Crystallization Behavior of Solution Deposited Lead Zirconate Titanate Thin Films

MISSING IMAGE

Material Information

Title:
Crystallization Behavior of Solution Deposited Lead Zirconate Titanate Thin Films
Physical Description:
1 online resource (148 p.)
Language:
english
Creator:
Mhin, Sung Wook
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Materials Science and Engineering
Committee Chair:
Jones, Jacob L
Committee Members:
Phillpot, Simon R
Myers, Michele V
Perry, Scott S
Biswas, Amlan

Subjects

Subjects / Keywords:
crystallization -- pzt -- thinfilms -- xrd
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:
Leadzirconate titanate (Pb(Zrx,Ti1-xO3, PZT) is awell-known piezoelectric and ferroelectric material. PZT in the form of thinfilms has current applications in capacitors, microsensors and microscalemechanical bugs. For the preparation of PZT thin films, chemical solutiondeposition is one of the most utilized techniques due to the low cost andaccurate control of stoichiometry in the thin film. During chemical solutiondeposition, PZT precursor solutions are deposited onto a substrate using spincoating, and after pyrolysis, amorphous thin films are obtained. The amorphousthin films are crystallized by heating to a higher temperature. Intermediate PtxPband fluorite phases are commonly observed prior to the formation of perovskitePZT thin films during crystallization. In addition, depending on the differentprocessing conditions during crystallization such as the atmosphericconditions, heating rates and different substrates, different texture can beobtained in PZT thin films. However, the mechanism of texture control in PZTthin films is not well understood, because it is difficult to track the changesduring fast crystallization from amorphous to perovskite PZT within a fewseconds. Thus, in situ x-raydiffraction (XRD) measurements are useful to observe the phase and textureevolution of PZT thin films during crystallization. Inthis dissertation, in situmeasurements of PZT thin films using laboratory- and synchrotron based XRD wereconducted to investigate the phase and texture evolution duringcrystallization. The stability of intermediate phases and perovskite PZT wasobserved during crystallization in different atmospheric conditions. Based onthese observations, a novel processing method was developed. Switching atmospheric conditions during crystallization ofPZT thin films suppressed the intermetallic PtxPb phase and promotedthe perovskite PZT phase. Further,based on the observations of phase and texture evolution using differentheating rates and substrates during crystallization, a mechanism for texturecontrol of PZT thin films is provided. 111 texture was observed in PZT on 111Pt electrodes. It is suggested that 111 Pt nucleates 111PZT directly. Similar trends for the formation of 111 texture were observedin PZT on an amorphized PbTiO3 (PTO) seed layer, which implies thatnucleation of 111 texture occurred on a 111 Pt electrode. Conversely,dominant 100 texture was observed in PZT on amorphized PTO duringcrystallization at slow heating rates. It is suggested that surface orhomogeneous nucleation occurs at slow heating rates to promote 100texture in PTO layer at low temperature, and 100 textured PZT isnucleated from 100 texture of PTO layer at higher temperature. Also,interdiffusion between the PTO layer and the PZT layer during crystallizationled to the formation of PZT with an inhomogeneous composition throughout thethin films. Unlike PZT on amorphized PTO, dominant 100 texture was observedin PZT on crystalline PTO regardless of heating rates during crystallization.Also, the formation of the PZT thin films was observed after phasetransformation of the PTO seed layer from tetragonal to cubic above the Curietemperature, which implies that formation of the 100 texture in PZT thinfilms was nucleated from the cubic PTO seed layer with 100 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 Sung Wook Mhin.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: Jones, Jacob L.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0045898:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 CRYSTALLIZATION BEHAVIOR OF SOLUTION DEPOSITED LEAD ZIRCONATE TITANATE THIN FILMS By SUNGWOOK MHIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TH E DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013

PAGE 2

2 2013 Sungwook Mhin

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGEMENTS Foremost, I would like to thank my advisor Jacob L. Jones for his support and guidance over the year s. Dr. Jones provided me the opportunity to work in his group and allowed me to work in various fields such as lead free materials, energy harvesting and thin films. I am very grateful to Dr. Jones for valuable discussion and care for this work. Also, I re spect Dr. Jones to have an eye for selection of group members. During my PhD course, I am grateful to my group members for their understanding, kindness and their care for me. I always felt confident because my group members supported me. Also, my PhD cour se was full of opportunities to learn and understand cultural differences and how to coexist together based on the experiences with group members from different countries such as United States, Australia, India, Turkey, Thailand, Brazil and China. I would like to thank Drs. Krishna Nittala and Jennifer Forrester for not only helping me to move my research forward, but also being my good friends and mentors. I would like to show my gratitude to Drs. Goknur Tutuncu, Dipankar Gosh, Elena Aksel, Shruti Seshadri Anderson Prewitt and Chris Fell for their support for my research. They always helped me and showed me their kindness and friendship. I am still dreaming to participate in another mudrun together. Also, I would like to thank to Tedi Marie Usher, Thanakor n Iamsasri, Tarielle Sanders, Giovanni Esteves, Brienne Johnson and Henry Aldridge for their sincere hearts and support. I hope we will meet again at conferences to talk about our memory in Dr. Jacob L. Jones group! I would like to acknowledge the help from my committee members, Dr. Simon Phillpot, Dr. Michele Manuel, Dr. Scott Perry and Dr. Amian Biswas for their time and guidance. Also, I am grateful to my previous committee member, Dr. Craciun, for

PAGE 5

5 training of different types of scans using X ray diff raction, which improved my understanding of X ray diffraction. This work was completed with help from different national laboratories and people. I am thankful to Drs. Geoff L. Brennecka and Jon Ihlefeld at Sandia National Laboratory. They provided many thin film samples for measurements. They were helpful in the discussion about PZT thin films crystallized in different atmospheric conditions. Also, I would like to express deepest gratitude to Drs. Ronald G. Polcawich and Luz M. Sanchez at Army Research Laboratory for their support and discussion. I would like to acknowledge Los Alamos National Laboratory for teaching and training me neutron scattering. I would also like to acknowledge Dr. Douglas S. Robinson for the technical expertise of in situ XRD set up at beamline 6 ID B, Advanced Photon Source, Argonne National Laboratories. Especially, I am so grateful for his kindness in helping my learning experience. I would like to thank my family and friends for their continuous support and encouragement over the years. Especially, I am so grateful to father, mother and younger sister who encouraged me to pursue a PhD degree in materials science and engineering. I love you all! Finally, I would like to thank the National Science Foundation, Sandia National La boratory and the Department of the Army for funding this research. This work is supported by NSF under DMR 1207293, the U.S. Department of the Army under W911NF 09 1 0435, and a UF Science for Life undergraduate award. Sandia National Laboratories is a mul ti program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S.

PAGE 6

6 AC04 94AL85000.

PAGE 7

7 TABLE OF CONTEN TS Page ACKNOWLEDGEMENTS ................................ ................................ ............................... 4 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 1. 1 Chemical Solution Deposition of PZT Thin Films ................................ .............. 18 1.1.1 Solution Preparation ................................ ................................ ................ 20 1.1.2 Spin Coating and Pyrolysis ................................ ................................ ...... 21 1.1.3 Crystallization ................................ ................................ .......................... 23 1. 2 Phase Evolution ................................ ................................ ................................ 25 1. 3 Texture E volution ................................ ................................ .............................. 27 2 MOTIVATION AND OUTLINE ................................ ................................ ................ 36 3 DETAILS OF GENERAL SAMPLE PREPARATION ................................ ............... 38 3.1 IMO PZT Solution Preparation and Deposition ................................ ................. 38 3.2 2 MOE PTO and PZT Solution Preparation and Deposition ............................. 39 3.2.1 Standard Processing ................................ ................................ ............... 39 3.2.2 Processing of Ceria Dispersed PZT Thin Films on PTO Seed Layer ...... 41 3.2.3 Pt/TiO x /SiO 2 /Si Substrate ................................ ................................ ........ 41 4 GE NERAL EXPERIMENTAL SETUP AND DATA EXTRACTION METHODOLOGIES ................................ ................................ ................................ 47 4.1 In situ Laboratory Based X ray Diffraction Setup and Geometry ...................... 47 4.2 Synchrotron X ray Diffraction Setup and Geometry ................................ .......... 48 4.3 Data Reduction and Representation ................................ ................................ 49 4.4 Temperature Calibration ................................ ................................ ................... 55 5 EFFECT OF CONSTANT AND SWITCHING ATMOSPHERIC CONDITIONS DURING CRYSTALLIZATION ................................ ................................ ................ 66 5.1 Literature Review ................................ ................................ .............................. 66 5.2 Experimental Procedure ................................ ................................ ................... 67 5.3 Results and Discussion ................................ ................................ ..................... 68 5.4 Summary ................................ ................................ ................................ .......... 71

PAGE 8

8 6 THE STABILITY OF THE LEAD TITANATE SEED LAYER BEFORE CRYSTALLIZATION OF PZT THIN FILMS ................................ ............................. 78 6.1 Literature Review ................................ ................................ .............................. 78 6.2 Experimental Procedure ................................ ................................ ................... 79 6.3 Results and Discussion ................................ ................................ ..................... 80 6.4 Summary ................................ ................................ ................................ .......... 83 7 CRYSTALLIZATION OF PZT THIN FILMS ON PLATINUM ELECTRODES .......... 88 7.1 Literature Review ................................ ................................ .............................. 88 7.2 Experimental Procedure ................................ ................................ ................... 89 7.3 Results and Discussion ................................ ................................ ..................... 90 7.4 Summary ................................ ................................ ................................ .......... 96 8 CRYSTALLIZATION OF PZT THIN FILMS ON AMORPHIZED PbTiO 3 SEED LAYERS ................................ ................................ ................................ ................ 105 8.1 Literature Review ................................ ................................ ............................ 105 8.2 Expe rimental Procedure ................................ ................................ ................. 106 8.3 Results and Discussion ................................ ................................ ................... 107 8.4 Summary ................................ ................................ ................................ ........ 114 9 CRYSTALLIZATION OF PZT THIN FILMS ON CRYSTALLINE PbTiO 3 SEED LAYER S ................................ ................................ ................................ ................ 122 9.1 Literature Review ................................ ................................ ............................ 122 9.2 Experimental Procedu re ................................ ................................ ................. 123 9.3 Results and Discuss ion ................................ ................................ ................... 124 9.4 Summary ................................ ................................ ................................ ........ 128 10 CONCLUSION S AND CONTRIBUTIONS ................................ ............................ 138 10.1 Conclusions ................................ ................................ ................................ .. 138 10.2 Contributions of Dissertation ................................ ................................ ......... 141 LIST OF REFERENCES ................................ ................................ ............................. 142 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 148

PAGE 9

9 LIST OF FIGURES Figure Page 1 1 Crystal structure of the ABO 3 perovskite PZT. ................................ .................... 31 1 2 Equ ilibrium phase diagram of PZT. ................................ ................................ .... 32 1 3 CSD of PZT thin films on a platinized substrate. ................................ ................ 33 1 4 Effect of pyrolysis temperature on the orientation of PZT 40/60 films after identical crystallization at 700C and proposed mechanism(s ) ] .......................... 34 1 5 Richardson Ellingham diagram of PbO and PbO 2 ................................ .......... 35 3 1 Preparation of IMO PZT solution. ................................ ................................ ....... 43 3 2 Preparation of 2 MOE PZT solution. ................................ ................................ ... 44 3 3 Preparation of 2 MOE PTO solution. ................................ ................................ .. 45 3 4 Micrograph and roughness of the Pt electrode before and after pyrolysis. ......... 46 4 1 In situ X RD setup ................................ ................................ ............................... 56 4 2 2D detector image s A) Representative 2D detector image. B) Calibration of geometric factors of XRD using Fit2D. ................................ ............................... 57 4 3 Intensity vs. 2 line plot. ................................ ................................ ..................... 58 4 4 Quantitative analysis of texture components. ................................ .................... 59 4 5 Summary of correction factors for 100 and 111 tex ture components. ................ 60 4 6 Intensity map of PZT thin films ................................ ................................ .......... 61 4 7 Representative evo lution plots. ................................ ................................ .......... 62 4 8 Temperature calibration using CeO 2 standard powder. ................................ ...... 63 4 9 Temperature profiles during crystallization of PZT thin films at different heating rates. ................................ ................................ ................................ ...... 64 4 10 Heating rates at 300C and 600C with respect to different voltage rates. ......... 65 5 1 In situ diffraction patterns of PZT thin films during crystallization under different atmospheric conditions ................................ ................................ ........ 73

PAGE 10

10 5 2 Scanning electron micrographs and XRD patterns of PZT thin films after crystallization in different atmospheres. ................................ .............................. 74 5 3 Schematic representation of thermal and atmospheric conditions for thin film crystallization in switching atmos pher es ................................ ........................... 75 5 4 XRD patterns of PZT thin films during crystallization in different switching atmospheric conditions ................................ ................................ ..................... 76 5 5 XRD patterns of PZT thin films after crystallization in different switching temperatures ................................ ................................ ................................ .... 77 6 1 PTO seed layer after pyrolysis of PZT layer at different temperature. ................ 84 6 2 Diffracti on patterns of PTO seed layer. ................................ .............................. 85 6 3 {110} Intensity maps of PTO seed la yer after pyrolysis of PZT layer at different temperatures. ................................ ................................ ....................... 86 6 4 Dark field micrograph of PZT thin films on a PTO seed l ayer after pyr olysis of PZT layer. ................................ ................................ ................................ .......... 87 7 1 XRD patterns of PZT thin films on Pt electrodes after crystallization with different heating rates. ................................ ................................ ........................ 97 7 2 Phase and texture evolution of PZT thin films on Pt electrodes with different heating rates. ................................ ................................ ................................ ...... 98 7 3 Intensity map of Pt a nd PtxPb at the maximum peak intensity during crystallization. ................................ ................................ ................................ ..... 99 7 4 The formation of Pt x Pb phase. ................................ ................................ ......... 100 7 5 Absence of (111) superlattice reflection of pyrochlore at 2 = 6.4 in synchrotron XRD patterns, confirming the phase is fluorite. ............................. 101 7 6 Integrated intensities of 111 fluorite at 2 range (9.8 2 1 0.5) over range between 90 and 90 ................................ ................................ ............. 102 7 7 Change in FWHM and normalized intensity of texture components during crystallization of PZ T with different heating rates ................................ ............ 103 7 8 Texture components of PZT thin films. ................................ ............................. 104 8 1 Phase and texture evolution plots of PZT thin films on an amorphized PTO seed layer during crystallization with different heating rates. ............................ 115

PAGE 11

11 8 2 Change in p eak intensities of (111) PtxPb, (111) PZT, (111) fluorite, (001) PZT, and (100) PZT during crystallization. ................................ ....................... 116 8 3 Phase e volution plot of a PTO seed layer during crystallization at the voltage rate (0.01 V/s). ................................ ................................ ................................ .. 117 8 4 Intensity shift of PZT th in films on amorphous PTO seed layer during crystallization. ................................ ................................ ................................ .. 118 8 5 Change in FWHM of texture components during crystallization of PZT with different heating rates. ................................ ................................ ...................... 119 8 6 Texture components of PZT thin films on amorphous PTO seed layer. .......... 120 8 7 Dark field micrograph of a PZT thin film on a PTO seed layer after crystallization of the PZT layer. ................................ ................................ ......... 121 9 1 Intensity ma p of crystallin e PTO seed layer ................................ .................... 130 9 2 Phase and texture evolution plots of PZT thin films on PTO seed layer during crystallization at different heating rates. ................................ ........................... 131 9 3 The formation of Pt x Pb phase. ................................ ................................ .......... 132 9 4 Absence of (111) superlattice peak in pyrochlore in PZT thin films during crystallization of PZT thin fi lms at different heating rates. ................................ 133 9 5 Formation of pyrochlore and perovskite PZT during crystallization of PZT layer at different heating rates. ................................ ................................ ......... 134 9 6 Texture evolution of PZT thin films during crystallization at different heating rates. ................................ ................................ ................................ ................ 135 9 7 Texture components of PZT thin films on crystalline PTO seed layer. ............ 136 9 8 Diffraction patterns of PZT thin films on PTO s eed layer after crystallization at different heating rates. ................................ ................................ ...................... 137

PAGE 12

12 Abstract of Dissertation Presented to the Graduate School of the Uni versity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CRYSTALLIZATION BEHAVIOR OF SOLUTION DEPOSITED LEAD ZIRCONATE TITANATE THIN FILMS By Sungwook Mhin August 2013 Chair: Jacob L. Jones Ma jor: Materials Science and Engineering Lead zirconate titanate ( Pb(Zr x Ti 1 x O 3 PZT) is a well known piezoelectric and ferroelectric material. PZT in the form of thin films ha s current applications in capacitor s microsensors and microscale mechanical bugs For the preparation of PZT thin films, chemical solution deposition is one of the most utilized techniques due to the low cost and accurate control of stoichiometry in the thin film. During chemical solution deposition, PZT precursor solutions are deposi t ed onto a substrate using spin coating, and after pyrolysis, amorphous thin films are obtained. The amorphous thin films are crystallized by heating to a high er temperature. Intermediate Pt x Pb and fluorite phases are commonly observed prior to the formati on of perovskite PZT thin films during crystallization. In addition depending on the different processing conditions during crystallization such as the atmospheric conditions, heating rates and different substrates, different texture can be obtained in PZ T thin films. However, the mechanism of texture control in PZT thin films is not well understood, because it is difficult to track the changes during fast crystallization from amorphous to perovskite PZT within a few seconds. Thus, in situ x ray diffractio n (XRD) measurements are useful to observe the phase and texture evolution of PZT thin films during crystallization.

PAGE 13

13 In this dissertation, in situ measurements of PZT thin films using laboratory and synchrotron based XRD were conducted to investigate the phase and texture evolution during crystallization. The stability of intermediate phases and perovskite PZT w as observed during crystallization in different atmospheric conditions. Based on the se observation s a new processing method was developed. Switch ing atmospheric conditions during crystallization of PZT thin films suppressed the intermetallic Pt x Pb phase and promoted the perovskite PZT phase. Further, b ased on the observation s of phase and texture evolution using different heating rates and substra tes during crystallization, a mechanism for texture control of PZT thin films is provid ed. 111 texture w as observed in PZT on 111 Pt electrodes. It is suggested that 111 Pt nucleates 111 PZT directly. Similar trends for the formation of 111 texture w ere ob served in PZT on an amorphized PbTiO 3 ( PTO ) seed layer, which implies that nucleation of 111 texture occurred on a 111 Pt electrode. Conversely, dominant 100 texture was observed in PZT on amorphized PTO during crystallization at slow heating rates. It is suggested that surface or homogeneous nucleation occurs at slow heating rates to promote 100 texture in Ti rich PZT at low temperature and 100 textured in Zr rich PZT is nucleated at higher temperature. Also, interdiffusion between the PTO layer and the P ZT layer during crystallization led to the formation of PZT with an inhomogeneous composition throughout the thin films. Unlike PZT on amorphized PTO, dominant 100 texture was observed in PZT on crystalline PTO regardless of heating rates during crystalliz ation. Also, the formation of the PZT thin films was observed after phase transformation of the PTO seed layer from tetragonal to cubic

PAGE 14

14 above the Curie temperature, which implies that formation of the 100 texture in PZT thin films was nucleated from the cu bic PTO seed layer with 100 texture.

PAGE 15

15 CHAPTER 1 INTRODUCTION In the late nineteenth century, the Curie brothers observed that the application of weight upon certain crystals, including quartz, resulted in the accumulation of char ge on the crystal surfaces. This phenomenon became known as the direct piezoelectric effect, and is defined as the linear coupling between the electrical and mechanical domains. Likewise, by the converse piezoelectric effect, a strain is induced as a resul t of an applied electric field. Piezoelectricity is only observed in crystalline dielectrics whose structure lacks a center of symmetry. Of the 32 possible point groups, based on crystallographic symmetry, 21 are non centrosymmetric and 20 are potentially piezoelectric [ 1 ]. The result of asymmetry in the structure, and therefore in potential, is a net displacement of the ions [ 2 ]. Piezoelectric materials can be characterized by several coefficients including: 1. d ij : Piezoelectric coefficient (C/N). Charge density (C/m 2 ) developed per given stress (N/m 2 ) due to the unique charge output coefficient of materials [ 3 ]. 2. K ij : Coupling coefficient. The ratio describes the conversion of mechanical energy to electrical energy. Piezoelectric coefficients with d ouble subscripts link mechanical and electrical quantities, where the first subscript gives the direction of the electric field associated with the applied voltage or charges generated. The second subscript indicates the mechanical stress or strain. The pi ezoelectric constant relating electric displacement produced by mechanical stress is termed by the d ij coefficient This is represented as

PAGE 16

16 (1 1) The d 33 coefficient denotes that force is applied along the 3rd axis and charges along the same axis ar e measured, while the d 31 coefficient indicates that the charges are m easured along the 3rd axis, but the direction of a pplied force is along 1st axis. Piezoelectric energy depends on different oper ation modes; compressive strain perpendicular to the elect rodes utilizes the d 33 mode, while a transverse strain parallel to electrodes exploit the d 31 mode. However, the power output from the d 33 mode is not a practical coupling mechanism for piezoelectric ity in the majority of application s [ 4 ]. Typically, in th e case of piezoelectric thin films, the piezoelectric elements are coupled in the transverse direction. Such an approach can provide mechanical amplification of the stress. Ferroelectricity of materials can be characterized by spontaneous polarization and the reversibility of this polarizatio n with an applied electrical fi e l d. [ 5 ] Also, phase transition s may occur from a high symmetry phase to a low symmetry ferroelectric phase. The temperature where this transition occurs is known as the Curie point and te mperature dependence of the dielectric constant above this temperature is explained by Curie Weiss law, (1 2 ) where r is the relative permittivity, 0 is the permittivity of vacuum, C is the Curie constant, and T 0 is the Curie temperature. Above the transition temperature, the crystal

PAGE 17

17 is in a paraelectric state. Conversely, the crystal is in the ferroelectric state bel ow the Curie temperature, which means a polar state with spontaneous polarization. In ferroelectric thin films, a reduction in permittivity is commonly observed due to several factors such as a remnant secondary phase or entrapped pores, grain size effect s clamping of the film by the substrat e. [ 6,7 ] A passive layer at the interface between the film and substrate also decreases the dielectric constant through a series capacitor arrangement with the thin films [ 8,9 ] In addition increased chemical disor der or formation of defects in thin films occurs during crystallization due to a comparatively lower crystallization temperature, which plays a role as local pinning center and decrease s the polar region response. [ 10 ] Pb(Zr x Ti 1 x )O 3 ( PZT ) is one of the mo st useful piezoelectric /ferroelectric materials due to its excellent dielectric, piezoelectric and ferroelectric properties. [ 11 18 ] The crystal structure of perovskite PZT is shown in Figure 1 1. The structure is usually depicted in its pseudocubic form. [ 19 ] The crystal structure (ABO 3 ) contains two cation sites in the crystal lattice. The larger cations (A) usually reside on the corners of the unit cell, while the smaller cations (B) are in the center of the unit cell. [ 1 9] The oxygen anions are located on face center positions. PZT is a solid solution between PbTiO 3 (PTO) and PbZrO 3 (PZO). T he phase diagram of PZT is presented in Figure 1 2. [ 3 ] The crystal structure of PZO is antiferroelectric rhombohedral while PTO is a ferroelectric tetragonal one W ith increasing PZO/PTO ratio in PZT, the tetragonality in PZT decreases until the crystal structure of PZT changes to a pure rhombohedral phase. The phase boundary between tetragonal and rhombohedral is called the morphotropic phase boundary (MPB). The

PAGE 18

18 PZO /PTO ratio for the MPB in PZT is known to be approximately 52/48. [ 3 ] Excellent piezoelectric and ferroelectric properties are observed in PZT at the MPB composition. [ 3 ] In particular, high dielectric constant and piezoelectric coefficient were observed i n PZT with the MPB composition. [ 3 ] These excellent properties are likely due to more polarization directions available in PZT at the MPB than those observed in PZT with a singular composition. [ 3 ] PZT in the form of film s have different properties, when compared to bulk PZT. Lower piezoelectric dielectric and ferroelectric properties are observed in PZT thin films [ 20 22 ] This is attributed to substrate clamping and residual stresses. [ 23 ] Clamped PZT thin films are not free to expand or contract in pla ne in response to an applied electric field which degrades the dielectric, ferroelectric, and electromechanical responses. Also, the crystallization of PZT thin films at high temperature can result in r esidual stresses The difference in thermal expansion between the films and the substrates generates residual stresses at room temperature. [ 24 25 ] 1. 1 Chemical Solution Deposition of PZT Thin Films The Chemical Solution D eposition (CSD) method is categorized based on different chemistry of precursors and dep osition methods. For ex ample, CSD methods include Metal Organic Decomposition (MOD) and sol gel methods. MOD precursors consist of metal atom s bonded to organic liquid via a bridging O, S, P or N atom which is insensitive to water The metal precursors in the liquid form are mixed with a solvent. The solution can be modified with different pH and viscosity. [ 26 ] The solution prepared using MOD precursors shows minimal aging because chemical precursors are insensitive to water. However, higher organic con tents in the MOD solution compared to

PAGE 19

19 sol gel based solutions could result in cracking of thin films during pyrolysis and crystallization steps. The sol gel process is a method used to produce inorganic oxides from chemical precursors. [ 1 9] A sol is a colloidal suspension of very small solid particles in a continuous liquid medium. A polymer is formed due to the interaction between particles within the sol for a sufficient time. When this polymer skeleton encloses the liquid, it is called a gel. [2 7 ] M etal alkoxides are commonly used as precursors in the sol gel method since they readily react with water. During solution preparation, alcohol exchange of metal alkoxide with solvent occurs which decreases the sensitivity of the metal alkoxides to hydrol ysis. S ince the m etal alkoxide precursors readily react with water, it is important to remove any trace water during preparation which ensure s little aging of the solution. The CSD method is also categorized based on different deposition methods suc h as sp in coating and aero sol deposition I n deposition of films using aerosol, chemical precursors are atomized to form an aerosol, and then deposited on a rotating substrate. Deposition can be enhanced by electrostatic charging of the aerosol. [ 28 ] The d esired t hickness can be obtained using different deposition time s and the number of deposited layers. However, i n the deposition of films using spin coating, the desired thi ckness can be obtained dependent on the number of deposited layers, spin speed and molar c oncentration of the chemical solution. S pin coating has been widely used for deposition of PZT thin films due to the cost effective, fast and simple method. Budd et al. developed the sol gel processing of PZT thin films demonstrating that PZT in the form o f thin films can have similar properties to bulk PZT, although these

PAGE 20

20 proper ties are usually diminished. [ 29 ] The sol gel method has been widely used to produce PZT thin films for various applications such as capacitors and m icroelectromechanical systems (MEMS) devices, and continues to be develope d for other applications. [ 30 32 ] The sol gel method consists of four steps: (1) precursor solution preparation, (2) deposition of the solution onto a substrate by spin coating, (3) pyrolysis of organic species at low temperature (300 C 500 C) to obtain amorphous films, and (4) crystallization of amorphous films to the desired perovskite phase PZT and densification at high temperature (600 C 700 C). [ 1 9] The CSD method is illustrated in Figure 1 3. In the cr ystallization step, texture control of PZT thin films is possible by utilizing different processing conditions. [ 33 ] In this chapter, details of each step in the sol gel method for PZT thin films are discussed and the crystallization behavior of PZT thin films with respect to different processing conditions provided in previous literature are reviewed 1.1 .1 Solution Preparation The final properties of PZT thin films depend on which route is used for PZT solution preparation. [ 34 ] 2 Methoxyethanol sol g el (2 MOE route) and inverted mixing order (IMO route) are mainly used to prepare the PZT solution. [ 1 9, 33 35 ] The 2 MOE route is based on the classical sol gel method and it was called 2 MOE route after the solvent, 2 Methoxyethanol. [ 35 36 ] For sol gel processing of PZT thin films, titanium and zirconium propoxides are used as metal alkoxide precursors for the B sites and lead (II) acetate is used for the A site to produce ABO 3 perovskite PZT. 2 MOE is used as a solvent for the process. B ecause the m etal alkoxides react with water readily, multiple refluxing and distillation are required to r emove any remnant water. [ 37 ] The experimental procedure for PZT solution preparation via the 2 MOE

PAGE 21

21 route is discussed in Chapter 3. The 2 MOE route has advantage s such as high controllability and reproducibility for PZT solution. Also, PZT solution prepared using the 2 MOE route has very little aging when no water exists in the solution. However, the 2 MOE route is regarded as a complex and time consuming process. Also, 2 MOE is referred to as a hazardous solution. The IMO route as a hybrid sol gel chelate method uses either alcohol exchange or chelation to decrease the sensitivity of th e metal alkoxide precursors. [ 33 38 ] The name of the IMO route came from the sequence of the addition of metal alkoxides during the process. In normal hybrid sol gel chelate route, the metal alkoxide for the A site is added first, and then the B site metal alkoxides are added, which is referred as sequenti al precursor addition (S PA). [ 39 ] However, in the IMO route, the B site precursor is added first, then the A site precursor is added. The experimental procedure for PZT solution preparation via the IMO route is discussed in Chapter 3. During solution preparation of PZT thin films via the IMO route, hydrolysis and chelation occurs. However, the main reaction is the chelation of alkoxides by acetic acid, which decreases the sen sitivity of metal alkoxides. [ 33 ] Also, esterification between the acetic acid and the alcohols in the solu tion occurs and thereby water is formed. The water in the solution allows continuous interaction between the precursors, which lea ds to aging of the solution. [ 40 ] 1.1 .2 Spin Coating and Pyrolysis When the PZT precursor solution is prepared, a few drops of the solution are deposited onto a cleaned substrate using a syringe with a 0.2 m filter. The amount of the solution depends on the substrate size and viscosity of the solution. A larger amount of the solution is required for a solution with higher visc osity and/or larger substrate to

PAGE 22

22 ensure full coverage. The filter is used to separate unreacted particles such as PbO from the solution, which may hinder homogeneous dispensing of the deposited solution on the substrate following the spin coating step. Aft er deposition of the solution, the substrate is spun at 2000 4000 rpm. Spinning the substrate serves to spreads the solution over the entire substrate. The final thickness of PZT thin films depends on spin speed and time. Commonly, thinner films are achi eved when a faster spin speed and longer time are used. As deposited coatings on the substrat e are amorphous. [ 41 ] After spin coating of the deposited solution onto the substrate, thin films are placed on a hot plate preheated at low temperatures (300 C 500 C) to decompose the organic species within the thin films. This is called pyrolysis Pyrolysis is defined as a thermochemical decomposition of organic material at elevated temperatures in the absence of oxygen. [ 1 9] During pyrolysis of the thin films the removal of organic species results in an amorphous PZT layer with a homogeneous distribution of cations on the substrate. However, pyrolysis of PZT thin films leads to a significant contraction of the films. This contraction causes a biaxial stress s tate that leads to t he formation of microcracks. [ 42 ] It is reported that thicker films are more susceptible to t he formation of microcracks. [ 42 ] Several strategies have been suggested to avoid the format ion of cracks in thin films. [ 38 ] One of the strate gies used for PZT thin films in this dissertation is the spin coating and pyrolysis of a thin layer several times until the desired thickness is achieved. Subsequently, the thin films are crystallized after several spin coating and pyrolysis steps. It is reported that the pyrolysis temperature is one of the important processing variables to influence the formation of intermediate phases and texture control of PZT

PAGE 23

23 thin f ilms during crystallization. [ 33 43 ] In previous literature, the formation of interme tallic Pt x Pb as a transient phase occurred during crystallization of PZT thin films, when the thin films were pyrolyzed at a low temperature (300 C). However, the formation of Pt x Pb was not observed during crystallization of PZT thin films when the thin fi lms were pyrolyzed at a higher temperature (400 C). [ 38 ] It was suggested that a small amount of organic species remained in amorphous PZT after pyrolysis at a low temperature, and further removal of the organic species was at the expense of oxygen within the thin films to lead to local reducing conditions at the Pt PZT interface, which promotes the formation of Pt x Pb during crysta llization of PZT thin films. [ 44 ] The effect of the pyrolysis temperature on the texture selection of PZT thin films during crys tallization is presented in Figure 1 2. [ 33 ] PZT thin films after pyrolysis below 420 C show dominant 111 texture after crystallization. Conversely, 100 texture is dominant in PZT thin films after crystallization, when PZT thin films are pyrolyzed above 42 0 C. Observations in previous literature suggest that heterogeneous nucleation of 111 texture of PZT occurs from the Pt electrode during crystallization (after pyrolysis at a low temperature), while homogeneous or surface nucleation of 100 texture of PZT o ccurs during crystallization (after pyro lysis at higher temperature). [ 33 ] 1.1 .3 Crystallization Amorphous PZT thin films after pyrolysis are heated to higher temperatures (600 C 750 C) at different heating rates for crystallization using a rapid thermal annealer (RTA) or a standard furnace or an infrared (IR) lamp. Under ideal conditions during crystallization, the amorphous layer transforms to the desired perovskite PZT phase. Since amorphous PZT is obtained after pyrolysis, nucleation and growth occurs during crystallization of PZT thin films. [ 1 9] Thus, the final microstructure of PZT thin

PAGE 24

24 films after crystallization may be governed by the nu cleation and growth process. [ 45 ] Due to the high surface to volume ratio of thin films, heterogeneous nucleatio n of PZT thin films is expected to be dominant at the subs trate thin film interface. [ 44 ] From standard nucleation and growth theory, the description of the energy barriers for homogeneous and heterogeneous nucleation are dependent on the driving force f or crystallization ( ). This may be represented as (1 3 ) (1 4 ) where is the interfacial energy and is a function related to the contact angle ( The function for a hemispherical nucleus can be described by (1 5 ) According to equations 1 1 and 1 2, the difference in the energy barrier required for nucleation events such as the interface, surface and bulk nucleation is defined by the surface energy ( driving force for crystallization ( ) and the contact angle ( with the substrate. [ 1 9] The function results in a lower energy barrier in equation 1 2, which means that heterogeneous nucleation is always preferred over homogeneous nucleation. However, homogeneous nucleation is as probable as heterogeneous nucleation with an increasing driving force for crystallization ( ). [ 1 9] Standard nucleation and growth theory implies that crystallization of PZT thin films at higher temperatures results in a lower driving force for crystallization ( ) due to and thereby heterogeneous nucleation is dominant during the crystallization of PZT. Thus, the crystall ization of PZT thin films is commonly carried out using RTA, which enables

PAGE 25

25 the crystallization of thin films at higher temperatures with fast heating rates. Conversely, when using a conventional furnace, the crystallization process begins during the heatin g to the final crystallization temperature. This can lead to more than one nucleation event duri ng crystallization. [ 38 ] It is attributed to the availability of sufficient energy to surmount the energy barriers of several nucleation events such as increasi ng te mperature of the thin films. [ 38 ] 1. 2 Phase Evolution When amorphous PZT transforms to crystalline PZT during crystallization, different intermediate phases such as Pt x Pb and fluorite/pyrochlore have been observed. [ 38 44 ] During the early stages of crystallization, a small amount of organic species remaining after pyrolysis are volatilized from the amorphous PZT film The volatilization of organic species is accompanied by the removal of oxygen in the thin films, which promotes local reducing condit i ons at the Pt PZT interface. [ 46 ] The reducing conditions lead to the reduction of Pb from the amorphous PZT thin films, and then the difference of Pb concentration between Pt and PZT enables the reduced Pb cations to diffuse into the Pt substrate. [ 44 ] T his diffusion can promote a reaction between Pt and Pb to form a transient intermetallic layer, Pt x Pb. [ 44 ] The volume of Pt x Pb phase depends on the selected heating rate during crystallization. [ 44 ] A decrease in the volume of the Pt x Pb phase was observed with a decrease in the heating rate during crystallization. [ 44 ] It is suggested that sufficient oxygen from the atmosphere diffuses into the Pt PZT interface at slow heating rates to decrease the extent of the reducing conditions resulting from the decom position of organic species [ 44 ]

PAGE 26

26 With the decomposition of the Pt x Pb phase, the formation of intermediate phases was commonly observed prior to the formation of perovskite PZT. [ 38 44 47 48 49 ] The formation of intermediate phases is attributed to limited long range diffusion during crystallization at com paratively low temperatures. [ 42 ] This limited long range diffusion hinders the formation of thermodynamically stable perovskite PZT until sufficiently hi gh temperatures are reached. [ 42 ] The intermediate phase was observed to be pyrochlore type structure of the form Pb 2 Ti 2 O 7 y where 0 y 1 which exhibits a wider range of stoichiometry than Pb(Zr x ,Ti 1 x )O 3 perovskite phase [ 50 ] However, there are ongoing debates on the crystal s tructure of the pyrochlore type intermediate phase during crystallization. Kwok et al. suggested that this intermediate phase is similar to the Pb 2 Ti 2 O 7 pyrochlore phase. [ 50 ] However, Wilkinson et al. and Lakeman et al. suggested this intermediate phase t o have the fluorite structure, Pb 2 Ti 2 O 7 x [ 51 52 ] the pyrochlore ( Fd3m ) and fluorite ( Fm3m ) structures are both closely related to the pyrochlore crystal structure described as a doubled (i.e., 2 a c x 2 a c x 2 a c ) fluorite structure through ordered cation si tes and anion vacancies in the 8a Wyckoff positions. [ 53 ] It is suggested that the stability of the fluorite/pyrochlore phase and final perovskite PZT depends on the oxidation state o f Pb during crystallization. [ 54 ] The oxidation state of Pb is closely re lated to the oxygen parti al pressure and temperature. [ 54 ] Based on the Richardson Ellingham diagram which considers the Gibbs free energy of the oxidation state of Pb with increasing temperature (Figure 1 5), the oxidation state of Pb can change from 4+ to 2+ above 550 C which is the temperature range of phase transformation from fluorite to perovskite PZT in previous literature. [ 55 ]

PAGE 27

27 Therefore, the oxidation state of Pb as 2+ can be stabilized above 550C during crystallization which maintain s the stoi chiometry of the perovskite, while the oxidation state of 4+ below 550 C can favor the fluorite phase. It implies that the intermediate pyrochlore/fluorite phase can be stabilized at a higher temperature when the oxidation state of Pb is stabilized as 4+ w ith higher oxygen partial pressure. The stability of the fluorite/pyrochlore phase and perovskite PZT also depends on Pb volatility during crystallization. Any deviation of Pb content from the ideal perovskite stoichiometry (Pb( Zr x ,Ti 1 x )O 3 ) intentionally or inadvertently due to volatilization during processing tends to stabilize the pyrochlore /fluorite phase. [ 33 56 ] During typical CSD PZT thin film preparation, the starting solutions are batched with excess Pb precursor to accommodate for any Pb loss dur ing processing of these types of films which ensures the formation of PZT without any remnant intermediate phase after crystallization [ 1 9 33 ] 1. 3 Texture E volution The term texture has several different and important meanings Surface texture, also kn own as surface finish, involves the characteri stics of a surface and has several components such as lay, surface roughness and waviness. [57] In engineering, surface texture is a measure of surface irregularities. There is also morphological texture. It is defined as the preferred element shape in a material, including fibers in a composite and anisometric grains. However, in materials science, crystallographic texture is defined as the distribution of crystallographic orientation s of polycrystalline materi als. [58] For example, when the orientations of different crystallites are distributed randomly, it is said that the material does not have texture. Conversely, when there is preferred lattice orientation of crystallites within a material, it is said that there is texture. The

PAGE 28

28 degree of texture depends on the percentage of pref erred orientation. Commonly, a degree of texture can be developed during thin film growth. Properties of materials highly depend on the different crystallographic texture. This disser tation is concerned wi th crystallographic texture only and thus is simply referred to as texture. Different piezoelectric and ferroelectric properties with respect to different texture of PZT th in films have been observed. [ 33 59 60 ] 111 textured PZT thi n films exhibit a lower coercive field ( ) than 100 and 110 textured PZT thin films while the highest piezoelectric coefficient can be achieved in 100 textured PZT thin films. [ 61 ] Utilizing the advantage of texture in PZT is important for var ious applications. For example, 111 texture of PZT thin films is desirable for capacitor application, because switching of the state of polarity is possible with lower power and fast speed [ 59 ] 100 texture in PZT thin films is preferred for sensor and MEM S devices, which requires excelle nt piezoelectric properties. [ 60 ] A number of efforts to control texture of PZT thin films have been carried out by altering different processing conditions such as pyrolysis temperature, different substrates, atmospheric co nditions, and heating rates. [ 33 43 46 54 62 ] For example, low pyrolysis temperature and fast heating rates for crystallization results in 111 texture of PZT thin films on a 111 Pt elect rode. [ 54 63 ] A high pyrolysis temperature and slow heating rates fo r crystallization leads to 100 texture in PZT thin films. [ 62 ] Unlike PZT thin films on Pt electrodes, fast heating rates result in 100 texture of PZT thin films on a 100 textured PTO seed layer. [ 64 ] These observations in the previous literature suggests that heterogeneous nucleation events are dominant in PZT thin films during crystallization at fast heating rates, while high pyrolysis temperature and slow heating

PAGE 29

29 rates lead to dominant homogeneous nucleation events during crystallization. Researchers are in good agreement concerning the experimental observations in the resulting texture of PZT thin films depending on different processing conditions. [ 33 43 46 54 62 ] However, there is disagreement about the mechanisms for texture control and their dependen ce on different processing c onditions. [ 43 46 63 65 66 67 ] 111 textured Pt electrodes have been widely used to promote 111 texture in PZT thin films based on the hypothesis that similar lattice parameters between Pt and PZT decrease the activation energy f or the formation of PZT on a Pt electrode v ia heterogeneous nucleation. [ 38 44 ] However, there are different mechanisms suggested for texture control of PZT thin films on a 111 Pt electrode. Based on several similarities that both the Pt electrode and Pt x P b exhibit, such as a face centered cubic (FCC) crystal structure and similar lattice parameters, the formation of 111 Pt x Pb was suggested to have an epitaxial relationship with the 111 Pt electrode. [ 44 68 ] The lattice parameter of the Pt x Pb (a 0 ~ 4.05 ) is closer to the lattice parameter of PZT (a 0 ~ 4.07 ) than Pt (a 0 ~ 3.92 ). [ 67 ] It was suggested that intermetallic 111 Pt x Pb nucleates 111 PZT during crystalliz ation at fast heating rates. [ 69 70 ] In addition, the formation of a PbO layer through d ecomposition of the intermetallic Pt x Pb phase was suggested to nucleate 100 texture in PZT thin films during crystalliz ation at slow heating rates. [ 71 72 ] The stability of intermediate pyrochlore/fluorite phase after pyrolysis at different temperature ha s been suggested to contro l texture of PZT thin films. [ 43 73 ] The low stability of the pyrochlore/fluorite phase formed after pyrolysis at low temperature was suggested to allow the growth of 111 textured PZT on a 111 Pt electrode. [ 74 ] Conversely, more s table pyrochlore/fluorite phase formed after pyrolysis at higher

PAGE 30

30 temperatures was suggested to lead the formation of 100 texture due to dominant growth with the lowest surface energy of the 100 plane in perovskite PZT. [ 74 ] Norga et al. and Voight et al. s uggested that 111 texture of the pyrochlore/fluorite phase is seeded at the Pt PZT thin film interface, and then the phase transformation from fluorite/pyrochlore to perovskite PZT occurs while maintaining the orientation relationship with the substrate. [ 75 76 ] The diffusion of Ti from the adhesion layer of a platinized Si substrate to PZT thin films has been commonly observed during crystallization. [64 66] The diffusion of Ti was suggested to control texture of PZT thin films during crystallization. Tani et al. suggested that Pt 3 Ti formed on the Pt electrode due to the diffusion of Ti from the substrate nucleate 111 texture of PZT thin films. [ 77 78 ] Muralt et al. suggested that Ti diffusion from adhesion layer leads to the formation of rutile TiO 2 on Pt electrode. [ 66 ] When the formation of the TiO 2 phase occurs, TiO 2 is nucleated heteroepitaxially from the Pt electrode and the TiO 2 transforms into the 111 perovskite nuclei. [ 80 ]

PAGE 31

31 Figure 1 1 Crystal structure of the ABO 3 perovskite PZT.

PAGE 32

32 Figure 1 2 Equil ibrium phase diagram of PZT. [ 3 ]

PAGE 33

33 Figure 1 3 CSD of PZT thin films on a platinized substrate.

PAGE 34

34 Figure 1 4 Effect of pyrolysis temperature on the orientation of PZT 40/60 films after identical crystall ization at 700 C and proposed mechanism(s) for the texture development. [ 33 ]

PAGE 35

35 Figure 1 5 Richardson Ellingham diagram of PbO and PbO 2

PAGE 36

36 CHAPTER 2 MOTIVATION AND OUTLINE Phase and texture of PZT thin films is dependent on different processing conditions during crystallization such as atmospheric conditions, heating rates and substrates. The trends of the resultant phases and texture of PZT thin films with respect to processing variables is consistent in previous reports. However, th ere is no consensus on the mechanism of phase and texture formation in PZT thin films due to a lack of understanding of different chemical and physical phenomena during crystallization. In order to investigate the mechanism of phase and texture formation measurements of phase and texture evolution with fast acquisition times during crystallization are required, and therefore two types of in situ XRD technique are introduced. Different nucleation mechanisms of texture in PZT thin films and their relations hip to different processing conditions are proposed. Also, different hypotheses on the formation of phase and texture in PZT thin films suggested by different researchers are examined based on the data obtained from in situ XRD measurements. This dissertat io n focuses on the effect of different processing variables such as atmospheric conditions, heating rates and the different substrates on phase and texture evolution of PZT thin films during crystallization. Details of the sample preparation using the CSD process and the methodology for in situ XRD data analysis are introduced in Chapters 3 and 4. A novel processing method for PZT thin films is described in Chapter 5. In situ XRD measurements during crystallization are used to examine interfacial interactio ns during crystallization in selected atmospheres. In Chapter 6, the stability of the PTO seed layer during deposition of the upper PZT layer is discussed. The effect of

PAGE 37

37 different substrates such as Pt and amorphous and crystalline PTO seed layers on PZT thin films at different heating rates is discussed in Chapters 7, 8 and 9. In Chapter 10, phase and texture evolution as a result of different processing conditions is summarized.

PAGE 38

38 CHAPTER 3 DETAILS OF GENERAL SAMPLE PREPARATION 3.1 IMO PZT Solution Prep aration and Deposition Various thin film samples were prepared using 2 different processing routes ; IMO and 2 MOE. [ 33 35 36 38 ] PZT solutions were spin coated on different substrates and pyrolyzed. This chapter provides the details of sample preparation u s ed in this work. Thin film samples of 0.35 M solution derived Pb(Zr 0.52 Ti 0.48 )O 3 with 20% Pb excess were prepared using the inverted mixing order (IMO) method detailed by Schwartz et al [ 1 9] 20% e xcess Pb was added to compensate for the loss of PbO du e to surface Pb loss and interface diffusion during crystallization. Also, it ha s been reported that PZT with 20% excess Pb showed fully crystallized perovskite PZT. [ 81 ] In order to prepare the IMO PZT solutions, zirconium butoxide (Acros Organics Ltd.) a nd titanium isopropoxide (Acros Organics Ltd.) were mixed using a stir bar for 5 min. Acetic acid and methanol were added and mixed for 5 min sequentially. Lead (IV) acetate (Acros Organics Ltd.) was added to the solution after dissolution. The solution wa s stirred and heated at 90C until the lead acetate dissolved fully. The c olor of the solution turn ed from a pale yellow to clear after the dissolution of lead (IV) acetate. In order to obtain 0.35 M PZT solutions, additional methanol and acetic acid were added to the solution twice more after dissolution of lead (IV) acetate. The solution was mixed using a stir bar for 2 min between each dilution step. The f low chart f or PZT solution preparation is listed in Figure 3 1. The PZT solution was dispensed usin g a syringe with a 0.45 m filter. The d ispensed PZT solution was spin cast for 30 s at 3000 rpm on square platinized silicon substrates layered as follows: Pt (170 nm)/ Ti (40 nm)/ SiO 2 (400 nm)/Si (SQI, Santa

PAGE 39

39 Clara, CA). The samples were then pyrolyzed a t 300C on a hotplate for 1 min. The spin coating and pyrolysis sequence was repeated twice more, for a total of three pyrolysis treatments per sample to achieve a thickness of 250 nm. The temperature of the hotplate was monitored using an external IR lase r digital thermometer (HDE B01, HDE, USA). 3.2 2 MOE PTO and PZT Solution Preparation and Deposition 3.2.1 Standard Processing PZT solutions with a 5% Pb excess were prepared using a procedure modified from Budd et al. [ 29 ] Lead trihydrate and 2 MOE solven t were mixed in 1 L flask inside a controlled atmosphere glove box to make lead acetate solution. Vacuum distillation was conducted on the flask using a rotary evaporator (Heidolph Laborata 4000) outside of the glove box after the flask was clamped shut. T he flask with lead acetate solution was located on 120C preheated silicon oil bath and rotated at 120 rpm in a positive pressure of nitrogen gas. Nitrogen gas was turned off and a vacuum was applied into the flask. The flask was rotated for 20 min with a 410 mbar vacuum. During rotation of the flask on the silicon oil bath, water vapor and 2 MOE solvent was evaporated and then condensed into cooling coils. The condensed drops were collected in a collection flask under the cooling coils. When the lead aceta te solution turn ed white foam under rotation, the velocity of rotation was decreased to 35 rpm and the flask was raised out of the silicon oil bath. Nitrogen gas was applied into the flask to remove vacuum. The flask was removed from the evaporator and sub merged in water until it reached room temperature. Prior to the vacuum distillation of lead acetate solution, Ti iso propoxide and 2 MOE were poured into another 125 mL flask sequentially and sealed inside the glove box. The Ti and Zr precursors in 2 MOE w ere mixed using stir bar.

PAGE 40

40 An additional 2 MOE was poured into the Ti/Zr solution and then poured into lead acetate solution inside glove box. This step was repeated one more time. The flask with PZT solution was sealed and removed from the glove box. The flask was moved to the evaporator for vacuum distillation. The flask was placed in a silicon oil bath preheated at 120C and rotated at 120 rpm for 3.5 h. A vacuum was applied into the flask for 5 min after 3.5 h and then nitrogen gas was applied for 5 min Formamide (4 volume %) was added into the PZT solution as a drying agent. The f low chart of 2 MOE derived P ZT solution preparation is listed in Figure 3 2. Based on the work of Muralt et al. [ 66 ], a 30% Pb excess PTO was deposited on the Pt electrode vi a the CSD method. Lead acetate trihydrate was mixed with 2 MOE into the 1 L flask. In another flask, Ti isopropoxide with 2 MOE was mixed using stir bar while vacuum distillation was performed on the lead acetate solution using a 4000 rotary evaporator (He idolph Laborata 4000). After the vacuum distillation of the lead acetate solution, the Ti iso propoxide solution was added to the lead acetate solution in the glove box. Additional vacuum distillation was conducted, similar to the process used to PZT solut ion. The f low chart of 2 MOE derived PTO solution preparation is listed in Figure 3 3. The PTO solution was spin coated for 45 s at a speed of 3000 rpm and then pyrolyzed at 350C for 2 min on a hot plate. Crystallization of pyrolyzed PTO was performed at 700C for 60 s using a RTA. This PTO layer was used as a seed layer for 2 MOE PZT thin films. PZT solution was spin coat on the PTO seed layer at a speed of 2000 rpm for 45 s and then pyrolyzed at 350C for 2 min on a hot plate. The spin coating and pyroly sis sequence was repeated twice, for a total of three pyrolysis

PAGE 41

41 treatments per sample. The thickness of the amorphous PZT layer on top of the PTO seed layer was 2 9 0 nm evidenced by the micrograph in Figure 6 4. 3.2.2 Processing of Ceria Dispersed PZT T hin F ilms on PTO Seed Layer Ceria (NIST, SRM674b) powders were used as a temperature standard to measure the actual temperature of 2 MOE derived solution deposited PZT thin films. In order to prepare Ceria dispersed PZT thin films, ceria powders were mixed wi th the prepared 2 MOE PZT solutions. The solutions were dispensed onto the PTO seed layer at a speed of 2500 rpm for 45 s and then pyrolyzed at 350C for 2 min on a hot plate. In order to prevent filtering dispersed ceria powders, a syringe filter was not used for spin coating of 2 MOE PZT solutions onto the PTO seed layer. The spin coating and pyrolysis of the PZT solution were repeated twice, for a total of three pyrolysis treatments. 3.2.3 Pt/TiO x /SiO 2 /Si Substrate The substrates were 4 inch diameter 1 00 Si wafers coated with thermally oxidized 500 nm thick SiO 2 thin films. A 0001 textured Ti layer was sputter deposited on thermally grown SiO 2 on Si wafer and annealed at 750C in an O 2 atmosphere in order to convert this layer into rutile TiO 2 with 100 texture. The 100 textured TiO 2 layer was used as a template to grow 111 textured Pt. A 100 nm thick Pt layer was sputter deposited onto the TiO 2 The process is detailed in Potrepka et al [ 82 ] The r oughness of the Pt electrode was observed using a profilo meter (Veeco Dektek 150). T he average roughness of the Pt substrate was 0.7 nm after deposition of the Pt electrode as shown in Figure 3 4 Also, a similar roughness value of the Pt electrode, 0.83 nm, was observed after s ubsequent heat treatment at 350C for 2 min three times, which is the

PAGE 42

42 same condition for pyrolysis of PZT thin films. It implies that pyrolysis of PZT thin films might not influence the roughness of the Pt electrode.

PAGE 43

43 Figure 3 1 Preparation of IMO PZT solution.

PAGE 44

44 Figure 3 2 Pr eparation of 2 MOE PZT solution.

PAGE 45

45 Figure 3 3 Preparation of 2 MOE PTO solution.

PAGE 46

46 Figure 3 4 Micrograph and roughness of the Pt electrode before and after pyrolysis.

PAGE 47

47 CHAPTER 4 GENERAL EXPERIMENTAL SETUP AND DATA EXTRACTION METHO DOLOGIES This chapter introduces the in situ XRD techniques that are utilized to investigate the phase and texture evolution of PZT thin films during crystallization. These techniques include both those utilizing laboratory and synchrotron X ray sources. Also, the methods used for XRD data analysis are discussed, including those used for phase identification and texture quantification of PZT thin films. 4.1 In situ Laboratory Based X ray Diffraction Setup and Geometry Diffraction patterns were recorded us ing a laboratory X ray diffractometer with a curved position sensitive detector (CPS 120, Inel, Artenay, France). Pyrolyzed PZT thin films were crystallized in a furnace attached to the diffractometer during heating. The Cu X ray wavelength wa The CPS detector is capable of measuring diffraction patterns over a wide range, which allows fast XRD pattern acquisition times. The incident X ray beam irradiates the sample through a Kapton sleeve on the furnace attached to the diffractom eter. Samples were loaded into an alumina crucible sample stage at a 10 angle to the incident X ray beam. Nitrogen, oxygen and air gases can be flowed through the furnace which is controlled by a flow meter. The volume of the furnace is approximately 500 cm 3 The experimental setu p used is shown in Figure 4 1 A The poles of planes contributing to the diffraction intensities are perpendicular to the sample surface in Bragg Brentano geometry ( /2 geometry). However, an asymmetric /2 geometry is used in the Inel CPS 120 diffractometer. This means that the direction of poles contributing to the diffraction intensities varies with the pole being measure d This type of diffractometer is benefici al for the observation of phase

PAGE 48

48 evolution of PZT thin films with increasing temperature because patterns can be acquired in a short time. However, different diffraction intensities observed with varying poles make it difficult to compare texture in thin fi lms quantitatively. 4.2 Synchrotron X ray Diffraction Setup and Geometry In situ XRD experiments using synchrotron X ray diffraction were conducted at the APS beamline 6 ID B. The experimental setup and geo metry is shown in Figures 4 1 B and C respecti vely. The X ray energy was 22.7 keV during the experiments, which ray beam on the thin films was controlled by changing the inclination of the samples relative to the beam. A l ow incidence angle (~ 0.5) of the X ray beam was used to maximize the interaction volume in thin films. The beam size was 1.5 mm x 149.9 m and the inclination angle was 2. The irradiated area on the surface of the thin films was 4.3 mm2. A 2D detector was used to record XRD patterns. The detector frame rates of the 2D detector were from 0.25 s 1 to 4 s 1 The 2D detector is capable of measuring diffraction intensities in the entire Debye Scherrer rings. [ 38 83 ] Therefore, phase and texture information f rom the thin films can be recorded in a single diffraction image, which allows quantitative and qualitative analysis of thin films. An IR lamp was located above the sample stage and a thermocouple was located on the sample stage. The heating rates from the IR lamp were controlled using different voltage rates (V/s). In the experimental setup, the XRD geometry is different from Bragg Brentano geometry ( /2 geometry), which is widely used for characterization of phase and texture information in thin films. In the XRD geometry used for in situ measurements, the direction normal to thin films ( n ) is not parallel to the scattering vector of diffracting plan es ( Q ) due to the incident angle between the sample and the incident X ray beam,

PAGE 49

49 as shown in Figure 4 1 C while n is parallel to Q in /2 geometry. Thus, the difference in the orientation of crystallites corresponding to diffraction intensities observed in two different XRD techniques can lead to dif ferent relative intensities. [ 38 ] 4.3 Data Reduction and Representation Representative XRD patterns from the diffractometer on beamline 6 ID B are illustrated in Figure 4 2 A XRD patterns from the 2D detec tor are referenced by azimuthal angle ( ) and radial distance ( ). The radial distance ( ) from beam center on the 2D detector is directly related to 2 angle and sample to detector distance ( D ) based on the following relation: (4 1) The azimuthal angle ( ) of 0 is in the vertical direction, and = 90 and 90 are towards the left and right direction of t he image, respectively. When extracting data from the 2D detector, the software Fit2D (Andy Hammersley, version 2004) is commonly used. This is different from the azimuthal angle ( 90 90) above, but is consistent with the geometry where in the film normal is most parallel to the angle defined as 0. The 2 and positions in a 2D detector image contribute to the diffraction intensity from the crystallites. [ 38 ] The diffraction intensities are distributed across the Debye Scher rer ring in nearly uniform (for polycrystalline materials) and/or highly non uniform ways (for textured materials). [ 38 ] Representative Debye Scherrer ri ngs illust rated in Figure 4 2 A provide an example of diffraction patterns of polycrystalline material s recorded using a 2D detector. Data reduction of the diffraction patterns recorded using the 2D detector for phase and texture analysis was conducted with different radial distance ( r ) and

PAGE 50

50 azimuthal angles ( ). [ 83 ] Prior to data reduction, geometric factors of the synchrotron XRD such as the distance from the sample to the 2D detector, the tilt angle of the detector, the rotation angle of tilting planes and beam center were refined from Debye Scherrer rings of standard CeO 2 dispersed in PZT thin films using the software Fit2D as shown in Figure 4 2 B [ 84 ] For phase identification of PZT thin films, the diffraction intensities were integrated over a selected range. For example, a section of Debye Scherrer rings integrated over the range 5 5 are shown in a representative 2D detector image in Figure 4 3 A The integration of intensities over a selected range can be represented as (4 2) In the selected section of the Debye Scherrer rings in Figure 4 3 A diffraction intensities integrated over the range are shown as the repr esentative plot in Figure 4 3 B In order to investigate texture information of PZT thin films, integration over a limited 2 range corresponding to different ( hkl ) planes was performed across a range. [ 27 ] The integration can be represented as (4 3) The intensity distribution over a range for a given ( hkl ) plane shows specific t exture components for a given ( hkl ) plane. Throughout the remainder of this dissertation, preferred orientation of an hkl pole will be referred to as the texture component hkl For example, given in Figure 4 4 A diffraction intensities from the 2 range

PAGE 51

51 selected across the range range at = 0, the {100} reflection is observed at = 45 and 45. In same manner, the {111} reflection can be detected at = 35.2 and 35.2. The angular relationship between different planes for a given ( hkl ) plane was obtained using a modeled (110) pole figure as shown in Figure 4 4 B [ 33 ,71 ] The angular relation ( ) between two planes can be represented as (4 4) A representative plot of the intensity distribution of the {110} refl ection is shown in Figure 4 4 C Peak fitting of the individual extracted refle ctions was performed using the Pearson VII function in the Matlab tool box. The curve fitting tool box in Matlab uses the least squares method of fitting data. [ 85 ] Fitting requires a parametric model that relates the measured data to the predicted model with various peak shape coefficients. [ 85 ] The result of the fitting process is an estimate of the model coefficients. In order to extract the coefficient estimates, the least squares method minimizes the summed squares of residuals. The residual ( ) is defined as the difference between the observed data ( ) and the fitted model ( ), and is represented as (4 5 ) The residual ( ) is identified as the difference between the data and the m odel. The summed residual square ( S ) is represented as (4 6 ) The Pearson VII function was used as a shape function to fit the reflections obtained from XRD, which provides a more accurate fit for the broad tailed curves than

PAGE 52

52 a Gauss ian curve or Lorentzian curve in the current study. The Pearson VII function, in general form, is defined as (4 7 ) where F is a measure of the peak width, and M is the tail decay rate. [ 86 ] Also, x is the peak position and x 0 is the id eal peak position. For example, representative results of the peak fitting on different texture components of PZT thin films over the azimuthal angle ( ) are shown in Figu re 4 4 D Using the least squares method and the Pearson VII function, f itted coefficients such as peak position ( ), shape parameter ( ), peak width ( ), and integrated intensity of each texture component were extracted. C onfidence bounds for the fitted coefficients of reflections w ere calc ulated using the curve fitting tool box in M atlab and t he coefficient confidence bounds are presented numerically Confidence bounds define the lower and upper values of the associated interval, and the width of the interval. [ 85 ] The width of the interva l indicates the uncertainty of the fitted coefficients. For example a wide interval for the fitted coefficients implies that more data are required when fitting before determining that the coefficients are accurate. That is, a few data points for the fitt ing procedure may lead to wider interval bounds for fitted coefficients. The confidence bounds are defined with a degree of certainty. In the current study 95% confidence bounds w ere chosen for peakfitting, which indicates that fitted coefficients have 9 5% chance that the new observation is actually contained within the lower and upper prediction bounds. The confidence bounds for fitted coefficients are given by

PAGE 53

53 (4 8 ) where b is the fitted coefficients, t depends on the confidence l evel (95% in current study) and Z is a vector of the diagonal elements from the estimated covariance matrix of the coefficient estimates, ( X T X ) 1 z 2 In a linear fit, X is the design matrix, while for a nonlinear fit X is the Jacobian of the fitted values with respect to the coefficients. X T is the transpose of X and z 2 is the mean squared error. [ 85 ] Representative fitted coefficients with confidence bounds obtained from different texture components are shown in Figure 4 4 D For example, in the peak of 111 texture component, the fitted value for the intensity is 3.32 x 10 5 counts The lower and upper bound s are 3.18 x 10 5 and 3.46 x 10 5 respectively. T he interval width between the low er bound and upper bound is 0.14 x 10 5 By default, the confidence lev el for the bounds is 95%. This indicates that there is 95% chance to observe the fitted intensity containing the lower and upper prediction bounds. The integrated intensities obtained from peak fitting were used to investigate the volume fraction of text ure components. [ 44 87 ] For the volume ( ) of a given { hkl } texture component, the intensity of the { hkl } texture component was first integrated with respect to the through the following equation: (4 9 ) The v olume fraction ( ) of a given texture component ( hkl ) was calculated using a formula that includes both the integrated intensity (Eq uation 4 9 ) and the correction factor ( ) of given { hkl } planes through the following equation: (4 10 )

PAGE 54

54 A correction factor ( ) was included to correct for the difference in the number of planes diffracting at each angle. Specifically, the intensities of the 100 texture components at = 45 and 45 observed in Figure 4 4 D are the result of diffraction from four {110} planes oriented 45 from the surface normal. However, there are two more {110} planes oriented 90 from the surface normal which results in the diffraction intensities of the 100 text ure components at = 90 and 90. Thus, for a film with only a 100 texture components, diffraction intensities would be present at = 90, 45, 45 and 90 with the multiplicity of peaks at 45 and 45 being double those at 9 0 and 90.[ 38 ] In the same manner, the intensities of the 111 texture components at = 35.3 and 35.3 observed in Figure 4 4 D are the result of diffraction from three {110} planes oriented 45 from the surface normal. However, there are t hree more {110} planes oriented 90 from the surface normal which results in the diffraction intensities of 111 texture components at = 90 and 90. Correction factors of 100 and 111 texture components are summarized in Figure 4 5 Another form of data presentation involved observation of the raw data with no polar plot to a rectangular contour plot to show texture information. This type of presentation is referre d to in this work as an intensity map For example, the section of the Debye Scherrer ring corresponding to the {110} reflection at the 2 over the range ( in Figure 4 6 Data reduction based on the methodology described above was performed for each of the 2D detector im ages collected during crystallization of PZT thin films. The

PAGE 55

55 reduced data was then stacked with respect to acquisition time to create a contour plot. Representative plots for phase and texture evolution with acquisition time are shown in Figure 4 7 Each p lot is called a pha se evolution plot (Figure 4 7 A ) and texture evolution plot (Figure 4 7 B ), respectively. 4.4 Temperature Calibration PZT thin films with an internal CeO 2 calibrant were heated using the IR lamp with the same voltage rate as that used for the in situ XRD measurements of samples without CeO 2 Voltage rates used for in situ XRD measurements were 0.005 V/s, 0.01 V/s, 0.1 V/s, 1 V/s, and instantaneous ramping. During heating, the 2 position of the (111) reflection of CeO 2 shifts to lower 2 due to thermal expansion. The 2 position shift of the (111) in CeO 2 as a function of time was extracted through peak fitting, and then converted to change of lattice parameter as a function of time. The expected change in the lattice parameter was calcu lated and compared with the experimental change in the lattice parameter as a function of temperature to extract the actual temperature as shown in Figure 4 8 [ 88 ] The temperature profile s obtained from this temperature calibration w ere used for phase an d texture evolution plots which are provided in Figure 4 9. Different heating rates for different voltage rates used for the in situ XRD measurements are summarized in Figure 4 10

PAGE 56

56 Figure 4 1 In situ XRD setup. A ) Laboratory based XRD setup used for in situ crystallization of PZT thin films in different atmospheres B ) Experiment al XRD setup at beamline 6 ID B C ) XRD geometry of synchrotron XRD at 6 ID B.

PAGE 57

57 Figure 4 2 2D detector image s A ) Represen tative 2D detector image B ) Calibration of g eometric factors of XRD using Fit2D.

PAGE 58

58 Figure 4 3 Intensity vs. 2 line plot. A ) Integrated intensities measured by the 2D detector at selected r range through ( B ) Representative intensity vs. 2 line plot.

PAGE 59

59 Figure 4 4 Quantitative analysis of texture components. A ) Azimuth al angle of different texture components with respect to the 110 pole figure of PZT thin films in a r epresentative 2D detector image B ) Angular relation of different texture com ponents for the 110 pole figure C ) Integrated intensities of {110} diffractio n intensities at 2 ra nge between 90 and 90 D ) Quantitative analysis of different texture components in intensity vs. line plot.

PAGE 60

60 Figure 4 5 Summary of correction factors for 100 and 111 texture comp onents.

PAGE 61

61 Figure 4 6 Intensity map of PZT thin films A 11.5 over range (between 90 and 90 ) in a 2D detector image B ) Representative {110} intensity map of a PZT thin film.

PAGE 62

62 Figure 4 7 Representative evolution plots. A ) Phase evolution plot B ) Texture evolution plot.

PAGE 63

63 Figure 4 8 Temperature calibration using CeO 2 standard powder. A) Shift of 2 peak in (111) CeO 2 vs. time B ) Change of d spacing of (111) CeO 2 vs. time C ) Lattice parameter change of CeO 2 vs. time D ) Temperature calibration using the change in lattice parameter as a function of time and temperature.

PAGE 64

64 Figure 4 9 Temperature profiles during crystallization of PZT thin films at diff erent heating rates.

PAGE 65

65 Figure 4 10 Heating rates at 300C and 600C with respect to different voltage rates.

PAGE 66

66 CHAPTER 5 EFFECT OF CONSTANT AND SWITCHING ATMOSPHERIC CONDITIONS DURING CRYSTALLIZATION 5.1 Lite rature R eview It has been repor ted that organic compounds in the amorphous PZT film are volatilized, and the reduced Pb cations and excess Pb can diffuse into the platinum substrate during crystallization [ 87 ] This diffusion can enable a reaction between Pt and Pb to form a transient i ntermetallic layer, Pt 3 Pb. [ 87 ] This layer disappears upon full crystallization of the films. Although Pb from Pt 3 Pb appears to be re oxidized and participate in crystallization at higher temperatures, some Pb may remain in the Pt electrode, which has the potential to create compositional gradients and/or affect the properties in significant ways. The remaining Pb can also diffuse further into the TiO x /SiO 2 adhesion layers to chemically react, resulting in de adhesion of the film from the substrate, or pote ntia l l y deleterious effects on underlying device structures when integrated with silicon logic. [ 33 ] Various processing methods for PZT thin films have been developed to mitigate detrimental interaction s between the films and substrate thereby improving pr operties. [ 89 91 ] Researchers studying base metal integration observed metal substrate oxidation between PZT thin films and a copper substrate, resulting in degradation of the ferroelectric properties. [ 89 90 ] It was suggested that careful solution chemist ry selection and the control of oxygen partial pressure can mitigate the oxidation of a metal substrate, and promote the formation of perovskite phase PZT. Therefore, control of the atmosphere during crystallization of PZT thin films on Pt electrode s could likewise prevent interaction s between films and substrates, as well as promote the formation of the perovskite phase.

PAGE 67

67 In this chapter the crystallization behavior of PZT thin films under constant flowing atmospheres (air, reducing, and oxidizing) is firs t analyzed using the in situ XRD technique. Benefits of in situ XRD include the ability to observe transient phases and phase transformations during crystallization. Based on the knowledge of how different atmospheric conditions promote different evolution of phases, the switching of atmospheric conditions during crystallization of PZT thin films is introduced. Varying the oxygen content by intentionally switching the atmosphere midway during crystallization allows for control over the stability of the fluo rite/pyrochlore phase and also destabilizes the formation of intermetallic phases. 5.2 Experimental Procedure Thin film samples of 0.35 M solution derived Pb(Zr 0.52 Ti 0.48 )O 3 with 20% Pb excess were prepared via the IMO route as described in C hapter 2. Cry stallization of the thin film samples was performed in a furnace installed on an Inel diffractometer. The curved position sensitive detector (CPS) on the Inel diffractometer allowed for concurrent measurement of diffraction intensities over a wide 2 range and diffraction patterns could be collected in ~ 60 s. Crystallization was performed by ramping the furnace to 700 C at an average heating rate of 25 C /min, with diffraction patterns continuously record ed during crystallization. Thus, each diffraction pat tern represents time/temperature averaged information during the 60 s acquisition time window. Temperature was measured using a thermocouple within the alumina sample holder. The total pressure in the furnace was 1 atm and the volume of the furnace was app roximately 500 cm 3 During heating, oxygen rich, oxygen poor, air and switching atmospheric condition gas flows were adjusted to 1 slm. Peak intensity profiles of Pt 3 Pb, pyrochlore and PZT were modeled using the Pearson VII function and

PAGE 68

68 appropriately modi fied to include diffracted intensities from both the Cu 1 2 of the incident X ray beam. The integrated intensities obtained through this fitting procedure were used to quantify the evolution of different phases with respect to temperature. [ 92 ] Conventional Bragg Brentano X ray diffraction (APD 3720, P hilips, USA) was also used to analyze the structure of the samples after crystallization. Microstructures of PZT thin films under different atmospheric conditions were observed using scanning electron microscopy (SEM). 5.3 Results and Discussion In situ XR D patterns of PZT films obtained under different atmospheric conditions (Fig ure 5 1) show different phase evolution during high temperature heat treatment. For crystallization in air (Fig ure 5 1 B ), the transient intermetallic phase, Pt 3 Pb, was observed a t a low temperature (350C). The diffraction intensity of Pt 3 Pb (111) decreased in conjunction with an increase in the intensity of a pyrochlore phase at 450C. Thereafter, as the pyrochlore phase decreased, the peak intensity of the perovskite phase incre ased at ~ 550C. In situ phase evolution in flowing nitrogen gas and flowing oxygen gas are shown in Fig ures 5 1 A and C respectively. The major differences in phase evolution of the PZT films under different atmospheric conditions relates to the existe nce of the transient intermetallic Pt 3 Pb layer. In flowing oxygen gas, no intermetallic Pt 3 Pb was observed during crystallization, while intermetallic Pt 3 Pb phase was observed at 330C and remained up to 700C in flowing nitrogen. The formation of the int ermetallic Pt 3 Pb phase was observed in air ( p O 2 0.21 atm) and nitrogen gas ( p O 2 10 7 atm) during crystallization of PZT thin films. From thermodynamic considerations, reducing conditions with oxygen partial pressures below 10 20 atm between 300C and 500C can reduce Pb 2+ to Pb 0 promoting the formation of

PAGE 69

69 Pt 3 Pb phase. [ 46 ] Alternatively, it has been suggested that combustion of remnant organic compounds in the film may result in highly reducing conditions at the Pt Pb interface. [ 63 ] The observation of Pt 3 Pb in air and nitrogen gas implies that availability of oxygen in both air and nitrogen gas does not compensate for the loss of oxygen during removal of organic compounds. In air, Pt 3 Pb is observed to disappear with the formati on of pyrochlore in the film. In nitrogen gas, Pt 3 Pb remains after the formation of pyrochlore, and the intermetallic phase is present to the maximum temperature during crystallization. This observation suggests that the oxygen deficient conditions present in nitrogen gas annealing can further stabilize the Pt 3 Pb phase. In flowing oxygen, Pt 3 Pb was not observed. The highly oxidizing atmosphere likely inhibits Pb reduction. XRD patterns and scanning electron micrographs of the film microstructures after crys tallization (Fig ure 5 2) show that the final structure of the PZT thin film crystallized in air and oxygen is perovskite PZT. In films crystallized in nitrogen, Pt 3 Pb and PZT are observed in the diffraction patterns of the crystallized samples. This furthe r complements the in situ observation that flowing nitrogen during crystallization stabilizes the intermetallic Pt 3 Pb phase. To further engineer the processing of thin films, the concept of switching atmospheres during crystallization was introduced in or der to both mitigate the diffusion of Pb during the initial stage of crystallization while controlling the oxidation state of Pb during phase formation and enhancing the instability of the pyrochlore phase. [ 89 ] It was hypothesized that a high oxygen parti al pressure up to an intermediate temperature during crystallization may prevent the diffusion of Pb into the substrate that usually promotes the Pt 3 Pb phase. Switching to a less oxidizing environment at an intermediate

PAGE 70

70 temperature was hypothesized to stab ilize the oxidation state of Pb to Pb 2+ rather than Pb 4+ thereby promoting the formation of the perovskite phase over pyrochlore. The intermediate temperatures to switch the atmosphere from oxygen gas to nitrogen gas were selected based on the formation t emperatures of the intermetallic Pt 3 Pb phase in air (Fig ure 5 1). Two switching temperatures (T 0 ), 350C and 450C, were selected and represent the previously measured formation and disappearance temperatures, respectively, of the intermetallic Pt 3 Pb. A s c hematic representation of thermal and atmospheric conditions for thin film crystallization in switching atmospheres is shown in Fig ure 5 3. In the switching atmosphere condition, oxygen gas flow was 1 slm up to a switching temperature (T 0 ) of 350 C or 450 C at which time nitrogen gas flow started immediately and remained flowing at 1 slm until the end of the experiment. Phase evolution of PZT thin films under switching atmospheric conditions is shown in Fig ure 5 4. The transient intermetallic Pt 3 Pb phase wa s observed when the switching temperature was 350C, while it was not observed at 450C. This implies that the 1 atm oxygen partial pressure up to 450C was sufficient to prevent the reduction of Pb that normally enables intermetallic formation. Also, the peak intensity of (111) pyrochlore and (111) PZT as a function of temperature shows perovskite phase PZT formation at 550C and the pyrochlore phase disappeared at 580C. The transient pyrochlore phase also appears to be present over a smaller temperature range in the 450C switched sample versus the 350C switched sample. These observations indicate that oxygen deficient conditions at higher temperatures can stabilize the oxidation state of Pb as 2+, thereby promoting the formation of perovskite rather tha n pyrochlore. XRD patterns of PZT thin films after crystallization under switching

PAGE 71

71 atmosphere conditions are shown in Fig ure 5 5. It was observed using XRD that both switching temperatures result in perovskite phase without a remnant pyrochlore phase. Whil e switching atmospheres is shown to affect the crystallization of PZT in the present work, the alteration of atmospheres during processing can also be considered a general route towards controlling the synthesis of materials. During heat treatment, the atm osphere surrounding materials can become locally oxidizing or reducing, influencing redox reactions that affect the final chemistry and microstructure of the film. In solution based processing, the atmosphere during heat treatment and crystallization is sh own to play a crucial role in the phase evolution and final texture of the thin film. [ 43 89 90 ] Switching atmospheres during film evolution is shown to prevent the formation of intermetallic layers. Whereas previous investigations of solution derived thin film processing utilized oxidation resistant substrates and low processing temperatures to avoid interfacial reactions, prevention of substrate oxidation and interfacial reactions by changing atmospheres during processing could enable the use of less nobl e and less expensive substrates. [ 93 ] Therefore, switching atmospheres may be considered as a route to control the phase evolution and be used as a tool for new substrate exploration when preparing solution derived thin films. 5.4 Summary Phase stability of Pt 3 Pb and perovskite PZT depends on the oxygen partial pressure during crystallization of amorphous PZT. During crystallization of PZT thin films, oxidizing conditions mitigate the formation of Pt 3 Pb, while flowing nitrogen can stabilize the Pt 3 Pb phase Based on the crystallization behavior of PZT thin films under different atmospheric conditions, the technique of switching atmospheres during

PAGE 72

72 crystallization has been introduced. Manipulation of oxygen partial pressure during crystallization resulted in perovskite PZT thin films while minimizing formation of Pt 3 Pb.

PAGE 73

73 Fig ure 5 1 I n situ diffraction patterns of PZT thin films during crystallization under different atmospheric cond itions A ) N 2 B ) air C ) O 2

PAGE 74

74 Fig ure 5 2 Scanning electr on m icrographs and XRD patterns of PZT thin films after crystallization in different atmospheres.

PAGE 75

75 Fig ure 5 3 Schematic representation of thermal and atmospheric conditions for thin film crystallization in switching atmospheres (T0 values used are 350C and 450C).

PAGE 76

76 Fig ure 5 4 XRD patterns of PZT thin films during crystallization in different swit ching atmospheric conditions A ) 350C B ) 450C.

PAGE 77

77 Fig ure 5 5 XRD patterns of PZT thin films after crystalli zation in different switching temperatures A ) 350C B ) 450C.

PAGE 78

78 CHAPTER 6 THE STABILITY OF THE LEAD TITANATE SEED LAYER BEFORE CRYSTALLIZATION OF PZT THIN FILMS 6.1 L iterature R eview C rystallographic texture and the mitigation of composition al gradients ha ve been used to improve properties of ferroelectric thin films. For example, PZT thin films with minimal Zr /Ti gradients through the film cross section have higher transverse piezoelectric coefficients and relative dielectric constant s compa red to PZT thin films with chemical gradient s [ 9 4 ] Additionally, 001 ty pe textured PZT thin films show higher transverse piezoelect ric coefficients compared to PZT with random orientation. [ 9 4 ] Different methods have been used to control the creation of d ifferent textures in PZT thin films. Pt metal is widely used as the underlying substrate for PZT thin films due to its high oxidation resistance and low diffusivity. [ 66 ] Highly 111 textured PZT can be obtained on a 111 textured Pt substrate due to good la ttice matching between 111 Pt and 111 PZT. In contrast h ighly 001 type PZT thin films have been synthesized using a 100 type textured PTO seed layer as a template. [ 9 5 9 6 ] The underlying hypothesis for this observation is that 100 type texture in the PTO seed layer can nucleate 001 type texture in the PZT due to good lattice matching. [ 9 5 ] Thus, stability of the underlying substrate is an important factor that can be used to transfer the preferred orientation from the substrate to PZT thin films during cry stallization. Prior to crystallization of the PZT thin films, PZT solution was deposited on the substrate and pyrolyzed at a low temperature to remove organic compounds within films. It has been reported that the removal of organic compounds during pyrolys is

PAGE 79

79 could absorb oxygen from PZT, reducing the partial oxygen pressure at the interface between the substrate and PZT. [ 44 ] Additionally, chemical gradients between the substrate and the PZT could lead to interdiffusion between them, which may degrade the s tability of the substrate and the thin films. In this chapter, high energy synchrotron XRD and TEM were used to determine the stability of a 100 textured PTO seed layer during deposition of a subsequent PZT layer. C hemical gradients through the PZT thin films were investigated using the EDX technique. The influence that the pyrolysis of PZT layer has on the stability of PTO seed layer is discussed 6.2 Experimental Procedure 100 textured PTO seed layer s were prepared by the CSD method using 2 MOE solutio ns as described in C hapter 2. PZT thin films were spin cast on a PTO seed la yer with 2500 rpm for 45 s and then pyrolyzed at different temperature for 2 mi n XRD measurements were performed on crystallized PTO and pyrolyzed PZT on PTO using laboratory bas ed X rays ( 1.5404 ) and high energy X rays at beamline 6 ID 0.546 19 ) The diffraction geometry for this setup is similar to the Debye Scherrer reflection geometry described in C hapter 4 In order to analyze phases of thin films, the measured intensitie s from the 2D detector were integrated within a 2 range from = 5 and 5. The resulting integrated diffraction intensities as a function of 2 are plotted a nd are shown in Figure 6 2 In order to extract information on the texture of PTO a nd pyrolyzed PZT on PTO, the diffracted intensities from a given { hk l} reflection were evaluated in a limited azimuthal an gle range. The {110} reflection in the azimuthal angle range ( ) were selected and plott ed as shown in Figure s 6 3 This

PAGE 80

80 representation is referred to as an intensity map. The intensity map shows intensities of the 100 texture components at = 11.6) w ere integrated over the azimu thal angle range ( performed on the {110} reflections over the azimuthal angle range in order to obtain texture streng t h of PTO Dark field cross sectional images, Selected Area Diffraction (SAED) patterns and E DX line profiles of PZT on PTO after pyrolysis at 350C were recorded using TEM (FEI Tecnai F20). Sample preparation including thinning, milling and lift out for TEM analysis was conducted using the Focus Ion Beam (FEI NOVA 600 dual beam FIB). [8 1 ] 6.3 Res ults and Discussion The s tability of the PTO seed layer after pyrolysis at different temperatures was investigated using lab oratory based XRD ( Figure 6 1 ) Strong 100 texture was observed in the PTO seed layer. Diffraction patterns of the PTO seed layer we re observed after pyrolysis of PZT layer up to 320C. However, no diffraction patterns of the PTO seed layer were observed after pyrolysis of the PZT layer at temperatures equal to and above 340C. The initial XRD measurements show that PTO no longer contr ibuted diffraction intensity either due to a small enough volume to prevent constructive interference and Bragg scattering, or removal of crystalline PTO altogether. A greater penetration depth is available using higher energy X rays at the synchrotron a nd was required to confirm that the underlying PTO layer was no longer diffracting. Diffraction patterns of the PTO seed layer after pyrolysis of the PZT layer at different temperatures were observed using synchrotron XRD. The c rystallized PTO seed layer i n Figure 6 2 show s strong 100 texture After pyrolysis of a PZT layer on the PTO seed layer at 350C, neither perovskite PTO nor PZT were observed in the XRD

PAGE 81

81 patterns. This is consistent with the data shown from the laboratory XRD that the PTO layer does not diffract at temperatures exceeding 340C. An additional benefit of the synchrotron measurement technique is that diffraction from planes tilted at large angles relative to the sample normal can be measured at angles on the detector In the {110} intensity map, strong 100 texture components at = 45 and 45 were observed in the crystallized PTO seed layer as shown in Figure 6 3. T he strength of the 100 texture component was high, as evidenced by the modeled FWHM of the 100 texture components measured as 5.7 The 100 texture components were observed after heating at 100 C and 300 C. However, n o diffraction intensity of PZT on the PTO seed layer after heating at 350C was observed in the {110} intensity map which agrees with the prior observation. This result implies that the absence of diffracted intensity in the PTO seed layer is not due to absorption by the pyrolyzed PZT layer on the PTO seed layer. Further, it is supported that PTO has become amorphized. TEM was used to investigate the microstructure and crystal structure of PZT thin films on PTO seed layer s after pyrolysis of PZT layer at 350C. In Figure 6 4 A a TEM dark field micrograph of a cross section of the pyrolyzed PZT on PTO is presented The thickness of the PTO and PZT were 3 3 nm and 290 nm, respectively. Also, a smooth interface between the PTO and PZT layers was observed. SAED patterns of the PTO layer indicate that the PTO layer is an amorphous phase as shown in Figure 6 4 B This observation demonst rates that the PTO layer amorphized after pyrolysis of the PZT layer. An EDX line profile through the cross section of the thin films shown in Figure 6 4 C indicates that both PTO and PZT have a homogeneous distribution of elements

PAGE 82

82 layer. An EDX line pro file through the cross section of the thin films shown in Figure 6 4 C indicates that both PTO and PZT have a homogeneous distribution of elements The average Zr/Ti ratio in the entire PZT layer is calculated as 53/47 which corresponds to the targeted i nitial stoichiometry Based on the XRD and TEM study, it i s suggested that the highly 100 oriented PTO seed layer amorphized after pyrolysis of the PZT layer A high degree of orientation can be disrupted in many ways by i mpurities and nonstochiometry A s mall change in the mosaicity of high preferred orientation could result in the disruption of constructive interference in X ray diffraction. A lthough the detail of mechanism is unclear at present one possible explanation for the absence of crystallinity m ight be due to PbO flux from the PTO seed layer to PZT. It has been reported that PbO flux can influence distortion of the lattice of PTO prepared by the hydrothermal method at low processing temperature s [ 98 ] The PTO seed layer can ab sorb small excess of Pb. The c oncentration of Pb in PTO is higher than that of PZT, and thereby the chemical gradient may drive the Pb in the PTO toward the PZT Another possible explanation is the dissolution of PbO from the PTO seed layer at higher pH of the PZT solution. P bO has its minimum solubility at a pH of 9 and PbO could be dissolved from PTO when deposited and pyrolyzed PZT solution on top of PTO seed layer. Dissolution of a sufficient quantity of PbO could disrupt the PTO lattice depending on pyrolysis parameters such as temperature, time and rate, which results in the disruption of X ray diffraction. T he amorphized PTO seed layer may affect the subsequent texture control of PZT thin films during crystallization Further, chemical gradients between amorphous PTO an d PZT could affect the composition of PZT thin films during crystallization. The

PAGE 83

83 crystallization behavior of PZT thin films on an amorphized PTO seed layer is discussed in C hapter 8 6.4 Summary Stability of the PTO seed layer after pyrolysis of PZT layer at different temperature s was investigated. A 100 textured PTO seed layer was used before the deposition of a subsequent PZT layer. At pyrolysis temperature s of the PZT layer above 340C 3, crystalline PTO seed layer tran sformed to an amorphou s phase. An a morphized PTO seed layer may influence the crystallization behavior of PZT thin films. The effect of amorphized PTO seed layer on crystallization of PZT thin films is discussed in Chapter 8.

PAGE 84

84 Figure 6 1 PTO seed layer after pyro lysis of PZT layer at different temperature.

PAGE 85

85 Figure 6 2 Diffraction patterns of PTO seed layer. A ) Diffraction patterns of PTO seed layer after pyrolysis of PZT layer at different temperature s B ) (001) and (100) reflection of PTO seed layer b efore and after pyrolysis at 350 C.

PAGE 86

86 Figure 6 3 {110} Intensity map s of PTO seed layer after pyrolysis of PZT layer at different temperature s

PAGE 87

87 Figure 6 4 Dark field micrograph of PZT thin films on a PTO seed lay er after pyrolysis of PZT la yer A ) cross sectional view B ) SAED patterns of PTO layer C ) EDX line profile

PAGE 88

88 CHAPTER 7 CRYSTALLIZATION OF PZT THIN FILMS ON PLATINUM ELECTRODES 7.1 Literature Review Literature studies ha ve shown that 111 textured PZT thin films near the MPB show l arger polarization with a maximum relative dielectric constant ( ) than 100 and 110 textured PZT thin films. [ 97 ] In addition 111 textured PZT thin films exhibit a lower coercive field ( ) than 100 and 110 textured PZT thin films. [ 99 ] The electrical properties exhibited by 111 textured PZT thin films are therefore desirable for Ferroelectric Random Access Memory (FeRAM) which require s switching of the state of polarity with lower power and fast speed. Platinum (Pt) metal is widely used as the underlying substrate for PZT thin films due to its high oxidation resistance and low diffusivity. [ 66 ] The texture of PZT thin films on a Pt electrode depends on processing conditions such as the heating rate and pyrolysis temperature. [ 66 67 1 00 101 ] 111 texture has been observed when fast heating rates are used. [ 53 62 ] On the other hand, 100 texture is observed with a slow heating rate. [ 62 ] The observation of texture in PZT thin films and its dependence upon heating rates is consistent among the published studies. Based on observation s of the orientation dependence on heating rate, several mechanisms for the nucleation in PZT thin films have been proposed. Liu et al. suggested that 111 PZT is nucleated when deposited on a 111 Pt electrode. [ 74 ] Huang et al. suggested that 111 Pt x Pb intermetallic phase nucleates 111 textured PZT, because lattice matching between 111 Pt x Pb and 111 PZT is better than between 111 Pt and 111 PZT. [ 69 70 ] Tani et al. suggested that a Pt 3 Ti intermetallic phase fo rms by

PAGE 89

89 diffusion of Ti from the adhesion layer. The Pt 3 Ti is said to nucleate 111 texture of PZT thin films. [ 78 ] Muralt et al. suggested that the formation of rutile TiO 2 phase due to diffusion of Ti from the interface to the surface of the Pt electrode c an transform into 111 perovskite PZT. [ 66 ] It is apparent that there is no consensus concerning the proposed mechanism for the nucleation of 111 texture. In this chapter, the crystallization behavior of PZT thin films was investigated on Pt bottom electro des using in situ XRD at a synchrotron X ray source. Phase and texture evolution as a function of temperature were observed with different heating rates. The mechanism for nucleation of the texture of PZT thin films with different heating rates is discusse d based on a quantitative analysis of texture components of PZT thin films on Pt electrodes. 7.2 Experimental Procedure Pyrolyzed 2 MOE derived PZT thin films on Pt electrodes were crystallized at different heating rates using an IR lamp. Heating rates for crystallization of 2 MOE derived PZT thin films were controlled by applying a voltage on an IR lamp. The voltage was controlled by using a digital to analog converter program providing a voltage signal between 0 V and 5 V to a solid state power controller (Control Concepts Inc. 1032A). During crystallization with different heating rates exposure times between 0.25 s and 8 s were used to record diffraction patterns in a 2D detector. Based on the method illustrated in Chapter 4 the integrated diffraction intensities versus 2 over a r of the integrated diffraction intensities were plotted as a function of 2 and time. Integrated intensities of the (111) Pt x Pb and (111) PZT during crystallization of PZT thin

PAGE 90

90 films were fit to a Pearson VII profile shape function via the curve fitting tool box in MATLAB to investigate the orientation relationship between the two phases. 4) of PZT were integrated through 2 over the azimuthal angular range ( during crystallization of PZT thin films on Pt electrodes. The collection of the integrated diffraction intensit ies were plotted a s a function of and time. Also, the intensities of through 2 over the azimuthal angular range ( the d iffracted intensit y of the fluorite phase reached a maximum. Integrated intensities and the FWHM of the texture components were also extracted using the Pearson VII profile shape function via the curve fitting tool box in MATLAB Background intensities w ere modeled using a linear polynomial function. The background intensities correspond to diffraction intensities from randomly oriented grains. Further details of in situ measurements using synchrotron XRD are described in Chapter 4 Diffraction patterns o f PZT thin films after crystallization at different heating rates were recorded using a laboratory X P owder). 7.3 Results and Discussion XRD patterns of PZT thin films after crystallization at different heating rate s recorded using a laboratory diffractometer with /2 geometry are shown in Figure 7 1. Strong er peak intensity of (111) PZT was observed after crystallization at different heating rates compared to randomly oriented polycrystalline PZT. It implies that the formation of 111 texture in PZT is preferred o n Pt electrodes.

PAGE 91

91 In order to investigate the effect of heating rates on the formation of 111 texture in PZT thin films, in situ synchrotron XRD patterns of PZT thin films during crystallization at different heating rates are shown in Figure 7 2. Strong pea k intensity of (111) Pt was observed at all heating rates. Amorphous PZT transformed to perovskite PZT via a pyrochlore/fluorite phase at all heating rates. A transient Pt x Pb phase was observed prior to the formation of the pyrochlore/fluorite phase at all heating rates. The formation of an intermetallic Pt x Pb phase has been propos ed due to Pb diffusion from amorphous PZT thin films into the Pt electrode. [ 44 ] During crystallization of PZT thin films, the elimination of organic compounds takes place within the PZT films, which leads to the formation of local reducing conditions at the Pt PZT interface. [ 44 54 ] These local reducing conditions allow reduction of the Pb precursor (Pb 2+ ) in PZT films to metallic Pb (Pb 0 ). [ 44 89 ] The metallic Pb diffuses into th e Pt electrode to form the Pt x Pb phase. It has been reported that the extent of the local reducing conditions depends on the degree of oxygen diffusion from the surrounding atmosphere to the interface between PZT and Pt. [ 44 89 ] Similar trends were observe d in the current study. The maximum intensity of (111) Pt x Pb decreased with decreasing heating rates, as observed in Figure 7 3. This implies that more oxygen can diffuse into the Pt PZT interface to compensate for the loss of oxygen during removal of orga nic compounds at slower heating rates. The only reflection of Pt x Pb observed in the phase evolution plot is the (111) (Figure 7 2) We hypothesized that 111 texture in the Pt x Pb phase is nucleated from the 111 Pt electrode. In order to examine this hypot hesis, the orientation relationship between Pt and Pt x Pb was investigated. The intensity map of the Pt x Pb and Pt at

PAGE 92

92 different heating rates is shown in Figure 7 3, when the intensity of Pt x Pb is at its greatest value. Similar non uniform intensity distribu tion was observed in the {111} Debye Scherrer rings of Pt and Pt x Pb over the range ( in the 2 range ( Diffraction intensity from both the {111} Pt and {111} Pt x Pb is present at angle ( 72 0 and 72 ) in the 2 range ( which corresponds to 111 texture components of cubic phase Also, non uniform intensity distribution at different { hkl } Debye Scherrer rings of Pt and Pt x Pb over the range ( indicates a cubic structure with 111 texture components. It might be interpreted that both Pt x Pb and Pt have a cubic structure with similar stacking of {111} planes. The orientation relationship between the Pt electrode and the intermetallic Pt x Pb was consistent across films crystallized using different heating rates. The transient 111 Pt x Pb intermetallic phase has been suggested to promote 111 texture of PZT thin films due to good lattice matching. [ 56,57 ] This hypothesis can be tested based on the in situ XRD observation that the peak intensities between (111) Pt x Pb and (111) PZT overlap with no discontinuity during crystallization. However, there was no overlap of integrated intensities between (111) Pt x Pb and (111) PZT, as observed in Figure 7 4. This implies that the formation of 111 PZT occurs after decomposition of the 111 Pt x Pb and further, that 111 Pt x Pb does not nucleate 111 PZT. The discontinuity of diffracted intensities between (111) Pt and (111) Pt x Pb was observed across films crysta llized using different heating rates. The above discussion is based on the assumption that Pt x Pb disappears completely when the reflection is no longer observed.

PAGE 93

93 No intermediate phases such as Pt 3 Ti and PbO were observed at any heating rate during crystall ization of PZT thin films (Figure 7 2). Tani et al. suggested that the diffusion of Ti from the underlying adhesion layer into Pt leads to the formation of intermetallic Pt 3 Ti phase. [ 78 ] Further, it was found that the intermetallic Pt 3 Ti phase promotes 11 1 texture of PZT thin films. [ 78 ] Chen et al. suggested that the formation of a PbO phase takes place upon the decomposition of the Pt x Pb phase, and the PbO phase in turn promotes 100 texture in the PZT thin films. [ 71 72 ] O ur observation s during crystalli zation of PZT thin films show no evidence that intermediate Pt 3 Ti and PbO influence the resulting texture of PZT thin films. An intermediate fluorite type phase was observed after decomposition of the intermetallic Pt x Pb phase during crystallization. Ther e are different results reported for the crystal structure of the intermediate phase prior to the formation of perovskite PZT thin films. Mainly, pyrochlore phase (A 2 B 2 O 7 x Fd3m ) and fluorite phase ( A 2 B 2 O 7 Fm3m ) were observed as an intermediate phase. [ 5 0,51 ] Pyrochlore phase can be described as a doubled fluorite structure. [87 ] Experimentally, the pyrochlore phase can be distinguished from the fluorite phase based on the observation of the (111) reflection of doubled unit cell in the XRD patterns. [ 87 1 02 ] However, t he fluorite phase was determined as an intermediate phase in the current study, because no diffracted intensities were observed at 2 = 6.4 corresponding to (111) reflection of the doubled unit cell in the pyrochlore as shown in Figure 7 5. The fluorite phase was observed as the intermediate phase at all heating rates. The difference of an intermediate phase during crystallization ma y be due to different chemical solution s Krishna et al. observed the (111) reflection of doubled unit cell in the XRD patterns, which implies the

PAGE 94

94 intermediate phase is pyrochlore phase during crystallization of PZT thin films when using IMO route. [87 ] Ho wever, 2 MOE route was used to prepare PZT thin films in current study. It is suggested that the formation of fluorite phase from amorphous PZT is preferred prior to the formation of perovskite PZT due to easier atomic rearrangement with lower energy. [ 1 9 ] The relative stability and preferred orientation of the intermediate fluorite phase was suggested to influence the final texture of PZT thin films. For example, Norga et al. proposed that heterogeneous nucleation of fluorite occurs at the interface betwee n PZT thin film and the Pt electrode, which implies that 111 textured fluorite phase is nucleated by the Pt electrode. [ 75 ] The fluorite phase with 111 texture is said to transform to perovskite PZT while maintaining an orientation relationship with the Pt electrode. [ 75 ] In order to investigate the orientation relationship between 111 fluorite phase and 111 PZT, the intensity of the {111} peak of the 2 rated through 2 for the range ( 6. A broad and weak intensity distribution of 111 fluorite phase over the range ( was observed prior to the formation o f perovskite PZT thin films at all heating rates, which implies that the fluorite phase is randomly oriented prior to the formation of PZT thin films. It is likely that the fluorite phase is not heterogeneously nucleated from the interface between Pt and P ZT thin films. Random orientation of the fluorite phase was also observed at different heating rates prior to the formation of the perovskite phase. Therefore, there is no evidence from the study that 111 fluorite phase influences 111 texture of PZT thin films.

PAGE 95

95 The effect of the Pt electrode on texture selection of perovskite PZT thin films has been proposed in previous work Nittala et al. suggested that the 111 Pt electrode can provide energetically favorable nucleation sites and decrease the nucleation energy, promoting the formation of 111 texture in PZT thin films. [ 87 ] The orientation relationship between the 111 Pt electrode and 111 PZT has previously been discussed using texture evolution plots of PZT thin films during crystallization using differe nt heating rates ( Figure 7 2 ) Strong 111 texture components with weak intensity of 100 texture components of PZT thin films were observed at all heating rates. The change in FWHM and normalized intensities of PZT thin films during crystallization with dif ferent heating rates are shown in Figure 7 7. T he FWHM of 1 11 texture components increased with slower heating rates. Conversely, t he FWHM of 100 texture components increased with faster heating rates. The FWHM and texture fraction of different texture co mponents at the maximum temperature during crystallization are shown in Figures 7 8 A and B respectively The FWHM of 111 texture components in PZT thin films is similar to that in Pt electrodes at all voltage rates, which reinforces the proposal that 111 texture of PZT thin films may be related to templated growth from 111 texture of the Pt electrode. [ 87 ] Conversely, the FWHM of 100 texture components in PZT thin films was much larger than 111 texture components of the Pt electrode at all voltage rate s. An increase in the FWHM of 100 texture components was observed across films crystallized with increasing voltage rates. In PZT thin films, as shown in Figure 7 8 B the texture fraction of 111 texture components in PZT thin films increased with increas ing voltage rates, which implies that faster heating rates may enable heterogeneous nucleation of 111 texture components

PAGE 96

96 of PZT thin films on 111 Pt electrodes. Conversely, the fraction of 100 texture components increased with decreasing voltage rates. The formation of 100 texture components may suggest surface nucleation which is balanced with competing 111 texture components. [ 87 ] 7.4 Summary Amorphous PZT thin films transformed to perovskite PZT via a transient Pt x Pb phase followed by a fluorite phase du ring crystallization. Both 111 Pt x Pb and the 111 Pt electrode have a cubic crystal structure and 111 Pt x Pb is proposed to nucleate in the 111 Pt electrode. There was no evidence that 111 Pt x Pb nucleates 111 PZT. Also, no orientation relationship between 11 1 fluorite and 111 PZT was observed. Strong 111 texture was observed at all heating rates during crystallization of PZT thin films. Based on the investigation of the texture components of Pt and PZT at different heating rates, it is concluded that 111 Pt m ay play a role as a template to grow 111 texture of PZT thin films.

PAGE 97

97 Figure 7 1 XRD patterns of PZT thin films on Pt electrode s after crystallization with different heating rates.

PAGE 98

98 Figure 7 2 Phase and texture evolution of PZT thin films on Pt electrodes with different heating rates.

PAGE 99

99 Figure 7 3 Intensity map of Pt and PtxPb at the maximum peak intensity during crystallization.

PAGE 100

100 Figure 7 4 The formation of Pt x Pb phase. A ) Representative phase evolution plot at instantaneous heating (7 B ) The formation of Pt x Pb during crystallization of PZT thin films at different heating rates.

PAGE 101

101 Figure 7 5 Absence of (111) superlattice reflection of pyrochlore at 2 = 6.4 in synchrotron XRD patterns confirming the phase is fluorite.

PAGE 102

102 Figure 7 6 Integrated intensities of 111 fluorite at 2 range ( 9.8 0.5 ) over range between 90 and 90 before the formation of perovskite PZT thin films.

PAGE 103

103 Figure 7 7 Change in FWHM and normalized inten sity of texture components during crystallization of PZT with different heating rates. 111 texture components are plotted in red and 100 texture components are plotted in black. Lines are added as a guide to the eye.

PAGE 104

104 Figure 7 8 Texture components of P ZT thin films. A ) FWHM of texture components at the maximum tem perature during crystallization B ) Texture fraction of PZT thin films at the maximum temperature during crystallization. Lines are added as a guide to the eye.

PAGE 105

105 CHAPTER 8 CRYSTALLIZATION OF PZT THIN FILMS ON AMORPHIZED PbTiO 3 SEED LAYERS 8.1 Literature Review A textured PTO seed layer is often used during processing a s a template to promote 100 texture in the PZT thin film. [ 64 95 96 ] Additionally it is well accepted that a seed layer is be neficial for the prevention of interdiffusion between the substrate and thin film s during crystallization. [ 103 ] However, amorphization of the PTO seed layer was observed after pyrolysis of PZT layer at a pyrolysis temperature of 350C (Chapter 6). The thi n films after pyrolysis are similar to an amorphous layer with compositional segregation (Ti rich layer on the Pt surface and Zr rich layer on top) on a Pt electrode. The amorphous layer can react with the Pt electrode, and the 111 texture of the Pt electr ode can be an important factor in the resulting texture of PZT thin films during crystallization. Compositional segregation in the amorphous layer potentially lead s to interdiffusion between PTO and PZT during crystallization, which results in an undesired composition of PZT thin films. It is hypothesized that the amorphized PTO layer and compositional segregation have a different influence on the crystallization behavior of PZT thin films when using different heating rates. In this chapter, the effect of d ifferent heating rates on the crystallization of PZT thin films on an amorphized PTO seed layer is studied using the in situ technique at a synchrotron X ray source. Phase and texture evolution were observed with different heating rates. The mechanism of n ucleation for texture of PZT thin films at different

PAGE 106

106 heating rates is discussed based on the quantitative analysis of texture components and microstructure of PZT thin films. 8.2 Experimental Procedure 2 MOE derived PZT thin films was deposited on an amorp hized PTO seed layer The PZT thin films were crystallized at different heating rates using an IR lamp. Heating rates for the crystallization of 2 MOE derived PZT thin films were controlled by applying a voltage on an IR lamp. The voltage was controlled u sing a digital to analog converter program providing a voltage signal between 0 V and 5 V to a solid state power controller (Control Concepts Inc. Model 1032A). During crystallization using different heating rates XRD patterns were recorded using a 2D det ector with exposure times between 0.25 s and 8 s. Using the method described in Chapter 4 the integrated diffraction intensities versus 2 over the A collection of the integrated diffraction intensities plotted as a function of 2 and time i s termed a d intensities in (111) Pt x Pb, (111) PZT, (111) Fluorite, (001) and (100) PZT were modeled using the Pearson VII profile shape function via the curve fitting tool box in MATLAB to investigate the phase transformations during crystallization. The intensitie azimuthal angular range ( of PZT thin films on an amorphized PTO seed layer. The collection of the integrated d iffraction intensities were plotted as a function of azimuthal angle ( ) and time. This is

PAGE 107

107 were modeled using a Pearson VII profile shape function via the curve fitting tool box in MATLAB to extract the integrated intensities and FWHM of different texture components. Background intensities were modeled using a linear polynomial function. The background intensities correspond to diffraction intensities from randomly oriented grains. Further details of the in situ measurements using synchrotron XRD are described in Chapter 4 Dark field cross sectional images, nano beam diffraction (NBD) patterns and EDX line profiles of PZT on an amorphized PTO seed layer af ter crystallization with an average heating rate of 30C/min (slowest heating rate) were recorded using TEM. Sample preparation including thinning, milling and lift out for TEM analysis was conducted using the Focus ed Ion Beam. [ 97 ] 8.3 Results and Discus sion Phase and texture evolution plots derived from synchrotron XRD data of PZT thin films on an amorphized PTO seed layer at different heating rates are shown in Figure 8 1 The relative low intensities of (111) Pt and (111) Pt x Pb in the phase evolution p lot obtained in the range ( is attributed to the fact that the normal to thin films is not parallel to the scattering vector of the diffract ing planes. P eak intensities of the (111) Pt and (111) Pt x Pb were observed in the phase evolution plot obtained in the range (70 Further details of the XRD geometry used in the current study are available in Chapter 4. In the phase evolution plots, no diffraction intensities from the crystalline PTO seed layer were observed in the ear ly stage of crystallization, which agrees with the amorphization of the PTO seed layer after pyrolysis of PZT thin films, as discussed in Chapter 6. The phase evolution sequence during crystallization is from

PAGE 108

108 amorphous to perovskite PZT via an intermetalli c Pt x Pb phase followed by an intermediate fluorite/pyrochlore phase. The presence of a transient Pt x Pb phase at all heating rates implies that the availability of an oxygen atmosphere does not compensate for the loss of oxygen during the removal of organic compounds. The intermediate phase was determined to be the fluorite phase, A 2 B 2 O 7 x because there was no diffraction intensity from the superlattice peak of (111) pyrochlore near 2 = 6.4 Changes in the peak intensities of (111) Pt x Pb, (111) PZT, (111) fluorite, (001) PZT, and (100) PZT during crystallization are shown in Figure 8 2 The only reflection of the Pt x Pb phase observed in the phase evolution plot is the (111). The (111) Pt x Pb was observed at 600 C at the fastest heating rate, while a lower f ormation temperature for the appearance of the (111) Pt x Pb was observed at slower heating rates. The formation of the (111) Pt x Pb reflection was observed at 450 C at the slowest heating rate. The formation temperature in the (111) of the Pt x Pb phase was ob served between 350 C and 450 C in previous literature. [ 47 ] The discrepancy in the formation temperature of the (111) Pt x Pb phase between the previous literature and the current study may be attributed with the different heating methods for the PZT thin fi lms. In previous literature, amorphous PZT thin films were placed on a hot plate at a fixed temperature, while continuous heating by IR lamp was used in the current study which is similar to RTA. Continuous heating with fast heating rates in the current st udy heated the amorphous PZT thin films to a higher temperature in a short time, which delays the formation of the Pt x Pb phase. It was reported that the crystallization temperature of PZT thin films is between 450 C and 600 C. [ 32 46 ] We hypothesized that the formation temperature of

PAGE 109

109 111 textured Pt x Pb and PZT can be overlapped at fast heating rates, and further 111 texture of Pt x Pb may nucleate 111 PZT due to good lattice matching. However, the peak intensities of the (111) Pt x Pb and (111) PZT did not over lap during crystallization at all heating rates. Similar trends were observed during the crystallization of PZT thin films on a Pt electrode ( Chapter 7 ) The discontinuity of the (111) intensities between Pt x Pb and PZT in both cases suggest that 111 Pt x Pb does not directly nucleate 111 perovskite PZT. Accompanying the decrease in peak intensity of (111) Pt x Pb, the peak intensity of the (111) fluorite phase increased at all heating rates, followed by the formation of the perovskite PZT phase, shown in Figur e 8 2 Similar trends f or the phase transformation from fluorite to perovskite PZT during crystallization were observed in both PZT thin films on amorphous PTO and Pt electrode s (Chapter 7). However, the peak intensity of (100) PZT increased with increasin g peak intensity of (111) fluorite in PZT thin films on amorphized PTO, while the peak intensity of (100) PZT increased w ith decreasing peak intensity of (111) fluorite in PZT thin films on a Pt electrode. It has been reported that higher Zr/Ti ratios in P ZT enhance the retention of fluorite phase prior to the formation of perovskite PZT during crystallization. [ 56 ] In other words, the crystallization temperature of perovskite phase PZT is higher when the Zr/Ti molar ratio is higher. In the current study, a lower formation temperature ( ~ 520 C) of the perovskite PZT thin films on an amorphized PTO seed layer was observed, compared to the crystallization temperature ( ~ 600 C) of PZT thin films on a Pt electrode (Chapter 7). This may indicate that a lower crys tallization temperature of PZT thin films on amorphized PTO is ascribed to the recrystallization of PTO or the crystallization of Ti rich PZT. However,

PAGE 110

110 recrystallization of a PTO seed layer is not likely in the current study because the fluorite phase was not observed during the crystallization of the PTO seed layer, as shown in Figure 8 3 Direct crystallization of perovskite PTO from the amorphous PTO without the fluorite phase was also observed in previous literature. [ 51 ] Conversely, the fluorite phase was observed prior to the formation of perovskite phase in PZT thin films on amorphized PTO, which supports the hypothesis that perovskite Ti rich PZT was crystallized at lower temperature. Accompanying the decrease in peak intensity of (111) fluorite, a shift of peak intensity from higher 2 (100) to lower 2 (001) was observed ( Figure 8 2) In other words, an increase in the peak intensities of (100) PZT w as observed prior to the increase in the peak intensity of (001) PZT. At room temperature the crystal structures of PTO and PZT used in c urrent study are tetragonal. However, both PTO and PZT above the Curie temperature (between 230 C (for Zr rich PZT) and 490 C (for Ti rich PZT)) transforms to a cubic structure T he observations of the (001) and (100) peaks are possibly attributed to the c rystallization of two different perovskite phases during crystallization. It is hypothesized that the 2 reflection shifts during crystallization are attributed to the crystallization of perovskite Ti rich PZT at lower temperature, and then crystallization of PZT with higher Zr concentration at higher temperature. [ 51 ] Previous studies have shown that PZT t hin films with different chemical compositions have different unit cell volumes. [ 104 105 ] Experimentally, a change in the unit cell volume of PZT thin films during crystallization can be measured using a 2D detector, as shown in Figure 8 4 A Debye Scher rer rings detected in the 2D detector can shift uniformly towards lower or higher 2 over the range (

PAGE 111

111 volume change during crystallization. For example, Ti rich PZT with a smaller unit cell volume can nucleate at a lower temperature (higher 2 ), while PZT with a higher Zr/Ti ratio can nuc leate at a higher temperature resulting in a larger unit cell volume (lower 2 ). Therefore, the volume change of PZT thin films from a smaller unit cell to a larger unit cell during crystallization could be shown as the shift of Debye Scherrer ring from hi gher angular angle (higher 2 ) to lower angular angle (lower 2 ) over the range ( In order to investigate the formation of perovskite PZT thin films with different chemical composition s during crystallization, ch anges in the integrated intensities of (001) and (100) as a function of temperature and a {110} intensity map of PZT thin films at different temperature s during crystallization at voltage rate of 0.005 V/s are shown in Figure 8 4 B and C respectively. At 500C (T1), no texture components were observed, which supports the observation of the dissolution of the PTO seed layer after pyrolysis of the PZT layer. At 516C (T2), 100 texture components were initially observed in the nucleating perovskite phase At 520C (T3), the intensity shift of texture components from higher to lower 2 was o bserved homogeneously over the entire measured a zimuthal angular range ( with a small unit cel l volume nucleate s first, and then PZT with a higher Zr / Ti ratio nucleates at a higher temperature during crystallization. At 556C (T4), no more intensity shift of texture components from higher to lower 2 was observed, which suggests no further composit ional change between Zr to Ti during the formation of PZT. At 700C (T5), increased strength of texture components in PZT thin films was observed at lower 2

PAGE 112

112 Texture evolution plots at different heating rates have been presented ( Figure 8 1) Correspondin g FWHM and normalized intensity of texture components as a function of time during crystallization are shown in Figure 8 5 There are no diffraction peaks in the early stage of crystallization, which supports the amorphization of the PTO seed layer after p yrolysis of upper PZT thin films. 100 110 and 111 texture components were observed at the fastest heating rate, while 100 and 111 texture components were observed at the slowest heating rate. Also, 111 texture appeared faster than 100 and 110 texture at f astest heating rate, while 100 texture appeared faster than 111 texture at the slowest heating rate. The FWHM and texture fraction of 100 110 and 111 texture components at the maximum temperature during crystallization with different voltage rates are su mmarized in Figures 8 6 A and B respectively. The FWHM of 111 texture components are similar to those of 111 Pt at all voltage rates, which implies that 111 texture of Pt nucleates 111 texture of PZT thin films. Conversely, the FWHM of 100 texture comp onents increased with increas ing heating rates. The texture fraction at different voltage rates is illustrated in Figure 8 6 B Dominant 100 texture was observed at low voltage rates, while 111 texture components were dominant at high voltage rates. Imagi ng in the TEM was carried out in order to confirm the observations described above. A cross sectional micrograph of a PZT thin film after crystallization at the slowest heating rate (Figure 8 7 A ) shows two distinct regions on the Pt electrode which were previously the amorphous PTO seed layer and PZT layer (Chapter 6) Both amorphous PTO and PZT have been crystallized. Also, the interfacial roughness increased between the PTO layer and PZT layer compared to the interfacial roughness

PAGE 113

113 between layers before crystallization (Figure 8 7 A ), which m ight be attributed to several factors such as dewetting or interdiffusion during crystallization. Nanobeam diffraction patterns of PTO and PZT are shown in Figures 8 7 B inset and D inset, respectively. The PTO an d PZT are tetragonal phase. EDX line profiles of Pb, Ti and Zr are provided in Figure 8 7 C No signal s for Pb, Ti or Zr w ere observed in the Pt layer. This indicates that no interdiffusion of elements between the substrate and the upper PTO/PZT layers h as occurred. A small amount of elemental Zr was detected in the PTO layer and a similar amount of elemental Pb was detected in both the PTO and PZT layers, which suggests that Zr from the PZT layer diffused into PTO layer and Ti rich PZT was crystallized n ear the Pt electrode. Also, a decrease in the Ti count and a similar increase in the Zr count with increasing distance were observed in the PZT layer, which implies that the Zr/Ti molar ratio increased throughout the cross section from the Pt thin film int erface to the surface with increasing temperature. The average Zr/Ti ratio s in the entire PTO and PZT layer were calculated as 10/90 and 51/49 which does not correspond to the initial stoichiometry of the thin films before crystallization T he crystalliza tion of Ti rich PZT appears to occur at a lower crystallization temperature, followed by the formation of Zr rich PZT across the thin film I t has been reported that dominant 100 texture is commonly observed in Zr rich PZT thin films. [ 67 ] Higher temperat ure is required to crystallize PZT with increasing Zr concentration, which leads to homogeneous n ucleation in PZT thin films. [ 51 ] In the case of PZT, the 100 plane has the lowest surface energy, and thereby dominant nucleation and growth of 100 texture oc cur. [ 107 ] Dominant 100 texture components observed in PZT thin films crystallized at slower heating rates may imply that sufficient

PAGE 114

114 time for interdiffusion between the PTO layer and PZT layer leads to increased Zr concentration towards surface and increas ed Ti concentration towards the PTO PZT interface which promotes PZT thin films with a higher Zr concentration. 8.4 Summary The effect of an amorphized PTO seed layer on the crystallization behavior of PZT thin films using different heating rates was stu died. Intermediate phases such as Pt x Pb and fluorite phase were observed prior to the formation of perovskite PZT thin films. Dominant 111 texture was observed at faster heating rates, while dominant 100 texture was observed at slower heating rates. It is suggested that 111 texture of the Pt plays a role in the nucleation of 111 PZT thin films. During the crystallization of PZT thin films, the amorphous PTO layer was recrystallized to form a Ti rich PZT layer through diffusion of Zr atoms from the PZT layer to the PTO layer. It is suggested that inter diffusion of Pb, Ti and Zr occurs between amorphous PTO and amorphous PZT during the crystallization of PZT, which can lead to chemical gradients in PZT thin films. Finally two different compositions in the am orphous layer lead to the formation of chemically inhomogeneous PZT thin films during crystallization.

PAGE 115

115 Figure 8 1 Phase and texture evolution plots of PZT thin films on an amorphized PTO seed layer during crystallization with different heating rates

PAGE 116

116 Figure 8 2 Change in peak intensities of (111) PtxPb, (111) PZT, (111) fluorite, (001) PZT, and (100) PZT during crystallization.

PAGE 117

117 Figure 8 3 Phase evolution plot of a PTO seed layer during crystallization at the voltage rate (0.01 V/s).

PAGE 118

118 Figure 8 4 Intensity shift of PZT thin films on amorphous PTO seed layer during crystallization. A ) Homogeneous shift of Debye Scherrer rings due to volume change in 2D detecto r. B ) Representative phase evolution plot in the range ( 2 of (001) and (100) are as a function of temperature C ) {110} intensity maps at different temperatures during crystallization.

PAGE 119

119 Figure 8 5 Change in FWHM of texture components during crystallization of PZT with different heating rates.

PAGE 120

120 Figure 8 6 Texture components of PZT thin films on amorphous PTO seed layer. A ) FWHM of texture components at the maximum tempera ture during crystallization. B ) Texture fraction of PZT thin films at the maximum temperature during crystallization.

PAGE 121

121 Figure 8 7 Dark field micrograph of a PZT thin film on a PTO seed layer after crystallization of the PZT layer A ) C ross sectional view B ) NBD pattern of a PTO layer C ) EDX line profile D ) NBD pattern of a PZT layer

PAGE 122

122 CHAPTER 9 CRYSTALLIZATION OF PZT THIN FILMS ON CRYSTALLINE PbTiO 3 SEED LAYER S 9.1 Literature Review Recent thin film research has focused on the use of seed layers that promote th e [106,107 ] Additionally, the seed layers reduce the crystallization temperature of upper thin films which results in the suppressi on of [ 106 107 ] PTO has several advantages as a se ed layer for the formation of perovskite PZT thin films. PTO has a similar crystal structure and lattice parameter to PZT with a Zr/Ti ratio of 52/48, which can lower the activation energy for nucleation of PZT, and thereby suppress the crystallization tem perature of PZT thin films. The lower crystallization temperature can r educe the volatilization of PbO at higher temperature s which enhances the crysta l linity of perovskite phase without any remnant Pb deficient phase. Additionally, texture control of the PZT is possible with the aid of a textured PTO seed layer. For example, Hiboux e t al. suggested 111 textured PTO seed layer promotes the formation of 111 texture in PZT thin films. [ 107 ] Also, 100 textured PZT thin films were obtained from a 100 textured PTO seed layer [ 107 ] It has been suggested that t he texture in the PTO seed layer between the Pt electrode and could provide sufficient nucleation sites to enhance the texture of perovskite PZT lms. [ 95 96 106 108 ] However, strategies to control final film texture are limited due to the lack of mechanistic understanding of crystallization in PZT thin film s using real time measurements This chapter presents the results and discussion of t he effect of a crystalline PTO seed layer on the crystallization behavior of PZT thin films using in situ XRD

PAGE 123

123 measurements at a synchrotron X ray source. The results obtai ned from in situ measurements reveal a fundamental understanding of crystallization behavior during heat treatments such as (1) multiple potential interactions with the substrate, (2) phase transformation from amorphous film to the desired homogeneous crys talline phase, and (3) texture evolution with respect to different heating rates 9.2 Experimental P rocedure Pyrolyzed 2 MOE derived PZT thin films on a crystalline PTO seed layer were crystallized at different heating rates using an IR lamp. Heating rate s for crystallization of 2 MOE derived PZT thin films were controlled by applying different voltage s on an IR lamp using a digital to analog converter program. During crystallization at different heating rates exposure times on the 6 ID B diffractometer a t the APS between 0.25 s and 6 s were used to record diffraction patterns in a 2D detector. Based on the method illustrated in C hapter 4 the integrated diffraction intensities versus 2 over a range ( 7 8 0 and 5 of the integrated diffraction intensities were extracted as a function of 2 and time to show phase evolution plot s Change s in the integrated intensities of the ( 111 ) Pt x Pb, (111) PZT, ( 111 ) f luorite, ( 001 ) an d ( 100 ) PZT during crystallization were modeled using the Pearson VII profile shape function via the curve fitting tool box in MATLAB through 2 over the azimuthal range ( crystallization of PZT thin films on a crystalline PTO seed layer. The collection of the integrated diffraction intensities were ex t racted as a function of azimuthal angle and time for texture evolution p lot s Integrated intensities of the texture components were fit to the Pearson VII profile shape function via the curve fitting tool box in MATLAB to

PAGE 124

124 extract texture information Background intensities were modeled using a linear polynomial function. The background intensities correspond to diffraction intensities from randomly oriented grains. D etails of in situ measurements using synchrotron XRD are described in C hapter 4 Diffraction patterns of the PZT thin films after crystallization at different heat ing rates were recorded using laboratory XRD. 9.3 Results and Discuss ion A { 110 } intensity map showing texture components of a PTO seed layer are shown in Figure 9 1 A N on uniform intensity for a {110} reflection over the range 90 90 was observed The peak intensities are observed in { 110 } intensity map at = 45 and 45 which corresponds to 100 texture components. The peak intensities at = 0 correspond to 110 texture components. Also, near ly uniform intensity over the range 90 90 was observed, which corresponds to random orientation. Based on th ese observations, it is clear that the PTO seed layer has comparatively stronger 100 texture with a rando m compo nent In order to investigate the templating effect of the Pt on the PTO seed layer, FWHMs of the texture components of the PTO seed layer were compared to that of the Pt elec trode, as shown in Figure 9 1 B The Pt electrode showed a strong tendency fo r 111 texture (Chapters 7 and 8). It was observed that FWHMs in the 100 and 110 texture components of the PTO seed layer are different from that of 111 Pt electrode. This may indicate that the Pt electrode does not provide nucleation sites for 100 and 110 of the PTO seed layer. Comparatively stronger intensity of 100 texture components is likely due to an inherent self orientation tendency in which the plane has the lowest surface

PAGE 125

125 energy. [ 107 ] In the case of PTO, the (100) is considered to be the lowest su rface energy plane [ 107 ] In situ XRD measurements were conducted in order to investigate the effect of a PTO seed layer on the crystallization behavior of PZT thin films. Phase and texture evolution during crystallization of a PZT layer were observed as d isplayed in F igure 9 2. During crystallization of PZT thin films at different heating rates the (111) peak of the Pt x Pb phase was observed first This does not support the idea that PTO is a prevention layer for diffusion of excess Pb from the PZT layer t o the Pt electrode Instead, this observation might imply that excess Pb in the PZT layer diffuse s into Pt (perhaps along grain boundar ies of the PTO seed layer ) The only reflection representing the Pt x Pb phase is (111) at 2 = 13.7 in the phase evolution plots. It is controversial that 111 Pt x Pb can nucleate 111 texture of PZT due to similar lattice parameters [ 69 70 ] The change in peak intensities of the ( 111 ) Pt x Pb and ( 111 ) PZT during crystallization i s shown in F igure 9 3. In the phase evolution plots, the (111) peaks of the Pt x Pb and PZT are located at a similar 2 position due to close lattice parameters, and thereby we can hypothesize that both (111) peaks from Pt x Pb and PZT could be overlapped with increasing temper ature where Pt x Pb nucleates PZT. However, no overlap of the peak intensities was observed between ( 111 ) Pt x Pb and ( 111 ) PZT, which implies that Pt x Pb does not nucleate PZT. It is also suggested that the crystalline PTO seed layer can become a barrier betwe en the Pt x Pb phase and the PZT layer to suppress the orientation relationship between them. Accompanying the decrease of peak intensity of the (111) Pt x Pb, the peak intensities of the intermediate phase increased during crystallization at all heating rate s

PAGE 126

126 There is debate concerning the crystal structure of the intermediate phase prior to the formation of perovskite PZT. A p yrochlore phase Pb 2 Ti 2 O 7 was suggested as an intermediate phase, while a fluorite phase Pb 2 Ti 2 O 7 x was suggested as an intermedia te phase. [ 50 51 52 ] Experimentally, the fluorite phase can be distinguished from the pyrochlore phase based on the lack of (111) reflection of doubled unit cell in pyrochlore in the diffraction patterns. As shown in F igure 9 4, intensity from the ( 111 ) py rochlore was not observed at 2 = 6.4 during crystallization at all heating rates which suggests that the intermediate phase observed in the current study is the fluorite phase. Changes in the peak intensities of the ( 111 ) fluorite ( 001 ) PZT and ( 100 ) PZT during crystallization are shown in F igure 9 5. At an early stage of crystallization, the 2 position of the ( 001 ) peak shifted to higher 2 while 2 position of the ( 100 ) peak shifted to lower 2 The ( 001 ) and ( 100 ) peaks converged to one peak at hi gher temperature which suggests that tetragonal PTO transform s to cubic above the C urie temperature. N o more increase of ( 100 ) peak intensity was observed after the phase transformation of PTO, which indicates that the transformation of PTO from tetragona l to cubic is complete. Subsequently, an increase in the peak intensit y of the ( 001 ) PZT was observed with a decrease in the peak i ntensity of the (111) fluorite phase Similar trends in the phase evolution were observed at all heating rates. Based on the quantitative results of the ( 001 ) PZT and ( 100 ) PZT in Figure 9 5 it is proposed that the cubic PTO seed layer above the C urie temperature nucleates PZT thin films The observation s of the phase evolution of PZT on PTO (Figure 9 5) are different from th e phase evolution of PZT on Pt electrode s (Chapter 7) A similar formation

PAGE 127

127 temperature of the fluorite phase was observed during the crystallization of PZT on the Pt electrode and PTO seed layer, respectively However, the subsequent formation of the PZT o n PTO was observed at the same time as the formation of the fluorite phase, while the formation of PZT on Pt was observed after fluorite phase disappeared. This implies that the crystallization temperature of PZT on Pt is higher than that of PZT on a PTO s eed layer The crystallization temperature of the PZT on PTO was 590 C, while crystallization temperature of the PZT on Pt was 650 C. Thus, it suggests that a PTO seed layer effective ly lower s the crystallization temperature of PZT thin films. Texture evo lution plots at different heating rates have been presented in Figure 9 2 B road 100 texture component s of PTO w ere observed at = 45 and 45 with nearly uniform intensity of random components over the whole range in the early stage of crystallization. The low scattered intensity of 100 texture components can be attributed to the small interaction volume of thin (34 nm) PTO seed layer. Additionally, an amorphous PZT layer can also absorb the diffracted X rays from PTO to cause a decrease of diffraction intensity of the PTO seed layer. Subsequent 100 texture components of PZT w ere observe d at = 45 and 45 at higher temperature Nearly u niform intensities across the whole range and the relatively weak intensity peak at = 0 are also observed, indicating the presence of random and 110 texture components. Similar trends in the formation of texture in the PZT were observed at all heating rates. Corresponding FWHM and normalized intensities of texture components as a function of time during crystallization are shown in Figure 9 6 Similar values of the FWHMs of 100 and 110 texture components of PZT were observed at all heating rates. The formation of 111 texture components was suppressed in PZT on PTO seed layer. It is

PAGE 128

128 likely that the absence of 111 texture in the PTO seed layer does not provide nucleation sites for 111 texture of PZT thin films. It is suggested that the orientation of the substrate below PZT thin films is an important factor th at leads to the development of the desired texture in PZT thin films. The FWHM s and texture fraction of PZT thin films at the maximum temperature during crystallization at different voltage rates are summarized in Figures 9 7 A and B respectively. The FWHM of texture components of PZT are similar to those of 1 00 P TO at all voltage rates, which implies that 1 00 texture of P TO nucleates 1 00 texture of PZT thin films regardless of heating rates The texture fraction at different voltage rates illustrated in Figure 9 7 B show that the texture fraction of 100 components are similar regardless of heating rates. Diffraction patterns of the PZT on a PTO seed layer prior to and following crystallization at different heating rates using laboratory XRD with /2 geometry are shown in F igure 9 8 Before crystallization of the PZT layer, the peak intensities of {100} and {110} of the PTO seed layer were observed. After crystallization, the {100} and {110} reflections of the PZT were observed at all heating rates. Th e peak intensities of (001) and (101) were much higher than that of (100) and (110) after crystallization at all heating rates, which implies the growth of c axis domains in the PZT occurred preferentially throughout the cross section. 9.4 Summary A cry stalline PTO seed layer was used as a template for PZT thin films. After pyrolysis of upper PZT thin films, crystalline PTO still remained. A morphous PZT thin films transformed to perovskite PZT via a Pt x Pb phase followed by a fluorite phase during crystal lization. There i s no evidence that 111 Pt x Pb nucleates 111 PZT. A phase

PAGE 129

129 transformation of PTO from tetragonal to cubic occurred, and then nucleation of PZT occurred from the cubic phase PTO seed layer. 100 texture components were observed at all heating r ates. Based on the investigation of the texture components of Pt and PZT at different heating rates, it is concluded that cubic phase 100 PTO may play a role as a template to grow 100 texture of PZT thin films. After crystallization, preferential growth of (001) in the PZT was observed, which implies that cubic PTO nucleates c axis domains in PZT preferentially during crystallization.

PAGE 130

130 Figure 9 1 Intensity map of crystalline PTO seed layer. A ) {110} intensity map of PTO seed layer afte r p yrolysis of PZT layer. B ) Comparison in the FWHM of texture components of PTO and Pt electrode.

PAGE 131

131 Figure 9 2 Phase and texture evolution plots of PZT thin films on PTO seed layer during crystallization at different heating rates.

PAGE 132

132 Figure 9 3 The formation of Pt x Pb phase. A ) Representative phase evolution plot at instantaneous heating (70 B ) The formation of Pt x Pb during crystallization of PZT thin films at different heating rates.

PAGE 133

133 Figure 9 4 Absence of (111) superlattice peak in pyrochlore in PZT thin films during crystallization of PZT thin films at different h eating rates confirming the phase is fluorite.

PAGE 134

134 Figure 9 5 Formation of pyrochlore and perovskite PZT during crystallization of PZT layer at different heating rates.

PAGE 135

135 Figure 9 6 Texture evolution of PZT thin films during crystallization at different heating rates.

PAGE 136

136 Figure 9 7 Texture components of PZT thin films on crystalline PTO seed layer. A ) FWHM of texture components at the maximum temperatur e during crystallizati on. B ) Texture fraction of PZT thin films at the maximum temperature during cryst allization.

PAGE 137

1 37 Figure 9 8 Diffraction patterns of PZT thin films on PTO seed layer after crystallization at different heating rates.

PAGE 138

138 CHAPTER 10 CONCLUSIONS AND CONTRIBUTIONS 10.1 Conclusions The effect of different processing con dition s such as atmospheric conditions, heating rates and different substrates on the crystallization behavior of PZT thin films was investigated using in situ lab oratory and synchrotron XRD measurements Overall, a similar sequence of phase evolution in P ZT thin films during crystallization was observed regardless of different processing conditions. During crystallization amorphous PZT thin films transformed to perovskite PZT phase via an intermetallic Pt x Pb phase followed by fluorite phase. During the e arly stage s of crystallization of PZT, decomposition of organic compounds within amorphous PZT occur s leading to local reducing conditions at the Pt PZT interface, and then Pb from amorphous PZT is reduced to metallic Pb (Pb 0 ). The formation of the interm etallic Pt x Pb phase is attributed to diffusion of the reduced Pb (Pb 0 ) from the amorphous PZT into the Pt electrode. The intermetallic Pt x Pb phase was observed to be highly 111 oriented. It is suggested that 111 texture of Pt x Pb is nucleated from 111 Pt. The formation and stability of the intermetallic Pt x Pb phase was observed to be highly depend ent on different atmospheric conditions during crystallization of PZT using lab ora tory based in situ XRD. In flowing N 2 gas, Pt x Pb remained after crystallization o f PZT thin films. Conversely, Pt x Pb phase was not observed in flowing O 2 gas. This implies that the extent of the local reducing conditions depends on the degree of oxygen diffusion from the surrounding atmosphere to the interface between PZT and Pt. In

PAGE 139

139 fl owing O 2 gas, however, high oxygen partial pressure can allow the average oxidation state of Pb to stabilize between 2+ and 4+, which i ndicates that overall stoichiometry of oxygen content favors the intermediate pyrochlore/fluorite phase, Pb 2 (Zr x Ti 1 x )O 7 y where 0 Therefore, switching atmospheric conditions during crystallization of PZT thin films was introduced to suppress the intermetallic Pt x Pb phase and to promote perovskite PZT phase. It was observed that a high oxygen partial pressure of O 2 gas suppressed the formation of the Pt x Pb phase, which implies that increased availability of oxygen at the Pt Pb interface mitigate s the local reducing conditions during the removal of organic compounds. Further, subsequent switching to N 2 gas at hi gher temperature s can stabilize the oxidation state of Pb as 2 + to promote the perovskite PZT phase, Pb(Zr x Ti 1 x )O 3 Different phase and texture evolution of PZT thin films during crystallization were observed depending on different heating rates and subs trates. Dominant 111 texture of PZT thin films was observed on 111 Pt electrode s at all heating rates, even though 100 texture increased slightly at slower heating rates. There is no evidence that intermediate phases such as 111 Pt x Pb and 111 fluorite nucl eate 111 texture of PZT thin films. It is suggested that direct nucleation of 111 preferred orientation from the Pt electrode to PZT occurs during crystallization. The formation of 100 texture at slower heating rates m ay be due to homogeneous nucleation or surface nucleation which is balanced with competing 111 texture. When amorphized PTO was used as a seed layer below PZT thin films, a more extreme change in texture of PZT thin films was observed with different heating rates during crystallization, compar ed to that observed in PZT thin films on Pt electrode s 111

PAGE 140

140 texture was preferred at faster heating rates, while 100 texture was preferred at slower heating rates. It is suggested that 111 texture of PZT thin films at faster heating rates is nucleated from 111 texture of the Pt electrode. A further observation in PZT thin films on amorphized PTO seed layer is that Ti rich PZT forms on the Pt electrode and Zr rich PZT forms near the thin film surface This suggest s that the interdiffusion of Pb, Ti and Zr b etween the amorphous PTO layer and the PZT layer occurs during crystallization. This results in a l ower crystallization temperature of PZT thin films on amorphized PTO which is attributed to the lower temperature required to produce Ti rich PZT, compared to the PZT thin films on Pt electrodes With increasing temperature, crystallization of perovskite PZT with a higher Zr/Ti ratio was observed. Dominant 100 texture in PZT thin films crystallized at slower heating rates may imply that 100 texture is homogen eously nucleated at slower heating rates due to higher Zr concentration in PZT In th is study, an amorphized PTO seed layer was determined as another parameter that influence s texture selection and chemical gradients of PZT thin films. A 100 textured PTO seed layer was used as a template for 100 textured PZT thin films based on the hypothesis that 100 PTO nucleates 100 PZT due to good lattice matching between them. A p hase transformation of the PTO seed layer from tetragonal to cubic was observed during cr ystallization. The formation of PZT thin films was observed after the phase transformation from tetragonal to cubic in the PTO seed layer. A l ower crystallization temperature of PZT was observed, which may suggest that the PTO seed layer lowers the activat ion energy for the formation of perovskite PZT. T here is no evidence that intermediate phases such as Pt x Pb nucleate 111 texture of PZT thin

PAGE 141

141 films. 100 texture of PZT was observed regardless of heating rates. It is suggested that 100 texture of PZT thin fi lms is nucleated from the cubic 100 PTO seed layer above the C urie temperature. Also, the growth of c axis domains of PZT is preferred on cubic PTO seed layer during crystallization. 10.2 Contributions of Dissertation This dissertation investigated the effect of different processing conditions on the crystallization behavior of PZT thin films using the in situ XRD technique T he major contributions of this work to the scientific community are explicitly listed here as a summary for the reader : 1. Switching atmosphere during crystallization as a new approach is presente d not only to control the interaction between the substrate and thin films, but also to promote the desired crystal structure without any remnant secondary phases. 2. Fluorite phase is determin ed as an interm ediate phase during crystallization of PZT thin film s prepared using 2 MOE solution. 3. It is proposed that 111 texture of PZT is nucleated from 111 texture of Pt substrate 4. Diffusion of Pb along the grain boundary of the Pt substrate is su ggest ed to form intermetallic Pt x Pb phase during crystallization of PZT 5. It is proposed that a c ubic phase PTO seed layer above the Curie temperature i s a template to nucleate and grow c domain PZT thin films preferentially.

PAGE 142

142 LIST OF REFERENCES [1] J. F. Nye, Physical Properties of Cryst als, ( Oxford Science Publishing, Clarendon, 1985 ) p. 352. [2] K. Uchino, Ferroelectric Devices, ( Marcel Dekker, New York 2000 ) p. 308 [3] B. Jaffe, W. R. Cook, Jr. and H. Jaffe, Piezoelectric Ceramic s Vol. 3 (Academic Press, London, 1971) p. 78. [4] H. A. Sodano, G. Park and D. J. Inman Strain 40 49 ( 2004 ) [5] K. Brinkman, Positional Order in Lead S c andium Tantalate (PST) as a <> for the Investigation of Relaxor Ferroelectric Behavior in Thi n Films Ph. D dissertation, cole Polytechnique F ederale de L ausanne (20 04 ). [6] P. Papet, J. P. Dougherty, and T. R. Shrout, J. Mater. Res. 5 2902 (1990). [7] G. Catalan, M. H. Corbett, R. M. Bowman, and J. M. Gregg, J. App. Phys. 91 2295 (2002). [8] J. H. Park, and S. Troiler Mckinstry, J. Mater. Res. 16 268 (2001). [9] L. Francis, Y. Oh, and D. Payne, J. Mater. Sci. 25 5007 (1990). [10] C. Parker, J. P. Maria, and A. Kingon, Appl. Phys. Lett. 81 340 (2002). [ 11 ] S. Roundy and P. K. Wright, Smart M ater. Struct. 13 1131 (2004). [ 1 2] S. Priya and D. J. Inman, Energy Harvesting Technologies ( Springer Verlag, New York, 2008). [ 1 3] I. Kanno, S. Fujii, T. Kamada, and R. Takayama, Appl. Phys. Lett. 70 1378 (1997). [ 1 4] P. Muralt, Int. Ferro. 17 297 (19 97). [ 1 5] N. Ledermann, P. Muralt, J. Baborowski, S. Gentil, K. Mukati, M. Cantoni A. Seifert, and N. Setter, Sensors Actuat A Phys 105 162 (2003). [ 1 6] R. Ramesh, Thin film ferroelectric materials and devices, Vol. 1 (Kluwer Academic Publishers Norwell massachusetts, 1997) p. 249. [ 1 7] O. Auciello, J. F. Scott, and R. Ramesh, Physics Today 51 22 (1998). [ 1 8] J. F. Scott, Science 315 954 (2007). [ 1 9] R. W. Schwartz, Chem. Mater. 9 2325 (1997).

PAGE 143

143 [ 20 ] R. J. Ong, T. A. Berfield, N. R. Sottos, and D. A. Payne, J. Eur. Ceram. Soc. 25 2247 (2005). [ 21 ] Y. L. Tu, M. L. Calzada, N. J. Phillips, and S. J. Milne, J. Am. Ceram. Soc. 79 441 (1996). [ 22 ] T. A. Berfield, R. J. Ong, D. A. Payne and N. R. Sottos, J. Appl. Phys. 101 024102 (2007). [ 23 ] S. Hiboux P. Muralt, and T. Maeder, J. Mater. Res. 14 4307 (1999). [ 24 ] R. A. Wolf and S. Trolier McKinstry, J. Appl. Phys. 95 ,1397 (2004). [ 25 ] G. W. Scherer, J. Sol Gel Sci. Techn. 8 353 (1997). [26] A. Kamyshny, J. Steinke, and S. Magdassi The Open Applied Ph ysics Journal 4 19 (2011). [27] C. J. Brinker and G. W. Scherer, Sol gel Science: The Physics and Chemistry of Sol gel Processing, (Academic press, Boston, 1990). [28] P. Marchand, I. A. Hassan, I. A. Parkin, and C. J. Camalt, Dalton Trans. 42 9406 (2013 ). [ 29 ] K. D. Budd, S. K. Dey, and D. A. Payne, Br. Ceram. Proc. 36 107 (1985). [ 30 ] Ferroelectric Thin Films Vols. 1 5, Mater. Res. Soc. Symp. Proc., 200 243 310 361 and 433 (1990 1996). [ 31 ] Integrated Ferroelectrics, Proc. of the International Symp osia on Integrated Ferroelectrics, 1 16 (1990 1997). [ 32 ] J. T. Evans, W. D. Miller, and J. A. Bullington. Ferroelectrics: A New Non Volatile Technology, (Krysalis Microelectronics Corp. Product Literature, 1987). [ 33 ] G. L. Brennecka, J. F. Ihlefeld, J. P Maria, B. A. Tuttle, and P. G. Clem, J. Am. Ceram. Soc. 93 3935 (2010). [ 34 ] R. W. Schwartz, J. A. Voigt, B. A. Tuttle, D. A. Payne, T. L. Reichert, and R. S. DaSalla, J. Mater. Res. 12 444 (1997). [35 ] R. W. Schwartz, T. Schneller, and R. Waser, C. R. Chim. 7 433 (2004). [ 36 ] R. W. Schwartz, T. J. Boyle, S. J. Lockwood, M. B. Sinclair, D. Dimos, and C. D. Buchheit, Int. Ferro. 7 259 (1995). [ 37 ] K. D. Budd and D. A. Payne, Institute of Physics Conference Series 13 (1989).

PAGE 144

144 [ 38 ] K. Nittala, In situ X ray Diffraction based Investigation of Crystallization in Solution Deposited PZT Thin Films, Ph. D dissertation, University of Florida, (2012). [ 39 ] G. H. Yi, Z. Wu, and M. Sayer, J. Appl. Phys. 64 2717 (1998). [ 40 ] T. J. Boyle, D. Dimos, R. W. Schwartz, T. M. Alam, M. B. Sinclair and C. D. Buchheit, J. Mater. Res. 12 1022 (1997). [ 41 ] R. W. Schwartz, Ph.D. Thesis, University of Illinois, (1989). [ 42 ] F. F. Lange, Science 273 903 (1996). [ 43 ] K. G. Brooks, I. M. Reaney, R. Klissurska, Y. Huang, L. Bursil l, and N. Setter, J. Mater. Res. 9 2540 (1994). [ 44 ] K. Nittala, S. Mhin, K. M. Dunnigan, D. S. Robinson, J. F. Ihlefeld, P. G. Kotula, G. L. Brennecka, and J. L. Jones, J. Appl. Phys. In press, (2013). [ 45 ] M. Ohring, Materials Science of Thin Films, (Ac ademic Press, San Diego, 2002) p. 783. [ 46 ] S. Y. Chen and I. W. Chen, J. Am. Ceram. Soc. 77 2337 (1994). [ 47 ] K. Nittala, G. L. Brennecka, B. A. Tuttle, and J. L. Jones, J. Mater. Res. 46 2148 (2011). [ 48 ] E. K. F. Dang and R. J. Gooding, Phys. Rev. Let t. 74 3848 (1995). [ 49 ] E. M. Griswold, L. Weaver, M. Sayer, and I. D. J. Calder. Mater. Res. 10 3149 (1995). [ 50 ] C. K. Kwok and S. B. Desu, J. Mater. Res. 9 (7), 1728 (1994). [ 51 ] A. P. Wilkinson, J. S. Speck, A. K. Cheetham, S. Natarajan, and J. M. Tho mas, Chem. Mater. 6 750 (1994). [ 52 ] C. D. E. Lakeman, Z. K. Xu, and D. A. Payne, J. Mater. Res. 10, 2042 (1995). [ 53 ] M. A. Subramanian, G. Aravamudan, and G. V. S. Rao, Prog. Solid State Chem. 15 55 (1983). [ 54 ] L. A. Bursill and K. G. Brooks, J. Appl. Phys. 75 4501 (1994). [ 55 ] I. Barin, Thermochemical Data of Pure Substances, (Wiley VCH, Weinheim,1995). [ 56 ] A. D. Polli, F. F. Lange, and C. G. Levi, J. Am. Ceram. Soc., 83 873 (2000). [57] C. Martin, The Surface Texture Book, (Thames and Hudson, Lond on, 2005), p. 256.

PAGE 145

145 [58] U. F. Kocks, T exture and Anisotropy: Preferred Orientations in Polycrystals and Their Effect on Materials Properties (Cambridge University Press., Cambridge, 2000), p. 676. [ 59 ] B. A. Tuttle, J. A. Voigt, T. J. Garino, D. C. Goodno w, R. W. Schwartz, D. L. Lamppa, T. J. Headley, and M. O. Eatough, Eight IEEE International Symposium on Applications of Ferroelectrics, Greenville, SC, USA (1992). [ 60 ] R. W. Whatmore, Q. Zhang, C. P. Shaw, R. A. Dorey, J. R. Alcock, Physica Scripta T129 6 (2007). [ 61 ] S. Trolier McKinstry and P. Muralt, J. Electroceram. 12 7 (2004). [ 62 ] G. L. Brennecka, C. M. Parish, J. Jones, B. A. Tuttle, J. S. Wheeler, and J. G. Ekerdt, 14th US Japan Seminar on Dielectric and Piezoelectric Materials, Welches, Oregon, (2009) p. 136. [ 63 ] S. Y. Chen and I. W. Chen, J. Am. Ceram. Soc. 77 2332 (1994). [ 64 ] L. M. Sanchez, D. M. Potrepka, G. R. Fox, I. Takeuchi, and R. G.Polcawich, MRS Online Proc. Library, 1299, mrsf10 1299 s04 09 (2011). [ 65 ] P. Muralt, J. Appl. Phys. 10 0 051605 (2006). [ 66 ] P. Muralt, T. Maeder, L. Sagalowicz, S. Hiboux, S. Scalese, D. Naumovic, R. G. Agostino, N. Xanthopoulos, H. J. Mathieu, L. Patthey, and E. L. Bullock, J. Appl. Phys. 83 3835 (1998). [ 67 ] S. Y. Chen and I. W. Chen, J. Am. Ceram. So c. 81 97 (1998). [ 68 ] D. Kaewchinda, T. Chairaungsri, M. Naksata, S. J. Milne, and R. Brydson, J. Eur. Ceram. Soc. 20 1277 (2000). [ 69 ] Z. Huang, Q. Zhang, and R. W. Whatmore, J. Mat. Sci. Lett. 17 1157 (1998). [ 70 ] Z. Huang, Q. Zhang, and R. W. Whatmor e, J. Appl. Phys. 85 7355 (1999). [ 71 ] W. Gong, J. F. Li, X. C. Chu, Z. L. Gui, and L. T. Li, Acta Mater. 52 2787 (2004). [ 72 ] S. Y. Chen, Mater. Chem. Phys. 45 159 (1996). [ 73 ] G. R. Fox and S. B. Krupanidhi, J. Mater. Res. 9 699 (1994). [ 74 ] Y. M. Li u and P. P. Phule, J. Am. Ceram. Soc. 79 495 (1996). [ 75 ] G. J. Norga, F. Vasiliu, L. Fe, D. J. Wouters, and O. Van der Biest, J. Mater. Res. 18 1232 (2003).

PAGE 146

146 [ 76 ] J. A. Voigt, B. A. Tuttle, T. J. Headley, and D. L. Lamppa, Mat. Res. Soc. Symp. Proc., Bos ton (1995). [ 77 ] K. Sreenivas, I. Reaney, T. Maeder, N. Setter, C. Jagadish, and R. G. Elliman, J. Appl. Phys. 75 232 (1994). [ 78 ] T. Tani, Ph.D Thesis, University of Illinois at Urbana Champaign (1993). [ 79 ] T. Maeder, L. Sagalowicz, and P. Muralt, Jpn. J. Appl. Phys. Part 1 37 2007 (1998). [ 80 ] K. Aoki, Y. Fukuda, K. Numata, and A. Nishimura, J. Appl. Phys. Part 1 34 192 (1995). [81] Z. Song, J. Gao, X. Zhu, L. Wang, X. Fu, and C. Lin, J. Mater. Sci. 36 4285 (2001). [ 82 ] D. M. Potrepka, G. R. Fox, L. M Sanchez, and R. G. Polcawich, MRS Online Proc. Library, Material Research Society, 1299, mrsf10 1299 s04 02 (2011). [ 83 ] B. B. He, Two Dimensional X Ray Diffraction, (John Wiley & Sons, Inc., 2009) 426 p. [ 84 ] G. Tutuncu, Y. Chang, S. Poterala, G.L. Mess ing, and J. L. Jones, J. Am. Ceram. Soc. 95 (8), 2653 (2012). [85] MATLAB version 7.10.0. Natick, Massachusetts: The MathWorks Inc., ( 2010 ) [86] P. S. Prevy, Adv. X ray Anal. 29 103 (1986). [ 87 ] K. Nittala, S. Mhin, J. L. Jones, D. S. Robinson, J. F. Ih lefeld and G. L. Brennecka, J. Appl. Phys. 112 104109 (2012). [ 88 ] S. Sameshima, M. Kawaminami, and Y. Hirata, J. Ceram. Soc. Japan 110 597 (2002). [ 89 ] M. D. Losego, L. H. Jimison, J. F. Ihlefeld, and J. P. Maria, Appl. Phys. Lett. 86 172906 (2005). [ 90 ] M. D. Losego, J. F. Ihlefeld and J. P. Maria, Chem. Mater. 20 303 (2008). [ 91 ] G. L. Brennecka, C. M. Parish, B. A. Tuttle, L. N. Brewer and M. A. Rodriguez, Adv. Mater. 20 1407 (2008). [ 92 ] J. E. Daniels, J. L. Jones and T. R. Finlayson, J. Phys. D: Appl. Phys., 39 5294 (2006). [ 93 ] C.B. Eom and S. Trolier McKinstry, MRS Bull. 37 1007 (2012). [ 94 ] F. Calame and P. Muralt, Appl. Phys. Lett. 90 062907 (2007).

PAGE 147

147 [ 95 ] S. Hiboux, P. Muralt, and N. Setter, Proc. MRS, 596 499 (1999). [ 96 ] G. L. Smith, J. S. Pulskamp, L. M. Sanchez, D. M. Potrepka, R. M. Proie, T. G. Ivanov, R. Q. Rudy, W. D. Nothwang, S. S. Bedair, C. D. Meyer, and R. G. Polcawich, J. Am. Ceram. Soc. 95 1777 (2012). [97 ] A. T. Chien, J. Sachleben, J. H. Kima, J. S. Speck and F. F. Lange J Mate r. Res 14 3303 (1999). [ 98 ] J. Mayer, L. A. Giannuzzi, T. Kamino, and J. Michael, MRS Bull. 32 400 (2007). [ 99 ] D. Ambika, V. Kumar, K. Tomioka, I. Kanno, Adv. Mat. Lett. 3 102 (2012). [ 100 ] J. Ricote, R. Poyato, M. Alguero, L. Pardo, M. L. Calz ada, D. Chateigner, J. Am. Ceram. Soc. 86 1571 (2003). [ 101 ] R. Poyato, M. L. Calzada, J. Ricote, L. Pardo, and B. Willing, Int. Ferro. 35 1807 (2001). [ 102 ] B. Morosin, A. C. Lawson, G. H. Kwei, J. A. Voight, and B. A. Tuttle, Mater. Res. Soc. Symp. Pro c. (1994). [ 103 ] Y. Lin, B. R. Zhao, H. B. Peng, B. Xu, H. Chen, F. Wu, H. J. Tao, Z. X. Zhao, and J. S. Chen, Appl. Phys. Lett. 73 2781 (1998). [ 104 ] G. Sh irane and A. Takeda, J. P hys S oc Jpn 7 (1952). [ 105 ] B. Ja e, R. S. Roth, and S. Marzullo, J. A ppl. Phys. 25 809 (1954). [ 106 ] J. Zeng, M. Zhang, L. Wang, and C. Lin, J. Phys.: Condens. Matter 11 1139 (1999). [ 107 ] A. Bose, PZT Thin Film Growth and Chemical Composition Control on F lat and N ovel T hree D imensional M icromachined S tructures for MEMS D evices Ph. D dissertation, cole Polytechnique F ederale de L ausanne (2011). [ 108 ] S. Hiboux and P. Muralt, J Eur Ceram Soc 24 159 3 (2004)

PAGE 148

148 BIOGRAPHICAL SKETCH Sungwook Mhin was born in July of 1981 in Seoul, Korea. He received a Bachelor of Enginee ring in Ceramic from the Hanyang University in August of 2006. He then received a Master of Engineering in Materials from Hanyang University in August of 2008. His two years as a master student was full of travelling for research. He went to Osaka Universi ty for the analysis of alumina using high resolution TEM in 2006. Also, h e had an opportunity to work at National Institute of Advanced Industrial Science and Technology (AIST) in Tsukuba ( Japan ) for total 8 months in March of 2007 He enjoyed learning res earch skills as well as Japanese. Also, he learned understanding different opinions through the communication with people who have different cultural background. His precious experience at the AIST in Japan led him to make a decision to pursue PhD of Engin eering. He joined the Department of Materials Science and Engineering at the University of Florida (UF) as a PhD student in 2009 and immediately joined oup His life at the UF was full of happiness. Also, he made precious friends around the world who have sunny smile like Florida. During his life at UF, he travelled across the states to perform his experiments and analyses at different national laboratory His precious experience to handle state of the art XRD machines at different national laboratory led him to understand microscopic view of ceramic materials during synthesis After his beautiful life for four years as a PhD student, h e received hi s PhD from U F on August of 2013