output = [azimuth_array ; data_azi_projection]; azi_proj_filename = strcat(filename_prefix,'__',num2str(run_counter 1), '_az_proj.txt'); fid_azi_projection = fopen(azi_proj_filename, 'w'); %writing the azimuthal projection into a text file for counter=1:1 :length(azi_projection_output) fprintf(fid_azi_projection, '%f %f', azi_projection_output(:,counter)); fprintf(fid_azi_projection, \ n');

PAGE 231

231 end fclose(fid_azi_projection); end function [] = azi_data_projection() %% reading names of files and extracting filename prefix file_data = dir('*.spr'); filename_prefix = regexp(file_data(1).name, '__', 'split'); filename_prefix = filename_prefix; %% %% % defining azimuthal angles and arra ys start_azimuth = 175; end_azimuth = 5; step_no = 170; azi_step = (end_azimuth start_azimuth)/(step_no 1); azimuth_array = start_azimuth:azi_step:end_azimuth; %% reading all files data for run_counter = 1:length(file_data) filename = strcat(filename_prefix, '__', num2str(run_counter 1), '.spr'); data = importdata(filename);

PAGE 232

232 data_intensity = data.data; clear data; [m,n] = size(data_intensity); % reset ting data_azi_projection to zeros data_azi_projection = zeros(1,m); for counter = 1:1:m data_azi_projection(counter) = sum(data_intensity(counter,:)); end azi_projection_output = [azim uth_array ; data_azi_projection]; azi_proj_filename = strcat(filename_prefix,'__',num2str(run_counter 1), '_az_proj.txt'); fid_azi_projection = fopen(azi_proj_filename, 'w'); %writing the azimuthal project ion into a text file for counter=1:1:length(azi_projection_output) fprintf(fid_azi_projection, '%f %f', azi_projection_output(:,counter)); fprintf(fid_azi_projection, \ n'); end fclose( fid_azi_projection); end function [time, attenuation] = rate_identifier() %% code to identify the heating rate from the name of the file % identifyers used: % heating rate identifier % inst inst % 1 V/s 1vps % 0.4 V/s 0p4vps % 0.1 V/s 0p1vps % 0.01 V/s 0p01vps

PAGE 233

233 % 0.005 V/s 0p005vps %% file_data = dir('*_ren_*'); file_identifier = regexp(file_data(1).name, '_', 'split'); rate_identifier = file_identifier [166] ; %% if strcmpi(rate_i dentifier, 'inst') time = 0.25; % time between acquisitions attenuation = 0.340481; else if strcmpi(rate_identifier, '1vps') time = 0.25; attenuation = 0.340481; else if strcmpi(rate_identifier, '0p4vps') time = 0.25 ; attenuation = 0.340481; else if strcmpi(rate_identifier, '0p1vps') time = 1; attenuation = 0.115928; else if strcmpi(rate_identifier, '0p01vps') time = 4; attenuation = 0.0394712; else if strcmpi(rate_identifier, '0p005vps') time = 8; attenuation = 0.0134392; end end end end end end function [] = move_azi_files() % function for moving all chi files and plots % into a separate folder

PAGE 234

234 %% % making new directory dos('mkdir azi_files_110'); % moving chi files dos('move *.spr azi_files_110 \ '); % moving az_proj files dos('move *_az_proj.txt azi_files_110 \ '); % moving the .mat files dos('move *.mat azi_files_110 \ '); % moving png and .fig files dos('move *.fig azi_files_110 \ '); dos('move *.png azi_files_110 \ ');

PAGE 235

235 APPENDIX B ANALYTICAL CORRECTIO N FOR BRAGG PLANE DI SPLACEMENT IN SYNCHROTRON IN SITU DIFFRACTION SETUP For flat plate geometry with a curved position sensitive detector, an analytical function for the correction of the apparent lattice parameter change due to sample displacement was demonstrated by Pramanick et al. [134] In this section, a similar anal ytical function to correlate between sample height displacement and the change in peak position is attempted. Due to thermal expansion of the stage, the sample is displaced along its normal direction. This displacement is designated as Due to the disp lacement along the sample normal, the position of the beam on the sample changes ( Figure B 1 ). Based on this, the following relationships can be derived: ( B 1 ) Since Bragg condition is always maintained, ( B 2 ) ( B 3 ) Due to the diffraction geometry, ( B 4 ) Combining ( B 3 ) and ( B 4 ) ( B 5 ) ( B 6 ) From Taylor series expansion of the LHS of ( B 3 ) we have

PAGE 236

236 ( B 7 ) Combining ( B 7 ) and ( B 3 ) ( B 8 ) ( B 9 ) ( B 10 ) Combining ( B 10 ) and ( B 8 ) ( B 11 ) For a cubic material Hence, ( B 12 ) Expanding and rearranging ( B 12 ) we get ( B 13 ) ( B 13 ) can be represented in the following form ( B 14 ) Where, ( B 15 ) and ( B 16 ) The above derived methodology indicates that the measured lattice parameter changes with the displacement along the sample normal and the incident angle of the

PAGE 237

237 X ray beam to the sample. To validate the analytical relationship, diffraction images ceria disp ersed in PZT thin films were collected at the following z height and incident angle values ( Table B 1). The apparent variation in the lattice parameter for different values of z height and eta are shown in Figure B 2 The lattice parameters were derived from the refined peak positions of the Ceria (111) peaks. Using the analytical relation for correction described in ( B 16 ) the exact lattice parameters were then est imated. The lattice parameters obtained after correction are shown in Figure B 3 While the analytical correction seems to decrease the variation and dispersion in th e lattice parameter values, complete correction for sample displacement and inclination is not obtained. Hence, the analytical correction bragg plane displacement due to for sample displacement was abandoned and an empirical approach for temperature estima tion was used.

PAGE 238

2 38 Table B 1 z height and eta matrix used for validating the derived analytical model Z Eta Eta Eta Eta Eta Eta Eta 0.00 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.01 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.02 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.03 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.04 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.05 1.0 1.5 2.0 2.5 3.0 3.5 4.0

PAGE 239

239 Figure B 1 Definition of terms for analytical flat plate correction Figure B 2 Variation of uncorrected lattice parameters of the (111) CeO 2 peak

PAGE 240

240 Figure B 3 Calculated lattice parameters using the calculated analytical function

PAGE 241

241 APPENDIX C MARCH 2010 EXPERIMENTAL RUN Two sets of samples were measured for each experimental condition. The phase evolution and azimuthal time plots for both sets are shown to demonstrate repeatability in measurements. C.1. Phase Evolution Plots

PAGE 242

242 C.2. (100) Azimuthal Time Plots

PAGE 243

243 APPENDIX D OCTOBER 2011 EXPERIMENTAL RUN Phase evolution and azimuthal time plots for two sets of samples are shown to demonstrate repeatability. D.1. PZT/Pt/Ti/SiO2/Si (SQI) D.1.1. Phase Evolution Plots

PAGE 244

244

PAGE 245

245 D.1.2. (100) Azimuthal Time Plots

PAGE 246

246

PAGE 247

247 D.2. PZT/Pt/TiOx/SiO2/Si (Sandia) D.2.1. Phase Evolution Plots

PAGE 248

248

PAGE 249

249 D.2.2. (100) Azimuthal Time Plots

PAGE 250

250

PAGE 251

251 D.3. PZT/Pt/ZnO/SiO2/Si (Sandia) D.3.1 Phase Evolution Plots

PAGE 252

252

PAGE 253

253 D.3.2. (100) Azimuthal Time Plots

PAGE 254

254

PAGE 255

255 D.4. PZT/Pt/TiOx/SiO2/Si (ARL) D.4.1. Phase Evolution Plots

PAGE 256

256

PAGE 257

257 D.4.2. (100) Azimuthal Time Plots

PAGE 258

258

PAGE 259

259 APPENDIX E (110) AZIMUTHAL SCAN PEAK FITTING RESULTS Error bars on all plots correspond to 95% confidence intervals. E.1. PZT/Pt/Ti/SiO2/Si (SQI)

PAGE 260

260 E.2. Pt/TiO x /SiO 2 /Si

PAGE 261

261 E.3. PZT/ Pt/ZnO /SiO2/Si

PAGE 262

262 E.4. PZT/ Pt/TiO 2 /SiO 2 /Si ( ARL )

PAGE 263

263 LIST OF REFERENCES [1] Brennecka G L, Ihlefeld J F, Maria J P, Tuttle B A, Clem P G 2010 J. Am. Ceram. Soc. 93 3935. [2] Dimos D, Mueller C H 1998 Annu. Rev. Mater. Sci. 28 397. [3] Muralt P 2000 IEEE Trans. Ultra. Ferro. Freq. Contr. 47 903. [4] Muralt P 1997 Int. Ferro. 17 297. [5] Ledermann N, Muralt P, Baborowski J, Gentil S, Mukati K, Cantoni M, Seifert A, Setter N 2003 Sensors and Actuators a Physical 105 162. [6] Ramesh R, Ed. Thin film ferroelectric materiasl and devices Vol. 1, Kluwer Academic Publishers, Boston 1997 [7] Auciello O, Scott J F, Ramesh R 1998 Physics Today 51 22. [8] Scott J F 2007 Science 315 954. [9] Moulson A J, Herbert J M 2003 Electroceramic s : materials, properties, applications (New York: J. Wiley) [10] Jaffe B, Cook W R, Jaffe H 1971 Piezoelectric Ceramics Academic Press Limited) [11] Nye J F 1957 Physical properties of crystals, their representation by tensors and matrices (Oxford,: Cl arendon Press) [12] Damjanovic D 1998 Reports on Progress in Physics 61 1267. [13] Sakabe Y 1997 Current Opinion in Solid State & Materials Science 2 584. [14] Pithan C, Schneller T, Shiratori Y, Majumder S B, Haegel F H, Dornseiffer J, Waser R 2006 Intern ational Journal of Materials Research 97 499. [15] Losego M D, Jimison L H, Ihlefeld J F, Maria J P 2005 Appl. Phys. Lett. 86 [16] Losego M D, Ihlefeld J F, Maria J P 2008 Chem. Mater. 20 303. [17] Ihlefeld J, Laughlin B, Hunt Lowery A, Borland W, Kingon A, Maria J P 2005 Journal of Electroceramics 14 95. [18] Kingon A I, Srinivasan S 2005 Nature Materials 4 233. [19] Narayanan M, Kwon D K, Ma B, Balachandran U 2008 Appl. Phys. Lett. 92 [20] Ma B H, Narayanan M, Tong S, Balachandran U 2010 J. Mater. Sci. 45 151.

PAGE 264

264 [21] Ong R J, Berfield T A, Sottos N R, Payne D A 2005 Journal of the European Ceramic Society 25 2247. [22] Tu Y L, Calzada M L, Phillips N J, Milne S J 1996 J. Am. Ceram. Soc. 79 441. [23] Berfield T A, Ong R J, Payne D A, Sottos N R 2007 J. Appl Phys. 101 [24] Chen H D, Udayakumar K R, Gaskey C J, Cross L E 1995 Appl. Phys. Lett. 67 3411. [25] Hiboux S, Muralt P, Maeder T 1999 J. Mater. Res. 14 4307. [26] Shelton C T, Kotula P G, Brennecka G L, Lam P G, Meyer K E, Maria J P, Gibbons B J, Ihlefe ld J F 2012 Advanced Functional Materials [27] Muralt P 2000 J. Micromech. Mircoengg. 10 136. [28] Izyumskaya N, Avrutin V, Gu X, Xiao B, Chevtchenko S, Yoon J G, Morkoc H, Zhou L, Smith D J 2007 Appl. Phys. Lett. 91 [29] Izyumskaya N, Alivov Y, Cho S J, Morkoc H, Lee H, Kang Y S 2007 Critical Reviews in Solid State and Materials Sciences 32 111. [30] Okada M, Tominaga K, Araki T, Katayama S, Sakashita Y 1990 Japanese Journal of Applied Physics Part 1 Regular Papers Short Notes & Review Papers 29 718. [31 ] Ramesh R, Aggarwal S, Auciello O 2001 Materials Science & Engineering R Reports 32 191. [32] Krupanidhi S B, Hu H, Kumar V 1992 J. Appl. Phys. 71 376. [33] Iijima K, Ueda I, Kugimiya K 1991 Japanese Journal of Applied Physics Part 1 Regular Papers Short Notes & Review Papers 30 2149. [34] Dey S K, Budd K D, Payne D A 1988 IEEE Trans. Ultra. Ferro. Freq. Contr. 35 80. [35] Haertling G H 1991 Ferroelectrics 116 51. [36] Yi G H, Sayer M 1991 American Ceramic Society Bulletin 70 1173. [37] Assink R A, Schwart z R W 1993 Chem. Mater. 5 511. [38] Schwartz R W, Voigt J A, Tuttle B A, Payne D A, Reichert T L, DaSalla R S 1997 J. Mater. Res. 12 444. [39] Schwartz R W, Schneller T, Waser R 2004 Comptes Rendus Chimie 7 433. [40] Schwartz R W 1997 Chem. Mater. 9 2325.

PAGE 265

265 [41] Schwartz R W, Boyle T J, Lockwood S J, Sinclair M B, Dimos D, Buchheit C D 1995 Int. Ferro. 7 259. [42] Brinker C J 1990 Sol gel science: the physics and chemistry of sol gel processing Academic press limited) [43] Budd K D, Payne D A 1989 Institute of Physics Conference Series 13. [44] Fukushima J, Kodaira K, Matsushita T 1984 J. Mater. Sci. 19 595. [45] Vest R W, Xu J J 1988 IEEE Trans. Ultra. Ferro. Freq. Contr. 35 711. [46] Vest R W, Zhu W 1991 Ferroelectrics 119 61. [47] Haertling G H 1991 Journa l of Vacuum Science & Technology a Vacuum Surfaces and Films 9 414. [48] Yi G H, Wu Z, Sayer M 1988 J. Appl. Phys. 64 2717. [49] Boyle T J, Dimos D, Schwartz R W, Alam T M, Sinclair M B, Buchheit C D 1997 J. Mater. Res. 12 1022. [50] Huang Z, Zhang Q, Whatmore R W 2001 Int. Ferro. 36 153. [51] Huang Z, Zhang Q, Whatmore R W 2002 Journal of Sol Gel Science and Technology 24 49. [52] Zhang Q, Huang Z, Whatmore R W 2002 Journal of Sol Gel Science and Technology 23 135. [53] Sengupta S S, Ma L, Adler D L, Payne D A 1995 J. Mater. Res. 10 1345. [54] Feth M P, Weber A, Merkle R, Reinohl U, Bertagnolli H 2003 Journal of Sol Gel Science and Technology 27 193. [55] Wilkinson A P, Xu J, Pattanaik S, Billinge S J L 1998 Chem. Mater. 10 3611. [56] Arcon I, Malic B, Kodre A, Kosec M 1999 Journal of Synchrotron Radiation 6 535. [57] Malic B, Arcon I, Kodre A, Kosec M 2006 J. Appl. Phys. 100 [58] Malic B, Kosec M, Arcon I, Kodre A 2005 Journal of the European Ceramic Society 25 2241. [59] Bornside D E Macosko C W, Scriven L E 1989 J. Appl. Phys. 66 5185. [60] Bornside D E, Macosko C W, Scriven L E 1987 Journal of Imaging Technology 13 122.

PAGE 266

266 [61] Lange F F 1996 Science 273 903. [62] Buchanan R C 2004 Ceramic materials for electronics (New York: Marcel Dekker) [63] Sreenivas K, Reaney I, Maeder T, Setter N, Jagadish C, Elliman R G 1994 J. Appl. Phys. 75 232. [64] Tani T 1993 PhD Thesis University of Illinois at Urbana Champaign [65] Fox G R, Troliermckinstry S, Krupanidhi S B, Casas L M 1995 J. Mater. Re s. 10 1508. [66] AlShareef H N, Dimos D, Tuttle B A, Raymond M V 1997 J. Mater. Res. 12 347. [67] Nam H J, Choi D K, Lee W J 2000 Thin Solid Films 371 264. [68] Maeder T, Sagalowicz L, Muralt P 1998 Japanese Journal of Applied Physics Part 1 Regular Papers Short Notes & Review Papers 37 2007. [69] Muralt P 2006 J. Appl. Phys. 100 051605. [70] Lakeman C D E, Xu Z K, Payne D A 1995 J. Mater. Res. 10 2042. [71] Levi C G 1998 Acta Materialia 46 787. [72] Norga G J, Vasiliu F, Fe L, Wouters D J, Van der Biest O 2003 J. Mater. Res. 18 1232. [73] Kosec M, Malic B 1998 Journal De Physique Iv 8 17. [74] Reaney I M, Brooks K, Klissurska R, Pawlaczyk C, Setter N 1994 J. Am. Ceram. Soc. 77 1209. [75] Wilkinson A P, Speck J S, Cheetham A K, Natarajan S, Thomas J M 1994 Chem. Mater. 6 750. [76] Levi C G 1999 Revista Mexicana De Fisica 45 30. [77] Kwok C K, Desu S B 1992 Appl. Phys. Lett. 60 1430. [78] Huang Z, Zhang Q, Whatmore R W 1998 J. Mat. Sci. Lett 17 1157. [79] Griswold E M, Weaver L, Sayer M, Calder I D 1995 J. Ma ter. Res. 10 3149. [80] Griswold E M, Sayer M, Weaver L 1995 Int. Ferro. 8 109. [81] Chen S Y, Chen I W 1994 J. Am. Ceram. Soc. 77 2332.

PAGE 267

267 [82] Brooks K G, Reaney I M, Klissurska R, Huang Y, Bursill L, Setter N 1994 J. Mater. Res. 9 2540. [83] Ellerkmann U, Schneller T, Nauenheim C, Bottger U, Waser R 2008 Thin Solid Films 516 4713. [84] Brennecka G L, Parish C M, Tuttle B A, Brewer L N, Rodriguez M A 2008 Adv. Mater. 20 1407. [85] Jaffe B, Cook W R, Jaff* H L C 1971 Piezoelectric ceramics (London, New York ,: Academic Press) [86] Kwok C K, Desu S B 1994 J. Mater. Res. 9 1728. [87] Es Souni M, Piorra A 2001 Materials Research Bulletin 36 2563. [88] Hu H, Peng C J, Krupanidhi S B 1993 Thin Solid Films 223 327. [89] Kwok C K, Desu S B, Vijay D P 1993 Ferroelectrics Letters Section 16 143. [90] Tuttle B A, Headley T J, Bunker B C, Schwartz R W, Zender T J, Hernandez C L, Goodnow D C, Tissot R J, Michael J, Carim A H 1992 J. Mater. Res. 7 1876. [91] Polli A D, Lange F F, Levi C G 2000 J. Am. Ceram. Soc. 83 873. [92] Chen J, Udayakumar K R, Brooks K G, Cross L E 1992 J. Appl. Phys. 71 4465. [93] Calzada M L, Alguero M, Pardo L 1998 Journal of Sol Gel Science and Technology 13 837. [94] Dang E K F, Gooding R J 1995 Physical Review Letters 74 3848. [95] Tani T, Payne D A 1994 J. Am. Ceram. Soc. 77 1242. [96] Lefevre M J, Speck J S, Schwartz R W, Dimos D, Lockwood S J 1996 J. Mater. Res. 11 2076. [97] Tuttle B A, Voigt J A, Garino T J, Goodnow D C, Schwartz R W, Lamppa D L, headley T J, Eatough M O, 1992, Eigh t IEEE International Symposium on Applications of Ferroelectrics Greenville, SC, USA [98] Whatmore R W, Zhang Q, Shaw C P, Dorey R A, Alcock J R 2007 Physica Scripta T129 6. [99] Seifert A, Lange F F, Speck J S 1995 J. Mater. Res. 10 680. [100] Kim J H, L ange F F 1999 J. Mater. Res. 14 4004. [101] Floquet N, Hector J, Gaucher P 1998 J. Appl. Phys. 84 3815.

PAGE 268

268 [102] Ricote J, Poyato R, Alguero M, Pardo L, Calzada M L, Chateigner D 2003 J. Am. Ceram. Soc. 86 1571. [103] Poyato R, Calzada M L, Ricote J, Pardo L, Willing B 2001 Int. Ferro. 35 1807. [104] Chen S Y, Chen I W 1998 J. Am. Ceram. Soc. 81 97. [105] Brennecka G L, Parish C M, Jones J L, Tuttle B A, Wheeler J S, G. E J, 2009, Proceedings of the US Japan meeting on Piezoelectrics [106] Chen S Y, Chen I W 1994 J. Am. Ceram. Soc. 77 2337. [107] Muralt P, Maeder T, Sagalowicz L, Hiboux S, Scalese S, Naumovic D, Agostino R G, Xanthopoulos N, Mathieu H J, Patthey L, Bullock E L 1998 J. Appl. Phys. 83 3835. [108] Bursill L A, Brooks K G 1994 J Appl. Phys. 75 4501. [109] Norga G J, Maes J, Coppye E, Fe L, Wouters D, Van der Biest O 2000 J. Mater. Res. 15 2309. [110] Kwok C K, Desu S B 1993 J. Mater. Res. 8 339. [111] Ohring M 2002 Materials Science of thin films (San Diego: Academic Press) [11 2] Chopra K L 1969 Thin film phenomena (New York: McGraw Hill) [113] Porter D A, Easterling K E 1992 Phase transformation in metals and alloys Taylor and Francis) [114] Flemings M C 1974 Solidification processing (New York,: McGraw Hill) [115] Habouti S Solterbeck C H, Es Souni M, Zaporojtchenko V 2008 J. Appl. Phys. 104 [116] Kaewchinda D, Chairaungsri T, Naksata M, Milne S J, Brydson R 2000 Journal of the European Ceramic Society 20 1277. [117] Huang Z, Zhang Q, Whatmore R W 1999 J. Appl. Phys. 85 73 55. [118] Gong W, Li J F, Chu X C, Gui Z L, Li L T 2004 Acta Materialia 52 2787. [119] Chen S Y 1996 Materials Chemistry and Physics 45 159. [120] Voigt J A, Tuttle B A, Headley T J, Lamppa D L, 1995, Mat. Res. Soc. Symp. Proc. Boston [121] Fox G R, Krupanidhi S B 1994 J. Mater. Res. 9 699.

PAGE 269

269 [122] Liu Y M, Phule P P 1996 J. Am. Ceram. Soc. 79 495. [123] Aoki K, Fukuda Y, Numata K, Nishimura A 1995 Japanese Journal of Applied Physics Part 1 Regular Papers Short Notes & Review Papers 34 192. [12 4] Bouregba R, Poullain G, Vilquin B, Murray H 2000 Materials Research Bulletin 35 1381. [125] Birkholz M 2006 Thin Film Analysis by X Ray Scattering (Weinheim: Wiley VCH) [126] Jones J L, Slamovich E B, Bowman K J 2004 J. Mater. Res. 19 3414. [127] He B B 2009 [128] Nittala K, Brennecka G L, Tuttle B A, Jones J L 2011 J. Mater. Sci. 46 2148. [129] Hammersley A 2004 [130] Liss K D, Schmoelzer T, Yan K, Reid M, Peel M, Dippenaar R, Clemens H 2009 J. Appl. Phys. 106 [131] Engler O, Randle V Introduction to texture analysis : macrotexture, microtexture, and orientation mapping (Boca Raton: CRC Press) [132] Kocks U F, Tome C N, Wenk H R 1998 Texture and anisotropy : preferred orientations in polycrystals and their effect on materials properties (Cambridge : Cambridge University Press) [133] Shackelford J F, Alexander W 2000 [134] Pramanick A, Omar S, Nino J C, Jones J L 2009 J. Appl. Crystallogr. 42 490. [135] Seltzer S M 1993 Radiation Research 136 147. [136] Aggarwal S, Madhukar S, Nagaraj B, Jenkins I G, Ramesh R, Boyer L, Evans J T 1999 Appl. Phys. Lett. 75 716. [137] Brennecka G L, Parish C M, Tuttle B A, Brewer L N 2008 J. Mater. Res. 23 176. [138] Jacobs R N, Salamanca Riba L 2003 J. Mater. Res. 18 1405. [139] Parish C M, Brennecka G L, Tuttle B A, Brewer L N 2008 J. Mater. Res. 23 2944. [140] Breval E, Wang C, Dougherty J P, Gachigi K W 2005 J. Am. Ceram. Soc. 88 437. [141] Calame F, Muralt P 2007 Appl. Phys. Lett. 90 [142] Parish C M, Brennecka G L, Tut tle B A, Brewer L N 2008 J. Am. Ceram. Soc. 91 3690.

PAGE 270

270 [143] Bradley D C, Mehrotra R C, Gaur D P 1978 Metal Alkoxides (London: Academic Press) [144] Nakamoto K 2009 Infrared and Raman spectra of inorganic and coordination compounds (Hoboken, N.J.: Wiley) [ 145] Coffman P R, Barlingay C K, Gupta A, Dey S K 1996 Journal of Sol Gel Science and Technology 6 83. [146] Atkins P W, De Paula J 2006 Atkins' Physical chemistry (New York: W.H. Freeman) [147] Parish C M, Brewer L N 2010 Microscopy and Microanalysis 16 259. [148] Parish C M, Brewer L N 2010 Ultramicroscopy 110 134. [149] Parish C M, Brewer L N 2009 Microscopy and Microanalysis 15 456. [150] Impey S A, Huang Z, Patel A, Beanland R, Shorrocks N M, Watton R, Whatmore R W 1998 J. Appl. Phys. 83 2202. [151] Bastani Y, Bassiri Gharb N 2012 Acta Materialia 60 6. [152] Selbach M S, Tybell T, Einarsrud M A, Grande T 2011 Appl. Phys. Lett. 98 [153] Yamano A, Kozuka H 2007 J. Am. Ceram. Soc. 90 3882. [154] Chen S Y, Chen I W 1994 J. Am. Ceram. Soc. 77 2332. [155] Griswold E M, Weaver L, McIntyre D S, Sayer M, Calder I D 1995 Int. Ferro. 10 123. [156] Rossini F D, Wagman D D, Evans W H, Prosen E J 1950 Annual Review of Physical Chemistry 1 1. [157] Chen S Y, Chen I W 1994 J. Am. Ceram. Soc. 77 2337. [158] Dippel A C Schneller T, Waser R, Park D, Mayer J 2010 Chem. Mater. 22 6209. [159] Huang Z, Zhang Q, Whatmore R W 1998 J. Mat. Sci. Lett 17 1865. [160] Hansen M, Elliott R P, Shunk F A, Anderko K 1958 Constitution of binary alloys (New York,: McGraw Hill) [161] Pot repka D M, Fox G, Sanchez L M, Polcawich R G, 2010, MRS Fall meeting Boston

PAGE 271

271 [162] Lee J H, Almer J, Aydiner C, Bernier J, Chapman K, Chupas P, Haeffner D, Kump K, Lee P L, Lienert U, Miceli A, Vera G 2007 Nuclear Instruments & Methods in Physics Research Section a Accelerators Spectrometers Detectors and Associated Equipment 582 182. [163] Lee J H, Aydiner C C, Almer J, Bernier J, Chapman K W, Chupas P J, Haeffner D, Kump K, Lee P L, Lienert U, Miceli A, Vera G 2008 Journal of Synchrotron Radiation 15 477. [164] Seifert A, Vojta A, Speck J S, Lange F F 1996 J. Mater. Res. 11 1470.

PAGE 272

272 BIOGRAPHICAL SKETCH The story of Krishna Nittala starts, as a first approximation, in a nondescript hospital in a small city in India where he was born The youngest of three children, his inquisitive nature manifested through the unhesitant dismantling of mechanical and electrical appliances during his childhood much to the angst of his parents. This fascination for all things mechani cal and electrical naturally steered him towards engineering in college. Years spent in college, however, proved to be much more valuable than anticipated due to the educational companionship provided by some extremely intelligent and resourceful friends. A brief stint into industry quickly followed college. Working in the semiconductor industry whetted his appetite to learn more about the material science behind the advanced technological devices. To satisfy his eagerness to learn, he joined the doctoral program in Department of Materials Science and Engineering at the University of Florida. During his doctoral research he initially tinkered with biomolecular motor proteins before setting his focus on understanding processing of ferroelectric thin films. T he perks of working on ferroelectric thin films included travelling and meeting great people. To the date this thesis was written relics l the havoc caused. While tempted to replace these items with newer and better versions, the sentimental value accrued through the years dissuades against any such action. This t works,