USF Libraries
USF Digital Collections

Growth and characterization of epitaxial thin films and multiferroic heterostructures of ferromagnetic and ferroelectric...

MISSING IMAGE

Material Information

Title:
Growth and characterization of epitaxial thin films and multiferroic heterostructures of ferromagnetic and ferroelectric materials
Physical Description:
Book
Language:
English
Creator:
Mukherjee, Devajyoti
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Pulsed Laser Deposition
Magnetic Anisotropy
Spintronics
Dilute Magnetic Semiconductor
Cobalt Ferrite
CoFe2O4
Lead Zirconium Titanate
PZT
Manganese or Vanadium Doped Zinc Oxide
ZnO:Mn
ZnO:V
LSMO
RKKY
Dissertations, Academic -- Physics -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Multiferroic materials exhibit unique properties such as simultaneous existence of two or more of coupled ferroic order parameters (ferromagnetism, ferroelectricity, ferroelasticity or their anti-ferroic counterparts) in a single material. Recent years have seen a huge research interest in multiferroic materials for their potential application as high density non-volatile memory devices. However, the scarcity of these materials in single phase and the weak coupling of their ferroic components have directed the research towards multiferroic heterostructures. These systems operate by coupling the magnetic and electric properties of two materials, generally a ferromagnetic material and a ferroelectric material via strain. In this work, horizontal heterostructures of composite multiferroic materials were grown and characterized using pulsed laser ablation technique. Alternate magnetic and ferroelectric layers of cobalt ferrite and lead zirconium titanate, respectively, were fabricated and the coupling effect was studied by X-ray stress analysis. It was observed that the interfacial stress played an important role in the coupling effect between the phases. Doped zinc oxide (ZnO) heterostructures were also studied where the ferromagnetic phase was a layer of manganese doped ZnO and the ferroelectric phase was a layer of vanadium doped ZnO. For the first time, a clear evidence of possible room temperature magneto-elastic coupling was observed in these heterostructures. This work provides new insight into the stress mediated coupling mechanisms in composite multiferroics.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Devajyoti Mukherjee.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004732
usfldc handle - e14.4732
System ID:
SFS0028034:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Growth and Characterization of Epitaxial Thin Films and Multiferroic Heterostructures of Ferromagnetic and Ferroelectric Materials by Devajyoti Mukherjee A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Physics College of Arts and Science University of South Florida Co-Major Professor: Sarath Witanachchi, Ph. D. Co-Major Professor: Pritish Mukherjee, Ph. D. Hariharan Srikanth, Ph. D. George S. Nolas, Ph. D. Lilia M. Woods, Ph. D. Date of Approval: September 8, 2010 Keywords: Pulsed Laser Deposition, Magneti c Anisotropy, Spintronics, Dilute Magnetic Semiconductor, Cobalt Ferrite, CoFe2O4, Lead Zirconium Titanate, PZT, Manganese or Vanadium D oped Zinc Oxide, ZnO:Mn, ZnO:V, LSMO, RKKY Copyright 2010, Devajyoti Mukherjee

PAGE 2

DEDICATION I dedicate this dissertation to my pare nts, Mr. Soumendra Nath Mukherjee and Mrs. Gouri Mukherjee, and my grandfat her, Late Benoy Krishna Chakraborty.

PAGE 3

ACKNOWLEDGEMENTS I would like to express my deepest gr atitude to my advisors, Dr. Sarath Witanachchi and Dr. Pritish Mukherjee, for supervising me throughout the years. I am indebted to them for numerous stimul ating discussions, critical comments, encouragement during the entire course of this work and for the inte rest they have shown in my academic and personal well-being. I would like to express my thanks to my committee members Dr. George S. Nolas, Dr. Lilia Woods and Dr. Hariharan Srikanth for serving on my dissertation committee. I would especially like to th ank Dr. Hariharan Srikanth for allowing me to use the magnetic measurement system at the Functio nal Materials Laborat ory (FML). I would also like to thank Marienette Morales a nd Dr. Manh-Huong Phan for their expertise during the measurements. I am also grateful to Dr. Srikanth for supplying the main components of the ferroelectric measurement sy stem that was assembled during this work. In this context, I would like to thank my colleagues, Robe rt Hyde, Dr. Tara Dhakal and Marek Merlak for their assistance during the set up process. During this work, I had a lot of handson experience in working with vacuum systems, pumps and lasers. I am indebted to Robert Hyde for patiently teaching me how to operate and maintain these systems. All th e dual laser deposited films have been grown under his guidance. The ICCD imag es included in this thesis have been captured by him.

PAGE 4

I am thankful to Dr. Tara Dhakal for his assistance in the magnetic and ferroelectric measurements; experiments on the V doped ZnO thin films and writing some of the manuscripts for publications. Life in the lab would have been definite ly mundane had it not been shared with my colleagues. To start with, I must thank Robert Hyde, Marek Merlak, Dr. Tara Dhakal, Ted Wangensteen, Dr. Gayan Dedigamuwa a nd Jason Rejman. I thank them for being around all the time, be it while partying or discussing academics or making fun. They have been an indispensable part since the time I joined the lab and shall continue to be so always. I would like to thank all the past and present staff at the Physics department for their assistance throughout the years. I would es pecially like to thank Mary Ann Prowant, Daisy Matos, Kimberly Carter, Su e Wolfe, and Evelyn Keeton-Williams. I am grateful to my uncle and aunt, Dr. Arunava Mukherjea and Dr. Swapna Mukherjea, for financially and emotionally supporting me throughout the years. I would like to thank my friend and co lleague, Dennis Pliutau, who has been an understanding room mate throughout the gradua te years. I would like to thank Dr. Anuja Datta who has been a great inspiration in my life. This work was supported in part by the National Science Foundation under the grants DMI-0217939, DMI-0078917 and by the Department of Defense under the grant W81XWH-07-1-0708.

PAGE 5

i TABLE OF CONTENTS LIST OF TABLES.............................................................................................................iv LIST OF FIGURES..........................................................................................................vii ABSTRACT.....................................................................................................................xx i CHAPTER 1: INTRODUCTION........................................................................................1 1.1. Materials Overview...........................................................................................3 1.1.1. Magnetic Materials............................................................................3 1.1.1.1. Spinel-type Ferrite: CoFe2O4..............................................3 1.1.1.2. Magnetic Anisotropy..........................................................6 1.1.2. Ferroelectric Materials.......................................................................8 1.1.2.1. Ferroelectric Perovskite: PZT.............................................9 1.1.3. Multiferroic Materials......................................................................11 1.1.3.1. Scarcity of Single Phase Materials...................................11 1.1.3.2. Magnetoelectric Effect......................................................12 1.2. Thin Film Multiferroics..................................................................................14 1.2.1. Single Phase Multiferroic Thin Films..............................................14 1.2.2. Composite Multiferroic Thin Films.................................................15 1.2.2.1. Horizontal Heterostructures..............................................17 1.2.2.2. Vertical Nanostructures....................................................18 1.2.2.3. Nano-particle Embedded Films........................................19 1.3. Chapter Summary...........................................................................................20 1.4. Dissertation Outline........................................................................................21 CHAPTER 2: METHODOLOGY.....................................................................................22 2.1. Pulsed Laser Deposition of Thin Films..........................................................23 2.2. Structural Characterization.............................................................................29 2.2.1. X-ray Diffraction.............................................................................29 2.2.2. Scanning Electron Microscopy........................................................41 2.2.3. Energy Dispersive X-Ray Spectroscopy..........................................43 2.2.4. Atomic Force Microscopy...............................................................44 2.3. Electrical Characterization..............................................................................46 2.4. Magnetic Characterization..............................................................................47 2.5. Ferroelectric Characterization.........................................................................49 2.6. Chapter Summary...........................................................................................53

PAGE 6

ii CHAPTER 3: CFO-PZT BI LAYER THIN FILMS..........................................................54 3.1. Cobalt Ferrite (CFO) Thin Films....................................................................55 3.1.1. Experimental Details........................................................................56 3.1.2. Results and Discussions...................................................................56 3.1.2.1. Structural Properties..........................................................56 3.1.2.1.1. Epitaxial CFO thin films....................................57 3.1.2.1.2. Residual Stress in CFO thin films......................61 3.1.2.2. Magnetic Properties..........................................................64 3.1.3. Conclusions......................................................................................69 3.2. Lead Zirconate Titanate (PZT) Thin Films.....................................................70 3.2.1. Experimental Details........................................................................71 3.2.2. Results and Discussions...................................................................71 3.2.2.1. Laser-target Interac tions and Plume Diagnostics.............72 3.2.2.1.1. Single KrF laser ablation...................................73 3.2.2.1.2. Dual (KrF and CO2) laser ablation.....................80 3.2.2.2. Structural Properties..........................................................90 3.2.2.2.1. LSMO/PZT/LSMO capacitor............................95 3.2.2.3. Ferroelectric Properties...................................................100 3.2.2.3.1. Fatigue Characterization..................................108 3.2.3. Conclusions....................................................................................110 3.3. CFO-PZT Bilayer Thin Films.......................................................................111 3.3.1. Experimental Details......................................................................111 3.3.2. Results and Discussions.................................................................113 3.3.2.1. Structural Properties........................................................113 3.3.2.2. Magnetic Properties........................................................124 3.3.2.3. Ferroelectric Properties...................................................129 3.3.3. Conclusions....................................................................................137 3.4. Chapter Summary.........................................................................................139 CHAPTER 4: DOPED ZINC OXIDE HETEROSTRUCTURES...................................140 4.1. Mn Doped ZnO Thin Films..........................................................................142 4.1.1. Experimental Details......................................................................143 4.1.2. Results and Discussions.................................................................144 4.1.2.1. Structural Properties........................................................144 4.1.2.2. Electrical Properties........................................................152 4.1.2.3. Magnetic Properties........................................................155 4.1.3. Theoretical Modeling.....................................................................160 4.1.3.1. RKKY exchange interaction...........................................161 4.1.3.2. Percolation of Bound Magnetic Polarons.......................168 4.1.4. Conclusions....................................................................................172 4.2. V Doped ZnO Thin Films.............................................................................173 4.2.1. Experimental Details......................................................................174 4.2.2. Results and Discussions.................................................................175 4.2.2.1. Laser-target Interac tions and Plume Diagnostics...........175 4.2.2.2. Structural Properties........................................................179

PAGE 7

iii 4.2.2.3. Electrical Properties........................................................181 4.2.2.4. Ferroelectric Properties...................................................182 4.2.3. Conclusions....................................................................................186 4.3. ZnO:Mn – ZnO:V Heterostructure...............................................................187 4.3.1. Experimental Details......................................................................187 4.3.2. Results and Discussions.................................................................188 4.3.2.1. Structural Properties........................................................188 4.3.2.2. Magnetic Properties........................................................190 4.3.3. Conclusions....................................................................................192 4.4. Chapter Summary.........................................................................................193 CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS...................................194 5.1. Oblique Angle Deposition of CFO Thin Films.............................................196 REFERENCES................................................................................................................203 APPENDIX A: POLYCRYSTALLINE CFO THIN FILMS..........................................218 APPENDIX B: LSMO ELECTRODES..........................................................................237 APPENDIX C: PLUME IMAGES..................................................................................246 APPENDIX D: PUBLICATIONS...................................................................................251 ABOUT THE AUTHOR...................................................................................END PAGE

PAGE 8

iv LIST OF TABLES Table 2.1. Summary of deposition conditions of a thin film named Sample A.................28 Table 3.1. Deposition conditions of CFO films on different substrates............................56 Table 3.2. Ratio of Co/Fe in CoFe2O4 target and thin films, obtained by EDS analysis.............................................................................................................57 Table 3.3. In plane and out of plane lattice parameters obtained from x-ray diffraction (XRD) peaks and the strain calculated using in plain lattice parameters. In plane st ress calculated using Young’s modulus Y= 1.5 x 1012 dyne/cm2. Anisotropy calculated using stress.......................................62 Table 3.4. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane and out-of-plane configurations for 200 nm thick CFO films on MgO and STO substrates. The symbols and denote the inplane and out-of-plane conf igurations respectively.........................................65 Table 3.5. Summary of maximum polarization (Pmax), remnant polarization (Pr), nominal switching voltage (Vc), coercive field (Ec) and leakage current density JL (A/cm2) for PZTSL and PZTDL films grown on MgO and STO substrates. Data measured at 9 V driving voltage at 1 Hz.....................105 Table 3.6. Deposition conditions of the layers in the CFO-PZT bilayer films that were grown on MgO and STO substrates......................................................112 Table 3.7. Deposition conditions of the laye rs in the CFO-PZT capacitor using LSMO top and bottom electrodes..................................................................113 Table 3.8. FWHM of rocking curves about CFO (400) plane, in-plane (a) and out-of-plane (a) lattice parameters obtai ned from XRD peaks, inplane ( ) and out-of-plane ( ) strains for CFO and CFO-PZT films on MgO and STO substrates..........................................................................119 Table 3.9. Summary of saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms and the coercivity (Hc) for CFO/PZT bilayer

PAGE 9

v and CFO single layer films measured at 300 K and 10 K. The in-plane and out-of-plane directions have been denoted by the symbols and respectively..............................................................................................125 Table 3.10. Summary of ma ximum polarization (Pmax), remnant polarization (Pr), nominal switching voltage (Vc), coercive field (Ec) and leakage current density JL (A/cm2) for CFO-PZT bilayer and PZT single layer films grown on MgO and STO substrates. Data measured at 9 V driving voltage at 1 Hz. .............................................................................................135 Table 3.11. Summary of crysta l structure of substrates and electrodes used for growth of bilayers an d single layers and lattice mismatches with respect to PZT................................................................................................137 Table 4.1. Deposition parameters for ZMO thin films. ZMO(600) and ZMO(10mT) are the same samples................................................................144 Table 4.2. FWHM of rocking curve about (002) plane of ZnO, FWHM of (002) peak of ZnO from -2 scan and average crysta llite size from Scherrer formula for ZMO films at vari ous growth temperatures...............................147 Table 4.3. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:Mn thin films on c-cut sapphire substrates grown at various growth temperatures at constant background oxygen pressure..........................................................................153 Table 4.4. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:Mn thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure.................154 Table 4.5. Magnetic properties at 10K and 300K for ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) films, respectively..............................................157 Table 4.6. Magnetic properties at 10K and 300K for ZMO(0m T), ZMO(10mT), ZMO(50mT) and ZMO(300mT) films, respectively.....................................159 Table 4.7. Summary of RKKY parameters for ZMO films, deposited at various temperatures...................................................................................................165 Table 4.8. Summary of RKKY parame ters for ZMO films under varying background pressures.....................................................................................168 Table 4.9. Deposition parameters for ZVO thin films. ...................................................175

PAGE 10

vi Table 4.10. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:V thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure................182 Table 4.11. Summary of polarization data for ZnO:V thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure................183 Table 4.12. Summary of polarization valu es for ZVO(500mT) film using LSMO electrodes.......................................................................................................186 Table 4.13. Deposition conditions of the layers in the ZnO:Mn-ZnO:V heterostructure on c-cu t sapphire substrate....................................................188 Table A.1. Lattice parameter (a ) and crystallite size (D) calculated using Scherrer formula from the XRD patterns for CFO target and the deposited films on Si and sapphire substrates.........................................................................222 Table A.2. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane and out-of-plane c onfigurations for 100 nm and 50 nm thick CFO films on Si substrates and 200 nm thick film of CFO on sapphire substrates. The symbols and denote the in-plane and outof-plane configurations respectively..............................................................224 Table A.3. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane configurations fo r 200nm, 100 nm and 50 nm thick CFO films on Si substrates. The symbol denotes the in-plane configurations................................................................................................227

PAGE 11

vii LIST OF FIGURES Figure 1.1.1. Typical hysteres is loop of a ferromagnetic material showing the saturation magnetization (Ms), remnant magnetization (Mr) and coercive field (Hc)..............................................................................................4 Figure 1.1.2. Schematic diagram of unit cell and subunits in CoFe2O4 inverse spinel structure (Adapt ed from Ref. [32])..........................................................5 Figure 1.1.3. (a) Ideal M H loop and (b) schematic diagram of magnetic moments for a thin film with the easy axis of magnetization along the film plane, respectively......................................................................................7 Figure 1.1.4. Typical hysteresi s loop of a ferroelectric material showing the saturation polarization (Ps), remnant polarization (Pr) and coercive field (Hc)............................................................................................................9 Figure 1.1.5. Schematic diagram of PZT un it cell showing off-centering of Zr/Ti ion under an external applied electr ic field (Adapted from C. Kittel [45])..................................................................................................................10 Figure 1.1.6. Schematic diagram showi ng the overlap required to achieve multiferroic properties in a single mate rial (Adapted from Ref. [47])............12 Figure 1.2.1. Schematic diagram of strain -mediated ME effect in a composite system consisting of ferromagnetic layer (pink) and ferroelectric layer (blue) (Adapted from Ref. [15]).......................................................................16 Figure 1.2.2. Schematic diagram of the three main kinds of ME composite nanostructures (a) horizontal he terostructures, (b) vertical nanostructures, and (c) nano-partic le embedded films, respectively. (Adapted from Ref [15])..................................................................................17 Figure 2.1.1. Schematic diagram of th e single laser deposition set up..............................23 Figure 2.1.2. Schematic diagram of the dual laser deposition set up.................................26 Figure 2.1.3. Photograph of the custom built PLD system at LAMSAT...........................27

PAGE 12

viii Figure 2.2.1. X-ray reflection from latti ce planes illustrating Bragg’s law.......................30 Figure 2.2.2. Schematic diagram of crystallite s in a (a) polycrystalline film with random orientation, (b) textured film with preferred weak orientation, (c) epitaxial film with orientation, respectively. (Adapted from Ref. [78])..................................................................................................................31 Figure 2.2.3. Photograph of Bruker D8 X-ray Diffractometer at LAMSAT showing the individual parts............................................................................32 Figure 2.2.4. Schematic diagram of X-ray reflections from the parallel crystal planes of a (100) oriented film grown on a (100) single crystal substrate. .........................................................................................................33 Figure 2.2.5. Schematic diagram of asymmetr ic scan or detect or scan performed about the (111) plane of an epitaxi al (100) oriented film grown on single crystal (100) substrate...........................................................................34 Figure 2.2.6. Schematic diagrams of film s grown under compressive and tensile strains...............................................................................................................35 Figure 2.2.7. (a) Photograph of Philips X’ Pert X-ray Diffractometer. (b) Schematic diagram showing the angles and with respect to the thin film sample and the incident X-ray beam.................................................37 Figure 2.2.8. Schematic diagram of the X -ray rocking curve measurement where the sample is tilted (rocked) a bout the plane of orientation.............................39 Figure 2.2.9. Schematic diagram of scan performed to determine the cubic symmetry of a (100) oriented thin film............................................................40 Figure 2.2.10. (a) Schematic diagram of electron gun column. (b, c, and d) Photographs of the JOEL SEM at LAMSAT with the major components identified......................................................................................42 Figure 2.2.11. (a) Photograph of the Oxford Instruments EDS detector that was attached to JOEL SEM. (b) Schematic diagram of signals generated when electron beam from SEM interacts with the sample..............................44 Figure 2.2.12. (a) Schematic diagram of the basic working principle of a typical AFM. (b) Photograph of Digita l Instruments III Atomic Force Microscope. .....................................................................................................45 Figure 2.3.1. Schematic diagrams of Van de r Pauw technique of thin film (a) resistivity and (b) Hall meas urements, respectively........................................46

PAGE 13

ix Figure 2.4.1. Photograph of the magnetic measure system, PPMS from Quantum Design..............................................................................................................47 Figure 2.5.1. Photographs of the home-m ade set-up used for ferroelectric measurements of thin films..............................................................................50 Figure 2.5.2. Block diagram of Sawyer-T ower circuit used for hysteresis measurement of ferroelectric thin film............................................................51 Figure 2.5.3. (a) Photograph of the standard FE capacitor used for calibration of the ferrotestor. (b) Schematic diagram of the cross-section of the standard FE capacitor.......................................................................................52 Figure 3.1.1. SEM images of CFO target surface..............................................................57 Figure 3.1.2. (a) AFM image of CFO film on MgO substrate. (b) Section analysis of CFO film on MgO. Ho rizontal and vertical distances between red markings are 3.008 m and 1.288 m, respectively........................................58 Figure 3.1.3. XRD patterns of (a) CFO pow der target, and epitaxial CFO films grown on (b) STO (100), and (c) MgO (100) substrates, respectively. The inset to (c) shows the details of MgO (200) (left) and CFO (400) (right) peaks.....................................................................................................59 Figure 3.1.4. Left column of graphs (blue) (a, b, c) represent the film grown on STO and the right column (red) (d, e, f) represents films grown on MgO. (a, and d) scan spectra from (311) CFO reflection. (b, and e) Rocking curves of CFO (400) peaks. (c, and f) Asymmetric scans of the (511) and (440) planes of the CFO films. .................................................60 Figure 3.1.5. Schematic diagrams show th e (a) unstressed CFO bulk powder (b) CFO lattice under in-plane tensile stress on MgO substrate with a > a and (c) CFO lattice under high in -plane compressive stress causing tetragonal distortion a> a..............................................................................63 Figure 3.1.6. XRD -2 scan about 2 = 43.006o by varying the angle from 0 to 15o keeping the at 0o on the CFO films on MgO (100) substrate. The legend shows the values of sin2 The inset shows the plot of d vs. sin2 and a linear fit to the data points............................................................64 Figure 3.1.7. (a, and c) M-H loops measured at 300 K and 10 K respectively of the 200 nm thin film grown on STO ( 100) for in plane and out of plane configuration. (b, and d) M-H lo ops measured at 300 K and 10 K respectively of the 200 nm thin film grown on MgO (100) for in plane and out of plane configurations........................................................................67

PAGE 14

x Figure 3.2.1. XRD patterns of stoichiometr ic PZT target and PZT target with excess PbO.......................................................................................................72 Figure 3.2.2. SEM images of the PZT target surface after irradi ation by 1000 laser pulses showing (a) the rect angular spot size, (b, a nd d) conical features near the center of the interaction site and (c) the unablated target surface, respectively.........................................................................................73 Figure 3.2.3 (a) SEM image of conical struct ures formed on a PZT target surface after repeated ablation by KrF excimer laser. (b) Histogram of at. % ratios of Pb/(Zr+Ti), Zr/(Zr+Ti) an d Ti/(Zr+Ti) at random locations on the cone tips and bodies...................................................................................74 Figure 3.2.4. SEM images of PZT target after irradiation by 1000 pulses of a KrF laser beam in 500 mT O2 ambient at laser fluences of (a) 1J/cm2, (b) 2J/cm2, (c) 3J/cm2, (d) 4J/cm2, (e) 5J/cm2, and (e) 6J/cm2, respectively, denoted as KrF 1J/cm2 to 5J/cm2.....................................................................76 Figure 3.2.5. Chemical compositions using EDS analysis of the ablated target surfaces as a function of KrF laser fluence for (a) stoichiometric PZT and (b) PZT (30 at.% PbO) target....................................................................77 Figure 3.2.6. Chemical compositions using EDS analysis of PZT films deposited on Si substrates at room temperature by varying the ambient O2 pressures...........................................................................................................78 Figure 3.2.7. (a) Schematic diagram of the arrangement of target and the ablated plume as viewed by the ICCD camera. (b) ICCD image the particulates ejected from the PZT (30 at. % excess Pb) target surface during ablation using high KrF fluence of 5 J/cm2. ........................................79 Figure 3.2.8. ICCD images of total visible emission spectra of single laser plumes varying the KrF flue nces as (a) 1 J/cm2 (b) 2 J/cm2 (c) 3 J/cm2, and (d) 4 J/cm2 under 500 mT pO2...............................................................................80 Figure 3.2.9. ICCD images of total visible emission spectra of CO2 laser ablated plumes with increasing CO2 fluences from 1 J/cm2 to 3 J/cm2........................81 Figure 3.2.10. KrF and CO2 pulse waveforms at (a) 250 ns, (b) 145 ns, (c) 100 ns, (d) 50 ns, (e) 0 ns, and (f) – 50 ns of peak-to-peak inter-pulse delays ( t), respectively..............................................................................................82 Figure 3.2.11. SEM images of laser-target in teraction sites on the stoichiometric PZT target using KrF (3 J/cm2) and CO2 (2 J/cm2) lasers by varying the peak-to-peak inter-pulse delay ( t) as (a) 250 ns (b) 145 ns, (c) 100 ns, (d) 50 ns, (e) 0 ns, and (f) -50 ns, respectively....................................84

PAGE 15

xi Figure 3.2.12. Chemical compositions using EDS analysis of the ablated stoichiometric PZT target surfaces using KrF (3 J/cm2) and CO2 (2 J/cm2) lasers by varying the peak-t o-peak inter-pulse delay ( t)....................85 Figure 3.2.13. SEM images of laser-target in teraction sites on the stoichiometric PZT target using CO2 fluence at 2 J/cm2 but varying the KrF fluence as (a) 1 J/cm2, (b) 2 J/cm2, (c) 3 J/cm2 and (d) 4 J/cm2, respectively keeping the peak-to-peak inter-pulse delay ( t) at 50 ns.................................86 Figure 3.2.14. SEM images of laser-target in teraction sites on the stoichiometric PZT target using CO2 fluence at 2 J/cm2 but varying the KrF fluence as (a) 1 J/cm2, (b) 2 J/cm2, (c) 3 J/cm2 and (d) 4 J/cm2, respectively keeping the peak-to-peak inter-pulse delay ( t) at 100 ns...............................86 Figure 3.2.15. Chemical compositions using EDS analysis of the ablated PZT (30 at. % excess PbO) target surfaces using CO2 fluence at 2 J/cm2 but varying the KrF fluence at the peak-to-peak delay ( t) of (a) 50 ns and (b) 100 ns, respectively.............................................................................87 Figure 3.2.16. Chemical compositions using EDS analysis of the ablated PZT target surfaces with excess PbO, us ing dual and single laser ablations...........88 Figure 3.2.17. ICCD images of total visibl e emission spectra of dual laser plumes varying the KrF flue nces as (a) 1 J/cm2 (b) 2 J/cm2 (c) 3 J/cm2, and (d) 4 J/cm2 but keeping the CO2 fluence at 2 J/cm2 and the inter-pulse delay at 100 ns under 500 mT pO2...................................................................89 Figure 3.2.18. Time integrated ICCD images of the visible emission spectra of laser-induced plumes in 500 mTorr ambient O2 for (a) dual-laser ablation using KrF and CO2 laser fluences of 3 J/cm2 and 2 J/cm2 respectively, with the inter-pulse pe ak-to-peak delay of 100 ns, and (b) the KrF single-laser ablation with a fluence of 3 J/cm2. The scale bar represents 10 mm.............................................................................................90 Figure 3.2.19. XRD scans of PZTSL films deposited using KrF fluence of (a) 2 J/cm2 and (b) 5 J/cm2 and a (c) PZTDL film using KrF 1 J/cm2 and CO2 2 J/cm2 ( t = 100 ns) on STO substrates respectively. The substrate peaks are denoted by *.....................................................................................91 Figure 3.2.20. Rocking curves about PZT (200) plane for single and dual laser grown PZT films on STO substrates................................................................92 Figure 3.2.21. AFM images of PZT films de posited using single laser ablation (a) at 2 J/cm2 and (b) at 5 J/cm2 and (c) dual laser ablation. All scan areas are 5 m x 5 m. .............................................................................................93

PAGE 16

xii Figure 3.2.22. SEM images of PZTSL films deposited at 550 C on STO substrates at (a) low fluence of 2 J/cm2 and (b) high fluence of 5 J/cm2, and (c) PZTDL film, respectively. (d) SEM imag e of details of one of the particulates on PZTSL film at 2 J/cm2..............................................................94 Figure 3.2.23. Schematic diagram of PZT th in film capacitor fabricated using LSMO top and bottom electrodes....................................................................95 Figure 3.2.24. XRD -2 scans of LSMO/PZT/LSMO capacitors on STO (100) substrates grown using (a) dual and (b) single-laser ablations, respectively. Inset (I) shows the de tails of STO (200) and LSMO (200) peaks around 2 values of 46 Inset (II) shows the rocking curve of the PZT (100) peak with FWHM value of 0.1 Inset (III) shows the rocking curve of the PZT (100) peak with the FWHM value of 0.5 The PZT ( l 00) (l = 1, 2, 3) reflections are denoted by , and the LSMO/STO (00l) peak s are denoted by *.......................................................96 Figure 3.2.25. AFM images of the PZT surface for the LSMO/PZT/LSMO capacitors on STO substrates deposited using (a) dual-laser ablation and (b) single-laser ablation. The Rrms surface roughness values for (a) and (b) are 1.6 nm and 11.5 nm, respectively..................................................97 Figure 3.2.26. XRD -2 scans of LSMO/PZT/LSMO capacitors on MgO (100) substrates grown by (a) dual and (b) single-laser ablation, respectively. Insets (I) and (II) in Figure 3.2.30 (a) and (b) show the details of PZT(100)/LSMO(100) peaks and MgO(200)/PZT(200)/LSMO(200) peaks, respectively. The small peaks denoted by are artifacts from the MgO substrates..........................................................................................98 Figure 3.2.27. XRD rocking curves a bout the PZT (100) plane for the LSMO/PZT/LSMO capacitors on MgO (100) substrates grown by (a) dual and (b) single-laser ablation, respectively................................................99 Figure 3.2.28. AFM images of the PZT surface for the LSMO/PZT/LSMO capacitors on MgO substrates depo sited using (a) dual-laser ablation and (b) single-laser ablation, respectively. The Rrms surface roughness values for (a) and (b) are 16.7 nm and 22.6 nm, respectively........................100 Figure 3.2.29. (a) P-E hysteresis loop and (b) XRD pattern for PZTSL film on LSMO/Si(100) substrate, respectively. .........................................................101 Figure 3.2.30. P-E hysteresis loops for PZTSL films at KrF fluences of (a, b) 2 J/cm2, (c, d) 3 J/cm2, and (e, f) 5 J/cm2 on STO and MgO substrates, respectively, measured at 9V driving voltage and varying the hysteresis period from 100 to 1000 ms. ........................................................103

PAGE 17

xiii Figure 3.2.31. P-E hysteresis loops for PZTSL and PZTDL films on STO and MgO substrates, respectively, at vari ous maximum driving voltages.....................104 Figure 3.2.32. P-E hysteresis loops measured with 9V driving voltage for PZTDL and PZTSL films on MgO substrates. The inset shows a schematic illustration of the capacitors with the thicknesses of LSMO and PZT layers being 100 nm and 500 nm, respectively. ............................................106 Figure 3.2.33. P-E hysteresis loops meas ured with 9V driving for PZTDL and PZTSL films on STO substrates. Th e inset shows a schematic illustration of the capacitors with the thicknesses of LSMO and PZT layers being 100 nm and 500 nm, respectively. ............................................107 Figure 3.2.34. Capacitor leak age current densities (JL) measured using a stress voltage of 9 V for a soak time of 1000 ms for PZTDL and PZTSL films grown on (a) MgO and (b) STO subs trates under same conditions, respectively....................................................................................................108 Figure 3.2.35. Results of positive-up negative-down (PUND) fatigue tests at 10 kHz using +/9 V read voltage s for LSMO/PZT/LSMO capacitors grown by dual-and single-laser ablation on (a) STO (b) MgO substrates. ......................................................................................................109 Figure 3.3.1. Schematic diagram of CFO-P ZT bilayer films deposited on MgO or STO substrates...............................................................................................111 Figure 3.3.2. Schematic diagram of the CFO-PZT bilayer capacitor fabricated using LSMO top and bottom electrodes........................................................112 Figure 3.3.3. SEM images of the PZT surf ace in CFO-PZT bilayer film grown on Si substrate.....................................................................................................113 Figure 3.3.4. SEM images of CFO-PZT b ilayer film grown on MgO substrate (cross-sectional view). ..................................................................................114 Figure 3.3.5. XRD -2 scans for single layer CFO and bilayer CFO-PZT films grown on MgO (a, and b) and STO (c and d) substrates, respectively. The insets to (a) and (b) show the details of the MgO(200), CFO (400) and PZT (200) peaks around 43 (2 )............................................................115 Figure 3.3.6. Left and right columns repr esent the films grown on STO and MgO substrates respectively. (a and b) are scan spectra from PZT (101) reflection in CFO-PZT bilayer film (c and d) and (e and f) are scan spectra from (311) CFO reflectio n in CFO-PZT bilayer and single layer CFO films respectively. (g a nd h) are rocking curves of CFO (400) peaks. (i and j) and (k and l) are asymmetric scans of (511) and

PAGE 18

xiv (440) CFO planes for single layer CFO films and bilayer layer CFOPZT films respectively. (i – l) Le ft peaks are from CFO (511) plane and the right peaks are from CFO (440) planes.............................................117 Figure 3.3.7. AFM images of (a) CFO su rface of CFO/MgO film, (b) top PZT surface of PZT/CFO/MgO film, (c) 3D rendition of part (b), (d) CFO surface of CFO/STO film and (e) top PZT surface of PZT/CFO/STO films. Scan areas are 1x1 m with z-height of 100 nm.................................121 Figure 3.3.8. XRD -2 scans for the CFO-PZT bilayer grown on single crystal MgO (100) substrate with a conducting LSMO top and bottom electrode layers. The PZT ( l 00) where l = 1, 2, and 3 peaks are denoted by *. The LSMO ( l 00) where l = 1, and 2 peaks are denoted by The inset shows the details of MgO (200) and CFO (400) peaks.........................122 Figure 3.3.9. XRD -2 scans for the CFO-PZT bilayer grown on single crystal STO (100) substrate with a co nducting LSMO top and bottom electrode layers. The PZT ( l 00) where l = 1, 2, and 3 peaks are denoted by *. The LSMO ( l 00) where l = 1, 2, and 3 peaks are denoted by The inset shows the details of LS MO (200), STO (200), PZT (200) and CFO (400) peaks............................................................................................123 Figure 3.3.10. M-H loops measured at (a, a nd c) 10 K and (b and d) 300 K for the CFO-MgO and PZT-CFO-MgO films, respectively. The in plane and out of plane denote directions for th e magnetic field applied parallel or perpendicular to the film plane, respectively.................................................126 Figure 3.3.11. M-H loops measured at (a, a nd c) 10 K and (b and d) 300 K for the CFO-STO and PZT-CFO-STO films, respectively. The in plane and out of plane denote directions for th e magnetic field applied parallel or perpendicular to the film plane, respectively.................................................127 Figure 3.3.12. Ferroelectric hysteresis l oops for CFO-PZT bilayer film on MgO substrate at 9V driving voltage in the frequency range 1 Hz to 10 Hz..........128 Figure 3.3.13. Frequency dependen ce of remnant polarization (Pr) and nominal voltage for switching (Vc) for CFO-PZT bilayer on MgO substrate.............131 Figure 3.3.14. Ferroelectric hysteresis loops for CFO-PZT bilayer film on STO substrate at 9V driving voltage in the frequency range 1 Hz to 10 Hz..........132 Figure 3.3.15. Frequency dependen ce of remnant polarization (Pr) and nominal voltage for switching (Vc) for CFO-PZT bilayer on MgO substrate.............133

PAGE 19

xv Figure 3.3.16. Ferroelectric hysteresis loops for CFO-PZT bilayer and PZT single layer films grown on MgO substrates under same conditions. Data measured at 9V driving voltage.....................................................................134 Figure 3.3.17. Ferroelectric hysteresis loops for CFO-PZT bilayer and PZT single layer films grown on STO substrates under same conditions. Data measured at 9V driving voltage.....................................................................135 Figure 3.3.18. Leakage current density in CFO-PZT bilayer and PZT single layer thin film capacitors grown on STO substrates under same conditions. Inset shows a schematic diagram of CFO-PZT bilayer during the leakage current density measurement............................................................136 Figure 4.1.1. SEM images of the Mn doped ZnO target surface.....................................145 Figure 4.1.2. XRD patterns of (a) Mn doped ZnO target (b) MnO2 powder and (c) undoped ZnO powder, respectively...............................................................145 Figure 4.1.3. XRD patterns of (Zn0.98Mn0.02)O films on sapphire substrates grown at room temperature, 200 C, 400 C and 600 C with a background oxygen pressure of 10 mT named as ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) respectivel y. The sapphire substrate peaks have been denoted by *..................................................................................146 Figure 4.1.4. Rocking curves about the ( 002) plane of ZnO for ZnO:Mn films grown at 200 C, 400 C and 600 C with a pO2 of 10 mT named as ZMO(200), ZMO(400), and ZMO(600), respectively...................................148 Figure 4.1.5 SEM images of (a) undoped ZnO film grown at 600 C, (b) ZMO(600), (c) ZMO(400) and (d ) ZMO(200), Mn doped ZnO films on c-cut sapphire substrates...........................................................................149 Figure 4.1.6. AFM images of Mn doped ZnO thin films grown at (a) 600 C and (b) room temperature under pO2 of 10mT. Scan areas are 1 m x 1 m. (c) and (d) are 3D projections for (a) and (b), respectively...........................150 Figure 4.1.7. XRD patterns for Mn dope d ZnO thin films on c-cut sapphire substrates deposited at 600 C with varying oxygen background pressure..........................................................................................................151 Figure 4.1.8. SEM images of undoped ZnO f ilms grown on (a, b) c-cut sapphire substrates and (c, d) Si (100) substr ates. Inset to (c) shows the details on one of the hexagonal facets of ZnO on the film........................................152 Figure 4.1.9. Resistivity ( ) and carrier c oncentration (nc) versus temperature measured for ZMO(600). ..............................................................................154

PAGE 20

xvi Figure 4.1.10. Magnetization measurement at 10 K for ZMO powder that was prepared by grinding a piece cut fr om the corresponding target...................155 Figure 4.1.11. M-H loops measured at (a) 10 K and (b) 300 K for undoped ZnO and ZMO films on c-cut sapphire subs trates both grown under same conditions.......................................................................................................156 Figure 4.1.12. M-H loops measured at (a) 10 K and (b) 300 K for ZMO(600) when the magnetic field was applied parallel (in-plane) and perpendicular (out-of-plan e) to the film surface............................................157 Figure 4.1.13. M-H loops measured at (a) 10 K and (b) 300 K for ZMO films deposited on c-cut sapphire substrates at RT and 600 C with a constant pO2 of 10 mT named as ZMO(RT) and ZMO(600) respectively....................................................................................................158 Figure 4.1.14. M-H loops measured at (a) 10 K and (b) 300 K for ZMO films deposited on c-cut sapphire substrates at 600 C by varying the pO2 from 0 mT to 300 mT.....................................................................................159 Figure 4.1.15. Resistivity ( ) versus temperature dependence of (a) ZMO(RT) and (b) ZMO(600)..........................................................................................161 Figure 4.1.16. Plot of RKKY exchange integral JRKKY(r) as a function of the average separation between magnetic spins (r). The inset shows the range of r when JRKKY > 0..............................................................................162 Figure 4.1.17. Initial magnetizati on curves at 10 K for (Zn0.98Mn0.02)O films deposited at various growth temper atures with constant background O2 pressure. The inset shows the free-spin Brillouin function BS (10K) for S=5/2 at 10 K.................................................................................................164 Figure 4.1.18. M versus 1/H (kOe-1) plots at 10 K for (Zn0.98Mn0.02)O films deposited at various growth te mperatures with constant pO2. (b) The solid lines are the linear fits fo r the data points for films named ZMO(RT), ZMO(200), ZMO(400) and ZMO(600)......................................165 Figure 4.1.19. Initial magnetizati on curves at 10 K for (Zn0.98Mn0.02)O films deposited at 600 C but varying the background O2 pressure........................167 Figure 4.1.20. M versus 1/H (kOe-1) plots at 10 K for (Zn0.98Mn0.02)O films deposited at 600 C but varying the background O2 pressure. The solid lines are the linear fits for the data points for films named ZMO(0mT), ZMO(10mT), ZMO(50mT) and ZMO(300mT)............................................168

PAGE 21

xvii Figure 4.1.21. Variation of the radius of polarons Rp in angstroms at different temperatures for ZMO(RT) and ZMO(600), (Zn0.98Mn0.02)O films deposited at RT and 600 C respectively......................................................171 Figure 4.2.1. Schematic diagram of the unit cell and neighboring atoms in V doped ZnO crystal structure...........................................................................174 Figure 4.2.2. SEM images of (a) unablated ZVO target surface (b) ablated surface at 2 J/cm2 and (c) ablated surface at 5 J/cm2..................................................176 Figure 4.2.3. Atomic % obtained from EDS analysis for the ZVO target under various laser fluences. On x-axis 0 J/cm2 implies the unablated target.........177 Figure 4.2.4. ICCD images of total visible emission spectra of single laser plumes varying the excimer fluences (F) from 1 to 4 J/cm2 under different gas pressures (pO2)...............................................................................................178 Figure 4.2.5. ICCD images of total visi ble emission spectra of laser ablated plumes under different gas pressures (pO2) and constant fluence of 2 J/cm2...............................................................................................................178 Figure 4.2.6. XRD -2 scans (a) ZVO(500mT), (b) ZVO(300mT), and (c) ZVO(100mT) films, respectively. The substrate peak is denoted by *.........179 Figure 4.2.7. XRD rocking curves about ( 002) plane of ZnO for (a) ZVO(500mT), (b) ZVO(300mT), and (c) ZVO(100mT) films, respectively........................180 Figure 4.2.8. AFM images V-doped ZnO thin films grown at 600 C but varying the background O2 pressure from (a) 100 mT, (b) 300 mT, and (c) 500 mT, named as ZVO(100mT), ZVO(300mT), and ZVO(500mT) respectively....................................................................................................181 Figure 4.2.9. P-V loops for ZVO(100mT), ZVO(300mT), and ZVO(500mT) films, respectively. Inset shows a sche matic diagram of ZVO capacitor................183 Figure 4.2.10. Polarization loops for Z VO (500mT) film. The inset shows a schematic diagram of LSMO /ZVO(500mT)/LSMO capacitor......................184 Figure 4.2.11. (a)Polarization and (b) capacitance vs driving voltage for V-doped ZnO thin film grown pO2 of 500mT, named as ZVO(500mT), respectively....................................................................................................185 Figure 4.3.1. Schematic diagram of ZnO:Mn-ZnO:V heterostructure............................187 Figure 4.3.2. XRD -2 scan of ZnO:Mn-ZnO:V heterostructure. The c-cut sapphire substrate peak is denoted by *. .......................................................189

PAGE 22

xviii Figure 4.3.4. Rocking curves about the (002) plane of ZnO for the ZnO:MnZnO:V heterostructure and the ZnO:Mn single layer film............................189 Figure 4.3.5. M-H loops measured at (a) 10 K and (b) 300 K for the individual layers, ZnO:Mn and ZnO:V, and th e ZnO:Mn-ZnO:V heterostructure, all grown on c-cut sapphire susbtrat es. The magnetic field was applied parallel to the film plane................................................................................191 Figure 4.3.6. Magnetization loops at 300 K before and after poling the ZnO:Mn/ZnO:V epitaxial heterostructure.....................................................192 Figure 5.1.1. Schematic diagram of th e oblique angle deposition and the shadowing effect (Adapted from Ref. [190]).................................................196 Figure 5.1.2. Schematic diagram of the grow th of columnar structures by oblique angle deposition (Adapted from Ref. [194])..................................................197 Figure 5.1.3. XRD patterns of CF O thin films deposited on Si (100) substrates at an oblique incident angle of 60 with different thicknesses as (a) 600 nm, (b) 400 nm, (c) 200 nm and (d) 100 nm, respectively. The substrate peak is denoted by *.......................................................................198 Figure 5.1.4. XRD patterns of CFO thin films deposited on Si under same conditions using (a) oblique and (b) normal incidence deposition, respectively....................................................................................................199 Figure 5.1.5. AFM images of CFO films with different thicknesses as (a) 200 nm using normal, and (b) 100 nm, (c) 200 nm, (d) 300 nm, (e) 400 nm, and (f) 600 nm using oblique inci dence deposition, respectively..................201 Figure 5.1.6. M-H loops for CF O films on Si substrates measured at (a) 300 K and (b) 10 K grown using normal and oblique incidence depositions. The magnetic field has been applied in-plane.......................................................202 Figure A.1. SEM images of CFO films on Si (100) substrates deposited at 450 C varying the background O2 pressure as (a) 0 mT, (b) 10 mT, (c) 40 mT, and (d) 60 mT, respectively...........................................................................219 Figure A.2. XRD patterns for (a) 200 nm, (b ) 100 nm and (c) 50 nm thick films of CFO grown on Si (100) s ubstrate, respectively.............................................220 Figure A.3. XRD -2 scan of CFO film grown on c-cut sapphire substrate..................221 Figure A.4. M-H loops measured at (a) 300 K and (b) 10 K for 100 nm thick textured polycrystalline CFO film grown on Si (100) for in plane and out of plane configuration..............................................................................223

PAGE 23

xix Figure A.5. M-H loops measured at (a) 300 K and (b) 10 K for 50 nm thick highly (111) textured CFO film grown on Si (100) for in plane and out of plane configuration........................................................................................225 Figure A.6. M-H loops measured at (a) 300 K and (b) 10 K for 200 nm thick highly (111) textured CFO film gr own on c-cut sapphire substrate for in plane and out of plane configurations........................................................226 Figure A.7. M-H loops measured at (a) 300 K and (b) 10 K for CFO films grown on Si (100) for different thicknesse s. The magnetic field was parallel to film plane...................................................................................................228 Figure A.8. M-H loops measured at (a ) 10 K, (b) 150 K and (c) 300 K for a 200 nm thick CFO film grown on Si (100) substrate. (d) The coercivity (Hc) as a function of temperature...................................................................229 Figure A.9. In-plane M-H loops measured at 300 K and 10 K for 50 nm thick CFO film grown on Si (100) substrate. Steps seen in thicker films (100 nm and above) due to low field switch ing are not present in this case..........230 Figure A.10. In plane M-H loops for CF O-Si 50 nm and 200 nm films measured at 10 K. The y-axis is normalized with the saturation magnetic field. The M-H loop of the (111) oriented film overlap s on the first step of the magnetic reversal seen on the 200 nm film..............................................231 Figure A.11. TDO measurement of CFO films on STO at 300 K with the film plane perpendicular to the DC fi eld. The line with solid square represent the field sweep from positive to negative, where as the solid circles represent the field sweep from negative to positive...........................234 Figure B.1. XRD patterns of (a) LSMO powder, and LSMO films on MgO (100) substrates deposited using pO2 of (a) 100 mT (LSMO-MgO 100mT), (b) 50 mT (LSMO-MgO 50 mT), a nd (c) 10 mT (LSMO-MgO 10 mT), respectively. The small p eak denoted by around 38 in (c) and (d) is an artifact from the MgO substrate................................................................239 Figure B.2. XRD patterns of (a) LSMO film on STO substrate and (b) STO substrates, respectively. The insets (I) and (II) to (a) shows the details of the STO(200)/LSMO(200) and STO(300)/LSMO(300) peaks. The inset to (b) shows the rocking cu rve performed about the STO (200) plane...............................................................................................................241 Figure B.3. SEM image of th e surface of LSMO film deposited on STO substrate.......241 Figure B.4. SEM images of LSMO top electrodes on PZT film grown using a shadow mask that produced 100 m diameter contacts.................................242

PAGE 24

xx Figure B.5. SEM images of PZT films on STO substrates de posited using dual laser ablation with a KrF fluence 1 J/cm2 and CO2 fluence of 2 J/cm2 with an interpulse delay t = 50 ns.............................................................243 Figure C.1. Time gated ICCD images usi ng 200 ns exposure time of single laser ablated plumes from PZT target using laser fluence of 2 J/cm2 under 500 mT ambient O2 gas..................................................................................247 Figure C.2. Plot of the highest intensities of the visible plumes captured at various time intervals as a function of the di stance from target surface and time. ........................................................................................................................248 Figure C.3. ICCD plume images using single and dual laser ablation under vacuum...........................................................................................................249 Figure C.4. ICCD images of CO2 laser ablated plumes captured at various time intervals showing the ej ection of particulates for the target surface using 500 s exposure times..........................................................................250

PAGE 25

xxi ABSTRACT Multiferroic materials exhibit unique properties such as simultaneous existence of two or more of coupled ferroic order pa rameters (ferromagnetism, ferroelectricity, ferroelasticity or their anti-fe rroic counterparts) in a singl e material. Recent years have seen a huge research interest in multiferroic materials for their potential application as high density non-volatile memory devices. Howeve r, the scarcity of these materials in single phase and the weak coupling of thei r ferroic components have directed the research towards multiferroic heterostruct ures. These systems operate by coupling the magnetic and electric properties of two materials, generally a ferromagnetic material and a ferroelectric material via strain. In this wo rk, horizontal heterostru ctures of composite multiferroic materials were grown and character ized using pulsed laser ablation technique. Alternate magnetic and ferroelectric layers of cobalt ferrite and lead zirconium titanate, respectively, were fabricated and the coupli ng effect was studied by X-ray stress analysis. It was observed that the interf acial stress played an important role in the coupling effect between the phases. Doped zinc oxide (ZnO) heterostructures were also studied where the ferromagnetic phase was a layer of manga nese doped ZnO and the ferroelectric phase was a layer of vanadium doped ZnO. For the fi rst time, a clear eviden ce of possible room temperature magneto-elastic coupling was obser ved in these heterostructures. This work provides new insight into the stress medi ated coupling mechanisms in composite multiferroics.

PAGE 26

1 CHAPTER 1: INTRODUCTION In 1865, James Clerk Maxwell had propos ed four equations governing the dynamics of electric fields, magnetic fields an d electric charges [1]. This was the first time that magnetic interactions and motion of electric charges, which were previously thought to be independent phenomena, were s hown to be intrinsically coupled to each other. This was followed by the some pion eering work, in 1888, by Rontgen [2] and in 1894, by Curie [3] that discovered the magnetoel ectric (ME) effect in solids. The ME effect in its most common definition de scribes the coupling between magnetic and electric fields in matter [4]. However, the se arch for materials that displayed ME effect was difficult and the field remained elusive fo r a long time. This was primarily due to the complex mechanisms involved and the different origins of ferroelectricity and magnetism in solids, which did not guara ntee a strong coupling between the two behaviors [5 8]. In the mean time, research on magnetic and el ectronic materials had found their way into every aspect of modern technology [9]. The huge amount of data generated by consumer electronics everyday is frequently stored in modern memory devices such as Magnetoresistive Random Access Memory (MRAM) and Ferroelectric Random Access Memory (FeRAM) [10, 11]. However, th e ever increasing trend towards device miniaturization led to renewed increased interest in combining the magnetic and electric properties into multifunctional materials, so that a single device element can perform

PAGE 27

2 more than one task [9]. Multiferroics materials might hold the future for a new generation of memory devices that could be electronically written and magnetically read [12]. Major discoveries such as the four state multiferro ic memory devices brought the research one step closer in that direction [13]. Recent years have seen a flurry of re search focused on multiferroic materials where both magnetism and ferroelectricity coex ist [4, 14 16]. The revival of modern multiferroics was possible not only due to recent theoretical breakthroughs in understanding the coexistence of magnetic and electrical orde ring but also advances in thin film growth techniques and experime ntal methods for observing magnetic and electric domains [4]. Probably, the most important factor that rejuvenated th e “quiet” and “abstruse” field of multiferroics was the broader re alization that combining magnetism and ferroelectricity in different ways in a single material could create new technical applications like the electric field cont rol of magnetic memory and vice versa.

PAGE 28

3 1.1. Materials Overview Before we go into the details of multiferroic behavior a review of the basic properties of magnetic and ferroelec tric materials is presented. 1.1.1. Magnetic Materials Magnetic materials exhibit long range al ignment of the atomic moments resulting in a spontaneous net magnetization (M) even in the absence of an external magnetic field (H) [14]. The important magnetic parameters can be obtained from the M versus H hysteresis loop as shown in Figure 1.1.1. The coercivity (Hc) is the reverse field that reduces M to zero from saturation magnetization (Ms). The remnant magnetization (Mr) is the value of M at H = 0. Magnetic memory devices should have high Hc and squareness (Mr/Ms) to ensure that a majority of do mains have similar switching fields. 1.1.1.1. Spinel-type Ferrite: CoFe2O4 The ferrites include the complete family of Fe-containing oxides such as the spinels (AFe2O4), garnets (AFe5O12), hexaferrites (AFe12O19), and orthoferrites (RFeO3) where A or R is a metal or a rare-earth element, respectively [14]. Among them the spinel-type ferrites have attracted consider able research interest over the years for possible applications in diverse fields [17 19 ]. Spinel ferrites have found applications in microwave magnetic devices [20 23], magnetooptic data storage ap pliances [24 26], and for potential biological applications in drug delivery [27, 28]. They have also found applications in magnetic tunne l junctions [29], magneto elas tic devices [30], spintronic devices and composite multiferroic heterostructures [31].

PAGE 29

4 Figure 1.1.1. Typical hysteresis loop of a ferromagnetic materi al showing the saturation magnetization (Ms), remnant magnetization (Mr) and coercive field (Hc). Cobalt ferrite (CFO), having a chemical formula CoFe2O4, is an important member of the spinel type ferrite family. CFO is a hard magnetic material with high degree of magnetic anisotropy and magnetostri ction [17]. The unique magnetic properties of this material can be understood from its electronic and structural configurations. The Co2+ and Fe3+ are located on the octahedral and tetr ahedral sites, respectively, as in an inverse spinel structure as shown schema tically in Figure 1.1.2. In a normal spinel structure (AB2O4), the A2+ cations occupy the tetrah edral sites whereas the B3+ cations occupy the octahedral sites. However, in an inve rse spinel structure, half of the octahedral coordination sites are occupied by A2+ cations and the remaining half as well as all the tetrahedral coordination s ites are occupied by the B3+ cations. A unit cell of CoFe2O4 HcMrMsMagnetization (M)Ma g netic Field ( H )

PAGE 30

5 (CFO) which is a f ace-centered cubic (fcc) struct ure with lattice parameter (ao) of 8.39 , consists of 8 formula units (Figure 1.1.2) [17]. The eight Fe3+ ions in tetrahedral sites are aligned antiferromagnetically with respect to the remaining eight Fe3+ ions via superexchange interactions mediated by oxygen ions. Thus the uncompensated Co2+ ions which have three unpaired electrons in thei r d-orbitals give a theoretical saturated magnetization value of 3 B per formula unit or per Co site. Figure 1.1.2. Schematic diagram of unit cell and subunits in CoFe2O4 inverse spinel structure (Adapted from Ref. [32]). However, the experimental value of the saturated magnetization in CFO is found to be around 4 B. This discrepancy between the theoretical and experimental values could be attributed to two factors [17]. Fi rst, the calculation was done by neglecting the contribution from the orbital motion of electrons. Secondly, the Fe3+ moments were Oxygen atoms B atoms (octahedral sites) A atoms (tetrahedral sites)

PAGE 31

6 assumed to be aligned perfectly anti-parallel, but in reality they may be canted. In addition, the distribution of diff erent ions may not be as perfect as assumed. This creates the intriguing magnetic properties of CFO thin films as compared to bulk powders. In this work, CFO was chosen as a promising magne tic material for growing thin films and heterostructures. 1.1.1.2. Magnetic Anisotropy When the magnetization (M) of a magnetic ma terial changes with the direction of the applied magnetic field (H), it is said to exhibit magnetic anisotropy. This behavior, primarily observed in thin films, makes th eir magnetic properties quite different from those of the bulk. In CFO thin films, due to their unique struct ural properties, the magnetic moments prefer to lie in an energetically favorable direction, called the easy axis of magnetization. Figure 1.1.3 (a) shows an idealized M versus H hysteresis loop for a thin film having the easy magnetization axis parallel to the film plane. When H is applied parallel to the film plane (in-plane) the M-H loop exhibits easy magnetization to saturation. Contrarily, when H is applied perpendicular to film plane (out-of-plane), a much larger field is necessary to fully magnetize the sample. Figure 1.1.3 (b) shows a schematic diagram of the direction of easy axis of magnetization with respect to the magnetic moments in the micr ostructure of the film. For thin films with uniaxial magnetic anisotropy, the overall magnetic anisotropy coefficient (Ku) is given by [33]: Ku = Kint + Ksh + K (1.1) where Kint is the intrinsic magneto-crystalline, Ksh is the shape, and K is the stress anisotropies, respectively.

PAGE 32

7 Figure 1.1.3. (a) Ideal M H loop and (b) sche matic diagram of magnetic moments for a thin film with the easy axis of magnetiz ation along the film plane, respectively. The microscopic origin of the intrinsic magne tocrystalline anisotropy is related to the spin state of the magnetic moments and by the symmetry of their arrangement in the crystal lattice that involves spin-lattice coupling. The Kint for a simple cubic symmetry crystal such as CFO is given by [34]: Kint = K1 (1 2 2 2 + 2 2 3 2 + 3 2 1 2) + K2 ( 1 2 2 2 3 2) + … (1.2) where Ki (i = 1, 2, …) is the anisotropy constant and 1, 2, and 3 are the direction cosines of M and the a, b, a nd c axes of crystal system. The source of the shape anisotropy is th e demagnetization energy associated with the shape of the sample. This favors the alignment of the moments along the largest extent of sample (parallel to the film). Fo r a film whose thickness is much less than its other dimensions, shape anisotropy is given by [34]: Ksh = 2 Ms 2 (1.3) Thin films invariably grow with in built stresses during the deposition process. The stress anisotropy is given by [34]: Easy axis of magnetization M H in-plane out-of-plane (a) (b)

PAGE 33

8 K = 3 s/ 2 (1.4) where is the stress and s is the magnetostriction coefficient. Magnetostriction is another property of magnetic materials associated with the change in dimensions upon magnetization [ 35, 36]. As a result, elastic strains are developed in the crystal latt ice that change the alignment of the magnetic moments and create the magnetoelastic eff ect in the material [37, 38]. 1.1.2. Ferroelectric Materials Ferroelectric materials exhibit long range a lignment of electric dipoles resulting in a net polarization under an applied electric fi eld. The unit cell of a ferroelectric crystal has a polar axis that causes s pontaneous electric dipole moment even in the absence of an electric field [34]. The existence of a spontaneous pol arization implies that there is a preferred special orientation in the crystal. Just like in magnetic materials the important parameters for a ferroelectric material can be obtained from the polarization (P) versus electric field (E) hysteresis loop as shown in Figure 1.1.4. Ferroelectric thin films have found app lications in micro-electro-mechanical (MEMs) devices, non-volatile memories, and high frequency electrical components [39]. An exhaustive review of applications can be found in Ref. [40] and the articles therein [40]. The principle of nonvolatile ferroelectr ic random access memories (FeRAMs) is based on the polarization reversal in ferroelectric thin film capacitors under the influence of an external applied electric field. The computational “0” and “1” are represented by the nonvolatile storage of the negative or positiv e remnant polarization state, respectively [40].

PAGE 34

9 EcPrPsPolarization (M)Electric Field (H) Figure 1.1.4. Typical hysteresis loop of a ferro electric material showing the saturation polarization (Ps), remnant polarization (Pr) and coercive field (Hc). 1.1.2.1. Ferroelectric Perovskite: PZT Ferroelectric perovskites, such as Pb(Zrx,Ti1-x)O3 (PZT), BaTiO3 (BTO), and BiFeO3 (BFO) have always attracted significant interest due to their simple crystal structure, better ferroelectric and mechan ical properties [14], and their potential application in ultrahigh density memory de vices [41]. Among them, PZT has remained the material of choice for its outstan ding ferroelectric properties [42]. Lead zirconium titanate (PZT) is a solid solution of PbTiO3 and PbZrO3 compounds [43]. The Curie temperature (Tc) of PZT can vary from 230 C to 490 C depending on the composition (i.e. 1 > x > 0). Above Tc, PZT is cubic (Figure 1.1.5) and has no electric dipole moment. However, below Tc, structural changes make the crystal non-centro-symmetric. The Ti/Zr ion shifts from its central location along one of several

PAGE 35

10 allowed directions (Figure 1.1.5). This slightly di storts the crystal lattice into a perovskite structure (tetragonal/rhombohedral shape), and produces an elec tric dipole. If it is then subjected to an electric field, the domain dipoles ali gn in the direction of the field causing polarization. The dipoles maintain this orient ation even after the field is removed which gives the remnant polarization. In this work, PZT has been selected as a ferroelectric material for growing thin films and heterostructures. Solid solutions of PZT with different compositions exhibit an unusual phase boundary which divides ferroelectric regions with differe nt structures. PZT shows anomalously high ferro electric response near such morphotropic phase boundary (MPB). For this reason, in this work, the PZT composition was kept constant at PbZr0.52Ti0.48O3 it is in the vicinity of the MPB [44]. Figure 1.1.5. Schematic diagram of PZT unit cell showing off-centering of Zr/Ti ion under an external applied electric fi eld (Adapted from C. Kittel [45]). Also, the direct epitaxial gr owth of PZT on Si substrates is inhibited due to interdiffusion and larger lattice mismatch [44]. Hence the PZT films have been grown on single crystal MgO and SrTiO3 substrates in this work. E Pb O Zr/Ti

PAGE 36

11 1.1.3. Multiferroic Materials By definition, multiferroic materials are those that simultaneously display more than one ferroic or corresponding antiferro ic properties such as ferroelectricity, ferromagnetism, ferroelasticity, or antife rroelectricity, and an tiferromagnetism [46]. Multiferroic materials can be divided into two categories. If the co-existence of the ferroic properties occur in the same materi al, it is called a single phase multiferroic. On the other hand, if the ferroic behaviors exist individually w ithin separate phases, it is called a composite multiferroic. 1.1.3.1. Scarcity of Single Phase Materials There is a scarcity of si ngle phase multiferroic materi als. Only a small group of materials that show magnetic and electric polarization are either ferromagnetic or ferroelectric. Within these select materials, a very small subgroup exhibits multiferroic behavior. The overlap required for multiferroic materials is shown schematically in Figure 1.1.6. In order to understand why these materials are so rare, one has to look into the microscopic mechanisms of magnetism and ferroe lectricity in these materials. Although the ferromagnetic and ferroelectric materials bo th exhibit similar behaviors in term of hysteresis, their origins are completely different. While on one hand ferroelectric materials are insulators with unfilled d orbitals, ferromagnets need conduction electrons and have incompletely filled d shells [14]. Thus there exists an apparent inconsistency between the usual mechanism of off-centering in a ferroelectric materials and the formation of magnetic order in a magnetic mate rial. This also explains the scarcity of ferroelectric-ferromagnetic multiferroics [48].

PAGE 37

12 Figure 1.1.6. Schematic diagram showing the overlap required to achieve multiferroic properties in a single material (Adapted from Ref. [47]). 1.1.3.2. Magnetoelectric Effect Magnetoelectric (ME) effect is the phe nomenon of the inducti on of magnetization by an electric field or polarization by a magne tic field [4, 14]. In general, the ME effect depends on temperature. From applications viewpoint, the real interest in multiferroic materials is in observation of the strong ME coupling at room temp erature. In general terms ME coupling typically refers to the lin ear ME effect. However, the effect could also be non-linear. This can be understood from the expansion of free energy (F) of a material given by: ... 2 1 2 1 ) (0 0 0 j i ij j i ij j i ij i S i i S iH E H H E E H M E P F H E F ... 2 1 2 1 k j i ijk k j i ijkE E H H H E (1.5) with E and H as the electric and magnetic fields, respectively. Magnetically Polarizable Ferroelectric Electrically Polarizable Ferromagnetic Multiferroic Magnetoelectric

PAGE 38

13 Differentiation leads to the polarization i iE F H E P ) ( ... 2 10 j i ijk k j ijk j ij j ij S iE H H H H E P (1.6) and to magnetization i iH F H E M ) ( ... 2 10 k i ijk j i ijk j ij j ij S iE E H E E H M (1.7) where S iPand S iMare the spontaneous polariza tion and magnetization, and and are the electric and magnetic sus ceptibilities. The coefficient in Equations (1.6, and 1.7) is designated as the linear ME coefficient. Experimentally, is generally expressed as E = E/ H, called the ME voltage coefficient. The SI unit of is (s/m), although often it is expressed in amore practi cal unit of (mV/cmOe).

PAGE 39

141.2. Thin Film Multiferroics One aspect of fundamental interest in mu ltiferroic research is the production of high quality thin films of multiferroic materials. Epitaxial thin films provide an additional degree of freedom compared to bulk sample s through the ‘strain engineering’ between the substrate and the film. Multiferroic thin f ilms and nanostructures ha ve been fabricated using a wide variety of growth techniques in cluding sputtering, spin coating, pulsed laser deposition, sol–gel processes, metal-organi c, chemical vapor deposition, and molecular beam epitaxy [14]. 1.2.1. Single Phase Multiferroic Thin Films Thin films of perovskite BiFeO3 (BFO) have received the most attention for their multiferroic behavior among all the single phase multiferroics [49]. BF O thin films have shown simultaneous ferroelectricity, ferroelas ticity and weak ferromagnetism [50, 51]. However, direct evidence of strong ME c oupling at room temperature has not been observed. The next important group includes hexagonal multiferroics formed by the ferroelectric antiferromagnetic manganites RMnO3 with R = Sc, Y, In, Ho, Er, Tm, Yb, Lu [4]. However, they show multiferroic behavior with high ferroelectric ordering temperature (> 590 K) and low magnetic or dering temperature ( 70 K – 120 K) making them still unsuitable for device applications. Thin films of BiMnO3 were also investigated for potential multiferroic applica tions. However, they showed ferroelectricity below 450K and ferromagnetism at 105 K [52], both temperatures being impractical. A number of other single phase multiferroic materials have been studied over the years. Some of the notable ones are BiCrO3 [53] displaying w eak antiferromagnetic-

PAGE 40

15 ferromagnetic coupling around 120 K – 140 K, or BiCoO3 or PbVO3 or double perovskites like Bi2NiMnO6 [4, 14]. However the ME coefficients ( E = E/ H) in most single phase materials are still small in magnitude ( E ~ 20 mV/cm Oe) for any potential applications [4]. Despite a concerted effort by a wide number of researchers, the search for a single phase multiferroic material showing strong coupling at room temperatures has proved to be a difficult one. 1.2.2. Composite Multiferroic Thin Films The scarcity of single phase multiferroic materials and the weak coupling of order parameters in the existing materials have directed research towards composite multiferroic thin films. These systems ope rate by coupling the magnetic and electric properties of two materials, generally a ferromagnetic material and a ferroelectric material via strain [14, 15]. The ME effect in composite materials is shown schematically in Figure 1.2.1. An in-plane magnetic field (H ) induces strain in the magnetic component due to the magnetostrictive effect, which is mechanically transferre d to the ferroelectric component inducing a dielectr ic polarization through the pi ezoelectric effect [15]. Conversely, an out-of-plane electric field (E) induces strain in the ferroelectric component due to the inverse piezoelectric e ffect, which is mechanic ally transferred to the magnetic component, inducing a change in magnetization [15]. Thus the ME effect is the cross interaction in the two phases. It is product of the magnetostrictive effect in the magnetic phase and the electrostrictive effect in the electric phase. The composite ME effect can therefore be described as follows [4, 15]: Direct ME effect = Electric Mechnical Mechanical Magentic (1.8)

PAGE 41

16Converse ME effect = Magnetic Mechanical Mechanical Electric (1.9) Figure 1.2.1. Schematic diagram of strain-m ediated ME effect in a composite system consisting of ferromagnetic laye r (pink) and ferroelectric layer (blue) (Adapted from Ref. [15]). There are three main kinds of ME com posite nanostructures that have been studied so far. As shown schematically in Figur es 1.2.2 (a, b, and c) they are denoted as 2-2 horizontal heterostructur es, 1-3 vertical nanostructu res, and 0-3 nano-particle embedded films, respectively [15]. The pair of numbers preceding the names refer to the dimensionalities of the two phases. The first kind is the 2-2 horizontal heterostructure with alternating ferroelectric (2 dimension) la yers and magnetic (2 dimension) layers, or simply a ferroelectric (or magnetic) thin film grown on a magnetic (or ferroelectric) + + + + ++++ ---Magnetostrictive Piezoelectric Ma g netic field Strain Strain Polarization Strain/ Magnetic field Polarization/ Strain H H E E Inverse piezoelectric Strain Electric field Strain Magnetization Electric field/ Strain Strain/ Magnetization

PAGE 42

17 substrate. The second kind is the 1-3 vertical heterostructure with one-phase nanopillars (1 dimension) embedded in a matrix of anot her phase (3 dimensions). Finally, the third kind is the 0-3 particulate na nocomposite films with magnetic particles (0 dimensions) embedded in a ferroelectric film matrix (3 dimensions). Figure 1.2.2. Schematic diagram of the three main kinds of ME composite nanostructures (a) horizontal heterostructures, (b) vertic al nanostructures, a nd (c) nano-particle embedded films, respectively. (Adapted from Ref [15]) The ferroelectric component materials often used include BaTiO3 (BTO), PbTiO3 (PTO), Pb(Zr,Ti)O3 (PZT) and BiFeO3 (BFO), and the magnetic component materials include CoFe2O4 (CFO), NiFe2O4 (NFO), Fe3O4, La1–xSrxMnO3 (LSMO) and metals. 1.2.2.1. Horizontal Heterostructures Horizontal multilayered films (Figure 1.2.2 a) of alternate ferroelectric perovskite and ferromagnetic spinel phases are the most widely investigated composites due to their ease of fabrication. In recent years a large number of composite horizontal multilayered structures have been investigated. These e xperiments have shown that these composite thin films have great potential for device applications. The observed ME effects were (a) (b) (c)

PAGE 43

18 comparable to those of their bulk counter parts. High quality ferroelectric-magnetic PZT/Ni0.8Zn0.2Fe2O4 multilayer films [54], BaTiO3/Fe3O4 and PZT/La0.7Sr0.3MnO3 bilayer films [55, 56], respectiv ely, have been reported to have coexisting ferroelectric and ferromagnetic behaviors. The ME coefficient in these composite films ranges from 3 mV/cm Oe to 30 mV/ cm Oe. Heterostructures of CFO and BaTiO3 have exhibited ME voltage coefficient E ( E/ H) of 66 mV/cmOe [57]. Recently, simple heterostructures of epitaxial Pb(Zr0.3Ti0.7)O3 films on La1.2Sr1.8Mn2O7 single crystal substrates have exhibited ME coefficient as high as 600 mV /cmOe depending on growth conditions [58]. One of the drawbacks of horizontal heterostruct ures is that the ME effects are weakened due to the large in-plane constraint (or clamping) from the substrate [59]. 1.2.2.2. Vertical Nanostructures One way to avoid the substrate clamping is to grow vertical nanostructures (Figure 1.2.2 b). There have been some reports of vertical nanostruc tured film in which ferrimagnetic CFO nanopillars are embedded in a BFO film [60, 61]. It was shown that an electric field could change the magnetic c onfiguration of the CFO pillars. The strength of the ME coupling between the ferroelectric matrix and the ferrimagnetic nanopillars, was estimated from the electric fiel d-induced change of magnetization ( M) which gave = M/ E = 0.126 Oe cm V–1. The mechanism responsible fo r the electric field-induced changes in magnetization was due to the inti mate lattice coupling between CFO and BFO. This was possible due to the th ree-dimensional heteroepitaxial growth in such vertical structures. In vertical nanostr uctures as described above, the shape of the piezoelectric matrix changes with the application of an el ectric field through the inverse piezoelectric

PAGE 44

19 effect (Figure 1.2.1). This al ters the magnetic anisotropy of the ferromagnetic pillars via magnetostriction giving rise to the observed ME effect. 1.2.2.3. Nano-particle Embedded Films Wan et al., [62] used a sol-gel techni que to grow composite films with CFO nanoparticles dispersed in PZT matrix (Figur e 1.2.2. c). Zhong et al. [63] employed an analogous method to grow compos ite films of CFO nanoparticle s in a ferroelectric phase of Bi3.15Nd0.85Ti3O12 (BNTO). Both films exhibited ferroelectric and ferromagnetic behaviors, and ME coefficients were measur ed in both studies. Murugavel et al. [64] prepared (100)-oriented PZT/NiFe2O4 composites on (001) SrTiO3 substrates by PLD where the NFO nanoparticles were randomly di spersed in the PZT matrix. However, the maximum values of E for these composites were of the order of 10 mV/cmOe which is lower than that reported for bulk PZT/NFO pa rticulate composites [1 8]. This might be due to lattice clamping effects by the substrate. In all these structures it interfacial strain was always found to play a crucial role that determined the effectiveness of th e coupling mechanism. The ultimate goal for multiferroic device fabrication would be a ne w single-phase or composite material with strong coupling between ferroel ectric and ferromagnetic order parameters making for simple control over the magnetic nature of the material with an applied electric field at room temperature. Multiferroic memories when fully functional would offer the possibility of combining the best qualitie s of FeRAMs and MRAM s: fast low-power electrical write operation, and non-de structive magnetic read operation.

PAGE 45

201.3. Chapter Summary In conclusion it can be said that high-qua lity thin films of composite multiferroic materials hold huge prospect for the rationa l design of new device functionalities. Preliminary experiments have shown that stro ng ME coupling at room temperatures can be achieved by the proper interfacial st rain engineering in these composite heterostructures. With this in mind, the work on the growth and characterization of 2-2 horizontal heterostructures of CFO and PZT was undert aken. CFO and PZT were chosen as the respective ferromagnetic and ferroelectric pha ses for their excellent room temperature magnetic and ferroelectric proper ties. Along the way, a single phase heterostructure using doped ZnO thin film was also fabricated and characterized for the first time. The following section gives a brief outline of the dissertation.

PAGE 46

211.4. Dissertation Outline Chapter 2 describes the experimental a nd measurement techniques used in this work. It includes a description of the process of pulsed laser deposition of the thin films and heterostructures. It also discusses the m easurement tools and how they were used to characterize the material prope rties of the deposited films. Chapter 3 and 4 describes the details of the results obtained from the e xperiments. Chapter 3 di scusses the role of epitaxial thin films in controlling the ferro magnetic/ferroelectric properties of CFO/PZT bilayer films. Chapter 4 desc ribes the interesting propertie s of doped ZnO thin films and how they can be manipulated to observe ma gneto-elastic coupling mechanisms. Chapter 5 will give a brief account of future directions and conclusions.

PAGE 47

22 CHAPTER 2: METHODOLOGY This chapter describes the experimental setups and measuremen t techniques used in this work. The entire work was done at the Department of Physics, University of South Florida (USF). Thin film fabrication using pulsed laser de position (PLD) and characterization was performed at Laborat ory for Advanced Material Science and Technology (LAMSAT) at USF. The PLD set up used in this work was assembled under the guidance of R. Hyde. The structural ch aracterization was done using X-ray diffraction (XRD), scanning electron microscope (SEM) an d atomic force microscope (AFM). Some of the XRD scans and the AFM scans were performed at Nanotechnology Research and Education Center (NREC) at USF. The ma gnetic measurements were performed at Functional Materials Laboratory (FML) at U SF. For the ferroelectric measurements a home-made micro-pole-station and ferro -tester set up was built at LAMSAT.

PAGE 48

23 2.1. Pulsed Laser Deposition of Thin Films. Pulsed laser deposition (PLD) is a versat ile technique for thin-film and multilayer research [65]. During the last decade PLD has be en extensively used to grow thin films of oxides and other complex materials [66]. Although the underlying ablation process is complex, PLD is conceptually and experimentally simple [65]. A ultra-violet KrF (wavelength of 248 nm, pulsed width of 20-25 ns) excimer laser beam is focused with an energy density (fluence) in the range 1 J/cm2 to 6 J/cm2 onto a rotating target inside a va cuum chamber as shown in the schematic diagram in Figure 2.1.1. The depos ition chamber is equipped with a multitarget “carousel” (changer) and a substrate heater. For multilayer deposition, multiple targets can be sequentially exposed to the in cident laser beam, thereby enabling the insitu growth of hetero-structures and super lattices with relativ ely clean interfaces. Figure 2.1.1. Schematic diagram of the single laser deposition set up. PLD is often described as a three-step process consisting of (i) laser-target interaction and vaporization of a target material, (ii) plasma plume formation and its Excimer Lase r Rotating Target Vacuum Chamber Substrate Heater M L Multi-target changer Outlet for O2 gas View port

PAGE 49

24 transport towards the substrate, and (iii) nucle ation of the ablated sp ecies and growth of thin film on the substrate surface [67]. When the laser beam strikes the rotating ta rget, the UV laser radiation is absorbed by the outermost layers of the target surface. This causes rapid h eating and vaporization of the target material to form dense plasma The highly energetic ab lated species travel towards the substrate, crystallizing into a f ilm with a composition typically identical to that of the target material [68]. PLD has several characteristics that di stinguish it from other film-growth processes and provide special advantages fo r the growth of oxides, doped semiconductors and multicomponent materials. Congruent (stoichiometric) transfer of materials. Under the proper choice of laser fluence films have the same composition as the target. This sets PLD apart from incongruent-transfer methods such as thermal evaporation and sputtering. Capability for reactive deposition in ambient gases. Since the deposition chamber does not require any electron beam or hot filament, ambient gases can be used. Energetic species of the ablated plasma react with ambient gas molecules facilitating the deposition of mu lticomponent ceramic materials. Growth of multilayered epitaxial heterostructures. These thin films are composed of layers of materials of different co mpositions, but all layers share a common continuous crystal structure. With the revival of multiferroics such layered structures have been researched a lot to investigate magneto-electric coupling between the layers [9].

PAGE 50

25 The ablation process has some characte ristic limitations depending on the materials being ablated. Particulates. PLD films invariably have some particulates with diameters ranging from 0.1 to 10 m (with most < 1 m). These particulates are undesirable for multilayer structures where they lead to unacceptable scattering mechanism and surface roughness of films [69, 70]. Uniform film thickness. Due to the highly forward directed nature of the laser ablated plume, PLD films are uniformly thick only in a narrow region. To overcome this problem, all the films in this work were grown on small area (1 cm x 1 cm) substrates where uniform thicknesses were achieved. Preferential ablation of elem ents for complex oxides. For multi-component oxides such as PZT laser ablation is complicated by the high volatili ty of one of the elements i.e. Pb. Due to the high vapor pressure of Pb in PZT, at high growth temperatures the films are often obtained as Pb depleted with poor ferroelectric properties. [42]. These shortcomings of PLD can be overcome using a Dual Laser Deposition (PLDDL) technique [71]. The dual laser ablation technique used in this work has been described in more details el sewhere [72 75]. Figure 2.1.2 s hows a schematic diagram of the PLDDL system. Briefly, the KrF excimer la ser output is combined with a CO2 laser (wavelength 10.6 m, pulse duration 500 ns) output. The laser beams are spatially and temporally overlapped on to a rectangular s pot size of 6 mm2 on the target. An intensified charge-coupled detector (ICCD ) system is aligned normal to the plume propagation direction to image the visible wavelength em ission from the laser induced plume. The

PAGE 51

26 excimer laser and ICCD camera are triggered through the digital dela y generator which is used to control the peak-to-peak (p-p) inter-pulse delay ( t) between the two lasers (Figure 2.1.2) [72 75]. Figure 2.1.2. Schematic diagram of the dual laser deposition set up. The p-p delay is adjustable in the range -70 ns to 250 ns, relative to the peak of the excimer laser pulse. The temporal prof iles of the two laser pulses are directly observed on a 400 MHz, 2 G-sample/sec os cilloscope through the use of a UV photodiode and an IR pyroelectric detector, thereby allowing a precise adjustment of the required inter-pulse delay [76]. The ICCD imaging and emission spectroscopy of the UV Detecto r Excimer Lase r Trig In Vacuum Chamber Substrate Heate r Rotating Target IR Detecto r CO2 Laser Sync Out 10.00nsDigital Delay Generator Oscillosco p e Tri g In ICCD Sync Control PC Lens ICCD 02004006008001000 t = Interpulse delay t Excimer pulse CO2 pulseVoltage (a.u)Time ( ns )

PAGE 52

27 ablated plumes in PLDDL are used in determining the optimum coupling of the laser outputs and optimizing the growth parameters for film deposition. Figure 2.1.3. Photograph of the custom built PLD system at LAMSAT. Excimer laser Load -lock Deposition Chamber TC pressure gauge Roughing pump Turbo-molecular pump Multi-target Changer Control PC Optics

PAGE 53

28 At the optimum coupling condition of laser energies enhanced plume excitation and expansion is observed that facilitates the growth of very smooth, particulate-free, and uniform films over large area. Figure 2.1.3 sh ows a photograph of the custom built PLD system at LAMSAT used in this work. The individual parts have been labeled. A general scheme followed for the deposition of all the thin film in this work. The target to substrate distance was always kept constant at 6 cm. Thin films were deposited on various single crystal s ubstrates such as Si (100), c-cut sapphire (Al2O3) (0001), MgO (100) and SrTiO3 (STO) (100). The substrates were firstly cleaned using ultrasonic cleaning in acetone, methanol, rinsed with de -ionized water, and then dried with high purity N2 gas before loading into the depositi on chamber. The base pressure in the deposition chamber was kept constant at 10-6 T. High purity O2 gas was used as the background ambient gas during deposition. Af ter the deposition, the films were cooled down to room temperature gradually in the same partial O2 pressure (pO2). In the following chapters the deposition conditions for the films will be summarized in a tabular form. As an exampl e Table 2.1 lists the deposition condition for a film named ‘Sample A’. Table 2.1. Summary of deposition conditions of a thin film named Sample A. SampleSubstrateLaser FluenceGrowth Temperatures O2 Pressure Film Thickness (J/cm2) Ts ( C)pO2 (mT) (nm) Sample ASi260010200

PAGE 54

29 2.2. Structural Characterization This section is devoted to explain the ba sic principles and the techniques that were used for the structural characterization of the thin films. Stru ctural characterization helped in optimizing the grow th conditions required for produ cing high quality thin films. It also helped in investigating the micro st ructural properties of the films such as crystallinity and surface morphology and how they were related to their material properties such as magnetizati on, polarization, conductivity, etc. Crystal structure and phase of the thin films were determined from the X-ray Diffraction (XRD) analysis. Chemical purity and stoichiometry were c onfirmed by the Energy Dispersive X-ray Spectroscopy (EDS). Surface morphology of the thin films was studied using the Scanning Electron Microscope (SEM) and the Atomic Force Microscope (AFM). 2.2.1. X-ray Diffraction X-ray diffraction (XRD) is a widely used, simple yet versatile tool that can all issues related to the crystal structure of thin films, including lattice parameters, identification of unknown material s, orientations of single cr ystalline films, preferred orientations of polycrystalline films, defects, stress, etc. Since the wavelength of X-rays (approx. 0.1 to 1 ) is comparable to the size of atoms, they are ideally suited for probing the structural arrangement of atoms in a wide range of materials. XRD is non destructive and does not require elaborat e sample preparation [77]. The basic principle is based on Bragg’s law given by: 2dsin = n (2.1) where is the angle of incidence of the X-rays, is the wavelength of the X-rays, and n is a positive integer representing the order of the diffraction peak. Figure 2.2.1 shows an

PAGE 55

30 incident beam of parallel X-rays impi nging the crystal surface at an angle and getting reflected from the parallel planes of atoms fo rmed by the crystal lattice of the material. The atoms, represented as spheres in all the figures, can be viewed as forming different sets of planes in the crysta l. Two consecutive reflected beams have a phase difference because they travel a different path. Constructive interference of the reflected rays occurs only when their path difference is an integer multiple of the wavelength of X-ray as given in Equation 2.1. A characteristic diffraction pattern is produced and plotted as intensity (I) versus 2 graph. The measured pattern can be then compared with a known database of reference patterns to determine the crystal structure of the film. Figure 2.2.1. X-ray reflection from latti ce planes illustrating Bragg’s law. Polycrystalline films generate XRD patterns with all the possible crystal orientations of the material similar to the bu lk powders. However, preferred oriented or textured films demonstrate XRD patterns with certain Bragg reflections more pronounced than the others. Figures 2.2.2 (a, b, and c) show schematic diagrams of randomly orientated crystallites in a polycrystalline film, textured film with preferred weak 2 d Incoming X-ray beam Reflected X-ray beam To the detector d sin

PAGE 56

31 orientation, and epitaxial film with str ong orientation, resp ectively. The preferred crystallographic directions lead to anisotropic properties in thin films [78]. Figure 2.2.2. Schematic diagram of crystallites in a (a) polycrystalline film with random orientation, (b) textured film with preferred weak orientat ion, (c) epitaxial film with strong orientation, respectively. (Adapted from Ref. [78]) The XRD scans were carried out with Cu K radiation using Bruker D8 Focus Xray Diffractometer equipped with a position sensitive Lynx Eye detector (PSD). Figure 2.2.3 shows a photograph of the equipment and its major components. The four major components in the system are: X-ray sour ce consisting of a Cu anode, LynxEye X-ray detector (five times more reso lution power than scintillati on detectors), goniometer with sample holder, and control computer. The sy stem can perform all linear scans such as 2 scans, 2 asymmetric or detector scans, and rocking curves. The following sections briefly describe the various types of scans that were performed during this work. During -2 scan, the angle of incidence ( ) and the angle of detection (2 ) are continuously varied but kept and 2 respectively, in a locked-coupled mode. Figure 2.2.4 shows an example of a -2 scan performed on an epit axial (100) f ilm grown on a (100) single crystal substrate. ( a ) ( c ) ( b )

PAGE 57

32 Figure 2.2.3. Photograph of Br uker D8 X-ray Diffractometer at LAMSAT showing the individual parts. X-ray source LynxEye Detector Goniometer Sample holder stage Monochromator Aperture slit system Detector slit system X-ray tube (Cu K ) K Filter 2

PAGE 58

33 The Bragg reflections occur only from th e (100) planes that are parallel to the substrate planes. The out-of-plane lattice parameter (a) can be calculated for the cubic system of crystals using the relation: 2 2 2 (hkl)l k h d a (2.2) which gives a= d(100) (i.e. d spacing for (100) plane). Figure 2.2.4. Schematic diagram of X-ray reflecti ons from the parallel crystal planes of a (100) oriented film grown on a ( 100) single crystal substrate. Asymmetric scans are used in order to cal culate the in-plane lattice parameter (a) in epitaxial films. In asymmetric scans the incident x-ray is set at a characteristic diffraction angle corresponding to a plane other than the epitaxial plane. The reflected beam is then traced by the detector which scans from 0 to 2 The XRD peak occurs when the Bragg condition is satisfied for the pl ane for which the incident angle is initially set. This is illustrated in Figure 2.2.5 which shows the asym metric scan being performed 2 Film (100) Substrate (100) d 100Incoming X-ray beam Reflected X-ray beam To the detector

PAGE 59

34 about the (111) plane for an epitaxial (100) fi lm grown on a (100) sing le crystal substrate. When the detector reaches the correct 2 values for the (111) plane, the Bragg reflection occurs. This allows for the calculation of d spacing (d(111)). The in-plane lattice parameter (a) is calculated using e quation (2.2) which gives a = 3d(111). Figure 2.2.5. Schematic diagram of asymmetric scan or detector scan performed about the (111) plane of an epitaxial (100) oriented film grown on single cr ystal (100) substrate. Thin films inevitably have in built stresses due to the deposition process. For an unstressed cubic cr ystal the bulk latt ice parameter (ao) is same in all directions. However, in a thin film the out-of-plane (a) and in-plane (a) lattice parameters could be different depending on whether the film grows with comp ressive or tensile (expansive) strains. As shown schematically in Figure 2.2.6, if the bulk lattice parameter (ao) of the material is greater than that of the substrate (asubstrate) or ao > asubstrate, then in order to match the (100) (111) 2 d(111) Detector Scan Incoming X-ray beam

PAGE 60

35 smaller lattice parameter of the substrate the film gets compressed in-plane and extended out-of-plane making a > a. In this case the film is said to be under in-plane compressive strain. The s ituation is opposite when ao < asubstrate, then the film gets extended in-plane and compre ssed out-of-plane making a < a. In this case the film is under in-plane tensile strain. Figure 2.2.6. Schematic diagrams of films gr own under compressive and tensile strains. The out-of-plane and in-plane strains ( ) at room temperature can be calculated by using the formula: o oa a a (2.3) a a a a a a a a ao asubstrate ao asubstrate aaaa (Compressive strain) (Tensile strain)

PAGE 61

36 where a is the out-of-plane (a) or in-plane (a) lattice parameters and ao is the bulk unstressed lattice parameter, respectively. Th e out-of-plane and in -plane and stresses ( ) in the film were calcula ted using Hook’s law, Y = / relating the in-plane strain values ( ) and the Young’s modulus (Y ) value for the material. The residual stress in a thin film is generally caused by the lattice mismatch between film and substrate or post-deposition processing or external influences like differences in thermal expansion coefficients between the film and the substrate, or the microstructure of the film [79]. The stress anal ysis is especially critical in multiferroic heterostructures since the magne to-electric effect is mediated through interfacial stress [9]. To quantify the residual stress in a thin film one can use the sin2 technique of stress measurement. The sin2 technique is a standard non-destructive technique to measure residual stress in a material. In this work the sin2 technique was performed using a Philips X’Pert X-ray Diffractometer equipped with a 3 dime nsional (pole-figure) goniometer as shown in Figure 2.2.7 (a). This allows the rotation of the sample in three degrees of freedom about the angles and as shown in Figure 2.2.7 (b). In sin2 technique, a highly textured film is scanned in the vicinity of a Bragg reflection plane. A shift in the Bragg reflection (2 ) is observed if a strained film is tilted by an angle [80]. The scan is performed in the continuous scan mode with the incident angle and the detector fixed at 2 for the plane under invest igation, while changing the angle at the same time but keeping the angle at 0 Due to the residual stress the lattice para meter and consequently the d-spacing of the parallel crystal planes ch ange causing a shift in the 2 peaks. When the film is under

PAGE 62

37 compression the 2 peaks show a shift to higher angles due to th e decrease in d-spacing following Bragg’s law. The 2 value for each angle was used to calculate the d values using Equation 2.1. The shift in d values ( d) are calculated using d = d –do where do is the unstrained lattice spacing of the film. Figure 2.2.7. (a) Photograph of Philips X’Pe rt X-ray Diffractometer. (b) Schematic diagram showing the angles and with respect to the th in film sample and the incident X-ray beam. X-ra y source Scintillation Detector Sam p le x y z Pole-figure Goniometer (a) (b) X-rays Detector Sample

PAGE 63

38 For biaxial residual stre ss the d-spacing versus sin2 graph follows a linear relationship. Hence the residual stress ( R) is calculated from the slope of the linear fit of d/do vs sin2 graph following the equation [80]: o Rd d Y 1 sin 1 2 12 (2.4) where do is the value at = 0 Y is the Young’s modulus and is the Poison’s ratio for the film. If the residual stress is not biaxial the sample has to be rotated to different angles and the sin2 technique has to be repeated at each angle. The equations used are as follows [80]: 22 11 2 0 0sin 1 E E d d d (2.5) 2 22 2 11sin cos (2.6) The sin2 technique is unique compared to othe r XRD scans as it gives a quantitative value for the residual stress. Rocking curves ( -scans) are used to estimate the degree of in plane orientation in textured films can be investigated usi ng rocking curves. The rocking curve scans are always performed about the plan e of preferred orientation. The detector is fixed on the 2 o position of the Bragg plane under investiga tion, and the sample is tilted (rocked) about the vicinity of the angle o as shown in Figure 2.2.8. Th e FWHM obtained from the rocking curve reveals the degree of film orie ntation. A narrow peak indicates of a high degree of in-plane orientation of the crystallites. For a perfect single crystal, the ideal

PAGE 64

39 rocking curve peak is a function [78]. The rocking curves of a highly orient ed thin films generally show a finite peak width (< 1 ). Figure 2.2.8. Schematic diagram of the X-ra y rocking curve measurement where the sample is tilted (rocked) about the plane of orientation. Azimuthal scans ( scans) are used to determine the crystal symmetry and epitaxial growth in thin films. The scans in this work were performed using the Philips X’Pert X-ray Diffractometer. Figure 2.2.9 (a) shows the ini tial set up for a typical scan to be performed to confirm the cubic symmetr y of an epitaxial (100) thin film. Although the epitaxial (100) planes give Br agg reflection at incident angle there are other sets of parallel planes for example (111) that could also satisfy Bragg condition provided the sample is mounted at some other incident angle o. In order that Bragg’s condition is satisfied for the plane (111) the sample is rotated by = 45 which is the angle between the (111) and (100) planes. o 2 o

PAGE 65

40 Figure 2.2.9. Schematic diagram of scan performed to determine the cubic symmetry of a (100) oriented thin film. 2 o (111) o (100) (111) 2 (a) (b)

PAGE 66

41 The angle is calculated using the formula: 2 2 2 2 2 2 1cose e e e e el k h l k h ll kk hh (2.7) where (hkl) and (hekele) are the Miller indices for the tw o planes under consideration for a cubic crystal system. Now with the incident beam fixed at o (Bragg angle for (111) plane) and the detector fixed at 2 o and = 45 the sample is rotated about the angle from 0 to 360 If there are only f our peaks in the spectra that are separa ted by equal intervals of 90 it confirms the four fold symmetry of th e cubic system and the in plane epitaxy. As shown later, both for cobalt ferrite and PZT the spectra peaks occur at 90 intervals, since both crystals exhi bit four fold symmetry. 2.2.2. Scanning Electron Microscope Scanning Electron Microscope (SEM) is one of the most widely used techniques to characterize the surface morphology and cross-se ction of the thin films. In this work all the imaging was performed using a JOEL JSM-6390LV SEM. Figures 2.2.10 (b, c, and d) show the major components of the SEM: the electron gun column, specimen chamber and control console. Figure 2.2.10 (a) s hows a schematic diagram of path of the electron beam inside the elec tron gun column. The electron gun column consists of an electron gun and magnetic lenses (Figure 2.2.1 0 a). The control cons ole consists of a LCD viewing screen and a control computer. At the base of the column is the specimen chamber (Figure 2.2.10 c) which is evacuated to about 10-6 Torr during operation. The electron gun consists of three components: tung sten wire filament serving as cathode, grid cap and anode (Figure 2.2.10 a).

PAGE 67

42 Figure 2.2.10. (a) Schematic diagram of electro n gun column. (b, c, and d) Photographs of the JOEL SEM at LAMSAT with the major components identified. The tungsten filament is heated by a current to a temperature of 2000-2700 K which results in thermionic emission of electrons and accelerates them to energy in the range 0.1–30 keV. A series of magnetic lenses focuses the electron beam to a small spot on the specimen. When the focused electron be am interacts with the specimen various signals including secondary el ectrons (SEs), backscattere d electrons (BSEs), Auger Specimen Chamber Electron Gun Column Electron Beam Anode Magnetic Lens BackScattered Detector Stage Sample Secondary Electron Detector Specimen Chamber Electron Beam Column Secondar y Electron Detector Stage (d) (a) (b) (c) Control Console Moveable Aperture Vacuum Tube

PAGE 68

43 electrons, X-rays, cathodoluminescence are ge nerated. SEs are mostly used for imaging as they mainly indicate the informati on about the specimen surface. However, BSE imaging provides compositional information as they interact with the positively charged nucleus of the specimen and are scattered at large angles (0 to 180 ). The JOEL JSM6390LV SEM is equipped with both the BS E and SE detectors. The machine has a maximum resolution power of 3 nm at an acceleration voltage of 30 kV and working distance (WD) of 8mm. The magnifica tion could be varied from 5x to 300,000x. 2.2.3. Energy Dispersive X-ray Spectroscopy Energy dispersive X-ray Spectroscopy (EDS ) is a type of el ectron spectroscopy, which uses the unique energy levels of the characteristic X-rays emitted from the atoms of the sample under investigation. Figure 2.2.11 (a) shows the major components of the EDS detector from Oxford Instruments INCA X-sight 7582M that was attached to the JOEL SEM described in Section 2.2.2. The worki ng principle for EDS is very simple. As shown in Figure 2.2.11 (b), when the high ener gy primary electron beam from the SEM interacts with the atoms of sample it creates an “interaction volume,” typically several microns in diameter. Several signals are ge nerated along with characteristic X-rays, which are fingerprints of the individual at oms encountered. These X-rays can penetrate through the material, allowing them to escap e and be detected by the EDS detector. Because the intensity of the individual X-ray is related to the quantity of the “parent atom” in the interaction volume, quantitative elemental analysis can be obtained from the sample. From the EDS spectra one can investigate the elementary composition of the materials and the presence of any foreign im purity in the sample. The stoichiometry of

PAGE 69

44 the sample is also tested from the elementary analysis of the EDS spectrum. In this work, the EDS scans were always performed using Cu calibration. Figure 2.2.11. (a) Photograph of the Oxford Inst ruments EDS detector that was attached to JOEL SEM. (b) Schematic diagram of signals generated when electron beam from SEM interacts with the sample. 2.2.4. Atomic Force Microscope Atomic Force Microscope (AFM) is a very high resolution scanning probe technique that can be used to analyze the su rface of the thin films. The AFM works much the same way as a phonograph or a profilometer, only on a much smaller scale [81]. As shown in Figure 2.2.12 (a), a very sharp tip (few nm) that is attached to a cantilever is dragged on the sample surface. During the scan, the probing tip is brought in close proximity to the sample surface and gets affected by Van der Waals forces between the atoms of tip and the surface. This causes a ve rtical deflection of the cantilever reflecting the topography of the surface. The deflection is recorded as a change in the direction of a Electron Beam Secondary electrons (10-100 nm) Back-scattered electrons (1-2 m) Auger electrons Transmitted electrons X-rays Cathodoluminescence Sample EDS Detector Dewar for Liquid N2 EDS Detector Electron Gun Column SEM (b) (a)

PAGE 70

45 reflected laser beam from the top of the cantilever picked up by a photodiode [81]. By collecting the height data for a succession of lines it is possible to form a three dimensional map of the surface features. Figure 2.2.12 (b) shows the major components of the Digital Instruments III Atomic Force Microscope that was used to scan the surfaces of the thin films in this work. Figure 2.2.12. (a) Schematic diagram of the basi c working principle of a typical AFM. (b) Photograph of Digital Instruments III Atomic Force Microscope. Laser Sample Cantilever & Tip Photodiode Laser Sample Feedback Electronics PZT Scanner Cantilever & Tip (b) (a) 3.4 mm 1.6 mm 0.3 mm 10 m

PAGE 71

46 2.3. Electrical Characterization The electrical properties of the films were measured using the Van der Pauw technique [82 84]. Since the thicknesses of the films were much smaller than their length and breadth, and they were grown on sy mmetrical (1 cm x 0.5 cm) substrates, the Van der Pauw method could be successfully applied. The electrical contacts were made with silver epoxy and gold wires. They were placed on the peri phery of the f ilm as shown in the schematic diagram in Figure 2.3.1 (a). Figure 2.3.1. Schematic diagrams of Van der Pauw technique of thin film (a) resistivity and (b) Hall measurements, respectively. During the Hall measurements the magnetic field (H) was applied perpendicular to the film plane as shown in the sc hematic diagram in Figure 2.3.1 (b). The electromagnet could provide a maximum field of 0.5 T. This was kept constant for all measurements. For measuring the temperatur e dependence of resi stivity and carrier concentration the samples were cooled in a continuous He flow cryostat which could reach temperatures as low as 30 K. A V A V H (a) (b) Thin Film

PAGE 72

47 2.4. Magnetic Characterization All the magnetic measurements in this work were performed using a commercial Physical Property Measurement System ( PPMS) from Quantum Design equipped with a Vibrating Sample Magnetometer. As shown in Figure 2.4.1, the PPMS consists of two major components: a liquid Helium Dewar with a 7 T longitudinal superconducting magnet and a temperature controller which can operate in the range 1.9 K to 400 K. The orientation of the thin film sample with respect to the magnetic field inside the PPMS can be varied. Figure 2.4.1. Photograph of the magnetic m easure system, PPMS from Quantum Design. The sample surface can be positioned either parallel or perpendicular to the applied magnetic field to measure the in-pla ne and out-of-plane magnetizations. Such Dewar for Liquid He Temperature Controller Sample Probe

PAGE 73

48 measurements are required for identification of the easy or hard axis of magnetization in thin films. If the easy axis of magnetization is parallel to the film plane then in-plane magnetization would shows saturation at lo wer applied fields than out-of-plane magnetization and vice versa. In this wor k, all the magnetization measurements were carried out at 10 K or 300 K in range of magnetic fields from -5 T to 5 T.

PAGE 74

49 2.5. Ferroelectric Characterization The ferroelectric properties of the thin films were measured using a specially designed and assembled experimental setup. As shown in Fi gure 2.5.1 the set-up consists of a Precision LC Materials Analyzer, an optical microscope with video monitor, a probe station, and control PC. The Precis ion LC Materials Analyzer from Radiant Technologies is the ferro-electric tester com ponent. The optical microscope is from Zeiss equipped with a Sanyo color CCTV camera and video monitor. The probe station includes sample stage holder, micro-ma nipulators and pole pieces. The micromanipulators on the probe station are KR N-01A ‘DC’ positioners from J-Micro Technology. The manipulators have a 0.3 inch m ovement range in x, y, and z directions which sets a limit to the separation between th e electrodes on the film. The manipulator’s base is magnetic which helps in reducing noise due to mechanic al vibrations during measurement. For the proper polarization of the ferroel ectric capacitors, very small electrodes were fabricated (approx. 100 m diameter) which could be seen only through the microscope and in the video monitor scr een (Figure 2.5.1). The n eedles like pole pieces have a 20 m tip and are made from tungsten or Pd alloy. They are used to connect the electrodes on the film to the ferro-teste r (Figure 2.5.1). The Preci sion LC ferro-tester includes many features such as +/10 V options, minimum pulse width of 50 s, minimum rise time @ 5 V of 40 s with 10 kHz Fatigue Frequency and minimum Hysteresis loop of 10s. The va rious tests that can be pr eformed include Hysteresis, Leakage Current, Charge, Retention, I (V), C (V), Fatigue, PUND, and imprint.

PAGE 75

50 Figure 2.5.1. Photographs of the home-made se t-up used for ferroelectric measurements of thin films. Optical Microscope Control PC Video Monitor Precision LC Materials Analyzer Probe Station Film Micrometer Pole Pieces Probe Station Micromani p ulators Film Surface Bottom Electrodes Top Electrodes (100 m diameter) Substrate Pole Pieces Video Screen x y z x z y Clockwise up Manipulator

PAGE 76

51 For Hysteresis measurement the Precision LC uses an in built modified SawyerTower circuit as shown in Figure 2.5.2 [85]. Figure 2.5.2. Block diagram of Sawyer-Tower ci rcuit used for hysteresis measurement of ferroelectric thin film. The circuit has been modified to avoid noise by applying virt ual ground [86] The FE thin film capacitor is usually fabricated using top and bottom electrodes as shown in Figure 2.5.2. By measuring the po tential (V) across a referen ce capacitor in series with the FE thin film capacitor (FEC) one can de termine the charge (Q) on the FEC using the relation [86]: Q = I dt = j C Voej t dt = CV (2.8) when the input voltage is V = Voej t and C is the capacitance of reference capacitor. When two capacitors are in series the ch arge on each capacitor must be the same (ideally), so the electric charges on the reference capacitor and the FEC are the same. V P Waveform Generator Display Reference Capacitor FE Thin film Ca p acitor P V Virtual Ground Top Electrode Bottom Electrode Ferroelectric Thin Film Substrate V t

PAGE 77

52 Since the reference capacitor has a very hi gh capacitance, most of the voltage drops across the FEC. Thus the signal V in the disp lay represents the volta ge across the sample. The P signal is proportional to the charge on the FEC. In the FEC there is a remnant polarization, i.e., the electric dipole moments remain aligne d in the direction it was poled by the applied field even after this field ha s been removed. Thus a plot of P versus V displays hysteresis [87]. In order to confirm the accuracy of th e polarization measurements using the custom built ferroelectric tester, the syst em was calibrated usi ng a standard capacitor (RT040903T018-SKTBD) from Radiant Te chnologies. Figure 2.5.3 (a) shows a photograph of the reference st andard capacitors. Figure 2.5.3 shows a schematic diagram of the cross-section. The standard capacitor is a PZT thin film grown on Si substrate having a top and bottom electrode configurati on. The known values of polarization of the standard capacitor provided in the specifica tion data sheets were compared with the values obtained from the measurement. The same values were obtained. This validated the polarization values obtain ed for the deposited films. Figure 2.5.3. (a) Photograph of the standard FE capacitor used for calibration of the ferrotestor. (b) Schematic diagram of the cro ss-section of the standard FE capacitor. Standard Capacitor (a) (b) Si wafer 550 mTop Electrode Bottom Electrode Ferroelectric material 0.25 m PZT protective layer 0.5 m SiO2 foundation layer

PAGE 78

53 2.6. Chapter Summary This chapter gave an outline of the experimental methods and procedures used in this work. High quality thin films were prep ared using PLD. Structural characterization was studied using XRD, SEM, EDS and AFM. Magnetic measurements were performed using PPMS. Measurements of ferroelectric properties required fabr ication of a homemade ferro-tester system. Following chapters elaborate on the results obtained from these experiments.

PAGE 79

54 CHAPTER 3: CFO-PZT BILAYER THIN FILMS In recent years there have been numerous reports of the magneto-electric (ME) effect in layered CFO-PZT structures mostly grown on Pt/Ti/SiO2/Si substrates [88-93] and some on MgO [94] or STO [ 95] substrates. It is considered that the ME effect in such composite structures is mediated mechanic ally by the mutual interaction between the constituent elastic, magnetic and electric components [89, 94]. Ho wever the structure property relationships in such composite th in films have remained elusive. Hence a systematic study on the growth of these multiferroic composite structures was undertaken. In this thesis, composite bilayered thin films consisting of CFO and PZT as ferromagnetic and ferroelectric phases, re spectively, were chosen and fabricated. Horizontal bilayer configura tion was selected for the ease of fabrication. Moreover the layered structures allowed fo r the effective separation of the conducting magnetic phase so as to enhance the polarization in the in sulating ferroelectric laye r. More importantly, the lattice strain and interlayer interaction was investigated with ease in such horizontal heterostructures [89, 96, and 97]. The coup ling effect between the CFO and PZT was investigated through the stress analysis of the constituent layers using high resolution Xray diffraction. The first two sections of the chapter discuss the CFO and PZT single layer films. The third section describes the CFO-PZT bilayer films.

PAGE 80

55 3.1. Cobalt Ferrite (CFO) Thin Films Cobalt ferrite, CoFe2O4, belongs to the family of spin el-type ferrites and is one of the important magnetic materials with high coercivity, moderate magnetization and highest magnetostriction coefficient [17, 35 38]. The magnetocrystalline anisotropy constant (K1) at room temperature for CFO is 2 x 106 erg/cm3, which is one of the highest for ferrites [98, 99]. However, the direction of anisotropy in CFO thin films can change depending on the growth conditions [100, 101], the choice of substrate, and whether the film is grown under compression or tensi on [102, 103]. Stress anisotropy comes from stresses produced in the CFO films duri ng the deposition process, or by thermal expansion differences, or lattice mismatch be tween substrates and films [37]. If grown under tension, CFO thin films have a tendency to grow with the easy axis perpendicular to the substrate which makes it an attractive material for applications in magneto-optic recording [17]. CFO has a very hi gh magnetostriction coefficient ( 100 = -200 x 10-6 to 590 x 10-6) [98]. For this reason CFO has been extensively explored as a promising magnetostrictive material for applications in actuators, sensors, and transducers [104 – 107]. CFO is considered as a key component for multiferroic multilayers or composites [96]. Epitaxial CFO/PZT composites [108] ha ve exhibited multiferroicity. CFO has also found applications in microwave devi ces due to its high resistivity (~107 .cm) and large permeability at high frequency (~1000MHz) [ 11]. Due to the good insulating properties and high Curie temperature, it has been used in magnetic tunn el junctions [109]. To summarize, CFO thin films have great potential for technological applications varying from magnetic recording to microwav e devices. Hence a systematic study of the magnetic properties of CFO thin would be beneficial.

PAGE 81

56 3.1.1. Experimental Details The CFO target was prepared by standard pressing-sintering technique. Weighed amount of high purity CFO powder was uniformly grinded and then pressed into a disc with a diameter of 30 mm and thickness of 5 mm This disc was sinter ed in a furnace at 1000 C for 2 hours to obtain a dense hard target. The target density wa s calculated to be 4.5 g/cm3. CFO films were grown on vari ous single crystal substrates such as Si (100), c-cut sapphire (Al2O3) (0001), MgO (100) and SrTiO3 (STO) (100). Table 3.1 summarizes the deposition conditions of the CFO films. Th e deposition rate under these conditions was 0.1 nm/s. Table 3.1. Deposition conditions of CF O films on different substrates. SampleSubstrateLaser FluenceGrowth Temperatures O2 Pressure Film Thickness (J/cm2) Ts ( C)pO2 (mT) (nm) CFO-SiSi (100)24501050, 100, 200 CFO-Al2O3Al2O3 (001) 245010200 CFO-MgOMgO (100)245010200 CFO-STOSTO (100)245010200 3.1.2. Results and Discussions The following sections describe the struct ure-property relations hips of epitaxial CFO thin film on MgO and STO substrates. A detailed discussion on polycrystalline CFO films grown on Si and Al2O3 substrates has been included in Appendix A. 3.1.2.1. Structural Properties The microstructure of the CFO target su rface was first stud ied. Figure 3.1.1 shows SEM images of the prepared target surface exhibiting tetragonal crystals of CFO.

PAGE 82

57 Figure 3.1.1. SEM images of CFO target surface. Table 3.2 shows the Co/Fe ratio in the CFO target and the deposited films on different substrates, obtained from the EDS analysis. The average Co/Fe ratio is close to 2 confirming the proper stoichiometry. Table 3.2. Ratio of Co/Fe in CoFe2O4 target and thin films, obtained by EDS analysis. SampleCo/Fe ratio CFO target1.96 0.11 CFOSi2.04 0.08 CFOMgO2.01 0.06 CFOSTO2.04 0.08 3.1.2.1.1. Epitaxial CFO thin films The small lattice mismatches of CFO (facecentered cubic, lattice parameter, a = 8.391 ) with MgO (face-centered-cubic, 2 x lattice parameter = 8.42 ); and STO (primitive cubic, 2 x lattice parameter = 7.81 ), respectively, allow for the epitaxial growth of CFO-MgO and CFOSTO films. The lattice mismatch at room temperature

PAGE 83

58 was calculated using the relation (as ao)/as (%) where as and ao are the lattice parameters of the substrate and the bulk powder. The cal culated values of lattice mismatches for CFO-MgO and CFO-STO are 0.34% and -7.4%, respectively. The surface morphology of CFO-MgO f ilm was examined using AFM. Figure 3.1.2 (a) shows an AFM image of the CFO surf ace. The as-deposited film exhibits a uniform and flat surface with a root-m ean-square value of surface roughness (Rrms) of 2.084 m. The average grain size is around 50 nm. Figure 3.1.2 (b) shows a section analysis of the film surface where the height of surface features is plotted against a horizontal line scan. It is evident that the surface features are small (<5 nm) indicating a highly flattened surface. Such a flattened surf ace could be indicativ e of a layer by layer growth mechanism as reported earlier [110]. Figure 3.1.2. (a) AFM image of CFO film on MgO substrate. (b) Section analysis of CFO film on MgO. Horizontal and vertical dist ances between red markings are 3.008 m and 1.288 m, respectively. Figures 3.1.3 (a, b, and c) show the XRD -2 scans for the CFO powder, CFOSTO and CFO-MgO films, respectively. The CFO powder XRD pattern (Figure 3.1.3 a) shows all the characteristic peaks of CFO [111 ]. The peaks match with the face-centered cubic phase with space group Fd-3m (227). The XRD patterns for CFO-STO and CFO-

PAGE 84

59 MgO films (Figures 3.1.3 b, and c) show only th e (400) peak of CFO with no trace of any impurity peaks even in the logarithmic scale. This demonstrates the epitaxial growth. The inset to Figure 3.1.3 (c) shows a close-up view for the CFO (400) and MgO (200) peaks. Due of the larger lattice mismatch (7.4 %) for CFO film grown on STO, CFO (400) peak is shifted significantly to lower values (Figure 3.1.3 b) comp ared to the corresponding powder peak (Figure 3.1.3 a). This peak shift indicates that the lattice structure of the CFO film lattice is possibly tetragonally distorted as reported earlier [112, 113]. 2030405060 102103104 42.843.043.3 CFO (400)(c)CFO-MgO2 ( de g) 102103104 STO (100) STO (200) CFO (400)(b)CFO-STO Intensity (a.u.)102103104 (331)(a)CFO Powder(440) (511) (422) (400) (222) (311) (220) (111) 2 (deg) Intensity (a.u.) CFO (400) Figure 3.1.3. XRD patterns of (a ) CFO powder target, and epita xial CFO films grown on (b) STO (100), and (c) MgO (100) substrates, respectively. The inset to (c) shows the details of MgO (200) (left) and CFO (400) (right) peaks.

PAGE 85

60 In order to verify the epitaxial relationship, scans and rocking curves were performed for the CFO-STO and CFO-MgO fi lms. Figures 3.1.4 (a, and d) show the scans for CFO films on STO and MgO substrates, respectively. 45135225315 -1.0-0.50.00.51.0 (a) (deg) 45135225315 (d) (deg)-1.0-0.50.00.51.0 (b)FWHM =0.915o (deg) (e)FWHM =0.076o (deg) 555759616365 Intensity (a.u.) Intensity (a.u.)(c)(440) (511) 2 (deg) Intensity (a.u.)53555759616365 (f)(440) (511) 2 (deg) Figure 3.1.4. Left column of gra phs (blue) (a, b, c) represen t the film grown on STO and the right column (red) (d, e, f) repr esents films grown on MgO. (a, and d) scan spectra from (311) CFO reflection. (b, and e) Rocki ng curves of CFO (400) peaks. (c, and f) Asymmetric scans of the (511) and (440) planes of the CFO films. The scan measurements were conducted us ing Bragg’s reflection from the (311) plane of CFO. In both cases the peaks of the spectra occur at intervals of 90

PAGE 86

61 confirming the cubic symmetry and in plane epitaxial growth. Figures 3.1.4 (b, and e) show the rocking curves ( -scans) about the (400) plan e of CFO-STO and CFO-MgO, respectively. The FWHM value for CFOMgO film is 0.076, which confirms an excellent crystallographic texture along (100) direction. However, FWHM in the case of CFO-STO film is larger (0.915) (Figure 3.1.4 b). This is due to the larger lattice mismatch between CFO and STO as mentioned earlier. Figures 3.1.4 (c and f) show the asymmetric scans of the CFO (511) and (440) planes. 3.1.2.1.2. Residual Stress in CFO thin films In the unstressed bulk CFO powder both the out-of-plane (a) and in-plane (a) lattice parameters are same in all directions. However in the CFO thin film due to the strains induced by lattice mismatches betw een substrates and the films, the aand a values are different. In order to study th e effect of strain on CFO films, the a and a were calculated. The out-of-plane lattice parameters (a) were calculated from the XRD -2 scans (Figure 3.1.3). The in-pla ne lattice parameters (a) were calculated from the asymmetric scans about the CFO (511) and ( 440) planes (Figures 3.1.4 c, and f). Table 3.3 summarizes all the lattice parameters and calculated stra ins. The in-plain strain ( ) was calculated by using the formula = (a ao)/ao where ao is the unstressed bulk lattice parameter for CFO (ao = 8.391 ). The in-plane stress ( ) was calculated using = Y where Y is Young’s modulus [98] for CFO along the (100) or in-plane direction (Y100 = 1.5 x 1012 dyne/cm2). Stress anisotropy constant (Ka) was estimated using Ka = (3/2) 100 where 100 for CFO [98, 100] is -590 x 10-6. From the listed values in Table 3.3 it is observed that in case of CFO-MgO film, a and a differ by only 0.15%. However a

PAGE 87

62 (8.401 ) is slightly larger than a (8.388 ) which implies that there is a small lateral stretching along the film surface. Table 3.3. In plane and out of plane lattice parameters obtained from x-ray diffraction (XRD) peaks and the strain calc ulated using in plain lattice parameters. In plane stress calculated using Young’s modulus Y= 1.5 x 1012 dyne/cm2. Anisotropy calculated using stress. Out-of-plane In-plane In-planeIn-plane Stress Anisotropy Samplelat. par.lat. par.StrainStress (Ka) aa (x109)(x106) () () (dyne/cm2) (erg/cm3) CFO-MgO 8.3888.401 0.0020.0013 0.00021.9 0.41.8 0.3 CFO-STO8.4868.297 0.009-0.011 0.001-16.5 0.114.6 1.4 This indicates that the CFO film on Mg O grows under subtle te nsile strain (0.13 %) with respect to bulk powder. On th e other hand for the CFO-STO film, the a (8.486 ) is larger than a (8.297 ) which implies that ther e is a compression along the film plane. The CFO-STO film grows with a larger compressive strain ( 1.1%) with respect to bulk powder. This difference between the a and a values results in a tetragonal distortion in the film grown on STO. Figur e 3.1.5 shows a schematic diagram of the different strain situations that the CFO fi lm undergoes under different lattice mismatches.

PAGE 88

63 Figure 3.1.5. Schematic diagrams show the (a) unstressed CFO bulk powder (b) CFO lattice under in-plane tensile stress on MgO substrate with a > a and (c) CFO lattice under high in-plane compressive st ress causing tetra gonal distortion a> a. To further emphasize the importance of stre ss anisotropy in epitaxial CFO films the sin2 technique (Section 2.2.1) of residual stress calcula tion was used. Figure 3.1.6 shows the 2 scans of CFO-MgO film for various tilts. The plot of lattice spacing d as a function of sin2 is shown in the inset of Figure 3.1.6. The good linear fit of data points indicates that the residual stress within th e area under scan is nearly homogeneous. The stress ( ) calculated from slope of the linear fit for CFO-MgO film is = -12.19 x 109 dyne/cm2. Due to the proximity of the substrate and the film peak, a background due to MgO substrate was subtracted fr om the slope. The values Y, and d0 used were as follows for CFO: Y = 1.5 x 1012 dyne/cm2, = 0.26 [114] and d0 = 2.0979 . The stress anisotropy (Ka) is obtained from Ka = (3/2) 100 where 100 for CFO is -590 x 10-6 and = -12.19 x 109 dyne/cm2, to be 10.79 x 106 erg/cm3. This value of Ka is larger than the magnetocrystalline anisotropy of bulk CFO (2 x 106erg/cm3). The value of Y and are CFO film on STO (compressive stress) CFO film on MgO (tensile stress) CFO bulk powder (unstressed) 8. 391 8. 391 8.428. 42 (a) (b) (c) 8. 388 8.401 In-plane stretchin g 7.81 7. 81 8. 297 8. 486 Out-of-plane Stretching

PAGE 89

64 specific for (100) direction and thus are good estimates for epitaxial films. A possible error in the calculation could be due to the bulk Poisson’s ratio ( ) used. However, by changing by 25%, the value of anisotropy (Ka) only changes by around 5%. The anisotropy field (Ha) of the film grown on MgO is de termined to be about 66.8 kOe, which is lower compared to that of the film grown on STO (Ha=72.5 kOe). 4243444 5 0 100 200 300 400 0.000.020.040.06 2.084 2.088 2.092 2.096 2.100 2.104 Intensity (a.u.) 5 deg 7 deg 9 deg 10 deg 13 deg 15 deg2 ( de g) d (angs.) sin2 Figure 3.1.6. XRD -2 scan about 2 = 43.006o by varying the angle from 0 to 15o keeping the at 0o on the CFO films on MgO (100) substrate. The legend shows the values of sin2 The inset shows the plot of d vs. sin2 and a linear fit to the data points. 3.1.2.2. Magnetic Properties The magnetic measurements were conducte d both at 300 K and 10 K in magnetic fields up to 50 kOe. The in-plane and out-of-plane c onfigurations symbolized by and respectively, represent the applicati on of the magnetic fields parallel and

PAGE 90

65 perpendicular to the film planes. All the hyste resis loops were obtain ed after subtracting the diamagnetic contributi on from the substrates. Figures 3.1.7 (a, and c) and (b, and d) show the M-H loops measured at 10 K and 300 K for CFO-STO and CFO-MgO films, re spectively. The films have the same thicknesses of 200 nm. The CFO-MgO film with a lattice mismatch of 0.34% shows outof-plane anisotropy with well saturated loop in the out-of-plane direction. On the other hand, the CFO-STO film with a lattice mismatch of 7.4% di splays in-plane anisotropy with well saturated loop in the in-plane direction. Table 3.4 summarizes the saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms and the coercivity (Hc) for all the samples at 300 K and 10 K. Table 3.4. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane and out-ofplane configurations for 200 nm thick CF O films on MgO and STO substrates. The symbols and denote the in-plane and out-of-pla ne configurations respectively. 300 K SampleMs Ms M r /Ms Hc Ms Ms M r /Ms Hc (emu/cm3) ( B/Co2+) (%)(kOe) (emu/cm3) ( B/Co2+) (%)(kOe) CFO-MgO304 52.437.63.8310 62.5<13.9< 0.03 CFO-STO478 53.853.93.5> 164> 1.3>21.63.0 10 K SampleMs Ms M r /Ms Hc Ms Ms M r /Ms Hc (emu/cm3) ( B/Co2+)(%)(kOe) (emu/cm3) ( B/Co2+)(%)(kOe) CFO-MgO>4973.9>41.310441 122.530.90.3 CFO-STO541 94.363.111> 164> 1.314.60.7 The Mr/Ms ratio provides an estimate of the degree of squareness of the loops. Although the M-H loop for CFO-MgO film sugges ts an out-of-plane anisotropy the easy axis of magnetization is not well defined because the Ms and Hc are smaller than Ms

PAGE 91

66 and Hc atypical of easy axis (Table 3.4). Howeve r, the easy axis of magnetization for CFO-STO film is along the plane of th e film (Figures 3.1.7 a, and c). The Ms value for CFO-STO film at 300 K is around 478 emu/cm3 which corresponds to 3.8 B per Co-site. The magnetic moment per Co-site was calcula ted taking into account the unit cell volume (V) of fcc structure where V =ao 3 = (8.391 )3 = 590.99 3 and 8 Co sites per unit cell. This value is around 2.6 B per Co-site in case of CFO-MgO film. However, at low temperature (10 K), the Ms is closer to the bulk value of the CFO (3 B per Co-site) [17]. The Mr/Ms is about 54% in CFO-STO film s uggesting a high degree of squareness at 300 K. The high degree of s quareness and coercive field Hc both at 300 K and 10 K for CFO-STO film indicate the strong in-plane magnetic anis otropy as reported earlier [113].On the other hand the M-H loops for CFO-MgO show lower Mr/Ms values compared to those for CFO-STO. According to earlier reports the magnetic anisotropy in CFO-MgO films is highly dependent on the film thickness [115]. The observed out-ofplane anisotropy in this case is consistent with previous reports for similar film thicknesses [33, 110, 116, and 117]. The difference in magnetic anisotropy for CFO-MgO and CFO-STO films could have arisen from the fact that the CFO film s grow with tensile strain on MgO substrate and undergo compression when grown on STO. Thus, the lattice mismatch stress may have played a significant role in the observed anisotropy.

PAGE 92

67-500 -250 0 250 500 (a)CFO-STO 300K in-plane out-of -plane 300K in-plane out-of -plane (b)CFO-MgO -50-2502550 -500 -250 0 250 500 CFO-STO in-plane out-of -plane (c)10KM (emu/cm3)-50-2502550 CFO-MgO in-plane out-of -plane (d)10KH (kOe) Figure 3.1.7. (a, and c) M-H loops measured at 300 K and 10 K respectively of the 200 nm thin film grown on STO (100) for in plan e and out of plane conf iguration. (b, and d) M-H loops measured at 300 K and 10 K respec tively of the 200 nm thin film grown on MgO (100) for in plane and out of plane configurations. To address this, the uniaxial anisotr opy was calculated by using the difference between the in-plane and out-of-plane magnetization values from Figure 3.1.7. The uniaxial magnetic anisot ropy is given by [38] dM H H Kin eff M out eff uS) (0 (3.1)

PAGE 93

68 whereNM H Hex eff N = demagnetization factor, M = magnetization and Hex is the external field applied and Ms is the saturation magnetization. The superscript out and in correspond to the out-of-plane and in-plane or ientation of the films with respect to the applied magnetic field. For the case of a thin film with uniform magnetization in all x, y and z direction, Equation (3.1) reduces to 2) ( 2 1out s in s in ex out s out ex uM M H M H K (3.2) The uniaxial magnetic anisotropy (Ku) calculated for the CFO-MgO film is Ku= 9.130.34 x 106 erg/cm3. This value of Ku is 3 times larger than the intrinsic magnetocrystalline anisotropy of bulk CFO (2 x 106erg/cm3) [98, 100]. The value of the anisotropy (Ku) thus obtained is very close to th at calculated earlier using the sin2 technique (Ku= 10.79 x 106erg/cm3). The consistency of thes e two independent methods clearly indicates that the la rge anisotropy seen in the CFO-MgO film arises from the presence of large stress due to the mismatch between the film and the substrate. For the case of the CFO-STO film, the out -of-plane magnetization is not saturated up to an applied field of 50 kOe (Figure 3. 1.7). Therefore, it is not precise to use Equation 3.2 to estimate the magnetic anisotropy for this film. Altern atively, the magnetic anisotropy of CFO-STO film was es timated by using the difference between the lattice parameter of bulk (ao = 8.39 ) CFO and the corresponding thin film (a = 8.297 ). The strain ( ) due to the difference between the ao and a is -0.011 (Table 3.3). The in-plane stress ( ) is -16.50 x 109 dyne/cm2 (Table 3.3). Magnetoela stic stress anisotropy constant (Ka) is 14.6 x 106 erg/cm3. This value is larger compared to that obtained for the film grown on MgO (100). This is consistent with the fact that stress is larger for the film

PAGE 94

69 grown on STO than for the film grown on MgO. It is the presence of larger stress in the film grown on STO that causes a larger shift in the (400) Bragg reflection compared to the bulk (Figures 3.1.6 a, a nd b). The anisotropy field (Ha) of the film grown on STO is estimated to be about 59.6 kOe, using the expression Ha=2Ka/Ms. 3.1.3. Conclusions To summarize, the magnetic anisotropies in epitaxial CFO thin films were investigated. Epitaxial films on MgO (100) with a lattice mismatch of 0.35% showed outof-plane anisotropy whereas the films on STO (100) with a lattice mismatch of 7.4% displayed in-plane anisotropy. Stress anisot ropy calculated from angle-dependent x-ray diffraction analysis confirmed that the change in anisotropy originates from the lattice mismatch strains.

PAGE 95

703.2. Lead Zirconium Titanate (PZT) Thin Films Ferroelectric thin films of PbZr0.52Ti0.48O3 (PZT) have attracted enormous research interest in recent years for a wide variety of applications in memory and microelectromechanical (MEMs) devices [39, 43]. PZT has also been used as a component of multiferroic structures [96]. Thin film deposition techniques, such as sputtering [119], pulsed laser deposition (PLD ) [120], molecular beam ep itaxy [121], metal-organic chemical vapor deposition [122], and sol-gel pr ocessing [123] have b een extensively used to fabricate PZT thin films and hetero-structures. PLD offers unique advantages in terms of st oichiometric transfer of material from multi-component targets to the as-deposited films; high deposition rate, and inherent simplicity for growth of multilayered structur es with precise control of composition and crystallinity (see Chapter 2, Section 2.1) However, laser ablation of PZT causes preferential evaporation of the volatile element Pb from the target. PZT films require deposition temperatures (Ts) between 500 C to 600 C for good crystallinity [124]. At such high temperatures the films are Pb depl eted due to the high vapor pressure of Pb [42]. This Pb deficiency is responsible for the coexistence of a meta-stable pyrochlore (non-ferroelectric) phase with the perovskite PZT struct ure, thereby degrading the ferroelectric properties of the films [43]. To compensate for the Pb loss, a common practice is to add excess PbO during the preparation of the targets. The composition of PZT films also depends strongly on the lase r energy density or fluence and partial pressure of ambient O2 (pO2) [43]. Generally, a high pO2 (~200-400 mTorr) and high KrF fluence (> 3.5 J/cm2) are used during deposition [44, 121]. At such high fluences and pO2, the highly forward directed laser-induced plume gives rise to non-uniform and

PAGE 96

71 particulate laden films which are undesirabl e in heterostructure growth. To overcome these adverse effects a dual-laser deposition (PLDDL) process was adopted [72 76] that significantly reduced the thic kness variation and particul ate density on the deposited films. This led to the growth of smooth uniform films with enhanced ferroelectric properties. 3.2.1. Experimental Details PZT films were grown on single crystal MgO (100) and SrTiO3 (STO) (100) substrates using both single (PLDSL) and dual laser deposition (PLDDL) (see Chapter 2, Section 2.1). The laser-target interactions were studied using tw o targets, namely a stoichiometric (PbZr0.52Ti0.48O3) PZT target and a PZT target with 30 atomic (at.) % excess PbO. Henceforth in the text, the nomenclature PZTST and PZTPbO will denote stoichiometric PZT (PbZr0.52Ti0.48O3) and PZT with 30 atomic (at.) % excess PbO. The PZT targets were purchased from Kurt J. Leskar Company (1” Dia. x 0.25” Thick, Density 7.41 g/cm3). All the films were deposited at 550 C under a pO2 of 500 mT using the PZTPbO target. 3.2.2. Results and Discussions The crystallinity of the PZT targets was characterized using XRD. Figure 3.2.1 shows the XRD scans for the PZTST and PZTPbO targets. Both the scans match with the tetragonal PbZr0.52Ti0.48O3 phase with no other impurity phases besides PbO. Due to tetragonal perovskite crystal structure of PZT and almost same lattice parameters (a=b= 4.036 , c= 4.146 ), the equivalent perpendicu lar planes are in close occurrence to each other in the XRD spectra (Figure 3.2.1).

PAGE 97

72202530354045505560 (001/100) (101/110) (111) (002/200) (102/201) (112/211)Intensity (arb. units)2 ( de g) PbO (002) PbO (220) PbO (020) PbO (111)PbZr0.52Ti0.48O3 (+ 30 at.% PbO) PbZr0.52Ti0.48O3 Figure 3.2.1. XRD patterns of stoichiometric PZT target and PZT target with excess PbO. 3.2.2.1. Laser-target interactio n and Plume Diagnostics In order to investigate the Pb depl etion during ablation, the laser-target interactions and the ablated plumes were studied. A genera l procedure was followed to systematically investigate the events that o ccur during the laser-target interaction. The target surface was irradiated by 1000 laser pul ses at a pulse rate of 10 Hz for various fluences in pO2 of 500 mT. The target was kept stati onary during this time. The repeated ablation by the focused laser beam created a rectangular 2 mm x 3 mm spot or lasertarget interaction site (Figure 3.2.2 a). Henceforth the nomenclature LTIS will be used to denote laser-target interactio n site, in the text. When viewed under SEM, uniformly distributed conical features were revealed near the center of the LTIS (Figure 3.2.2 b). Multiple EDS scans were performed estimate the average composition of the target after ablation. The conical structures were imaged in detail as shown in Figure 3.2.2 (d). The physical transformation of the target surface after ablation can be seen by comparing

PAGE 98

73 Figure 3.2.2 (d) to the unablated target surf ace in Figure 3.2.2 (c), both imaged under the same magnifications. Figure 3.2.2. SEM images of the PZT target surface after irradiat ion by 1000 laser pulses showing (a) the rectangular spot size, (b, and d) conical features near the center of the interaction site and (c ) the unablated target surface, respectively. 3.2.2.1.1. Single KrF laser ablation Figure 3.2.3 (a) shows a typical SEM image of the PZTST target surface near the center of a LTIS after irradi ation by a single KrF laser at a low fluence of 1 J/cm2. The conical structures are charac terized by a smooth, globular c one-tip and a distinct conebody underneath. All the cones have a common or ientation and point in the direction of the incoming laser beam. EDS measurements were performed randomly at various locations on different cone tips and cone bodies. 2 mm 3 mm (a) (b) (c) (d)

PAGE 99

74 Figure 3.2.3 (a) SEM image of conical struct ures formed on a PZT target surface after repeated ablation by KrF excimer laser. (b ) Histogram of at. % ratios of Pb/(Zr+Ti), Zr/(Zr+Ti) and Ti/(Zr+Ti) at random locations on the c one tips and bodies. Figure 3.2.3 (b) shows a histogram of at. % ratios of Pb/(Zr+Ti), Zr/(Zr+Ti) and Ti/(Zr+Ti) obtained from EDS an alysis. It is seen that the at. % ratios of Pb are significantly lower in the cone tips as compared to the cone bodies. However the Zr and Ti ratios remain almost the same. In other words, the cone tips are extremely Pb deficient as compared to their bodies. This suggests th at the cone tip originates from the melt and solidification of the cone on th e top end and after the laser pulse. One of the theories that (a) (b)Cone-body Cone-tipAtomic % (arb. units) Pb/(Zr+Ti) Ti/(Zr+Ti) Zr/(Zr+Ti)Cone Tip Cone Body

PAGE 100

75 explain these cone formations is the so-calle d ‘impurity shielding mechanism’ [68]. This mechanism attributes the cone formation to vaporization-resistant impurities which in this case are Zr (melting point, mp =1865 C) and Ti (mp = 1668 C), which have higher melting temperatures as compared to Pb (mp = 327.5 C). The materials surrounding the impurities (i.e. Pb) are preferentially removed during ablation. As a result of this erosionformation process, cones grow in length as the laser exposure increa ses [125]. This effect generally leads to the non-stoi chiometry in the films [126]. Figure 3.2.4 shows the SEM images of LTISs on the PZTST target surface after irradiation by the KrF laser with increasing fluences from 1 to 6 J/cm2, denoted as KrF 1 to 6 J/cm2. Severe surface melting and formation of columnar structures are observed. Two types of cones can be identified: one t ype with a well defined cone tip and body at fluences KrF 1 to 2 J/cm2 (Figures 3.2.4 a, and b) as e xplained earlier, and the other type having only the cone tip at fluences KrF 3 to 5 J/cm2 (Figures 3.2.4 c to f).Unlike the cones formed at low fluences (Figures 3.2.4 a, and b) which point towards the incoming laser beam, the cones formed at higher fluences (Figures 3.2.4 c to f) are oriented with the cone axis almost perpendicular to the ta rget surface. Also the space between the cone tips is larger than those at 1 and 2 J/cm2. These differences suggest that the formation mechanism of the cones in Figures 3.3.4. (c f) is different from the impurity shielding mechanism and are not associated with preferential ablation. Several models have been proposed to explain the cones formed at higher fluences (Figures 3.2.4 c to f). Among them, the hydrodynamic sputtering mechanism proposed by Kelly et al. [127] and the capilla ry wave instability mechanism proposed by Brailovsky et al. [128] are the most acceptable ones. All these models attribute the cone

PAGE 101

76 formation to instabilities on the melted ta rget surface created by the laser pulse. The motion of the melted surface creates aspirates which subsequently solidify to form the cones. The SEM images of the PZTPbO target after ablation also showed similar structures. Figure 3.2.4. SEM images of PZT target af ter irradiation by 1000 pulses of a KrF laser beam in 500 mT O2 ambient at laser fl uences of (a) 1J/cm2, (b) 2J/cm2, (c) 3J/cm2, (d) 4J/cm2, (e) 5J/cm2, and (e) 6J/cm2, respectively, denoted as KrF 1J/cm2 to 5J/cm2. Figures 3.2.5 (a, and b) show the vari ation of at. % ratios of Pb/(Zr+Ti), Zr/(Zr+Ti) and Ti/(Zr+Ti), obtained from EDS, at the LTISs for the PZTST and the KrF 2J/cm2 KrF 4J/cm2 KrF 1J/cm2 KrF 3J/cm2 KrF 5J/cm2 KrF 6J/cm2 (b) (c) (d) (e) (f) (a)

PAGE 102

77 PZTPbO targets, respectively. Here 0 J/cm2 represent the target compositions before ablation. For the PZTST target (Figure 3.2.5 a), the at % ratio of Pb decreases from 0.93 at 0 J/cm2 to 0.66 at 1 J/cm2 and reaches saturation at higher fluences. However, the Zr and Ti ratios remain almost constant. On the other hand, for the PZTPbO target, the at. % ratio of Pb decreases from 1.3 at 0 J/cm2 to 0.83 at 1 J/cm2 with almost no change in the Zr and Ti ratios. However, above 3 J/cm2, the at. % ratio of Pb reaches saturation at 1.0, with the Zr and Ti ratios still the same. This suggests that above the threshold fluence of 3 J/cm2, congruent ablation occurs from the PZTPbO target. Hence the PZT films were grown using the excess PbO target. 012345 0.4 0.6 0.8 1.0 1.2 1.4 0.4 0.6 0.8 1.0 (b) Pb/(Zr+Ti) Zr/(Zr+Ti) Ti/(Zr+Ti)KrF laser Fluence (J/cm2)(a) Atomic % (arb. units) Pb/(Zr+Ti) Zr/(Zr+Ti) Ti/(Zr+Ti) Figure 3.2.5. Chemical compositions using EDS analysis of the ablated target surfaces as a function of KrF laser fluence for (a) stoi chiometric PZT and (b) PZT (30 at.% PbO) target. In order to study the effect of ambient gas pressure on the film stoichiometry, PZT films deposited on Si substrates at room temperature by varying the pO2 in the

PAGE 103

78 chamber. Figure 3.2.6 shows the variation of at. % of Pb, Zr, and Ti on the PZT films obtained from EDS analysis. It is evident that high pO2 (~500mT) is facilitates more Pb incorporation in the deposited PZT films. 0100200300400500 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Zr/(Ti+Zr) Ti/(Ti+Zr)Ambient O2 pressure (mT) Pb/(Ti+Zr)Atomic % (arb. units) Figure 3.2.6. Chemical compositions using ED S analysis of PZT films deposited on Si substrates at room temperature by varying the ambient O2 pressures. The visible emissions from the laser-i nduced plumes were captured using ICCD imaging. The ICCD imaging system was a ligned normal to the plume propagation to image the axial (on-axis) and transverse plume expansion. Figure 3.2.7 (a) shows a schematic diagram of the arrangement of targ et and the ablated plumes as imaged by the ICCD camera. The FWHMs of the plumes were measured 1 cm away from the target as shown in Figure 3.2.7 (a). Initial studies of the plasma plumes showed that using the detector with zero gain the visible plumes lasted for about 11 s. Figure 3.2.7 (b) shows an ICCD image of the particulates ejected from the PZTPbO target surface during ablation

PAGE 104

79 using high KrF fluence of 5 J/cm2. These particulates get deposited on the film surface as molten spherical droplets. This is called “splashing”. The particles ejected from the target become visible only after the intense plume disappears (duration time ~ 11 s). Thus the image was obtained using a 500 s gate or exposure time on the camera. Figure 3.2.7. (a) Schematic diagram of the arra ngement of target and the ablated plume as viewed by the ICCD camera. (b) ICCD image the particulates ejected from the PZT (30 at. % excess Pb) target surface during ablation using high KrF fluence of 5 J/cm2. In order to capture the total emitted light from the plumes, a 20 s exposure time was set in the camera. Figure 3.2.8 shows the ICCD images for the total visible emission spectra of the single KrF laser-ablated plumes at various fluences in pO2 of 500 mT. With higher fluence the plasma is more excited as evident from the axial and transverse expansion of the plumes. To summarize, from the single KrF laser ablation studies it is observed that the preferential evaporation of Pb from the target is suppresse d at high fluences (> 3 J/cm2) where the high energy facilitate s the congruent evaporation of materials from the target surface. Further, high O2 ambient during deposition restrict s the Pb loss. The results are consistent with earl ier reports [43, 44]. Rotating Target 1cm (a) Plasma Plume (b)

PAGE 105

80 Figure 3.2.8. ICCD images of total visibl e emission spectra of single laser plumes varying the KrF fluences as (a) 1 J/cm2 (b) 2 J/cm2 (c) 3 J/cm2, and (d) 4 J/cm2 under 500 mT pO2. 3.2.2.1.2. Dual (KrF and CO2) laser ablation The most crucial requirement in PLDDL is the correct choice of the KrF and CO2 laser fluences. From the PLDSL study, a KrF fluence of 3 J/cm2 was chosen. The CO2 fluence was chosen such that it was lower than the ablation threshold of the target material but enough to melt the target surface [72]. Figure 3.2.9 shows a series of ICCD images of total visible spectra of CO2 laser ablated plumes with increasing CO2 fluences from 1 J/cm2 to 3 J/cm2. It is evident that at 3 J/cm2 ablation occurs. Thus, a CO2 fluence of 2 J/cm2 was chosen. For all PLDDL films throughout this work, it will be assumed that the CO2 fluence is 2 J/cm2 if not mentioned. KrF 1J/cm2 1cm 1cm 1cm 1cm KrF 2J/cm2 KrF 3J/cm2 KrF 4J/cm2 (a) (d) (c) (b)

PAGE 106

81 Figure 3.2.9. ICCD images of to tal visible emission spectra of CO2 laser ablated plumes with increasing CO2 fluences from 1 J/cm2 to 3 J/cm2. In order to find the optimum condition for the efficient coupling of the CO2 and KrF laser outputs, the laser-target interaction were studied using PLDDL with a KrF fluence of 3 J/cm2 but varying the peak-to-peak (p-p) inter-pulse delay ( t). Figure 3.2.10 shows the KrF and CO2 pulse waveforms at various p-p inter-pulse delays Figure 3.2.10 (c) represents the situation when the KrF pulse arriving at the target about 50 ns after the onset of the CO2 pulse. CO2 1 J/cm2 CO2 2 J/cm2 CO2 3 J/cm21cm 1cm 1cm (a) (c) (b)

PAGE 107

82 0250500750 0250500750 t = 250 ns t (b) (a) KrF CO2t = 145 ns (c) KrF CO2 KrF CO2t = 100 ns (e) (d) KrF CO2 KrF CO2t = 50 ns KrF CO2t = 0 nsTime (ns)Intensity (arb. units) (f)t = -50 ns Figure 3.2.10. KrF and CO2 pulse waveforms at (a) 250 ns (b) 145 ns, (c) 100 ns, (d) 50 ns, (e) 0 ns, and (f) – 50 ns of p eak-to-peak inter-pulse delays ( t), respectively. Figure 3.2.11 shows SEM images of the LTISs on the PZTST target at various p-p inter-pulse delays ( t). The surface features are distinctly different depending on the delay ( t). For t = 250 ns and 145 ns (Figures 3. 2.11 a, and b), the ablated surface shows typical conical features indicating pref erential evaporation. At such large delay times (145 ns < t < 250 ns), the CO2 laser energy is totally absorbed by the preceding KrF induced plasma, increasing the plasma temperature [72]. The higher plasma

PAGE 108

83 temperature coupled with the high pO2 keeps the molten region on the target surface in a liquid state for a larger time. During this time, the volatile Pb es capes from the molten pool leaving behind columnar stru ctures. However, from Figures 3.2.11 (c, and d) it is observed that surface features indicate congruent ablation similar to high fluence PLDSL (Figures 3.2.4 c to f). This probably indicate s an optimum coupling of the laser energies. At delay times (50 ns < t < 100 ns), the rising edge of CO2 pulse (Figures 3.2.10 c, and d) first preconditions the target to produce a shallow, transient, molten layer, from which the slightly delayed KrF pulse then initia tes the ablation. The receding end of the CO2 pulse is absorbed by the KrF induced plasma to produce higher ionization. However, the situation is quite different for delay times in the range -50 ns < t < 0 ns (Figures 3.2.10 e, and f). When the CO2 pulse arrives at the target much earlier than the KrF pulse, it completely melts the target surface and faci litates preferential evaporation of Pb. The delayed KrF pulse ablates this completely liquid target surface with low Pb content. This is confirmed by analyzing the chem ical composition using EDS of the LTISs as shown in Figure 3.2.12. The at. % of Pb at the LTIS is the least in the range of delay times, -50 ns < t < 0 ns, as explained earlier. For a delay time t > 50 ns, the at. % of Pb reaches saturation. The Zr and Ti content remains almost constant. The SEM and EDS analysis of the PZTPbO also showed similar results. Thus from the study of p-p interpulse delay times in PLDDL it is concluded that the optimum coupling of the laser outputs and congruent ablation condition is ac hieved in the range, 50 ns < t < 100 ns. This was supported by ICCD imaging of the plasma plum es at various delays. It was observed that the transverse cross-sectional expansion of th e plumes was the largest in the range, 50 ns < t < 250 ns.

PAGE 109

84 Figure 3.2.11. SEM images of laser-target in teraction sites on the stoichiometric PZT target using KrF (3 J/cm2) and CO2 (2 J/cm2) lasers by varying the peak-to-peak interpulse delay ( t) as (a) 250 ns, (b) 145 ns, (c) 100 ns, (d) 50 ns, (e) 0 ns, and (f) -50 ns, respectively. t = 250 ns (a) t = 145 ns (b) t = 100 ns (c) t = 50 ns (d) t = 0 ns (e) t = -50 ns (f)

PAGE 110

85-50050100150200250 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Zr/(Ti+Zr) Ti/(Ti+Zr) Atomic % (arb. units)Interpulse dela y ( t ) ( ns ) Pb/(Ti+Zr) Figure 3.2.12. Chemical compositions using EDS analysis of the ablated stoichiometric PZT target surfaces using KrF (3 J/cm2) and CO2 (2 J/cm2) lasers by varying the peak-topeak inter-pulse delay ( t). The effect of the varying the KrF laser fluence but keeping the CO2 fluence at 2 J/cm2 was studied using 50 ns and 100 ns dela ys. Figures 3.2.13 and 3.2.14 show SEM images of LTISs on the PZTST target at 50 ns and 100 ns delay, respectively. From the target surface features in Fi gures 3.2.13 and 3.2.14, it is obvious that a minimum KrF fluence of 3 J/cm2 is required for congruent ablati on. However, overall there is no difference in the surface morphologies betwee n the two delay times as seen in Figures 3.2.13 and 3.2.14. The surface features for the LTISs on the PZTPbO target also showed similar structures and followed the same trend.

PAGE 111

86 Figure 3.2.13. SEM images of laser-target in teraction sites on the stoichiometric PZT target using CO2 fluence at 2 J/cm2 but varying the KrF fluence as (a) 1 J/cm2, (b) 2 J/cm2, (c) 3 J/cm2 and (d) 4 J/cm2, respectively keeping the peak-to-peak inter-pulse delay ( t) at 50 ns. Figure 3.2.14. SEM images of laser-target in teraction sites on the stoichiometric PZT target using CO2 fluence at 2 J/cm2 but varying the KrF fluence as (a) 1 J/cm2, (b) 2 J/cm2, (c) 3 J/cm2 and (d) 4 J/cm2, respectively keeping the peak-to-peak inter-pulse delay ( t) at 100 ns. KrF 1 J/cm2 CO22 J/cm2 (d) (c) (b) (a) KrF 1 J/cm2 CO22 J/cm2 KrF 1 J/cm2 CO22 J/cm2 KrF 1 J/cm2 CO22 J/cm2 t = 100 ns t = 100 ns t = 100 ns t = 100 ns KrF 1 J/cm2 CO22J/cm2 (d) (c) (b) (a) KrF 1 J/cm2 CO22 J/cm2 KrF 1 J/cm2 CO22 J/cm2 KrF 1 J/cm2 CO22 J/cm2 t = 50 ns t = 50 ns t = 50 ns t = 50 ns

PAGE 112

87 Figures 3.2.15 (a, and b) show chemi cal compositions of the LTISs on PZTPbO target at delay times of 50 ns and 100 ns, respectively. Although there are no structural differences as observed in SEM images (Figures 3.2.13 and 3.2.14) it is evident from Figure 3.2.15 that the Pb depletion is highe r for 50 ns delay above a fluence of 3 J/cm2. Therefore, the interpulse delay was fixed as 100 ns in PLDDL. 01234 0.4 0.6 0.8 1.0 1.2 1.4 0.4 0.6 0.8 1.0 1.2 1.4 t = 100 ns (b) t = 50 ns (a) KrF laser Fluence (J/cm2)Atomic % (arb. units) Pb/(Ti+Zr) Zr/(Ti+Zr) Ti/(Ti+Zr) Figure 3.2.15. Chemical compositions using ED S analysis of the ablated PZT (30 at. % excess PbO) target surfaces using CO2 fluence at 2 J/cm2 but varying the KrF fluence at the peak-to-peak delay ( t) of (a) 50 ns and (b) 100 ns, respectively. Figure 3.2.16 compares the chemical com positions obtained from EDS analysis of the ablated PZTPbO target surfaces using PLDDL and PLDSL. It is evident that the at. % of Pb loss is less in PLDDL compared to PLDSL. From the laser-target interaction study, it can be concluded that PLDDL minimizes the Pb loss during ablation.

PAGE 113

8801234 0.40 0.45 0.50 0.55 0.60 0.8 0.9 1.0 1.1 1.2 1.3 Dual laser ablation Pb/(Zr+Ti) Zr/(Zr+Ti) Ti/(Zr+Ti) Single laser ablation Pb/(Zr+Ti) Zr/(Zr+Ti) Ti/(Zr+Ti)KrF laser Fluence (J/cm2)Atomic % (arb. units) Figure 3.2.16. Chemical compositions using EDS analysis of the ablated PZT target surfaces with excess PbO, using dual and single laser ablations. This can be associated with the more expanded plume in PLDDL due to higher ionization [72]. Figure 3.2.17 shows the ICCD images for the total visible emission spectra of the PLDDL ablated plumes at various KrF fluences. With higher fluence the plasma is more excited as evident from the ax ial and transverse expansion of the plumes with higher laser ener gies. Further the PLDDL plumes (Figure 3.2.17) have broader transverse expansion as compared to PLDSL plumes (Figure 3.2.7).

PAGE 114

89 Figure 3.2.17. ICCD images of total visible emission spectra of dual laser plumes varying the KrF fluences as (a) 1 J/cm2 (b) 2 J/cm2 (c) 3 J/cm2, and (d) 4 J/cm2 but keeping the CO2 fluence at 2 J/cm2 and the inter-pulse dela y at 100 ns under 500 mT pO2. Figures 3.2.18 (a, and b) shows the ICCD images of the laser-induced plasma plumes for the PLDDL and (b) PLDSL, respectively, using KrF fluence of 3 J/cm2. In order to compare the images, the intensities were normalized to the PLDSL plume (Figure 3.2.18 b). The overall total intensity of the PLDDL plume was 145% greater than that of PLDSL emission. Also, the broader transverse ex pansion, measured on-axis 1 cm from the target (Figure 3.2.7 a), of the plume in dua l-laser (28.0 mm FWHM) compared to single– laser (18.7 mm FWHM) clearly exhibits coupling of the laser energies in dual-laser ablation. This allowed the growth of more uniform film over a larger area and led to enhanced film properties. KrF 1J/cm2 CO2 2J/cm2 t = 100ns 1cm 1cm 1cm 1cm KrF 1J/cm2 CO2 2J/cm2 t = 100ns KrF 1J/cm2 CO2 2J/cm2 t = 100ns KrF 1J/cm2 CO2 2J/cm2 t = 100ns (a) (c) (b) (d)

PAGE 115

90 Figure 3.2.18. Time integrated ICCD images of the visible emission spectra of laserinduced plumes in 500 mTorr ambient O2 for (a) dual-laser ablation using KrF and CO2 laser fluences of 3 J/cm2 and 2 J/cm2 respectively, with the inte r-pulse peak-to-peak delay of 100 ns, and (b) the KrF single-lase r ablation with a fluence of 3 J/cm2. The scale bar represents 10 mm. 3.2.2.2. Structural Properties The single and dual laser processed PZT films were characterized using XRD, SEM and AFM. Henceforth in the text, the nomenclature PZTDL and PZTSL refer to duallaser and single-laser deposited PZT films, resp ectively. Figures 3.2.19 (a, b, and c) show XRD scans of PZTSL films deposited using KrF fluence of 2 J/cm2 and 5 J/cm2 and a PZTDL film using KrF 1 J/cm2 and CO2 2 J/cm2 ( t = 100 ns) on STO substrates, respectively. The films are hi ghly epitaxial with peaks onl y from the [100] planes of tetragonal PZT structure. There are no obser ved peaks from secondary phase formation even in the log-scale intensity. From Figures 3.2.19 (a, and b) it is can be concluded that the crystallinity PZTSL films is improved with higher fluence as evident from the higher peak intensities at 5 J/cm2 (Figure 3.2.19 b) compared to 2 J/cm2 (Figure 3.2.19 a). However, the PZTDL film (Figure 3.2.19 c) shows im proved crystallinit y with the high peak intensities although grown at lower Kr F fluence. All the films have the same thickness of about 350 nm.

PAGE 116

912030405060708 0 102103104105106 102103104105106 10210310410510 6 (c) 2 ( de g) * (300) (200) (100)(b) (a)Intensity (counts) Figure 3.2.19. XRD scans of PZTSL films deposited using KrF fluence of (a) 2 J/cm2 and (b) 5 J/cm2 and a (c) PZTDL film using KrF 1 J/cm2 and CO2 2 J/cm2 ( t = 100 ns) on STO substrates, respectively. The substrate peaks are denoted by *. In order to further confirm the in plan e epitaxy, rocking curves were performed about the PZT (200) plane for the PZTSL at 5J/cm2 and PZTDL films as shown in Figure 3.2.20. Both films have excellent in plane epitaxy indicated by th e narrow FWHM of the rocking curves. The FWHMs of the rocking curves for the PZTSL at 5J/cm2 and PZTDL films are 0.536 and 0.541 respectively. EDS analysis of the films revealed that the at. % ratio of Pb for PZTSL films deposited at 2J/cm2 and 5 J/cm2 was 0.39 and 0.85, respectively. However, the PZTDL film had high Pb content of 0.84.

PAGE 117

92 -1.5-1-0.500.511.5 (deg)Intensity (arb. units) Single laser Dual laser Figure 3.2.20. Rocking curves about PZT ( 200) plane for single and dual laser grown PZT films on STO substrates. Figures 3.2.21 (a, b, and c) s how AFM images of the of PZTSL films deposited using KrF fluence of 2 J/cm2 and 5 J/cm2 and PZTDL film using KrF 1 J/cm2 and CO2 2 J/cm2 ( t = 100 ns) on STO substrates, respectively. The PZTSL film at 5 J/cm2 (Figure 3.2.21 b) exhibits a roug her surface and high density of larg er particulates as compared to the PZTSL film at 2 J/cm2 (Figure 3.2.21 a). The root-mean-square surface roughness (Rrms) values for PZTSL films at 5 J/cm2 and 2 J/cm2 are 12 nm and 3 nm, respectively. However, a drastic reduction in particulate density and su rface roughness is observed in PZTDL film (Figure 3.2.21 c). The Rrms value PZTDL film is 2 nm. This confirms the effectiveness of dual laser ablation in gr owing high Pb content, particulate free and smooth PZT films with the desired perovs kite structure and no impurity phases.

PAGE 118

93 Figure 3.2.21. AFM images of PZT films de posited using single laser ablation (a) at 2 J/cm2 and (b) at 5 J/cm2 and (c) dual laser ablation. All scan areas are 5 m x 5 m. Figures 3.2.22 (a, b, and c) show SEM images of PZTSL films grown using low fluence of at 2 J/cm2, high fluence of 5 J/cm2, and PZTDL film deposited using KrF 1 J/cm2 and CO2 2 J/cm2 ( t = 100 ns) on STO substrates, respectively. Although the PZTSL film at 2J/cm2 (Figure 3.1.22 a) shows a relatively clean surface, the low Pb content degraded the ferroelectric prope rties. On the other hand, the PZTSL film at 5 J/cm2 (Figure 3.2.22 c) had good crystallinity (F igure 3.2.19) and the proper Pb content, however the high density of particulates on the film surface make it unsuitable for heterostructure growth. On the contrary, the PZTDL film not only had the good crystallinity and proper Pb c ontent, it exhibited low part iculate density compared to

PAGE 119

94 PZTSL at 5 J/cm2. Figure 3.2.22 (d) shows the details of one of the particulates of PZTSL film at 2 J/cm2. The spherical shape of the particle is a clear indication that it was “splashed” as a molten droplet on the film surface [66]. Thus, th e origin of these particulates is from the ta rget surface during ablation. Figure 3.2.22. SEM images of PZTSL films deposited at 550 C on STO substrates at (a) low fluence of 2 J/cm2 and (b) high fluence of 5 J/cm2, and (c) PZTDL film, respectively. (d) SEM image of details of one of the particulates on PZTSL film at 2 J/cm2. Although the PLDDL films deposited using low KrF fluence of 1 J/cm2 as described above, exhibited particulate fr ee surface with high Pb content, their ferroelectric properties were not as high as films deposited at high KrF fluence. Thus, in

PAGE 120

95 order to enhance the ferroelectric properties, PZTDL films were deposit ed using a high KrF fluence of 3 J/cm2, keeping other parameters constant. For comparison, PZTSL films were also deposited under the same conditions. Further to test the polarization of the PZT capacitors, metallic oxide La0.7Sr0.3MnO3 (LSMO) electrodes were used. The growth and characterization of LSMO electrode s are described in Appendix B. 3.2.2.2.1. LSMO/PZT/LSMO capacitor The following paragraphs describe stru ctural properties of LSMO/PZT/LSMO capacitors grown on STO and MgO substrates A schematic diagram of the capacitor structure is shown in Figure 3.2.23. The th icknesses of the indi vidual LSMO and PZT layers are 100 and 500 nm, respectively. Figure 3.2.23. Schematic diagram of PZT thin film capacitor fabricated using LSMO top and bottom electrodes. The crystallinity of the PZT capacitors wa s examined by XRD. Figures 3.2.24 (a, and b) show the XRD patterns for LSMO/PZTDL/LSMO/STO and LSMO/PZTSL/LSMO/STO, respectively. In both cases, only strong ( l 00) ( l = 1, 2, and 3) diffraction peaks of LSMO and PZT are observe d along with those of the single-crystal STO (100) substrate. This implies that the indi vidual layers have an epitaxial relationship with each other. The films are highly text ured without any secondary impurity phases MgO / SrTiO3 (100) substrate LSMO ( 100 ) ( 100 nm ) PZT ( 100 ) ( 500 nm )

PAGE 121

96 even in log scale. The PZT peaks were inde xed with the tetragonal phase with space group P4mm (99). The (200) planes of LSMO and STO substrate are shown in detail in the inset (I) of Figure 3.2.24 (a). Inset (II) of Figure 3.2.24 (a) and inset (III) of Figure 3.2.24 (b) show rocking curves ( scans) about the PZT (100) plane for PZTDL and PZTSL, respectively. The smaller FWHM value for PZTDL (0.1 ) as compared to PZTSL (0.5 ) relates to better in-plane orientation for the dual-laser deposited film. 20304050607080 101102103104105 101102103104105 46.5046.75 10.410.8 10.410.8 * *LSMO (200) STO (200)(III)2 (deg)FWHM 0.5oFWHM 0.1o*(I) (II) PZT (100) PZT (200) PZT (300) (deg) (deg)PZT (100) PZT (100)(b)Intensity (counts)2 ( de g) * (a) Figure 3.2.24. XRD -2 scans of LSMO/PZT/LSMO capacitors on STO (100) substrates grown using (a) dual and (b) si ngle-laser ablations, respectively. Inset (I) shows the details of STO (200) and LSMO (200) peaks around 2 values of 46 Inset (II) shows the rocking curve of the PZT (100) peak with FWHM value of 0.1 Inset (III) shows the rocking curve of the PZT (100) p eak with the FWHM value of 0.5 The PZT ( l 00) (l = 1, 2, 3) reflections are denoted by , and the LSMO/STO (00l) peaks are denoted by *.

PAGE 122

97 Figures 3.2.25 (a, and b) show AF M images of the surfaces of PZTDL and PZTSL films on STO substrates, respectively. Clearly, PZTDL (Figure 3.2.25 a) exhibits a smoother surface and smaller grain size with a root mean square (Rrms) roughness value of 1.6 nm compared to 11.5 nm for PZTSL (Figure 3.2. 25 b). Figure 3.2.25. AFM images of the PZT su rface for the LSMO/PZT/LSMO capacitors on STO substrates deposited using (a) dual-laser ablation and (b) single-laser ablation. The Rrms surface roughness values for (a) and (b) are 1.6 nm and 11.5 nm, respectively. Since the surface features on the PZTDL film were much smaller those on the PZTSL film, the z-height for the AFM scan for PZTDL had to be set as two times smaller than that for the PZTSL (Figures 3.2. 25 a, and b), in or der to make the surface features

PAGE 123

98 observable. The surface features for PZTDL have been shown in detail for a smaller scan area in the lower panel of Figure 3.2.25. Figures 3.2.26 (a, and b) show the XRD -2 scans for LSMO/PZTDL/LSMO/MgO and LSMO/PZTSL/LSMO/MgO, respectively. 20304050607 0 43444546 43444546 212223 212223 Intensity (arb. units)* *(b) 2 (deg)PZT (101)(II) (II) (I) (I)PZT (300) PZT (300) MgO (200) MgO (200) LSMO (200) LSMO (200) PZT (200) PZT (200) LSMO (100) LSMO (100) PZT (100) PZT (100)(a) Figure 3.2.26. XRD -2 scans of LSMO/PZT/LSMO capacitors on MgO (100) substrates grown by (a) dual and (b) single-lase r ablation, respectively. Insets (I) and (II) in Figure 3.2.30 (a) and (b) show the details of PZT(100)/LSMO(100) peaks and MgO(200)/PZT(200)/LSMO(200) peaks, respectiv ely. The small peaks denoted by are artifacts from the MgO substrates.

PAGE 124

99 The XRD patterns indicate an epitaxial relationship between the films and the substrate. The insets (I) and (II) in Figure 3.2.26 (a) show the details of PZT(100)/LSMO(100) peaks and MgO(200)/PZT(200)/LSMO(200) peaks, respectively. The corresponding peaks for the PZTSL films have been shown in the inset (I) and (II) in Figure 3.2.26 (b). The better in-plane epitaxy in the PZTDL film compared to PZTSL film is confirmed from the rocking curves about the PZT (100) plane as shown in Figure 3.2.27. 10.010.511.011.512. 0 0 200 400 600 800 1000 0 2000 4000 6000 8000 10000 Single Laser Ablation PZT (100) (b)Intensity (counts) ( de g) Dual Laser Ablation PZT (100) (a) FWHM = 0.5oFWHM = 0.8o Figure 3.2.27. XRD rocking curves about th e PZT (100) plane for the LSMO/PZT/LSMO capacitors on MgO (100) substrates grown by (a) dual and (b) sing le-laser ablation, respectively.

PAGE 125

100 Although both the films have the same thickness (500 m), the higher intensity counts and lower FWHM value (Figure 3.2.27 a) in the PZTDL-MgO films demonstrate the better crystallinity compared to PZTSL-MgO film. Figures 3.2.28 (a, and b) show AF M images of the surfaces of PZTDL and PZTSL films on MgO substrates, respectively. The smoother surface of PZTDL film is not as apparent as seen earlier for PZT-STO films (F igure 3.2.25). Nevertheless, the root mean surface roughness values is smaller (Rrms = 16.7 nm) for PZTDL as compared to PZTSL (Rrms = 22.6 nm). Figure 3.2.28. AFM images of the PZT su rface for the LSMO/PZT/LSMO capacitors on MgO substrates deposited using (a) dual-lase r ablation and (b) single-laser ablation, respectively. The Rrms surface roughness values for (a) and (b) are 16.7 nm and 22.6 nm, respectively. 3.2.2.3 Ferroelectric Properties The requirements of a good ferroelect ric thin film cap acitor for memory applications include hi gh remnant polarization (Pr), low coercive field (Ec) allowing operation at low voltages, well saturated squa re hysteresis loop, short switching times, good retention, and fatigue properties [40, 129 ]. The following section will describe the

PAGE 126

101 ferroelectric properties of epitaxial PZT films on STO and MgO s ubstrates. However, polycrystalline PZT films were also deposited on LSMO/Si(100) substrates for comparison. Figure 3.2.29 (a) shows the room temperature polarizatio n versus electric field (P-E) hysteresis loop for a 400 nm thick PZTSL film on LSMO/Si(100) substrate denoted as PZTSL(400 nm)-Si. 203040506 0 0 1000 2000 3000 4000 5000 -1000-50005001000 -40 -30 -20 -10 0 10 20 30 40 (b)(201)/(210)Si (100)* LSMO (018)(211)LSMO (202)(111)LSMO (104)(101) (110)Intensity (arb. units)2 (deg)Ec= 180 kV/cm Pr = 24 C/cm2Pmax = 36 C/cm2Driving Voltage (40 V)(a) PZTSL(400nm)-Si Polarization (C/cm 2 )Electric Field (kV/cm) Figure 3.2.29. (a) P-E hysteresis loop and (b) XRD pattern for PZTSL film on LSMO/Si(100) substrate, respectively. The maximum polarization (Pmax) of 36 C/cm2, Pr of 24 C/cm2 and Ec of 180 kV/cm was observed using a ma ximum driving voltage of 40 V. Figure 3.2.29 (b) shows

PAGE 127

102 the XRD pattern of the film exhibiting th e polycrystalline nature. The results are consistent with earlier reports on PZT/LSMO/Si films [130, 131]. Figure 3.2.30 shows the P-E loops for PZTSL films deposited at varying KrF fluences of 2 J/cm2, 3 J/cm2, and 5 J/cm2 on STO and MgO substrates. All the films have the same thickness of 500 nm. The measurem ents were performed with 9 V driving voltage at various hysteresis periods from 100 ms to 1000 ms. From Figures 3.2.30 (a, and b) it is evident that the PZTSL-STO and PZTSL-MgO films at low KrF fluence of 2 J/cm2 have poor ferroelectric behavi or. As described earlier, th is is due to the poor Pb content in the films. PZTSL films deposited at higher fluenc es (Figures 3.2. 30 c to f) have better P-E hysteresis loops with the Pr values much higher than polycrystalline PZT films (Figure 3.2.30). The polarization values increase with KrF fluence from 3 J/cm2 to 5 J/cm2. This is possibly due to the higher Pb co ntent in the films grown with increasing fluences. The PZTSL films grown using KrF fluence of 5 J/cm2 (Figures 3.2.30 e, and f) exhibit the highest po larization, however, the P-E loops changed drastically with the change of hysteresis periods or frequency which makes them not suitable for device application. Thus, it appears that PZTSL films at KrF 3 J/cm2 (Figures 2.2.30 c, and d) exhibit the most acceptable ferroelectric beha vior with better frequency response, among all the films. However, from Figures 3.2.30 (b, and d) it is seen that the loops are positively biased. This voltage shift means that an internal bias field has built up in the PZTSL film which is again inappr opriate for device applicati ons [132]. All these adverse effects were overcome in the PZTDL films.

PAGE 128

103 Figure 3.2.30. P-E hysteresis loops for PZTSL films at KrF fluences of (a, b) 2 J/cm2, (c, d) 3 J/cm2, and (e, f) 5 J/cm2 on STO and MgO substrates, respectively, measured at 9V driving voltage and varying the hys teresis period from 100 to 1000 ms. (a) (b) (c) (d) (e) (f) -200-160-120-80-4004080120160200 -10 -8 -6 -4 -2 0 2 4 6 8 10 100 ms 300 ms 500 ms 700 ms 900 ms 1000 msPZTSL-STO KrF 2 J/cm2Polarization (C/cm2)Electric Field ( kV/cm ) -200-160-120-80-4004080120160200 -20 -15 -10 -5 0 5 10 15 20 100 ms 300 ms 500 ms 700 ms 900 ms 1000 msPZTSL-MgO KrF 2 J/cm2Polarization (C/cm2)Electric Field ( kV/cm ) -200-160-120-80-4004080120160200 -100 -80 -60 -40 -20 0 20 40 60 80 100 PZTSL-STO KrF 3 J/cm2 100 ms 300 ms 500 ms 700 ms 900 ms 1000 msPolarization (C/cm2)Electric Field ( kV/cm ) -200-160-120-80-4004080120160200 -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 PZTSL-MgO KrF 3 J/cm2 100 ms 300 ms 500 ms 700 ms 900 ms 1000 ms Electric Field ( kV/cm ) Polarization (C/cm2)-200-160-120-80-4004080120160200 -100 -80 -60 -40 -20 0 20 40 60 80 100 100 ms 300 ms 500 ms 700 ms 900 ms 1000 msPZTSL-STO KrF 5 J/cm2Polarization (C/cm2)Electric Field ( kV/cm ) -200-160-120-80-4004080120160200 -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 PZTSL-MgO KrF 5 J/cm2 100 ms 300 ms 500 ms 700 ms 900 ms 1000 msPolarization (C/cm2)Electric Field ( kV/cm )

PAGE 129

104 The following paragraphs will describe the ferroelectric properties of PZTDL films deposited on MgO and STO substrates using KrF fluence of 3 J/cm2 and CO2 fluence of 2 J/cm2 with 100 ns p-p delay. For comparison PZTSL films were also grown under the same conditions. Figure 3.2.31 shows the P-E loops for PZTSL and PZTDL capacitors deposited on MgO and STO substrates. -100 -50 0 50 100 -150-75075150 -100 -50 0 50 100 -150-75075150 1 V 3 V 5 V 7 V 9 V(d) (c) (b) (a)PZTDLMgO PZTSLMgO 1 V 3 V 5 V 7 V 9 V 1 V 3 V 5 V 7 V 9 VPZTSLSTO 1 V 3 V 5 V 7 V 9 VElectric Field (kV/cm)Polarization (C/cm2)PZTDLSTO Figure 3.2.31. P-E hysteresis loops for PZTSL and PZTDL films on STO and MgO substrates, respectively, at va rious maximum driving voltages.

PAGE 130

105 All the films have the same thicknesses (500 nm) and LSMO top bottom electrode configurations. All the loops were measured using a rang e of driving voltages from 1 V to 9 V at 1Hz. From Figure 3.2.34 it is evident that the P-E loops for PZTDL films (Figures 3.2.31 b, and d) show well satura ted square behavior at lower voltages as compared to PZTSL films (Figures 3.2.31 a, and c). Further the PZTDL films show enhanced polarization compared to PZTSL films. Figures 3.2.32 shows the P-E loops for PZTDL-MgO and PZTSL-MgO films measured at an applied voltage of 9 V. Sim ilarly, Figures 3.2.33 shows the P-E loops for PZTDL-STO and PZTSL-STO films at 9V. The polarization values have been summarized in Table 3.5. The coercive field (Ec) was calculated using Ec = |(Vc(+)+Vc(-))|/2d where Vc is the nominal voltage required for switching. Table 3.5. Summary of maximum polarization (Pmax), remnant polarization (Pr), nominal switching voltage (Vc), coercive field (Ec) and leakage current density JL (A/cm2) for PZTSL and PZTDL films grown on MgO and STO subs trates. Data measured at 9 V driving voltage at 1 Hz. Sample SubstratePmax P r Vc EcJL ( C/cm2) ( C/cm2) (V)(kV/cm) (A/cm2) (+) (-) (+)(-) x 10-9PZTSLMgO6445-392.09-1.9140.00.43 PZTDLMgO9677-752.45-1.9944.40.27 PZTSLSTO6834-451.99-1.5835.70.43 PZTDLSTO10891-922.65-1.6743.20.20 From the listed values in Table 3.5 it can be seen that the Pmax values at 9 V for the PZTDL films have increased about 50% from their values for the PZTSL films. The Pr values have also almost doubled for PZTDL films. The Pr values for PZTSL (Table 3.5) are consistent with earlier reports for LSMO /PZT/LSMO capacitors grown by PLD [133,

PAGE 131

106 134]. For the PZTDL-STO film, a Pr value of 91 C/cm2 was observed. This is the first observation of such high value for remnant polarization in in-situ grown PZT films. Reported Pr values vary from 15 to 54 C/cm2 [42, 135] for PLD gr own PZT films with one of the highest Pr values reported as 70 C/cm2 for PLD grown PZT films, but required ex-situ post annealing at 750 C in air by rapid thermal annealing (RTA) technique [136]. -200-160-120-80-4004080120160200 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 LSMO electrodes PZT (500 nm) LSMO (100 nm) MgO(100) substrate PZTDLMgO PZTSLMgOPolarization (C/cm2)Electric Field ( kV/cm ) Figure 3.2.32. P-E hysteresis loops measur ed with 9V driving voltage for PZTDL and PZTSL films on MgO substrates. The inset shows a schematic illustration of the capacitors with the thicknesses of LSMO and PZT laye rs being 100 nm and 500 nm, respectively. The P-E loop for PZTSL-STO in Figure 3.2.33 shows a slight shift towards positive voltages as a consequence of high bui lt-in field [133]. However, the P-E curve for PZTDL did not show any such asymmetric be havior which could result in imprint failures.

PAGE 132

107-200-160-120-80-4004080120160200 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 LSMO electrodes PZT (500 nm) LSMO (100 nm) STO(100) substrate PZTDLSTO PZTSLSTOPolarization (C/cm2)Electric Field ( kV/cm ) Figure 3.2.33. P-E hystere sis loops measured with 9V driving for PZTDL and PZTSL films on STO substrates. The inset shows a schematic illustration of the capacitors with the thicknesses of LSMO and PZT layers be ing 100 nm and 500 nm, respectively. Leakage currents in ferroelectric thin film capacitors are measured by applying a DC voltage until the current level stabilizes [86]. High leakage currents are unacceptable as they create anomalous e ffects such as distortion in hysteresis loops [86]. Leakage current density (JL) implies current per unit area of the electrode. Leakage current densities (JL) in all PZTDL and PZTSL capacitors were measured by applying a stress voltage of 9 V for a period of 1 s (soak time). The values have been shown in Table 3.5. Figure 3.2.34 shows a comparison of the leakage current densities (JL) for PZTDL and PZTSL capacitors on MgO and STO substrates as mentioned earlier. The PZTDL capacitors have lower leakage curr ent densities compared to PZTSL capacitors.

PAGE 133

10802004006008001000 10-1010-9 10-1010-910-8 (b) PZTSLSTO PZTDLSTOJL(A/cm2)Time (ms) (a) PZTSLMgO PZTDLMgO Figure 3.2.34. Capacitor leakage current densities (JL) measured using a stress voltage of 9 V for a soak time of 1000 ms for PZTDL and PZTSL films grown on (a) MgO and (b) STO substrates under same conditions, respectively. 3.2.2.3.1. Fatigue Characterization The fatigue measurements were performed using standard positive-up negativedown (PUND) testing mode to simulate fe rroelectric memory operation. During the PUND test, a series of stress measurement pe riods is applied to the capacitor. The stress is in the form of a 10 kHz switching pulse wa veform of 0.01 ms pulse width. The pulse waveform is similar to a square wave except that the maximum positive and negative voltages are applied for only a fraction of the period of the waveform. During the remaining portion of the waveform, the sample is held at zero volts. The waveform is applied for a specified period, and then the PUND measurement is made. In each subsequent stress period, the duration (and therefore the number of switching cycles) of

PAGE 134

109 the stress waveform is doubled. The PUND ch aracterization for the films was performed using peak voltages +9 V and –9 V which simu late the read and write voltages of the memory device. The final polarization of th e device after PUND test is the difference between the (+/-) switching polarization (Psw) and linear non-switching polarization (Pns) values when a (+/-) read voltage (9 V/-9V) was applied to the capacitors. For device application the value of (Psw – Pns) should be large enough for signal detection. Figures 3.2.35 (a, and b) show the fatigue (degradation) properties of PZTDL and PZTSL films on STO and MgO substrate, respectively. 1031041051061071081091010-100 -50 0 50 100 -100 -50 0 50 100 PZTDLMgO PZTSLMgO(b) PZTDLSTO PZTSLSTOPolarization (C/cm2)Switching cycles(a) Figure 3.2.35. Results of positive-up negati ve-down (PUND) fatigue tests at 10 kHz using +/9 V read voltages for LSMO/PZT/ LSMO capacitors grown by dual-and singlelaser ablation on (a) STO (b) MgO substrates.

PAGE 135

110 From Figure 3.2.35 it is evident that there is no change in the polarization for PZTDL films even after 109 switching cycles. However there are small variations in the polarization for PZTSL films. From the PUND test it is clear that PZTDL films exhibited a better fatigue response compared to PZTSL. Possible explanation could be associated with a reduced number of oxygen vacancies at electrode-film interface for PZTDL. 3.2.3. Conclusions In summary, using the dual-laser abla tion process, PZT films have been successfully grown with the proper stoichio metry and particulate-free smooth surfaces. The films have the desired perovskite struct ure and no impurity phases. These films have superior ferroelectric properties as compared to films grown by single laser ablation. This technique could be generalized to all multi -component thin film growth with a high volatility element which would lead to non-stoi chiometric transfer of materials in other processes.

PAGE 136

1113.3. CFO-PZT Bilayer Thin Films 3.3.1. Experimental Details CFO-PZT bilayer thin films were grown on 1cm x 0.5 cm single crystalline MgO (100) and SrTiO3 (STO) (100) substrates from compressed powder targets of CFO and Pb(Zr0.52Ti0.48)O3. Henceforth, the nomenclature CFO/MgO and CFO/STO for CFO single layer thin films on MgO and STO substrates respectively, will be used in the text. Similarly, the CFO-PZT bilayer thin film s grown on MgO and STO substrates will be referred as, PZT/CFO/MgO and PZT/CFO/STO, respectively. A schematic diagram of the structure is shown in Figure 3.3.1. Figure 3.3.1. Schematic diagram of CFO-P ZT bilayer films deposited on MgO or STO substrates. CFO/MgO and CFO/STO films of simila r thicknesses were prepared under the same experimental conditions for co mparison. For the PZT/CFO/MgO and PZT/CFO/STO thin films, the CFO layer was deposited at 450 C and 10 mTorr O2 pressure for a 200 nm layer thickness. The subsequent PZT layer of same thickness was deposited at 550 C, 300 mTorr O2 pressure. Table 3.6 summarizes the deposition conditions for the individual CFO and PZT layers. MgO / SrTiO3 (100) substrate CFO (100) (200 nm) PZT (100) (200 nm)

PAGE 137

112 Table 3.6. Deposition conditions of the layers in the CFO-PZT bilayer films that were grown on MgO and STO substrates. Layers inLaser FluenceGrowth Temperatures O2 Pressure Thickness CFO-PZT bilayer (J/cm2) Ts ( C)pO2 (mT) (nm) CFO245010200 PZT2550300200 For the polarization measurements, CFO-PZ T bilayer capacitors were fabricated using conducting LSMO top and bottom electr odes. Figure 3.3.2 shows the schematic diagram of the LSMO/CFO/PZT/LSMO capac itor grown on MgO (100) and STO (100) substrates. Figure 3.3.2. Schematic diagram of the CF O-PZT bilayer capacitor fabricated using LSMO top and bottom electrodes. Table 3.7 summarizes the deposition conditions for the individual layers. The PZT layer was deposited using dual laser abla tion. In Section 3.2 it was illustrated how the ferroelectric properties of PZT could be improved using dual laser ablation. The top LSMO electrodes were deposit ed in-situ using a shadow mask that produced 100 m diameter contacts. MgO / SrTiO3 (100) substrate LSMO ( 100 ) ( 100 nm ) CFO (100) (200 nm) PZT ( 100 ) ( 500 nm )

PAGE 138

113 Table 3.7. Deposition conditions of the laye rs in the CFO-PZT capacitor using LSMO top and bottom electrodes. Layers inLaser FluenceGrowth Temperatures O2 Pressure Thickness CFO-PZT capacitor (J/cm2) Ts ( C)pO2 (mT) (nm) LSMO (bottom electrode)260010100 CFO245010200 PZTUV:3, IR:2550500500 LSMO (top electrodes)260010100 3.3.2. Results and Discussions 3.3.2.1. Structural Properties Figure 3.3.3 shows SEM images of the P ZT top layer in PZT/ CFO/Si film. The surface is dense and shows distinct grain coal escing (at higher magnification) due to the high growth temperature. Figure 3.3.3. SEM images of the PZT surf ace in CFO-PZT bilayer film grown on Si substrate.

PAGE 139

114 Figure 3.3.4 (right panel) shows the cro ss-section of a PZT/CFO/MgO film. The image is obtained by first milling a wedge patt ern (left panel) on the film surface using Focussed Ion Beam (FIB) and then looking at the exposed layers using SEM. Figure 3.3.4. SEM images of CFO-PZT b ilayer film grown on MgO substrate (crosssectional view). The small lattice mismatch (0.04 %) betw een PZT (tetragonal perovskite, lattice parameters, a=b= 4.036 , c= 4.146 ) and CF O (face-centered cubic, lattice parameter, a = 8.391 ) as well as betw een CFO and the substrates a llows for the growth of the epitaxial films. Figures 3.3. 5 (a, and b) show the XRD -2 spectra for PZT/CFO/MgO and CFO/MgO, respectively. XRD spectra for PZT/CFO/STO and CFO/STO are shown in Figures 3.3.5 (c, and d), respectively. In all the samples the single phase nature and epitaxial relationship with the substrates ar e observed. The XRD peak in CFO is assigned to the (400) plane, corresponding to the facecentered cubic (fcc) phase of CFO with space group Fd-3m (227). For PZT/CFO/MgO a nd PZT/CFO/STO films, the PZT peak is indexed to the (100) plane of tetragonal P ZT with space group P4mm (99) (Figures 3.3.5 a, and c). Due to the small lattice mism atch between MgO (face-centered-cubic, 2 x Protective Pt layer Wedge pattern PZT top surface PZT(100) CFO ( 100 ) MgO (100)

PAGE 140

115 lattice parameter = 8.42 ) and CFO, the MgO (200) and CFO (400) peaks are in close occurrence to each other in the -2 spectra. Insets to the Figu res 3.3.5 (a, and b) visibly show the MgO and CFO peaks in the samples. 1520253035404550556065 42.542.843.043.3 42.542.843.043.3 (d)CFO (400)CFO (400) STO (200) STO (200)Intensity (a.u.)2 (deg) (c)PZT (100) PZT (200) STO (100) STO (100) (b)MgO (200) CFO (400) CFO (400) PZT (200) PZT (100) (a)CFO MgO PZT2 (deg) MgO CFO2 (deg) Figure 3.3.5. XRD -2 scans for single layer CFO and bilayer CFO-PZT films grown on MgO (a, and b) and STO (c, and d) substrates respectively. The insets to (a) and (b) show the details of the MgO (200), CF O (400) and PZT (200) peaks around 43 (2 ). However, due to the larger lattice mismatch between STO (primitive cubic, 2 x lattice parameter = 7.81 ) and CFO, the (4 00) peaks of CFO are shifted significantly (Figures 3.3.5 c and d) as compared to the peaks of polycrystalline CFO [17]. The out-of-

PAGE 141

116 plane lattice parameter (a) for CFO was calculated from the XRD -2 scans (Table 3.8). In order to verify the in-plane epitaxia l relationship and cubic symmetry for the CFO and PZT layers, scans were performed. Figures 3.3.6 (a, and b) show the scan spectra from the PZT (101) plane for PZT/ CFO/STO and PZT/CFO/MgO, respectively. Figures 3.3.6 (c, and d) show the scan spectra from the CFO (311) plane for PZT/CFO/STO and PZT/CFO/MgO, respectivel y. Figures 3.3.6 (e) and (f) show the scan spectra from the CFO (311) plane for CFO/STO and CFO/MgO, respectively. In all cases, the peaks in the spectra occur at intervals of 90 suggesting the four-fold cubic symmetry and cube-on-cube growth. Figures 3. 3.6 (g, and h) show the rocking curves ( scans) about the CFO (400) planes for PZT/CFO/STO and CFO/STO, and PZT/CFO/MgO and CFO/MgO, respectively. The small full-width at half maximum (FWHM) values (< 1 ) of the rocking curves confirm a good degree of in-plane orientation for CFO in all the samples (see Table 3.8 for FWHM values). However the (400) texture is sharper in the films grow n on MgO which may be attributed to the smaller lattice mismatch between CFO and MgO. In addition, th e degree of (400) texturing of CFO weakens slightly in P ZT/CFO/STO and PZT/CFO/MgO compared to PZT/STO and PZT/MgO, respectiv ely. Figures 3.3.6 (i, and j) show the asymmetric scans of (511) and (440) planes of CFO for CF O/STO and CFO/MgO, respectively. Figures 3.3.6 (k, and l) show the similar as ymmetric scans for PZT/CFO/STO and PZT/CFO/MgO, respectively. For Figures 3.3.6 (i l) the left peaks are from CFO (511) plane and the right peaks are from CFO ( 440) planes. The averag e in-plane lattice

PAGE 142

117 parameters (a) for CFO in the samples were calculated from the asymmetric scans shown in Figures 3.3.6 (i l) (Table 3.8). (j) (i) (b) (a) (f) (e) (d) (c) 45135225315 Intensity (a.u.) Intensity (a.u.) (deg)Intensity (a.u.) (deg)45135225315 -1.0-0.50.00.51.0 (h) CFO CFO-PZT(g) (deg)-1.0-0.50.00.51.0 CFO CFO-PZT (deg) (l) (k) 5658606264 2 ( de g) 5658606264 2 (deg) Figure 3.3.6. Left and right columns re present the films grown on STO and MgO substrates respectively. (a and b) are scan spectra from PZT (101) reflection in CFOPZT bilayer film. (c and d) and (e and f) are scan spectra from (311) CFO reflection in CFO-PZT bilayer and single layer CFO films re spectively. (g and h) are rocking curves of CFO (400) peaks. (i and j) and (k and l) are asymmetric scans of (511) and (440) CFO planes for single layer CFO films and bilayer layer CFO-PZT films respectively. (i – l) Left peaks are from CFO (511) plane and the right peaks are from CFO (440) planes.

PAGE 143

118 The strain ( ) in the CFO layer was calculated by using the formula = ( a ao)/ ao, where a is out-of-plane (a) or in-plane (a) lattice parameter and ao is the bulk unstressed lattice parameter of CFO ( ao = 8.39 ) [17]. The in-plane stress ( ) in the film was calculated using the relation = Y where is the in-plane strain ( ) and the Young’s modulus value for CFO (Y = 1.5 x 1012 dyne/cm2) [98]. Table 3.8 summarizes the lattice parameters and strains calcu lated for the out-of-plane and in-plane configurations. From the strain values listed in Table 3.8 it is seen that the CFO/MgO films grows with slight in-plane tensile ( = 0.0015) and out-of plane compressive ( = -0.0005) strains. On the other hand, the CF O/STO film grows with larger in-plane compressive ( = -0.0116) and out-of plane tensile ( = 0.0124) strains. This can be attributed to the different lattice mismatches of CFO with MgO and STO substrates. The lattice mismatch at room temperatur e was calculated using the relation (as – ao)/as (%) where as is the lattice parameter of the substrat e. The calculated values for CFO/MgO and CFO/STO are 0.36 % and 7.8 %, respectively. It is also observed that the in-plane lattice parameter (a) of CFO for the PZT/CFO/MgO film (a = 8.294 ) is smaller than that of the CFO/MgO film (a = 8.403 ). This suggests that possi bly with the deposition of the PZT layer on top, the CFO layer experiences an in-plane compression that compels it to match its a to the smaller lattice pa rameter of PZT (a = b = 4.036 , c = 4.146 ). As a consequence the in-plane strain and consequently the stress is amplified in PZT/CFO/MgO. However, an opposite tren d is observed for the films grown on STO substrates. The CFO/STO film is already hi ghly strained due to the large mismatch between the STO substrate and CFO. With P ZT layer on top of it the PZT/CFO/STO film is slightly relaxed to a lower strain state.

PAGE 144

119 Table 3.8. FWHM of rocking curves about CFO (400) plane, in-plane (a) and out-ofplane (a) lattice parameters obtained from XRD peaks, in-plane ( ) and out-of-plane ( ) strains for CFO and CFO-PZT f ilms on MgO and STO substrates. FWHM of Out-of-planeIn-planeIn-plane SampleRocking curve a straina strainstress ( )()() x 109 (dyne/cm2) CFO/MgO0.0768.386-0.00058.403 0.0040.0015 0.0005 2.3 0.6 PZT/CFO/MgO0.3218.338-0.00628.294 0.007-0.0114 0.0008 -17.1 0.1 CFO/STO0.9158.4940.01248.292 0.005-0.0116 0.0006-17.5 0.09 PZT/CFO/STO0.9868.4790.01068.330 0.002-0.0071 0.0002 -10.7 0.03 In order to analyze the surface morphologies of the thin films and predict their mechanisms of growth, AFM was employed. Figure 3.3.7 (a) illustrates an AFM image of the CFO top layer for CFO/MgO film. Th e image reveals a very smooth and compact surface with a root mean square roughness (Rrms) value of 2.084 nm and small grain size with relatively uniform size distribution. As reported earlier such a flattened surface could be indicative of a laye r-by-layer growth mechanism [110]. Figure 3.3.7 (b) shows an AFM image of the PZT top layer for PZT/ CFO/MgO film. The PZT layer is relatively less smooth with Rrms value of 11.456 nm and larger grain size as compared to CFO/MgO. Uniform grain size distribution is also observed for PZT. The surface exhibited a texturing which is probably reminiscent of the epitaxial growth. Additionally, the grains appear to be preferentially elongated in one in -plane orientation. This effect can be correlated to the larger difference in the a (8.294 ) and a (8.338 ) values of the CFO layer in PZT/CFO/MgO (Table 3.8) From Figure 3.3.7 (c), which is a 3 dimensional projection of the P ZT top layer shown in Figure 3. 3.7 (b), distin ct cusp (or valleys) and dome features are observed on the film surface. It has been reported earlier that such cusps are as sociated with high stress concentr ation regions [137] which increase

PAGE 145

120 of the surface energy and gets compensated by the decrease of strain energy via elastic strain relaxation in the films. This strain relaxation of the top PZT layer suggests that probably the bottom CFO layer is more strained. The films grown on STO substrates showed distinctly different surface morphol ogies. Figure 3.3.7 (d) shows the surface of CFO/STO film. The surface appears rougher than CFO/MgO with Rrms value of 7.502 nm. It also consists of grains with various sh apes and sizes. This coul d be attributed to the island growth mode [110, 113]. Figure 3.3.7 (e) shows the PZT top surface for PZT/CFO/STO film. The various grain si zes with larger grain growth and Rrms value of 22.683 nm still conform to the island growth mechanism.

PAGE 146

121 Figure 3.3.7. AFM images of (a) CFO surf ace of CFO/MgO film, (b) top PZT surface of PZT/CFO/MgO film, (c) 3D rendition of part (b), (d) CFO surface of CFO/STO film and (e) top PZT surface of PZT/CFO/STO films. Scan areas are 1x1 m with z-height of 100 nm.

PAGE 147

122 The following paragraphs discuss the stru ctural properties of CFO-PZT bilayer films (see schematic in Figure 3.3.2) usi ng LSMO bottom electrodes grown on MgO and STO substrates (LSMO/MgO and LSMO/STO) used for the ferroelectric measurements. The small lattice mismatch es (8 % for LSMO-MgO and 0.009 % for LSMO-STO) between LSMO (pseudo-cubic, a = 0.387 nm ), MgO (cubic, a = 0.421 nm) and STO (cubic, a = 0.3905 nm) allowed for the growth of epitaxial LSMO bottom electrodes on MgO and STO substrates. Figure 3.3.8 shows the XRD -2 scan for the LSMO/PZT/CFO/LSMO film grow n on MgO substrate. [138]. 20304050607 0 0 1x1042x1043x10 4 42434445 103104105 MgO (200) PZT (200) CFO (400) Intensity (counts)2 (deg)CFO(400)/MgO(200) LSMO (100) LSMO (200)* ** PZT (100) PZT (200) PZT (300)Intensity (counts)2 (deg) Figure 3.3.8. XRD -2 scans for the CFO-PZT bilaye r grown on single crystal MgO (100) substrate with a conduc ting LSMO top and bottom electrode layers. The PZT ( l 00) where l = 1, 2, and 3 peaks are denoted by *. The LSMO ( l 00) where l = 1, and 2 peaks are denoted by The inset shows the details of MgO (200) and CFO (400) peaks. Only the CFO (400) peak and the PZT ( l 00) where l = 1, 2, and 3 peaks are observed. No impurity or additional orientat ions of CFO or PZT are observed. This

PAGE 148

123 confirms that the film is si ngle phase and epitaxial in na ture. The MgO (200) peak and CFO (400) peak have been shown in deta il in the inset to Figure 3.3.8. Figure 3.3.9 shows the XRD -2 scan for the CFO-P ZT bilayer grown on LSMO/STO substrate (see schematic diagram in Figure 3.3.2). It is again evident that the structure is highly epitaxia l in nature with only ( l 00) where l = 1, 2, and 3 orientations from the constituent layers. Due to the sma ll lattice mismatch between STO, LSMO, PZT and CFO, the (200) peaks of STO, PZT a nd LSMO, and the (400) peak of CFO have been shown in detail in th e inset to Figure 3.3.9. 203040506070 0 1x1042x1043x1044x10 4 42434445464748 102103104105 LSMO (200) STO (200) PZT (200) CFO (400) Intensity (counts)2 (deg)CFO (400) LSMO/STO (100) LSMO/STO (200) LSMO/STO (300)* ** PZT (100) PZT (200) PZT (300)Intensity (counts)2 ( de g) Figure 3.3.9. XRD -2 scans for the CFO-PZT bilaye r grown on single crystal STO (100) substrate with a conduc ting LSMO top and bottom electrode layers. The PZT ( l 00) where l = 1, 2, and 3 peaks are denoted by *. The LSMO ( l 00) where l = 1, 2, and 3 peaks are denoted by The inset shows the details of LSMO (200), STO (200), PZT (200) and CFO (400) peaks.

PAGE 149

1243.3.2.2. Magnetic Properties Figure 3.3.10 shows the magnetization (M ) magnetic field (H ) hysteresis loops for PZT/CFO/MgO and CFO/MgO, respectively. Similarly, Figure 3.3.11 shows the M-H loops for PZT/CFO/STO and CFO/STO, resp ectively. The in-plane and out-of-plane configurations symbolized by and respectively, represent the application of the magnetic fields parallel and perpendicular to the film planes. The hysteresis loops were acquired after the removal of the diamagne tic contribution from the substrates. In addition since the thickness of CFO layer was kept constant in all the films, the magnetization values were only normalized to the volume of the CFO layer assuming no magnetic contribution from the PZT layer. Table 3.9 summarizes the saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms and the coercivity (Hc) for all the samples at 300 K and 10 K. The Mr/Ms ratio provides an esti mate of the degree of squareness of the loops. The magnetic measur ements were performed both at 300 K and 10 K to emphasize the consistency of the underlying mechanisms. The Ms values for CFO films on STO substrates are larger than those on Mg O substrates [28]. The Ms value for the PZT/CFO/STO film is about 5.8 B per Co2+ site which is much higher than the theoretical bulk value of 3 B per Co2+ site [17]. From Figure 3.3.10 it is observed that the magnetization of CFO reduces in PZT/CFO/MgO film as compared to that in CFO/MgO film bot h at 300 K and 10 K. Around 25 % decrease in the Ms values is observed for CFO/MgO film with the deposition of the PZT top layer top both in th e in-plane and out-of-plane directions at 300 K (Table 3.9). However the Hc values still remain about the same. The out-of-plane anisotropy exhibited by the CFO/MgO film is cl early seen in the M-H loops measured at

PAGE 150

125 10 K (Figures 3.3.10 a, and c). The in-plane magnetization does not show any saturation even at 50 kOe while on the other hand the out-plane magnetization shows well behaved saturation. This behavior is preserved in the M-H loops of PZT/ CFO/MgO films. This suggests that the easy axis of magnetizatio n of CFO/MgO film does not change with deposition of the PZT top layer. In short, th e net effect of addition of the PZT layer on top of CFO/MgO film is an observed decreas e in magnetization with nominal change in coercivity and squareness. Table 3.9. Summary of saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms and the coercivity (Hc) for CFO/PZT bilayer and CFO single layer films measured at 300 K and 10 K. The in-plane and out-of-plane directions have been denoted by the symbols and respectively. 300 K SampleMs Ms M r /Ms Hc Ms Ms M r /Ms Hc (emu/cm3) ( B/Co2+)(%)(kOe) (emu/cm3) ( B/Co2+)(%)(kOe) CFO/MgO305 42.437.63.8310 62.5<13.9< 0.03 PZT/CFO/MgO228 21.825.81234 51.8<5.9<0.03 CFO/STO478 53.853.93.5> 164> 1.3>21.63.0 PZT/CFO/STO592 54.753.63.41901.541.50.1 10 K SampleMs Ms M r /Ms Hc Ms Ms M r /Ms Hc (emu/cm3) ( B/Co2+)(%)(kOe) (emu/cm3) ( B/Co2+)(%)(kOe) CFO/MgO>497>3.9>41.310441 123.530.90.3 PZT/CFO/MgO>400>3.2>39.64353 32.818.10.3 CFO/STO541 94.363.111> 164> 1.3>14.60.7 PZT/CFO/STO728 95.865.6112061.628.80.5

PAGE 151

126-500 -250 0 250 500 -50-2502550 -500 -250 0 250 500 -50-25-02550 (a)in-plane CFO/MgO PZT/CFO/MgO10 KM (emu/cm3) (b)in-plane CFO/MgO PZT/CFO/MgO300 K (c)out-of-plane CFO/MgO PZT/CFO/MgO10 KH ( kOe ) (d)300 K out-of-plane CFO/MgO PZT/CFO/MgO Figure 3.3.10. M-H loops measured at (a, an d c) 10 K and (b and d) 300 K for the CFOMgO and PZT-CFO-MgO films, respectively. The in plane and out of plane denote directions for the magnetic field applied para llel or perpendicular to the film plane, respectively. In contrary, magnetization of CFO/STO film increases w ith the addition of PZT on top as shown in Figure 3.3.11. Around 25% and 34% increase in the in-plane Ms can be estimated for PZT/CFO/STO at 300 K and 10 K, respectively (Table 3.9). The out-ofplane Ms also increases by 25% at 10 K in the PZT/CFO/STO film as compared to CFO/STO film. The CFO/STO film exhibits strong in-plane anisotropy with well

PAGE 152

127 saturated loops in the in-plane direction (Figur es 3.3.11 a, and b) a nd almost no saturation in the out-of-plane direction (Figures 3.3.11 c, and d). However from Figure 3.3.11 (d) it is evident that the out-of -plane magnetization at 300 K for PZT/CFO/STO shows well saturated behavior with an almost double Mr/Ms ratio (Table 3.9) compared to CFO SL film on STO. A similar trend is obs erved at 10 K (Figure 3.3.11 c). -800 -600 -400 -200 0 200 400 600 800 -50-2502550 -200 -100 0 100 200 -50-25-02550 (a) 10 K in-plane CFO/STO PZT/CFO/STO(b) CFO/STO PZT/CFO/STO300 K in-plane(c) CFO/STO PZT/CFO/STOout-of-plane 10 KM (emu/cm3)H ( kOe ) (d) 300 K CFO/STO PZT/CFO/STOout-of-plane Figure 3.3.11. M-H loops measured at (a, an d c) 10 K and (b and d) 300 K for the CFOSTO and PZT-CFO-STO films, respectively. The in plane and out of plane denote directions for the magnetic field applied para llel or perpendicular to the film plane, respectively.

PAGE 153

128 This possibly indicates a reorientation in the direction of the easy axis of magnetization for CFO/STO film with the depo sition of the PZT top layer. Thus the net effect on the magnetic properties of CFO/STO f ilm with PZT top layer is an increase in magnetization and possible change in direction of magnetic anisotropy. The effect of stress ( ) on the magnetization of a magnetostrictive material can be understood from the following thermodynamic relation [38, 139]: B H l l 4 1 1 (3.1) where B is the magnetic induction is the stress, l is the length of the material and H is the external applied magnetic field. From equation (3.1) it can be assumed that the magnetization is decreased (increased) by tens ion (compression) if the magnetostriction ( l / l ) is negative (positive) when is positive (negative). Since CFO is a negative magnetostrictive material the magnetization would be reduced by stress (tensile). The CFO/MgO film is under in-plane tensil e stress (Table 3.8), due to the lateral stretching along the film plan e which results in the a (8.403 ) being larger than a (8.386 ). With the addition of a top PZT la yer the in-plane residual stress may be increased which results in reduced magnetizat ion in PZT/CFO/MgO. On the contrary, the CFO/STO film is under large in-pla ne compressive stress with a (8.494 ) larger than a (8.292 ). With the deposition of the PZT layer, the stress is released making the a (8.330 ) in PZT/CFO/STO larger than th at in CFO/STO. Thus strain relaxation enhanced the magnetization in PZT/CFO/STO film. The results matched well with the negative magnetostriction of CFO.

PAGE 154

129 An increase in the residual stress due to the top PZT layer on the CFO layer and its effect on magnetization ha s been reported earlier on th e polycrystalline films grown on silicon [89]. Sim et. al. had suggested that the residual stress in the CFO films on Si could be intrinsic and associated with the or ientation change or defect incorporation or non-equilibrium phase formation. However the stress mechanisms were complicated and difficult to quantify [89]. In this work the residual stress could be estimated due to epitaxial growth. The suppression of the Ms with higher stress ( ) can also be understood qualitatively in terms of magnetic strain energy density, which is given by E = 3/2 s sin2 where s is the magnetostrictive coe fficient at saturation, and is the angle between the stress ( ) and the saturation magnetization Ms. When s is negative (as in the case for CFO in tension), the domains te nd to align perpendicular to the axis of tension. This maximizes the energy since is 90o, which in turn makes the domains unstable. Thus the magnetization will be d ecreased as observed in our experiment. In addition, since the out-of-plane Ms in all the films follows the same trend as in-plane Ms, there is no change in magnetic anisotropy in bilayers as compared to single layers. 3.3.2.3. Ferroelectric Properties Figure 3.3.12 shows the ferroelectric (FE) hysteresis loops for CFO-PZT bilayer films on LSMO/MgO substrates. All the data was measured using a driving voltage of 9 V that was required for the pr oper saturation of the loops bu t changing the frequency in the range 1Hz to 10 Hz. The best FE response is observed at 10 Hz with a well saturated and square shaped loop with Pmax = 127 C/cm2, Pr = + 121 C/cm2 and -120 C/cm2, and nominal voltages required for switching, Vc = + 4.72 V and – 4.53 V, respectively.

PAGE 155

130-10-8-6-4-20246810 -250 -200 -150 -100 -50 0 50 100 150 200 250 10 Hz 5 Hz 2 Hz 1.25 Hz 1 HzPolarization (C/cm2) V olta g e ( V ) Figure 3.3.12. Ferroelectric hysteresis loops for CFO-PZT bilayer film on MgO substrate at 9V driving voltage in the frequency range 1 Hz to 10 Hz. The hysteresis loops at lower frequencies in the range 1 Hz to 2 Hz loops are tilted with Pmax < Pr. This is due to the presence the intermediate CFO dielectric layer along with the ferroelectric PZT layer. Such behavior was not observed for the PZT-MgO films (see Section 3.2). The intermediate CFO layer has a lower dielectric constant than PZT and separates the bound charges that are due to polarization from compensating the charges on the bottom LSMO electrode. Due to the incomplete compensation a depolarizing field (Ed) develops across the PZT la yer during measurement. The depolarizing field is given by Ed = Pd/( dt) where d and d are the thickness and the complex permittivity of the dielectric layer, and P and t are the polarization and thickness of the ferroelectric layer [140]. The complex permittivity ( d) is dependent on the frequency of the applied field ( ) by the relation d = + i ( / ) where and are the

PAGE 156

131 conductivity and real part of permittivity of the medium, respectively. At low frequencies the contribution from the complex part of d is enhanced and effects the Ed. This is manifested as the tilting of hyste resis loops at lo w frequencies. Figure 3.3.13 shows the frequency dependence of Pr and Vc at 9 V driving voltage. The similar values of + Pr and – Pr are associated with the symmetric electrode configuration. Further, in th e range 1 Hz to 1 kHz, the Pr values decrease monotonically indicating typical ferr oelectric behavior. 100101102103-250 -200 -150 -100 -50 0 50 100 150 200 250 100101102103 -10 -8 -6 -4 -2 0 2 4 6 8 10 + Pr Pr Vc (V) Pr (C/cm2) Vc + VcFrequenc y f ( Hz ) Figure 3.3.13. Frequency dependence of remnant polarization (Pr) and nominal voltage for switching (Vc) for CFO-PZT bilayer on MgO substrate. The Vc values increase almost linearly in th e frequency range 1 Hz to 100 Hz with a slight drop at 1 kHz. Such linear beha vior has recently been reported for PZT films grown on different substrates [141]. The linear dependence of coercive field (Ec) with switching field frequency ( ) is associated with the vi scous motion of domain boundaries

PAGE 157

132 given by Ec = Vc/d = (Po h/ ) where Po is the spontaneous polarization, is the viscosity of domain wall motion, d and h/ 2 are the half widths of domain wall and crystallite, respectively. Figure 3.3.14 shows the ferroelectric hys teresis loops for CFO-PZT bilayer film on STO substrates measured at 9 V driving vo ltage but changing the frequency from 1 Hz to 10 Hz. Well saturated and square shaped hy steresis loops are observed at 1 Hz with Pmax = 185 C/cm2, Pr = + 182 C/cm2 and -180 C/cm2, and Vc = + 5.85 V and – 4.95 V, respectively. -10-8-6-4-20246810 -250 -200 -150 -100 -50 0 50 100 150 200 250 10 Hz 5 Hz 2 Hz 1.25 Hz 1 HzPolarization (C/cm2) V olta g e ( V ) Figure 3.3.14. Ferroelectric hysteresis loops fo r CFO-PZT bilayer film on STO substrate at 9V driving voltage in the frequency range 1 Hz to 10 Hz. Figure 3.3.15 shows the frequency dependence of Pr and Vc at 9 V driving voltage. The monotonically decreasing Pr values and linearly increasing Vc values with increasing frequency again indicate typi cal ferroelectric behavior.

PAGE 158

133 100101102103-250 -200 -150 -100 -50 0 50 100 150 200 250 100101102103 -10 -8 -6 -4 -2 0 2 4 6 8 10 + Pr Pr Vc (V) Pr (C/cm2) + Vc VcFrequenc y f ( Hz ) Figure 3.3.15. Frequency dependence of remnant polarization (Pr) and nominal voltage for switching (Vc) for CFO-PZT bilayer on MgO substrate. Figure 3.3.16 shows the P-E loops for th e CFO-PZT bilayer (BL) and PZT single layer (SL) films grown on LSMO/MgO substr ates. Similarly, Figure 3.3.17 shows the PE loops for the CFO-PZT bilayer (BL) and PZT single layer (SL) films grown on STO substrates. All the presented loops were m easured using triangular pulses at 1000 ms hysteresis period (i.e. 1 Hz fr equency) at 9 V driving voltage. The thickness (d) of the PZT layer was kept constant at 500 nm both in the BL and SL films.

PAGE 159

134-200-160-120-80-4004080120160200 -200 -150 -100 -50 0 50 100 150 200 CFO-PZT bilayer PZT single layerPolarization (C/cm2)Electric Field ( kV/cm ) Figure 3.3.16. Ferroelectric hysteresis loops for CFO-PZT bilayer and PZT single layer films grown on MgO substrates under same conditions. Data measured at 9V driving voltage. The polarization hysteresis values for the bilayer (BL) and single layer (SL) on different substrates have been summari zed in Table 3.10. The coercive field (Ec) has been calculated using Ec = |(Vc(+)+Vc(-))|/2d. The leakage current density (JL) has been measured at 9 V for 1000 ms soak time. It is clear from Figures 3.3.16 and 3.3.17 that there is a huge enhancement in the polarizat ion values for the BLs as compared to the SLs.

PAGE 160

135-200-160-120-80-4004080120160200 -200 -150 -100 -50 0 50 100 150 200 CFO-PZT bilayer PZT single layer Electric Field ( kV/cm ) Polarization (C/cm2) Figure 3.3.17. Ferroelectric hysteresis loops for CFO-PZT bilayer and PZT single layer films grown on STO substrates under same conditions. Data measur ed at 9V driving voltage. Table 3.10. Summary of maximum polarization (Pmax), remnant polarization (Pr), nominal switching voltage (Vc), coercive field (Ec) and leakage current density JL (A/cm2) for CFO-PZT bilayer and PZT single layer film s grown on MgO and STO substrates. Data measured at 9 V drivi ng voltage at 1 Hz. Sample SubstratePmax P r Vc EcJL ( C/cm2) ( C/cm2) (V)(kV/cm) (A/cm2) (+) (-) (+)(-) x 10-9CFO-PZTMgO127121-1204.72-4.5392.51.99 PZTMgO9375-722.64-1.9946.30.27 CFO-PZTSTO185182-1805.85-4.95108.08.61 PZTSTO10891-922.62-1.6943.10.20

PAGE 161

136 The Pmax values for CFO-PZT BL on MgO and STO increased by almost 37 % and 70 % respectively as compared to the SL PZT films. The Ec values also increased in the BLs which is attributed to the larger depol arizing fields in the BLs due to the presence of the intermediate CFO layer. Figure 3.3.18 shows a comparison of the leakage current densities (JL) for CFOPZT BL and PZT SL capacitors on STO subs trates measured under constant voltage stress of 9V applied for a soak time of 1000 ms (Table 3.10). Although the BL films exhibited higher polarization values compared to the SL films, the leakage current densities in the BLs were higher than in the SLs. 02004006008001000 10-1010-910-810-7 STO (100) substrate LSMO (100) (100 nm) CFO (100) (200 nm) PZT (100) (500 nm) V Stress Voltage = 9 V CFOPZT bilayer PZT single layerLeakage Current density (A/cm2)Stress Time (ms) Figure 3.3.18. Leakage current density in CFO-PZT bilayer and PZT single layer thin film capacitors grown on STO substrates under same conditions. Inset shows a schematic diagram of CFO-PZT bilayer during the leakage current density measurement.

PAGE 162

137 One possible explanation for the observe d enhancement of polarization in BLs could be due to the difference in stress in the PZT layer grown on CFO in BL and on LSMO on SL. The lattice mismatches of th e substrates and electrodes with PZT are summarized in Table 3.11. Due to the sma ller lattice mismatch between PZT and CFO (0.04 %) as compared to PZT and LSMO (4.1 %), the PZT layer in CFO-PZT BL film is less strained which lead to higher polarization values. Table 3.11. Summary of crystal structure of substrates and el ectrodes used for growth of bilayers and single layers and lattice mismatches with respect to PZT. MaterialCrystal structureLattice parameters ()Lattice mismatch (%) (a, b, c)with respect to PZT PbZr0.52Ti0.42O3 (PZT) tetragonal perovskitea=b=4.036,c=4.146La0.7Sr0.3MnO3 (LSMO)pseudocubic perovskitea=b=c=3.876-4.1 CoFe2O4 (CFO)face centered cubica=b=c=8.3910.04 MgOcubic rock salta=b=c=4.2164.3 SrTiO3 (STO)cubic perovskitea=b=c=3.905-3.4 From the above ferroelectric measurements there is a clear evidence of ferroelastic coupling in these CFO-PZT BL films. The cha nge in polarization in the CFO-PZT BLs with respect to the SL PZT film could be attributed to the interfacial stress between the layers. The change in the ferroelectric property due to the introduction of a CFO layer shows a possibility of magnetoelectric (ME) coupling in CFO-PZT bilayer. 3.3.3. Conclusions To summarize, the epitaxial bilayer film s of CFO and PZT we re grown on single crystalline MgO and STO substrates. The f ilms grown on MgO (100) showed better inplane epitaxy compared to the films grown on STO (100). In-plane strain on the film grown on MgO increased with the PZT layer on top of the CFO layer. This increase in

PAGE 163

138 stress in the bilayer resulted in lowered sa turation magnetization compared to the single layer CFO-MgO film. In the case of the film grown on STO (100), the strain was already high even in the single-layer CFO film. With the bilayer CFO-PZT film on STO, the inplane strain reduced slightly, which in turn resulted in higher sa turation magnetization. The strain-modulated magnetism in these bi layers leads to the possibility of magnetoelectric coupling sought in multiferroics. The ferroelectric properties of the bilayers showed huge enhancements in the pol arization as compared to the single layer. This could be attributed to stress mediated ferroelastic coupling which could again lead to a possible magnetoelectric (ME) coupling in these structures.

PAGE 164

1393.4. Chapter Summary In conclusion, this chapter describe d the ferromagnetic and ferroelectric properties of epitaxial CFO and PZT thin film s, respectively. In case of CFO films it was observed that the epitaxial st rain played an important ro le in governing the magnetic anisotropies. The growth of high quality P ZT films required a thr ough investigation of the problem of Pb depletion. Using the dual la ser process, finally enhanced ferroelectric properties were observed. After the successful optimization of the growth parameters of the single layers, the horizontal composite CFOPZT bilayered structures were fabricated. Simultaneous observation of ferromagnetis m and ferroelectric ity confirmed the multiferroic nature of the structure. X-analysis of interfacial stress and their correlation to the observed properties showed a clear evid ence of magneto-elas tic coupling in the heterostructures. This indicated a possibility of ME effect in these multiferroic structures. However in order to prove ME effect the magnetic properties have to be studied under an applied electric field and vice versa.

PAGE 165

140 CHAPTER 4: DOPED ZINC OXIDE HETEROSTRUCTURES This chapter describes the growth and pr operties of doped Zinc oxide (ZnO) thin films. ZnO is a direct wide band gap (3.4 eV ) semiconductor that has sustained research interest for quite some time because of its multifunctionality like room temperature (RT) ferromagnetism (FM) [142], and piezoelectricit y [143]. ZnO has foun d applications in optoelectronics [144], gas sensing [ 145, 146], photocatalysis [147 149], and UVphotoelectronics [150 152]. It has also b een used as a transparent conducting oxide [153], and as electrodes for photoelectrochemical cells [ 154 156]. The most important feature for this wide scale research in ZnO is the simplicity of its growth in different forms using diverse physical and chemical methods. PLD technique has been very successful for growing ZnO thin films with defined dopant concentrations and complex morphologies for various applications [157]. ZnO thin films have preferred grow with a c-axis orientation irrespective of the type of substrate used [158]. As long as the same hexagonal symmetry is present in the surface-un it cell of the substrate compared to the cplane of ZnO, i.e. the (0001) crystallographi c plane, the epitaxial relationship between substrate and the film is maintained. This allows for the growth of epitaxial ZnO films on c-cut sapphire (Al2O3) substrates although the lattice mismatch between sapphire and ZnO is as large as 26 %. This large mi sfit hinders a pseudomorphic growth and consequently results in defects and film im perfections that aff ect their properties.

PAGE 166

141 In recent times, it has been reported th at the ferromagnetic properties of ZnO can be enhanced by 3d transition metal doping pr oducing RT dilute ma gnetic semiconductors (DMS) [159]. Among the transition metals that have been doped, Manganese (Mn) has attracted more attention due its complete so lid solubility in ZnO and the ability to introduce a strong magnetic moment when doped in ZnO [160]. On the other hand, it was also reported that the piezoelectric propert ies of ZnO can be greatly enhanced by doping it with vanadium (V) [161, 162]. From th e viewpoint of material constituents, multiferroic materials in single phase have always attracted more attention than composites due to the direct ME coupling between the two order parameters and less interfacial defects [14]. In this viewpoint, the single phase structure and the observation of simultaneous RT ferromagnetic and ferroel ectric properties make the Mn and V doped ZnO thin films an attractive candidate for a potential multiferroic device. In this work, the doped ZnO heterostructures have been grown fo r the first time to test the magnetoelastic coupling. The following sections describe the grow th and ferromagnetic and ferroelectric characterization of Mn doped and V doped ZnO thin films respectively. The final section describes the fabrication and properties of the Mn and V doped ZnO heterostructure. Evidence of possible magneto-electric c oupling in these structures is shown.

PAGE 167

142 4.1. Mn doped ZnO Thin Films Almost a decade ago Dietl et al. [163] ha d first made the theoretical predictions about RT FM in Mn doped ZnO (ZMO) and the potential applica tions as spintronic devices. This had started an extensive research effort in ZMO thin films. However diverse experimental results ranging from paramagnetism [144], spin glass [165], antiFM [166], low temperature FM [167] to even RT FM [168 170] have been reported over the years. Sharma et al. [168] who were th e first to report the RT FM had shown that Mn when doped nominally in ZnO was in Mn2+ valence state and the FM was carrier induced. In contrast Kundaliya et al. [169] la ter suggested that the FM was due to an oxygen-vacancy-stabilized Mn2-xZnxO3phase rather than carrier induced interaction among separate Mn atoms in ZnO matrix. Subse quently, Garcia et al. [170] showed that the RT FM in the ZMO was associ ated with the coexistence of Mn3+ and Mn4+ via a double-exchange mechanism. Such conflicti ng arguments produced no consensus on the origins of FM in ZMO systems. A fundamental understanding of the mechanism of magnetism in these materials is essential in order to achieve desired properties for use in device applications. Further experimental repor ts on the effects of growth conditions such as substrate temperature (Ts) and oxygen pressure (pO2) in controlling the magnetic properties of ZMO films grow n by pulsed laser deposition (P LD) suggested that higher saturation magnetization (Ms) was observed in the conducting films grown at a Ts of 500600 C under low pO2 [171 173]. These reports indica ted that the observed FM was very sensitive to sample pr eparation, crystalline quality, residual defects and resulting carrier concentrations. In 2005, Coey et al. [174] proposed a defect mediated bound magnetic polaron (BMP) model to explain the FM in ZMO. In recent years such defect

PAGE 168

143 mediated FM was experimentally observed [173, 175]. More recently, Calderon et al. [176 178] proposed a model based on two complementary magnetic mechanisms – the BMP percolation at low temperatures in insulating samples and the indirect RKKY exchange mechanism in more conducting sa mples at temperatures where substantial thermally activated carriers were present in the impurity band. However, there have been no direct observations that have validated this model. In th is work both experimental and theoretical investigations s uggested that the dual FM mo del [176] may finally resolve some of the controversies associated w ith the origins of FM in ZMO systems. 4.1.1. Experimental Details The ZMO target was fabricated by conve ntional solid state reaction technique. Stoichiometric amounts of high purity ZnO and MnO2 powders (both 99.99% pure) were mixed, calcined at 400 C for 12 h followed by uniaxial is ostatic cold pressing at 200 MPa and finally was sintered for 12h at 1000 C in air in order to make a hard ceramic target. Care was taken so that no impurities were introduced during the mixing and sintering process. EDS analysis on different regions of the target surface was performed to estimate the atomic % of Mn and also to check the uniform distribution of Mn in the target. The average atomic % Mn in the target was determined to be 2.0 0.4 %. ZMO films were grown on 10 x 5 mm2 c-cut sapphire (Al2O3) (001) substrates by varying the substrate temperatures (Ts) and background O2 pressures (pO2). Table 4.1 shows the deposition conditions for the sa mples. For ZMO films grown by varying Ts, the pO2 was kept constant at 10 mT. The best crystalline sample was achieved at 600 C. In order to optimize the deposition conditi ons ZMO films were also grown by varying pO2, with Ts at 600 C. Corresponding to each ZMO film, an undoped ZnO film was

PAGE 169

144 deposited under the same conditions for comparison. The film thicknesses were measured using a profilometer also shown in Table 4.1. Table 4.1. Deposition parameters for ZMO th in films. ZMO(600) and ZMO(10mT) are the same samples. SampleSubstrateLaser FluenceGrowth Temperatures O2 Pressure Film Thickness (J/cm2) Ts ( C)pO2 (mT) (nm) ZMO(RT) Al2O32RT1045316 ZMO(200) Al2O322001045425 ZMO(400) Al2O324001045813 ZMO(600) Al2O326001040360* ZMO(0mT) Al2O326000 (~10-3)121024 ZMO(10mT) Al2O326001040360* ZMO(50mT) Al2O3260050118242 ZMO(300mT) Al2O32600300117916 4.1.2. Results and Discussion 4.1.2.1. Structural Properties The prepared ZMO target was observed under SEM to reveal the microstructure. Figure 4.1.1 shows the ZMO targ et surface under different magnifications. The target constituted of hexagonal crystals of about 5 m size in average. Presence of hexagonal crystals suggests absence of secondary phases. XRD was performed to confirm the crystallinity of the ZMO targ et, and the undoped ZnO and MnO2 powders that were mixed to prepare the target. Figur es 4.1.2 (a, b, and c) show the XRD -2 scans of the ZMO target, MnO2 powder, and undoped ZnO powder, respectively. The XRD patterns for both the ZMO target (Figure 4.1.2. a) and the undoped ZnO powder (Figure 4.1.2. c) match with the hexagonal lattice structur e with space group P63mc (186) and lattice parameters a = 3.253 , and c = 5.213 , respectively.

PAGE 170

145 Figure 4.1.1. SEM images of the Mn doped ZnO target surface. 20304050607080 ZnO powder(201) (112) (103) (110) (102) (101) (002) (100)Intensity (arb. units)2 (deg)*Mn2O3 peaks* * *MnO2 powder(301) (311) (002) (211) (210) (111) (101) (110) MnO2 (c) (b) (a)Zn0.98Mn0.02O target(202) Figure 4.1.2. XRD patterns of (a) Mn doped ZnO target (b) MnO2 powder and (c) undoped ZnO powder, respectively. In order to rule out the presence of ot her impurity phases of Mn in the ZMO thin films a careful XRD analysis wa s conducted. Figure 4.1.3 shows the -2 XRD scans of

PAGE 171

146 ZMO films deposited at increa sing growth temperatures (Ts) of RT, 200 C, 400 C and 600 C, labeled as ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) respectively. 20406080100 102103104105 103104105 103104105 103104105 103104105 Undoped ZnO2 (deg) ZMO(600) ZMO(RT) ZMO(400) ZMO(200)Intensity (counts) (004) (002)* Figure 4.1.3. XRD patterns of (Zn0.98Mn0.02)O films on sapphire subs trates grown at room temperature, 200 C, 400 C and 600 C with a background oxygen pressure of 10 mT named as ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) respectively. The sapphire substrate peaks have been denoted by *. The films are highly textured with no obs erved peaks (within the resolution limits of XRD) from secondary phase formation of oxides of Mn or ZnMnO alloys which could lead to anti-ferromagnetic cluster formation. Th e log-scale for intensit y has been used to

PAGE 172

147 magnify the low intensity. Table 4.2 shows the observed peak shifts to lower angles in the doped samples which can possibly indicate Mn incorporation into Zn lattice sites. However since the ionic radii for Mn2+ (0.91 ) and Zn2+ (0.83 ) are fairly close to each other, the peak shift due to Mn substitution in ZnO lattic e is extremely small for low % of doping. It can be calculated that for 2 atom ic % Mn doping into ZnO there is ~ 0.01 increase in lattice parameter if Mn2+ substitutes the Zn sites. The increase in the peak intensities in Figure 4.1.3 for ZMO films of similar thicknesses suggest that the crystallinity improved with higher Ts. Table 4.2. FWHM of rocking curve about (002) plane of ZnO, FWHM of (002) peak of ZnO from -2 scan and average crystallite size fr om Scherrer formula for ZMO films at various growth temperatures. SampleGrowth temperatures2 ( )FWHM ( ) of FWHM ( ) of Crystallite size Ts ( C)Rocking curve -2 curve D (nm) ZMO(RT)RT 34.35n.a0.28041 ZMO(200)200 C34.552.3070.21754 ZMO(400)400 C34.501.3990.21454 ZMO(600)600 C34.500.4170.21156 UndopedZnO600 C34.550.4040.23250 To confirm the improved crys tallinity, rocking curves were performed about the ZnO (002) peak for the samples. Due to th e poor crystallinity of ZMO(RT), no rocking curve spectrum was obtained. Figure 4.1.4 shows the narrowing of the rocking curves for samples deposited at higher Ts. From Table 4.2 a slight increase (by ~0.01 ) in FWHM from rocking curves for ZMO(600) film is observed as compared to the undoped ZnO film grown under same conditions. This is pres umably due to the strain induced from the occupation of Mn ions at Zn ion sites.

PAGE 173

148 The average crystallite size in the films assuming spherical grains ( D =4/3 L ) were calculated using the Scherrer formula, L = 0.94 / B cos where L is the coherence length, B is the FWHM of (002) peak of ZnO in this case, = 1.54439 wavelength of CuK Xrays used and is the angle of diffraction. The aver age grain size increased with higher Ts and better crystall inity (Table 4.2). -3-2-10123 0.0 0.2 0.4 0.6 0.8 1.0 ZMO(600) ZMO(400) ZMO(200)Intensity (arb. units) ( de g) Figure 4.1.4. Rocking curves about the (002) plane of ZnO for ZnO:Mn films grown at 200 C, 400 C and 600 C with a pO2 of 10 mT named as ZMO(200), ZMO(400), and ZMO(600), respectively. The surface microstructures of the ZMO films were studied using SEM. Figure 4.1.5 shows SEM images of (a) undoped ZnO film grown at 600 C, (b) ZMO(600), (c) ZMO(400), and (d) ZMO(200), respectively. The improved crystallinity and grain agglomeration are observed with higher growth temperatures.

PAGE 174

149 Figure 4.1.5 SEM images of (a) undoped ZnO film grown at 600 C, (b) ZMO(600), (c) ZMO(400) and (d) ZMO(200), Mn doped Zn O films on c-cut sapphire substrates. AFM was used to analyze the surface mo rphologies of the films. Figures 4.1.6 (a, and b) show AFM images of the ZMO(600) and ZMO(RT) surfaces, respectively. Clear evidence of grain agglomera tion at higher temperatures is observed. The average grain size calculated for ZMO(600) is 228 nm whereas that for ZMO (RT) is 54 was nm. Figures 4.1.6 (c, and d) show 3D projecti ons of (a) and (b) re spectively. ZMO(600) exhibits a rougher surf ace with the root m ean square roughness (Rrms ) value of 14.69 nm while that for ZMO(RT) is 4.01 nm. The surface features on ZMO(RT) (z-height 10 nm) are much smaller than those for ZMO(600) (z-height 100 nm). The AFM analysis again confirmed the improved crystallinity and surface morphology with higher Ts. Similar structural features have been repo rted earlier for ZMO films [167 –171].

PAGE 175

150 Figure 4.1.6. AFM images of Mn doped ZnO thin films grown at (a) 600 C and (b) room temperature under pO2 of 10mT. Scan areas are 1 m x 1 m. (c) and (d) are 3D projections for (a) and (b), respectively. The crystallinity of ZMO films deposited at varying background oxygen pressures was analyzed. Figure 4.1.7 shows the XRD pa tterns for the ZMO films deposited at 600 C but varying the pO2 from 0 mT to 300 mT. The XRD pa tterns indicate that the films are single phase and c-axis preferred oriented with no impurity Mn-oxide phases. However since the film thicknesses were larger (> 1 m) than the critical thickness

PAGE 176

151 required for maintaining the epitaxial relatio nship, at higher depos ition pressures (100 mT, 300 mT), the films show polycrystalline nature. 304050607080 2 Theta (degrees)(0110)0 mT(0004) Al2O3 (0006) (0002)50 mT(0103)100 mTIntensity (arb. units) 300 mT Figure 4.1.7. XRD patterns for Mn doped ZnO thin films on c-cut sapphire substrates deposited at 600 C with varying oxygen background pressure. Surface microstructure of epitaxial ZnO films grown under different pO2 were compared with polycrystalline ZnO films depos ited on Si (100) substrates under the same conditions. Figure 4.1.8 shows SEM images of undoped ZnO films deposited on sapphire (a, b) and Si (100) (c, d) substrates. The left panel images (a, c) represent films grown at low pO2 of 10 mT while the right panel images (b, d) represent thos e at 300 mT. It is clear that the grain size increases with higher pO2 on both substrates and the hexagonal faceting is lost due to large grain agglomera tion. However the epitaxial ZnO films (a, b)

PAGE 177

152 have more compact and flattened surface as co mpared to the polycrystalline films (c, d) which appear to be porous. The growth morphologies of ZnO films on different substrates have been studie d in details elsewhere [179]. Figure 4.1.8. SEM images of undoped ZnO f ilms grown on (a, b) c-cut sapphire substrates and (c, d) Si (100) substrates. Inset to (c) shows the details on one of the hexagonal facets of ZnO on the film. 4.1.2.2. Electrical Properties The electrical properties of ZMO films were investigated using st andard Van der Pauw configuration [82, 83]. Table 4.3 lists the resistivity ( ), carrier concentration (nC), and Hall mobility ( H) measured at room temperature for ZMO films deposited at various Ts, and constant pO2 of 10 mT. pO2= 10 mT pO2= 10 mT pO2= 300 mT pO2= 300 mT

PAGE 178

153 Table 4.3. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:Mn thin films on c-cut sapphire substrates grown at various growth temperatures at cons tant background oxygen pressure. SampleGrowth temperatures ( cm)nC (cm-3) H (cm2/Vs) ZMO(RT)RT 1.36 x 104 7.78 x 1012 58.81 ZMO(200)200 C9.692.89 x 10172.23 ZMO(400)400 C8.333.88 x 10171.93 ZMO(600)600 C0.462.51 x 10185.45 Undoped ZnO600 C0.021.04 x 1018330.19 All the measured films show n-type conduction. The films become more conducting with increasing growth temperatures Conductivity in ZnO films is associated with donor-type defect sites such as intrinsic oxygen vacancies or Zn interstitials that are created during the deposition process [180]. These donor-type defects give the n-type carriers in the films. With higher Ts and low pO2 (10 mT), less O2 is incorporated into the films giving rise to more O2 vacancies and more free carriers and higher conductivity [180]. Generally in doped samples there is an overall increase of because the charged dopants (Mn2+) act as scattering sites for electrons [181]. This explains the orders of magnitude increase in for ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) as compared to undoped ZnO thin film ( ~ 10-2 cm). ZMO(RT) is highly insulating with very low density of free carriers (Table 4.3). This is probably associated with the poor crystalline nature of the film which make s the system strongly disordered. Table 4.4 shows the resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature for ZMO and undoped ZnO films deposited at 600 C, and varying pO2 from 0 mT to 50 mT. The films show ntype conduction decrea ses conductivity with higher background O2 pressure. Increasing the background O2 pressure during the

PAGE 179

154 deposition facilitates th e incorporating of O2 into the films and hence decreasing the number of O2 vacancies. Thus the free carrier dens ity decreases making the films more insulating. Table 4.4. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:Mn thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure. SampleO2 pressure (mT) ( cm)nC (cm-3) H (cm2/Vs) ZMO(0mT)00.32 1.92 x 1018 10.06 ZMO(10mT)100.462.51 x 10185.49 ZMO(50mT)5084.86.40 x 10161.15 Undoped ZnO00.019.56 x 1017653.76 Undoped ZnO100.02 1.04 x 1018330.19 Undoped ZnO5017.501.04 x 101634.27 Figure 4.1.9 shows the variation of and nc as a function of temperature for ZMO(600). An increase in nc with higher temperatures is observed. 50100150200250300 0 10 20 30 40 50 60 70 0.0 0.5 1.0 1.5 2.0 2.5 3.0 ZMO(600) Resistivity (-cm)Temperature (K) Carrier conc. nc(x1018 cm-3) Figure 4.1.9. Resistivity ( ) and carrier concentration (nc) versus temperature measured for ZMO(600).

PAGE 180

155 4.1.2.3. Magnetic Properties The magnetic properties of bulk ZMO powder were measured. Figure 4.1.10 shows the magnetization versus magnetic file d (M-H) hysteresis loop measured at 10 K for bulk ZMO powder. The straight line curv e indicates a diamagnetic nature. This confirms that the observed FM in ZMO systems is manifested only in thin film form [182]. During the magnetic measurements, the samples were handled very carefully to avoid any trace magnetic contamination. They were always handled with gloves using non-magnetic tweezers and kept in closed dry c ontainers. The substrates were cleaned by standard acetone-methanol ultrasonication. The samples were loaded in non-magnetic capsules using non-magnetic tape in the PPM S probe. The substrates were checked for magnetic contamination before deposition. The background di amagnetic contribution due to the sapphire substrates has been corrected consistently in all the presented curves. -10-50510 -0.010 -0.008 -0.006 -0.004 -0.002 0.000 0.002 0.004 0.006 0.008 0.010 10 K Zn0.98Mn0.02O bulk powderM (emu/g)Field (kOe) Figure 4.1.10. Magnetization measurement at 10 K for ZMO powder that was prepared by grinding a piece cut from the corresponding target. To investigate the role of Mn doping in enhancing the FM of ZMO films, M vs H loops were first measured for the film ZMO(600) and the undoped ZnO film grown under

PAGE 181

156 the same conditions (Ts = 600 C, pO2 = 10 mT). The saturation magnetization (Ms) for ZMO(600) is four times higher than undoped ZnO film, both showing RT FM as shown in Figure 4.1.11. FM in undoped ZnO films has b een associated with defects that are mostly located at the film surface and interf ace between the substrat e and the film [182]. However Mn doping into ZnO plays a very important role in improving the magnetic properties of these ZMO systems. -60-40-200204060 -4 -3 -2 -1 0 1 2 3 4 -60-40-200204060 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 (a)10 K ZnO Zn0.98Mn0.02OM (emu/cm3)H (kOe)(b) ZnO Zn0.98Mn0.02O 300 KM (emu/cm3)H (kOe) Figure 4.1.11. M-H loops measured at (a ) 10 K and (b) 300 K for undoped ZnO and ZMO films on c-cut sapphire substrates both grown under same conditions. The magnetic anisotropy in ZMO films was studied by measuring M versus H loops at 10 K and 300 K by applying the ma gnetic field both parall el (in-plane) and perpendicular (out-of-plane) to the film plane. However such measurements for ZMO(600) showed no noticea ble variation in the Ms and Hc at 10K as shown in Figure 4.1.12(a). Absence of anisotropic ferromagnetism suggests a weak contribution from magnetic point defects which was very large for Co doped ZnO thin films as reported by Venkatesan et al. [174, 183]. FM in ZMO systems at low temperatures (10 K) is intrinsic and not governed by defects. However, the different Ms values at 300 K (Figure 4.1.12.

PAGE 182

157 b) for the in-plane and out-of plane configurations indicates that defects play a role in governing the magnetic behavior at higher temperatures. -60-40-200204060 -6 -4 -2 0 2 4 6 -60-40-200204060 -2 -1 0 1 2 (a) ZMO(600) in-plane out-of-plane 300 KM (emu/cm3)H ( kOe ) (b) ZMO(600) in-plane out-of-plane 10 KM (emu/cm3)H ( kOe ) Figure 4.1.12. M-H loops measured at (a) 10 K and (b) 300 K for ZMO(600) when the magnetic field was applied parallel (in-plane) an d perpendicular (out-of-plane) to the film surface. Figure 4.1.13 show M-H loops measured in -plane at (a) 10 K and (b) 300 K for ZMO(RT) and ZMO(600), respecti vely. Table 4.5 lists the Ms and Hc values for all the samples. An increase in the 300 K, Ms from 1.09 0.02 to 1.67 0.02 emu/cm3 and Hc from ~120 Oe to ~250 Oe is observed as the Ts is increased from RT to 600 C. Table 4.5. Magnetic properties at 10 K a nd 300 K for ZMO(RT), ZMO(200), ZMO(400) and ZMO(600) films, respectively. Samples on10 K300 K c-cut Al2O3 Growth temperatures MsHcMsHcsubstrates Ts ( C) (emu/cm3) (Oe) (emu/cm3) (Oe) ZMO(RT)RT2.17 0.031101.09 0.02120 ZMO(200)2002.51 0.051201.12 0.01125 ZMO(400)4002.93 0.021202.02 0.01130 ZMO(600)6003.08 0.121801.67 0.02250 Undoped ZnO6000.94 0.021201.36 0.02305

PAGE 183

158 The increase in Hc from ZMO(RT) to ZMO(600) mo st likely is related to the increase in grain size with higher Ts shown earlier [167-171]. The Hc and Ms for films ZMO(200) and ZMO(400) also follow the same trend. -60-40-200204060 -2 -1 0 1 2 -4 -3 -2 -1 0 1 2 3 4 300 K(b)M (emu/cm3) M (emu/cm3)H (kOe) ZMO(RT) ZMO(600)10 K(a) ZMO(RT) ZMO(600) Figure 4.1.13. M-H loops measured at (a) 10 K and (b) 300 K for ZMO films deposited on c-cut sapphire substrates at RT and 600 C with a constant pO2 of 10 mT named as ZMO(RT) and ZMO(600) respectively. Figure 4.1.14 shows M-H loops measured in -plane at (a) 10 K and (b) 300 K for samples ZMO(0mT) ZMO(10mT) ZMO(50m T) and ZMO(300mT), respectively. The magnetization is reduced with increasing pO2. The magnetic properties have been summarized in Table 4.6.

PAGE 184

159 Table 4.6. Magnetic properties at 10 K and 300 K for ZMO(0mT), ZMO(10mT), ZMO(50mT) and ZMO(300mT) films, respectively. Samples onBackground10 K300 K c-cut Al2O3 O2 pressureMsHcMsHcsubstrates(mT) (emu/cm3) (Oe) (emu/cm3) (Oe) ZMO(0mT)02.75 0.048001.47 0.03203 ZMO(10mT)103.08 0.121801.67 0.02250 ZMO(50mT)502.45 0.0310100.47 0.02265 ZMO(300mT)3001.16 0.021050.33 0.03120 -60-40-200204060 -2 -1 0 1 2 -4 -3 -2 -1 0 1 2 3 4 ZMO(0mT) ZMO(10mT) ZMO(50mT) ZMO(300mT)(a)10 KM (emu/cm3)H (kOe)(b) ZMO(0mT) ZMO(10mT) ZMO(50mT) ZMO(300mT)300 KM (emu/cm3) Figure 4.1.14. M-H loops measured at (a) 10 K and (b) 300 K for ZMO films deposited on c-cut sapphire substrates at 600 C by varying the pO2 from 0 mT to 300 mT.

PAGE 185

160 4.1.3. Theoretical Modeling This section describes the “dual mechanis m of FM” in ZMO thin films. The dual model is a combination of percolation of bound magnetic polarons (BMPs) at lower temperature and Ruderman–Kittel–Kasuya–Yos ida (RKKY) FM at higher temperature. The analysis is based on the experimental data on ZMO films grown at different temperatures named as ZMO(RT), ZM O(200), ZMO(400) and ZMO(600). The most probable mechanism for the observed magnetis m at low temperatures (10 K) in the highly insulating sample ZMO(RT) is due to th e percolation of BMPs In contrast, films deposited at 200 C, 400 C and 600 C, which are more conducting, it is the indirect RKKY mechanism which dominates the magnetic behavior [176]. Since both mechanisms depend on therma lly generated carriers, the activation energies ( E) for ZMO(RT) and ZMO(600) were es timated by fitting the resistivity ( ) curves to the expression, = oexp( E/kBT). Figure 4.1.15 shows the ln( ) versus 1000/T curves for ZMO(RT) and ZMO(600). The E was calculated from the slopes of the linear fits of ln( ) versus 1000/T data points. The value of E for ZMO(RT) and ZMO(600) are ~6.4 meV and ~42.4 meV respectiv ely. The smaller value of E in ZMO(RT) indicates that it has a shallower donor level than in ZMO(600). Since Mn doping into ZnO does not create additional donors, the available dono rs are associated with defect sites. The larger density of defects associated with the poor crystalline quality of ZMO(RT) contributed to intermediate impurity bands which effectively reduced the E.

PAGE 186

1613.33.43.53.63.73.83.94.0 -1.8 -1.7 -1.6 -1.5 7.40 7.41 7.42 7.43 7.44 7.45 7.46 (b)y = 511.23 x + 2x10-13R2 = 0.9987E = 42.4 meVln ()1000/T (K -1 ) ZMO(600) Linear fit(a) E = 6.4 meVy = 74.44 x + 7.159 R2 = 0.9731 ZMO(RT) Linear fitln () Figure 4.1.15. Resistivity ( ) versus temperature dependence of (a) ZMO(RT) and (b) ZMO(600). 4.1.3.1. RKKY exchange interaction The RKKY potential energy is given by URKKY (r) = JRKKY (r) S1. S2 where JRKKY is the exchange integral and S1 and S2 are the interacting magnetic spins [45]. The RKKY exchange integral JRKKY has the form, JRKKY( r ) = (Vo cos2kf r )/ r3 where Vo is the strength of RKKY interaction, kf is the Fermi wave vector and r is the separation between the magnetic spins For FM, when URKKY < 0 i.e. energy minimum, JRKKY >0, since S1. S2 = 1 [45]. The range of the average separation (r ) between randomly oriented magnetic dopant spins that interact via FM RKKY mechanis m can be estimated by plotting the exchange

PAGE 187

162 integral JRKKY as a function of r for Vo = 1 and assuming the intrinsic Fermi energy Ef = 1.65 eV to be half the band gap for ZnO (Eg = 3.3 eV). This assu mption is valid because the carrier concentration (nc) does not much alter in ZMO(600) (nc = 2.51 x 1018) due to Mn doping as compared to undoped ZnO (nc = 1.04 x 1018) (Table 4.3). The Fermi wave vector (kf) was calculated from the relation Ef = (h2/ 8 2m) kf 2 which came out to be 6.57 x 109 (nm-1). Figure 4.1.16 shows the variation of JRKKY(r) as a function of r. 0.00.51.01.5 -100 -50 0 0.30.40.50.60.70.80.91.01.1 r (nm)JRKKY (a.u.)JRKKY (a.u.) r (nm)r = 0.96 nm r = 0.44 nm Figure 4.1.16. Plot of RKKY exchange integral JRKKY(r) as a function of the average separation between magnetic spins (r). The inset shows the range of r when JRKKY > 0. From the plot it is observed that JRKKY > 0 in the ranges r < 0.11 nm and 0.36 nm < r < 0.59 nm and 0.84 nm < r < 1.07 nm. For 2 atomic % doping of Mn in ZnO the average separation between Mn2+ spins was found to be r ~ 1.05 nm, which is based on the relation r = (1/ni)1/3 where ni = 8.438 x 1020 cm-3 (i.e. density of Mn2+ spins). Thus r

PAGE 188

163 falls within the range of FM RKKY interacti on. This implies that at such dilute doping levels RKKY interaction would give FM. F. W. Smith [184] in his work on the magnetization of dilute magnetic alloys showed that the measured magnetization in ZnMn or CuMn alloys is dominated by impurity-impurity interaction vi a RKKY potential. In a high magnetic field (H) such that g BH >> kBT and niVo the approach to saturation of the magnetization (M) follows the relation: M = g BSni [1 2(2S + 1)niVo/3g BH] (4.1) where ni is the concentration of the active magnetic impurities (Mn2+) interacting via RKKY, Vo is the strength of RKKY interaction and S = 5/2 is the spin of magnetic impurity [168]. However, in this case g BH ~ kBT but is >> niVo. Figure 4.1.17 shows the approach to saturation of M as H is increased from 0 to 50 kOe. The inset and also shown as the black solid line in Figure 4.1.17 show the free-spin magnetization described by paramagnetic moments given by the expression [45]: M = nig BSBS(x) with x = g BSH/kBT (4.2) where the magnetic impurity density ni = 8.438 x 1020 cm-3 for 2 atomic % Mn doping, BS(x) is the Brillouin function with spin S = 5/2, the Lander factor g = 2 as expected for Mn2+ B is the Bohr magneton, and kB is the Boltzmann constant. The experimental curves show ferromagnetic behavior by reaching saturation more rapidly as a function of applied field as compared to free-spins. This indicates that the Mn2+ spins are not free and interact via a carrier mediated mechanism.

PAGE 189

16401020304050 0 1 2 3 4 5 0501001500 20 40 ---nigBSBS(x)S = 5/2 (Mn2+)10 KM (emu/cm3)H (kOe)M (emu/cm3)H ( kOe ) ZMO(RT) ZMO(200) ZMO(400) ZMO(600) nigBSBS(x) Figure 4.1.17. Initial magnetizatio n curves at 10 K for (Zn0.98Mn0.02)O films deposited at various growth temperatures with constant background O2 pressure. The inset shows the free-spin Brillouin function BS (10K) for S=5/2 at 10 K. Figure 4.1.18 shows the linear f its of M versus 1/H data for 20 kOe < H < 50 kOe. From the slopes and intercepts of the linear fits of M versus 1/H data points at saturation measured at 10 K, for ZMO(RT), ZMO(200), ZMO(400) and ZMO(600), ni and Vo were calculated using Equation 4.1. The calculated va lues have been summarized in Table 4.7. From Figure 4.1.18 and the values in Table 4.7 it is observed that the intercepts and slopes increase slightly from ZMO(200) to ZMO(600). This shows that the number of magnetic impurities interac ting via RKKY mechanism in creases with higher Ts and better crystallinity.

PAGE 190

1650.0150.0200.0250.0300.0350.0400.04 5 1.5 2.0 2.5 3.0 3.5 10 K ZMO(RT) ZMO(200) ZMO(400) ZMO(600)M (emu/cm3)1/H ( kOe-1 ) Figure 4.1.18. M versus 1/H (kOe-1) plots at 10 K for (Zn0.98Mn0.02)O films deposited at various growth temperatures with constant pO2. (b) The solid lines are the linear fits for the data points for films named ZMO( RT), ZMO(200), ZMO(400) and ZMO(600). Table 4.7. Summary of RKKY parameters for ZMO films, deposited at various temperatures. SampleInterceptSlope niVo niVo (meV) (1019 cm-3) (10-37 ergcm3) ZMO(RT)2.3741103145.126.750.0216 ZMO(200)2.8173141666.086.580.0250 ZMO(400)3.1209147826.735.590.0235 ZMO(600)3.3896153997.314.940.0226 However the slope of the linear fit for ZMO(RT) is different than the others showing a smaller value for ni. The calculated value of ni for ZMO(600) is ni ~7.31 x 1019 cm-3. It is much lower than the magnetic impur ity density for 2 atomic % doping of Mn in

PAGE 191

166 ZnO which is 8.438 x 1020 cm-3. In other words only ~8 % of the total number of Mn2+ ions in the system are in volved in RKKY mediated FM. From Figure 4.1.13 (a) for ZMO(600) the Ms is ~3.1 emu/cm3 which corresponds to an average magnetic moment of 0.4 B/Mn2+ assuming a uniform Mn ion distribution. This is again 8 % of the theoretical value which is 5 B/Mn2+ when all the Mn spins are aligned [168]. Thus only about 8 % of the inco rporated Mn atoms at substitutional Zn sites contribute to FM via RKKY mechanism. The remaining Mn atoms possibly occupy interstitial defect sites or get accumulated at the grain boundaries. However, the contribution of interstitial Mn ion to FM vi a a defect mediated mechanisms such as BMP has been shown to be very small [174], a nd therefore does not account for the observed magnetism in our highly crystalline ZMO (600) films. Further, the carrier density (nc) at 10 K for ZMO(600) estimated using nc(T) = ncoexp( E/kBT) for the measured values of nc (300K) and E is nc ~ 2.01 x 1016 cm-3. Since nc << ni or in other words the mean distance between the ca rriers is greater than that betw een the spins, RKKY interaction would give FM as also predicted by Pri our et al. [185]. The average value of Vo (5.96 +/0.85) x 10-37 ergcm3 is consistent with those observed by Smith. Moreover, the average value of niVo 0.02 meV (< kBT), the spin-spin interaction energy for RKKY, is smaller than the thermal energy at 10 K (0.863 meV) ensuring the availability of enough carriers to mediate RKKY FM even at such low temperatures. The above analysis has been also extended to ZMO films grown at 600 C but varying the pO2. Figure 4.1.19 shows the magnetization curves for samples ZMO(0mT), ZMO(10mT), ZMO(50mT), and ZMO(300mT).

PAGE 192

16701020304050 0 1 2 3 10 K ZMO(0mT) ZMO(10mT) ZMO(50mT) ZMO(300mT)M (emu/cm3)H ( kOe ) Figure 4.1.19. Initial magnetizatio n curves at 10 K for (Zn0.98Mn0.02)O films deposited at 600 C but varying the background O2 pressure. Figure 4.1.20 shows the linear f its of M versus 1/H data for 20 kOe < H < 50 kOe. From the slopes and intercepts of the linear fits of M versus 1/H data points at saturation measured at 10 K, for ZMO(0mT), ZM O(10mT), ZMO(50mT) and ZMO(300mT), ni and Vo were calculated using Equation 4.1. The calcu lated values have been summarized in Table 4.8.

PAGE 193

1680.020.030.040.050.060.07 0.5 1.0 1.5 2.0 2.5 3.0 3. 5 10 K ZMO(RT) ZMO(200) ZMO(400) ZMO(600) 1/H ( kOe-1 ) M (emu/cm3) Figure 4.1.20. M versus 1/H (kOe-1) plots at 10 K for (Zn0.98Mn0.02)O films deposited at 600 C but varying the background O2 pressure. The solid lines are the linear fits for the data points for films named ZMO(0mT), ZMO(10mT), ZMO(50mT) and ZMO(300mT). Table 4.8. Summary of RKKY parameters for ZMO films under varying background pressures. SampleInterceptSlope niVo niVo (meV) (1019 cm-3) (10-37 ergcm3) ZMO(0mT)2.9276106486.314.580.0181 ZMO(10mT)2.3933111155.167.150.0231 ZMO(50mT)1.63356865.63.529.480.0209 ZMO(300mT)1.31817678.92.8410.60.0289 4.1.3.2. Percolation of Bound Magnetic Polarons (BMPs) BMPs are formed in ZMO thin films by the alignment of the spins of Mn2+ with that of localized electrons. The electrons ar e localized to some poi nt defects (i.e. the

PAGE 194

169 magnetic impurity Mn2+, and the oxygen vacancies created during deposition) by electrostatic interaction within an electron confinement radius ( aB). Within this radius of influence the electron interacts an ti-ferromagnetically with the Mn2+ forming a BMP. The size of the BMP increases as the temperatur e decreases eventually overlapping with neighboring BMPs. This overlap aligns the Mn2+ spins forming ferromagnetic clusters. The FM emerges when an “infinite cluster” (of the size of the system) is formed, i.e. when percolation of BMPs occur [176]. In order to explain the FM in ZM O(RT) by the BMP model, the electron confinement radius (aB) was calculated using aB = (mo/me*)a where is the static dielectric constant, me* = 0.28mo is the electron effective mass and a = 0.52 is the Bohr radius [176]. First, the value of was computed from the cap acitance measurements of the films using a Hewlett Packard 4192A Hi gh Frequency Impedance Analyzer at 1 MHz and 1 V rms oscillation voltage. The value of for ZMO(600) was 1.95 0.36 but for ZMO(RT) the value was higher than the impedance limit of the instrument. Secondly, for ZMO(RT) and ZMO(600) were computed using the Brus equation (Equation 4.3) [186, 187] where the change in band gap was equated to the activation energy ( E ) measured earlier and the crystal radius was calculated from Scherrer formula as in Table 4.2. The Brus equation is given by: R 4 1.8e m 1 m 1 8R h Eo 2 h e 2 2 (4.3) where E is the carrier activation energy which is same as the change (decrease) in the band gap of the material due to doping, R is the crystal radius, me = 0.28 mo and mh = 0.59 mo are the effective masses of the electron in the conduction band and the hole in the valence band (considered independe nt of Mn content) respectively, o is the permittivity

PAGE 195

170 of free space, e and h are the electronic ch arge and Planck’s constant, respectively, mo is the free electron mass and is the static dielectric constant of the material. The calculated values for ZMO(RT) and ZMO(600) are 11.3 and 2.06 respectively. The electron confinement radii for ZMO(RT) and ZMO(600) are ~ 20 and ~ 4 respectively. The radius value for ZMO(RT) is larger owing to the shallower defect level ( E ~ 6.4 meV) [176]. The BMP model is valid only in th e low carrier density regime where ncaB 3 << 1 and when ni >> nc This is true for both ZMO(RT) and ZMO(600). Following Calderon et al. [176] the temperature depend ence of the radius of the polaron (RP) was plotted using the following equation RP(T) (aB/2) ln (sS|J| (a0/aB)3/kBT) (4.4) where s = the spin of carriers (electrons ), S = 5/2 the spin of magnetic dopant (Mn2+), using J 1 eV the local exchange coupling betw een carrier spin and the magnetic Mn moments [176] and a0 3= 47.77 3 the unit cell volume for ZnO. Figure 4.1.21 show the curves for the ZMO(RT) and ZMO(600). For ZMO(RT), polarons only start forming at a temperature of Tinitiation ~ sS|J| (a0/aB)3/kB ~ 75 K above which there are no polarons in the system. However in ZMO(600) po larons are always present even at room temperature. The size of the BMPs in ZMO(RT) increases drastically from ~0.7 at 70 K to ~20 at 10 K whereas for ZMO(600) the increase in size is gradual from ~7 at 300 K to ~13 at 10 K. The magnetic dopant spins at a distance r < RP tend to align with the localized carrier spin. The average separation between randomly oriented magnetic dopant spins is found to be ~ 10 , which is based on the relation r = (1/ni)1/3 where ni = 8.438 x 1020 cm-3 for 2 atomic % doping of Mn in ZnO. With lowering of temperature, polar ons start overlapping with neighboring BMPs

PAGE 196

171 forming FM clusters which keep growing in size to form an infinite cluster (of the size of the system) i.e. when the BMP percolation occurs. 050100150200250300 -15 -10 -5 0 5 10 15 20 25 No Polarons Tinitiation= 75KPolaron overlap Polaron formationRP(T) ()Temperature (K) ZMO(RT) ZMO(600) Figure 4.1.21. Variation of the radius of polarons Rp in angstroms at different temperatures for ZMO(RT) and ZMO(600), (Zn0.98Mn0.02)O films deposited at RT and 600 C respectively. The percolation radius (rperc) for ZMO(RT) at 10 K is ~ 5 m calculated from rperc 0.86/nc 1/3 [18] where nc (10K) was estimated using nc(T) = nco exp( E/kBT) from nc(RT) and E values measured earlier. The rperc in ZMO(RT) is larger than the thickness of the film and implied long range FM at 10 K. The rperc for ZMO(600) at 10 K is 0.03 m which is much smaller than the film thickness to mediate large FM via BMP percolation throughout the sample. From Figure 4.1.21 it was observed that for ZMO(600), Rp r, in the range 10 K – 300 K. This implies that the BMPs are always

PAGE 197

172 overlapping with their spins aligned with the carriers. If there are substantial thermally activated carriers, the BMP scen ario extrapolates to RKKY FM. 4.1.4. Conclusions To summarize, Mn doped ZnO films have been grown using PLD. The ferromagnetic properties have been investigated. The observed FM in Zn0.98Mn0.02O films shows characteristics of both intrinsic and carrier mediated mechanisms. The experimental results show a strong correlation between effec tive carrier densities due to different growth conditions and the FM in the samples. The da ta is consistent with the dual ferromagnetic theoretical model propos ed for dilute magnetic semiconductors (DMS). Although these results are encouragin g for a fundamental understanding of the origins of the observed ferromagnetic phases in DMS, higher values of Ms at RT would be required for any potentia l device application.

PAGE 198

173 4.2. V doped ZnO Thin Films ZnO has the strongest piezoelectric response among the tetrahedrally bonded semiconductors [143] which makes it a suitable material for technol ogical applications that require strong electromechanical coupl ing such as sensors and actuators [188]. Recently, Yang et al. [161, 162] has reporte d an electromechanical coefficient ( d33) in 2.5 atomic % V-doped ZnO to be as high as 11 0 pC/N. This value is an order higher compared to the d33 coefficient of bulk ZnO [189] whic h is 9.9 pC/N. Another significant property due to V doping is reported to be the switchable Ps indicated by a butterfly like displacement (D) versus electric field (E) graph [161]. Yang et. al. proposed a microscopic explanation for the enhancement of piezoresponse in ZVO systems. The dominant effect of applying el ectric field on the wurtzite structure of ZnO is the rotation of the bonds that are non-collinear with the polar c-axis, (i.e., Zn2-O1 bonds as shown schematically in Figure 4.2.1) towards the di rection of applied fi eld and thus producing strain. The V ions which repl ace the Zn sites are in 5+ valence state [162]. When V5+ substitutes Zn2+ sites, due to their higher positive charge, they make the V-O1 (Figure 4.2.1) bonds more polar than Zn2-O1 bonds. Hence V-O1 bonds rotate more easily in applied field enhancing the piezo response. In short, doping Zn2+ sites by V5+ creates a mixed valency as well as strain in the or iginal ZnO hexagonal stru cture because of the reduced ionic size of V5+ and higher positive charge. The mixed valency creates charge polarity between Zn-O and V-O bonds. The charge polarity and rotati on of the nonlinear V-O bonds with respect to the Zn-O bonds under electri c field gives enhanced ferroelectricity [161].

PAGE 199

174 Figure 4.2.1. Schematic diagram of the unit cell and neighboring atoms in V doped ZnO crystal structure. Ferroelectricity (FE) is exhibited in insu lating films. As described earlier in Section 4.1, ZnO films grow with intrinsic oxygen deficiencies when deposited under low background O2 pressure (pO2). However the films grown under high pO2 (> 100 mT) are more insulating and ideal for exhibiting Ps. Thus the effect of pO2 in the growth and properties of ZVO f ilms were studied. 4.2.1. Experimental Details The ZVO target was prepared by standa rd solid state re action method. Highly pure (99.99%) powders of ZnO and V2O5 were mixed in stoichiometric proportion and well ground before being calcined in air for 6 hrs at 600C. The calcined mixture was ground again and cold pressed und er a pressure of 90 MPa into a 1 inch diameter and 0.25 inch thick target. The presse d target was sintered at 1000 C in air for 12 h EDS analysis on different regions of the target su rface confirmed the unifo rm distribution of V in the target. The average atomic % of V in the target was 2.0 0.6 %. c axis V O1O1 O1Zn2O2O2Zn2 Zn1 a axis b axis

PAGE 200

175 A series of ZVO films were grown on c-cut sapphire (Al2O3) (001) substrates at 600C by varying the background O2 pressure from the Zn0.98V0.02O ceramic target. Table 4.9 shows the deposition parame ters used in this study. Table 4.9. Deposition parameters for ZVO thin films. SampleSubstrateLaser FluenceGrowth Temperatures O2 Pressure Film Thickness (J/cm2) Ts ( C)pO2 (mT) (nm) ZVO(100mT) Al2O32600100103310 ZVO(300mT) Al2O3260030097125 ZVO(500mT) Al2O3260050095324 ZVO(500mT)LSMO/Si2600500130020 4.2.2. Results and Discussions 4.2.2.1. Laser-target interactio ns and Plume Diagnostics The ZVO target surface was irradiated by 1000 KrF laser pulses for a range of laser fluences from 1 J/cm2 to 5 J/cm2 under 500 mT pO2. After ablati on, the surface morphologies and the compositions of the lasertarget interaction sites were examined by SEM and EDS. Figures 4.2.2 (a, b, c, and d) show SEM images of the ZVO target surface before ablation (KrF 0 J/cm2), after ablation at 2 J/cm2 (KrF 2 J/cm2), after ablation at 3 J/cm2 (KrF 3 J/cm2), and after ablation at 5 J/cm2 (KrF 5 J/cm2), respectively. From Figure 4.2.2 (b) it is observed that the ablated target surface at 2 J/cm2 has melted exposing some voids and cracks. On the other hand, from Figure 4.2.2 (c) shows that at 5 J/cm2 the ablated target surface e xhibits distinct large circular cavities possibly formed by the release of gas (vapor) from the target surface due to subsurface boiling. Consequently, films deposited at 5 J/cm2 showed increased particulate density on the surface.

PAGE 201

176 Figure 4.2.2. SEM images of (a) unablated ZVO target surface (b) ablated surface at 2 J/cm2 and (c) ablated surface at 5 J/cm2. Figure 4.2.3 shows the variation of atomic % of Zn, V and O in the ablated target for different laser fluences, obtained from EDS analysis. It is evident that the stoichiometry remains almost constant for al l the different fluences, that implies that congruent evaporation of the elements occur above the ablation threshold. (c) (a) (b) (d) KrF 0 J/cm2 KrF 2 J/cm2 KrF 3 J/cm2 KrF 5 J/cm2

PAGE 202

177012345 0.0 0.5 1.0 1.5 2.0 45.0 47.5 50.0 52.5 55.0 Oxygen Zinc VanadiumAtomic % Laser Fluence ( J/cm2 ) Figure 4.2.3. Atomic % obtained from EDS an alysis for the ZVO target under various laser fluences. On x-axis 0 J/cm2 implies the unablated target. Figure 4.2.4 shows the ICCD images of th e total visible light emitted from the plasma plumes for different laser fluences (F) under pO2 of 0 mT and 500 mT. In order to capture the complete visible emission, the camera was set at 20 s exposure time and 25 frames were integrated each having a 200 ns st ep size to get the final image. From Figure 4.2.4 it is clear that the plume expansion profil es along the axial direction increased with higher laser fluence both at 0 mT and 500 mT However due to the high background gas pressure the transverse expans ion profiles at 500 mT are rest ricted as compared to those at 0 mT. This indicates that the kinetic energi es of the ablated species are probably larger along the axial direction at 500 mT compared to those at 0 mT.

PAGE 203

178 1c 1cm 1cm 1cm 1cm 1cm 1cm 1cm 1cm pO2 = 0 mT F = 1 J / cm2 pO2 = 0 mT F = 2 J / cm2pO2 = 0 mT F = 3 J / cm2pO2 = 0 mT F = 4 J / cm2pO2 = 500 mT F = 1 J / cm2 pO2 = 500 mT F = 2 J / cm2pO2 = 500 mT F = 3 J / cm2pO2 = 500 mT F = 4 J / cm2 Figure 4.2.4. ICCD images of total visible emission spectra of single laser plumes varying the excimer fluences (F) from 1 to 4 J/cm2 under different gas pressures (pO2). The plume profiles were also studied for different background oxygen pressure and at constant laser fluence. Figure 4.2.5 show ICCD images of the ablated plumes at 2 J/cm2 under varying pO2 of 100 mT, 300 mT, and 500 mT. Figure 4.2.5. ICCD images of to tal visible emission spectra of laser ablated plumes under different gas pressures (pO2) and constant fluence of 2 J/cm2. The images show transverse plume c onfinement and represent the deposition conditions for the as-grown ZVO films (Table 4.9). The relative intensities for the plumes under 100 mT, 300 mT and 500 mT are 100 %, 62 % and 45 % respectively. The 300mT 1cm 500mT 1cm 100mT 1cm

PAGE 204

179 FWHMs of the plume intensity profiles at 2 cm from the target we re 1.10 cm for 100 mT, 0.71 cm for 300 mT and 0.54 cm for 500 mT respectively. 4.2.2.2. Structural Properties Figure 4.2.6 shows the XRD -2 scans of ZVO films, named as ZVO(100mT), ZVO(300mT), and ZVO(500mT), respectively. The films are highly textured with only (002) orientation. There are no observed peaks from secondary phase formation of oxides of V or other impurities within the resolu tion limits of XRD.The in plane epitaxial relationship was verified by rocking scans ab out the (002) plane of ZnO as shown in Figure 4.2.7. The narrow FWHM (< 1 ) confirms the high degree of texturing in all the samples. However, the slightly larger FWHMs for ZVO(300mT) and ZVO(500mT) as compared to ZVO(100mT) implies a less preferentia l orientation. 30405060708 0 (c)ZVO(100mT)2 ( de g) (b) (a)ZVO(500mT) ZVO(300mT)Intensity (arb. units)ZnO (004)*ZnO (002) Figure 4.2.6. XRD -2 scans (a) ZVO(500mT), (b) ZVO(300mT), and (c) ZVO(100mT) films, respectively. The substrate peak is denoted by *.

PAGE 205

18016.016.517.017.518.0 (c)FWHM = 0.56oZVO(100 mT) (deg)(a)FWHM = 0.87oZVO(300 mT) Intensity (arb. units)(b) FWHM = 0.76oZVO(500 mT) Figure 4.2.7. XRD rocking curves about (002) plane of ZnO for (a) ZVO(500mT), (b) ZVO(300mT), and (c) ZVO(100mT) films, respectively. Figures 4.2.8 (a, b, and c) show AFM images of ZVO(100 mT), ZVO(300 mT), and ZVO(500 mT) films, respectively. It is obs erved that the particulate density on the film surface increases with higher pO2. From Figure 4.2.8 (c) it is apparent that the large number of micron size droplets visible on the ZVO(500mT) film surface possibly came from the target during ablation. To confirm th at the particles are not any foreign object, EDS was performed on top of the droplets. Th e EDS spectra showed the presence of only Zn and O confirming that they came from the target. The root mean square surface roughnesses (Rrms) for ZVO(100mT), ZVO(300mT) and ZVO(500mT) were 17.5 nm, 33.5 nm and 55.8 nm, respectively. The Rrms increased by three times as the pO2 increased from 100 mT to 500 mT.

PAGE 206

181 Figure 4.2.8. AFM images V-doped ZnO thin films grown at 600 C but varying the background O2 pressure from (a) 100 mT, (b) 300 mT, and (c) 500 mT, named as ZVO(100mT), ZVO(300mT), a nd ZVO(500mT) respectively. 4.2.2.3. Electrical properties The electrical properties were measured us ing the Van der Pauw technique. Table 4.10 summarizes the room temperature resistivities ( ) and carrier concentrations (nc) for ZVO films. As observed earlier for ZM O films, in this case also the increases and the nc decreases with higher pO2. The for ZVO(500mT) was higher than the instrumental limits. For this reason, the and nc for a ZVO film grown at 200 mT named ZVO (200mT) has been included in Table 4.10. It emphasized the trend in the electrical properties in ZVO film s as a function of pO2.

PAGE 207

182 Table 4.10. Resistivity ( ), carrier concentration (nC) and Hall mobility ( H) measured at room temperature of ZnO:V thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure. Sample O2 Pressure Sheet resistance ResistivityCarrier Conc.Mobility pO2 (mT)RS( ) ( cm)nC (cm-3) H (cm2/Vs) ZVO(100mT)1002.76 x 1028.82 x 10-32.66 x 101926.51 ZVO(200mT)2002.38 x 1057.613.08 x 10172.67 ZVO(300mT)3001.19 x 1074.53 x 1023.76 x 10153.65 ZVO(500mT)500> 108> 104-4.2.2.4. Ferroelectric properties The ferroelectric properties for the ZVO films on insulating Al2O3 substrates were tested using a co-planar electrode configuration. A schematic diagram of the capacitor has been shown in the inset to Figure 4.2.9. Gold palladium (Au Pd) electrodes were sputter coated on top of the film using a shadow mask with 500 m holes. Figure 4.2 9 shows the polarization (P ) loops for ZVO(100mT), ZVO(300mT), and ZVO(500mT) capacitors, respectively measured at 1 kHz. Table 4.11 summarizes the P values for the ZVO films.The remnant polarization (Pr) for ZVO(500mT) (0.25 C/cm2) is consistent with the reported value (0.2 C/cm2) for ZVO films deposited on Si (111) substrates [161, 162].

PAGE 208

183-6-4-20246 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 V Au-Pd electrodes ZVO film Al2O3 substrate ZVO(500mT) ZVO(300mT) ZVO(100mT)Polarization (C/cm2)Drive voltage (V) Figure 4.2.9. P-V loops for ZVO(100mT), ZVO(300mT), and ZVO(500mT) films, respectively. Inset shows a sche matic diagram of ZVO capacitor. Table 4.11. Summary of polarization data fo r ZnO:V thin films on c-cut sapphire substrates grown at 600 C with varying background O2 pressure. SampleMaximum PolarizationRemnant PolarizationCoercive Field Pmax ( C/cm2)P r ( C/cm2)EC (kV/cm) ZVO(100mT)0.010.00452.05 ZVO(300mT)0.480.14.4 ZVO(500mT)0.830.244.9 The high polarization value for ZVO(500mT) can be associated the reduction in donor type defects like O2 vacancies created during deposition at high pO2. Less number of defects within the film can inhibit the current percolation paths and consequently

PAGE 209

184 reduce the leakage current acro ss the capacitor. This allows the measurement of P by applying higher driving voltage s without causing a dielectric break down in the capacitor. The ferroelectric properties of ZVO film s deposited on Si (100) substrates were also investigated using LSMO top and bottom electrodes (inset Figure 4.2.10). The LSMO electrodes were deposit ed in-situ using PLD at 600 C and 10 mT pO2. Figure 4.2.10 shows the polarization loops of LSMO /ZVO(500mT)/LSMO capacitors measured at 1 kHz. Compared to the co-planer Au-Pd electrodes, the LSMO top-bottom electrodes allow applying higher driving voltages and c onsequently measuring higher polarization. -10-8-6-4-20246810 -2 -1 0 1 2 V ZVO(500mT) (1.3 m) LSMO electrode (100 nm) Si (100) substrate Drive voltage (V)Polarization (C/cm2) 2 V 4 V 5 V 6 V 9 V Figure 4.2.10. Polarization loops for ZVO (500 mT) film. The inset shows a schematic diagram of LSMO/ZVO(500mT)/LSMO capacitor. Figure 4.2.11 (a) shows the frequency depe ndence of polarization at a driving voltage of 5 V for the LSMO/ZVO(500mT)/ LSMO capacitor. The nominal switching field (Vc) increased with lower frequency consiste nt with typical ferroelectric behavior.

PAGE 210

185 Lower frequency allows more time for the dipoles to align with the applied field and thus increases the polariza tion. Figure 4.2.11 (b) shows the cap acitance vs voltage loops for the ZVO(500mT) capacitor. -6.0 -4.5 -3.0 -1.5 0.0 1.5 3.0 4.5 6.0 -6-4-2024 6 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 (b) (a) Drive voltage (V) Frequency 10 Hz 100 Hz 1000 Hz Polarization (C/cm 2 ) Capacitance (nF) Figure 4.2.11. (a)Polarization and (b) capacitan ce versus driving voltage for V-doped ZnO thin film grown pO2 of 500mT, named as Z VO(500mT), respectively. The behavior is characteristic of a lea ky capacitor which is possibly due to the defects in the film which produce percol ation paths for the current. Table 4.12 summarizes the polarization data obtaine d for the ZVO(500mT) films using LSMO electrodes.

PAGE 211

186 Table 4.12. Summary of polarization values for ZVO(500mT) film using LSMO electrodes. Driving VoltageFrequencyPmaxP r (+)P r (-)Vc (+)Vc (-) (V)(Hz) ( C/cm2) ( C/cm2) ( C/cm2) (V)(V) 210000.4880.109-0.1180.384-0.396 410000.9110.254-0.2270.701-1.017 510001.1290.349-0.2920.773-1.42 51001.8490.848-0.7651.366-1.885 5103.6943.313-3.4472.602-3.107 610001.3570.456-0.3640.736-1.856 910002.0550.838-0.5720.447-3.382 91003.4761.95-1.722.174-4.035 4.2.3. Conclusions To summarize, V doped ZnO films have b een grown using PLD. The ferroelectric properties have been measured. Ferroelectri c switching is observed at around 4 V driving voltage. Higher saturation polarization is ach ieved for films grown at higher oxygen pressure. This is related to the more insulating nature of the films at high pressures. However larger polarization needs to be achieved for potential device application.

PAGE 212

187 4.3. ZnO:Mn-ZnO:V Heterostructure A schematic diagram of the ZnO:Mn/ZnO:V heterostructure has been shown in Figure 4.3.1. It consists of three layers. Fi rst, a 300 nm thick conductive ZnO layer was deposited on c-cut sapphire substrate to serve both as a bottom electrode during polarization measurement and a buffer laye r for the epitaxial growth. Second, a 1.2 m ZnO:Mn layer was deposited at the conditions that gave the highest magnetization in ZMO films. Third, a 1.1 m ZnO:V layer was deposited at the conditions that gave the highest polarization in ZVO fi lms. Finally, 0.5 mm diameter Au electrodes were sputter coated to serve as top electrode s during polarization measurements. Figure 4.3.1. Schematic diagram of Zn O:Mn-ZnO:V heterostructure. 4.3.1. Experimental Details The individual layers were grown under the conditions which gave the highest ferromagnetic and ferroelectric response duri ng the growth of the single layers as discussed in section 4.1 and 4.2. The deposition conditions for the in-situ grown individual layers have been summarized in Table 4.13. c-cut Sapphire (001) substrate ZnO (002) (300 nm) ZnO:Mn (002) (1.2 m) ZnO:V (002) (1.1 m) Au electrodes

PAGE 213

188 Table 4.13. Deposition conditions of the layers in the ZnO:Mn-ZnO:V heterostructure on c-cut sapphire substrate. Layers inLaser FluenceGrowth Temperatures O2 Pressure Film Thickness heterostructure (J/cm2) Ts ( C)pO2 (mT) (nm) ZnO260010300 ZnO:Mn2600101200 ZnO:V26005001100 4.3.2. Results and discussions 4.3.2.1. Structural properties Figure 4.3.2 shows the XRD -2 scan of ZnO:Mn-ZnO:V heterostructure. The high counts from the peak intensities confirm the good crystalline quality. The heterostructure is highly textur ed since the major peaks are only from the [002] planes of ZnO. Some low intensity peaks from other orie ntations of ZnO are al so observed. This is associated to the large thickness (approximately 2.6 m) of the whole structure where the epitaxial relationship among the individual laye rs is difficult to maintain throughout the structure. On the other hand, no impurity p eaks from secondary phase segregation are observed even in the logarithmic scale. This implies that the structure is single phase. The degree of in-plane orientation in the hete rostructure was measured by rocking curves performed about the (002) plane of ZnO. Fi gure 4.3.2 shows the rocking curves about the ZnO:Mn-ZnO:V heterostructure and the Zn O:Mn single layer film. The small FWHM (1.158 ) for the heterostructure confirms th e high degree of in-plane texturing.

PAGE 214

18930405060708 0 101102103104105106107 *(004) (103) (110) (002)Intensity (counts)2 (deg) Figure 4.3.2. XRD -2 scan of ZnO:Mn-ZnO:V hetero structure. The c-cut sapphire substrate peak is denoted by *. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 -2-1012 (deg)Intensity (a.u.) ZnO:Mn-ZnO:V ZnO:Mn Figure 4.3.4. Rocking curves about the (002) plane of ZnO for the ZnO:Mn-ZnO:V heterostructure and the Zn O:Mn single layer film.

PAGE 215

190 The slight relaxation of the preferential orientation in the heterostructure as compared to single ZnO:Mn film (FWHM = 0.962 ) is again due to the large thickness and different growth conditions of the individual layers. 4.3.2.2. Magnetic Properties The magnetic properties were for both th e heterostructure and the individual layers. Figure 4.3.5 shows the in-plane M-H loops measured at 10 K and 300 K for the ZnO:Mn-ZnO:V heterostructure, ZnO:Mn single layer, and ZnO:V single layer, respectively. Room temperature (RT) FM is observed in the heterostructure which is important for device applicati on purpose. The magnetization in ZnO:V layer is similar to the undoped ZnO film which implies that V do ping does not contribute to the observed FM in ZnO films. The ZnO:Mn single layer film exhibits the highest magnetization which is related higher carrier concentration in the film. The magnetization in the heterostructure is lower than that in ZnO:Mn single layer. One of the possible explanations is that carrier concentration in the heterostructure is lower. During the growth of the heterostru cture, the ZnO:V layer is deposited at high pO2 (500 mT) which possibly allows more O2 incorporation into the ZnO:Mn layer and reduce s the carriers. The electric field control of magnetiza tion was investigated by measuring the magnetization before and after poling the ZnO:Mn ZnO:V heterostructure. The ferroelectric ZnO:V top layer was poled by a pplying 4 V of d.c. bias voltage across the coplanar Au electrodes (Figure 4.3.1). Figure 4.3.6 shows the M versus H loops at 300 K for the heterostructure before and after poling. A large dr op in the magnetization after poling indicates a possible magnetoelectric co upling at room temperature. This decrease

PAGE 216

191 in magnetization by applying elec tric field can be attributed to a probable interaction between the magnetic moments in ZnO:Mn layer and the electric dipoles in ZnO:V layer. -60-40-200204060 -2 -1 0 1 2 -4 -3 -2 -1 0 1 2 3 4 H ( kOe ) M (emu/cm3)(b)300 K ZnO:Mn ZnO:V ZnO:Mn-ZnO:V ZnO:Mn ZnO:V ZnO:Mn-ZnO:VM (emu/cm3)(a)10 K Figure 4.3.5. M-H loops measured at (a) 10 K and (b) 300 K for the individual layers, ZnO:Mn and ZnO:V, and the ZnO:Mn-ZnO:V heterostructure, all grown on c-cut sapphire susbtrates. The magnetic field wa s applied parallel to the film plane.

PAGE 217

192-60-40-20020406 0 -2 -1 0 1 2 ZnO:Mn-ZnO:V bilayer Before poling After poling300 KM (emu/cm3)H ( kOe ) Figure 4.3.6. Magnetization loops at 300 K befo re and after poling the ZnO:Mn/ZnO:V epitaxial heterostructure. 4.3.3. Conclusions To summarize, multiferroic layered ZnO:Mn-Z nO:V heterostructure has been fabricated using PLD. The heterostructure exhibite d simultaneous magnetic and ferroelectric properties. Possibility of magneto-elastic coupling through a cro ss interaction between the layers has been observed. It is probable that such a coupling mechanism could arise from the difference in carrier concentration across the interface of the two phases. Further research in this direction is required to enha nce the saturation values for practical device application.

PAGE 218

193 4.4. Chapter Summary In conclusion, this chapter described that the magnetic and ferroelectric properties of ZnO thin films may be enhanced by pr oper doping. It was observed that both the ferromagnetism in Mn doped ZnO films and the ferroelectricity in V doped ZnO films depended strongly on the carrier concentratio n in the films. Finally the single phase multiferroic heterostructure was fabricated and characterized. Possibility of magnetoelastic coupling was confirmed in the heterost ructure. Since such structures have been grown for the first time, lack of theoretical background inhibited th e progress. The exact mechanism of magnetoelastic coupling in not clearly understood. The difference of carrier concentration across the interface probably play ed an important role. Further research is needed to enhance magnetiza tion and polarization va lues for a potential device application.

PAGE 219

194 CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS In this thesis the horizontal heterostructures were studied for two systems namely the CFO/PZT bilayer films and the ZnO:Mn /ZnO:V structures. On one hand, it was shown that the interfacial strain played an important part in the observed magnetic and ferroelectric properties of the CFO/PZT f ilms. On the other, it was probably the difference of carrier concentration across the interface in ZnO:Mn/ZnO:V films that lead to the possible magnetoelastic coupling. Both systems showed great potential as a composite multiferroic material. Fabrication of epitaxial CFO/PZT multila yers was also attempted during the thesis. However in order to maintain the epitaxial relationships between the layers, the individual layer thickness had to be limited to tens of nm. At such small thicknesses the PZT layers could not be polarized due to substrate clam ping effects. Increasing the layer thickness destroyed the epitaxy. Growth of CFO/PZT multilaye rs can be been possible using techniques like molecular beam epitaxy (MBE) where deposition of layers at an atomistic scale is possible. In order to increase the magnetism in Mn doped ZnO thin films, higher doping percentages were also attempte d. However, higher doping of Mn into ZnO (6 at. % Mn in ZnO) proved to be futile due to the precipi tation of anti-ferromagnetic Mn oxide phases

PAGE 220

195 in the films. Consequently, the saturated magnetization in the hi gher doped ZnO films was lower than in Zn0.98Mn0.02O films. An important feature that was shown duri ng the course of this thesis was the growth ZnO nanorods on different substrates using PLD (see Appendix D: Publications). The nanorods allow more surface area for better interfacial interactio n of the different phases. Since ZnO can be doped with Mn or V to enhance their magnetic and ferroelectric behavior and studying the coup ling effects in such doped nano-structures could be a future direction of work. Another future direction for composite mu ltiferroic heterostructures could be the ferrimagnetic nano-pillars of CFO embedded in a ferroelectric matrix of PZT as described in Chapter 1 (Section 1.2.2.2). Alt hough there have been some reports of vertical nanostructured films in which CFO nanopillars were embedded in a BiFeO3 film [60, 61], such structures have not been studied in details. In order to grow the CFO nanopillars the oblique angle deposition can be used [190 – 192]. The fo llowing section will give a brief account of the preliminary experiments done in this direction.

PAGE 221

196 5.1. Oblique Angle Deposition of CFO Thin Films Oblique angle deposition using laser ablati on has been reported to be successful in the growth of nano-columnar structures thr ough the shadowing effect and the low adatom mobility [190 192]. During oblique angle depos ition the substrate is tilted so that the angle formed between the surface normal of the target and the surface normal of the substrate is oblique. A schematic diagram of the set up is show n in Figure 5.1.1. The angle shown in the Figure 5.1.1 is the oblique angle. Figure 5.1.1. Schematic diagram of the oblique angle depositi on and the shadowing effect (Adapted from Ref. [190]). As shown schematically in Figure 5.1.2, for normal incidence deposition, nucleation sites or island of materials deposited on the substrates, grow faster in the lateral direction in the substrate plane and ev entually cover up the su rface of the substrate leading to a polycrystalline film with no pr eferred orientation. However, for oblique angle deposition, due to the str ong shadowing effect [193] the islands grow taller in the vertical direction; due to pr eferred orientation of the ma terial concerned. The initial

PAGE 222

197 islands shadow a considerable amount of surface area in the film (Figure 5.1.2). This results in the enhancement of the preferre d growth direction a nd possible columnar growth in the initial deposited layer [194]. Figure 5.1.2. Schematic diagram of the growth of columnar structures by oblique angle deposition (Adapted from Ref. [194]). Due to the increased surface of such co lumnar ferromagnetic layer, with the ferroelectric material deposited on top would lead to strain magnification in the interfaces of the two phases for better coupling between the magnetic and ferroelectric phases. The height and orientations of the columns could be manipul ated by changi ng the oblique angle. A series of CFO films were deposited on Si (100) substrates using an oblique angle ( ) of 120 between the substrate and target normal directions. The substrate temperature and ambient O2 pressure were kept at 450 C and 10 mT respectively. Figure

PAGE 223

198 5.1.3 shows the XRD patterns of oblique angle deposited CFO film w ith the thicknesses varying as 100 nm, 200 nm, 400 nm, and 600 nm, respectively. 10203040506 0 *100 nm(333) (400) (222)2 ( de g) *200 nm(222) (111) (311) (422) (d) (c) (b)400 nm(220) (333) (a)600 nm(111)Intensity (arb. units) Figure 5.1.3. XRD patterns of CFO thin films deposit ed on Si (100) s ubstrates at an oblique incident angle of 60 with different thicknesses as (a) 600 nm, (b) 400 nm, (c) 200 nm and (d) 100 nm, respectively. Th e substrate peak is denoted by *. It is observed that there is preferred growth in the ( 111) direction for 100 nm and 200 nm films (Figures 5.1.3 c, and d). However with increasi ng thicknesses the preferred growth is lost, result ing in polycrystalline bulk like XRD pattern.

PAGE 224

199 Figures 5.1.4 (a, and b) show XRD patt erns of CFO films both deposited under the same conditions as mentioned above but using oblique and normal incidence deposition respectively. The film thickness was 200 nm. 0 50 100 150 200 250 10203040506 0 0 200 400 600 800 1000 (a) Obligue incidence deposition(111) (b) Normal incidence deposition*(333) (222) (333) (222) (311) (311) (111)Intensity (arb. units) 2 ( de g) Figure 5.1.4. XRD patterns of CFO thin f ilms deposited on Si under same conditions using (a) oblique and (b) normal in cidence deposition, respectively. On one hand when a normal incident CFO film of thickness about 200 nm shows no preferred orientation, on the other, the film deposited at oblique incidence still shows preferred growth in the (111) plane. This proves that pref erred direction of growth is enhanced by oblique incidence.

PAGE 225

200 Figure 5.1.5 shows AFM images of the surface of the CFO films deposited on Si substrates at oblique incidence for variuos th icknesses. As the films thicknesses increase the grain boundaries get less distinct and eventually gets all smeared up after 600 nm. Comparing this with the XRD data in Figure 5.1.2 it can be inferred that there is initial preferred growth which gets destro yed as the film thickness increases. Figure 5.1.6 shows in-plane M-H loops for CFO deposited under the same conditions but using oblique and normal in cidence deposition respectively. The film thickness was 200 nm. It can be seen that th e in-plane magnetization is reduced in the oblique angle deposited film. This could indicate that the moments are aligned perpendicular to the film plane for the obli que angle deposited film. However, further investigations are required to confirm the results. It can be concluded that ob lique angle deposition facili tates preferred growth and probable columnar structures. The enhancemen t of preferred growth by oblique angle of deposition has been demonstrated for the firs t time for CFO. This is promising because the ferroelectric layer could be deposited on top of the ob lique angle deposited film. The deposition of the ferroelectric layer should be done using normal incidence deposition. This would make the top layer flat for pol arization measurements. This oblique angle deposition technique could be used in future to grow new composite multiferroic structures.

PAGE 226

201 Figure 5.1.5. AFM images of CFO films with different thicknesses as (a) 200 nm using normal, and (b) 100 nm, (c) 200 nm, (d) 300 nm, (e) 400 nm, and (f) 600 nm using oblique incidence deposition, respectively.

PAGE 227

202-50-2502550 -800 -600 -400 -200 0 200 400 600 800 -800 -600 -400 -200 0 200 400 600 800 in-plane Normal incidence Oblique incidence(b)10 KM (emu/cm3)H (kOe)in-plane Normal incidence Oblique incidence(a)300 K M (emu/cm3) Figure 5.1.6. M-H loops for CFO films on Si substrates measured at (a) 300 K and (b) 10 K grown using normal and oblique incidence depositions. The magnetic field has been applied in-plane.

PAGE 228

203 REFERENCES [1] J. C. Maxwell, Phil. Trans. R. Soc. Lond. 155 459 (1865). [2] W. C. Rontgen, Annalen der Physik 271 264 (1888). [3] P. Curie, J. Physique 3 393 (1894). [4] M. Fiebig, J. Phys. D: Appl. Phys. 38 R123 (2005). [5] S. W. Cheong, and M. Mostovoy, Nature Materials 6 13 (2007). [6] F. Jona, and G. Shirane, Ferroelectric Crystals (Dover, New York, 1993). [7] N. A. Hill, J. Phys. Chem. B 104 6694 (2000). [8] D. I. Khomskii, Bull. Am. Phys. Soc. C 21 002 (2001). [9] N. A. Spaldin, and M. Fiebig, Science 309 391 (2005). [10] J. Akerman, Science 308 508 (2005). [11] J. F. Scott, Science 315 954 (2007). [12] J. F. Scott, Nature Materials 6 256 (2007) [13] M. Gajek, M. Bibes, S. Fusil, K. Bou zehouane, J. Fontcuberata, A. Barthelemy, and A. Fert, Nature Materials. 6 296 (2007). [14] L.W. Martin, Y.-H. Chu, R. Ramesh, Materials Science and Engineering R 68, 89 (2010). [15] Y. Wang, J. Hu, Y. Lin, and Ce -Wen Nan, Npg Asia Materials 2 61 (2010). [16] A. Loidl, H. von Loehneysen, and G.M. Kalvius, J. Phys. Condens. Matter. 20 430301 (2008).

PAGE 229

204 [17] A. Goldman, Modern Ferrite Technology (Van Nostrand Reinhold, New York, 1990). [18] T. Yamaguchi and M. Abe, Ferrites Proceeding Of International Conference in Ferrites (Centre for Academic Publications, Japan, 1992). [19] R. Valenzuela, Magnetic Ceramics (Cambridge University Press, United Kingdom, 1994). [20] B.Y. Wong, X. Sui, D. E. Laughlin, and M. H. Kryder, J. Appl. Phys. 75 5966 (1994). [21] G. F. Dionne, D. E. Oates, D. H. Temme, and J. A. Weiss, IEEE Trans. Microwave Theor. Tech. 44 1361 (1996). [22] H. Watanabe, S. Iida, and M. Sugimoto, Ferrites Proceeding Of International Conference (Centre for Academ ic Publications, Japan, 1980). [23] C. L. Hogan, Bell Syst. Tech. J. 31 1 (1952). [24] A. K. Giri, E. M. Kirkpatrick, P. M oongkhamklang, and S. A. Majetich. Appl. Phys. Lett. 80 2341 (2002). [25] R. J. Gambino, T. Suzuki, Magneto-optical Recording Materials (IEEE Press, New York, 2000). [26] F. J. Cadieu, Permanent magnetic thin films (Academic Press, San Diego, 1992).) [27] R. Langer, Science 249 1527 (1990). [28] K. Butter, P. H. H. Bomans, P. M. Frederik, G. J. Vroege, and A. Phlips, Nature Materials 2 88 (2003). [29] J. S. Moodera, L. R. Kinder, T. M. Wong, R. Meservey, Phys. Rev. Lett. 74 3273 (1995).

PAGE 230

205 [30] J. Zhou et al. Appl. Surface Science 253 7456 (2007). [31] Y. Suzuki, Annu. Rev. Mater. Res. 31 265 (2001). [32] C. N. R. Rao, and J. Gopalakrisnan, New directions in solid state chemistry 2nd ed. (Cambridge University Press, Cambridge, 1997). [33] A. Lisfi, J. C. Lodder, E. G. Keim and C. M Williams, Appl. Phys. Lett. 82 76 (2003). [34] J. I. Gersten, and F. W. Smith, The Physics and Chemistry of Materials (John Wiley & Sons, New York, 2001). [35] E. W. Lee, Rep. Prog. Phys. 18, 184 (1955). [36] E. Kneller, Ferromagnetismus (Springer, Berlin, 1962). [37] D. Sander, Rep. Prog. Phys 62, 809 (1999). [38] D. C. Jiles, and D. L. Atherton, J. Phys. D: Appl. Phys. 17, 1265 (1984). [39] J. F. Scott, Science 315 954 (2007). [40] N. Setter, D. Damjanovic, L. Eng, G. Fox, S. Gevorgian, S. Hong, A. Kingon, H. Kohlstedt, N. Y. Park, G. B. Stephenson, I. Stolitchnov, A. K. Taganstev, D. V. Taylor, T. Yamada, and S. Streiffer, J. Appl. Phys. 100 051606 (2006). [41] J. F. Scott, C. A. Paz de Araujo, Science 246 1400 (1989). [42] J. Schwarzkopf, and R. Foranari, Progress in Crystal Growth and Characterization of Materials 52 159 (2006). [43] N. Izyumskaya, Y. I. Alivov, S. J. Cho, H. Morko, H. Lee, and Y. S. Kang, Crit. Rev. Solid State Mater. Sci. 32 111 (2007). [44] T. J. Zhu, L. Lu, and M. O. Lai, Appl. Phys. A 81 701 (2005). [45] C. Kittel, Introduction to Solid State Physics 7th ed. (John Wiley & Sons, New York, 1996).

PAGE 231

206 [46] H. Schmid, Ferroelectrics 162 317 (1994). [47] W. Eerenstein, N. D. Mathur, and J. F. Scott, Nature 442 759 (2006). [48] N.A. Hill, J. Phys. Chem. B 104 6694 (2000). [49] J. Wang et al., Science 299 1719 (2003). [50] S. M. Skinner, IEEE Trans. Parts, Mater. Packaging 6 68 (1970). [51] F. Kubel, and H. Schmid, Acta Crystallogr. B 46 698 (1990). [52] A. Sharan, J. Lettieri, Y. Jia, W. Tia n, X. Pan, D.G. Schlom V. Gopalan, Phys. Rev. B 69 214109 (2004). [53] T. Atou, H. Chiba, K. Ohoyama, Y. Yamaguichi, Y. Syono, J. Solid State Chem. 145 639 (1999). [54] S. Ryu, J. H. Park, H. M. Jang, Appl. Phys. Lett. 91 142910 (2007). [55] M. Ziese, A. Bollero, I. Panagiotopo ulos, N. Moutis, Appl. Phys. Lett. 88 212502 (2006). [56] Y. G. Ma, W. N. Cheng, M. Nin g, C. K. Ong, Appl. Phys. Lett. 90 152911 (2007). [57] Y. Zhang, C. Deng, J. Ma, Y. Lin, C.-W. Nan, Appl. Phys. Lett. 92 062911 (2008). [58] T. Wu, M. A. Zurbuchen, S. Sa ha, R. V. Wang, S. K. Stre iffer, and J. F. Mitchell, Phys. Rev. B 73 134416 (2006). [59] C.-W. Nan, G. Liu, Y. Lin, H. Chen, Phys. Rev. Lett. 94 197203 (2005). [60] T. Maruyama et al., Nature Nanotech. 4, 158 (2009). [61] H. Zheng et al., Nano Lett. 6 1401 (2006).

PAGE 232

207 [62] J. G. Wan, X. W. Wang, Y. J. Wu and J. M. Liu, Appl. Phys. Lett. 86 122501 (2005). [63] X. L. Zhong, J. B. Wang, M. Liao, G. H. Huang, S. H. Xie, Y. C. Zhou, Y. Qiao, and J. P. He, Appl. Phys. Lett. 90 152903 (2007). [64] H. Ryu et al., Appl. Phys. Lett. 89 102907 (2006). [65] D. H. Lowndes, D. B. Geohegan, A. A. Puretzky, D. P. Norton, and C. M. Rouleau, Science 273 898 (1996). [66] D. B. Chrisey, and G. K. Hubler, Pulsed laser Deposition of Thin Films (Wiley, New York, 1994). [67] M. Okuyama, and Y. Hamakawa, Int. J. Eng. Sci. 29 391 (1991). [68] S.R. Foltyn, Pulsed Laser Deposition of Thin Films D. B. Chrisey and G. K. Hubler eds. (Wiley, New York, pp. 89 1994). [69] L. C. Chen, Pulsed laser Deposition of Thin Films (Wiley, New York, pp. 167, 1994). [70] D. H. Lowndes, Laser Ablation and Desorption J. C. Miller, and R. F. Haglund eds, Experimental Methods in the Physics of Sciences (Academic, NewYork, pp. 475, 1998). [71] T. Venkatesan, and S. M. Green, The Industrial Physicist 2 22 (1996). [72] S. Witanachchi, K. Ahmed, P. Sakthi vel, P. Mukherjee, Appl. Phys. Lett. 66 1469 (1995). [73] P. Mukherjee, S. Chen, and S. Witanachchi, Appl. Phys. Lett. 74 1546 (1999). [74] P. Mukherjee, S. Chen, J. B. Cuff, P. Sakt hivel, and S. Witanach chi, J. Appl. Phys. 91 1828 (2002); 91 1837 (2002).

PAGE 233

208 [75] S. Witanachchi, and P. Mukherj ee, J. Vac. Sci. Technol. A 13 1171 (1995). [76] P. Mukherjee, J. B. Cuff, and S. Witanachchi, Rev. Sci. Instrum. 72 2380 (2001). [77] A. Guinier, X-Ray Diffraction (Freeman, San Francisco, 1963). [78] M. Birkholz, Thin film analysis by X-ray scattering (Wiley-VCH, Weinheim, 2006). [79] Y. Pauleau, Vacuum 61 175 (2001). [80] I. Noyan, and J. Cohen, Residual Stress: Measurement by Diffraction and Interpretation (Springer Verlag, New York, 1987). [81] R. Waser, Nanoelectronics and Information Technology (Wiley-VCH, Weinheim, 2003). [82] L. J. Van der Pauw, Philips Res. Rep. 13 1 (1958). [83] L. J. Van der Pauw, Philips Tech. Rev. 20 220 (1958). [84] Hall Effect Measurements, National Institute of Standards and Technology, (Retrieved from http://www.eeel.nist.gov). [85] C. B. Sawyer, and C. H. Tower, Phys. Rev. 35 269 (1930). [86] S. Hong, Nanoscale Phenomena in Ferroelectric Thin Films (Kluwer Academic Publishers, Boston, 2004). [87] M. Dawber, I. Farnan, and J. F. Scott, Am. J. Phys. 71 (8) 819 (2003). [88] Z. Li, Y. Wang, Y. Lin, and C. Nan, Phys. Rev B 79 180406(R) (2009). [89] C. H. Sim, Z. Z. Pan, and J. Wang, J. Appl. Phys. 105 084113 (2009). [90] H. He, J. Ma, Y. Lin, and C. W. Nan, J. Appl. Phys. 104 114114 (2008). [91] H. -C. He, J. Wang, J. -P. Zhou, a nd C. -W. Nan, Adv. Funct. Mater. 17 1333 (2007).

PAGE 234

209 [92] J. –P. Zhou, H. –C. He, Y. Zhang, C. –Y Deng, Z. Shi, C. –W. Nan, Appl. Phys. A 89 553 (2007). [93] J.-P. Zhou, H. He, Z. Shi, and C. -W. Nan, Appl. Phys. Lett. 88 013111 (2006). [94] N. Ortega, A. Kumar, R. S. Katiyar and C. Rinaldi, J. Mater. Sci. 44 5127 (2009). [95] Y. Zhang, Z. Li, C. Deng, J. Ma, Y. Lin, and C.-W. Nan, Appl. Phys. Lett. 92 152510 (2008). [96] R. Ramesh, and N. A. Spaldin, Nat. Mater. 6 21 (2007). [97] C. Nan, M. I. Bichurin, S. Dong, D. Vieh land, and G. Srinivasan, J. Appl. Phys. 103 031101 (2008). [98] R. M. Bozorth, E. F. Tilden, and A. J. Williams, Phys. Rev. 99, 1788 (1955). [99] D. J. Craik and J. Derek, Magnetic Oxides (Wiley-Interscience, London, p. 327, 1975). [100] P. D. Thang, G. Rijnders, D. H. A. Blank, J. Magn. Magn. Mater. 310 2621 (2007). [101] P. C. Dorsey, P. Lubitz, D. B. Chrisey, and J. S. Horwitz, J. Appl. Phys. 79 6338 (1996). [102] Y. Suzuki, G. Hu, R. B. Van Dover, R. J. Cava, J. Magn. Magn. Mater. 191 1 (1999). [103] G. Hu, J. H. Choi, C. B. Eom, V. G. Harris, and Y. Suzuki, Phys. Rev. B 62 R779 (2000). [104] N. Ranvah, I. C. Nlebedim, Y. Melikhov, J. E. Snyder, D. C. Jiles A. J. Moses, P. I. Williams, F. Anayi, and S. –H. Song, IEEE Trans. Mag. 44 3013 (2008).

PAGE 235

210 [105] J. A. Paulsen, A. P. Ring, C. C. H. Lo, J. E. Snyder, and D. C. Jiles, J. Appl. Phys. 97, 044502 (2005). [106] Y. Chen, J. E. Snyder, K. W. Dennis, R. W. McCallum, and D. C.Jiles, J. Appl. Phys. 87 5798 (2000). [107] M. Inoue, S. Yamamoto, N. Fujita, and T. Fujii, IEEE Trans. Mag. 23 3334 (1987). [108] N. Ortega, P. Bhattacharya, R. S. Katiyar, P. Dutta, A. Manivannan, M. S. Seehra, I. Takeuchi, and S. B. Majumder, J. Appl. Phys. 100 126105 (2006). [109] A.V. Ramos, M. J. Guittet, J. B. Moussy, R. Mattana, C. Deranlot, F. Petroff, and C. Gatel, Appl. Phys. Lett. 91 122107 (2007). [110] W. Huang, J. Zhu, H. Z. Zeng, X. H. Wei, Y. Zhang, and Y. R. Li, Appl. Phys. Lett. 89 262506 (2006). [111] R. S. Hassan, N. Viart, C. Ulhaq-Bouillet, J. L. Loison, G. Versini, J. P. Vola, O. Crgut, G. Pourroy, D. Muller and D. Chateigner, Thin Solid Films 515 2943 (2007). [112] P. C. Dorsey, P. Lubitz, D. B. Chrisey, and J. S. Horwitz, J. Appl. Phys. 79 (8) 6338 (1996). [113] W. Huang, L. X. Zhou, H. Z. Zeng, X. H. We i, J. Zhu, Y. Zhang, and Y. R. Li, J. Cryst. Growth 300 426 (2007). [114] S. R. Murthy, Cryst. Res. Technol. 25 461(1990). [115] Lisfi, and C. M. Williams, J. Appl. Phys. 93 8143 (2003). [116] Lisfi, C. M. Williams, L. T. Nguyen, J. C. Lodder, A. Coleman, H. Corcoran, A. Johnson, P. Chang, A. Kumar, and W. Morgan, Phys. Rev. B 76 054405 (2007).

PAGE 236

211 [117] G. Hu, J. H. Choi, C. B. Eom, V. G. Harris, and Y. Su zuki Phys. Rev. B 62 779 (2000). [118] J. H. Yin, J. Ding, B. H. Liu, J. B. Yi, X. S. Miao, and J. S. Chen, J. Appl. Phys. 101 09K509 (2007). [119] B. Xiao, V. Avrutin, H. Liu, . zgr, H. Morko, and C. Lu, Appl. Phys. Lett. 93 052913 (2008). [120] D. Walker, P. A. Thomas, and S. R. Collins, Physica Status Solidi AApplications and Materials Science 206 1799 (2009). [121] P. S. Anderson, S. Guerin, B. E. Hayden, M. A. Khan, A. J. Bell, Y. Han, M. Pasha, K. R. Whittle, and I. M. Reaney, Appl. Phys. Lett. 90 202907 (2007). [122] T. Fujisawa, H. Nakaki, R. Ikariyama, T. Yamada, M. Ishikawa, H. Morioka, and H. Funakubo, J. Appl. Phys. 105 061614 (2009). [123] A. Talin, S. M. Smith, S. Voight, J. Finde r, K. Eisenbeiser, D. Penunuri, Z. Yu, P. Fejes, T. Eschrich, J. Curless, D. Convey, and A. Hooper, Appl. Phys. Lett. 81 1062 (2002). [124] C. Yang, J. Liu, S. Zhang, Z. Chen, Mater. Sci. Eng. B 99 356 (2003). [125] X. Y. Chen, and Z. G. Liu, Appl. Phys. A 69 523 (1999). [126] S. R. Foltyn, R. C. Dye, K. C. Ott, E. Peterson, K. M. Hubbard, et. al., Appl. Phys. Lett. 59 594 (1991). [127] R. K. Kelly, and A. Miotello, Pulsed Laser Deposition of Thin Films ed. by D. B. Chrisey, and G. K. Hubler (Wiley, New York, p. 55, 1994). [128] B. Brailovsky, S. V. gaponov, and V. I. Luchin, Appl. Phys. A 61 81 (1995).

PAGE 237

212 [129] S. K. Pandey, A. R. James, C. Prakash, T. C. Goel, and K. Zimik, Mater. Sci. Eng. B 112 96 (2004). [130] S. Habouti, A. Lahmar, M. Dietze, C. H. Solterbeck, V. Zaporojtchenko, and M. Es-Souni, Acta Materialia 57 2328 (2009). [131] C. H. Jeon, C. S. Kim, K. B. Han, H. S. Jhon, and S. Y. Lee, Mater. Sci. Eng. B 109 141 (2004). [132] J. Lee, R. Ramesh, V. G. Keramidas, W. L. Warren, G. E. Pike, and J. T. Evans, Appl. Phys. Lett. 66 1337 (1995). [133] S. Dussan, A. Kumar, J. F. Scott, and R. S. Katiyar, Appl. Phys. Lett. 96 072904 (2010). [134] F. Chen, Q. Z. Liu, H. F. Wang, F. H. Zhang, and W. Wu, Appl. Phys. Lett. 90 192907 (2007). [135] N. Ortega, A. Kumar, and R. S. Katiyar, J. Appl. Phys. 104 074114 (2008). [136] W. Wu, K. H. Wong, C. L. Choy, a nd Y. H. Zhang, Appl. Phys. Lett. 77 3441 (2000). [137] D.E. Jesson, S.J. Pennycook, J.M. Bari beau, D.C. Houghton, Phys. Rev. Lett. 71, 1744 (1993). [138] M. Spankova, S. Chromik, I. Vavra, K. Se dlackkova, P. Lobotka, S. Lucas, and S. Stancek, Appl. Surface Science 253 7599 (2007). [139] R. M. Bozorth, Ferromagnetism 6th Printing (D. Van Nostrand Company Inc., New York, 1961) [140] D. Damjanovic, The Science of Hysteres is, ed. by I. Mayergoyz, and G.Bertotti (Elsevier, Amsterdam, 2005).

PAGE 238

213 [141] S. Sidorkin, L. P. Nesterenko, S. V. Ry abtsev, and A. A. Sidorkin, Phys. Solid State 51 1348 (2009). [142] Q. Xu, H. Schmidt, S. Zhou, K. Potzger, M. Helm, H. Hochmuth, M. Lorenz, A. Setzer, P. Esquinazi, C. Meinecke, and M. Grundmann, Appl. Phys. Lett. 92 082508 (2008). [143] A. D. Corso, M. Posternak, Raffaele, and A. Baldereschi, Phys. Rev. B 50 10175 (1994). [144] R. D. Vispute et. al, Appl. Phys. Lett 73 348 (1998). [145] Q. H. Li, T. Gao, Y. G. Wang, a nd T. H. Wang, Appl. Phys. Lett. 86 123117 (2005). [146] G. Sberveglieri, Sensors Actuators B 23 103 (1995). [147] M. Dutta, and D. Basak, Nanotechnol. 20 475602 (2009). [148] N. Kislov, J. Lahiri, H. Verma, D. Y. Goswami, E. Stefanakos, M. Batzill, Langmuir 25 3310 (2009). [149] F. Xu, Z. -Y. Yuan, G. -H. Du, T. -Z. Ren, C. Bouvy, M. Halasa, B. -L. Su, Nanotechnol. 17 588 (2006). [150] S. Singh, P. Thiyagarajan, K. M. Kant, D. Anita, S. Thirupathiah, N. Rama, B. Tiwari, M. Kottaisamy, and M. S. R Rao, J. Phys. D: Appl. Phys. 40 6312 (2007). [151] C. Klingshirn, Phys. Status Solidi B 224 3027 (2007). [152] U. Ozgur, Y. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Dogan, V. Ayrutin, S. J. Cho, and H. Morkoc, J. Appl. Phys. 98 041301 (2005). [153] M. Batzill, and U. Diebold, Prog. Surf. Sci. 79 47 (2005).

PAGE 239

214 [154] H. E. Unalan, D. Wei, K. Suzuki, S. Dala l, P. Hiralal, H. Matsumoto, S. Imaizumi, M. Minagawa, A. Tanioka, A. J. Flewitt, W. I. Milne, G. A. J. Amaratunga, Appl. Phys. Lett. 93 133116 (2008). [155] H. H. Chen, A. D. Pasquier, G. Saraf, J. Zhong, Y. Lu, Semicond. Sci. Technol. 23 045004 (2008). [156] W.J. Lee, H. Okada, A. Wakahara, and A. Yoshida, Ceram. Int. 32 495 (2006). [157] Y. Y. Villanueva, D. Liu, and P. T. Cheng, Thin Solid Films 501 366 (2006). [158] F. Claeyssens, C. L. Freeman, N. L. Alla n, Y. Sun, M. N. R. Ashfold, and J. H. Harding, J. Mater. Chem. 15 139 (2005). [159] F. Pan, C. Song, X. J. Liu, Y. C. Ya ng, F. Zeng, Mater. Sci. Engineering R 62 1 (2008). [160] Z. W. Liu, S. G. Yang, R. V. Rama nujan, and C. K. Ong, Mater. Lett. 62 1255 (2008). [161] Y. C. Yang, C. Song, X. H. Wang, F. Zeng, and F. Pan, Appl. Phys. Lett. 92 012907 (2008). [162] Y. C. Yang, C. Song, F. Zeng, F. Pan, Y. N. Xie, and T. Liu, Appl. Phys. Lett. 90 242903, (2007). [163] T. Dietl, H. Ohno, F. Matsukura, J. Cibert, and D. Ferrand, Science 287 1019 (2000). [164] X. M. Cheng, and C. L. Chien, J. Appl. Phys. 93 7876 (2003). [165] T. Fukumura, Z. Jin, M. Kawasaki, T. S hono, T. Hasegawa, S. Koshihara, and H. Koinuma, Appl. Phys. Lett. 78 958 (2001).

PAGE 240

215 [166] G. Lawes, A. R. Risbud, A. P. Ramirez, and R. Seshadri, Phys. Rev. B 71 045201 (2005). [167] S. W. Jung, S. J. An, G. C. Yi, C. U. Jung, S. Lee, and S. Cho, Appl. Phys. Lett. 80 4561 (2002). [168] P. Sharma, A. Gupta, K.V. Rao, F. J. Owe ns, R. Sharma, R. Ahuja, J. M. Osorio Guillen, B. Johansson, and G. A. Gehring, Nature Mat. 2 673 (2003). [169] D. C. Kundaliya, S. B. Ogale, S. E. Lofland, S. Dhar, C. J. Metting, S. R. Shinde, Z. Ma, B. Varughese, K. V. Ramanujachary, L. Salamanca-Riba, and T. Venkatesan, Nat. Mater. 3 709 (2004). [170] M. A. Garca, M. L. Ruiz-Gonzles, A. Quesada, J. L. Costa Krmer, J. F. Fernandz, S. J. Khatib, A. Wennberg, A. C. Caballero, M. S. Martn-Gonzles, M. Villegas, F. Briones, J. M. Gonzles -Galbet, and A. Hernando, Phys. Rev. Lett. 94 217206 (2005). [171] A. K. Pradhan, K. Zhang, S. Mohanty, J. B. Dadson, D. Hunter, J. Zhang, D. J. Sellmyer, U. N. Roy, Y. Cui, A. Burger S. Mathews, B. Joseph, B. R. Sekhar, and B. K. Roul, Appl. Phys. Lett. 86 152511 (2005). [172] M. Suchea, S. Christoulakis, M. Katharakis, G. Kiriakidis,N. Katsarakis, E. Koudoumas, Appl. Surface Science 253 8141 (2007). [173] S. Ramachandran, J. Narajan, and J. T. Prater, Appl. Phys. Lett. 88 242503 (2006). [174] J. M. D. Coey, M. Venkatesan, and C. B. Fitzgerald, Nat. Mat. 4 173 (2005). [175] Q. Xu, H. Schmidt, L. Hartmann, H. Hochmuth, M. Lorenz, A. Setzer, P. Esquinazi, C. Meinecke, and M. Grundmann, Appl. Phys. Lett. 92 092503 (2007).

PAGE 241

216 [176] M. J. Calderon, and S. Das Sharma, Annals of Physics 322 2618 (2007). [177] Xu Zuo, Soack-Dae Yoon, Aria Yang, WenHui Duan, Carmine Vittoria, and Vincent G. Harris, J. Appl. Phys. 105 07C508 (2009). [178] Z. Yang, M. Biasini, W. P. Beyermann, M. B. Katz, O. K. Ezekoye, X. Q. Pan,Y. Pu, J. Shi, Z. Zuo, and J. L. Liu, J. Appl. Phys. 104 113712 (2008). [179] H. Verma, D. Mukherjee, S. Witanachch i, P. Mukherjee, a nd M. Batzill, J. Crystal Growth 312 2012 (2010). [180] K Ellmer, J. Phys. D: Appl. Phys. 34 3097 (2001). [181] W. M. Hlaing Oo, L. V. Saraf, M. H. Engelhard, V. Shutth anandan, L. Bergman, J. Huso, and M. D. McCluskey, J. Appl. Phys. 105 013715 (2009). [182] N. H. Hong, J. Sakai, and V. Brize, J. Phys.: Condensed Matter 19, 036219 (2007). [183] M. Venkatesan, C. B. Fitzgerald, J. G. Lunney, and J. M. D. Coey, Phys. Rev. Lett. 93 177206 (2004). [184] F. W. Smith, Phys. Rev. B 10 2980 (1974); 14 241 (1976). [185] D. J. Priour, Jr., E. H. Hwang, a nd S. Das Sarma, Phys. Rev. Lett. 92 117201 (2004). [186] L. Brus, J. Phys. Chem. 90 2555 (1986). [187] C. K. Ghosh, K. K. Chattopadhyay, a nd M. K. Mitra, J. Appl. Phys. 101 124911 (2007). [188] T. Shibata, K. Unno, E. Makino, Y. Ito, and S. Shimada, Sens. Actuators, A 102 106 (2002). [189] M. H. Zhao, Z. L. Wang, and S. X. Mao, Nano Lett. 4 587 (2004). [190] H. Wang, and Y. P. Zhao, J. Vac. Sci. Technol. B 24(3) 1230 (2006).

PAGE 242

217 [191] Y. Ma, F. Liu, M. Zhu, J. Liu, H. Wang, Yi Yang, and Y. Li, Nanotechnol. 20 275201 (2009). [192] Y. Shim, and J. G. Amar, Phys. Rev. Lett. 98 046103 (2007). [193] K. Robbie, and M. J. Brett, J. Vac. Sci. Technol. A 15 1460 (1997). [194] T. Karabacak, A. Mallikarjunan, J. P. Singh, D. Ye, G. C. Wang, and T. M. Lu, Appl. Phys. Lett. 83 3096 (2003).

PAGE 243

218 APPENDIX A: POLYCRYSTALLINE CFO THIN FILMS

PAGE 244

219 Appendix A: Polycrystal line CFO thin films Polycrystalline CFO thin films were gr own on Si and c-cut sapphire substrates. Figures A.1 (a, b, c, and d) show SEM im ages of CFO-Si films deposited at 450 C varying the background O2 pressure (pO2) as 0 mT, 10 mT, 40 mT and 60 mT, respectively. The surface appears to be sm ooth and relatively clean with very few particulates in the film deposited at 10 mT (Figure A.1. b) which was chosen as pO2 during deposition. The film thicknesses were 200 nm. Figure A.1. SEM images of CFO films on Si (100) substrates deposited at 450 C varying the background O2 pressure as (a) 0 mT, (b) 10 mT, (c) 40 mT, and (d) 60 mT, respectively. Figures A.2 (a, b, and c) show the XRD -2 scans of CFO-Si films grown under the same conditions for different thicknesses of 200 nm, 100 nm and 50 nm, respectively. The XRD pattern of the 200 nm thick film (Figure A.2 a) demonstrates a bulk-like polycrystalline nature (see Figure 3.1.6 for powder XRD). (a) (c) (b) (d)

PAGE 245

220 Appendix A (continued) The observed peaks are indexed to the f ace-centered cubic phase of CFO with a space group of Fd-3m (227) and la ttice parameter a = 8.391 . 2030405060 (440) (333) (222) (111)50 nm 2 ( de g) (111) (440) (333) (222) (311) Si(100) Si(100)100 nm Intensity (a.u.) (c) (b) (a)(440) (422)200 nm(333) (222) (311) (111) Figure A.2. XRD patterns for (a) 200 nm, (b) 100 nm and (c) 50 nm thick films of CFO grown on Si (100) substrate, respectively. However, the XRD pattern of the 100 nm thick film (Figure A.2 b) shows a (111) crystallographic texture as indicated by the high relative intensity of the series of [111] peaks compared to the (311) peak (highest peak in the bulk XRD pattern). The XRD pattern of the 50 nm thick film (Figure A.2 c) shows only a st rong (111) text ure. It has been reported earlier that below a critical fi lm thickness, the (111) orientation is the preferred growth direction fo r CFO-Si thin films [1, 2].

PAGE 246

221 Appendix A (continued) Figure A.3 shows the XRD -2 scan of CFO film grown on c-cut sapphire (Al2O3) (0001) substrate. The pattern shows only a strong (111) texture with no addition orientations of CFO although the film thickness is 200 nm. 2030405060 Sapphire substrate* (333)*(222) (111)Intensity (a.u.)2 ( de g) Figure A.3. XRD -2 scan of CFO film grown on c-cut sapphire substrate. It has been reported that spinel-type ferr ite thin films grow with a strong (111) orientation on Al2O3 substrates [3, 4]. The hexa gonal surface unit cells of Al2O3 substrate provide triangular lattice nucle ation sites for the CFO (111) plane. This gives a small lattice mismatch of 2.8 % between CFO and Al2O3 along the (111) direction. The average crystallite sizes (D) in CFO films were calculated from the broadening of the XRD -2 peaks using the Scherrer formula [5, 6].

PAGE 247

222 Appendix A (continued) The Scherrer formula estimates the crys tallite diameter D assuming spherical crystallites as D = 4/3L where L is given by Bcos L (A.1) Here L is the coherence le ngth of reflected X-rays, is the particle shape factor (for spherical particles, = 0.94), is the wavelength of X-rays (for CuK = 1.54439 ), B is the FWHM of -2 peak and is the angle of diffraction. Table A.1 summarizes the crystallite sizes (D) calculated using Equation A.1. The lattice parameters (a) were also calculated from the XRD -2 peaks and listed in Table A.1. Table A.1. Lattice parameter (a) and crystall ite size (D) calculated using Scherrer formula from the XRD patterns for CFO target a nd the deposited films on Si and sapphire substrates. SampleFilm ThicknessLattice parameterCrystallite size (nm)a ()D (nm) CFO target-8.391 0.00462 9 CFO-Si2008.402 0.01435 7 CFO-Si1008.391 0.0535 10 CFO-Si 508.423 0.00328 5 CFO-Al2O32008.429 0.00943 17 The lattice parameters for the 200 nm and 100 nm thick CFO-Si films (Table A.1) are very close to that for the CFO target (or bulk powder). The unchangeable lattice parameters suggest that there are no detectab le stresses related to lattice distortions in these films. However the lattice parameter in the 50 nm thick CFO-Si film is larger than the CFO powder (Table A.1) which implies th at the film is under a tensile stress.

PAGE 248

223Appendix A (continued) Further, in all the samples the crystallite sizes are in the range 28-40 nm which is smaller than the critical single domain size for CFO ( 40 nm) [7]. Figure A.4 shows the in-plane and out-of -plane M-H loops at 300 K and 10 K for the (111) textured 100 nm thick CFO-Si films. -50-2502550 -800 -600 -400 -200 0 200 400 600 800 -800 -600 -400 -200 0 200 400 600 800 (b)10 K in-plane out-of-planeM (emu/cm3)H (kOe) (a)300 K in-plane out-of-plane M (emu/cm3) Figure A.4. M-H loops measured at (a) 300K and (b) 10 K for 100 nm thick textured polycrystalline CFO film grown on Si (100) for in plane and out of pl ane configuration. The magnetic properties have been summ arized in Table A.2. The magnetic behavior of CFO-Si(100nm) film is almost isotropic with similar Ms and Ms values.

PAGE 249

224Appendix A (continued) At 300 K, the values are about 4 and 4.5 B per Co site (Table A.2), respectively, which is higher with the bulk value (4 B per Co site) [8]. This could be related to the textured growth and unstressed state of the film as confirmed by XRD (Table A.2).The higher Ms in CFO-Si film compared to the bulk CFO is possibly due to the smaller grain size as also confirmed by XRD crysta llite size calculation (Table A.2). Table A.2. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane and out-ofplane configurations for 100 nm and 50 nm th ick CFO films on Si substrates and 200 nm thick film of CFO on sapphire substrates. The symbols and denote the in-plane and out-of-plane configurations respectively. 300 K Sample Ms Ms Mr/Ms Hc Ms Ms Mr/Ms Hc (emu/cm3) ( B/Co2+) (%)(kOe) (emu/cm3) ( B/Co2+) (%)(kOe) CFO-Si (100nm)502 124.030.10.5564 64.515.80.4 CFO-Si (50nm)616 144.913.80.2>631>5.3>13.70.3 CFO-Al2O3 (200nm)342 52.712.20.3>325>2.6>7.60.1 10 K Sample Ms Ms Mr/Ms Hc Ms Ms Mr/Ms Hc (emu/cm3) ( B/Co2+) (%)(kOe) (emu/cm3) ( B/Co2+) (%)(kOe) CFO-Si (100nm)632 35.047.22441 122.530.90.3 CFO-Si (50nm)570 54.515.30.2>2401.9>25.40.3 CFO-Al2O3 (200nm)473 53.844.35>2251.8>14.20.2 Figure A.5 shows the M-H loops at 300 K a nd 10 K for the highly (111) textured 50 nm thick CFO-Si film. The M-H loop shows strong in-plane anisotropy. This behavior is consistent with earlier repor ts which suggest th at below a critical film thickness, the shape anisotropy in CF O overcomes the magnetocrystalline anisotropy [9].

PAGE 250

225Appendix A (continued) This aligns the magnetic moments along th e film plane i.e (111) crystallographic plane for CFO-Si(50 nm) film. The magnetic properties have been summarized in Table A.2. -10-50510 -800 -600 -400 -200 0 200 400 600 800 -800 -600 -400 -200 0 200 400 600 800 (b)10 K in-plane out-of-planeM (emu/cm3)H ( kOe ) (a)300 K in-plane out-of-plane M (emu/cm3) Figure A.5. M-H loops measured at (a) 300K and (b) 10 K for 50 nm thick highly (111) textured CFO film grown on Si (100) for in plane and out of plane configuration. Figure A.6 shows the M-H loops 200 nm thick CFO-Al2O3 film. The M-H loops show strong in-plane anisotropy.

PAGE 251

226Appendix A (continued) The magnetic properties have been summar ized in Table A.2. The magnetization in CFO-Al2O3(200nm) film is similar to CFO-Si( 50nm) film except for the higher Hc and Mr/Ms ratio at 10 K. -50-2502550 -600 -400 -200 0 200 400 600 -600 -400 -200 0 200 400 600 10 K(b) in-plane out-of-planeM (emu/cm3)H (kOe)300 K(a) in-plane out-of-planeM (emu/cm3) Figure A.6. M-H loops measured at (a) 300K and (b) 10 K for 200 nm thick highly (111) textured CFO film grown on c-cut sapphire substrate for in plane and out of plane configurations. The thickness dependence of magnetization for CFO-Si films is shown in Figure A.7. Table A.3 lists the in-plane magnetic properties for all the samples.

PAGE 252

227Appendix A (continued) The Ms increased and Hc decreased with decreasing film thickness. The results are consistent with ea rlier reports [10]. Table A.3. Saturation magnetization (Ms), ratio of remnant magnetization (Mr) to Ms (squareness), and coercive field (Hc) measured at 300 K and 10 K for in-plane configurations for 200nm, 100 nm and 50 nm thick CFO films on Si substrates. The symbol denotes the in-plane configurations. 300 K SampleMs Ms M r /Ms Hc (emu/cm3) ( B/Co2+)(%)(kOe) CFO-Si (200nm)416 23.327.61.2 CFO-Si (100nm)502 124.030.10.5 CFO-Si (50nm)616 144.913.80.2 10 K SampleMs Ms M r /Ms Hc (emu/cm3) ( B/Co2+)(%)(kOe) CFO-Si (200nm)368 92.959.210 CFO-Si (100nm)632 35.047.22 CFO-Si (50nm)>5704.5>15.30.2 Low field switching in Hysteresis loops An interesting feature seen in the low temperature M-H loops of the CFO films is the low magnetic field switching (Figure A.7)). Step-like features due to such switching occur symmetrically on both sides of the zero fi eld line. The steps seen in the M-H loops are similar to those observed in composite thin films [11] and in bilayers composed of hard and soft magnetic materials [12]. There could be several explanations for the low field switching such as the effect of substratefilm interface [11] or th e characteristic of a preferred orientation.

PAGE 253

228Appendix A (continued) -50-2502550 -600 -400 -200 0 200 400 600 800 -800 -600 -400 -200 0 200 400 600 800 50 nm 100 nm 200 nm 10 K(b)M (emu/cm3)H ( kOe ) 50 nm 100 nm 200 nm(a)300 K M (emu/cm3) Figure A.7. M-H loops measured at (a) 300K and (b) 10 K for CFO films grown on Si (100) for different thicknesses. The magne tic field was parallel to film plane. In order to investigate this, the temp erature dependence of magnetization was measured. Figures A.8 (a, b, and c) show the in-plane M-H loops for the 200 nm thick CFO film on Si substrate meas ured at different temperatur es from 10 K to 300 K. The low field switching is clearly visible at low te mperatures (10 K) due to the increase in coercivity (Hc) resulting from the reduced thermal influence on the alignment of magnetic moments. Although the M-H loop at 300K (Figure A.8 c) is narrower it still

PAGE 254

229 shows the step like switching when is magnified in the vicinity of the zero field regions. This confirms that low field switching is i ndependent of temperature. The temperature dependence of coercivity (Hc) is shown in Figure A.8 (d). In the case of a thin film with uniaxial symmetry, the temperature dependence of Hc can be fit to the following relation [13]: 2 11B CO CT T H H (A.2) -50-2502550 -500 -250 0 250 500 (a)M (emu/cm2)10K-50-2502550 H (kOe) (b)H (kOe)150K-50-2502550 (c)H (kOe)300K0100200300 500 2500 4500 6500 8500 (d) Hc (Oe)Temperature (K) Figure A.8. M-H loops measured at (a) 10 K, (b) 150 K and (c) 300 K for a 200 nm thick CFO film grown on Si (100) subs trate. (d) The coercivity (Hc) as a function of temperature.

PAGE 255

230Appendix A (continued) From the slope of the fit shown in Figur e A.8 (d), the blocking temperature (TB) is deduced to be about 355 K, which is consis tent with earlier reports for CFO (~390K) [13]. The increase of coercivity with loweri ng temperature was also observed for the case of the 100 nm thick film, which has bot h (111) and (311) dominant peaks. However, the switching behavior is not observed in the M-H loops for the highly (111) textured 50 nm thick CFOSi film, even when magnified in the vicinity of the zero fields at 300 K and 10 K, as shown in Fi gure A.9. This implies that the low field switching probably arises only in thicker films from an interfacial layer at the filmsubstrate interface as repo rted earlier [14, 15]. -5-4-3-2-1012345 -1.0 -0.5 0.0 0.5 1.0 No low field switching CFO-Si(50 nm) M/MsH (kOe) 10K 300K Figure A.9. In-plane M-H loops measured at 300K and 10 K for 50 nm thick CFO film grown on Si (100) substrate. Steps seen in thicker films (100 nm and above) due to low field switching are not present in this case.

PAGE 256

231Appendix A (continued) To affirm this, the in-plane M-H loops at 10 K for both the 50 nm and 200 nm films are overlapped as shown in Figure A. 10. The M-H loop for the CFO-Si(50 nm) film lies exactly on the onset of the magnetic sw itching for the CFO-Si (200 nm) film, on both sides of the field sweep (Figure A.10). -50-2502550 -1.0 -0.5 0.0 0.5 1.0 10 K CFO-Si(200 nm) CFO-Si(50 nm)M/MsH (kOe) Figure A.10. In plane M-H loops for CFO-Si 50 nm and 200 nm films measured at 10K. The y-axis is normalized with the saturati on magnetic field. The M-H loop of the (111) oriented film overlaps on the first step of the magnetic reversal seen on the 200 nm film. From Figure A.6 (b) it is observed that the in-plane magnetization for the 200 nm thick (111) textured CFO-Al2O3 film also exhibits a step-like switching feature. This eliminates the possibility that the low fi eld switching is associated with the (111) orientation. Thus the low field switching c ould arise mainly from the interface between the substrate and the film.

PAGE 257

232Appendix A (continued) Growth of interfacial laye r at the substrate-film in terface is common in magnetic thin films [16]. The magnetic moments within the interfacial layer are probably aligned parallel to the film plane and are magnetical ly decoupled from the rest of the film. The degree of the alignment of the magnetizati on at the interface possibly depends on the interfacial stress and the thickness of the interf acial layer. The in-plane anisotropy of the interfacial layer facilitates the magnetic reversal when the applied field is parallel to the film plane. This explains why the low fiel d switching is observed only in the in plane MH loops but not in the out of plane configurations. The low field step like features discusse d above are different from the narrowing feature of the M-H hysteresis l oops at low fields for CFO-Si (100nm) film. This behavior has also been observed for CFO nano-particles and can be attributed to the small grain size in the films [17]. Probing magnetic switching using RF transverse susceptibility A quantitative analysis of the field streng th required to align the moments in the interfacial layer is discusse d in the following paragraphs. While the low field switching feature is clearly observed in the M-H loops, it is difficult to determine precisely magnetic switc hing field and particul arly its dependence on measuring temperature. Thus a very sensitiv e and precise RF transverse susceptibility (TS) technique was utilized to probe magne tic switching and its temperature dependence in CFO films. The RF transverse sus ceptibility technique developed by Srikanth [18] et al., is based on a sensitive, se lf-resonant tunnel diode oscilla tor (TDO) incorporated into the PPMS.

PAGE 258

233Appendix A (continued) It has been validated over the years to be an excellent t echnique for probing effective magnetic anisotropy and switching in a wide range of magnetic materials ranging from thin films [19], nanop articles [20] to single crystals [21]. In TS experiments, a small perturbing AC or RF field (< 10 Oe) is applied to the sample in addition to the swept dc magnetic field. Since the sample is pl aced in an inductive RF coil that is part of a self-resonant circuit, the shift in the res onant frequency with varying dc magnetic field and/or temperature give a direct measure of the change in i nductance and hence the sample susceptibility. The change in transverse susceptibility with dc magnetic field has been expressed by: % 100 ) ( ) ( ) ( %max max H H HT T (A.3) where Hmax is the maximum applied d.c. magnetic field. Theoretically, the transverse susceptibility spectrum in a unipol ar field scan from positive to negative saturation should consist of three singularities of which two occur at the anisotropy fields ( Hk) and one at the switching field (Hs) [22, 23]. However, it has been experimentally shown that depending upo n the magnetic nature of the sample and the orientation in which the sample is introdu ced in the inductive coil, either anisotropy peaks or switching peaks could be prominently observed. In the present study, RF transverse suscep tibility measurements were conducted at different temperatures (T = 300K, 200K a nd 10K) for the CFO films grown on MgO and STO. The maximum dc field applied was 5 kOe. The results are disp layed in Figure A.11 and its inset. It can be clea rly seen from Figure A.11 that the TS curves show only switching peak, but no anisotropy peaks are detected.

PAGE 259

234Appendix A (continued) The absence of anisotropy peaks can be understood due to the fact that the anisotropy fields of the films investigat ed were beyond the maximum applied dc magnetic field. However, it wa s possible to detect the low field switching present in the cobalt ferrite films. The change of TS peak position with te mperature (see inset of Figure A.11) clearly reveals that for the CFO films i nvestigated, the switching field increases as the temperature is decreased. For the film grown on MgO, the switching fields at 300K, 200K and 10 K were 70Oe, 80 Oe and 115 Oe respectively. In case of the CFO film grown on STO, the switching fields obtaine d for 300K, 200K and 10K were 80 Oe, 90 Oe and 135 Oe respectively. -1300-800-3002007001200 0.01 0.02 0.03 0.04 010020030070 90 110 130 CFO on STO T=300K / max (%)Field ( Oe ) Film on MgO Film on STOTemperature (K)Switching Field(Oe) Figure A.11. TDO measurement of CFO film s on STO at 300K with the film plane perpendicular to the DC field. The line with solid square represent the field sweep from positive to negative, where as the solid circle s represent the field sweep from negative to positive.

PAGE 260

235Appendix A (continued) The larger values of switching field for the case of the film grown on STO may be reconciled with the fact that the lattice mismat ch and thus the stress is larger in this sample. The stress due to lattice mismatch would be more significant at the interface. This indicates that the origin of the low fiel d switching steps seen in the M-H loops of the CFO films is associated with the stress at the interface betwee n the substrate and the film. References [1] A. Raghunathan, I. C. Nlebedim, D. C. Jiles, and J. E. Snyder, J. Appl. Phys. 107, 09A516 (2010). [2] J. H. Yin, J. Ding, B. H. Liu, X. S. Miao, J. S. Chen, J. Magn. Magn. Mater. 310, 2537 (2007). [3] M. T. Johnson, P. G. Kotula, C. B. Carter, J. Cryst. Growth 206, 299 (1999). [4] Y. Yamamoto, H. Tanaka, and T. Kawai, Jpn. J. Appl. Phys. 40, L545 (2001). [5] B. D. Cullity, Elements of X-ray Diffraction, 2nd ed. (Addison-Wesley, New York, 1978). [6] H. P. Klug, and L. E. Alexander, X-ray Diffraction Procedures for Polycrystalline and Amorphous Materials 2nd ed. (John Wiley & Son Inc., New York, 1974). [7] C. N. Chinnasamy, B. Jeyadevan, K. Shinoda, K. Tohji, D. J. Djayaprawira, M. Takahashi, R. Justin Joseyphus, and A. Narayanasamy, Appl. Phys. Lett. 83, 2862 (2003). [8] A. Goldman, Modern Ferrite Technology, 2nd ed. (Springer, New York, 2006). [9] M. C. Terzzoli, S. Duhalde, S. Jacob, L. Steren, C. Moina, J. Alloys and Compounds 369, 209 (2004). [10] D. T. Margulies, F. T. Parker, M. L. Rudee, F. E. Spada, J. N. Chapman, P. R. Aitchson, and A. E. Berkowitz, Phys. Rev. Lett. 79, 5162 (1997). [11] R. Hajndl, J. Sanders, and H. Srikanth, J. Appl. Phys. 93, 7999 (2003).

PAGE 261

236Appendix A (continued) [12] Y. Suzuki, R. B. Van Dover, E. M. Gyorgy, J. M. Phillips, and R. J. Felder, Phys. Rev. B 53, 14016 (1996). [13] Natl. Bur. Stand. (U.S.) Monogr. 25, Sec. 9, 22 (1971). [14] J. Pommier, P. Meyer, G. Penissard, J. Fe rre, P. Bruno, and D. Renard, Phys. Rev. Lett. 65, 2054 (1990). [15] R. Allenspach, M. Stampanoni, and A. Bischof, Phys. Rev. Lett. 65, 3344 (1990). [16] A. Biswas, M. Rajeswari, R. C. Srivastava Y. H. Li, T. Venkatesan, R. L. Greene and A. J. Millis, Phys. Rev. B 61, 9665 (2000). [17] C. Vazquez-Vazquez, M. A. Lopez-Quin tela, M. C. Bujan-Nunez, and J. Rivas, J. Nanopart. Res. DOI 10.1007/s11051-010-9920-7 (2010). [18] H. Srikanth, J. Wiggins, and H. Rees, Rev. of Sci. Inst. 70, 3097 (1999). [19] N. A. Frey, S. Srinath, H. Srikanth, M. Varela, S. Pennycook, G. X. Miao and A. Gupta, Phys. Rev. B 74, 024420 (2006). [20] P. Poddar, J. L. Wilson, H. Srikanth, D. F. Farrell, and S. A. Majetich, Phys. Rev. B 68, 214409 (2003). [21] G. T. Woods, P. Poddar, H. Srikanth and Y. M. Mukovskii, J. Appl. Phys. 97, 10C104 (2005). [22] A. Aharoni, E. H. Frei, S. Shtrikman, a nd D. Treves, Bull. Res. Counc. Isr., Sect. F 6A, 215 (1957). [23] L. Spinu, H. Srikanth, A. Gupta, X. W. Li, and G. Xiao, Phys. Rev. B 62, 8931 (2000).

PAGE 262

237 APPENDIX B: LSMO ELECTRODES

PAGE 263

238Appendix B: LSMO electrodes PZT capacitors fabricated using noble me tal electrodes like Pt or Au exhibit a significant loss of switching polar ization with repeated swit ching cycles also know as ‘ferroelectric fatigue in thin films’ [1]. This loss of switchable polarization limits the lifetime of a ferroelectric device, such as ferroelectric nonvolatile memory, where the write and read cycles both rely on ferroelect ric switching. [1].To overcome this problem metallic oxide electrodes such as RuO2, La0.5Sr0.5CoO3, YBBCu3Oy, LaNiO3, IrO2 and La0.7Sr0.3MnO3 (LSMO) are often used [2 8]. Fati gue is caused by the defects due to oxygen vacancies in the PZT films. [1]. It ha s been suggested that the oxide electrodes act as a sink for oxygen vacancies so that they do not pile-up at the film-electrode interface [3] [9]. This reduces the fatigue [10]. LSMO is a meta llic oxide that has attracted great interest recently due to it s colossal magnetoresistance [11]. Epitaxial LSMO films are highly conducting with a room temperature resistivity of 300 -cm [12]. PZT capacitors with LSMO top and bottom electrodes (LSMO/PZT/LSMO) have shown better fatigue endurance compar ed to other oxide electrodes [8]. The LSMO bottom electrodes were first de posited as a thin layer on the MgO and STO substrates prior to the PZT layer deposition. PLDSL was used with a KrF fluence of 2 J/cm2. The LSMO target was purchased from Kurt J. Leskar Company with a composition of La0.7Sr0.3MnO3. The LSMO films were deposited at 600 C by varying the ambient oxygen pressure pO2 from 10 mT to 100 mT for 100 nm thickness. The small lattice mismatch (8 %) be tween LSMO (pseudo-c ubic, a = 0.387 nm) and MgO (cubic, a = 0.421 nm) allowed for th e epitaxial growth of LSMO on MgO [13]. However, it was observed that the pO2 influenced the epitaxial growth.

PAGE 264

239Appendix B (continued) Figures B.1 (a, b, c, an d d) show XRD patterns of LSMO powder, and LSMO films deposited at 100 mT, 50 mT, and 10 mT of pO2 on MgO substrates, respectively. 203040506 0 LSMO-MgO 10 mTIntensity (arb. units)2 ( de g) MgO (200)* LSMO (100) LSMO (200) LSMO-MgO(d)50 mT MgO (200)*LSMO powder LSMO-MgO 100 mT (214) (122) (024) (202) (113) (104) (110) (012)MgO (200)(c) (b) (a) Figure B.1. XRD patterns of (a) LSMO powder, and LSMO films on MgO (100) substrates deposited using pO2 of (a) 100 mT (LSMO-MgO 100mT), (b) 50 mT (LSMOMgO 50 mT), and (c) 10 mT (LSMO-MgO 10 mT ), respectively. The small peak denoted by around 38 in (c) and (d) is an artifact from the MgO substrate. The XRD pattern for the 100 mT deposited film (Figure B.1. b) exhibits a polycrystalline nature similar to that of the LSMO powder (Figure B.1 a).

PAGE 265

240Appendix B (continued) The peaks are indexed with the rhombohe dral LSMO structure with a space group R-3c (167) (a = 5.5 , c = 13.37 ) [14]. Ho wever, the LSMO films deposited at 50 mT and 10 mT pO2 (Figure B.1 c, and d) are highly text ured in the (100) direction with no other orientations observed. There is a peak shift to lower angles as compared to Figure B.1 (a). The peaks are indexed to the LSMO pseudo-cubic perovskite structure (a = 3.87 ). The results are consistent with previous reports [13]. From the above study, the pO2 during deposition of LSMO-MgO films was fixed at 10 mT. The similar crystal structure between LSMO (pseudo-cubic, a = 0.387 nm) and STO (cubic, a = 0.3905 nm) and the extremel y small lattice mismatch (0.009 %) allow for the growth of almost perfectly epitaxi al LSMO films on STO substrates [15, 16]. Figures B.2 (a, and b) show XRD patterns fo r LSMO-STO film and STO substrate. The XRD patterns are almost identi cal indicating the highly epitax ial nature of the film. The insets (I) and (II) to Figure B.2 (a) shows the details of the STO(200)/LSMO(200) and STO(300)/LSMO(300) peaks. The inset to Figure B.2 (b) shows the rocking curve performed about the STO (200) plane. The FWHM of the rocking curve is 0.268 which confirms the high degree of in orientation in STO. Figure B.3 shows SEM image of LSMO-S TO film (bottom electrode for PZT capacitors). The surface appear s to be dense and compact indicative of a layer by layer growth mode as reported earlier [16]. The average grain size is less than 100 nm. Further, no particulates are observed which could de grade the growth of the subsequent PZT layer.

PAGE 266

241Appendix B (continued) 20304050607080 46.446.646.8 72.472.672.873.0 23.023.524.0 LSMO/STO (300) LSMO/STO (200) LSMO/STO (100) Intensity (arb. units)STO (300) STO (200) STO (100) (deg) 2 (deg) 2 (deg)STO (200)(b)LSMO (200) STO (200) LSMO (300) STO (300)2 (deg)(II) (I) (a) Figure B.2. XRD patterns of (a) LSMO film on STO substrate and (b) STO substrates, respectively. The insets (I) and (II) to (a) shows the details of th e STO(200)/LSMO(200) and STO(300)/LSMO(300) peaks. The inset to (b) shows the rock ing curve performed about the STO (200) plane. Figure B.3. SEM image of the surface of LSMO film deposited on STO substrate.

PAGE 267

242Appendix B (continued) After the deposition of the PZT layer, the top LSMO electrodes were deposited in-situ using a shadow mask that produced 100 m diameter contacts as shown in the SEM image in Figure B.4. Figure B.4. SEM images of LSMO top elec trodes on PZT film grown using a shadow mask that produced 100 m diameter contacts. An interesting feature was observed in PZTDL films deposited on STO substrates using a KrF fluence 1 J/cm2 and CO2 fluence of 2 J/cm2 with an interpulse delay t = 50 ns as described in Chapter 3, Section 3.2 (F igure 3.2.10 f). In this case, the initial CO2 pulse melted the target surface, followed by the KrF pulse which ablated the materials from the molten pool. This condition was comp letely opposite to that required for particulate free films. In this condition, huge amount of particulates are expected to get ejected from the target surface and get deposited on the film surface. PZT film LSMO bottom electrodes LSMO top electrodes LSMO film on STO substrate 100 m

PAGE 268

243Appendix B (continued) Figures B.5 (a to d) show SEM images at increasing magnification of the surface of PZTDL-STO film deposited using t = 50 ns. Figure B.5. SEM images of PZT films on STO substrates depos ited using dual laser ablation with a KrF fluence 1 J/cm2 and CO2 fluence of 2 J/cm2 with an interpulse delay t = 50 ns. At lower magnification (Figure B.5 a) the su rface appears particulate laden and rough. However at higher magnifications (Figure B.5 b and c) it appears that the particulates on the surface are uniformly distributed and have a distinct shape.

PAGE 269

244Appendix B (continued) Unlike the ‘splashed’ molten droplets seen earlier (Figure B.5 d), these particulates appear to be cubic or tetragonal just like PZT crystals. This suggests that the particulates are a result of crystallization on the film surface. The mono-dispersed arrangement (Figure B.5 d) coupled with th e epitaxial and pure phase nature revealed from XRD make these films interesting. This could imply that the dual laser ablation process can be used not only to remove par ticulates but also to deliberately deposit particulates for specific application. These f ilm did not show any fe rroelectric behavior and further research is required which is not under the premise of this work. References [1] D. Dimos, W. L. Warren, and H. N. Al-Shareef, Thin Film Ferroelectric Materials and Devices ed. by R. Ramesh (Kluwer Academic Publishers, Boston, p. 199, 1997). [2] H. N. Al-Shareef, O. Auciello, and A. I. Kingon, J. Appl. Phys. 77, 2146 (1995). [3] M. S. Chen, T. B. Wu, and J. M. Wu, Appl. Phys. Lett. 68, 1430 (1996). [4] R. Ramesh, W. K. Chan, B. Wilkens, H. Gilchrist, T. Sands, J. M. Tarascon, D. K. Fork, J. Lee, and A. Safari, Appl. Phys. Lett. 61, 1537 (1992). [5] R. Ramesh, H. Gilchrist, T. Sands, V. G. Keramidas, R. Haakenaasen, and D. K. Fork, Appl. Phys. Lett. 63, 3592 (1993). [6] I. Stolichnov, A. Tagantsev, N. Setter, J. S. Cross, and M. Tsukada, Appl. Phys. Lett. 74, 3552 (1999). [7] T. Nakamura, Y. Nakao, A. Kamisawa and H. Takasu, Appl. Phys. Lett. 65, 1522 (1994). [8] F. Chen, Q. Z. Liu, H. F. Wang, F. H. Zhang, and W. Wu, Appl. Phys. Lett. 90, 192907 (2007).

PAGE 270

245Appendix B (continued) [9] H. N. Al-Shareef, B. A. Tuttle, W. L. Wa rren, T. J. Headley, D. Dimos, J. A. Voight, and R. D. Nasby, J. Appl. Phys. 79, 1013 (1996). [10] J. J. Lee, C. L. Thio, and S. B. Desu, J. Appl. Phys. 78, 5073 (1995). [11] R. von Helmolt, J. Wecker, B. Holzapfel, L. Schultz, and K. Samwer, Phys. Rev. Lett. 71, 2331 (1993). [12] J. M. D. Coey, M. Viret, L. Ranno, and K. Ounadjela, Phys. Rev. Lett. 75, 3910 (1995). [13] M. Spankova, S. Chromik, I. Vavra, K. Sedlackova, P. Lobotka, S. Lucas, and S. Stancek, Appl. Surface Science 253, 7599 (2007). [14] Z. Zhang, R. Ranjith, B. T. Xie, L. You, L. M. Wong, S. J. Wang, J. L. Wang, W. Prellier, Y. G. Zhao, and T. Wu, Appl. Phys. Lett. 96, 222501 (2010). [15] H. Zhang,Y. Anb, Z. H. Maib, H. B. Lub, K. Zhaob, G. Q. Panc, R. P. Lic, and R. Fan, Physica B 403, 2008 (2008). [16] A. M. Haghiri-Gosnet, J. Wolfman, B. Mercey, Ch. Simon, P. Lecoeur, M. Korzenski, M. Hervieu, R. Desfeux, and G. Baldinozzi, J. Appl. Phys. 88, 4257 (2000).

PAGE 271

246 APPENDIX C: PLUME IMAGES

PAGE 272

247Appendix C: Plume Images In order to study the visibl e laser ablated plume propaga tion towards the substrate as a function of time, ICCD images were cap tured after various time intervals with the respect to a reference zero time. Figure C. 1 shows time gated ICCD images using 200 ns exposure time of single laser ablated plumes fr om PZT target using a laser fluence of 2 J/cm2 under 500 mT ambient O2 gas. The position of the target is denoted by the white dotted rectangular box on the left hand side of each image. It is clearly seen that the relative separation between the plume and the target increases with time and reaches a constant separation after 6 s. Figure C.1. Time gated ICCD images using 200 ns exposure time of single laser ablated plumes from PZT target us ing laser fluence of 2 J/cm2 under 500 mT ambient O2 gas. The highest intensities from the plume propagation images were plotted as a function of distance from the target and time as shown in Figure C.2. 2( s) 1 cm 1 ( s) 1 cm 0 ( s) 1 cm 4 ( s) 1 cm 6 ( s) 1 cm 8 ( s) 1 cm

PAGE 273

248Appendix C (continued) It can be seen that the intensity decreases almost exponentially with time and distance from the target surface. 0246810 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.00.51.01.52.02.5 Distance from tar g et surface ( cm ) Time (s)Intensity (arb. units) Figure C.2. Plot of the highest intensities of the visible pl umes captured at various time intervals as a function of the distance from target surface and time. Figure C.3. shows ICCD images (20 s exposure time, 200 ns step size) of total visible emission spectra of singl e laser plumes (a d) varying the excimer (KrF) fluences from 1 to 4 J/cm2 and dual laser plumes (e –f) varying excimer fluences keeping 2 J/cm2 CO2 (IR) fluence and 100 ns peak to peak inter-pulse delay under vacuum. The plumes using dual laser ablation show broader transverse crosssection similar to the ones discussed in Chapter 3 (Section 3.2). However, unlike the ones in Chapter 3 which were under high ambient, the most intense part is closer to the target surface.

PAGE 274

249Appendix C (continued) Figure C.3. ICCD plume images using si ngle and dual laser ablation under vacuum. 1cm (a) KrF 1J/cm2 1cm (b) KrF 2J/cm2 1cm (c) KrF 3J/cm2 1cm (d) KrF 4J/cm2 1cm 1cm 1cm 1cm KrF 1J/cm2 CO2 2J/cm2 t = 100ns KrF 1J/cm2 CO2 2J/cm2 t = 100ns KrF 1J/cm2 CO2 2J/cm2 t = 100ns KrF 1J/cm2 CO2 2J/cm2 t = 100ns (e) (g) (f) (h)

PAGE 275

250Appendix C (continued) Figure C.4 shows ICCD images of CO2 laser ablated plumes captured at different time intervals using a CO2 fluence of 3 J/cm2. It can be seen that particulates are ejected from the target surface at 3 J/cm2 and they become visible only after the intense plume disappears. Figure C.4. ICCD images of CO2 laser ablated plumes captured at various time intervals showing the ejection of particulat es for the target surface using 500 s exposure times. t = 0 s t = 500 s t = 1000 s t = 1500 s 1cm 1cm 1cm 1cm

PAGE 276

251 APPENDIX D: PUBLICATIONS

PAGE 277

252Appendix D: Publications Journal Articles [1] D. Mukherjee, T. Dhakal, H. Srikanth, P. Mukherjee, and S. Witanachchi, “Evidence for carrier-mediated magnetis m in Mn-doped ZnO thin films”, Phys. Rev. B 81, 205202 (2010). [2] T. Dhakal*, D. Mukherjee*, P. Mukherjee, R. Hyde, M. H. Phan, H. Srikanth, and S. Witanachchi, “Magnetic anisotr opy and field-switching in cobalt ferrite thin films deposited by pulsed la ser ablation”, J. Appl. Phys. 107, 053914 (2010). (*These authors c ontributed equally) [3] H. Verma, D. Mukherjee, S. Witanachchi, P. Mukherjee, and M. Batzill, “Comparative study of ZnO thin film and nano-pillar growth on YSZ(111) and sapphire (0001) substrates by pulsed laser deposition”, J. Cryst. Growth 312, 2012 (2010). [4] D. Mukherjee, T. Dhakal, R. Hyde, P. Mukherjee, H. Srikanth, and S. Witanachchi, “Role of epitaxy in contro lling the magnetic a nd magnetostrictive properties of cobalt ferrite-PZT bilayers”, J. Phys. D: Appl. Phys. (under review). Referred Conference Proceedings [1] D. Mukherjee, R. Hyde, T. Dhakal, H. Srikanth, P. Mukherjee and S. Witanachchi, “Investigation of the Pb depletion in single and dual pulsed laser deposited epitaxial PZT thin films and thei r structural characterization”, Mater. Res. Soc. Symp. Proc. 1199, 37 (2010).

PAGE 278

253Appendix D (continued) [2] T. Dhakal, D. Mukherjee, R. Hyde, H. Srikanth, P. Mukherjee, and S. Witanachchi, “Enhancement in ferroelectri city in V-doped ZnO thin film grown using laser ablation” Mater. Res. Soc. Symp. Proc. 1199, 44 (2010). [3] D. Mukherjee, T. Dhakal, H. Srikanth, P. Mukherjee, and S. Witanachchi, “Growth of Epitaxial ZnO:Mn/ZnO:V Heterostructures and Ferroelectricferromagnetic Characterization,” Mater. Res. Soc. Symp. Proc. 1161, 02 (2009), also appeared in “Novel Materials and Devices for Spintronics: MRS Proceeding Volume 1183 (2009)”. [4] R. Hyde, M. Beekman, D. Mukherjee, G. Nolas, P. Mukherjee, and S. Witanachchi, “Growth and characterizati on of germanium-based type I clathrate thin films deposited by pulsed laser abla tion”, Advances in Electronic Ceramics, Ceramic Engineering and Science Proceed ings, Edited by: C. Randal, Hua-Tay Lin, K. Koumoto, and P. Clem, Vol. 28, (2007) [5] S. Witanachchi, R. Hyde, M. Beekman, D. Mukherjee, P. Mukherjee, and George S. Nolas, “Synthesis and Char acterization of Bulk and Thin Film Clathrates for Solid State Power Conversion Applications”, Proc. 25th Int. Conference on Thermoelectrics, p. 44, Viena, Austria, Aug. 2006. [6] S. Witanachchi, R. Hyde, M. Beekman, D. Mukherjee, P. Mukherjee, and G. S. Nolas, “Synthesis and Characterization of Bulk and Thin Film Clathrates for Solid State Power Conversion Applications”, 25th International Conference on Thermoelectrics, IEEE Proceedings (2006).

PAGE 279

254Appendix D (continued) CONFERENCE PRESENTATIONS Oral Presentations [1] “Dual-laser Deposition of Stoichiometric PZT/CoFe2O4 Epitaxial Heterostructures” D. Mukherjee, R. Hyde, T. Dhakal, H. Srikanth, P. Mukherjee, and S. Witanachchi (2010 MRS Sp ring Meeting, San Francisco) [2] “Growth of Epitaxial ZnO:Mn/ZnO:V Heterostructures and Ferroelectricferromagnetic Characterization,” D. Mukherjee, T. Dhakal, H. Srikanth, P. Mukherjee, and S. Witanachchi (2009 MRS Spring Meeting, San Francisco). Poster Presentations [1] “Investigation of the Pb depletion in single and dual pulsed laser deposited epitaxial PZT thin films and thei r structural characterization”, D. Mukherjee, R. Hyde, T. Dhakal, H. Srikanth, P. Mukhe rjee, and S. Witanachchi (2009 MRS Fall Meeting, Boston). [2] “Enhancement in ferroelectricity in Vdoped ZnO thin film grown using laser ablation” T. Dhakal, D. Mukherjee, R. Hyde, H. Srikanth, P. Mukherjee, and S. Witanachchi (2009 MRS Fall Meeting, Boston). [3] “Growth of epitaxial CoFe2O4/PZT heterostructures and ferroelectricferromagnetic characterization” D. Mukherjee, T. Dhakal, R. Hyde, P. Mukherjee, H. Srikanth, and S. Witan achchi (2008 MRS Fall Meeting, Boston).

PAGE 280

255Appendix D (continued) [4] “Growth of epitaxial CoFe2O4/PZT heterostructures and ferromagnetic characterization” D. Mukherjee, T. Dhakal, R. Hyde, P. Mukherjee, H. Srikanth, and S. Witanachchi (2008 Poster Sympos ium & Competition, “Global Challenges for the 21st Century, USF).

PAGE 281

ABOUT THE AUTHOR Devajyoti Mukherjee earned his B.S. degree (1st class with dist inction, silver medal) and M.S. degree (1st class) in Physics in 2002 and 2004, respect ively, from Jadavpur University, Kolkata, India. In 2004, he was awarded a Junior Research Fellowship by the Council of Scientific and Industrial Research University Grants Commission (CSIR-UGC), Govt. of India. In 2005, he came to USA and joined the Physics Department at the University of South Florida (USF). He was awarded the Duckwall Summer Fellowship by the Physic s department, USF, two times during the course of his dissertation. He has published several pee r-reviewed journal articles and scientific papers at conferences in the ar eas of material science, thin films and multiferroics.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004732
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Mukherjee, Devajyoti.
0 245
Growth and characterization of epitaxial thin films and multiferroic heterostructures of ferromagnetic and ferroelectric materials
h [electronic resource] /
by Devajyoti Mukherjee.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: Multiferroic materials exhibit unique properties such as simultaneous existence of two or more of coupled ferroic order parameters (ferromagnetism, ferroelectricity, ferroelasticity or their anti-ferroic counterparts) in a single material. Recent years have seen a huge research interest in multiferroic materials for their potential application as high density non-volatile memory devices. However, the scarcity of these materials in single phase and the weak coupling of their ferroic components have directed the research towards multiferroic heterostructures. These systems operate by coupling the magnetic and electric properties of two materials, generally a ferromagnetic material and a ferroelectric material via strain. In this work, horizontal heterostructures of composite multiferroic materials were grown and characterized using pulsed laser ablation technique. Alternate magnetic and ferroelectric layers of cobalt ferrite and lead zirconium titanate, respectively, were fabricated and the coupling effect was studied by X-ray stress analysis. It was observed that the interfacial stress played an important role in the coupling effect between the phases. Doped zinc oxide (ZnO) heterostructures were also studied where the ferromagnetic phase was a layer of manganese doped ZnO and the ferroelectric phase was a layer of vanadium doped ZnO. For the first time, a clear evidence of possible room temperature magneto-elastic coupling was observed in these heterostructures. This work provides new insight into the stress mediated coupling mechanisms in composite multiferroics.
590
Advisor: Sarath Witanachchi, Ph.D.
653
Pulsed Laser Deposition
Magnetic Anisotropy
Spintronics
Dilute Magnetic Semiconductor
Cobalt Ferrite
CoFe2O4
Lead Zirconium Titanate
PZT
Manganese or Vanadium Doped Zinc Oxide
ZnO:Mn
ZnO:V
LSMO
RKKY
690
Dissertations, Academic
z USF
x Physics
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4732