Bio-Inspired Nacre-Like Ceramic Matrix Composites

BIO-INSPIRED NACRE-LIKE CERAMIC MATRIX COMPOSITES

Jiacheng Huang

APPROVED BY SUPERVISORY COMMITTEE:

______Majid Minary, Chair

______Dong Qian

______Yonas Tadesse

______Rodrigo Bernal Montoya

Jiacheng Huang

To my family

BIO-INSPIRED NACRE-LIKE CERAMIC MATRIX COMPOSITES

JIACHENG HUANG, BE, MS

DISSERTATION

Presented to the Faculty of

The University of Texas at Dallas

in Partial Fulfillment

of the Requirements

for the Degree of

DOCTOR OF PHILOSOPHY IN

MECHANICAL ENGINEERING

THE UNIVERSITY OF TEXAS AT DALLAS

December 2018

ACKNOWLEDGMENTS

I would like to thank my doctoral mentor, Dr. Majid Minary, for his much appreciated financial

support, workspace, and guidance throughout the course and research of my PhD work. I would

also like to thank my committee members, Dr. Dong Qian, Dr. Yonas Tadesse and Dr. Rodrigo

Bernal Montoya, for their suggestions and support. I am grateful for all my coauthors,

collaborators, and colleagues that helped make this work possible. More importantly, I appreciate

the continued support and encouragement from my family, especially, my wife Huiwen Shi.

I acknowledge Dr. Zhe Xu (UTD), Seyed Soheil Daryadel (UTD), Dr. Ali Behroozfar (UTD),

Seyed Reza Morsali (UTD), Salvador Moreno (UTD), Dr. Mahmoud Baniasadi (UTD), Dr. Xi

Yang (UTD), Dr. Enlong Yang (Visiting Professor in UTD from JXU, China), Martha Serna

(UTD), Dr. Tingge Xu (UTD), Xuemin Wang (UTD) , Dr. Winston Layne (Nanotech, UTD), Dr.

Gordon Pollack, all cleanroom staff (UTD), Dr. Rajarshi Banerjee (UNT), Thomas W. Scharf

(UNT) , William S Rubink (UNT), and Hunter Lide (UNT) for their insightful discussions,

assistance with experimental equipment and computational simulations.

November 2018

BIO-INSPIRED NACRE-LIKE CERAMIC MATRIX COMPOSITES

Jiacheng Huang, PhD The University of Texas at Dallas, 2018

Supervising Professor: Dr. Majid Minary, Chair

Understanding the role of the ductile polymer phase in mechanical behavior of bioinspired hybrid

composites is an important step toward development of materials with damage tolerant properties.

In this dissertation,fabrication and characterization of a bioinspired lamellar composite by

incorporation of a semicrystalline polymer into a freeze-casted scaffold is reported. The elastic

modulus and ductility of the polymer phase can be changed by more than three and fifty five times,

respectively, in addition to forty two folds decrease in modulus of toughness, by thermal annealing

post-processing, after infiltration into the freeze-casted ceramic scaffold. The results show that although polymer phase affects the fracture toughness and flexural behavior of the composite, the drastic changes in mechanical properties of the polymer phase has only marginal effects on the

resulted properties of the composite. In situ SEM experiments and finite element simulation were

used to investigate the deformation mechanism and the effect of the polymer phase on the

distribution of stress in the fabricated composites.

Additionally, a metal-ceramic composite comprised of ~82 vol. % alumina (Al2O3) and ~18 vol.

% nickel (Ni) was fabricated via co-assembly of alumina micro-platelets with Ni particles using

the freeze-casting process followed by the spark plasma sintering (SPS). The SPS processing with

a custom-designed temperature-pressure history resulted in formation of elongated Ni phase

between the lamellar-ceramic phase. Results of the mechanical characterization showed that

inclusion of Ni improveed the flexural strength of the composite by more than 47% compared to

the lamellar ceramic. Additionally, the crack initiation (KIC) and crack growth toughness

increaseed by 20% and 47%, respectively. The inclusion of softer Ni phase did not compromise

the indentation modulus and indentation hardness of the composite compared to the pure ceramic.

Infiltration of a molten metal phase into a ceramic scaffold to manufacture metal-ceramic

composites often involves high temperature, high pressure, and expensive processes. Low-cost processes for fabrication of metal-ceramic composites can substantially increase their applications in various industries. In this disseration, electroplating (electrodeposition) as a low-cost, room- temperature process is demonstrated for infiltration of metal (copper) into a lamellar ceramic

(alumina) scaffold. Estimation shows that energy consumption of this process is less than a few percent of the conventional molten metal infiltration process. Characterization of mechanical properties showed that metal infiltration enhanced the flexural modulus and strength by more than

50% and 140%, respectively, compared to the pure lamellar ceramic. More importantly, metal infiltration remarkably enhanced the crack initiation and crack growth resistance by more than

230% and 510% compared to the lamellar ceramic. The electrodeposition process for development of metal-ceramic composites can be extended to other metals and alloys that can be electrochemically deposited, as a low-cost and versatile process.

vii

TABLE OF CONTENTS

ACKNOWLEDGMENTS ...... v ABSTRACT ...... vi LIST OF FIGURES ...... ix LIST OF TABLES ...... xiii CHAPTER 1 INTRODUCTION ...... 1 CHAPTER 2 EXPERIMENTS AND METHODS ...... 9 2.1 Fabrication and characterization of Lamellar Ceramic Semicrystalline-Polymer Composite .. 9 2.2 Fabrication and characterization of alumina-nickel composite ...... 14 2.3 Fabrication and characterization of Alumina-Copper Composite ...... 18 2.4 Experiment Calculation and simulation ...... 24 CHAPTER 3 LAMELLAR CERAMIC SEMICRYSTALLINE-POLYMER COMPOSITE FABRICATED BY FREEZE-CASTING...... 28 3.1 Results and Discussion ...... 28 3.2 Conclusions ...... 45 CHAPTER 4 ALUMINA-NICKEL COMPOSITE PROCESSED VIA CO-ASSEMBLY USING FREEZE-CASTING AND SPARK PLASMA SINTERING ...... 47 4.1 Results and Discussion ...... 47 4.2 Conclusions ...... 58 CHAPTER 5 MANUFACTURING OF METAL-CERAMIC COMPOSITE THROUGH INFILTRATION OF COPPER INTO FREEZE-CAST ALUMINA SCAFFOLD BY ELECTROPLATING ...... 60 5.1 Results and Discussion ...... 60 5.2 Conclusions ...... 71 CHAPTER 6 OUTLOOK ...... 73 REFERENCES ...... 76 BIOGRAPHICAL SKETCH ...... 82 CURRICULUM VITAE

viii

LIST OF FIGURES

Figure 1.1. Illustration of the toughening mechanisms of in nacre [16]...... 2

Figure 2.1. Illustration of the vacuum infiltration process. Pictures of the samples before polymer infiltration, during polymer infiltration and after polymer infiltration...... 10

Figure 2.2. A photo of the mechanical experiment under the optimal microscope...... 12

Figure 2.3. The illustration of the freeze-casting setup and a photograph of the main parts of the freeze-casting setup...... 15

Figure 2.4. The schematic of the sample cross-section for the electrodeposition process. Red arrows indicate that the deposition starts from the bottom of the ceramic scaffold all the way to the top and finally forms a copper cap at the top the scaffold...... 19

Figure 2.5. The illustration of the electrodeposition process. Copper ions flow freely through the ceramic scaffold until are deposited onto the conductive layer (the cathod) under the ceramic scaffold...... 21

Figure 2.6. 2D plane strain FE model of ceramic platelet-P(VDF-TrFE) composite. Ceramic platelets are shown in yellow, and the polymer is shown in navy. This structure corresponds to a ceramic volume fraction of 49%, a polymer volume fraction of 18% and void fractions of 33%, similar to the experimental values...... 26

Figure 3.1. (A) Schematic representation of the freeze-casting setup. (B) Illustration of the compression process. (C) Illustration of the fabrication process of the composite, including directional freeze-casting, compression, sintering, and vacuum infiltration. (D) Photographs of the freeze-casted scaffold...... 29

Figure 3.2. (A) Atomic structure of a unit of P(VDF-TrFE). (B) Schematic of the semicrystalline morphology of P(VDF-TrFE) before and after annealing. (C) Stress-strain responses of the polymer before and after annealing. The inset shows the magnified view of the low strain region...... 30

Figure 3.3. Mechanical properties of P(VDF-TrFE) polymer extracted from tensile experiment. (A)-(D) Elastic modulus, ductility (failure strain), tensile toughness, and strength...... 31

Figure 3.4. The SEM images show the structure of the composites. (A)-(C) are SEM images of the ceramic scaffold. (D)-(F) are SEM images of the as-made composite. (G)-(I) SEM images of the annealed composite. Scale bars are (A), (D) and (G) 20 m, (B), (E) and (H) 4 m, (C), (F) and (I) 1 m...... 32 µ µ µ

Figure 3.5. The SEM image and EDS maps for an as-made composite. (A) The SEM image of an as-made composite. (B) The overlay EDS map (Aluminum: yellow; Carbon: Red; Silicon: purple). (C)-(E) The EDS map for aluminum, carbon and silicon, respectively. Scale bars are 2 μm...... 33

Figure 3.6. SEM images of the as-made and annealed composites with and without surface functionalization of platelets. (A) and (B) are as-made and annealed composites without ATES surface functionalization, respectively. (C) and (D) are as-made and annealed composites with ATES surface functionalization. (E) and (F) show cross-section of an as- made composite without and with surface functionalization, respectively. Clearly, the surface functionalization results in better interface between the platelets and the polymer...... 34

Figure 3.7. (A)-(B) flexure strength and flexure failure strain of the samples...... 36

Figure 3.8. Characterization of the fracture toughness of the fabricated composites. (A) Representative load-displacement responses from the single edge notched beam (SENB) fracture tests. Note that the measured specimens differ slightly in dimensions. (B) The crack growth resistance curves (R-curves) obtained from the SENB responses. (C) Fracture toughness KIC vs flexural strength (n = 3). The vertical and horizontal error bars are the standard deviation...... 38

Figure 3.9. Fracture toughness responses of non-functionalized specimens. (A) Load– displacement responses from single edge notched beam (SENB) fracture tests. Note that the measured specimens differ slightly in dimensions. (B) Crack growth resistance curves (R-curves) obtained from the SENB fracture tests in A...... 39

Figure 3.10. Snap-shots from the in situ SEM fracture toughness test for the (A) as-made composite, (E) annealed composite and (H) ceramic scaffold. Details of fracture section are shown in (B, C, D) for as-made composite, (F) and (G) for annealed composite and (I, J) for ceramic scaffold. Scale bars are (A, E, H) 400 m, (B) 50 m, (C) 3 m, (D) 1 m and (F, G, I, J) 2 m. Green arrows indicate the stretched polymer, red arrows indicate the fractured platelets...... µ ...... µ µ ...... µ 40 µ Figure 3.11. 2D plane strain FE model of ceramic platelet-P(VDF-TrFE) composite. Ceramic platelets are shown in yellow, and polymer is shown in navy. This structure corresponds to a ceramic volume fraction of 49%, a polymer volume fraction of 18% and void fractions of 33%, similar to the experimental values...... 42

Figure 3.12. The von-Mises stress partition at 0.2% applied nominal strain. (A-C) Ceramic scaffold. (D-F) As-made composite. (G-I) Annealed composite...... 44

Figure 3.13. Simulated von Mises stress distribution in the scaffold and composites at (A) 0.2% strain, and (B) 1.0% strain...... 46

Figure 4.1. The fabrication process of the metal - ceramic composite. The process includes directional freeze-casting, freeze-drying, compaction and stacking, followed by the spark plasma sintering (shown in Figure 4.2)...... 48

Figure 4.2. (A) Temperature- pressure profile during the SPS process. (B) The schematic of the internal structure of the composite during the SPS process...... 49

Figure 4.3. (A) - (C) Lamellar ceramic; (D) - (F) metal-ceramic composite. The brighter regions in the composite are nickel phase and the darker regions are the alumina phase. (A) and (D) are SEM images of the samples before sintering, (B) and (E) are cross-section images, (C) and (F) are top-view images...... 50

Figure 4.4. (A) Flexural stress-strain responses from the 3-point flexural experiment. (B) Comparison of the flexural strength. (C) Comparison of the strain at the maximum stress. Fracture toughness experiments were also carried out to quantify and compare the fracture resistance properties of the lamellar ceramic and metal-ceramic composite. At least three samples were tested for each sample type. For this purpose, the single-edge notched beam (SENB) was used. The dimensions of the specimens (width, length, thickness, and pre-notch length) followed the ASTM standard E1820...... 52

Figure 4.5. Fracture toughness characterization results. (A) Representative load–displacement responses from the single-edge notched beam (SENB) fracture experiment. (B) The crack growth resistance (R-curve) responses obtained from the SENB experiment. (C) The fracture toughness KIC versus the flexural strength. The vertical and horizontal error bars are the standard deviation...... 54

Figure 4.6. SEM micrographs from the fractured area of the SENB samples. (A)-(F) The lamellar ceramic and (G)-(L) metal-ceramic composite. Red arrows in (H) indicate the crack deflection...... 56

Figure 4.7. Nanoindentation responses of the samples. The insets in (A) and (B) show the orientation of the platelets with respect to the indenter tip. (A) and (B) are representative force vs. indentation depth responses. (C) and (D) show the indentation modulus and indentation hardness, respectively...... 57

Figure 4.8. Ashby plots for (A) fracture toughness vs. density and (B) flexural modulus vs. flexural strength...... 58

Figure 5.1. The illustration of the electrodeposition setup and photograph of the setup...... 61

Figure 5.2. The fabrication process of the electrodeposited metal-ceramic composite. (A) The process steps include directional freeze-casting of alumina micro-particles and freeze- drying, compression, sintering, and metal infiltration by electroplating. The photographs in the bottom panel of A show actual specimens. The original alumina ceramic scaffold has a white-grey color, and once it was infiltrated with Cu, it appears as orange color. (B)

The schematic illustrates electroplating of Cu inside the pores of the ceramic scaffold. The bottom side of the ceramic scaffold is placed on the cathode surface. Blue, grey, and orange colors indicate alumina, silica, and Cu, respectively...... 61

Figure 5.3. False-colored SEM images of the electrodepositioned sample, cross-section area. Orange area indicates the copper while the gray and blue areas show the ceramic scaffold...... 63

Figure 5.4. (A)-(B) Cross-section scanning electron microscope (SEM) images of the ceramic scaffold before sintering and after sintering, respectively. (C) Cross-section SEM image of the metal-ceramic composite. Lighter regions are metal (Cu) and the darker regions are ceramic micro-platelets. The inset in C demonstrates the electrical conductivity of the composite...... 63

Figure 5.5. (A) Flexural stress-strain responses from the 3-point flexural experiment. (B) The flexural modulus versus the flexural strength. The vertical and horizontal error bars are the standard deviation. (C) Comparison of the strain at the maximum stress. The error bars are the standard deviation...... 66

Figure 5.6. Fracture toughness characterization results. (A) Representative load-displacement responses from the single-edge notched beam (SENB) fracture experiment. (B) The crack growth resistance (R-curve) responses obtained from the SENB experiment. (C) The fracture toughness KIC versus the flexural strength. (D) The comparison between the crack initiation toughness and the crack growth toughness. The error bars are the standard deviation. The arrows and the numbers show % improvement of crack growth vs. crack initiation toughness...... 67

Figure 5.7. SEM micrographs from the fractured area of the metal-ceramic composite after SENB experiment. (A) Multiple crack deflections ahead of the crack tip. (B) - (G) SEM images show various toughening mechanisms in the composite that include (B) crack branching, (C) friction, (D) and (E) platelet pullout, and (F) and (G) plasticity and failure of the metal phase. The scale bar in (B) - (G) is 2 m...... 69

Figure 5.8. Ashby plots for (A) the fracture toughness KICµ verses density, and (B) for modulus vs. strength [49]. Properties of the composite in this work is overlaid on the plots using a red solid circle...... 70

Figure 6.1. Ashby plot for fracture toughness vs. modulus. Modulus vs fracture toughness...... 73

Figure 6.2. Ashby plot for specific modulus vs. specifc strength...... 74

Figure 6.3. SEM images of the sample before the electrodeposition and after the electrodeposition of metal ductile layer...... 75

xii

LIST OF TABLES

Table 3.1. Mechanical property of the ceramic, composite and anneal composite with surface functionalization...... 35

xiii

CHAPTER 1

INTRODUCTION

High strength and toughness in most conventional man-made structural materials are

mutually exclusive [1]. Natural biocomposites such as nacreous part of sea-shells and bone use

sophisticated design combining stiff biominerals and soft biopolymers in hybrid multiscale

hierarchical architecture with intricate interfaces to realize simultaneous superior strength and

toughness, which result in “damage-tolerant” structural materials [2-14]. More specifically, in

“nacre” several toughening mechanisms including mineral bridges, friction between nano-

asperities on mineral surface, deflection of cracks at the biopolymer- aragonite interfaces, plastic

deformation of the biopolymer and possible platelet pull-out have been proposed [15-17].

Incorporation of several or all of these toughening mechanisms are required to achieve such

performance in synthetic hybrid composites [18, 19]. Layer-by-layer and vacuum filtration

approaches have been demonstrated to fabricate lamellar composite films [20-26]. The quest

toward development of damage-tolerant synthetic materials is driven by their numerous

applications as lightweight structural materials [8, 27-29]. Freeze-casting and magnetically assisted slip-casting have been two of the major routes to achieve macroscale hierarchical lamellar scaffold, similar to the brick-and-mortar structure in “nacre” [30-37]. In these methods, the porous

scaffold is subsequently infiltrated with a polymer phase or monomers that are converted to

polymer in situ to achieve a hybrid composite.

The polymer phase in a hybrid composite imparts various toughening mechanisms such as

crack bridging, plastic deformation, void formation, etc., depending on physical properties of the

polymer, such as its stiffness, strength, toughness, and ductility [38, 39]. Although the properties

Figure 1.1. Illustration of the toughening mechanisms of in nacre [16]. Reprinted by permission from Springer Nature: Springer, Nature (Bioinspired Structural Materials, Ulrike G.K. Wegst, Hao Bai, Eduardo Saiz, Antoni P. Tomsia & Robert O. Ritchie) Copyright 2015.

of the ceramic phase, and its microstructure such as porosity, and their overall impact on the behavior of composites have been extensively investigated, study of the role of different polymer phases have been limited in comparison. Recently, the role of polymer phase in the mechanics of nacre-like composites was investigated by employing three different polymers, including a soft and weak elastomer, a strong, stiff and brittle thermoplastic, and a tough polymer of intermediate strength and stiffness [40]. Along this line, it would be desirable to achieve various mechanical properties within the same polymer system, for example by post-processing or various cross-

linking. This would allow for systematic study of the role of the polymer phase in the deformation

mechanism and mechanical properties of the nacre-like composites.

In this dissertation, a semicrystalline polymer was used as the ductile phase for infiltration

of freeze-casted ceramic scaffold, and it was shown that mechanical properties of the polymer phase can be tuned by thermal annealing after infiltration. Thermal annealing process leads to enhancement of the crystalline phase of the polymer, which results in three times enhancement in the elastic modulus, more than fifty five folds decrease in ductility, and forty two folds decrease in the modulus of toughness. The mechanical properties of the fabricated composites were evaluated using flexure, and fracture toughness experiments, combined with in situ and scanning electron microscopy (SEM) experiments and numerical simulation.

Inclusion of a ductile metal phase into a ceramic matrix can improve its fracture toughness, in addition to other possible desirable properties such as electrical conductivity [41-43]. Such metal-ceramic composites have applications and market demand in various industries including automotive, aerospace, oil and defense, in products such as high performance wear-resistance parts, cutting tools, light-weight structural composites, and aero-engine components [44, 45].

Metal inclusion can be either in the form of randomly distributed particles, layered

(lamellar), or nature-inspired brick-and-mortar architecture. Based on the desired structure, there are various processes for preparation of metal-ceramic composites. They include (but not limited to) stir-casting, tape-casting and tape-sintering, squeeze-casting, thermal spraying of multilayered ceramic-metal, hot pressing, freeze-casting, and spark-plasma sintering, among others [41-44, 46-

49]. In certain cases, a combination of these processes is used to fabricate the metal-ceramic composite. Often it is desired that the metal phase to have small volume percentage (ideally less

than 20%) to achieve a good balance of strength and toughness in the composite. This requirement

imposes further processing challenges, in particular for lamellar and brick-and-mortar architecture, since the pore size to be filled with metal mortar is small.

The overall toughening mechanism of metal inclusion into a ceramic is believed to be a combination of several factors including (but not limited to) energy dissipation associated with the plastic deformation of the metal phase, crack tip blunting, deflection of the crack at the metal- ceramic interface, and reduction of stress concentration by distribution of the crack to a larger area

[41-43]. In the case of lamellar and brick-and-mortar architecture, the ductile metal phase can also function similar to a lubricant, which relieves stress concentration by allowing for limited sliding.

The ductile metal phase may also provide crack bridging as an extrinsic crack-tip shielding mechanism [47, 50]. These mechanism overall are manifested in the form of a rising R-curve behavior.

In several early works alumina-Ni composites have been reported, in which Ni lamina or

Ni inclusions were used for toughening of alumina [41-43]. For example, tape casting and hot- pressing was used to fabricate a Ni/alumina composite in which the Ni-lamina had similar thickness to alumina [41]. It was observed that the strength and toughness of the laminates were greatly improved in comparison to that of monolithic alumina. Similarly, it was reported that inclusion of ~13% of nickel (particle size of ~0.5 microns) doubled the fracture toughness of alumina in alumina/Ni composite. The composite was prepared by the selective reduction process, in which powder mixtures of alumina and Ni oxide were reduced selectively to alumina and Ni

[43]. The toughening mechanics was identified as the plastic deformation of Ni particles concluded from the observed strained Ni particles in the post-failure analysis. An alternative approach to

introduce metal inclusions into ceramic matrix is pre-coating ceramic micro-platelets with a metal using electroless plating, followed by subsequent assembly of the platelets. For example, a Ni- alumina composite was reported in which initially Ni was coated on the surface of alumina micro- platelets using the electroless plating. Subsequently, the Ni-coated micro-platelets were aligned by a rotating magnetic field, taking advantage of ferromagnetic properties of Ni. In the final step, the assembled Ni-coated ceramic scaffold (green body) was sintered using spark plasma sintering

(SPS) [49].

There have also been several reports on “nacre-like” composites and lamellar ceramic- metal composites [31, 47, 48, 50]. The metal mortar is often introduced into a porous ceramic scaffold using melt infiltration (also called liquid metal infiltration) process. Infiltration of metals into ceramic scaffold with small pore size may be inherently challenging because of the poor wetting between most metals and ceramics. The infiltration process becomes increasingly difficult as the volume of the ceramic scaffold increases. An alumina-copper composite containing 90% ceramic, and 10% copper has been also reported, in which copper flakes were co-assembled with alumina platelets and subsequently densified [31].

The RROLM (“rolling of randomly orientated layer-wise materials”) manufacturing methodology was recently introduced. This process enables a layer-wise aligned distribution of micro-scale intermetallic particles within a ductile Al matrix [51]. The results showed 70% and

90% improvements in strength and toughness with respect to the control sample.

Despite favorable properties of metal-ceramic composites, they have not achieved full potential in application, much due to processing cost and challenges [44, 45, 52-54]. There are various processes for fabrication of metal-ceramic composites. They include (but not limited to) stir-

casting, tape-casting and tape-sintering, squeeze-casting (melt infiltration), thermal spraying, hot pressing, freeze-casting and spark-plasma sintering, among others [41-44, 46-48, 52, 55, 56]. Melt infiltration requires high-temperature, and high-pressure because of the poor wetting between most metals and ceramics. The infiltration process becomes increasingly difficult as the volume of the ceramic scaffold increases. Low-cost processes for fabrication of metal-ceramic composites can substantially increase their applications in various industries including automotive, aerospace, oil and defense, in products such as high performance wear-resistance parts, cutting tools, light-weight structural composites, and aero-engine components.

The metal phase in the composite can be either in the form of randomly distributed particles, layered (lamellar), or nature-inspired brick-and-mortar architecture. Inclusion of the ductile metal phase into a ceramic matrix generally improves the fracture toughness of the ceramic, in addition to other possible desirable properties such as electrical conductivity [42, 43]. The overall toughening mechanism of metal inclusion into ceramic is believed to be a combination of several factors including the plastic deformation of the metal phase, crack tip blunting, deflection of the crack at the metal-ceramic interface, and reduction of stress concentration by distribution of the crack to a larger area [41-43]. In the case of lamellar and brick-and-mortar architecture, the ductile metal phase can also function similar to a lubricant, which relieves stress concentration by allowing for limited sliding. The ductile metal phase may also provide crack bridging as an extrinsic crack-tip shielding mechanism [31, 47, 48, 50, 56]. These mechanisms are overall manifested in the form of a rising R-curve behavior, which shows the crack growth toughness versus the crack length.

Nickel-alumina composites, in which Ni lamina or Ni inclusions were used for toughening

of alumina, have been reported [41-43]. Tape-casting and hot-pressing was used to fabricate

Ni/alumina composites [41]. It was observed that the strength and toughness of the laminates were

greatly improved in comparison to that of monolithic alumina. In another study, a metal-ceramic

composite was prepared by the selective reduction process, in which powder mixtures of alumina

and Ni oxide were reduced selectively to alumina and Ni [43]. An alternative approach to introduce metal inclusions into ceramic matrix is pre-coating ceramic micro-platelets with a metal using electroless plating, followed by subsequent assembly of the platelets. For example, a Ni-alumina composite was reported in which initially Ni was coated on the surface of alumina micro-platelets using electroless plating. Subsequently, the Ni-coated micro-platelets were aligned by a rotating magnetic field, taking advantage of ferromagnetic properties of Ni. In the final step, the assembled

Ni-coated ceramic scaffold (green body) was sintered using spark plasma sintering (SPS)[49].

Electrodeposition (electroplating) is an electrochemical process that has been used for

decades for applications in various industries, notably surface coating and electronics. More

recently, electrodeposition has been used as a versatile process for nanomaterial synthesis such as

nanoparticles and template-assisted synthesis of nanowires, as well as for microscale additive

manufacturing of metals [51, 57, 58]. In this disseration, electrodeposition and freeze-casting

processes were combined to fabricate a metal-ceramic composite with a fine laminated

microstructure. Copper was electroplated at room-temperature into lamellar alumina scaffold that

was fabricated by freeze-casting. In electroplating, copper ions (Cu2+) dissolved in an aqueous

electrolyte transport through the gaps inside the ceramic scaffold, and are reduced to solid metal

phase by accepting electrons. Estimation shows that energy consumption of this process is less

than few percent of the conventional molten metal infiltration process, in which large amount of energy is applied to melt the metal and squeeze it under high pressure into preform ceramic scaffold.

CHAPTER 2

EXPERIMENTS AND METHODS

2.1 Fabrication and characterization of Lamellar Ceramic Semicrystalline-Polymer Composite

2.1.1 Materials

Ceramic micro-platelets (Rona Flair® White Sapphire) were obtained from Merck Ltd.

Silica nanoparticle were obtained from Allied high tech products, Inc. Methanol (CAS number 67-

56-1, >= 99.9% purity), ATES (3-Aminopropyl, triethoxysilane, 99%, CAS Number 919-30-2) and DMF (N, N-Dimethylformamide, CAS 68-12-2) were purchased from Sigma Aldrich. P(VDF-

TrFE) was obtained from Piezotech S.A.S, France. Acetone (BDH, Cat No. BDH1101-1LP) was used to dilute the mixture of P(VDF-TrFE) and DMF. Alginate (PROTANAL® LF10/60FT) was purchased from FMC Corporation.

2.1.2 Preparation of the ceramic slurry

70 ml water-based slurry comprised of 16.5 g sapphire platelets, 2.5 g silica suspension and 1.75 g alginate was prepared. Mixture was initially mixed by a vortex mixer for 30 minutes, and was further stirred by a hotplate stirrer for an additional 12 hours.

2.1.3 Freeze-casting and freeze-drying

The slurry was poured into an acrylic mold with a copper bottom plate. Mold was then placed on top of the cold finger of a homemade freeze casting setup. The other end of the cold finger was submerged into a liquid nitrogen container. The frozen slurry was dried in a

LABCONCO FreeZone2.5 freeze dryer for 48 hours.

2.1.4 Vacuum infiltration process

Figure 2.1. Illustration of the vacuum infiltration process. Pictures of the samples before polymer infiltration, during polymer infiltration and after polymer infiltration.

In the vacuum infiltration process, polymer solution or ATES solution waspoured onto the

sintered ceramic scaffold inside a vacuum chamber. After that, the vacuum chamber was connected to one atmosphere pressure. The atmosphere pressure pushed the solution into the scaffold and created a fully infiltrated composite polymer ceramic composite.

2.1.5 Sintering

Freeze-dried samples were compressed by a Model 4386 Bench Top Laboratory Manual

Press (Carver, Inc.) under 2 tons-force at room temperature for one hour. Subsequently, a ST-

1700C-445 high temperature box furnace (Sentro Tech Company) was used to sinter the ceramic

scaffold. The process was started by preheating the compressed sample at 600 to remove the

organic (Alginate) phase. The preheated sample was further compressed by a manual℃ press under

one ton-force for 30 minutes to further reduce the porosity of the sample. Subsequently, the

compressed sample was sintered in the furnace at 1600 furnace for three hours.

℃ 2.1.6 Preparation of P(VDF-TrFE) solution and polymer-ceramic composite

10 g of 25% P(VDF-TrFE) solution was comprised of 2.5 g of P(VDF-TrFE) and 7.5 g

DMF. The mixture was stirred by a hotplate/stirrer for 12 hours at 70 °C. The sintered ceramic

scaffold was cut into smaller pieces by a diamond saw. Vacuum infiltration process was used to

infiltrate the ATES solution into sintered ceramic to functionalize surfaces of the ceramic platelets.

To prepare the ATES solution, 15 ml of DI water was mixed with 5 ml of methanol (3:1volume ratio) in a beaker. To this solution, 2 ml ATES was added and stirred for 1 hour on a hotplate/stirrer. Ceramic samples were then infiltrated in ATES solution at 40 for 1 hour.

Surface functionalized scaffold was dried in an oven at 40 for 2 days. Then, another℃ vacuum infiltration step was applied to infiltrate P(VDF-TrFE) solution℃ into the scaffold. Polymer infiltrated samples were dried in the oven at 50 for 48 hours. Annealed composites were obtained by heat-treatment of the infiltrated composite℃ at 135 for 4 hours, following annealing the process for P(VDF-TrFE) [59, 60]. ℃

2.1.7 Preparation of test specimens

Specimens for fracture toughness test and flexure test were prepared by a diamond saw, followed by grinding sequentially with 180, 320, 600, and 4000 sand papers. Notches on SENB

(single-edge notched beam) specimens were generated by a diamond saw followed by sharpening

by a razor blade.

2.1.8 Mechanical characterization

Figure 2.2. Aphoto of the mechanical experiment under the optimal microscope.

3-point flexural test and in-situ SEM fracture toughness experiments were performed to investigate mechanical properties of the ceramic scaffold and composites. The in-situ SEM test

was performed using a MTI/FULLAM SEMTester (MTI Instruments, Inc.) inside a LEO 1530 VP

FE-SEM. For in-situ SEM fracture toughness test, specimen size followed the ASTM standard for

fracture toughness test [61]. The fracture toughness samples were ~4.2 mm wide, ~2 mm in depth,

and 20~30 mm long. Test span was 17.2 mm. Displacement speed was 0.01 mm/min. For 3-point

flexural test, samples were ~2 mm in width, 0.5~1 mm in depth and ~20 mm in length. Test span

was 9.8 mm. Loading speed was 0.01 mm/min.

2.1.9 Estimation of the porosity

The porosity of the ceramic scaffold was estimated using the total mass , volume

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ratio of the solid content, total volume of the sample, and the density of each𝑚𝑚 component,

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 according to equations 1-5, below: 𝑉𝑉

= + (1)

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑝𝑝𝑝𝑝 𝑙𝑙𝑙𝑙 𝑚𝑚 = 𝑚𝑚 × 𝑚𝑚 + × (2)

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑝𝑝𝑝𝑝 𝑝𝑝𝑝𝑝 𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙 𝑚𝑚 In the 𝑉𝑉above𝜌𝜌 equatgions𝑉𝑉 𝜌𝜌 , subscriptions pl and lp are for the platelet (Al2O3) and liquid

phase (SiO2), respectively. Ceramic scaffold comprised of 90 vol% of platelets and 10 vol% of

liquid phase (SiO2).

= 90% × × + 10% × × (3)

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑝𝑝𝑝𝑝 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑙𝑙𝑙𝑙 𝑚𝑚 = 𝑉𝑉 𝜌𝜌 𝑉𝑉 𝜌𝜌 %× × (4) 𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑉𝑉𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 90 𝜌𝜌𝑝𝑝𝑝𝑝+10% 𝜌𝜌𝑙𝑙𝑙𝑙 = (5) 𝑉𝑉𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡−𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑉𝑉 3 3 For Al2O3 a density of 3.98 g/cm and for silica a density of 2.196 g/cm were used [62,

63]. Porosity of the infiltrated composite was calculated as following:

= (6)

𝑐𝑐ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝜌𝜌 𝜌𝜌 = − 𝜌𝜌 × (7)

𝑚𝑚𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝜌𝜌𝑐𝑐ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑉𝑉𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡′

= (8) 𝑚𝑚𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑉𝑉𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝜌𝜌𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 = × (1 ) (9)

𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑉𝑉 𝑉𝑉 ′ =− 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 (10) 𝑉𝑉𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡−𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶−𝑉𝑉𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝐶𝐶−𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃−𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑉𝑉𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 The porosity of the composite was calculated using the total mass and total volume

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 after polymer infiltration. For P(VDF-TrFE), a density of 1.87𝑚𝑚 g/cm3′ was used. The

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑉𝑉estimation′ showed that the ceramic scaffolds had a porosity of ~49%. After polymer infiltration,

the composites had a porosity of ~33%, and a polymer content of ~18 vol%.

2.2 Fabrication and characterization of alumina-nickel composite

2.2.1 Materials preparation

Ceramic micro-platelets (Rona Flair® White alumina) were obtained from Merck Ltd.

Alginate (PROTANAL® LF10/60FT) was purchased from FMC Corporation. Nickel powder with

an average particle size of ~2-3 micron was purchased from Atlantic Equipment Engineers, Inc.

Initially, 11.7 g alumina platelets and 5.8 g nickel powder were thoroughly dry-mixed for 2 days

using a roller mixer with 8-mm-diameter tungsten carbide milling balls under a nitrogen gas

environment. Finally, ~70 ml DI water-based slurry comprised of 17.5g alumina platelets and

nickel power and 1.75 g alginate was prepared. The slurry was stirred on a hotplate using a

magnetic stir bar for more than 12 hours. Copper plating electrolyte was purchased from Transene

Inc.

2.2.2 Freeze-casting

In the freeze-casting process, the slurry was poured into an acrylic mold with a copper bottom plate. The mold was then placed on top of the cold finger of a homemade freeze-casting

setup. The other end of the cold finger was submerged into a liquid nitrogen container. The frozen

slurry was dried in a Labconco FreeZone2.5 freeze dryer for 48 hours. The freeze-cast green body was then compressed in the direction perpendicular to the direction of the growth of ice crystals to reduce the porosity of the green body. After compression, the green body was cut into several pieces and was stacked into a thicker sample.

Figure 2.3. The illustration of the freeze-casting setup and a photograph of the main parts of the freeze-casting setup.

2.2.3 Spark plasma sintering

The spark plasma sintering (SPS) process was used on the stacked sample with a model

10-4 SPS unit from Thermal Technology LLC, Santa Rosa, CA, USA. The specimen was placed in a 20-mm-diameter graphite die lined with graphite foil. Two graphite punches with 19.7-mm-

diameter were inserted from the top and bottom to seal the scaffold inside the graphite die with

two pieces of graphite foil placed between the sample and the punch. The graphite foil provided a

better seal and prevented the sample from sticking to the die and punches. The temperature was

measured by an infrared pyrometer targeting the base of a pre-cut bore in the die. The pre-cut bore

was centered on the side of the die 5 mm away from the inner wall. Since oxidation of Ni was

undesirable, the entire SPS process began with a vacuum purge to 2×10-2 torr. The burn-out and

sintering was conducted in an argon environment. The final processed sample was a 20-mm-

diameter pellet with a thickness of approximately 3.5 mm. Both the lamellar ceramic and the metal-

ceramic composite went through all the processes elaborated above.

2.2.4 Specimen preparation

Specimens for the fracture toughness experiment and 3-point flexure experiment were

prepared by a diamond saw, followed by grinding sequentially using sand papers with grit sizes

180, 320, 600, and 4000. Notches on the SENB (single-edge notched beam) specimens were

generated by a diamond saw followed by sharpening by a razor blade [31, 49, 64]. Each SENB

sample was further polished using a polishing cloth and 1 µm diamond suspension using a M-

Prep 5 polisher (Allied High Tech Products, Inc.).

2.2.5 Mechanical characterization

Three-point flexural test and fracture toughness experiments were performed to investigate mechanical properties of the fabricated samples. The experiments were performed using a

MTI/Fullam SEMTester (MTI Instruments, Inc.). For the fracture toughness test, specimen size followed the ASTM standard E1820 [61]. The fracture toughness samples were ~2.5 mm wide,

~1.25 mm in thickness, and 20 mm long. The test span was 9.8 mm. For the 3-point flexural test, samples were ~1.1 mm in width, ~0.6 mm in depth, and 20 mm in length. Test span was 9.8 mm.

The displacement speed was 0.1 mm/min in both experiments.

2.2.6 Estimation of the porosity

The porosity of the composite sample was estimated using the total mass , volume

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 ratio of the solid content, total volume of the sample, and the density of each𝑚𝑚 component,

𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 according to equations 11-14, below: 𝑉𝑉

= + (11)

𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑚𝑚𝑆𝑆𝑆𝑆 𝑚𝑚𝑁𝑁𝑁𝑁 = × + × (12)

𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑉𝑉𝑆𝑆𝑆𝑆 𝜌𝜌𝑆𝑆𝑆𝑆 𝑉𝑉𝑁𝑁𝑁𝑁 𝜌𝜌𝑁𝑁𝑁𝑁 in which, subscriptions Sa and Ni are for sapphire platelet (Al2O3) and nickel (Ni), respectively.

Metal-ceramic composite sample comprised of ~82 vol. % of alumina platelets and ~18 vol. % of

nickel particles.

= 82% × × + 18% × × (13)

𝑚𝑚𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑉𝑉𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝜌𝜌𝑆𝑆𝑆𝑆 𝑉𝑉𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝜌𝜌𝑁𝑁𝑁𝑁

= (14) 𝜌𝜌𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆−𝜌𝜌𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝜌𝜌𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 3 3 For Al2O3 a density of 3.98 g/cm and for nickel a density of 8.91 g/cm were used. The

estimation showed that the lamellar ceramic had a porosity of ~8.22 vol. % with an overall density

of 3.65 g/cm3, the metal-ceramic composite had a porosity of ~11.29 vol. % with an overall density

of 4.32 g/cm3.

2.3 Fabrication and characterization of Alumina-Copper Composite

2.3.1 Materials preparation

Ceramic micro-platelets (Rona Flair® White alumina) were obtained from Merck Ltd.

Alginate (PROTANAL® LF10/60FT) was purchased from FMC Corporation. Copper plating

electrolyte was purchased from Transene Inc. 70 ml DI water-based slurry comprised of 16.5 g alumina micro-platelets, 1 g silica nanoparticles, and 1.75 g alginate was prepared. The slurry was stirred on a hotplate using a magnetic stir bar for more than 12 hours.

2.3.2 Freeze-casting

For the freeze-casting process, the slurry was poured into an acrylic mold with a copper bottom-plate. The mold was then placed on top of the cold finger of a homemade freeze-casting

setup. The other end of the cold finger was submerged into a liquid nitrogen container. After

freezing, the frozen slurry was dried in a Labconco FreeZone2.5 freeze dryer for 48 hours. The

freeze-cast green body was then compressed in the direction perpendicular to the growth direction

of ice crystals to reduce the porosity of the green body. After compression, the green body was cut

into several pieces and was stacked into a thicker sample.

2.3.3 Sintering

Freeze-casted samples were compressed and heated inside a ST-1700C-445 high temperature box furnace (Sentro Tech). The process was started by preheating the compressed sample at 600 to remove the organic (Alginate) phase. The preheated sample was further compressed to ℃reduce the porosity. Subsequently, the compressed sample was sintered in the furnace at 1600 for 3 hours.

℃ 2.3.4 Electrodeposition

The sintered ceramic scaffolds were cut into ~2 mm thin pieces by a precision low-speed saw (ALLIED HIGH TECH PRODUCT. INC). The ceramic scaffold and the glass slide were coated with gold on their one side by sputter coating, and then assembled with their conductive

surfaces facing each other. After that, the area around the ceramic scaffold on the glass was sealed by an epoxy glue. The assembly was immersed in an electrolyte bath. During the vacuum infiltration process, copper sulfate electrolyte was poured onto the sample in a vacuum chamber.

The chamber was then connected to the atmospheric pressure to facilitate infiltration of the electrolyte into the scaffold. The ceramic scaffold was connected to the cathode and a copper mesh was connected to the anode of the potentiostat. A reference electrode was used to form a 3- electrode cell during the electrodeposition process. Copper sulfuric acid solution was used as the electrolyte bath (TRANSENE COMPANY INC), and the electrodeposition process was controlled by a VERSASTAT4-200 (ADVANCED MEASUREMENT TECHNOLOGY). A constant DC voltage of -0.2 V was applied during electrodeposition. In this work, the electrodeposition process was continuously run for several days until the scaffold was fully infiltrated by copper.

2.3.5 Calculation of the process energy

In conventional metal-ceramic composite synthesis process, the furnace is heated several hundred degrees above the melting point of the metal phase to facilitate metal infiltration into the ceramic scaffold. Furthermore, the furnace needs to be water-cooled throughout the synthesis process, and the process is conducted under high vacuum condition. Here, an estimation of energy consumption is made for comparison of molten metal infiltration and electrodeposition processes.

A process step based on the literature for other metals was assumed [65] . It was considered a furnace (with a rated power of 40 kW) to be heated up to ~1280 °C, at 10 °C/min ramp rate, and held at that temperature for 60 minutes. The total energy consumption (W) can be estimated as:

= + , in which and indicate the energy consumption during

𝑊𝑊 𝑊𝑊ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑊𝑊ℎ𝑜𝑜𝑜𝑜𝑜𝑜 𝑊𝑊ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑊𝑊ℎ𝑜𝑜𝑜𝑜𝑜𝑜

the heating phase and the hold phase, respectively, (W = 298 MJ). Based on the literature, the

estimated mass and volume of the infiltrated metal was assumed to be 843 and 7550 g, 3 respectively. Hence, the consumed specific energy and energy density for the infiltrated𝑐𝑐𝑐𝑐 metal can

be calculated as: = ~39 / and = ~350 / . In 𝑊𝑊 𝑊𝑊 3 𝑚𝑚 𝑉𝑉 the electrodeposition𝑆𝑆𝑆𝑆𝑆𝑆 process,𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒under𝑒𝑒𝑒𝑒 the applied𝑘𝑘 potential𝐽𝐽 𝑔𝑔 𝐸𝐸of𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 - 0.2 V,𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 the average current𝑘𝑘 during𝑘𝑘 𝑐𝑐𝑐𝑐 the

process was ~10 mA. The power (P) and the total energy consumption (W) during

electrodeposition process was calculated as: = × = 0.2 × 0.01 = 0.002 , and = ×

= 0.002 × 7 × 24 × 3600 = 1209.6 . The𝑃𝑃 volume𝑈𝑈 𝐼𝐼 and mass of the electrodeposited𝑊𝑊 𝑊𝑊 copper𝑃𝑃

was𝑡𝑡 0.75 and 6.71 g. Hence, the specific𝐽𝐽 energy and energy density was calculated to be: 3 𝑐𝑐𝑐𝑐 = ~180 / , and = ~1.6 . This 𝑊𝑊 𝑊𝑊 𝑘𝑘𝑘𝑘 3 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑚𝑚𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝐽𝐽 𝑔𝑔 𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐 estimation shows that the energy consumption in the electrodeposition process is more than 200- fold smaller than the molten metal infiltration process.

2.3.6 Specimen preparation

Specimens for the fracture toughness experiment and 3-point flexure experiment were

prepared by a diamond saw, followed by grinding sequentially using sand papers with grit sizes of

180, 320, 600, and 4000. Notches on SENB (single-edge notched beam) specimens were generated by a diamond saw followed by sharpening by a razor blade [31, 49, 64]. Each SENB sample was further polished using a polishing cloth and 1 µm diamond suspension using a M-Prep 5 polisher

(Allied High Tech Products, Inc.).

2.3.7 Estimation of the porosity

The porosity of the electrodeposited metal-ceramic composite was estimated using the total

density of the ceramic scaffold , volume ratio of the ceramic solid content, density of the

ce CMC composite, and the density of eachρ component, according to equations 15-22, below: ρ

m = m + m (15)

Ce Al Si m = V × + V × (16)

Ce Al ρAl Si ρSi Subscriptions Ce, Al and Si are for ceramic phase, alumina platelets (Al2O3), and silica nanoparticles (SiO2), respectively. Subscriptions of CMC and Cu are for the electrodeposited

metal-ceramic composite and copper phase, respectively. Ceramic phase comprised of 90 vol. %

of alumina platelets and 10 vol. % of silica nanoparticles.

m = 90% × V × + 10% × V × (17)

Ce Ce ρAl Ce ρSi = 0.9 + 0.1 (18)

ρCe ρAl ρSi Porosity = 1 (19) ρScaffold Scaffold − ρCe Vol. % = (20) ρCMC−ρScaffold Cu ρCu Porosity = 1 Vol. % Vol. % (21)

CMC − Cu − Ce Porosity = 1 Vol. % (1 Porosity ) (22)

CMC − Cu − − Scaffold

For alumina a density of 3.98 g/cm3, for silica a density of 2.196 g/cm3 and for copper a

density of 8.96 g/cm3 were used. The estimation showed that the electrodeposited metal - ceramic

composite had a ceramic phase of 49 vol. %, metal phase of 44 vol. % and a porosity of 7 vol.

%. The lamellar ceramic and polymer≈ -ceramic composite were≈ used as the reference with an

identical ceramic volume ratio (49 vol. %). The polymer-ceramic composite had 19 vol. % of

polymer and 32 vol. % of porosity.

2.3.8 Mechanical characterization

The three-point flexural test and fracture toughness experiments were performed to investigate mechanical properties. The experiments were performed using a MTI/Fullam

SEMTester (MTI Instruments, Inc.). For the fracture toughness experiment, specimen size

followed the ASTM standard [61]. The fracture toughness specimens were ~2.5 mm wide, ~1.25

mm in thickness, and 20 mm long. The test span was 9.8 mm. For the 3-point flexural test, samples

were ~0.4 mm in width, ~0.4 mm in depth, and 20 mm in length. Test span was 9.8 mm. The

displacement speed was 0.1 mm/min in both experiments.

2.4 Experiment Calculation and simulation

2.4.1 Three-point flexural test

Flexure test calculation followed the ASTM standard for flexural test [66]. Flexural stress was calculated via the following eqn. = , where is stress (MPa) in the outer surface at 3𝑃𝑃𝑃𝑃 𝑓𝑓 2 𝑓𝑓 the midpoint, P is the load (N) at a given𝜎𝜎 point2𝑏𝑏𝑑𝑑 on the load𝜎𝜎 -deflection curve, L is the support span

(mm), b and d are the width (mm) and the depth (mm) of beam, respectively. Flexural strain was

calculated via the eqn. = , where is the strain in the outer surface, D is the maximum 6𝐷𝐷𝐷𝐷 2 𝑓𝑓 𝐿𝐿 𝑓𝑓 deflection (mm) of the center𝜖𝜖 of the beam,𝜖𝜖 L is the support span (mm), and d is the depth.

2.4.2 R-curves from SENB experiment

Change in compliance was used to indirectly determine the length of the crack of SENB sample. The change of compliance is often obtained during the cyclic loading. However, cyclic loading was not suitable for the samples tested in this work [31, 33, 47]. The compliance C was

determined from C = u/f, in which u is displacement, and f is the force at each point after crack

initiation. Crack length was calculated from equation:

= + (23) 𝑊𝑊−𝑎𝑎𝑛𝑛−1 𝐶𝐶𝑛𝑛−𝐶𝐶𝑛𝑛−1 𝑎𝑎𝑛𝑛 𝑎𝑎𝑛𝑛−1 2 𝐶𝐶𝑛𝑛 in which W is the thickness of the specimen, a and c represent the crack length and

compliance at n-th and (n-1)-th step. To assess the different mechanisms during the propagation

of stable crack through the specimen, the J-integral was calculated as a function of crack extension,

following the procedure previously reported in the literature [31]. The J-integral comprised of an

elastic contribution Jel and a plastic contribution Jpl.

( ) J = 2 2 (24) el K 1−v E . J = J + pl pl 1 (25) pl pl 1 9 An −An−1 1−an−an−1 n � n−1 �bn−1� � B �� − bn−1 � Here, K is the stress intensity factor, E and are the elastic modulus and Poisson’s ratio of respectively. accounts for toughening mechanisms𝑣𝑣 during fracture propagation and was

𝐾𝐾𝐽𝐽𝐽𝐽 calculated from eqn. = (J + J )E . el pl 𝐾𝐾𝐽𝐽𝐽𝐽 �

2.4.3 Finite element (FE) simulation

εmax = 1%

x axis of symmetry

Several SEM images of the cross-section of the composite were used to construct a representative 2D microstructure with dimensions of 44 μm (width) × 55 μm (length). This geometry was used to set up a 2D FE model. Uniaxial tension scenario was realized by applying two symmetric boundary conditions at the right and bottom edges of the model, and a prescribed displacement boundary at the top edge of the model. The maximum displacement was 0.55 μm, corresponding to a maximum normal engineering strain in the Y-direction of 1%. Material properties for the ceramic and polymer phases were assigned to the segmented areas obtained from the image analysis, as indicated by different colors in Figure 2.4. For the case of the pure ceramic scaffold no polymer phase was included. The ceramic was considered as linear elastic with an

elastic modulus of 350 GPa. For polymers, elastoplastic models were used. The elastic moduli of the as-made and annealed polymers were 445.3 MPa and 1521.4 MPa. The FE model was discretized by 34,496 and 49,863 2D plane strain elements, for the ceramic scaffold and composites respectively, using equal element size. Implicit analysis was conducted using the

ABAQUS Standard solver.

CHAPTER 3

LAMELLAR CERAMIC SEMICRYSTALLINE-POLYMER

COMPOSITE FABRICATED BY FREEZE-CASTING1

3.1 Results and Discussion

Figure 3.1 schematically illustrates the fabrication process of the lamellar composite.

Directional freezing of the ceramic slurry was performed by pouring the slurry into a rectangular acrylic mold placed on a copper cold finger. During the freeze-casting process, ceramic platelets in the slurry accumulated in the space between the growing ice crystals. After freeze-casting process, the ice phase was sublimed by freeze-drying, leaving a negative replica of the ice crystal pattern. Figure 3.1D shows photographs of the freeze-casted scaffold. After this step, cold pressing and sintering processes were applied to the freeze-dried scaffold. Subsequently, several of the scaffolds were surface-functionalized with ATES and infiltrated with the polymer phase.

A thermoplastic, semicrystalline polymer was used as the ductile phase in the composite.

Its mechanical properties can be modified by thermal annealing from very ductile and soft to relatively stiff and brittle (Figure 3.2C). More specifically, theP(VDF-TrFE) polymer was used for infiltration into the lamellar ceramic scaffold. The atomistic model of a unit of this polymer is shown in Figure 3.2A. P(VDF-TrFE) is a piezoelectric polymer with linear chains and its semicrystalline morphology is composed of folded chain crystallite domains immersed in an amorphous phase (Figure 3.2B). Thermal annealing at the paraelectric phase (the temperature

1 From Jiacheng Huang, Zhe Xu, Salvador Moreno, Seyedreza Morsali, Zhong Zhou, Soheil Daryadel, Mahmoud Baniasadi, Dong Qian, Majid Minary-Jolandan, “Lamellar Ceramic Semicrystaline-Polymer Composite Fabricated by Freeze Casting.” Copyright © 2017 by John Wiley Sons, Inc. Reprinted by permission of John Wiley & Sons, Inc.

between the Curie temperature and the melting temperature) results in marked improvement in the

crystallinity of this polymer [59]. The thermal energy during annealing provides sufficient mobility

for the amorphous polymer chains to form folded chain crystallites domains. For example, two

hour annealing at ~ 135 °C results in ~45% enhancement in crystallinity [59].

A C Controller Sample slurry

Mold

Thermocouple

Band heater

Copper finger

Liquid Nitrogen Mixed slurry Freeze casting

Compression Sintering Vacuum infiltration D of polymer

Freeze cast green body

Compression grip

A representative stress-strain response from as-made and annealed polymer films is shown in Figure 3.2C. The as-made polymer shows a very large ductility (~630%), while the annealed polymer fails at a relatively small strain (~11%). The inset shows the magnified view of the low

strain region of the stress-strain response. Detailed mechanical properties of the polymer is shown

in Figure 3.3. The improvement in crystallinity by thermal annealing results in ~3.5 times

improvement in the elastic modulus (~1.5 GPa for annealed polymer, compared to 0.45 GPa for

as-made polymer), at the expense of significant reduction in ductility (~11% for annealed polymer, compared to more than 630% for the as-made polymer). The tensile toughness of the as-made

polymer was ~51 J/g as compared to ~1.2 J/g for the annealed polymer. An interesting observation

is that the as-made polymer exhibited larger strength compared to the annealed one. This could be

attributed to the alignment of polymer chains in the as-made polymer after a large strain, which

resembles a drawing process. The aligned polymer chains resulted in the observed strain hardening

behavior and carried larger stress before failure.

A C H

P(VDF-TrFE) C

B As-made

Annealed

crystallite Amorphous

Mechanical properties of three types of samples, which include the “ceramic scaffold”

without polymer phase, the scaffold infiltrated by polymer phase referred to “as-made composite”, and the scaffold infiltrated by polymer phase and subsequently annealed, referred to as “annealed composite” were characterized. Figure 3.4 shows SEM cross-section images of the three types of composites. In low magnification images in 3.4A, D, and G it can be observed that most platelets are fairly aligned. In higher magnification images, it can be observed that adjacent ceramic platelets are sintered creating lamellar structure with porous regions between the lamellae (Figure

3.4B and C). The fabricated ceramic scaffold has a porosity in the range of 49% - 55%. After infiltration with the polymer phase, the porosity reduced to 33% - 37%.

A B 1521.4 ± 188.1 630.8 ± 20.2

11.3 ± 3

445.3 ± 74.3

C D 51 ± 13 35.1 ± 7.1

24.3 ± 6.3

1.2 ± 0.7

Figure 3.3. Mechanical properties of P(VDF-TrFE) polymer extracted from tensile experiment. (A)-(D) Elastic modulus, ductility (failure strain), tensile toughness, and strength.

Figure 3.5 shows the EDS (energy dispersive X-ray spectroscopy) map of a section of the

fabricated composite. Silicon from silica particles and carbon from polymer phase are observed

between sintered platelets (Alumina), shown by different colors. The polymer phase is pointed by

green arrows in Figure 3.4F and I. For the as-made composite, the polymer phase coated the

ceramic platelets fairly uniformly and formed polymer bridges between platelets. The polymer

phase after annealing showed more of a fibrillar structure wrapping around the platelets.

A B C

D E F D E F

G H I

Figure 3.4. The SEM images show the structure of the composites. (A)-(C) are SEM images of the ceramic scaffold. (D)-(F) are SEM images of the as-made composite. (G)-(I) SEM images of the

annealed composite. Scale bars are (A), (D) and (G) 20 m, (B), (E) and (H) 4 m, (C), (F) and (I) 1 m. µ µ µ A C

Figure 3.6 shows SEM images of as-made and annealed composites with and without surface functionalization of platelets. It is noted that surface functionalization of the ceramic

platelets had a significant effect in adhesion between the platelets and the polymer. This can be observed in cross-section SEM images in Figure 3.6E and F below.

A B

C D

E F

Figure 3.6. SEM images of the as-made and annealed composites with and without surface functionalization of platelets. (A) and (B) are as-made and annealed composites without ATES surface functionalization, respectively. (C) and (D) are as-made and annealed composites with ATES surface functionalization. (E) and (F) show cross-section of an as-made composite without and with surface functionalization, respectively. Clearly, the surface functionalization results in better interface between the platelets and the polymer.

Mechanical properties of the ceramic scaffold and the infiltrated composites were investigated via 3-point flexural test, and fracture toughness test (Figure 3.7 - 3.9, and Table 3.1).

Results showed that both polymer infiltration and thermal annealing had considerable effects on

mechanical properties and behavior of the composites. Flexural strength for ceramic scaffold, as- made composite and annealed composite were 115.6 ± 18.2 MPa, 131.1 ± 7.8 MPa and 155.2 ±

5.4 MPa, respectively (Figure 3.7A). Failure strain under flexure were 0.17 ± 0.05 %, 0.3 ± 0.08

%, and 0.32 ± 0.04 %, for the ceramic scaffold, as-made composite and annealed composites, respectively (Figure 3.7B). The results indicated that the infiltrated polymer did have positive influence on flexural properties of the composite. Statistical analysis (one way ANOVA, p <0.05) followed by a Tukey test further confirmed significant changes in flexural strength and failure strain between the ceramic scaffold and annealed composite. Addition of the annealed polymer phase improved the flexural strength by 34% and failure strain by 96% compared to the ceramic scaffold. Compared to the ceramic scaffold, addition of the mechanically softer as-made polymer phase (without thermal annealing) facilitated larger failure strain, however, increase in strength was not statistically significant. The stiff polymer phase (annealed) likely affected stress distribution to reduce the effects of stress concentration, resulting in higher flexural strength and failure strain, as will be discussed in the FE simulation below.

Table 3.1. Mechanical property of the ceramic, composite and anneal composite with surface functionalization.

Ceramic scaffold As-made composite Annealed composite Flexural strength [MPa] 115.15 ± 24.5 131.14 ± 7.83 155.15 ± 5.42 Strain at failure [%] 0.17 ± 0.05 0.30 ± 0.08 0.32 ± 0.04 Fracture toughness 1.37 ± 0.19 1.89 ± 0.26 1.95 ± 0.04

Representative load-displacement responses for the composite with and without surface functionalization from SENB fracture toughness tests are shown in Figure 3.8A and Figure 3.9A, respectively. Ceramic scaffolds and composites without surface functionalization exhibited

Figure 3.7. (A)-(B) flexure strength and flexure failure strain of the samples. catastrophic failure with a sharp drop in load, while crack propagation in composites with surface functionalization was stable indicated by the gradual decrease in load. Note that the samples had slightly different dimensions. Quantitative analysis on fracture toughness test was performed via

KIC and KJC calculation based on SENB load-displacement data. In terms of crack growth

resistance, functionalized composites exhibited a rising R-curve (crack resistance curve) vs. crack

extension (Figure 3.8B), while ceramic scaffold and composites without surface functionalization

showed limited crack resistance corresponding to their catastrophic failure (Figure 3.9B). All data

points shown in the crack growth resistance plots are within the ASTM valid range of crack

extension ( = 0.25 , in which b is the initial material ligament) [40, 61]. Data points within

𝑚𝑚𝑚𝑚𝑚𝑚 the ASTM ∆standard𝑎𝑎 represent𝑏𝑏 material properties and rule out dimensional effects.

SEM images (Figure 3.6) show that polymer forms better adhesion with functionalized platelets. For these samples, because of the stronger bond between the polymer and ceramic scaffold, polymer experienced large deformation before rupture or delamination, which resulted in a rising R-curve and larger fracture toughness. On the other hand, in composites with a weaker bond between the polymer and ceramic scaffold, polymer tends to delaminate quickly without experiencing significant deformation, hence dissipates less energy during the crack propagation.

This resulted in limited R-curve and smaller fracture toughness.

The relation between the fracture toughness at crack initiation KIC and flexural strength is

1/2 shown in Figure 3.8C. KIC for ceramic scaffold was 1.37 ± 0.19 MPa.m ; it increased to 1.89 ±

0.26 MPa.m1/2 and 1.95 ± 0.04 MPa.m1/2 for functionalized as-made composite and annealed composite, respectively. Despite the non-significant performance difference between as-made

composite and annealed composite, the fracture toughness of the infiltrated composites were

considerably larger than the ceramic scaffold. Similarly, there was no significant difference

between the R-curve of the annealed vs. as-made composite, in Figure 3.8B. Interestingly, both

the fracture toughness and the flexural strength showed smallest range of variation (standard

deviation) for the annealed composite (Figure 3.8C).

Figure 3.8. Characterization of the fracture toughness of the fabricated composites. (A) Representative load-displacement responses from the single edge notched beam (SENB) fracture

tests. Note that the measured specimens differ slightly in dimensions. (B) The crack growth resistance curves (R-curves) obtained from the SENB responses. (C) Fracture toughness KIC vs flexural strength (n = 3). The vertical and horizontal error bars are the standard deviation.

To observe the deformation mechanism of the composites, in-situ SEM SENB experiments were conducted (Figure 3.10). For the ceramic scaffold, the crack extended in a nearly straight line from the crack tip. For the composites, the crack took several short deviations from the straight

A B C D

E F G

H I J

scaffold. Scale bars are (A, E, H) 400 m, (B) 50 m, (C) 3 m, (D) 1 m and (F, G, I, J) 2 m. Green arrows indicate the stretched polymer, red arrows indicate the fractured platelets. µ µ µ µ µ line. Detailed SEM images of each sample were acquired after the SENB test. Stretched polymer

was observed on the fracture surface of the composites, which shows that the polymer was

stretched during the crack propagation process, forming fiber-like shape (Figure 3.10B, C, D, F,

G). This abundance of polymer fibers on the fracture surface of the composite specimens shows

that the rising R-curve behavior of the composite can be attributed to the energy dissipation by

polymer deformation during crack growth, which can reduce the stress concentration. On the other

hand, for the ceramic scaffold, platelet fracture was observed; flat cross-section of sintered

platelets are pointed by red arrows in Figure 3.10I, J. Although the layered structure of the ceramic

scaffold mediated slight crack deflections, however, as was observed from its R-curve behavior,

the ceramic scaffold has limited crack growth resistance. The polymer phase, however, bridged

the crack planes and was stretched (Figure 3.10B), which resulted in rising R-curve behavior through the crack bridging mechanism.

To investigate distribution of stress in the specimens, a 2D FE model was constructed from the representative SEM images (Figure 3.11). Material properties for the ceramic and polymer

phases were assigned to the segmented areas obtained from the image analysis, as indicated by

different colors in Figure 3.11. For the case of the pure ceramic scaffold no polymer phase was

included. Figure 3.11 shows the representative 2D FE model with dimensions of 44 μm (width) ×

55 μm (length). Uniaxial tension loading was realized by applying two symmetric boundary

conditions at the right and bottom edges of the model, and a prescribed displacement boundary at

the top edge of the model. The maximum displacement was 0.55 μm, corresponding to a maximum

normal engineering strain in the Y-direction of 1%. The ceramic was considered as linear elastic

with an elastic modulus of 350 GPa. For polymers, elasto-plastic models were used. The elastic

moduli of the as-made and annealed polymers were 445.3 MPa and 1521.4 MPa, respectively. The

stress-strain responses up to 10% strain of the polymer phase were converted to the yield stress

versus plastic strain. These yield stress curves were used to describe the isotropic plastic hardening

behavior of the polymer phase.

εmax = 1%

x axis of symmetry

Figure 3.11. 2D plane strain FE model of ceramic platelet-P(VDF-TrFE) composite. Ceramic platelets are shown in yellow, and polymer is shown in navy. This structure corresponds to a ceramic volume fraction of 49%, a polymer volume fraction of 18% and void fractions of 33%, similar to the experimental values.

The von Mises stress distribution under tensile deformation from the FE simulation was evaluated. von Mises stress was chosen because the plastic behavior of the polymer phase was described by a von Mises stress-based yield function. The stress partition to different parts of the model is visualized in Figure 3.12, at the nominal global strain of 0.2%. The 0.2% strain was chosen in reference to the flexure test failure strain of 0.17% for the ceramic scaffold (Figure 3.7B).

From Figure 3.12, we can see that in all the three cases, stress contour maps exhibited similar

pattern: stresses below 100 MPa were found in the polymer phase (for the composite) and the

branches of the ceramic platelets; stresses between 100 MPa and 1000 MPa were mainly carried

by the ceramic scaffold, in particular in the branches that link all the way through the model as

illustrated in Figure 3.12B, E, and H; and a small level of stress concentration larger than 1000

MPa was observed in the ceramic ligaments connecting the platelets. Comparing the two

composites, the stress maps show marginal differences despite the three times difference in elastic

moduli between the as-made and annealed polymers.

In order to further quantify the von Mises stress distribution within the polymer and

ceramic phases during mechanical loading, stress distribution histograms are plotted in Figure

3.13. The abscissa is stress in logarithmic scale, and the ordinate is frequency, which is the number

of elements presenting a stress in the respective range, normalized by the total number of elements

in each model. In the histogram of composites at 0.2% global strain shown in Figure 3.13A, the

first peak appeared at 10 – 20 MPa, which is the stress mainly carried by the polymer phase up to

the respective yield stress of the polymer. The second peak was observed at stress in the range of

100 – 200 MPa, which is exclusively carried by the ceramic phase. In contrast, there was only one

peak in the stress range of 100 – 200 MPa for the ceramic scaffold. Furthermore, the addition of

polymer phase enabled a more uniform stress distribution in the composite than in the scaffold.

This consequently would reduce stress concentrations, which may result in enhanced fracture

toughness. At 0.2% strain, the difference between the stress distribution in the as-made polymer and annealed polymer was negligible. At an increased global nominal strain of 1% plotted in Figure

3.13B, the composite containing annealed polymer showed distinct first peak with higher stress than the composite containing as-made polymer, and the corresponding stress values were

correlated to the maximum yield stress of these two polymers in the studied strain range, i.e., 22

MPa and 14 MPa.

0-100MPa 100-1000MPa >1000MPa Ceramic scaffold scaffoldA B C

60% 39.4% 0.6%

As-made composite D E F

70.3% 29.3% 0.4%

Annealed composite G H I

69.9% 29.7% 0.4%

Figure 3.12. The von-Mises stress partition at 0.2% applied nominal strain. (A-C) Ceramic scaffold. (D-F) As-made composite. (G-I) Annealed composite.

3.2 Conclusions

In this chapter, fabrication and characterization of a lamellar ceramic-polymer composite

was reported. The composite was fabricated using the freeze-casting process infiltrated by a semicrystalline polymer. Annealing of the polymer phase allows for changing its mechanical properties from soft (E ~ 445 MPa) and ductile (εf ~ 630%) to relatively stiff (E ~ 1.5 GPa) and

brittle (εf ~ 11%). The flexure test, and fracture toughness measurement, along with in situ SEM

observation of the mechanical behavior were used to investigate the change in mechanical

properties of the composite as the result of change in the mechanical properties of the polymer.

The results show that addition of the polymer phase enhances flexural strength, failure strain under

flexure, and fracture toughness and crack extension resistance behavior of the composite. In

addition, the surface functionalization of the ceramic scaffold had a significant effect in the

effectiveness of the polymer phase in the composite by enhancing the interfacial adhesion between

the polymer and the ceramic. More importantly, however, the results showed that more than ~three

times change in the elastic modulus, and more than fifty times change in ductility of the polymer

phase had only marginal effects on the behavior of the composite.

A Ceramic scaffold B Ceramic scaffold Composite-as made Composite-as made Composite-annealed Composite-annealed carried by ceramic

carried by Frequency Frequency polymer

0.01 0.1 1 10 100 1000 10000 1 10 100 1000 10000 von-Mises Stress (MPa) von-Mises Stress (MPa)

Figure 3.13. Simulated von Mises stress distribution in the scaffold and composites at (A) 0.2% strain, and (B) 1.0% strain.

CHAPTER 4

ALUMINA-NICKEL COMPOSITE PROCESSED VIA CO-ASSEMBLY USING

FREEZE-CASTING AND SPARK PLASMA SINTERING

4.1 Results and Discussion

Figure 4.1 - 4.2 illustrate the fabrication process of the metal - ceramic composite. Details of the process are presented in the materials and methods chapter. Briefly, alumina micro-platelets

(with an average in-plane size of ~ 6 μm) and nickel powder (particles with size of ~2-3 μm) were thoroughly mixed in a container filled with Argon gas. A roller mixer with tungsten carbide milling balls was used for mixing. The mixed powder was then made into a slurry by adding DI water.

Alginate was added as a rheology modifier to prevent the suspension from sedimentation during the freeze-casting process. Directional-freezing of the slurry was performed by pouring the slurry into a mold placed on a copper cold finger. During the freeze-casting process, ice crystals start growing from the copper plate, forming vertical laminated ice. The laminated ice crystals repel the particles in the mixture into the gaps between ice lamella. After freeze-casting process, the ice was sublimed by freeze-drying, leaving a negative replica of the ice lamellar pattern of the powder mixture.

The freeze-dried scaffold was then compressed, cut and stacked, followed by the spark plasma sintering process. SPS is a low-voltage, direct current (DC) activated, pressure-assisted sintering process, and has been widely applied for materials processing in recent years [33, 35].

Unlike the conventional hot-press where the heat is produced using the heating element outside of the sample, SPS produces heat by joule heating within the sample by passing a large current through a conductive sample. SPS can also produce heat through the conductive graphite die and

the punch around a non-conductive sample. During the SPS process under high pressure and high temperature, Ni particles melt and flow between alumina platelets, and form the ductile phase

(Figure 4.2).

During the SPS process, the pressure was initially set to 20 MPa, and the temperature was held at 600 for 3 hours to burn-off the Alginate binder. Then, the temperature was raised to

1250 with℃ a ramping rate of 100 /min and held at that temperature for 10 minutes. (Note that the melting℃ point of Ni is 1453 °C).℃ Subsequently, the temperature was slowly decreased with a cooling rate of ~3 /min while the pressure was increased with a ramping rate of 1 MPa/min. The reason for this slow℃ cooling and pressing process was to gradually squeeze the softened nickel and

form a layer of the metal phase between the alumina platelets. This temperature-pressure history during the SPS process is shown in Figure 4.2.

Figure 4.2. (A) Temperature- pressure profile during the SPS process. (B) The schematic of the internal structure of the composite during the SPS process.

To evaluate the effect of the Ni phase in the composite, a lamellar ceramic was also

fabricated using the identical process, without the Ni powder. SEM images of the lamellar ceramic

(Figure 4.3A-C) and metal-ceramic composite (Figure 4.3D-F) are shown in Figure 4.3. Nickel

particles can be observed in the metal-ceramic composite before the sintering process (Figure

4.3D). The initial distribution of Ni particles is crucial to the properties of the composite. An even

distribution before sintering process will facilitate more even distribution of Ni phase in the

sintered composite. Alumina platelets and nickel particles were assembled into a lamellar structure

during the freeze-casting process. The pressure applied during the SPS process further reduces the

gap between platelets and improves their alignment. In addition, Ni melts and gets squeezed in

between the laminated alumina. In SEM images in Figures 4.3E and 4.3F, the brighter regions are

Ni phase and the darker areas are alumina platelets. Overall, it can be observed that the Ni phase

is distributed rather uniformly between the alumina platelets. Additionally, Ni regions are

elongated in the alignment direction of the platelets, perpendicular to the loading direction in the

SPS process.

The 3-point flexural test was performed to quantify and compare the strength of the

lamellar ceramic and metal - ceramic composite. The flexure test was performed under displacement control mode with a displacement rate of 0.1 mm/min. Calculations for the flexural test followed the ASTM standard D790. The flexural stress was calculated using = , 3𝑃𝑃𝑃𝑃 2 𝑓𝑓 2𝑏𝑏𝑑𝑑 where is stress (MPa) in the outer surface of the beam at the mid-span of the beam, P𝜎𝜎 is the load

𝜎𝜎𝑓𝑓 50

(N) at a given point on the load-deflection curve, L is the support span, b and d are the width and

the depth of the beam, respectively. Flexural strain was calculated using = , where is the 6𝐷𝐷𝐷𝐷 2 𝑓𝑓 𝐿𝐿 𝑓𝑓 strain in the outer surface, D is the maximum deflection of the beam’s mid𝜖𝜖 -span, L is the 𝜖𝜖support span, and d is the depth.

The representative flexural stress-strain responses for the lamellar ceramic and metal - ceramic composite are shown in Figure 4.4A. At least four samples were tested for each type. Both materials showed a linear behavior up to the failure point. The maximum force was used for the calculation of flexural strength. Flexural strengths of the lamellar ceramic and metal - ceramic composite were 159.7 ± 57.5 MPa and 235.5 ± 38.2 MPa, respectively (Figure 4.4B). Strains at maximum stress for the lamellar ceramic and metal - ceramic composite were 0.159 ± 3.8 ×10-4

% and 0.165 ± 2.8 ×10-4 %, respectively (Figure 4.4C). The results indicate that inclusion of the

Ni phase did have a positive influence on the composite’s flexural strength. Composite’s strength

improved by ~47% compared to the lamellar ceramic. The Ni phase affects the stress distribution

by bearing fractions of the load and reducing the stress concentration, and hence, it increases the

strength of the composite [40, 64].

The crack growth resistance curve (or the R-curve) was extracted from the fracture force-

displacement responses. The R-curve represents fracture toughness vs. crack extension.

Interestingly, both the lamellar ceramic and metal-ceramic composite exhibited a rising R-curve

behavior (Figure 4.5B). A rising R-curve behavior indicates presence of intrinsic toughness

mechanisms that increase the fracture toughness of the material as the crack length extends. The

rising R-curve in the ceramic originates from its lamellar structure, which enables limited sliding

between platelets and platelet pull-out as the crack extends. These mechanisms may also result in the deflection of the crack path. The metal-ceramic composite showed a larger fracture toughness with more stable R-curve behavior, an indication that Ni inclusion in the composite provides an additional toughening mechanism for fracture.

The fracture toughness (KIC) or the crack initiation fracture toughness is obtained from the

R-curve at the initial point when the crack extension length is zero. Figure 4.5C shows the relation

between the KIC and the flexural strength for the lamellar ceramic and the metal-ceramic

/ composite. KIC for the lamellar ceramic was 2.54 ± 0.66 MPa. m , and it increased to 3.05 ± 1 2 0.19 MPa. m / for the metal-ceramic composite. Inclusion of Ni enhanced both the fracture 1 2 toughness and the strength. Additionally, the smaller standard deviation of the composite indicates

a more stable and predictable behavior compared to the lamellar ceramic.

After the fracture toughness experiment, SEM images were acquired from the fractured area of the specimens to identify fracture mechanisms. Interestingly, both the lamellar ceramic and the metal-ceramic composite maintained their integrity after the fracture, which enabled observation of the crack path in the post-failure specimens. Both materials showed a zigzag crack path, which is remarkable considering that they have more than 80 vol. % of brittle constituents.

In particular for the lamellar ceramic, the structured architecture of the aligned platelets effectively deflected the crack multiple times in the crack tip (Figure 4.6A-B), which may be the mechanism

for the rising R-curve. The majority of the crack path passes among the platelets, although some

level of platelet failure was also observed (Figure 4.6C-F).

A similar crack deflection was observed ahead of the crack tip in the metal-ceramic

composite. In the composite, there were additional toughening mechanisms compared to the

lamellar ceramic. Multiple micro-cracks were formed along the main crack, which contributed to

more energy dissipation during the crack growth (Figure 4.6G-H). Platelet pull-out was observed

in higher magnification SEM images (Figure 4.6I-J). Additionally, plastic deformation and failure of the Ni phase were observed along the crack path (Figure 4.6K-L). Generally, an intermediate bonding (interfacial strength) between the metal phase and the ceramic phase is desired [43]. If the

bonding is too weak, the crack will propagate along the interface, and contribution from the

ductility of the metal phase to composite toughness will be limited. On the other hand, if the

interface is very strong, only limited fraction of the metal phase will contribute to the toughness.

For an intermediate bonding strength, both interface debonding and deformation of the metal phase

will contribute to the toughness. The observed platelet pull-out and deformation of the metal phase

may indicate an intermediate interfacial strength in the fabricated alumina-Ni composite.

Nanoindentation experiments were performed to obtain and compare the indentation

modulus and indentation hardness. The indentation test was run in the load-control mode. The

maximum force was set to 200 mN, and the loading rate was 1 mN/s. The peak hold time was

120s. Figure 4.7A and B show the indentation force vs. indentation depth responses for the cases

that the loading direction was parallel and perpendicular to the planes of the alumina platelets,

respectively.

Figure 4.6. SEM micrographs from the fractured area of the SENB samples. (A)-(F) The lamellar ceramic and (G)-(L) metal-ceramic composite. Red arrows in (H) indicate the crack deflection.

The effect of Ni inclusion is more pronounced for the case of indentation perpendicular to the plane of the platelets (Figure 4.7B). For the lamellar ceramic multiple deflections in force is observed, while the metal-ceramic composite shows a smooth indentation response. The results for the nanoindentation modulus and hardness are shown in Figures 4.7C and 4.7D. The elastic modulus and hardness of both materials are statistically the same. Only when the loading was perpendicular to the plane of the platelets, the hardness of the metal-ceramic composite was statistically smaller than the lamellar ceramic. Also, both materials showed larger modulus and hardness when the indentation load was in the plane of the platelets. Overall, indentation results show that inclusion of Ni does not result in significant decreases in the modulus and hardness of the composite compared to the lamellar ceramic, although Ni has smaller hardness and modulus compared to alumina. This is desirable since the main purpose of inclusion of a small volume

percentage of ductile phase (Ni) was to increase the fracture toughness of the composite without a

significant compromise in other properties such as the elastic modulus, strength and hardness.

Figure 4.8A and B show Ashby plots for fracture toughness vs. density and flexural modulus vs. flexural strength. Obtained mechanical properties of the fabricated Ni-Alumina composite is superimposed on the plots. For more improved properties several points can be

considered for future investigations. Further decrease in the volume fraction of the metal phase

and its more uniform distribution may contribute to improved fracture toughness, elastic modulus

and strength of the composite. Improved alignment of the ceramic platelets in the freeze-casting

process may also enhance mechanical properties of the composite.

Figure 4.8. Ashby plots for (A) fracture toughness vs. density and (B) flexural modulus vs. flexural strength.

4.2 Conclusions

Co-assembly of metal micro-particles and ceramic micro-platelets using the freeze-casting process is a promising route for inclusion of metals into ceramics. The designed temperature- pressure profile during the SPS process results in formation of a ductile metal phase between ceramic micro-platelets. The results show that inclusion of Ni into alumina enhances fracture toughness and strength of the composite without significant compromise of the indentation modulus and hardness, as compared to the pure lamellar ceramic. Improved properties can be observed in the R-curve behavior of the composite. Post-failure analysis of the fractured specimens

revealed that combinations of crack deflection, formation of micro-cracks, platelet pull-out, and plastic deformation of the metal phase contributed to the enhanced fracture toughness of the composite.

CHAPTER 5

MANUFACTURING OF METAL-CERAMIC COMPOSITE

THROUGH INFILTRATION OF COPPER INTO FREEZE-CAST

ALUMINA SCAFFOLD BY ELECTROPLATING

5.1 Results and Discussion

Figure 5.1 and 5.2A illustrates the fabrication process of the metal-ceramic composite.

Details of the process are provided in the materials and method section. The mixture of alumina micro-platelets (with an average in-plane size of ~4-8 μm, and thickness in the range of ~450 ±

100 nm) and silica nanoparticles were made into a slurry by adding DI (deionized) water. Alginate was added as a rheology modifier to prevent the suspension from sedimentation during the freeze- casting process. Directional freezing of the slurry was performed by pouring the slurry into a mold placed at top of a copper cold finger. During the freeze-casting process, ice crystals start growing from the copper plate, forming vertical laminated ice and repel the suspended powder mixture into the gaps between the ice crystals. After the freeze-casting process, the ice phase was sublimed by freeze-drying, leaving a negative replica of the ice crystal pattern of the powder mixture. The alginate also functions as a soft binder between the alumina platelets and silica nanoparticles in the freeze-casted green body. After the compression step to reduce the gaps between the ceramic lamella, the freeze-casted scaffold was heated in two steps: the alginate was removed in the first lower temperature step, and then the ceramic scaffold was thoroughly sintered in the second high temperature step.

The sintered ceramic scaffold was cut into ~1.3 mm thin sections by a diamond saw

followed by ultrasonic cleaning in DI water. Then, it was dried in a vacuum oven at 40

overnight. Subsequently, one side of the scaffold and a glass slide were coated with a thin layer ℃of

Figure 5.1. The illustration of the electrodeposition setup and photograph of the setup.

Figure 5.2. The fabrication process of the electrodeposited metal-ceramic composite. (A) The process steps include directional freeze-casting of alumina micro-particles and freeze-drying,

compression, sintering, and metal infiltration by electroplating. The photographs in the bottom panel of A show actual specimens. The original alumina ceramic scaffold has a white-grey color, and once it was infiltrated with Cu, it appears as orange color. (B) The schematic illustrates electroplating of Cu inside the pores of the ceramic scaffold. The bottom side of the ceramic scaffold is placed on the cathode surface. Blue, grey, and orange colors indicate alumina, silica, and Cu, respectively.

gold (Au) by sputter deposition. The scaffold was attached to the glass slide on their conductive

surfaces. The side of the scaffold on the glass slide functions as the cathode during the

electroplating process. For electroplating, a reference electrode was used to form a three-electrode

electrochemical cell, in which a Cu mesh was connected to the anode electrode, and the ceramic

scaffold was connected to the cathode electrode of the potentiostat. Vacuum infiltration was

applied to assist full infiltration of the electrolyte into the ceramic scaffold. For electroplating of

Cu, a constant DC voltage of -0.2 V was applied. During the electrodeposition process, copper

starts to grow from the gold-coated side of the scaffold (the cathode), and continues to grow and

fill up the gaps inside the lamellar ceramic scaffold (Figure 5.2B) to form the metal-ceramic

composite with fine laminated copper-alumina microstructure.

As shown in Figure 5.3, copper filled-up the ceramic scaffold and formed a copper cap at

the top of the scaffold. The cross-section scanning electron microscope (SEM) images of the

ceramic scaffold before sintering and after sintering, as well as the cross-section SEM image of the composite are shown in Figure 5.4. The freeze-casting and cold-pressing steps form a compact laminated structure of alumina micro-platelets and silica nanoparticles, and the alginate functions as the binder (Figure 5.4A). Figure 5.4B and C show the ceramic scaffold before and after the electrodeposition process. Note that the SEM image in Figure 5.4B was obtained from a fractured specimen, while the SEM image of the composite was obtained from a polished surface. In the

composite, the lighter regions are metal (Cu) and the darker regions are alumina micro-platelets.

As can be observed, the Cu phase fully filled up the gaps between ceramic phase, and no porosity was observed in the SEM images (Figure 5.4C).

Figure 5.3. False-colored SEM images of the electrodepositioned sample, cross-section area. Orange area indicates the copper while the gray and blue areas show the ceramic scaffold.

Figure 5.4. (A)-(B) Cross-section scanning electron microscope (SEM) images of the ceramic scaffold before sintering and after sintering, respectively. (C) Cross-section SEM image of the

metal-ceramic composite. Lighter regions are metal (Cu) and the darker regions are ceramic micro- platelets. The inset in C demonstrates the electrical conductivity of the composite.

A pure lamellar ceramic with the identical process to the composite without the metal infiltration step was also prepared, to compare properties of the metal-ceramic composite with the pure lamellar ceramic. Additionally, in mechanical properties in Figures 5.5 and 5.6, properties of

a polymer-ceramic composite with the identical ceramic scaffold are included. In this case, the

ceramic scaffold was infiltrated with a polymer, PVDF-TrFE (Polyvinylidene fluoride-

trifluoroethylene), instead of metal. This allowed to directly compare the behavior of the polymer vs. metal as the ductile phase in hybrid composites.

The 3-point flexural test was performed to quantify and compare the flexural properties of the electrodeposited metal-ceramic composite, and the lamellar ceramic, as shown in Figure 5.5.

The flexure test was performed in displacement control mode with a displacement rate of 0.1 mm/min. Calculations for the flexural test followed the ASTM standard D790 [66]. The flexural

stress was calculated using = , where is the stress in the outer surface of the beam at 3𝑃𝑃𝑃𝑃 𝑓𝑓 2 𝑓𝑓 mid-span, P is the load at a 𝜎𝜎given2 point𝑏𝑏𝑑𝑑 on the 𝜎𝜎load-deflection curve, L is the support span, b and

d are the width and the depth of the beam, respectively. Flexural strain was calculated using =

𝜖𝜖𝑓𝑓 , in which is the strain in the outer surface, D is the maximum deflection of the center of the 6𝐷𝐷𝐷𝐷 2 𝐿𝐿 𝑓𝑓 beam, L is the𝜖𝜖 support span, and d is the thickness of the beam.

The representative flexural stress-strain responses for metal-ceramic composite, polymer-

ceramic composite, and lamellar ceramic are shown in Figure 5.5A. At least four samples were

tested for each type. The lamellar ceramic and polymer-ceramic composite showed a linear

behavior up to the failure point, followed by an abrupt fracture. The metal-ceramic composite showed a linear behavior, followed by a small region of nonlinear behavior right before the fracture. While there was a large drop in force after the fracture point, overall the metal-ceramic

composite showed a more gradual failure compared to the other two materials.

showed a linear behavior, followed by a small region of nonlinear behavior right before the

fracture. While there was a large drop in force after the fracture point, overall the metal-ceramic

composite showed a more gradual failure compared to the other two materials.

The maximum force was used for the calculation of flexural strength. The flexural strength

for the metal-ceramic composite, polymer-ceramic composite, and lamellar ceramic were 278.7 ±

21.8 MPa, 143.14 ± 14.46 MPa and 115.6 ± 24.5 MPa, respectively. Flexural modulus for the

metal-ceramic composite, polymer-ceramic composite, and lamellar ceramic were 116.4 ± 15.59

GPa, 53.5 ± 9.94 GPa and 76.5 ± 18.21 GPa, respectively (Figure 5.5B). Strain at the maximum stress for the metal-ceramic composite, polymer-ceramic composite, and lamellar ceramic were

0.27 ± 0.044 %, 0.32 ± 0.044 % and 0.163 ± 0.034 %, respectively (Figure 5.5C). The results indicate that inclusion of the second phase did have a positive influence on the flexural strength of the composite, while the strong metallic phase (Cu) exhibited more pronounced improvement. The strength of the electrodeposited metal-ceramic composite improved by 141% compared to the lamellar ceramic, in comparison to the polymer-ceramic composite’s ~24% improvement. Copper is capable of bearing more load compared to the polymer, which reduces the stress concentration in the ceramic phase, and therefore increases the strength of the composite by delaying the failure

[49, 64].

To quantify and compare the fracture resistance properties of the metal-ceramic composite, polymer-ceramic composite, and the lamellar ceramic, the fracture toughness experiment was carried out. At least three samples were tested for each condition. For this purpose, the single-edge

notched beam (SENB) was used. The dimensions of the specimens (width, length, thickness, and

pre-notch length) followed the ASTM standard E1820. Figure 5.6A shows representative load-

displacement responses from the SENB fracture toughness experiment. The pure lamellar alumina

exhibited a sharp drop in force after the maximum load, while the electrodeposited metal-ceramic

composite and the polymer-ceramic composite showed a gradual failure after the maximum load.

The gradual failure in fracture experiment shows the direct role of the ductile phase in enhancing

the fracture toughness of the material.

The crack growth resistance curves (or the R-curve) were extracted from the load- displacement responses. The R-curve represents fracture toughness vs. crack extension (Figure

5.6B). A rising R-curve behavior indicates the presence of intrinsic fracture toughness mechanisms that increase the fracture toughness of the material as the crack length extends. Interestingly, all three materials exhibited a rising R-curve behavior, however, the metal-ceramic composite showed a more pronounced rising curve compared to the other two materials. The rising R-curve in the ceramic originates from its lamellar structure, which enables limited crack deflection and micro cracking as the crack extends. The metal-ceramic composite showed a significantly larger fracture toughness with a more stable R-curve behavior, indications that copper inclusion in the composite provides better toughening effect and introduces additional toughening mechanisms, as discussed in the following.

The fracture toughness (KIC) or the crack initiation fracture toughness is obtained from the

R-curve at the initial point when the crack extension length is zero. Figure 5.6C shows the relation

between KIC and the flexural strength for the three materials. KIC for the lamellar ceramic was 1.36

± 0.18 MPa. m / , and it increased to 1.77 ± 0.18 MPa. m / , and 4.58 ± 0.66 MPa. m / , for the 1 2 1 2 1 2 polymer-ceramic composite, and the electrodeposited metal-ceramic composite, respectively.

Figure 5.6D shows the comparison between the crack initiation toughness (KIC) and the crack growth toughness. The crack growth toughness was defined based on the plateau region of the R-curve. Clearly, the metal-ceramic composite has a much larger crack growth toughness compared to the other two materials. The crack growth toughness was 1.72 ± 0.25 MPa. m / , 1 2 2.36 ± 0.15 MPa. m / , and 10.52 ± 0.86 MPa. m / for the lamellar ceramic, polymer-ceramic 1 2 1 2 composite, and metal-ceramic composite, respectively.

µ 69

The percentage increase between crack initiation toughness and crack growth toughness is shown by numbers in Figure 5.6D. The much pronounced R-curve behavior of the metal-ceramic composite results in ~130% increase from crack initiation toughness to crack growth toughness, compared to only ~33% and 27% for the polymer-ceramic composite and the lamellar ceramic.

After the fracture toughness experiment, SEM images were acquired from the fractured area of the specimens to identify the underlying toughening mechanisms. Both electrodeposited metal-ceramic composite and polymer-ceramic composite maintained their integrity after the fracture, while the lamellar ceramic snapped into two pieces. The electrodeposited metal-ceramic composite showed a zigzag crack path, in which the structured architecture of the aligned platelets as well as the copper phase effectively deflected the crack multiple times ahead of the crack tip

(Figure 5.7A). In addition to crack deflection, other toughening mechanisms were observed.

Figure 5.8. Ashby plots for (A) the fracture toughness KIC verses density, and (B) for modulus vs. strength [49]. Properties of the composite in this work is overlaid on the plots using a red solid circle.

These mechanisms shown in Figure 5.7(B)-(G) included crack branching, friction, platelet pullout, and deformation and plasticity of the metal phase. The majority of the crack path passes among the platelets, although some level of platelet failure was also observed.

Figure 5.8A and B show Ashby plots for the fracture toughness vs. density and elastic modulus vs. strength. The obtained mechanical properties of the fabricated Cu-Alumina composite are superimposed on the plots. For further improvements for the properties of the composite several points can be considered. Further decrease in the volume fraction of the metal phase and better constituent distribution may contribute to improved fracture toughness, and strength of the composite. Improved alignment of the ceramic platelets in the freeze-casting process is also expected to enhance composite’s mechanical properties. Other metals that can be electrodeposited such as nickel and even refractory metals can be examined as the ductile phase. Alloys with higher strength may also contribute to enhanced mechanical properties.

5.2 Conclusions

A low-cost, room-temperature process for infiltration of metals into pre-formed ceramic scaffold for manufacturing of metal-ceramic composite was demonstrated. This process is based on reduction of metal ions to metal inside the gaps of the ceramic scaffold by electrodeposition from an aqueous bath. Compared to the conventional molten metal infiltration process for preparation of metal-ceramic composites, this process consumes ~2 orders of magnitude less energy. Electroplating is a scalable and low-cost process, and has been used in various industries for centuries. There are many metals and even alloys that can be electroplated. Hence, the

presented process in this disseration can be readily expanded to various metals and alloys for different applications. This low-cost process may expand applications of metal-ceramic composites in various industries including automotive, aerospace, oil, and defense.

CHAPTER 6

OUTLOOK

The fracture toughness of the fabricated compites in this work were improved when metal was used as the ductile phase compared to the polymer. Additionally, the electrodeposited metal phase composite had better fracture toughness compared to the nickel particle metal phase composite (Figure 6.1.). However, the normalized strength and modulus showed certain level of decrease (Figure 6.2.). The large volume ratio of the metal phase drastically increased the overall density of the composite, causing the reduced specific strength and specific modulus.

Figure 6.1. Ashby plot for fracture toughness vs. modulus. Modulus vs fracture toughness.

Figure 6.2. Ashby plot for specific modulus vs. specifc strength.

One way to mitigate such effect is to reduce the volume ratio of the metal phase as well as sample’s porosity. One of the key attributes of the nacre’s microstructure is the extremely thin layer of the ductile phase between the well-aligned platelets. Such microstructure ensures both outstanding mechanical properties and light-weight simultanously. Creating such thin layer of metal phase between ceramic platelets was proven to be extremely challenging. However, the challenge can be overcome through the electrodeposition process, since the metal ions can flow readily though any gaps and slits inside the ceramic scaffold. Preliminary work was conducted in this dissertation that shows possibility of creating such thin metal mortar layer using the electrodeposition process (Figure 6.6). This process can be further expanded toward development of tough and ligh-weight composites.

Figure 6.3. SEM images of the sample before the electrodeposition and after the electrodeposition of metal ductile layer.

REFERENCES

1. Ritchie, R.O., The conflicts between strength and toughness. Nature materials, 2011. 10(11): p. 817.

2. Gao, H., et al., Materials become insensitive to flaws at nanoscale: lessons from nature. Proceedings of the national Academy of Sciences, 2003. 100(10): p. 5597-5600.

3. Fratzl, P. and R. Weinkamer, Nature’s hierarchical materials. Progress in materials Science, 2007. 52(8): p. 1263-1334.

4. Ritchie, R.O., In pursuit of damage tolerance in engineering and biological materials. MRS Bulletin, 2014. 39(10): p. 880-890.

5. Yaraghi, N.A., et al., A sinusoidally architected helicoidal biocomposite. Advanced Materials, 2016. 28(32): p. 6835-6844.

6. Kamat, S., et al., Structural basis for the fracture toughness of the shell of the conch Strombus gigas. Nature, 2000. 405(6790): p. 1036.

7. Koester, K.J., J. Ager Iii, and R. Ritchie, The true toughness of human cortical bone measured with realistically short cracks. Nature materials, 2008. 7(8): p. 672.

8. Naleway, S.E., et al., Structural design elements in biological materials: application to bioinspiration. Advanced Materials, 2015. 27(37): p. 5455-5476.

9. Li, L. and C. Ortiz, A natural 3D interconnected laminated composite with enhanced damage resistance. Advanced Functional Materials, 2015. 25(23): p. 3463-3471.

10. Amini, S., et al., The role of quasi-plasticity in the extreme contact damage tolerance of the stomatopod dactyl club. Nature materials, 2015. 14(9): p. 943.

11. Schwiedrzik, J., et al., In situ micropillar compression reveals superior strength and ductility but an absence of damage in lamellar bone. Nature materials, 2014. 13(7): p. 740.

12. Peterlik, H., et al., From brittle to ductile fracture of bone. Nature materials, 2006. 5(1): p. 52.

13. Weaver, J.C., et al., The stomatopod dactyl club: a formidable damage-tolerant biological hammer. Science, 2012. 336(6086): p. 1275-1280.

14. Chen, P.-Y., et al., Structure and mechanical properties of selected biological materials. Journal of the mechanical behavior of biomedical materials, 2008. 1(3): p. 208-226.

15. Meyers, M.A., J. McKittrick, and P.-Y. Chen, Structural biological materials: critical mechanics-materials connections. science, 2013. 339(6121): p. 773-779.

16. Wegst, U.G., et al., Bioinspired structural materials. Nature materials, 2015. 14(1): p. 23.

17. Barthelat, F., Z. Yin, and M.J. Buehler, Structure and mechanics of interfaces in biological materials. Nature Reviews Materials, 2016. 1(4): p. 16007.

18. Corni, I., et al., A review of experimental techniques to produce a nacre-like structure. Bioinspiration & biomimetics, 2012. 7(3): p. 031001.

19. Cheng, Q., et al., Learning from nature: constructing integrated graphene-based artificial nacre. ACS nano, 2015. 9(3): p. 2231-2234.

20. Bonderer, L.J., A.R. Studart, and L.J. Gauckler, Bioinspired design and assembly of platelet reinforced polymer films. Science, 2008. 319(5866): p. 1069-1073.

21. Xu, Z., et al., Bioinspired Multifunctional Ceramic Platelet‐Reinforced Piezoelectric Polymer Composite. Advanced Engineering Materials, 2017. 19(2): p. 1600570.

22. Richardson, J.J., M. Björnmalm, and F. Caruso, Technology-driven layer-by-layer assembly of nanofilms. Science, 2015. 348(6233): p. 2491.

23. Chen, K., et al., Binary synergy strengthening and toughening of bio-inspired nacre-like graphene oxide/sodium alginate composite paper. Acs Nano, 2015. 9(8): p. 8165-8175.

24. Beese, A.M., et al., Defect‐Tolerant Nanocomposites through Bio‐Inspired Stiffness Modulation. Advanced Functional Materials, 2014. 24(19): p. 2883-2891.

25. Walther, A., et al., Large-area, lightweight and thick biomimetic composites with superior material properties via fast, economic, and green pathways. Nano letters, 2010. 10(8): p. 2742-2748.

26. Podsiadlo, P., et al., Ultrastrong and stiff layered polymer nanocomposites. Science, 2007. 318(5847): p. 80-83.

27. Studart, A.R., Towards high‐performance bioinspired composites. Advanced Materials, 2012. 24(37): p. 5024-5044.

28. Mirkhalaf, M., A.K. Dastjerdi, and F. Barthelat, Overcoming the brittleness of glass through bio-inspiration and micro-architecture. Nature communications, 2014. 5: p. 3166.

29. Espinosa, H.D., et al., Tablet-level origin of toughening in abalone shells and translation to synthetic composite materials. Nature communications, 2011. 2: p. 173.

30. Deville, S., et al., Freezing as a path to build complex composites. Science, 2006. 311(5760): p. 515-518.

31. Le Ferrand, H., et al., Magnetically assisted slip casting of bioinspired heterogeneous composites. Nature materials, 2015. 14(11): p. 1172.

32. Naglieri, V., et al., On the development of ice-templated silicon carbide scaffolds for nature-inspired structural materials. Acta Materialia, 2013. 61(18): p. 6948-6957.

33. Bouville, F., et al., Strong, tough and stiff bioinspired ceramics from brittle constituents. Nature materials, 2014. 13(5): p. 508.

34. Bai, H., et al., Bioinspired large-scale aligned porous materials assembled with dual temperature gradients. Science advances, 2015. 1(11): p. e1500849.

35. Porter, M.M., et al., Magnetic freeze casting inspired by nature. Materials Science and Engineering: A, 2012. 556: p. 741-750.

36. Deville, S., E. Saiz, and A.P. Tomsia, Ice-templated porous alumina structures. Acta Materialia, 2007. 55(6): p. 1965-1974.

37. Libanori, R., R.M. Erb, and A.R. Studart, Mechanics of platelet-reinforced composites assembled using mechanical and magnetic stimuli. ACS applied materials & interfaces, 2013. 5(21): p. 10794-10805.

38. Munch, E., et al., Tough, bio-inspired hybrid materials. Science, 2008. 322(5907): p. 1516-1520.

39. Mao, L.-B., et al., Synthetic nacre by predesigned matrix-directed mineralization. Science, 2016. 354(6308): p. 107-110.

40. Niebel, T.P., et al., Role of the polymer phase in the mechanics of nacre-like composites. Journal of the Mechanics and Physics of Solids, 2016. 96: p. 133-146.

41. Chen, Z. and J.J. Mecholsky Jr, Toughening by metallic lamina in nickel/alumina composites. Journal of the American Ceramic Society, 1993. 76(5): p. 1258-1264.

42. Sun, X. and J. Yeomans, Microstructure and fracture toughness of nickel particle toughened alumina matrix composites. Journal of materials science, 1996. 31(4): p. 875- 880.

43. Tuan, W.-H. and R. Brook, The toughening of alumina with nickel inclusions. Journal of the European Ceramic Society, 1990. 6(1): p. 31-37.

44. Slawik, T., et al., Metal–Ceramic Layered Materials and Composites Manufactured Using Powder Techniques. Advanced Engineering Materials, 2014. 16(10): p. 1293- 1302.

45. Rosso, M., Ceramic and metal matrix composites: Routes and properties. Journal of Materials Processing Technology, 2006. 175(1-3): p. 364-375.

46. Mousavian, R.T., et al., Fabrication of aluminum matrix composites reinforced with nano-to micrometer-sized SiC particles. Materials & Design, 2016. 89: p. 58-70.

47. Launey, M.E., et al., A novel biomimetic approach to the design of high-performance ceramic–metal composites. Journal of the Royal Society Interface, 2010. 7(46): p. 741- 753.

48. Roy, S., B. Butz, and A. Wanner, Damage evolution and domain-level anisotropy in metal/ceramic composites exhibiting lamellar microstructures. Acta Materialia, 2010. 58(7): p. 2300-2312.

49. Xu, Z., et al., Bioinspired Nacre‐Like Ceramic with Nickel Inclusions Fabricated by Electroless Plating and Spark Plasma Sintering. Advanced Engineering Materials, 2018. 20(5): p. 1700782.

50. Wilkerson, R.P., et al., A Novel Approach to Developing Biomimetic (“Nacre‐Like”) Metal‐Compliant‐Phase (Nickel–Alumina) Ceramics through Coextrusion. Advanced Materials, 2016. 28(45): p. 10061-10067.

51. Heim, F.M., Y. Zhang, and X. Li, Uniting Strength and Toughness of Al Matrix Composites with Coordinated Al3Ni and Al3Ti Reinforcements. Advanced Engineering Materials, 2018. 20(1): p. 1700605.

52. Liu, J., J. Li, and C. Xu, Interaction of the cutting tools and the ceramic-reinforced metal matrix composites during micro-machining: a review. CIRP Journal of Manufacturing Science and Technology, 2014. 7(2): p. 55-70.

53. Ohnabe, H., et al., Potential application of ceramic matrix composites to aero-engine components. Composites Part A: Applied Science and Manufacturing, 1999. 30(4): p. 489-496.

54. Kaczmar, J., K. Pietrzak, and W. Włosiński, The production and application of metal matrix composite materials. Journal of materials processing technology, 2000. 106(1-3): p. 58-67.

55. Khosroshahi, N.B., et al., Mechanical properties of rolled A356 based composites reinforced by Cu-coated bimodal ceramic particles. Materials & Design, 2015. 83: p. 678-688.

56. Liu, Q., et al., Fabrication of a new SiC/2024Al co-continuous composite with lamellar microstructure and high mechanical properties. Journal of Alloys and Compounds, 2014. 585: p. 146-153.

57. Lombardi, I., et al., Template assisted deposition of Ag nanoparticle arrays for surface- enhanced Raman scattering applications. Sensors and Actuators B: Chemical, 2007. 125(2): p. 353-356.

58. Daryadel, S., et al., Localized Pulsed Electrodeposition Process for Three-Dimensional Printing of Nanotwinned Metallic Nanostructures. Nano letters, 2017. 18(1): p. 208-214.

59. Xu, Z., et al., Evolution of electromechanical and morphological properties of piezoelectric thin films with thermomechanical processing. Polymer, 2016. 106: p. 62-71.

60. Baniasadi, M., et al., Thermo-electromechanical behavior of piezoelectric nanofibers. ACS applied materials & interfaces, 2016. 8(4): p. 2540-2551.

61. Fatigue, A.C.E.o. and Fracture, Standard test method for measurement of fracture toughness. 2011: ASTM International.

62. Yeh, T.S. and M.D. Sacks, Effect of particle size distribution on the sintering of alumina. Journal of the American Ceramic Society, 1988. 71(12): p. C‐484-C‐487.

63. Lide, D.R., CRC handbook of chemistry and physics. 2012: CRC Boca Raton.

64. Huang, J., et al., Lamellar Ceramic Semicrystalline‐Polymer Composite Fabricated by Freeze Casting. Advanced Engineering Materials, 2017. 19(8): p. 1700214.

65. Chao, W., et al., The effect of Ti addition on the microstructure and fracture toughness of BN-Al composite materials synthesized by vacuum infiltration. Archives of Metallurgy and Materials, 2013. 58(2): p. 509-512.

66. Properties, A.S.D.o.M. Standard test methods for flexural properties of unreinforced and reinforced plastics and electrical insulating materials. 1997. American Society for Testing Materials.

BIOGRAPHICAL SKETCH

Jiacheng Huang was born in Shanghai, China. Jiacheng entered the Shanghai University of

Engineering Science, China in 2008, he earned his bachelor’s degree with a major in automobile engineering in 2012. In August 2012, he left China and enrolled as a master’s student at The

University of Texas at Dallas, Richardson, Texas. After studying two years at UTD, he received his master’s degree with a major in mechanical engineering in May 2014. From 2015 to 2018 he was employed as a teaching/research assistant working for Dr. Majid Minary at the Nano-bio lab at UTD.

CURRICULUM VITAE

Jiacheng Huang

Department of Mechanical Engineering Erik Jonsson School of Engineering and Computer Science The University of Texas at Dallas 800 West Campbell Rd., Richardson, TX 75080-3021 Email: [email protected]

EDUCATION Ph.D. in Mechanical Engineering Sep 2015-Dec 2018 The University of Texas at Dallas, Dallas-TX Dissertation title: Bio-Inspired Nacre-Like Ceramic Matrix Composites

M.S. in Mechanical Engineering Sep 2012-May 2014 The University of Texas at Dallas, Dallas-TX Thesis: Electro-mechanical Characterization of Piezoelectric Ropes and Yarns

B.E. in Automobile engineering Sep 2008-June 2012 Shanghai University of Engineering Science, Shanghai, China

RESEARCH EXPERIENCE Graduate Research Assistant Sep 2015-Nov 2018 Department of Mechanical Engineering, UTD Dallas, US

Manufacturing of Metal-Ceramic Composite through Infiltration of Copper into Freeze-Cast Alumina Scaffold by Electroplating • Fabricated copper ceramic composite by freeze casting and electrodeposition. • Characterized the composite using three-point flexural experiment and fracture toughness experiment. • SEM investigation of composite microstructure and toughening mechanism.

Alumina-Nickel Composite Processed via Co-assembly using Freeze-Casting and Spark Plasma Sintering • Fabricated Nickel alumina composite by spark plasma sintering, conducted and analyzed all experimental data • Characterized the composite using three-point flexural experiment and fracture toughness experiment. • SEM investigation of composite microstructure and toughening mechanism. Lamellar Ceramic Semicrystalline-Polymer Composite Fabricated by Freeze Casting • Fabricated thin Nickel alumina composite by freeze casting and vacuum infiltration, conducted and analyzed all experimental data • Characterized the composite using in-situ SEM compression experiment, three- point flexural experiment and in-situ SEM fracture toughness experiment. • SEM investigation of composite microstructure and toughening mechanism.

Research Volunteer Sep 2012-May 2014 Department of Mechanical Engineering, UTD Dallas, US

RESEARCH SKILLS Microstructural Characterization • Scanning electron microscopy • Fourier transform infrared • Atomic force microscopy spectroscopy • Piezo-response force microscopy • Raman spectroscopy • Nano-indentation/ in situ Nano flip • Differential scanning calorimetry • Instron/ in situ mechanical Fabrication measurement • Spark plasma sintering • Electro-spinning • Sputter coating • Electro-less plating • Freeze casting • Electro plating • Atomic layer deposition • Vacuum infiltration Engineering and Programming Software

• Solidworks • Ansys • National Instruments Labview • Matlab

TEACHING EXPERIENCES The University of Texas at Dallas, Dallas-TX - Teaching assistant - Mechanics of Materials Fall 2018 - Teaching assistant - Mechanics of Materials Fall 2016 - Introduction to Nanostructured Materials Spring 2016

PROFESSIONAL MEMBERSHIP • The American Society of Mechanical Engineers

CONFERENCE PRESENTATION - Jiacheng Huang, Zhe Xu, Majid Minary-Jolandan , “Bio-Inspired Nacre-Like Ceramic with Nickel Inclusions Fabricated by Freeze Casting and Spark Plasma Sintering”, ASME 2018 MSEC conference, College station, TX, June 18th, 2018.

PUBLICATIONS

1. Jiacheng Huang, Majid Minary-Jolandan. “Manufacturing of Metal-Ceramic Composite through Infiltration of Copper into Freeze-Cast Alumina Scaffold by Electroplating”, ACS applied materials & interfaces, submitted. 2. Jiacheng Huang, William S Rubink, Hunter Lide, Thomas W. Scharf, Rajarshi Banerjee, Majid Minary-Jolandan. “Alumina-Nickel Composite Processed via Co-assembly using Freeze-Casting and Spark Plasma Sintering”, Advanced Engineering Material, submitted. 3. Jiacheng Huang, Zhe Xu, Moreno Salvador, Seyed Reza Morsali, Zhong Zhou,

Mahmoud Baniasadi, Dong Qian, and Majid Minary-Jolandan. “Lamellar Ceramic- Semicrystalline Piezo Polymer Composite Fabricated by Freeze Casting”, Advanced Engineering Materials 19 (8), 1700214. 4. Mahmoud Baniasadi, Jiacheng Huang, Zhe Xu, Salvador Moreno, Xi Yang, Jason Chang, Manuel Angel Quevedo-Lopez, Mohammad Naraghi, and Majid Minary-Jolandan. "High-performance coils and yarns of polymeric piezoelectric nanofibers."ACS applied materials & interfaces 7, no. 9 (2015): 5358-5366. 5. Zhe Xu, Jiacheng Huang, Cheng Zhang, Soheil Daryadel, Ali Behroozfar, Brandon McWilliams, Benjamin Boesl, Arvind Agarwal, Majid Minary- Jolandan. “Bioinspired Brick-and-Mortar Ceramic-Metal Composite Fabricated by Electro-less Plating and Spark Plasma Sintering”, Advanced Engineering Materials 20 (5), 1700782. 6. D Lingam, AR Parikh, J Huang, A Jain, M Minary-Jolandan. “Nano/microscale pyroelectric energy harvesting: challenges and opportunities”, International Journal of Smart and Nano Materials 4 (4), 229- 245 7. Enlong Yang, Zhe Xu, Lucas K Chur, Ali Behroozfar, Mahmoud Baniasadi, Salvador Moreno, Jiacheng Huang, Jules Gilligan, and Majid Minary- Jolandan. “Nanofibrous Smart Fabrics from Twisted Yarns of Electrospun Piezo Polymer”, ACS applied materials & interfaces 9 (28), 24220-24229