<%BANNER%>

Synthesis and Characterization of Electroceramic Foams for 3-3 Piezoelectric Composites

Permanent Link: http://ufdc.ufl.edu/UFE0024086/00001

Material Information

Title: Synthesis and Characterization of Electroceramic Foams for 3-3 Piezoelectric Composites
Physical Description: 1 online resource (84 p.)
Language: english
Creator: Wucherer, Laurel
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: characterization, composite, foam, piezoelectric, synthesis
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: We investigated processing-structure-property-performance interrelationships of barium titanate foam in 3-3 piezocomposites for sensor applications. The first area of investigation focuses on the tailoring of the foam microstructure via processing parameters. Barium titanate foams were synthesized via direct foaming method using commercial powders. In order to control the microstructure and by extension the properties, synthesis parameters such as foaming agent, composition, sintering time, and sintering temperature were varied to determine their influence. Microstructure was classified in terms of strut characteristics, average grain size, average cell window size, porosity, and grain boundary integrity. The two systems that fabricated homogeneous structures favorable to ultrasonic sonar applications were a laboratory developed and a commercial polyurethane system, both silicon-free. When sintered at 1400 degrees Celsius for 8 hours they produced thick, dense struts and uniform cell window-size distributions, though with different characteristics. The foam prepared with the laboratory synthesized polyurethane had a porosity of 90 percent, an average grain size of 44 microns, and an average cell window size of 67 microns, while the foam prepared with the commercial polyurethane had a porosity of 87 percent, an average grain size of 20 microns, and an average cell window size of 99 microns. The second area of investigation was the resulting mechanical properties of the foams. They were measured using confined compression testing to determine if the foams had acceptable properties for their intended application and could withstand handling. The mechanical properties (compressive modulus and collapse stress) determined from the experimental data increased with increasing density. Additionally, the lab developed PU foam showed a higher mechanical strength due to a higher average density. The data was then fit with a phenomenological model. The focus of this work is on the elastic response due to the application requirements. The final area of investigation is the infiltration of the foam with polymer to create a 3-3 piezocomposite. Polymer selection was based on a criterion of material properties including viscosity, cure time, cure temperature, and hardness. The samples were infiltrated with Epofix epoxy using Epovac infiltration equipment which is a vacuum impregnation system that infiltrates all empty space in the foam created by the cell window network with the epoxy. Using SEM, it was determined that the foam was completely infiltrated, creating a 3-3 composite. Preliminary electrical characterization of the piezoelectric properties indicated that the hydrostatic figure of merit of the composite was calculated to be lower than that of bulk barium titanate and bulk lead zirconate titanate. This is due to the low compliance of the polymer chosen for the composite and the lower piezoelectric properties of barium titanate in comparison to PZT and the low volume fraction of piezoelectric material in the composite. Based on these results, the next step is to create a 3-3 composite with lead zirconate titanate to determine if the piezoelectric properties are further enhanced in this system. From the areas investigated above, it is concluded that the 3-3 piezoelectric composite is achievable.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Laurel Wucherer.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Nino, Juan C.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024086:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024086/00001

Material Information

Title: Synthesis and Characterization of Electroceramic Foams for 3-3 Piezoelectric Composites
Physical Description: 1 online resource (84 p.)
Language: english
Creator: Wucherer, Laurel
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: characterization, composite, foam, piezoelectric, synthesis
Materials Science and Engineering -- Dissertations, Academic -- UF
Genre: Materials Science and Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: We investigated processing-structure-property-performance interrelationships of barium titanate foam in 3-3 piezocomposites for sensor applications. The first area of investigation focuses on the tailoring of the foam microstructure via processing parameters. Barium titanate foams were synthesized via direct foaming method using commercial powders. In order to control the microstructure and by extension the properties, synthesis parameters such as foaming agent, composition, sintering time, and sintering temperature were varied to determine their influence. Microstructure was classified in terms of strut characteristics, average grain size, average cell window size, porosity, and grain boundary integrity. The two systems that fabricated homogeneous structures favorable to ultrasonic sonar applications were a laboratory developed and a commercial polyurethane system, both silicon-free. When sintered at 1400 degrees Celsius for 8 hours they produced thick, dense struts and uniform cell window-size distributions, though with different characteristics. The foam prepared with the laboratory synthesized polyurethane had a porosity of 90 percent, an average grain size of 44 microns, and an average cell window size of 67 microns, while the foam prepared with the commercial polyurethane had a porosity of 87 percent, an average grain size of 20 microns, and an average cell window size of 99 microns. The second area of investigation was the resulting mechanical properties of the foams. They were measured using confined compression testing to determine if the foams had acceptable properties for their intended application and could withstand handling. The mechanical properties (compressive modulus and collapse stress) determined from the experimental data increased with increasing density. Additionally, the lab developed PU foam showed a higher mechanical strength due to a higher average density. The data was then fit with a phenomenological model. The focus of this work is on the elastic response due to the application requirements. The final area of investigation is the infiltration of the foam with polymer to create a 3-3 piezocomposite. Polymer selection was based on a criterion of material properties including viscosity, cure time, cure temperature, and hardness. The samples were infiltrated with Epofix epoxy using Epovac infiltration equipment which is a vacuum impregnation system that infiltrates all empty space in the foam created by the cell window network with the epoxy. Using SEM, it was determined that the foam was completely infiltrated, creating a 3-3 composite. Preliminary electrical characterization of the piezoelectric properties indicated that the hydrostatic figure of merit of the composite was calculated to be lower than that of bulk barium titanate and bulk lead zirconate titanate. This is due to the low compliance of the polymer chosen for the composite and the lower piezoelectric properties of barium titanate in comparison to PZT and the low volume fraction of piezoelectric material in the composite. Based on these results, the next step is to create a 3-3 composite with lead zirconate titanate to determine if the piezoelectric properties are further enhanced in this system. From the areas investigated above, it is concluded that the 3-3 piezoelectric composite is achievable.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Laurel Wucherer.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Nino, Juan C.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-06-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024086:00001


This item has the following downloads:


Full Text

PAGE 1

1 SYNTHESIS AND CHARACTERIZATION OF ELECTROCERAMIC FOAMS FOR 3 3 PIEZOELECTRIC COMPOSITES By LAUREL WUCHERER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEG REE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Laurel Wucherer

PAGE 3

3 ACKNOWLEDG MENTS First of all, I would like to acknowledge my advisor, Dr. Juan C. Nino, for his support and guidance. His high standards and constant en courage ment to strive for the best led me to many opportunities during my college career and for my future professional career. I thank Dr. Ghatu Subhash for his invaluable suggestions and rewarding discussions. I would also like to thank my other committe e members (Dr. Wolfgang Sigmund and Dr. Jacob Jones) for their time. Further, I would like to thank Professor Enrico Traversa and Fr ancesco Basoli for their innovative work in the field of ceramic foams which gave me the motivation and fundamentals to begi n my research I would also like to convey my heartfelt thanks to the former and current group members of NRG (Nino Research Group): Samantha Yates, Shobit Omar, Lu Cai, Wei Qiu, Peng Xu, Mohammed Elshennawy, Satyajit Phadke, Marta Giachino, Kevin Tierne y, Donald Moore, Kevin Lee, Louis Perez, Julian Gomez and Adam Wilk f or providing me with an excellent research environment and helpful advice. I also wish to acknowledge all of my friends at the University of Florida (UF) with whom I expe rienced ALL of G ainesville during the last 4 and half years. Without Nikki and Kristen, the triumvirate would not have prevailed. I am especially grateful to UF’s Danza Dance Company which gave me the creative outlet to pursue my passion of dance during my college years. F inal “cheers” go to my former roommate Michelle Motola, who endur ed my long hours and mood swings during the hard times Finally, I am grateful to my entire family especially my mother Pat Poole, my father Carl Wucherer a nd my “mini me” Devon, who provi ded me with the intelligence and confidenc e to take advantage of all life has to offer.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .............. 3 LIST OF TABLES ................................ ................................ ................................ ........................ 6 LIST OF FIGURES ................................ ................................ ................................ ..................... 7 ABSTRACT ................................ ................................ ................................ ................................ 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ................... 11 1.1 Statement of Problem and Motivation ................................ ................................ .............. 11 1.2 Scientific Approach ................................ ................................ ................................ .......... 11 2 BACKGROUND ................................ ................................ ................................ .................... 14 2.1 Piezoelectricity ................................ ................................ ................................ ................. 14 2.2 Piezoelectric Ceramics ................................ ................................ ................................ .... 14 2.3 Ceramic Polymer Piezoelectric Composites ................................ ................................ .... 15 2.4 Ceramic Foams ................................ ................................ ................................ ............... 17 2.5 In Situ Foam Processing ................................ ................................ ................................ 18 3 FOAM SYNTHESIS AND MICROSTRUCTURAL ANALYSIS ................................ ............... 26 3.1 Introduction ................................ ................................ ................................ ...................... 26 3.2 Experimental Procedure ................................ ................................ ................................ .. 27 3.3 Results and Discussion ................................ ................................ ................................ ... 30 3.3.1 Effect of Sintering Time and Temperatu re ................................ ................................ 30 3.3.2 Effect of Polyurethane System ................................ ................................ .................. 32 3.3.3 Effect of Composition ................................ ................................ ................................ 34 3.4 Conclusion ................................ ................................ ................................ ....................... 35 4 MECHANICAL CHARACTERIZATION ................................ ................................ .................. 44 4.1 Introduction ................................ ................................ ................................ ...................... 44 4.2 Experimental Procedure ................................ ................................ ................................ .. 45 4.3 Results and Discussion ................................ ................................ ................................ ... 46 4.3.1 Confined Compression Testing ................................ ................................ ................. 46 4.3.2 Model Fit ................................ ................................ ................................ ................... 50 4.4 Conclusion ................................ ................................ ................................ ....................... 52

PAGE 5

5 5 POLMYER INFILTRATION AND PRELIMINARY ELECTRICAL CHARACTERI ZATION ...... 61 5.1 Introduction ................................ ................................ ................................ ...................... 61 5.2 Experimental Procedure ................................ ................................ ................................ .. 61 5.2.1 Polyme r Infiltration ................................ ................................ ................................ .... 61 5.2.2 Electrical Characterization ................................ ................................ ......................... 62 5.3 Results and Discussion ................................ ................................ ................................ ... 63 5.3.1 Polymer Infiltration ................................ ................................ ................................ .... 63 5.3.2 Electrical Characterization ................................ ................................ ......................... 64 5.4 Conclusion ................................ ................................ ................................ ....................... 65 6 SUMMARY AND OUTLOOK ................................ ................................ ................................ 69 6.1 Summary ................................ ................................ ................................ ......................... 69 6.2 Outlook ................................ ................................ ................................ ............................ 72 APPENDIX A IMAGE ANALYSIS ................................ ................................ ................................ ................. 74 B RAMAN SPECTROSCOPY ................................ ................................ ................................ ... 77 REFERENCES ................................ ................................ ................................ ......................... 81 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ........ 84

PAGE 6

6 LIST OF TABLES Table p age 2 1 Piezoelectric coefficients of selected ceramics (in pC/N) ................................ ................. 21 2 2 Dielectric and piezoelectric properties of 3 3 composites 23 ................................ .............. 21 3 1 Image analysis grain data ................................ ................................ .............................. 37 3 2 Image analysis cell window data ................................ ................................ .................... 37 3 3 Mercury porosimetry data ................................ ................................ .............................. 37 3 4 Porosity measurements by helium pycnometer ................................ .............................. 37 4 1 Porosity and density values for BaTiO 3 foams ................................ ............................... 54 4 2 Mechanical properties for BaTiO 3 foam ................................ ................................ .......... 54 4 3 Properties of vari ous ceramic foams and commercial PZT ................................ ............. 54 5 1 Material properties of polymer considered for infiltration ................................ ................ 67 5 2 Piezoelectric coeffi cients of composite and bulk materials ................................ ............. 67 B 1 Raman spectra of (left) bulk barium titanate and (right) barium titanate at different temperatures. ................................ ................................ ................................ ................. 7 8 B 2 Raman s pectra comparison between barium titanate foam and pellet (top left) at 30 o C, (top right) at 100 o C and (bottom left) 150 o C and the difference in peak center (foam peak center – pellet peak center). ................................ ................................ ........................... 7 8

PAGE 7

7 LIST OF FIGURES Figure p age 2 1 Barium titanate unit cell with barium (red), oxygen (blue), and titanium (green) atoms .... 21 2 2 Three of the ten connectivity patterns used for composites by Newnham et al. 1 .............. 2 2 2 3 Comparison of HFOM for different ceramic polymer composites ................................ .... 2 2 2 4 Scanning electron microscopy image depicting component of the foam microstructure 2 3 2 5 Schematics of the three common foaming methods: (1) replica, (2) sacrificial phase and (3) direct foaming ................................ ................................ ................................ ........... 2 3 2 6 Comparison of microstructure in (a) SiOC foam, (b,c) Al 2 O 3 foam and (d) BaTiO 3 foam synthesized via direc t foaming method ................................ ................................ .......... 24 2 7 Chemical reaction of polyurethane system ................................ ................................ ...... 2 4 2 8 Synthesis steps for PU5 ................................ ................................ ................................ 25 3 1 Weaire Phelan s chematic ................................ ................................ .............................. 38 3 2 Direct fomaing schematic using barium titanate powder ................................ ................. 38 3 3 FE SEM images of PU5 sample synthesized with (A) Voranate M220 and (B) PAPI 27 both sintered at 1400 o C for 4 hours ................................ ................................ ............... 38 3 4 Processing parameters chart ................................ ................................ ......................... 39 3 5 FE SEM images of (A) BaTiO 3 foam synthesized using PU1 and (B) foam synthesized using PU5 ................................ ................................ ................................ ...................... 40 3 6 FE SEM micrographs of PU5 20 vol% samples sintered at differenct temperature for 4 hours in low magnification ................................ ................................ .............................. 40 3 7 FE SEM micrographs of PU5 20 vol% samples sintered at differenct temperature for 4 hours in low magn ification ................................ ................................ .............................. 41 3 8 FE SEM micrographs of PU5 30 vol% samples sintered at different times A) 4 hours, B) 8 hours and C) 12 hours ................................ ................................ ................................ ... 41 3 9 FE SEM images at 500X and 5000X of (A,B) PU1 (C,D ) PU2 (E,F) PU3 (G,H) PU4 and (I,J) PU5 foams ................................ ................................ ................................ .............. 42 3 10 FE SEM images of PU2 foam samples at (A) 30, (B) 40, (C) 50 and (D) 60 vol% ........... 43 3 11 FE SEM images of PU2 foam samples as (A) 20, (B) 3 0, (C) 40 and (D) 50 vol% .......... 43 4 1 Confinement cell and foam sample used for confined compression testing .................... 55

PAGE 8

8 4 2 Stress strain response of BaTiO 3 foam produced from LPU system ............................... 55 4 3 Stress strain response of BaTiO 3 foam produced from CPU system .............................. 56 4 4 Mechanical properties of BaTiO 3 foam as a function of density ................................ ...... 56 4 5 Stress strain curves comparing density and PU ................................ ............................. 57 4 6 Compression of LPU foam with respective stress strain curves ................................ ..... 57 4 7 Compression of CPU foam with respective stress strain curves ................................ ..... 58 4 8 Measured data and model fit for BaTiO 3 foam produced by LPU ................................ ... 58 4 9 Measured data and model fit for BaTiO 3 foam produced by CPU ................................ ... 59 4 10 Model parameters as a f unction of density ................................ ................................ ..... 59 4 11 Comparison of mechanical properties determined from the (M ) fited data and (E ) experimental data as a function of density ................................ ................................ ..... 60 4 12 Model estimated ultimate str ength of BaTiO 3 foams as a function of density ................. 60 5 1 Epovac system for polymer impregnation ................................ ................................ ...... 67 5 2 SEM images of the infiltrated composite: (light) BaTiO 3 (dark) Epofix ........................... 67 5 3 Capacitance and loss of composite as a funcation of frequency ................................ ..... 68 5 4 Permittivity (real and imaginary) of composite ................................ ................................ 68 5 5 Theoretical HFOM as a functio n of ceramic volume fraction ................................ ............ 68 A 1 Micrograph of PU2 foam with the grain boundaries digitally enhanced ......................... 7 5 A 2 Micrograph with enhanced contrast ................................ ................................ ............... 75 A 3 Micrographs with the threshold adjusted to measure (left) grain size and (right) cell window size ................................ ................................ ................................ ................... 7 6 B 1 Raman spectra of (left) bulk barium titanate and (right) barium titanate at dif ferent temperatures. ................................ ................................ ................................ ............... 79 B 2 Raman spectra comparison between barium titanate foam and pellet (top left) at 30 o C, (top right) at 100 o C and (bottom left) 150 o C and the difference in peak center (foam peak center – pell et peak center). ................................ ................................ ............................ 80

PAGE 9

9 Abstract of Thesis Presented to the Graduate School Of the University of Florida in Partial Fulfillment of the Requirem ents for the Degree of Master of Science SYNTHESIS AND CHARACTERIZATION OF ELECTROCERAMIC FOAMS FOR 3 3 PIEZOCO MPOSITES By Laurel Wucherer December 2008 Chair: Juan C. Nino Major: Materials Science and Engineering We investigated processing structure property performance interrelationships of barium titanate foam in 3 3 piezocomposites for sensor applications Th e first area of investigation focuses on the tailoring of the foam microstructure via processing parameters. B arium titanate foams were synthesized via direct foaming method using commercial powders. In order to control the microstructure and by extension the properties, synthesis parameters such as foaming agent, composition, sintering time, and sintering temperature were varied to determine their influence. Microstructure was classified in terms of strut characteristics, average grain size, average cell window size, porosity, and grain boundary integrity. The two systems that fabricated homogeneous structures favorable to ultrasonic sonar applications were a laboratory developed and a commercial polyurethane system, both silico n free. When sintered at 1 400 degrees C elsius for 8 hours they produced thick, dense struts and uniform cell window size distributions, though w ith different characteristics. The foam prepared with the laboratory synthesized polyurethane had a porosit y of 90 percent, an average gra in size of 44 microns and an a verage cell window size of 67 microns while the foam prepared with the commercial pol yurethane had a porosity of 87 percent, an average grain size of 20 microns and an a verage cell window size of 99 microns

PAGE 10

10 The second a rea of investigation was the resulting mechanical properties of the foams. They were measured using confined compression testing to determine if the foams had acceptable properties for their intended application and could withstand handling The mechanical properties (compressive modulus and collapse stress) determined from the experimental data increased with increasing density. Additionally, the lab developed PU foam showed a higher mechanical strength due to a higher average density. The data was then fi t with a phenomenological model. The focus of this work is on the elastic response due to the application requirements. The final area of investigation is the infiltration of the foam with polymer to create a 3 3 piezocomposite. Polymer selection was base d on a criterion of material properties including viscosity, cure time, cure temperature, and hardness. The samples were infiltrated with Epofix epoxy using Epovac infiltration equipment which is a vacuum impregnation system that infiltrates all empty spac e in the foam created by the cell window network with the epoxy. Using SEM, it was determined that the foam was completely infiltrated, creating a 3 3 composite Preliminary electrical characterization of the piezoelectric properties indicated that the hy d rostatic figure of merit of the composite was calculated to be lower than that of bulk barium titanate and bulk lead zirconate titanate. This is due to the low compliance of the polymer chosen for the composite and the lower piezoelectric properties of ba rium titanate in comparison to PZT and the low volume fraction of piezoelectric material in the composite Based on these results, the next step is to create a 3 3 composite with lead zirconate titanate to determine if the piezoelectric properties are furt her enhanced in this system. From the areas investigated above, it is concluded that the 3 3 piezoelectric composite is achievable.

PAGE 11

11 CHAPTER 1 INTRODUCTION 1.1 Statement of Problem and Motivation F oams, specifically ceramic based, offer unique and favorable properties but their use has been limited by the inability to predict and control the resulting microstructure during processing, especially using a direct foaming method since the process relies on the incorporation of gas bubbles to create the foam. By establishi ng process microstructure property interrelationships, the microstructure of the foams, and inherently the properties, can be tailored for many different applications. The purpose of this work is to explore the electroceramic properties of BaTiO 3 foam for electromechanical applications. The intent is to use BaTiO 3 as a test material and eventually, create piezoelectric composites based on materials such as PZT and piezoelectric polymers. This piezo composite could be used in sonar and ultrasound application by creating an open cell structure that has the potential to exhibit 3 3 connectivity by creating a three dimensional network of cell windows and ceramic struts Therefore the investigation of the synthesis of BaTiO 3 foams using the direct foaming method and ceramic powder precursor for engineering of piezocomposites is of significant interest. 1.2 Scientific Approach The literature review can be briefly summarized as follows: Newnham et al. 1 published a pioneerin g work in the area of piezoelectric polymer composites, which included the composite connectivity notation currently used and internationally accepted This notation is further discussed in Chapter 3. One of the most common types of composites fabricated d ue to its increased enhancements of piezoelectric property and successful processing is the 1 3 composite made from piezoelectric rods in a passive polymer matrix whose key engineering aspects for design have been identified through extensive experimental and theoretical work led primarily by Haun et al. 2 3 and by Smith et al. 4 5 6 7 However, Skinner et al. 8 was the first to

PAGE 12

12 report piezocomposites with 3 3 connectivity using a lost wax method with coral as the starting material. Other sacrificial methods were investigated by Shrout et al. 9 and Rittenmyer et al. 10 From this work, it was discovered that larger p iezoelectric coefficients were obtained using soft polymer as the second phase because its higher elastic compliance promoted stress transfer to the ceramics, which is highlighted in the hyd rostatic figure of merit, discussed in Chapter 2 In the field of ceramic foams, the first major break though was the patent by Schwartzwalter et al. 11 for polymer foam replication method 40 years ago. Since then, the research has expanded to incorporate many different proces sing methods, which have been reviewed by Gauckler et al 12 and many different ceramics including open cell and closed cell SiOC foams by Colombo et al. 13 cordierit e foams by Costa Oliviera et al. 14 and open cell Al 2 O 3 foams by Peng et al. 15 After careful review of the literature, it is clear that ceramic foams and piezocompos ites have been studied in the past but even though areas such as microstructure, processing and properties have been investigated individually, very little has been done to understand the interrelationships between these individual areas. Specifically elec troceramic foams have not been thoroughly investigated to realize their potential to fabricate 3 3 composites for piezoelectric sensors. Therefore, the scientific objective of this thesis is to fabricate foamed piezoelectric ceramic polymer composites with 3 3 connectivity and controlled microstructure, and to investigate the effect that a tailored ceramic mi crostructure has on the electromechanical properties of the composite. In order to accomplish this, further investigation in areas such as processing, microstructural optimization, mechanical property characterization, polymer infiltration and electrical property characterization is required. In processing for example, very little is known about the influence of variation in processing parameters on th e final microstructure, or the reproducibility of the structure using direct foaming method since gas incorporation is not easy to control. While there are previous

PAGE 13

13 publications on the design requirements of 1 3 composites, there is insufficient work on th e design requirements of 3 3 composites and for piezoelectric sonar sensors, specifically an optimal microstructure for both. Experiments regarding these issues will be conducted. Once an optimum foam microstructure is fabricated, it is necessary to evalu ate the reproducibility and standardization of the process ; therefore, multiple foamed samples from the same mixture batch will be fabricated and characterized to assess the sample to sample variation. Additionally, the mechanical properties of the foam wi ll be measured using confined compression testing and fitted using a phenomenological model established by Subhash et al. 16 specifically for the mechanical behavior of foams. After determining the ideal precurso rs and processing conditions for synthesizing the foams, polymer will be infiltrated into the ceramic foam skeleton to create a 3 3 composite. Special attention will be given to polymer selection as it will determine the magnitude of piezoelectric property enhancement. A common vacuum and pressure assisted technique will be employed to achieve the complete infiltration of the ceramic foam. 17 18 Once a 3 3 composite has been achieved, the electrical properties will be measured including dielectric, ferroelectric and piezoelectric characterization A hydrostatic figure of merit (HFOM) of the piezocomposite wil l be determined and compared to the HFOM of other piezocomposit es and piezoelectric materials

PAGE 14

14 CHAPTER 2 BACKGROUND The present chapter presents a brief summary of some of the theoretical background required for understanding and analyzing the investigation presented in the subsequent chapters. As a summary, it is not intended to cover the background of the research areas in their entirety, but rather to present a work frame for the rest of the thesis. 2.1 Piezoelectricity Piezoelectricity can be defined as the linear coupling between mechanical stress, X and electric po larization, P (referred to as direct e ffect) or as the coupling of mechanical strain, x and applied electric field, E (converse effect). Piezoelectricity is a third rank tensor property since it relates the polarization vector, P to stress, X a second r ank tensor. Tensor notation set aside, the principal piezoelectric coefficient, d describes the direct effect as well as the converse effect. This implies that the units of d are [C/N] or [m/V]. Typical values for useful piezoelectric materials range from quartz) to ~ 800 pC/N (lead zirconate titanate, PZT ceramics). The converse piezoelectric effect of material is useful for actuation applications, while the direct piezoelectric effect can be exploited in sensing applications, directly related to this work. Naturally occurring piezoelectric materials include minerals such as tourmaline and quartz. Quartz is widely used in frequency control and time keeping applications. In the area of polycrystalline materials the most important family of piezo electrics is that of perovskite based (e.g. BaTiO 3 PZT, etc) ceramics. An abbreviated overview of pie zoelectric ceramics is discussed in the following section. 2.2 Piezoelectric Ceramics For over 50 years since the report on the piezoelectric properties of BaTiO 3 and PZT, by Roberts 19 and Jaffe et al. 20 respectively, perovskites have been at the forefront of transducer technology. To date, barium titanate ceramics are used in applications as diverse as underwater sonar, biomedical ultrasound, multilayer actuators for fuel injection, bimorph pneumatic valves

PAGE 15

15 and picoliter pumps for inkjet printers. BaTiO 3 ( Figure 2 1 ) crystallize s with the perovskite structure. The larger Ba atoms (red) occupy the corners of the unit cell and establish a twelve fold coordination with oxygen (blue) which occupies the face centers. Together they form a cubic close packed arrangement with a typical lattice parameter around 4 21 Titanium ions (green) occupy the center of the unit cell and establish an octahedral coordination with oxygen. The high temperature form of BaTiO 3 is cubic (space group Pm 3 m) an d on cooling a phase transformation occurs in which the atoms are displaced to a tetragonal phase (space group P4mm) BaTiO 3 exhibits phase transformations which cause ferroelectric and paraelectric behavior depending on the temperature and structure. It h as unique electronic properties including a high dielectric constant ( 140 0), piezoelectricity (d 33 = 190 pCN 1 ), ferroelectricity (T c = 130 C), and pyroelectricity (p = 20 nC/cm 2 K) In addition, it exhibits good mechanical strength and a high melting t emperature (T m = 1625 C) 4 Typical piezoelectric coefficients for other polycr ystalline ceramics are shown in Table 2 1 2.3 Ceramic Polymer Piezoelectric Composites In general, the aim of composite technology is to combine materials in such a way that the composite exhibits optimum properties and performance far exceeding the behavior of each of its indivi dual members. Based on the work by Newnham et al. 1 ( C hapter 1) using a simple cube model, two phase composites were labeled by a two number notation, the first defining how the ceramic phase is connected in space and the second number indicating how the polymer is connected. There are ten con nectivity patterns in a two phase composite, ranging from a 0 0 unconnected checkerboard pattern to a 3 3 pattern in which both phases are three dimensionally self connected. Some of the images from the now legendary figure from Newnham’s work are reproduc ed i n Figure 2 2 From these configurations, by far the most widely used for sonar, medical imaging and non destructive testing applications is the 1 3 type composites made from piezoelectric rods in a passive poly mer matrix.

PAGE 16

16 Of particular interest and direct relevance to this work, however, is the previous work on 3 3 piezoelectric composites. Expanding on the literature review in Chapter 1, Shrout et al. 8 developed a si mplified fabrication process involving the burn out of polymer spheres referred to as BURPS (BURned out Plastic Spheres). Using this method with polymethyl methacrylate as the sacrificial polymer, Rittenmyer et al. 10 sintered porous PZT skeletons and later impregnated them with a stiff vinylcychlohexane dioxide epoxy and a soft silicone elastomer. It was found that the larger piezoelectric coefficients were obtained for the softer polymer because its higher elast ic compliance promotes stress transfer to the ceramic. It was found that the hydrostatic figure of merit (HFOM) for hydrophone applications d h g h was more than a hundred times than that of PZT ceramics. HFOM is defined as HFOM = d h g h ( 2 1 ) d h = d 33 + 2d 31 ( 2 2 ) g h = d h 33 T ( 2 3 ) Figure 2 3 shows the difference in HFOM between different piezocomposites reviewed by Cross 22 Besides th e hydrostatic sensitivity, additional advantages of these composites include low dielectric constant, high compliance, mechanical flexibility and low density which will enable improved matching of acoustic impedance to water (1.569 MRayls) 4 A summary of the dielectric and piezoelectric properties of 3 3 type composites is shown in Table 2 2 after Gururaja et al. 23 It has to be said that the figure of merit of these composites, although high, are at least an order of magnitude below other composite configurations such as diced PZT polymer composite (20,000 x 10 15 m 2 /N); however this process is complicated and relies heavily on the minimizatio increase the figure of merit while maintaining a simple processing method is desired. The 3 3 piezocomposite is engineered by infiltrating the cell window network completely with polymer so that each component has 3 dimensional connectivity. Composites with 3 3 connectivity are desired because they offer a higher piezoelectric enhancement

PAGE 17

17 2.4 Ceramic Foams In the past, metallic and polymeric foams have commonly been used in industry due to high mech anical strength and ductility, respectively. However, porous ceramics have become more interesting in an industrial sense because they have low density, low thermal conductivity, high specific strength, high thermal shock resistance, and high surface area. 24 Depending on their structure, c eramic foams can be used in corrosive filtration systems thermal insulation, acoustic receiving systems, scaffolds for bone tissue engineering and solid oxide fuel ce lls 25 The growing scientific and industrial interest is the result of the properties of this material form such as low thermal conductivity, low dielectric constant, low density, high permeability, etc. In addition, the incorporation of fillers allows for the fabrication of structures with tailored thermal, electrical, magnetic, and mechanical properties. 24 Technically speaking, ceramic foams are a specific class of porous mat erials that consists of a three dimensional array of polyhedral cells and a total porosity greater than 45 vol% 26 The cells can be surrounded by ceramic walls and struts (defined in next section), a 3 D interco nnected structure is formed with open porosity and the material is referred to as open cell foam. As such, open cell foams can be considered as 3 3 composites between ceramic and air, and as the backbone for 3 3 ceramic composites in general. They are, the refore, a key component of this work. Remarkably, despite the enormous success and recent progress in ceramic foam science and technology, there has been no systematic study regarding the control of the ceramic microstructure. In general, research that has dealt with structural control has focused on the control of the cell and strut configuration, or the foam mesostructure. While a controlled mesostructure is essential to the mechanical integrity and connectivity of the foam, the microstructure (i.e. grain size, grain boundary configuration, etc) plays an essential role in the general properties of the ceramic foam. This is especially important in electroceramics such as piezoelectrics and ferroelectrics where grain size and grain boundary configuration dir ec tly

PAGE 18

18 affect among others, the dielectric permittivity and domain wall mobility, which in turn affects poling and the overall piezoelectric response of the material. Before continuing it is imperative to clarify the terminology used throughout this thesi s to describe the micro and meso structure of the foam. T here are three major components ( Figure 2 4 ) : the ceramic strut, grain and cell window. The ceramic strut consists of a number of gr ains and is the in tercon nected ceramic network that will eventually be the ceramic phase of the composite. The cell windows are the interconnected cell window phase, and will eventually be the polymeric phase of the composite. Foam structures with this interconnected cell window network, are considered open cell foams. Pores in foams, refer to the holes in the struts either in the grains or at the grain boundaries. Cells refer to the hemispherical voids consisting of cell windows and cell walls, which are struts within the cell. T he mesostructure is classified in terms of strut, cell walls, cells and cell windows. The microstructure is classified in terms of grain size, grain boundary configuration and porosity. Both must be understood to control the overall structure of the foam. The shape and size of these features are determined by processing parameters. This will be discussed in the next section and more thoroughly in Chapter 3. 2.5 In Situ Foam Processing There are three common foaming methods ( Figure 2 5 ) : replica, sacrifi cial template and direct foaming Replica method (1) is the impregnation of a cellular structure with a ceramic suspension or precursor solution which forms to the morphology of the original cellular structure. While i t’s a convenient method, it produces low strut stability and limited cell window size due to the limitations of the template. Sacrificial template method (2) is the preparation of a composite consisting of a continuous ceramic matrix and homogeneously disp ersed particles which are extracted to produce pores. These foams have a wide range of porosities and higher mechanical strength then that of the replica method. Di rect foaming method (3) relies on the stabilization and setting of wet foams (either by surf actant or particles) produced by directly

PAGE 19

19 incorporating air bubbles into a ceramic suspension using a foaming agent such as a polyurethane system. This method eliminates the need to extract a sacrificial phase before sintering and produces the highest mec hanical strength of the three methods while still maintaining a wide range of pore sizes and porosities 12 Expanding on these three methods, a wide range of processes for the production of ceramic foams have be en demonstrated and have been recently reviewed. Perhaps the most widely used method is the sponge replica technique in which a polymer sponge is impregnated with ceramic slurry and followed by burn out of the polymer and high temperature sintering of the remaining ceramic. Bao et al. 27 have demonstrated a variation of this technique in which a polymer precursor for the ceramic, or preceramic, is used. Recently direct foaming of sol gel ceramic precursors without the use of sacrificial templates has been reported. These processing techniques have been used to synthesize several ceramic foams including alumina 28 SiOC 13 ( Figure 2 6 a) and PZT 14 similar structure is seen for the Al 2 O 3 foam by Colombo and Hellmann in Figure 2 6 c Further, the detail of the strut in the Al 2 O 3 ceramic foam indicates a hollow internal structure ( Figure 2 6 b). In contrast, the direct foaming process typically leads to open cell structures with small cell windows as seen in Figure 2 6 d. This work utilizes a direct foaming technique that Traversa and collaborators 29 employed and de veloped to successfully fabricate NiO YSZ foams for anodes in solid oxide fuel cells. The technique is based on the in situ polymerization of a polyurethane system (PU) loaded with ceramic powder with the addition of surfactants, followed by a precisely co ntrolled firing schedule. Due to the instable nature of the vapor liquid interface created during the foaming process, the polyurethane systems are usually created using particles or surfactants to stabilize this interface. However, most commercial systems are surfactant stabilized systems. The ceramic powder is dispersed separately in the two components of a commercial polyurethane system. The two components are mixed to induce polymerization and the mix is placed in an

PAGE 20

20 open Teflon mold. The chemical r eact ion of the PU is shown in Figure 2 7 The reaction produces carbon dioxide which form bubbles within the structure and will eventually produce cells or cell windows after sintering. Figure 2 8 shows a summary of the processing steps for this technique using a PU developed by Traversa et al. 29 at the University of Rome – Tor Vergata. The PU consists of methyl diphenyl diisocyanate (MDI) and po lyethylene glycol (PEG) / 1, 4 Diazabicyclo (2.2.2)octane (1,4 Dabco) which constitute the PU Tween 80 is added as a pore minimizer and water acts as a catalyst. In addition, a sim ilar method is employed using commercially purchased PU systems rather tha n lab developed as the once discussed above. The difference between the two systems is the type of diisocyanate and polyol used which determine the resulting microstructure.

PAGE 21

21 Figure 2 1 Barium titanate unit cell with barium (red), oxygen (blue) and titanium (green) atoms. Table 2 1 Piezoelectric coefficients of s elected ceramics (in pC/N) Polycrystalline Ceramics d 33 d 31 d 13 d 32 d 24 BaTiO 3 4 190 78 260 PZT 5H 41 593 274 741 PZT 8 41 225 27 330 PbNb 2 O 6 60 43 180 24 170 Na 0.5 K 0.5 NbO 3 127 51 306 Table 2 2 Dielectric and piezoelectric properties of 3 3 composites 23 Composites (kg/m 3 ) K 33 d 33 (pC/N) g 33 (mVm/N) g h (mVm/N) d h g h (10 15 m 2 /N) Coral Replamine 3300 50 100 225 140 5040 PZT Spurrs epoxy 4500 620 150 28 20 2200 PZT silicone rubber 4000 450 200 50 45 8100 Porous PZT 3840 200 190 130 50 4500 Barium titanate 6020 1 400 190 2.7 93

PAGE 22

22 Figure 2 2 Three of the ten connectivity patterns commonly used for com posites by Newnham et al. 1 Figure 2 3 Comparison of HFOM for different ceramic polymer composites

PAGE 23

23 Figure 2 4 Scanning election microscopy image depicting components of the foam microstructure. Figure 2 5 Schematics of the three common foaming methods : (1) replica, (2) sacrificial phase and (3) direct foaming.

PAGE 24

24 Figure 2 6 Com parison of microstructure in (a) SiOC foam, (b,c) Al 2 O 3 foam and d) BaTiO 3 foam synthesized via direct foaming method. Figure 2 7 Chemical reaction of polyurethane system

PAGE 25

25 Figure 2 8 Synt hesis steps for PU5.

PAGE 26

26 CHAPTER 3 FOAM SYNTHESIS AND MICROSTRUCTURAL ANALYSIS 3.1 Introduction Foams can be classified by the s tructure and size of their cell windows. Cell windows in closed cell foams are not interconnected and contribute to their higher compr essive strength. Open cell f oams contain interconnected cell window networks and are commonly used in composite materials. The Weaire Phelan structure ( Figure 3 1 ) is theorized to be the ideal unit cell of a perfe ctly ordered foam, consisting of fitted polyhedra with both hexagonal and pentagonal faces which minimize surface area and maximize packing. 30 There are three classifications of pore size: micro meso and mac roporosity (less than 2 nm, between 2 nm and 50 nm, and greater than 50 nm, respectively). 32 The structure of the foam determines its properties and potential applications. It is characterized in terms of strut characteristics, cell p arameters and pore parameters. Cell parameters include grain size and grain boundary integrity of the grain which creates the cell. Pore parameters include cell window size, cell wind ow distribution, and porosity. Strut characterist ics are classified by relative density and thickness. It is difficult to classify one “ideal” microstructure for all applications because different applications require different properties and therefore different structures. For example, from a mechanica l point of view a smaller grain size and more grain boundaries can lead to increased strength similar to the Hall Petch effect seen in metals due to dislocation pinning at the grain boundaries. 33 Additionally, lower porosities are preferred to increase compressive strength and the elastic modulus. 34 However, for piezoelectric composites, an increase in porosity would lead to an increase in volume fraction of polymer which would enhance the piezoelectric properties of the composite by transferring stress to the ceramic phase. Grain boundary integrity, dense and sturdy struts without sharp edges and a uniform cell window distribution are desired for both applications. 24 26 By adjusting the synthesis parameters, the microstructure can be controlled to produce various properties for numerous applications.

PAGE 27

27 These synthesis parameters in clude foaming method, composition, sintering time, and sintering temperature. The compound BaTiO 3 (Chapter 1) has interesting properties including piezoelectric properties. As a foam, BaTiO 3 can demonstrate a wider range of properties for additional appl ications, but has not been investigated in detail like other ceramics. For example Colombo et al. synthesized flexible and semi rigid silicon oxycarbide foams from preceramic polymer and a polyurethane system and sintered them at 1200 C 25 A silicon free commercial polyurethane system and novel laboratory polyurethane system (Chapter 2) have been tailored for barium titanate foam synthesis via direct foaming method. The effects of the synthesis parameters on the resulting mi crostructures are presented in this chapter 3.2 Experimental Procedure The direct foaming method used a liquid, two component polyurethane system. The fabrication of a polyurethane system (PU) was based on the reaction between a diisocyanate and a polyol ( Figure 3 2 ) which occurs in the presence of a catalyst. To develop a foam structure, water was added as a foaming agent to react with the isocyanate group during the polymerization to produce carbon dioxide, which expands the polymer matrix producing a cellular structure. To obtain a ceramic loaded polymeric foam, a ceramic powder was dispersed in the liquid reagents prior to polymerization. BaTiO 3 powder (Aldrich US ) with an average grain size of 2 m and a purity level of 99.99% was weighed to produce specific composition s of 20 vol%, 30 vol%, 40 vol%, 50 vol% or 60 vol%. The same powder was used to test five different polyurethane systems. The five polyurethane systems tested were :

PAGE 28

28 1. PU1 Rigid commercial system (Prochima, Italy) 2. PU2 Rigid silicon free commercial system (Smooth on, NJ, US) 3. PU3 Flexible silicon free commercial system (Smooth on, NJ, US) 4. PU4 Rigid diisocyanate concentrated commercial system (Cytec, US) 5. PU5 Rigid silicon fre e laboratory system (University of Rome Tor Vergata, Italy) The commercial PU systems were combined in ratios of 1:1, polyol to diisocyanate. PU5 samples were prepared using a laboratory developed polyurethane system as described in C hapter 2 29 Due to chemical importing restrictions, the original polymeric MDI used in the laboratory grade system, Voranate M220, could not be purchased in the United States. Instead, its US equivalent chemical, PAPI 27, had to be u sed to continue the experiments. Figure 3 3 shows the field emission scanning electron microscopy ( FE SEM ) micrographs of the foams obtained with the two chemicals. No major differences in microstructure were obser ved. It was concluded that the chemicals were equivalent for this experiment. The foams were stirred until highly viscous (~ 2 minutes after combining) using an electric screwdriver and stirring attachment and set aside to harden and cure. When dry, t he f oams were cut into 2 cm 3 rectangular samples The newly cut green sample s were placed in a 1700 C horizontal tube furnace (CM Furnaces, Bloomfield US) designed for air flow to prevent sample reduction and furnace corrosion caused by the burn out of the polyurethane system. The samples were heated at 60 C/hour to 600 C to prevent damage to the struts and held for 1 hour to completely remove the polyurethane from the foam. Once the polyurethane is removed, the system is completely ceramic. It was then he ated at 1 60 C/hour to a specific sintering temperature and held for either 4, 8 or 15 hours. The temperature was lowered at 140 C /hour to room temperature. All samples were fired in air with the flow rate maintained at 10 cm 3 /minute. A flow chart is show n in Figure 3 4 to organize all of the potential variables. Depending on the type of structure that is

PAGE 29

29 produced from each of these variables, some were eliminated, which simplified the experiments. This will be dis cussed in the next section. FE SEM was used to determine the effect of synthesis parameters on relative grain and cell window sizes and relative grain and cell window size distributions. Four images of different magnification (100X, 500X, 1000X, and 5000X ) were taken of each sample at different locations of the sample. Three samples of each PU system were characterized to determine how repeatable the process was This was a very important part of the project to determine the interrelationship between proce ssing, structure and properties of the foams. If the structure was not repeatable, these relationships would be insignificant. Optical analysis was performed using Image J software on the resulting SEM images Parameters determined from this technique were average cell window area an d average grain area. T hree images for each variable configuration (i.e. PU2, 30 vol% 1400 o C for 8 hours) at 1 000X magnification were used in the analysis. In addition images were taken and two different depths into the sample to ensure a homogeneous structure. Each of these depths was analyzed as well. T he grain boundaries and cell windows were outlined in black using an image editing software, while the grains remained light gray. The threshold of the image was then adjusted to highlight the dark regions or lights regions based on extreme contrast. For example, if the grains were being analyzed, the threshold settings would make the grains red and the grain boundaries and cell windows white. The macro, “Analyze Particl es”, wa s used to scan the image, line by line, counting red as “1” and white as “0” and starting a recount every time a white pixel was encountered The program then sums the count and presents it in a table with statistics including average and standard deviatio n. The resulting number of pixels per red region was then converted to microns per grain using the image magnification to determine an average grain area or cell window area. From these values an average cell window and grain size were estimated assuming t hat the grains and the cell windows were circular (size =

PAGE 30

30 1/2 ) Using this technique an estimate of the grain and cell window size s was determined. Cell size was not calculated since the resulting microstructures did not produce cells. More descriptio n on this analysis is given in Appendix A The values of the cell window sizes were more accurately measured using a helium pycnometer (Quantachrome Ultrapyc 1000 Gas Pycnometer) and mercury porosimetry (MP) Helium pycnometry measures the true density and calculates the open porosity which is important to create the 3 3 piezoelectric composite. MP characterizes the pore size distribution taking into account both the pores in the grains and struts and the cell windows. Raman spectroscopy was performed to de termine if residual stress was present in the foam structure, compared to bulk barium titanate. The results are shown in Appendix B. All of the results from the characterization techniques were compiled to determine the ultimate processing parameters that would produce and open structure with dense struts and no sharp wall edges. 3.3 Results and Discussion Figure 3 5 shows the difference in microstructure between PU1 and PU5 foams. The presence of Si rich contaminants ( Figure 3 5 A ) created platelets at the grain boundaries. These contaminants could be detrimental for mechanical properties because it decreases the mechanical strength of the grain boundaries and electrical applica tions by decreasing the charge mobility at the grain boundary Therefore, the Si free PU5 system was successfully developed to avoid contaminants ( Figure 3 5 B ) and Si free commercial systems were purchased In orde r to understand the process structure relationship, each synthesis parameter was examined to determine its influence on the foam microstructure. 3.3.1 Effect of Sintering Time and Temperature The microstructure is influenced significantly more by sintering te mperature than by sintering time (4, 8 and 12 hours) Experiments varying in sintering temperature were optimized for PU5 laboratory grade (using Voranate M220), 20 vol% samples sintered at 4 hours. The sintering

PAGE 31

31 temperatures tested were 1250, 1300, 1350 a nd 1400 C. High and low magnification SEM images of this system are in Figure 3 6 and Figure 3 7 respectively A w ell defined open cell microstructure is maintained at lowe r sintering t emperatures (1250 and 1300 C) while at higher sintering temperatures the cell window s sometimes lack spherical geometry which is also controlled by sintering time and composition. The cell window size and porosity decreased with increasing temperature. H igh temperature samples are more robust during handling than low temperature samples. This is because the low temperature samples do not have well sintered structures which weaken the mechanical resistance. Because of its more uniform cell window and grain size distributions, handling capabilities and well sintered structure, a sintering temperature of 1400 C was used for all other experiments involving different synthesis parameter variations. The foams synthesized via direct foaming method with PU2 and PU5 follow conventional mechanisms regarding sintering and foam formation. The stages of sintering (initial, intermediate and final) can also be seen in Figure 3 7 At lower temperature (initial), there are no neck formations, but the contact area between particles increases. As the temperature increases (intermediate), continuous cell window and cell channels form and necks increase. Finally at 1400 o C, smaller cell windows between the grains are eliminated and grains are consolid ated. Note the well distributed grain sizes in the 1400 o C samples in Figure 3 6 D and Figure 3 7 D, respectively. The microstructure exhibits normal grain growth as the grain size is prop ortional to the square root of the sintering time As expected, there is a difference in grain size between samples sintered for 4 hours and samples sintered for 8 hours. At 4 hours, the grain size is 25 to 75 % smaller than at 8 hours. There is less unsin tered material at higher sintering times. However, there is little difference in microstructure between 8 hours and 15 hours except in a slightly increased grain size 5 to 10 % A sintering time of 8 hours for all foams was primarily

PAGE 32

32 chosen because it pr oduced a homogeneous microstructure in less than 2 days of total processing time. The differences in microstructure are shown in Figure 3 8 for PU5 foams. 3.3.2 Effect of Polyurethane System Of all the processing para meters, the choice of polyurethane system had the most significant impact on the resulting microstructure of the foam. It determined the configurations and connectivity of the foam while the other parameters such as composition and sintering temperature af fect the size and shape of the grains, struts and cell windows. Because each PU was compositionally different, so were the foams. PU1 ( Figure 3 9 A and B ), PU3 ( Figure 3 9 E and F ) and PU4 ( Figure 3 9 G and H ) foams all had cells with some closed porosity (> 5%). PU2 ( Figure 3 9 C and D) and PU5 ( Figure 3 9 I and J) foams created very open structures (closed porosity < 5%) which prevented them from creating cells; therefore the structures only consisted of struts and windows. PU1, PU2 and PU5 all exhibited a unique structure characteristic which differs fr om the microstructure of foams After preliminary research, some of the PU systems were eliminated because they did not pr ovide homogeneous foam structures in terms of uniform cell, grain and cell window sizes and uncontaminated grain boundaries PU1 foams contained rod contaminants at the grain boundaries thin struts, and did not produce uniform grain size of cell window si ze distributions. PU3 foams while very strong comparatively to the other foams, produced low porosity and closed cell structures which were not conducive to 3 3 composites. PU4 foams demonstrated abnormal grain growth with a contrast between very large gr ains (>20 ) surrounded by hundreds of very small grains (<2 ) and sharp cell wall edges which contributed to their sensitivity during hand l ing. This is due to the increased stress concentrations created at sharp and small areas such as the wall edges. Thick, dense and rounded struts prevent these stress concentrations, making the foams easier to handle. The final PU systems investigated were

PAGE 33

33 PU2 and PU5 foams The remaining characterization techniques including image analysis, helium pycnometry and MP w ere used to only characterize these final PU systems. Although both maintain shape during handling and produce homogeneous microstructures, PU2 and PU5 foams have considerably different microstructures. PU2 foams ha ve smaller grain sizes (20 – er and more interconnected cell windows (87 – consisting of many smaller grains ( Figure 3 9 C and D ) all of which increa se its resistance to damage during handling and mechanical strength PU5 fo ams have larger grain sizes (35 – – struts consisting of a few large grains ( Figure 3 9 I and J) Due to the smaller cell window size and larger grain size, PU5 foams are 3% denser and therefore can be handled easily as well. Image analysis data and MP data reflect these trends in grain cell window and pore size shown in Table 3 1 and Table 3 2 The minimum values report ed for the MP data in Table 3 3 reflect the pore sizes in the strut and the maximum values reflect the cell window sizes. Additionally PU2 exhibits well defined grain boundaries and hexagonal grain terraces, indicating growth in the <111> direction PU5 samples have smoother and more abnormally shaped grains. The terraces on the faces of the grains could possibly indicate over sintering and could potentially decrease mechanical strength. The well defined grain boundaries are prefe rred for higher mechanical strength because they indicate a that the foams have reached the final stage of sintering since the atoms have had more time to diffuse away from grain boundaries, producing deeper and well defined grain boundaries. However, the smooth, abnormally shaped grains could potentially decrease the stability for the foam. It is proposed that the difference in microstructure between the foams is derived from the difference in chemical composition of the PU systems The type of diisocyanat e, polyol and the inclusion of pore minimizers or catalysts can all affect the microstructure. There are many different types of diisocyanates and polyols which can be used to create polyurethane. For example, m ethyl diphenyl diisocyanate (MDI), used in P U5, is a symmetric diisocyanate which produces very rigid foams. It reacts very quickly

PAGE 34

34 which limits the number and size of bubbles introduced into the foam. Pore minimizers, such as Tween 80, and catalysts also have the same effect. The inclusion of an M DI and Tween 80 make PU5 foams denser that PU2. Unfortunately, the exact chemical composition of most commercially available PU systems and their components are not publicly available. In the work by Colombo et al. 13 SiOC foams were created using preceramic polymers and a PU similar to PU5. Flexible, semi rigid and rigid foams were synthesized and characterized. These foams had cellular structures similar to the flexible foams made with PU3, but exhibited incre ased porosity and homogeneity. In comparison to PU2 and PU5 foams, the SiOC foams have interconnected cell structure and more closed cells while the foams focused on in this paper have an interconnected cell window structure and open cells. While the strut s are denser in the SiOC foams due to the formation of cell walls, they also have very sharp edges, which could lead to a decrease in mechanical strength. Both the SiOC and BaTiO 3 foams demonstrate macroporosity. 3.3.3 Effect of Composition In both types of fo ams, the most homogeneous microstructures in terms of cells, grains and cell windows were found at 30 vol% of barium titanate. Below or above this value, the integrity, defined as microstructure homogeneity of the foams, decreased. Increasing ceramic conce ntration resulted in increased abnormal grain growth, increased amount of unsintered 2 ). These sa mples usually result in an average grain area that does not reflect the actual microstructure since it includes a lar ge range of grain sizes, from less than 3 suggested that the difference in grain size caused by the difference in composition is due to how the powder is dispersed into the PU. At l ower compositions, the powder can be dispersed homogeneously with more room to grow at the same rate during sintering. However, if the composition is too low, the material does not sinter as thoroughly because contact between particles is limited and at hi gher compositions, oversaturated PU doesn’t allow the particles to

PAGE 35

35 sinter and grow at constant rates. Porosity is influenced by ceramic content. An increase in ceramic content results in a decrease in porosity as seen in Table 3 3 and Table 3 4 There is a more significant decrease in cell window size and increase in grain size with increasing BaTiO 3 content in PU2 than in PU5 shown in both the SEM images in Figure 3 10 and Figure 3 11 and the image analysis data in Table 3 1 and Table 3 2 Additionally, uniform cell window size distributions and open structures are more c ommonly produced at 30 vol% than any other composition. Moreover, the PU mixture becomes inhomogeneous and oversaturated at higher BaTiO 3 concentrations ( Figure 3 11 C and D) whic h cause unsintered particles to cont aminate the grain boundaries. However, increased ceramic concentration produces stronger and less brittle foams. At 20 vol%, PU5 foams are very brittle and break during handling. Comparatively, at 50 vol%, PU2 foams are very durable and can withstand exten sive handling. A ceramic content of 30 vol% give a good compromise between these two properties. Physically, PU2 and PU5 foams behave similarly when handled. 3.4 Conclusion A surfactant stabilized direct foaming method was used to create BaTiO 3 based cerami c foams. The composition, polyurethane system sintering temperature and sintering time were varied to determine their influence on the microstructure of the foam and ultimately the properties and applications of the foam After determining that Si based P U systems produced contaminants at the grain boundary and inhomogeneous grain size distributions and flexible PU systems produced closed cell structures, research was focused on Si free, rigid PU systems: one commercially available and the other laboratory developed. After additional initial testing, some of the other variables were eliminated. The primary synthesis variables were polyurethane system (PU2 – commercial and PU5 – laboratory) and composition (20 – 60 vol% ceramic). Both the rigid commercial a nd rigid laboratory systems produced foams with uniform cell window size distributions and dense, thick struts. However, the commercial system’s struts consisted of many small grains which should have better mechanical properties while the

PAGE 36

36 laboratory syst em struts had fewer but larger grains. The rigid laboratory developed system demonstrated uniform cell window distr ibution, variable porosity (75 87 %), dense struts, contaminant free grain boundaries at concentrations less that 50 vol% and mechanical in tegrity. The commercial silicon free system also had a uniform cell window size distribution, smaller grain size, larger cell window size and also demonstrated dense struts and variable porosity (82 – 93 %). Lower ceramic concentrations (20, 30 and 40 vol% ) are ideal for electrical applications because of the interconnected cell window network and increased porosity. However, they are more brittle than foams with higher ceramic concentrations (50 and 60 vol% ) which are more favored in mechanical application s. Foams with 30 vol% ceramic demonstrated a homogeneous microstructure with a balance of properties for both mechanical and electrical applications. They have high porosity (~ 88 %), uniform and smaller grain size, are relatively strong and not easily dam aged. A 30 vol% foam, synthesized using the laboratory developed rigid polyurethane system and a 30 vol% commercial silicon free system, sintered for 8 hours at 1400 C, produced different microstructures which have the potential to be used in electromech anical applications.

PAGE 37

37 Table 3 1 Image analysis grain data PU Composition ( vol%) Average Grain Area 2 ) Standard Deviation 2 ) Average Grain PU2 30 1244 100 20 PU2 40 1563 121 22 PU2 50 1662 366 23 P U2 60 2290 298 27 PU5 20 3849 384 35 PU5 30 7178 285 44 PU5 40 6165 327 48 PU5 50 8495 327 52 Table 3 2 Image analysis cell window data PU Composition (vol%) Average Grain Area 2 ) Standard Deviation 2 ) Average Grain PU2 30 1244 100 20 PU2 40 1563 121 22 PU2 50 1662 366 23 PU2 60 2290 298 27 PU5 20 3849 384 35 PU5 30 7178 285 44 PU5 40 6165 327 48 PU5 50 8495 327 52 Table 3 3 Mercury porosimetry data Sample Mi Porosity (%) PU2 30 8 4.9 111 92 PU2 40 8 4.7 97 86 PU5 20 8 4.6 80 93 PU5 30 8 4.3 69 88 Table 3 4 Porosity measurements by helium pycnometer Polyurethane System Composition (vol%) Poros ity (%) PU2 30 93 PU2 40 90 PU2 50 88 PU2 60 82 PU5 20 87 PU5 30 85 PU5 40 83 PU5 50 75

PAGE 38

38 Figure 3 1 Weaire Phelan schematic. Figure 3 2 Direct foaming schematic using barium titanate powder Figure 3 3 FE SEM images of PU5 sample synthesize with (A) Voranate M220 and (B) PAPI 27 both sintered at 1400 o C for 4 hours Voranate PAPI A Mag = 500 X Voranate 20 m B Mag = 500 X Papi 20 m

PAGE 39

39 Figure 3 4 Proce ssing parameters chart

PAGE 40

40 Figure 3 5 FE SEM images of (A) BaTiO 3 foam synthesized using PU 1 and (B) foam synthesized using PU 5. Figure 3 6 FE SEM micrographs of PU5 20 vol% samples sintered at different temperatur es for 4 hours in low magnification. A Mag = 5000 X Commerical PU 2 m B Mag = 5000 X Research PU 2 m A Mag = 5000 X Commerical PU 2 m B Mag = 5000 X Research PU 2 m PAPI A Mag = 500 X 1250 o C 20 m C Mag = 500 X 1350 o C 20 m B Mag = 500 X 1300 o C 2 0 m B Mag = 500 X 1300 o C 2 0 m D Mag = 500 X 1400 o C 2 0 m

PAGE 41

41 Figure 3 7 FE SEM micrographs of PU5 20 vol% samples sintered at different temperatures for 4 hours in high magnification. Figure 3 8 FE SEM micrograph s of PU5 30 vol% samples sintered at different times A) 4 hours, B) 8 hours and C) 12 hours PAPI A Mag = 500 0 X 1250 o C 2 m C Mag = 5 0 00 X 1350 o C 2 m B Mag = 5000 X 1300 o C 2 m B Mag = 5000 X 1300 o C 2 m D Mag = 5000 X 1400 o C 2 m

PAGE 42

42 Figure 3 9 FE SEM images at 500X and 5000X of (A,B) PU1 (C,D) PU2 (E,F) PU3 (G,H) PU4 and (I,J) PU5 foams

PAGE 43

43 Figure 3 10 FE SEM images of PU2 foam samples at (A) 30, (B) 40, (C) 50 and (D) 60 vol% sintered at 1400 C for 8 hours. Figure 3 11 SEM images of PU5 foam samples at (A) 20, (B) 30, (C) 40 and (D) 50 vol% sin tered at 1400 C for 8 hours PAPI A Mag = 10 00 X 30 vol % 1 0 m C Mag = 10 00 X 50 vol % 1 0 m B Mag = 1 000 X 40 vol % 10 m B Mag = 1 000 X 40 vol % 10 m D Mag = 1 000 X 60 vol % 10 m PAPI A Mag = 10 00 X 20 vol % 1 0 m C Mag = 10 00 X 40 vol % 1 0 m B Mag = 1 000 X 30 vol % 10 m B Mag = 1 000 X 30 vol % 10 m D Mag = 1 000 X 50 vol % 10 m

PAGE 44

44 CHAPTER 4 MECHANICAL CHARACTERIZATION 4.1 Introduction As discussed in Chapter 3, BaTiO 3 foams were synthesized using the direct foaming method based on two different rigid, Si free polyurethane systems, a commercially purchased system (P U2 = CPU) and a laboratory developed system (PU5 = LPU). It was found that the microstructure of the ceramic foam could be controlled by adjusting the polyurethane system, ceramic composition, sintering time and sintering temperature. An open structure foa m with dense, rounded struts composed of a few larger grains (20 – at a high sintering temperature (in comparison to bulk) and an intermediate sintering time 34 The optimal ceramic content wa s determined to be 30 vol% and the PU made up the remaining 70 vol% It was found that the variability in density for these foams was not significant. Although these foams are primarily intended for sensing application, the ir ability to withstand stress an d deformation is important for handling purposes as well as to ensure mechanical integrity at high stress levels during service. Among many desirable properties, the compressive behavior is of prime importance In literature, mechanical propert ies of 1 3 35 and 0 3 22 composites have been reported but very little has been reported on mechanical p roperties of 3 3 composites. A typical 1 3 composite consists of rods in a matrix in which c eramic rods are held parallel by a passive polymer matrix. 35 F or underwater sensor application it is important that the acoustic impedance of the composite be similar to that of water (1.569 MRayls). 4 While recent breakthroughs have been made with 1 3 composites in achieving this low acoustic impedance 5 3 3 composites offer a good balance between a polymer with low acoustic impedance and a piezoelectric c eramic with high mechanical strength and piezoelectric properties. However, for good mechanical stability, the strengths of both the ceramic foam and the polymer must be enhanced.

PAGE 45

45 In this study the mechanical properties of the ceramic foam synthesized via the process previously discussed in Chapter 3 were measured For the piezo sensor application, the primar y focus is on the elastic response because reversible deformation i s required for the sensor to operate. In this work it is shown that the mechanical properties of the foam fabricated by the above mentioned method are comparable to other ceramic foams and therefore, possibly suitable for sensor applications. For polymeric foams, Liu and Subhas h 36 proposed a phenomenological model that can capture the entire compressiv e stress strain response In general, t he stress strain response of porous materials subjected to compressive loads has three distinct zones; (i) an elastic region, (ii) a plateau region and (iii ) a densification region. The plateau region is associated wit h the collapse of the porous, cellular structure which gives the foam its energy absorption ability. As mentioned previously, f or the piezo sensor application, the primary focus is on the elasti c region due to the requirement of reversible deformation. Therefore, only the early portion of the stress strain response that includes elastic and partial collapse strains were used for modelling and property determination. 4.2 Experimental Procedure BaTiO 3 foam samples were synthesized using two rigid, Si free polyurethane systems: (i) commercially purchased (CPU previously PU2 ) from Smooth On (Easton, PA) and (ii) laboratory developed system (LPU previously PU5 ) by Traversa et al. 29 at the University of Rome – Tor Vergata. After the previously reported processing steps, 28 32 vol% BaTiO 3 was foamed and sintered for mechanical testing. 34 BaTiO 3 foams produced from t he above two systems were cut into 12.7 mm diameter specimens using a tape cast hole punch. Five specimens were made for each system. The density of the foam was systematically varied by varying the ceramic content from 28 – 32 vol%. It was earlier found t hat the variability in density from one specimen to another was small (0.08%) when the initial ceramic content was kept constant ; therefore, only one sample was

PAGE 46

46 made for each density and the difference in mechanical properties of foams produced from the tw o PU systems was determined The apparent density and open cell porosity of each specimen was measured using helium pycnometry ( Table 4 1 ) It is shown that the LPU foams were denser due to a smaller expansion during the foaming p rocess but they also have a larger change in density with corresponding change in ceramic content compared to CPU foams Uniaxial compression experiments were performed at room temperature using a MTS universal testing machine (UTM). Each specimen was mar ked with a grid of horizontal lines to track the uniformity of deformation during the test. Previously it was noted that when these foam specimens were loaded without any lateral confinement, they deformed unevenly along the length of the specimen. 37 This behavior was due to premature collapse of large porous cells that lie along the outer surface. The collapse the led to instabilities resulting in buckling of the entire specimen. Such buckling modes are also o bserved in commercial brittle polymer foams. To avoid this mode of failure it was decided to test the specimens in a transparent confinement cell as shown in Figure 4 1 The transparent cell was made of acrylic an d consisted of a central cylinder with an end cap that can be screwed to the bottom of the cylinder. The ceramic foam specimen is placed inside the cell and pushed to the bottom by a cylindrical rod. The entire assembly is placed in the MTS machine and loa ded at a displacement rate of 1.27 mm/min resulting in a strain rate of 0.001/s. It is assumed that the strain of the acrylic rod is negligible. The experimental data was then fit using the phenomenological model proposed by Subhash and Liu 36 further discussed in Section 4.3.3. 4.3 Results and Discussion 4.3.1 Confined Compression Testing Because the foams were fabricated using the process described in Chapter 3 the microstructure is similar. The ceramic content was va ried around an average of 30 vol% to determine the effect of density on the mechanical properties of the foams. Both PU foams are

PAGE 47

47 resilient during handling and produce homogenous microstructures as determine d from SEM and image analysis previously reported by the autho r and collaborators 34 CPU foams had grains while LPU foams ha d and struts consisting of a few large grains (45 ). 34 The compressive engineering stress stra in responses of the different ceramic foams are shown in Figure 4 2 and Figure 4 3 Because the focus of the work was on the elastic region of the respon se, the samples were strained below 60% (to the yield point) Similar to any cellular material behavior, the ceramic foams exhibited an initial elasti c response followed by a cell collapse process which results in the plateau region. If the foams were compressed further, a densification response probably would have been observed. The stress strain response was strongly dependent on initial foam density. Both the initial slope and the collapse stress increased with density for both PU systems The slopes were highly non linear and therefore the stiffness was incrementally calculated and averaged over the elastic range. This stiffness is known as the colla C ). Mechanical properties extracted from the above plots are summarized in Table 4 2 The compression modulus (E C ) and collapse st C ) of the CPU and L PU foams determined from the above stress strain responses ar e plotted with respect to density in Figure 4 4 Although b oth material properties increase with increas ing density for both foams a significantly higher rate of increase is observed in foam produced by LPU. As ex pected, the LPU foam had on average a higher modulus and col lapse stress due to higher density The mechanical properties for both of the ceramic foams follow a linear trend with density. A better comparison of mechanical properties of foams can be seen in Figure 4 5 where the stress st rain curves at two different foam densities are com pared Clearly, for approximately similar foam density, a greater collapse stress is observed in LPU foam than in CPU foam in additi on to a greater increase in properties with a similar increase in density In addition, at both densities, it can be

PAGE 48

48 seen that the mechanical properties of the foams made with LPU are higher than the foams made with CPU. As mentioned earlier, the fo ams wer e marked with grid lines to track the uniformity of deformation and collapse the collapse process Figure 4 6 and Figure 4 7 illustrate the deformation behavior of foams at various strain levels. Note that all these lines remain relatively straight throughout the deformation process. In foams with relatively large cell window sizes and inhomogeneous microstructures these lines become nonlinear and the spacing between the li nes becomes uneven very early on in the deformation process due to the collap se of isolated large cells 36 Such features were not observed in the BaTiO 3 foams because the cell window size is relatively small and uniform and therefore, the spacing between any two lines remains relatively constant along the length. However, the spacing between the lines decreases with strain as the cells start to collapse uniformly There is one noticeable difference between the d eformation behavior of LPU and CPU foams. In the LPU foams, the deformation is not uniform along the length of the specimen. Note that the line spacing on the top decreases more rapidly than the spacing at the bottom. This indicated that the de formation in this foam is more progressive rather than uniform throughout the specimen. The deformation progresses from the top of the specimen to the bottom of the specimen with increasing load, i.e., there is a progressive collapse of cells with load. On the other h and, the deformation of CPU foams is comparatively more uniform ( Figure 4 7 ) as indicated by the continued reduction in the spacing of most of the lines with increasing strain. Table 4 3 com pares the properties of the selected BaTiO 3 foam to two other ceramic foams of similar densities and a bulk material commonly used in commercial sensors 41 The open cell, macrocellular SiOC fo am was synthesized by Colombo et al. via direct foaming method using a PU similar to LPU 38 The collapse strength of the SiOC foam s is higher while the densities are similar because the structures are different. The SiOC foams have more closed cell porosity and a much smaller cell window size 3 foams have significantly more

PAGE 49

49 open porosity and a larger window size (40 – 80 ). These features make the BaTiO 3 foam a better candidate for a p iezocomposite. The L PU foam at the same density has a higher collapse stress due to dense r struts, while the cordierite foams have hollow struts caused by the replica process. 39 The BaTiO 3 foam struts have an a and consist of a few described in the previous chapter. In comparison the SiOC have struts consisting of many small grains which are greater tha According to Li et al. 40 the thickness of the cell wall or strut is directly correlated with the strength of the foam; therefore, all other being equal, it is expected for the strength of the oth er foams to be higher due to thicker struts than the BaTiO 3 foam. Most piezo sensors i n the market today are made of bulk electroceramic mate rials such as PZT and have higher mechanical strength. The foams presented here will be infiltrated with a polymer to enhance mechanical properties and possibly piezoelectric properties depending on the polymer selected. The required mechanical strength of a piezoelectric mat erial for sensor applications is yet to be reported. W hile the mechanical properties of the BaT iO 3 foam skeleton are not yet comparable to o ther composites, such as the 1 3 ceramic rod composites by Smith et al. 5 (E~31 GPa for 25 % PZT rod composite) the ceramic foam could still be acceptable for sensor applications and will be determined once a prototype piezocomposite sensor has been fabricated. It must be noted that the mechanical properties of the foams presented here are not yet optimized. The primary focus is to produce 3 3 piezocomposites with enha nced piezoelectric properties before optimizing the mechanical properties and ensure that the material can withstand the stress required for this application. While the stress required may vary with application, s uggestions for increasing the mechanical st rength of the foam include increasing the ceramic content to 40 vol% or modifying the grain size by altering the sintering temperature. In addition infiltration with polymer is expected to increase the compliance of the compressive

PAGE 50

50 behavior and enhance t he relaxation behavior of the composite, while lowering the acoustic i mpedance of the overall sensor For such applications it is i mportant to maintain a balance between mechanical properties and electrical properties of the piezocomposite. Therefore the c urrent method of producing foams and the resulting microstructure are expected to produce an acceptable balance between the two properties, thus making it suitable for sensor applications. 4.3.2 Model Fit The stress strain behavior of the foams discussed here reveals an elastic response followed by an inelastic response that consists of the cell collapse in the foam. If the compressive load was continue, the foams would have exhibited a densification regime which is typical of any cellular material. This comple x behavior was captured using the following phenomenologic al model by Liu and Subhash 36 = A e 1 B + e + k e C ( e 1 ) ( 4 1 ) The first term in the model c aptures the elastic response and the cell collapse process whereas the second term captures the densification response. However, the interest is only in the early part of the response consisting of the elastic and cell collapse regimes, as shown in Figure 4 8 and Figure 4 9 Only t he first term in the above phenomenological model was used to fit these strain strain curves. Recently Walters et al. 37 extended the above model to both compression and tensile regimes and were able to determine several other parameters through extensive experimentation on five different density foams. The advantage of such a model is that one can determi ne the response of foam at intermediate d ensities that were not originally available. The model can also calculate tensile strength of the foam as will be describe below. In the above equation the parameter A represents the yield stress (or collapse stre ss) and parameter B represents the ratio of ultimate tensile strength to collapse stress. 37 Parameters and describe the plateau region of the stress strain response For > a hardening like

PAGE 51

51 behavior is observed, = represents the perfectly plastic response and < represents a softening like behavior. By fitting the experimental data with the first term in the model equation, the model parameters described above ( ) can be obtained. The r elationship between the model parameters and foam mechanical properties were determined by Walter et al. 37 as follows: C ( 4 2 ) C ult ( 4 3 ) E = A 1 + B ( 4 4 ) Figure 4 8 and Figure 4 9 show the comparison betw een the experimental data and the model fit for the LPU and the CPU BaTiO 3 foams respectively As shown, the model accurately captures both the elastic and plateau regions; therefore with the parameters determined from the model, the compressive modulus and collapse stress can be accurately calculated using the equations listed above. Figure 4 10 illustrates the difference in model parameter as a function of density. The difference in trends for and between CPU and LPU is purely mathematical and due to the fact that the collapse stresses for CPU are under 1 MPa and the collapse stresses for LPU are above 1 MPa. This is also the reason why the values for and are much higher for the ceramic foams compared to and for the polymer foams 37 Compared to the model parameter s obtained by Subhash et al. 36 when testing polymer foams, some of the trends are di A increases and B remains relatively constant with increasing density. The values of the mechanical properties obtained from the model are comparable to the values d etermined previously from the measured data as seen in Figure 4 11 Clearly, the model captures the experimental data well. Therefore, the model can now be extended to other density foams with the range of densitie s tested here because the trends in the model

PAGE 52

52 parameters are now known. We can also determine the ultimate tensile strength of these foams with the above equation. Note that, unlike the compression response where large strains are noted, brittle foams exhi bit small strain in tension before failure. This extended experimental work by Walter et al. 37 has determined the relationship for structural foams. Figure 4 12 show s t he expected ultimate tensile stresses ( ult ) of the ceramic foams as determined using the above equation. As expected, the tensile strength of the BaTiO 3 foams is significantly less than the compressive strength (three times) due to the inherent brittle behavio r of ceramics. The ultimate tensil e strength increases with density. Interestingly, it is noted that t he tensile strength of the LPU foams increase linearly whereas the tensile strength of the CPU foams increase nonlinearly. More in depth experimentation and model analysis is required to f urther validate these values in this class of ceramics. The properties determined here provide a starting point for optimization of the foams in future studies were the ceramic foams will be infiltrated with polymeric materials for enhanced piezoelectric e ffect. 4.4 Conclusion The mechanical properties of BaTiO 3 foams synthesized via direct foaming method using a commercial and a laboratory developed rigid, Si free polyurethane system s were determined using confined compre ssion testing There was an increase in modulus and collapse strength with increasing density in both foams. T he LPU foam s were denser compared to the CPU foams (0.82 g/cm 3 vs. 0.51 g/cm 3 ), and hence had a higher modulus (6.92 MPa) and collapse strength (0.91 MPa) than the CPU foam (4.16 MPa and 0.47 MPa respectively ). The mechanical properties of LPU foam are more strongly dependent on density than CPU foams. The mechanical pr operties of both foams were comparable with the mechanical properties of other ceramic foams such as SiOC and Cordier ite.

PAGE 53

53 The phenomenological model proposed by Liu and Subhash 36 captured the elastic and plateau regions of the compressive behaviour reasonably well. The model parameters determined also compare well with the ex perimentally determine properties. Based on this work, it is suggested that these ceramic foams have good mechanical properties and the potential for electromechanical applications such as piezo sensors where they can be infiltrated with polymer to create a piezo composite with enhanced the piezoelectric and mechanical properties.

PAGE 54

54 Table 4 2 Mechanical properties for LPU foam Ceramic C ontent (vol%) E C (MPa) C (MPa) CPU LPU CPU LPU 32 4.99 9.69 0.93 1.25 31 4.71 7.77 0.60 1.07 30 4.64 6.69 0.54 1.06 29 4.06 6.08 0.51 1.02 28 3.67 4.39 0.39 0.78 Average 4.41 6.92 0.59 0.91 StDev 0.53 1.97 0.21 0.32 Table 4 3 Proper ties of various ceramic foams and commercial PZT Parameter BaTiO 3 Foam LPU BaTiO 3 Foam CPU SiOC Foam Macrocellular 38 Cordierite Foam 39 Bulk PZT 41 3 ) 0.43 0.42 0.4 0.34 7.6 E (MPa) 4.39 3.78 66,000 c (MPa) 0.78 0.47 3.2 0.73 > 517 Table 4 1 Porosity and density values for LPU foams Ceramic Content (vol%) Porosity (%) Density (g/cm 3 ) CPU LPU CPU LPU 32 87. 0 80.0 0.78 1.20 31 88.0 84.7 0.72 0.92 30 91.0 85.3 0.54 0.88 29 93.0 86.3 0.42 0.82 28 93.7 92.8 0.38 0.43 Average 90.6 85.8 0.57 0.82 StDev 2.95 4.6 0.18 0.28

PAGE 55

55 Figure 4 1 C onfinement cell and foam sample used for confined compression testing 0 10 20 30 40 50 60 70 80 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.43 g/cm 3 0.82 g/cm 3 0.88 g/cm 3 0.92 g/cm 3 1.2 g/cm 3 Stress (MPa) Strain (%) Figure 4 2 Stress strain response of BaTiO 3 foam produced from LPU system.

PAGE 56

56 0 10 20 30 40 50 60 70 80 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.72 g/cm 3 0.78 g/cm 3 0.54 g/cm 3 0.42 g/cm 3 0.38 g/cm 3 Stress (MPa) Strain (%) Figure 4 3 Stress strain response of BaTiO 3 foam produced from CPU system. 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 CPU E C CPU C LPU E C LPU C Linear Fit MPa Density (g/cm 3 ) Figure 4 4 Mechanical properties of BaTiO 3 foam as a function of density

PAGE 57

57 0 10 20 30 40 50 60 70 80 90 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 LPU (0.43 g/cm 3 ) CPU (0.42 g/cm 3 ) LPU (0.82 g/cm 3 ) CPU (0.78 g/cm 3 ) Stress (MPa) Strain (%) Figure 4 5 Stress strain curves comparing density and PU Figure 4 6 Compression of LPU foam with respective stress strain curves

PAGE 58

58 Figure 4 7 Compression of CPU foam with respective stress strain curves 0 10 20 30 40 50 60 70 80 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Data Model 0.43 g/cm 3 0.88 g/cm 3 0.82 g/cm 3 0.92 g/cm 3 1.2 g/cm 3 Stress (MPa) Strain (%) Figure 4 8 Measur ed data and model fit for BaTiO 3 foam produced by LPU.

PAGE 59

59 0 10 20 30 40 50 60 70 80 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Stress (MPa) Strain (%) Data Model 0.72 g/cm 3 0.78 g/cm 3 0.54 g/cm 3 0.42 g/cm 3 0.38 g/cm 3 Figure 4 9 Measured data and model fit for BaTiO 3 foam produced by C PU 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45 LPU CPU A B and A and B (MPa) Density (g/cm 3 ) A B Figure 4 10 M odel parameters as a function of density.

PAGE 60

60 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 0 2 4 6 8 10 12 14 16 E-CPU E C E-CPU C E-LPU E C E-LPU C M-CPU E C M-CPU C M-LPU E C M-LPU C MPa Density (g/cm 3 ) Figure 4 11 Comparison of mechanical properties determined from the (M ) fitted d ata and (E ) experimental data as a function of density 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 CPU LPU ult (MPa) Density (g/cm 3 ) Figure 4 12 Model estimated ultimate tensile strength of BaTiO 3 foams as a function of density.

PAGE 61

61 CHAPTER 5 POLYMER INFILTRATION AND PRELIMINARY ELECTRICAL CHARACTERIZATION 5.1 Introduction The ultimate goal of this work is to c reate a 3 3 composite. Up until now, the research has been focused on creating the electroceramic skeleton from BaTiO 3 which can be considered a composite composed of ceramic and air. By incorporating a second phase into the cell window network of the foa m, the properties can be tailored for specific applications. For ultrasound sensors the properties required are elastic mechanical properties and high HFOM (high d 33 and d 31 ). Since it is known that BaTiO 3 has a high dielectric permittivity and a low HFOM ( 93 x 10 15 m 2 /N) 22 the polymer must have properties that will improve the properties of the composite. Although the polymer might not be piezoelectric, its high compliance allows it to transfer stress to the ceramic, enhancing piezoelectric properties overall. 20 Acoustic impedance depends on the applications. For underwater sonar sensors, the composite should have an acoustic impedance comparable to water (1.569 MRa yls) and for medical ultrasound applications it should be comparable to body tissue. 4 Ceramic piezoelectrics have high acoustic impedances which differ greatly from the acoustic impedance of water while polymers have low acoustic impedances which can mediate the high difference when coupled with the ceramic 20 Besides polymer selection, the infiltration process is also important. A simple process that achieves complet e infiltration at room temperature is desired. Vacuum assisted impregnation is that process. 17 18 Coupled with the right polymer, the 3 3 composite is achieved. 5.2 Exp erimental Procedure 5.2.1 Polymer Infiltration Previously synthesized LPU foams are cut into either 8 x 8 x 5 mm plates for dielectric and piezoelectric characterization or 25.4x2 mm cylindrical specimens for acoustic impedance measurements. These dimensions a re based on application frequency and literature. For piezoelectric measurements, the samples are cut in compliance with preferred dimensions for

PAGE 62

62 optimum piezoelectric measurements. 41 The samples are then put in to a polymer container and set in the vacu um chamber of the Epovac (Logitech Ltd, Glasgow, Scotland, U.K. ) shown in Figure 5 1 The pressure of the chamber is lowered to 0 MPa and held for 3 minutes to ensure that all moisture has been removed from the sample. After, the polymer is poured into the polymer containing the sample very slowly. Once the sample is completely covered, pressure is raised to atmosphere and the sample is removed. After the polymer is cured, t he sample is polished to remove excess polymer and vacuum grease. Scanning electron microscopy (SEM) and optical microscopy are used to characterize the composite. 5.2.2 Electrical Characterization For electrical characterization, the plates were coated (Au/Pd ) and electroded using silver paint. The measured as a function of frequency using a precision LCR meter connected and computer interf ace. From this data, the relative dielectric constant r ’), also known as the real part of the permittivity, and the imaginary part of the r ”) could be calculated using the equations 21 : r = Ct 0 A ( 5 1 ) r = r tan ( 5 2 ) The samples were then poled in jojoba oil at 60 o C under an electric field of 30 kV/cm. 42 Average piezoelectric coefficients for the materi als (d 33 and d 31 ) were measured using a standard piezoelectric meter system. From these measurements the HFOM (d h g h ) could be calculated and compared to the HFOM of bulk BaT iO 3 bulk PZT and other piezoelectric composites. The impedance of the composite w as measured to determine the resonant frequencies which can then be used to calculate the electromechanical coupling coefficients (k eff k 33 k h etc).. Resonance measurements are performed using a precision impedance analyzer. The results are discussed he re.

PAGE 63

63 5.3 Results and Discussion 5.3.1 Polymer Infiltration The important properties to consider when choosing which polymer to use for infiltration are viscosity, cure time, cure temperature and hardness. Epoxies used in casting and injection molding were first co nsidered because of their low viscosities and ability to be molded. The cure time had to be long enough so that the polymer didn’t cure before completely infiltrating the sample but short enough to be practical for commercial application. In addition, a po lymer that cures at room temperature makes the process simpler and prevents the structures from transitioning at higher temperatures. Finally, a low hardness and dielectric constant are desired for the piezoelectric application. Table 5 1 lists the polymers and the respective properties that were considered for infiltration taken from the company websites There has been previous work on composites which used the polymer polyvinyl difluoride (PVDF) because it is o ne of the few pol ymers which is piezoelectric. While this is an interesting property for piezocomposites, using PVDF to infiltrate is a difficult process because the solvent is usually a complex organic and the solution does not have a low viscosity, which is why PVDF was not chosen for the first batch of composites, but may be used later to optimize the electrical properties. Based on the properties listed above and availability, Epofix was chosen for infiltration. In addition, it also commonly used in SEM to increase con trast. The resulting SEM images of the composite are shown in Figure 5 2 From these images, it can be seen that the foam system has been completely infiltrated and that the ceramic foam is a homogeneous phase. Unf ortunately, the SEM image cannot show the 3 3 connectivity of the composite since it is a 2 dimensional technique. While it can be assumed that the composite has 3 3 connectivity because the foam originally had 3 3 connectivity between the foam and air, an other characterization technique will need to be used in order to create 3 dimensional images of the

PAGE 64

64 composite. Regardless, a 3 3 composite has been achieved between ceramic foam and polymer. 5.3.2 Electrical Characterization Figure 5 3 show s the capacitance and loss of the composite as a function of frequency from 40 Hz to 110 MHz respectively Using the equation mentione d above, the dielectric permittivity of the material is calculated and also shown as a function of frequency from 40 Hz to 11 0 MHz in Figure 5 4 The typical frequency range for ultrasound applications is above human hearing which is 20 kHz. 4 As seen from the figure, the dielectric constant decreases with increasing frequency. Therefore, for ultrasound applications the piezoelectric properties increase due to the decreasing dielectric constant. Table 5 2 lists the piezoelectri c coefficients including d 33 d 31 and d h g h for the composite. It is assumed that r ~ r ’. The static dielectric constant (at 1 kHz) was used to calculate the HFOM. The theoretical stat ic dielectric constant was calculated using the rule of mixtures assu ming the volume fraction of ceramic is about 7 vol%, and compared to the experimental value This value is 144 .6, which is similar to value determined experimentally The term 33 in the HFOM equation is the dielectric constant multiplied by the permittivi ty of free space. From the measurements and calculations, the piezocomposite has a lower HFOM than bulk BaTiO 3 and bulk PZT. This is could be due to a couple of factors. Firstly, the volume of fraction of piezoelectric material is only about 10 vol%. An i ncrease in the HFOM is expected with an increase in the amount of piezoelectric material in the ceramic. A theoretical HFOM as a function of vol ume fraction has shown in Figure 5 5 It is observed that the HFOM pea ks at about 10 vol% and that the theoretical HFOM is much higher than the measured HFOM at 7 vol%. This is most likely due to the poling conditions. T he polymer limits the temperature and electric field that can be used to pole the composite. However, the focus of this work was to successfully create a 3 3 composite and measure the piezoelectric properties, which has been

PAGE 65

65 achieved. The next step is to improve these properties further through material selection, using piezoelectric materials such as PZT and PVDF as mentioned above. The impedance of the composite was measured; however no resonant frequencies were observed indicating that the composite acted like a capacitor. This is due again to the choice of materials and to the low ceramic volume fraction in the composite which would be the source of the resonance. Future work consists of attempting these measurements again with a higher ceramic content composite made of PZT. 5.4 Conclusion Using BaTiO 3 foam synthesized via direct foaming method as discussed in previous chapters, a composite was fabricated which has 3 3 connectivity. Different polymers were considered for infiltration based on their viscosity, cure time, cure temperature, hardness and dielectric constant. The purpose of the polymer in the com posite is to decrease the dielectric constant, acoustic impedance and brittle behavior of the foam while enhancing the HFOM. Based on these properties, Epofix resin commonly used for mounting samples for SEM characterization due to its high phase contrast was chosen. The samples were infiltrated using Epovac equipment which is a vacuum impregnation system. Using this technique, polymer infiltrated all empty space in the foam. Using SEM, it was confirmed that the foam was completely infiltrated, creating a 3 3 composite with the ceramic foam. However, only a 2D image could be taken using this tech ni q ue. Therefore, a 3D image will be achieved later to show the interconnecting phases. After the composite was fabricated, preliminary electrical characterization measured the piezoelectric and dielectric properties of the material. After the samples were poled at 60 o C for 30 minutes under a 30 kV/cm field, the piezoelectric coefficients (d 33 d 31 and d h g h ) were measured using a piezoelectric meter. The HFOM of th e composite ( 53.7 x 10 15 m 2 /N) did not exceed the HFOM of bulk BaTiO 3 or bulk PZT. This is due to the low compliance of the polymer chosen for infiltration and this knowledge will be used in determining a different polymer for the

PAGE 66

66 composite in the next ex periments. This is also due to the low volume fraction of piezoelectric material in composite. In addition, resonance measurements were unsuccessful since resonance was not observed in the composite again due to the material choice and low ceramic content. However, the 3 3 piezoelectric composite has been successfully fabricated. Now, the composite can be fabricated with better piezoelectric materials such as PZT and PVDF, using the processes developed in this work.

PAGE 67

67 Table 5 1 Material properties of polymers considered for infiltration Polymer Viscosity (cps) Hardness Cure Time (h) Cure Temperature Dielectric Constant EP5340 (Eager Polymers) 600 88D 24 25 o C 4.57 PVDF (Kynar) 3.1E6 78D ~ 10 Epofix (S truers) 50 75D 8 25 o C ~ 5 Figure 5 1 Epovac system for polymer impregnation Figure 5 2 SEM images of the infiltrated composite: (light) BaTiO 3 (dark) Epofix. Table 5 2 Piezoelectric coefficients of composite and bulk materials Material (kg/m 3 ) r ’ d 33 (pC/N) d 31 (pC/N) d h g h (10 15 m 2 /N) BaTiO 3 piezocomposite 1737 156 45 18 58.7 Bulk BaTiO 3 19 6200 1400 190 78 93 Bulk PZT 8 41 7600 225 27 10 0

PAGE 68

68 0 20 40 60 80 100 120 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018 Frequency (MHz) Capacitance (pF) 0 10 20 30 40 50 Loss (tan ) Figure 5 3 Capacitance and loss of composite as a function of frequency 0 20 40 60 80 100 120 0 20 40 60 80 100 120 140 160 180 200 Frequency (MHz) r 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 r Figure 5 4 Permittivity (real and imaginary) of composite. 0.0 0.2 0.4 0.6 0.8 1.0 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 HFOM Volume Fraction of BaTiO 3 Figure 5 5 Theoretical HFOM as a function of ceramic volume fraction

PAGE 69

69 CHAPTER 6 SUMMARY AND OUTLOOK 6.1 Summary A surfactant stabilized direct foaming method was used to create BaTiO 3 based ceramic foams. The composition, polyurethane system and sintering time were var ied to determine their influence on the microstructure of the foam and ultimately the properties and applications of the foam After determining that Si based PU systems produced contaminants at the grain boundary and inhomogeneous grain size distributions and flexible PU systems produced closed cell structures, research was focused on Si free, rigid PU sys tems: one commercially purchased and the other laboratory developed. After additional initial testing, some of the other variables were eliminated. The p rimary synthesis variable s were polyurethane system (CPU – commercial and LPU – laboratory) and composition (20 – 60 vol% ceramic). Both the rigid commercial and rigid laboratory systems produced foams with uniform cell window size distributions and dense thick struts. However, the commercial system’s struts consisted of many small grains, while the laboratory system struts had fewer but larger grains. The rigid laboratory developed system demonstrated uniform cell window distribution, variable porosity i n relation to ceramic composition (75 87 %), dense struts, contaminant free grain boundaries at concentrations less that 50 vol% and mechanical integrity. The commercial silicon free system also had a uniform cell window size distribution, smaller grain size, larger cell window size and also demonstrated dense struts and variable porosity in relation to ceramic composition (82 – 93 %). Lower ceramic concentrations (20, 30 and 40 vol% ) are ideal for electrical applications because of the interconnected cel l window network and increased porosity. However, they are more brittle than foams with higher ceramic concentrations (50 and 60 vol% ) which are more favored in mechanical applications. 30 vol% foams demonstrated a homogeneous microstructure with a balance of properties for both mechanical and electrical applications. They

PAGE 70

70 have high porosity (~ 88 %), uniform and smaller grain size, are relatively strong and not easily damaged. Raman spectroscopy indicated that compressive residual stress was created in the structure during the foaming process which could have affected the mechanical properties. This was observed by the shifts in the Raman modes of the spectra in comparison to bulk barium titanate. A 30 vol% foam, synthesized using the laboratory developed rigid polyurethane system and a 30 vol% commercial silicon free system, sintered for 8 hours at 1400 C, produced different microstructures which have the potential to be used in electromechanical applications. The mechanical properties of BaTiO 3 foams sy nthesized via direct foaming method using a commercial and a laboratory developed rigid, Si free polyurethane system s were determined using confined compre ssion testing There was an increase in modulus and collapse strength with increasing density in both foams. T he LPU foam s were denser compared to the CPU foams (0.82 g/cm 3 vs. 0.51 g/cm 3 ), and hence had a higher modulus (6.92 MPa) and collapse strength (0.91 MPa) than the CPU foam (4.16 MPa and 0.47 MPa respectively ). The mechanical properties of LPU fo am are more strongly dependent on density than CPU foams. The mechanical pr operties of both foams were comparable with the mechanical properties of other ceramic foams such as SiOC and Cordierite. The properties determined here provide a starting point for optimization of the foams in future studies were the ceramic foams will be infiltrated with polymeric materials for enhanced piezoelectric effect. The phenomenological model proposed by Liu and Subhash 36 captu red the elastic and plateau regions of the compressive behaviour reasonably well. Using this model, the properties and stress strain behaviour of intermediate densities can be calculated without being measured. These properties include compressive modulus (E C C ) and the ultimate ult ). The properties determined by the experimental data and the properties determined by the model were very similar, which verified the model’s capabilities to accurately determine the mechanic al properties of the ceramic foam.

PAGE 71

71 Finally, using BaTiO 3 foam, a composite was fabricated which has 3 3 connectivity. Different polymers were considered for infiltration based on their viscosity, cure time, cure temperature, hardness and dielectric constan t. The purpose of the polymer in the composite is to decrease the dielectric constant, acoustic impedance and brittle behavior of the foam while enhancing the HFOM. Based on these properties, Epofix epoxy commonly used for mounting samples for SEM characte rization due to its high phase contrast was chosen. The samples were infiltrated using Epovac equipment which is a vacuum impregnation system that successfully infiltrated all empty space in the foam. Using SEM, it was determined that the foam was complet ely infiltrated, creating a 3 3 composite with the ceramic foam. However, only a 2D image could be taken using this technique Therefore, a 3D image will be achieved later to show the interconnecting phases. After the composite was fabricated, preliminar y electrical characterization measured the piezoelectric and dielectric properties of the material. After the samples were poled at 60 o C for 30 minutes under a 30 kV/cm field, the piezoelectric coefficients (d 33 d 31 and d h g h ) were measured using a piezoe lectric meter. The HFOM of the composite (53.7 x 10 15 m 2 /N) did not exceed the HFOM of bulk BaTiO 3 or bulk PZT. This is due to the low compliance of the polymer chosen for infiltration and this knowledge will be used in determining a different polymer for the composite in the next experiments. The experimental HFOM is lower than the calculated theoretical HFOM. However, the 3 3 piezoelectric composite has been successfully fabricated. Now, the composite can be fabricated with better piezoelectric materials such as PZT and PVDF, using the processes developed in this work. In addition, resonance measurements were unsuccessful since resonance was not observed due to the material choice and low ceramic volume fraction. Now, the composite can be fabricated with better piezoelectric materials such as PZT and PVDF, using the processes developed in this work.

PAGE 72

72 6.2 Outlook While there have been many accomplishments achieved in this work including a unique electroceramic foam structure and 3 3 piezocomposites, there is still a lot of research to do in order to optimize and fully characterize the 3 3 piezocomposite. These include synthesizing composites with higher piezoelectric material volume fraction, resonance measurements, acoustic impedance measurements and composi te structure characterization using a 3D imaging technique such as TEM. In addition, characterization of the composite mechanical properties and eventually the optimization of the mechanical properties and electrical properties should be achieved in the fu ture. The acoustic impedance of the composite is measured using equipment consisting of a speaker, wave guide and sample holder. The samples are cut into a 25.4 mm diameter cylinder, mounted onto the sample holder, and placed at the end of the wave guide. The acoustic impedance is measured at 20 kHz to simulate ultrasound conditions. The measured acoustic impedance is then compared to the acoustic impedance of water and tissue to determine if it had acceptable properties of ultrasonic sensor applications e ither underwater or in medicine. Additionally, the effective acoustic piezoelectric modulus will be measured using the same samples from the acoustic impedance measurements. The samples are poled and vibrated using a laser vibrometer. As stated in Chapter 5, an additional imaging technique is desired to characterize the composite structure because SEM can only show that the phases are interconnected in two dimensions, while this work is focused on achieving connectivity in 3 dimensions. In addition, there has been work by Wachsman et al. on quantifying 3 3 connectivity. In most cases, connectivity is characterized by imaging techniques and is only qualitative. However, this new work has been able to quantify this property by creating relationships between t he number of nodes and struts in either phase of the composite.

PAGE 73

73 As stated previously, there has been little research on the required (minimum) mechanical properties of piezoelectrics. Future work will determine this by characterizing the compressive and r elaxation behavior of the composites. In addition, the piezocomposite structure will be optimized to enhance mechanical properties and electrical properties. This includes using different materials for the composite such as PZT and PVDF to observe their ef fect on the HFOM. Now that the 3 3 composite has been fabricated, the freedom to experiment with other materials for different applications is availabl e.

PAGE 74

74 APPENDIX A IMAGE ANALYSIS The steps used to perform image analysis on the FE SEM images of the foams made by the author are described in this appendix. 1. Micrographs were taken at 1000X magnification and different locations on the sample. The images used did not have scale bars or labels on them. This prevents the software from calculating th e label which would affect the data output. 2. Using an image editing software, such as Photoshop, the grain boundaries are digitally enhanced to increase contrast in the image as shown in Figure A 1 3. Image J softwa re is opened and the image is opened in the program. 4. Image>Type>32 bit. This changes the image from an RGB image to a 32 bit image. If the image is not converted, it cannot be analyzed. 5. Image>Adjust>Brightness/Contrast. The contrast is adjusted to the e xtreme so that the grains are whit e and grain boundaries and cell windows are black as shown in the Figure A 2 below. 6. Adjust>Threshold. Adjust the threshold so that the feature to be measured is red. For example, if the grain size is being measured, then the grains should be red. Images for both grain size measurements and cell window size measurements are shown in Figure A 3 7. Analyze>Analyze Particles. As stated in Chapte r 3, the macro then calculates the number of red pixels and then prints out the results and summary of the calculation which include the average area. 8. Because the macro reports the area of the feature, the size of the feature must be determined assuming t hat the grains and pores are circular. As seen from the above figures, this is not the case and this method gives only an estimation of the feature sizes. 9. Based on the scale bar and magnification, the feature sizes which are calculated in pixels/feature are then converted to microns/feature which is the actual feature size.

PAGE 75

75 Figure A 1 Micrograph of PU2 foam with the grain boundaries digitally enhanced Figure A 2 Micrograph with enhanced con trast.

PAGE 76

76 Figure A 3 Micrographs with the threshold adjusted to measure (left) grain size and (right) cell window size.

PAGE 77

77 APPENDIX B RAMAN SPECTROSCOPY Raman spectroscopy was performed on the foam synthesized via direct foaming. This was done to determine if residual stress had formed in the structure during the foaming process. Measurements were taken at room temperature, 100 o C, and 150 o C to capture the phase transition from cubic to tetragonal at around 130 o C The measured spectra were then compared to the Raman spectra of bulk BaTiO 3 pellets to determine if there were shifts in the peak center which would indicate residual stress. Table B 1 and Table B 2 are the character tables for cubic and tetragonal barium titanate, respectively, based on group theory. The high temperature cubic phase has no Raman active modes while the lower temperature tetragonal phase has seven Raman ac tive modes and six infrared active modes. All measurements were taken using a Reinshaw Biotherm Raman microscope and temperature controlled sample holder. The experiments were part of a collaboration with Professor Ian Reaney and Dr. Honza Pokorny at the University of Sheffield. Figure B 1 is the measured Raman spectra. It is observed that due to background noise, nine modes were required to fit the data. While group theory estimates that there are no Raman active modes in the cubic phase, it is not uncommon to see tetragonal phase peaks in the supposedly cubic structure. This is due to local tetragonal phases at the atomic level which are detected in the Raman spectroscopy. Figure B 2 shows the Raman spectra and c hange in peak center between the pellet and foam as a function of peaknumber for each temperature. The major modes to notice are peak numbers 3, 6, 7 and 9 as they are prominent peaks in the spectra. According to Figure B 2 there is a significant differenc e in the peak centers of 6, 7 and 9 at 150 o C between bulk barium titanate and the foam. This indicates that there is residual stress formed during the foaming process, creating tensile stresses which cause the bonds to stretch and tighten thus vibrating a t higher frequencies. Interestingly, peak number 3 does not appear in the foam at 150 o C.

PAGE 78

78 Table B 1 Character table for ABO 3 cubic structure Distribution of Degrees of Freedom Number of normal modes h O A and B cations h O 1(a), 1(b) O anion h D 4 3 (c) Acoustic Modes Rotational Modes Lattice Modes Selection Rules A 1 g 0 0 0 0 0 Raman A 1u 0 0 0 0 0 Inactive A 2g 0 0 0 0 0 Inactive A 2u 0 0 0 0 0 Inactive E g 0 0 0 0 0 Raman E u 0 0 0 0 0 Inactive F 1g 0 0 0 1 0 Inactive F 1u 2 2 1 0 3 Infrared F 2g 0 0 0 0 0 Raman F 2u 0 1 0 0 1 Inactive ) ( 3 = 1 IR F u Table B 2 Character table for ABO 3 tetragonal structure Distributi on of Degrees of Freedom Number of normal modes v C 4 A, B, and O 1 v C 4 1(a), 1(b) O 2 v v C 2 2(c) Acoustic Modes Rotational Modes Lattice Modes Selection Rules A 1 3 1 1 0 3 Raman and Infrared A 2 0 0 0 1 0 Inactive B 1 0 1 0 0 1 Raman B 2 0 0 0 0 0 Raman E 3 2 1 1 3 Raman and Infrared ) ( 3 + ) ( + ) ( 3 = 1 1 IR R E R B IR R A

PAGE 79

79 30 o C 0 200 400 600 800 1000 0 100000 200000 300000 400000 Count Raman Shift (cm -1 ) Data Peak fit Modes 100 o C 0 200 400 600 800 1000 0 200000 400000 600000 800000 Count Raman Shift (cm -1 ) Data Peak fit Modes 150 o C Figure B 1 Raman sp ectra of (left) bulk barium titanate and (right) barium titanate at different temperatures 0 200 400 600 800 1000 0 50000 100000 150000 200000 250000 300000 Count Raman Shift (cm -1 ) Data Peak fit Modes 0 200 400 600 800 1000 0 50000 100000 150000 200000 250000 300000 Count Raman Shift (cm -1 ) Data Peak fit Modes 0 200 400 600 800 1000 0 50000 100000 150000 200000 250000 300000 Count Raman Shift (cm -1 ) Data Peak fit Modes 0 200 400 600 800 1000 0 100000 200000 300000 400000 500000 Count Raman Shift (cm -1 ) Data Peak fit Modes

PAGE 80

80 Figure B 2 Raman spectra comparison between barium titanate foam and pellet (top left) at 30 o C, (top right) at 1 00 o C and (bottom left) 150 o C and the difference in peak center (foam peak center – pelle t peak center) 0 200 400 600 800 1000 0 100000 200000 300000 400000 500000 600000 Count Raman Shift (cm -1 ) Foam Pellet 0 200 400 600 800 1000 0 100000 200000 300000 400000 500000 Count Raman Shift (cm -1 ) Foam Pellet 0 200 400 600 800 1000 0 200000 400000 600000 800000 Count Raman Shift (cm -1 ) Foam Pellet 0 1 2 3 4 5 6 7 8 9 10 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 in Peak Center from Pellet 30 o C 100 o C 150 o C Peak Number

PAGE 81

81 REFERENCES 1. R.E. Newnham, D.P. Skinner and L.E. Cross, “Connectivity and Piezoelectric Pyroelectric Composites”, Mater. Res. Bull., 13 [5] 525 53 6 (1978). 2. M.J. Haun, P.Moses, T.R. Gururaja, W.A. Schulze and R.E. Newnham, “Transversely Reinfoced 1 3 and 1 3 0 Piezoelectric Composites”, Ferroelectrics, 49 [1] 259 264 (1983). 3. M.J. Haun, R.E. Newnham, ”An Experimental and Theoretical Study of 1 3 and 1 3 0 Piezoelectric PZT Polymer Composites for Hydrophone Applications”, Ferroelectrics, 68 [1] 123 139 (1986). 4. W.A. Smith, A.A. Shaulov, and B.A. Auld, ”Design of Piezocomposites for Ultrasonic Transducers”, Ferroelectrics, 91 [1] 155 162 (1989). 5. W.A. Smith and B.A. Auld, “Modeling 1 3 Composite Piezolelectrics – Thickness Mode Oscillations”, IEEE T Ultrason. Ferr., 38 [1] 40 47 (1991). 6. W.A. Smith, “Modeling 1 3 Composite Piezoelectrics – Hydrostatic Response”, IEEE T Ultrason. Ferr., 40 [1] 41 49 (199 3). 7. W.A. Smith, “Composite Piezoelectrics Utilizing a Negative Poisson Ratio Polymer”, United States Patent 5,334,903 (1994). 8. D.P. Skinner, R.E. Newnham, and L.E. Cross, “Flexible Composite Transducers”, Mater. Res. Bull., 13 [6] 599 607 (1978). 9. T.R. Sh rout, W.A. Schulze, and J.V. Biggers, “Simplified Fabrication of PZT Polymer Composites”, Mater. Res. Bull., 14 1553 1559 (1979). 10. K. Rittenmyer, T. Shrout, W.A. Schulze, and R.E. Newnham, “Piezoelectric 3 3 Composites”, Ferroelectrics, 41 [1] 323 329 (1982). 11. K. Schwartzwalder, H. Somers, and A.V. Somers, “Method of Making Porous Ceramic Articles,” United States Patent 3,090,094 (1963). 12. L.J. Gauckler, A.R. Studer, U.T. Gonzenbach, E. Tervoort, “Processing Routes to Macroporous Ceramics: A Review”, J. Am. Ceram. Soc., 89 [6] 1771 1789 (2006). 13. P. Colombo and M. Modesti, “Silicon Oxycarbide Foams from a Silicone Preceramic Polymer and Polyurethane”, J. Sol Gel Sci. Techn., 14 [3] 103 111 (1999). 14. R. Ramesh, H. Kara, and C.R. Bowen, “Finite E lement Modeling of Dense and Porous Piezoceramic Disc Hydrophones”, Ultrasonics, 43 [3] 173 181 (2005). 15. J.J.C. Busfield, J.R.G. Evans, Z. Fan, H.X. Peng. “Microstructure of Ceramic Foams”. J. Eur. Ceram. Soc., 20 [7] 807 813 (2000).

PAGE 82

82 16. Q. Liu, G. Subhash, A Phenomenological Constitutive Model for Foams under Large Deformations", Poly. Eng. and Sci., 44 [3] 463 473 (2004). 17. E. Ozcivici and R.P. Singh, “Fabrication and Characterization of Ceramic Foams Based on Silicon Carbide Matrix and Hollows Alumina Silic ate Spheres”, J. Am. Ceram. Soc., 88 [12] 3338 3345 (2005). 18. G.J. Qi, C.R. Zhang, H.F. Hu, F. Cao, S.Q. Wang, Y.B. Cao, and Y.G. Jiang, “Preparation of Three Dimensional Silica Fiber Reinforced Silicon Nitride Composites Using Perhydropolysilazane as Precu rsor”, Mater. Lett., 59 [26] 3256 3258 (2005). 19. S. Roberts, “Dielectric and Piezoelectric Properties of Barium Titanate”, Phys. Rev. 71 [12] 890 895 (1947). 20. B. Jaffe, R.S. Roth, and S. Marzullo, “Piezoelectric Properties of Lead Zirconate Lead Titanate So lid Solution Ceramics”, J. Appl. Phys., 25 [6] 809 810 (1954). 21. A.J. Moulson, J.M. Herbert, Electroceramics 2 nd Ed. John Wiley & Sons Ltd. (2003). 22. L.E. Cross, “Ferroelectric Ceramics: Tailoring Properties for Specific Applications”, Ferroelectric Ceramic s, Ed N Setter and E L Colla (1993). 23. T.R. Gururaja, A. Safari, R.E. Newnham and L.E. Cross, “Piezoelectric Ceramic Polymer Composites for Transducer Applications,” Electronic Ceramics edited by L.M. Levinson, 92 128 (2004). 24. D.J. Green, and P. Colombo, “ Cellular Ceramics: Intriguing Structures, Novel Properties and Innovative Applications,” Mater. Res. Bull., 28 [4] 296 300 (2003). 25. P. Colombo and J.R. Hellman, “Ceramic Foams from Preceramic Polymers”, Mater. Res. Innov., 6 [5 6] 260 272 (2006). 26. D.M. Liu “Influence of Porosity and Pore Size on the Compressive Strength of Porous Hydroxyapatite Ceramic”, Ceram. Int., 23 [2] 135 139 (1997). 27. X. Bao, M.R. Nangrejo, and M.J. Edirisinghe, “Preparation of Silicon Carbide Foams Using Polymer Precursor Solutions” J. Mater. Sci., 35 [17] 4365 4372 (2000). 28. P. Colombo and E. Bernardo, “Macro and Micro cellular Porous Ceramics from Preceramic Polymers”, Compos. Sci. Techno., 63 [16] 2353 2359 (2003). 29. A. Rainer, F. Basoli, S. Licoccia, and E. Traversa “Foaming of F illed Polyurethanes for Fabrication of Porous Anode Supports for IT SOFC” J. Am. Ceram. Soc., 89 [6] 1795 – 1800 (2006). 30. D. Weaire, The Kelvin Problem: Foam Structure of Minimal Surface Area CRC Press. (1996). 31. R.M. Stroud, J.W. Long, J.J. Pietron, E.M. Lu cas, M.L. Anderson, K.E. Swyder Lyons, C.I. Merzbacher, D.R. Rolison. “Mesoporous, Microporous and Nanowired: Electron Microscopy of Aerogel Composites”, Microsc. Microanal., 8 [2] 324 328 (2002).

PAGE 83

83 32. X.Y. Deng, X.H. Wang, Z.L. Gui, L.T. Li and I.W. Chen, “Gra in Size Effects on the Hardness of Nanograin BaTiO 3 Ceramics”, J. Electroceram., 10 44 47 (2007). 33. R.W., Siegel, G.E. Fougere, “Mechanical Properties of Nanophase Materials”, NATO Conference, Corfu, Greece, 20 Jun – 2 Jul (1993). 34. L. Wucherer, J. C. Nino F. Basoli, E. Traversa, “Synthesis and Characterization of BaTiO 3 based Foams with Controlled Microstructure” Int. J. Appl. Ceram. Tech. (In press 2008). 35. W.A. Smith and B.A. Auld, “Modeling 1 3 Composite Piezoelectrics: Hydrostatic Response”, Ultra. Ferro IEEE, 40 [1] 40 47 (1991). 36. G. Subhash, Q. Liu, X. L Gao, "Quasistatic and High Strain Rate Uniaxial Compressive Response of Polymeric Structural Foams," Int. J. Impact. Eng., 32 [7] 1113 1126 (2006). 37. T.R. Walter, A. Richards and G. Subhash, “A Unified Phenomenological Model for Compression and Tension Response of Polymeric Foams,” ASME J. of Eng. Mater. and Tech. (in press 2008). 38. P. Colombo, J.R. Hellmann and D.L Shelleman, “Mechanical Properties of Silicon Oxycarbide Ceramic Foams,” Comp. Sci. Tech., 84 [10] 2245 2251 (2001). 39. F.A. Costa Oliveria, S. Dias, M. Fatima Vaz, J.Cruz Fernandes, “Behavior of Open Cell Cordierite Foams under Compression” J. Eur. Ceram. Soc. 26 [1 2] 179 186 (2006). 40. K. Li, X.L. Gao, and G. Subhash, “Effects of Cell Shape and C ell Wall Thickness Variations on the Elastic Properties of Two Dimensional Cellular Solids”, Int. J. Sol. and Struct., 42 [5 6] 1777 1795 (2005). 41. Morgan Technical Ceramics Piezoelectrics Morgan Tec hnical Ceramics, Bedford, OH. http://www.morganelectrocer amic.com/piezomaterials/index.html 42. T. Ogawa, “Poling Field Dependence of Ferroelectric Properties in Barium Titanate Ceramics”, Jpn. J. Appl. Phys., 40 5630 5633 (2001).

PAGE 84

84 BIOGRAPHICAL SKETCH Laurel Wucherer was born in 1986, in Orlando, Florida. Both of her parents are CPAs and her sister, Devon, is cur rently a senior in high school. Since visiting the University of Flo rida (UF) in the spring of 2004, she became fascinated with material science after seeing the interesting research and facilities avail able. After starting her studies, Laurel wanted to experience the best of research and industry and the 3 2 program helped her achieve that by finishing her undergraduate and advanced studies simultaneously. In Dece mber she will be receiving her b achelor ’s and m aster’s degrees in material science and engineering. During that time, she joined Dr. Juan C. Nino’s research group, where she became well acquainted with the life of a research scientist including the frustrations, triumphs and interesting work hour s. In order to maintain her sanity, Laurel became very involved in many extracurricular activities including a sorority, tennis club and dance company. These activities offered her a creative outlet and the opportunity to continue hobbies she enjoyed growi ng up. Laurel has thoroughly enjoyed her time at UF and would not have preferred any other university for her college career. After graduation, she will begin her professional career as a materials engineer at the Walt Disney Company in Orlando, Florida.