<%BANNER%>

Enhanced Electrochemical Properties of Lithium Iron Phosphate as a Cathode Material for Lithium Ion Rechargeable Batteries

Permanent Link: http://ufdc.ufl.edu/UFE0044991/00001

Material Information

Title: Enhanced Electrochemical Properties of Lithium Iron Phosphate as a Cathode Material for Lithium Ion Rechargeable Batteries
Physical Description: 1 online resource (157 p.)
Language: english
Creator: Lee, Jungbae
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

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

Notes

Abstract: Intrinsically poor conductivity of LFP has hindered the realization of its high theoretical capacity (170mAhg-1). In an attempt to solve these issues, in this work we have investigated the enhancement in electrochemical properties of LFP by process modification and surface modification. Firstly, the electrochemical improved performance of bare LFP with surfactant processing is introduced. The addition of surfactant as a dispersing agent during vibratory ball milling of LiFePO4 (LFP) precursor showed better size uniformity, morphology control, and reduced particle size when anionic surfactant (Avanel S-150) was used. Electrodes fabricated from LFP particles by solid state reaction involving vibratory milling showed a 22% decrease in capacity after 50 cycles, whereas the performance of electrode prepared by surfactant processed LFP showed only 3% loss in capacity. Secondly, in order to improve the electrical conductivity of LFP cathode, metal (Cu nano-flakes with very high surface area) with a polyethylene glycol (PEG) as carbon source and dispersant was incorporated in the cathode by ball milling Cu nano-flakes. Uniformly dispersed Cu flakes subsequently transformed to CuO during the calcination process. Interestingly, Cu flakes was used as a catalyst for transforming carbon from disordered to graphitic carbon from the analysis of ID/IG ratio with the help of Cu flakes during calcinations process. The Cu incorporated LFP composite cathode showed a high capacity of 161 mAhg-1, displayed excellent high rate and cyclic performance. Lastly, In order to improve the electrical conductivity of LFP cathode with a consideration of a reduced cost of the coating material, metal oxide was employed. This idea was originated from the transformation of metal to metal oxide as mentioned in the above. ZnO/Carbon was incorporated in the cathode by ball milling ZnO, PEG and LFP particles together. Herein, the catalytic property of ZnO for carbon transformation was confirmed again through the analysis of ID/IG ratio. The uniformly dispersed carbon and ZnO on the surface of LFP led to a good electronic contact between the LFP grains. Thus, an excellent high rate performance up to 10C was successfully achieved.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jungbae Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Singh, Rajiv K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044991:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044991/00001

Material Information

Title: Enhanced Electrochemical Properties of Lithium Iron Phosphate as a Cathode Material for Lithium Ion Rechargeable Batteries
Physical Description: 1 online resource (157 p.)
Language: english
Creator: Lee, Jungbae
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

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

Notes

Abstract: Intrinsically poor conductivity of LFP has hindered the realization of its high theoretical capacity (170mAhg-1). In an attempt to solve these issues, in this work we have investigated the enhancement in electrochemical properties of LFP by process modification and surface modification. Firstly, the electrochemical improved performance of bare LFP with surfactant processing is introduced. The addition of surfactant as a dispersing agent during vibratory ball milling of LiFePO4 (LFP) precursor showed better size uniformity, morphology control, and reduced particle size when anionic surfactant (Avanel S-150) was used. Electrodes fabricated from LFP particles by solid state reaction involving vibratory milling showed a 22% decrease in capacity after 50 cycles, whereas the performance of electrode prepared by surfactant processed LFP showed only 3% loss in capacity. Secondly, in order to improve the electrical conductivity of LFP cathode, metal (Cu nano-flakes with very high surface area) with a polyethylene glycol (PEG) as carbon source and dispersant was incorporated in the cathode by ball milling Cu nano-flakes. Uniformly dispersed Cu flakes subsequently transformed to CuO during the calcination process. Interestingly, Cu flakes was used as a catalyst for transforming carbon from disordered to graphitic carbon from the analysis of ID/IG ratio with the help of Cu flakes during calcinations process. The Cu incorporated LFP composite cathode showed a high capacity of 161 mAhg-1, displayed excellent high rate and cyclic performance. Lastly, In order to improve the electrical conductivity of LFP cathode with a consideration of a reduced cost of the coating material, metal oxide was employed. This idea was originated from the transformation of metal to metal oxide as mentioned in the above. ZnO/Carbon was incorporated in the cathode by ball milling ZnO, PEG and LFP particles together. Herein, the catalytic property of ZnO for carbon transformation was confirmed again through the analysis of ID/IG ratio. The uniformly dispersed carbon and ZnO on the surface of LFP led to a good electronic contact between the LFP grains. Thus, an excellent high rate performance up to 10C was successfully achieved.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jungbae Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2012.
Local: Adviser: Singh, Rajiv K.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044991:00001


This item has the following downloads:


Full Text

PAGE 1

1 ENHANCED ELECTROCHEMICAL PROPERTIES OF LITHIUM IRON PHOSPHATE AS A CATHODE MATERIAL FOR LITHIUM ION RECHARGEABLE BATTERIES By JUNGBAE LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 20 12

PAGE 2

2 20 12 Jungbae Lee

PAGE 3

3 T o my loving family

PAGE 4

4 ACKNOWLEDGMENTS First of all, I would like to thank Jesus, for his ready made plan for me I would like to acknowledge my advisor, Dr. Rajiv K. Singh, for his support and guidance. His faith in me helped me move forward through the tough times in my research and was a constant source of motivation. His virtues and mentorship encouraged me to get an achievement and helped me expand my potential. I would like to thank Dr. Brij Moudgil for his invaluable suggestions and fruitful discussions. I would also like to thank my other committee m embers (Dr. Stephen Pearton, Dr. David Norton and Dr. Changwon Park) for their precious time, valuable guidance and constructive comments for providing feedback about my research. and current group members: Dr. Purushottam Kumar, Jinhyung Lee, Minfei Xue, Jongcheol Kim and Dib jyoti Das for providing me excellent research environment and being Eric research. I had really enjoyed working with them for the last two years. I also wish to acknowledge all my friends at the University of Florida with whom I had a wonderful t ime in Gainesville. Matthew Camaratta, Rui Quing, Jaewoong Lee, Sungwook Mhin, Sunghwan Yeo, Wooram Yoon, Hyuksoo Han, Jiho Park, Hyeongeun Yoo, Soojin Paik and Yoonseon Heo. All of them had provided me a great friendship here in UF. Also, I would like to thank my elder sisters and brothers in law for their encouraging words for keeping me enthusias tic all the time my elder brother and his wife who spent much time worrying and praying about my circumstances during my

PAGE 5

5 Ph.D work. I would like to express my s incere gratitude to my parents for their solid faith and giving love in me and encouraging me to be ambitious. For their endless support father is not here, I believe that he is shedding tears with a great pleasure on the sky for consideration and support. Above all I would like to thank my soul mate, the only my lovely wife, Kyeyoung, for h er endless love, encouragement, quiet patience and support toward me. Thanks to her devotion, I could focus on my work without any difficulties. Their love and belief in me has shaped me into who I am today. I cannot imagine ong support behind me.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LIST OF FIGURES ................................ ................................ ................................ ........ 10 ABSTRACT ................................ ................................ ................................ ................... 13 C H A PTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 1.1 Problem and Motivation ................................ ................................ .................... 15 1. 2 Organization o f Dissertation ................................ ................................ .............. 16 2 BACKGROUND AND LITERATURE REVIEW ................................ ....................... 18 2. 1 Fundamentals of Battery ................................ ................................ ................... 18 2. 1.1 Oxidation and Reduc tion ................................ ................................ ......... 18 2. 1.2 Major Characteristics ................................ ................................ ............... 19 2 1 2 .1 Cell voltage ................................ ................................ .................... 19 2 1 2 2 Energy density ................................ ................................ ............... 20 2 1 2 3 Theoretical capacity ................................ ................................ ....... 20 2. 2 Rechargeable Batteries ................................ ................................ .................... 20 2. 3 Lithium Ion Batteries ................................ ................................ ......................... 22 2. 4 Current State of Arts ................................ ................................ ......................... 25 2. 4.1 Crystal Structure of Lithium Iron Phosphate ................................ ............ 25 2. 4.1 Current Efforts for Improved Electrochemical Properties ......................... 27 2 4 .1.1 Doping methods ................................ ................................ ............. 27 2 4 .1. 2 Coating methods ................................ ................................ ............ 31 2 4 .1. 3 Particle size reduction ................................ ................................ .... 33 2 4 .1. 4 Synthesis methods ................................ ................................ ......... 34 2. 5 Solid State Methods ................................ ................................ .......................... 35 2. 6 Electrochemical Characterizations ................................ ................................ .... 36 2. 6.1 Potential Intermittent Titration Technique (PITT) ................................ ..... 36 2. 6.2 Alternative Current Impedance (ACI) ................................ ....................... 38 3 EXPERIMENTAL PROCEDURES ................................ ................................ .......... 47 3. 1 Powder Synthesis ................................ ................................ ............................. 47 3. 1 .1 Material Selection ................................ ................................ .................... 47 3. 1 2 Powder P rocessing ................................ ................................ .................. 47 3. 1 3 Conventional S intering ................................ ................................ ............ 48 3. 2 Sample P reparation f or C haracterization ................................ .......................... 48

PAGE 7

7 3. 2 1 X R ay D iffraction ................................ ................................ ..................... 48 3. 2 2 Scanning E lectron M icroscopy ................................ ................................ 48 3. 2 3 Transmission E lectron M icroscopy ................................ .......................... 49 3. 2 4 X Ray Photoelectron Spectroscopy ................................ ......................... 49 3. 2 5 Raman Spectroscopy ................................ ................................ .............. 49 3. 2 6 Specific Surface Area and Density Measurements ................................ 50 3 2 7 Particle Size and Size Distribution Measurements ................................ .. 50 3. 3 Electrochemical Measurements ................................ ................................ ........ 50 3. 3 1 Cathode Preparation and Galvanostatic Measurement ........................... 50 3. 3 2 Experimental Setup for Alternative Current Impedance ........................... 51 3. 3 3 Electrical Conductivity Measurement ................................ ....................... 52 4 OPTIMIZATION OF BARE ACTIVE MATERIAL AS A CATHODE ......................... 55 4. 1 Experimental ................................ ................................ ................................ ..... 56 4. 1 .1 Synthesis of Active Material ................................ ................................ .... 56 4. 1 2 Material Characterization ................................ ................................ ......... 57 4. 1 3 Electrochemical Characterization ................................ ............................ 57 4. 2 Results and Discuss ion ................................ ................................ ..................... 58 4. 2 1 Crystal Structure and Particle Morphology ................................ .............. 58 4. 3 2 Electrochemical Characteristics ................................ .............................. 60 4. 3 Summary ................................ ................................ ................................ .......... 62 5 ELECTROCHEMICAL PERFORMANCE OF SURFACTANT PROCESSED LFP AS A CATHODE MATERIAL ................................ ................................ .................. 73 5 1 Experimental ................................ ................................ ................................ ..... 74 5 1 .1 Synthesis of Surfactant Processed Active Material ................................ 74 5 1 2 Material Characterization ................................ ................................ ......... 75 5 1 3 Electrochemical Characterization ................................ ............................ 76 5 2 Results and Discussion ................................ ................................ ..................... 76 5 2 .1 Phase Analysis of Prepared Particles ................................ ...................... 76 5 2 2 Particle Size Distribution and Specific Surface Area ............................... 78 5 2 3 Electrochemical Characteristics ................................ .............................. 80 5 3 Summary ................................ ................................ ................................ .......... 83 6 ELECTROCHEMICAL ENHANCEMENT OF LFP AS A CATHODE MATERIAL BY INCORPORATING CU NANO FLAKES ................................ ............................ 94 6 1 Experimental ................................ ................................ ................................ ..... 95 6 1 .1 Synthesis of Cu Nano Flake Inco rporated Active Material ....................... 95 6 1 2 Material Characterization ................................ ................................ ......... 96 6 1 3 Electrochemical Characterization ................................ ............................ 97 6 2 Results and Discussion ................................ ................................ ..................... 98 6 2 1 Thermal Analysis ................................ ................................ ..................... 98 6 2 2 Crystal Structure and Particle Morphology ................................ .............. 98 6 2 3 Carbon Structural Analysis (I D /I G ) ................................ .......................... 101

PAGE 8

8 6 2 4 Electrochemical Characteristics ................................ ............................ 102 6 3 Summary ................................ ................................ ................................ ........ 105 7 ZINC OXIDE INCORPORATED LFP FOR HIGH RATE ELECTROCHEMICAL PERFORMANCE ................................ ................................ ................................ .. 119 7 1 Experimental ................................ ................................ ................................ ... 120 7 1 .1 Synthesis of Zinc Oxide Incorporated Active Material ........................... 120 7 1 2 Material Characterization ................................ ................................ ....... 121 7 1 3 Electrochemical Characterization ................................ .......................... 121 7. 2 Results a nd Discussion ................................ ................................ ................... 122 7. 2 .1 Crystal Structure and Particle Morphology ................................ ............ 122 7. 2 2 Carbon Structural Analysis (I D /I G ) ................................ .......................... 125 7. 2 3 Electro chemical Characteristics ................................ ............................ 127 7. 3 Summar y ................................ ................................ ................................ ........ 131 8 CONCLUSIONS AND FUTURE WORK ................................ ............................... 144 8.1 Conclusions ................................ ................................ ................................ .... 144 8.2 Future Work ................................ ................................ ................................ .... 147 LIST OF REFERENCES ................................ ................................ ............................. 148 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 157

PAGE 9

9 LIST OF TABLES Table Page 2 1 Comparisons of three typical rechargeable batteries ................................ .......... 21 2 2 Comparison data among various Lithium base batteries ................................ .... 25 4 1 Comparisons of lattice parameters and crystallite size of all samples ................ 59 4 2 Particle size data for volume of all samples ................................ ........................ 60 5 1 Comparisons of lattice parameters and crystallite size ................................ ....... 78 5 2 Specific surface area (SSA) and density of samples ................................ .......... 79 5 3 EIS parameters of the samples ................................ ................................ .......... 83 6 1 Comparisons of crystallite size of all samples ................................ .................... 99 6 2 Specific surface area(SSA) of samples throu gh BET measurement ................ 100 7 1 Comparisons of FWHM and crystallite size of all samples ............................... 123 7 2 Specific surface area(SSA) of samples through BET measurement ................ 124 7 3 Raman spectra parameters of samples ................................ ............................ 126 7 4 EIS parameters of the samples ................................ ................................ ........ 131

PAGE 10

10 LIST OF FIGURES Figure page 2 1 Schematic of reaction between A(IV) and B(II) with electron flow through external circuit (only X can put through ( ) ion selective membrane) ................. 40 2 2 Energy density comparison of various batteries ................................ ................. 40 2 3 Charge and discharge diagram of LiFePO 4 /graphite battery .............................. 41 2 4 Cycling behavior at 55 o C of an o ptimized LFP/C composite electrode at a rate of C/10 ................................ ................................ ................................ ......... 41 2 5 Crystal structure compariso n ................................ ................................ .............. 42 2 6 Schematic of comparison of partial and homogeneous carbon coating as electrons path ................................ ................................ ................................ ..... 43 2 7 Particle morphology comparison according to the various synthesis methods ... 44 2 8 Vibratory milling and schematic of milling process ................................ ............. 45 2 9 Equivalent electrical circuit of an ele ctrochemical cell for ACI ............................ 46 2 10 Nyquist plot for an electrochemical cell ................................ .............................. 46 3 1 Materials comparison. ................................ ................................ ........................ 53 3 2 Schematic of electrochemical measurement using 2016 coin type cell .............. 54 3 3 Schematic of Alternative Current Impedance (ACI) measurement experimental setup ................................ ................................ ............................. 54 4 1 Schematic diagram of LFP synthesis process ................................ .................... 63 4 2 XRD patterns of LFP calcined at from 600~750 o C after vibratory ball milling .... 64 4 3 Rietveld refinement s using Fullprof program of LFP ................................ ........... 66 4 4 Typical FE SEM images of LFP ................................ ................................ .......... 67 4 5 Particle size distribution (PSD) ................................ ................................ ........... 68 4 6 Comparison of mean particle size distribution (MPSD) of volume percent and specific surface area (SSA) according to the calcination temperature ................ 69 4 7 Electrochemical properties ................................ ................................ ................. 70

PAGE 11

11 4 8 Cycle performance comparison between LFP w/o and w/ surfactant at C/10 ..... 71 4 9 Structure transformation during charge and discharge ................................ ....... 72 5 1 Structural Formula of Avanel S 150 anionic surfactant ................................ ....... 85 5 2 Schematic diagram of surfactant processed LFP synthesis ............................... 85 5 3 XRD patterns of LFP precursor after vibratory ball milling and before heat treatment ................................ ................................ ................................ ............ 86 5 4 XRD pattern comparison of LFP w/ and w/o surfactant prepared by solid state method ................................ ................................ ................................ ....... 87 5 5 Rietveld refinement using Fullprof program of LFP ................................ ............ 88 5 6 Typical FE SEM images ................................ ................................ ..................... 89 5 7 P article size distribution ................................ ................................ ...................... 90 5 8 Initial and 50 th charge/discharge capacity comparison between LFP w/ and w/o surfactant at C/10 ................................ ................................ ......................... 91 5 9 Cycle performance comparison between LFP w/o and w/ surfactant at C/10 ..... 92 5 10 EIS spectra of the LFP prepared by vibratory ball milling w/o and w/ surfactant in the frequency range between 100 kHz and 10 mHz ...................... 93 6 1 Schematic diagram of LFP and Cu incorporated LFP synthesis ....................... 107 6 2 TGA curves of the PEG, bare LFP and LFP/PEG precursor with 10 wt.% ....... 107 6 3 XRD analysis compar ison ................................ ................................ ................ 108 6 4 FE SEM image of Cu nano flake ................................ ................................ ...... 109 6 5 FE SEM images (The insets ind icate magnified image(x100,000) ................... 111 6 6 XPS spectra ................................ ................................ ................................ ...... 112 6 7 Raman spe ctra of LFP, LFP/PEG, and LFP/PEG/Cu (The insets indicate resolved raman spectra using Gaussian distribution fitting) ............................. 113 6 8 PITT measurement data of LFP/PEG/Cu (3wt. %) between 3.44 and 3.45V (The inset indicates ln ( I t /A) vs. t (s)) ................................ ............................... 114 6 9 Chemical diffusio n coefficient comparisons of LFP, LFP/PEG, and LFP/PEG/Cu using PITT method between 3.4 and 3.58V ................................ 115

PAGE 12

12 6 10 Electrochemical properti es ................................ ................................ ............... 116 6 11 Cycle performance comparison between bare LFP and LFP/PEG/Cu discharge capacity at C/10 ................................ ................................ ............... 117 6 12 Electrochemical Impedance Spectra of bare LFP and LFP/PEG/Cu ................ 118 7 1 Schematic diagram of LFP and ZnO coated LFP synthesis ............................. 132 7 2 XRD patterns of LFP, LFP/PEG and LFP/PEG/ZnO ................................ ........ 132 7 3 FE SEM image of ZnO ................................ ................................ ..................... 133 7 4 FE SEM images and EDX spectrum ................................ ................................ 135 7 5 Surface analysis ................................ ................................ ............................... 136 7 6 XPS spectra ................................ ................................ ................................ ...... 137 7 7 Raman spectra of LFP, LFP/PEG, and LFP/PEG/ZnO ................................ ..... 138 7 8 Diff usion coefficient measurements ................................ ................................ .. 139 7 9 Chemical diffusion coefficient comparisons of LFP, LFP/PEG, and LFP/PEG//ZnO using PITT method between 3.4 and 3.58V ............................ 140 7 10 Electrochemical voltage profiles ................................ ................................ ....... 141 7 11 Electrochemical test comparison ................................ ................................ ...... 142 7 12 Electrochemical Impedance Spectra of bare LFP, LFP/PEG and LFP/PEG/ZnO (The insets indicate an equivalent circuit for left hand side and LFP/PEG/ZnO (2wt. %) for right hand side) ................................ ...................... 143

PAGE 13

13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ENHANCED ELECTROCHEMICAL PROPERTIES OF LITHIUM IRON PHOSPHATE AS A CATHODE M A TERIAL FOR LITHIUM ION RECHARGEABLE BATTERIES By Jungbae Lee December 20 12 Chair: Rajiv K. Singh Major: Materials Science and Engineering I ntrinsically poor conductivity of LFP has hindered the realization of its hig h theoretical capacity (170 mAh g 1 ). In an attempt to solve these issue s in this work we have investigated the enhancement in electrochemical properties of LFP by p rocess modification and surface modification. Firstly, the electrochemical improved performance of bare LFP with surfactant processing is introduced. The addition of surfactant as a dispersing agent during vibratory ball milling of LiFePO4 (LFP) precursor showed better size uniformity, morphology control, and reduced particle size when anionic surfactant (Avanel S 150) was used. Electrodes fabricated from LFP particles by solid state reaction involving vibratory milling showed a 22% decrease in capacity after 50 cycles, whereas the performance of electrode prepared by surfactant processed LFP showed only 3% loss in capacity. Secondly, i n order to improve the electric al conductivity of LFP cathode, metal ( Cu nano flakes w ith very high surface area ) with a polyethylene glycol (PEG) as carbon source and dispersant was incorporated in the cathode by ball milling Cu nano flakes. Uniformly dispersed Cu flakes subsequently transformed to CuO during the calcination process. Inter estingly, Cu flakes was used as a catalyst for transforming carbon from disordered to graphitic carbon from the analysis

PAGE 14

14 of I D /I G ratio with the help of Cu flakes during calcinations process. The Cu incorporated LFP composite cathode showed a high capacity of 161 mAhg 1, displayed excellent high rate and cyclic performance. Lastly, In order to improve the electrical conductivity of LFP cathode with a consideration of a reduced cost of the coating material metal oxide was employed. This idea was originated from the transformation of metal to metal oxide as mentioned in the above. ZnO/Carbon was incorporated in the cathode by ball milling ZnO, PEG and LFP particles together. Herein, the catalytic property of ZnO for carbon transformation was confirmed again t hrough the analysis of I D /I G ratio. T he uniformly dispersed carbon and ZnO on the surface of LFP led to a good electronic contact between the LFP grains. Thus, an excellent high rate performance up to 10C was successfully achieved.

PAGE 15

15 CHAPTER 1 INTRODUCTION 1. 1 Problem a nd Motivation With the increase in demand for energy one of the greatest challenge s is to make a renewable re source of energy due to limited oil resources and issues of global warming originated from fuel energy us e 1 2 To replace the fuel energy, there are many kinds of renewable resource of energy such as wind power, hydraulic power, nuclear power, geothermal power and solar power energy. Based on these renewable energies, many kinds of energy storages such as batterie s, fuel cells, solar cells have been developed Among them, n ext generation of consumer products from mobile electronics to electric vehicles (EV)s or hybrid electric vehicles (HEV)s demands high energy density rechargeable batteries. Lithium Ion Batteries (LIB) s are the current choice for high energy density and light weight energy storage devices. Industry experts expected lithium ion batteries to be the promising energy storage of the global market for use in hybrid electric vehicles (HEV)s because li ght weight l i thium ion batteries with high power density and energy density can store electricity significantly 3 4 Actuall y lithium ion battery market has grown incredibly from mobile to car industry. Specially, the market value of lithium ion batteries as power suppliers for EV is estimated $15.9 billion in 2019 which is 5 times as compared with the value of $3.4 billion in 2013 acco rding to the estimation report by JP Morgan. The value of $15.9 billion as mentioned above is only about the market of lithium ion batteries for EV s Thus, it is natural that so many researchers and companies have participated in competency for developing standard power sources. As a result, the performance of lithium ion battery has been improved incredibly. However, for the application of power supplier for EV there are still many

PAGE 16

16 requirements of researching about designing light weight, high energy densi ty and high power density material with a consideration of safety under a severe condition in order to use the batteries all around the world. Above all, reducing price per unit cell is key point for the application of lithium ion batteries. Therefore, the abundant materials such on earth can be the candidates for a cost effective material as electrode material s in lithium ion batteries. However, the first marketed lithium ion cathode material was LiCoO 2 (LCO) and it has been currently used as a cathode mat erial because LCO has a high theoretical capacity of 280 mAhg 1 Though LiCoO 2 (LCO) is currently the principal cathode material for commercial LIB, it suffers from many disadvantages such as toxicity, high cost of the rare Co mineral resources, and unreliability from explosive reactions at high temperatures. Among the candidates to replace LCO, lithium iron phosphate (LFP) offers highly promising properties e.g. high theoretical capacity, eco friendly, cheap, and thermally stable structure 5 1. 2 Organization o f Dissertation The purpose of this study is to understand the fundamental of electrochemical performance enhancement of LiFePO 4 m aterial for the application of l ithium ion secondary batteries. In Chapter 2, an overview of fundamental of lithium ion batteries (LIBs) and literature review about current state of ar t for improving electrochemical properties of LiFePO 4 (LFP) were introduced. In C hapter 3, experimental procedures and characterization techniques were discussed for investigating physical, chemical, and electrochemical properties of LFP materials employed in this work. For the deepest and reasonable approach of bare LFP material, the study about the bare LFP material was discussed in C hapter 4. Based on this s tudy presented in C hapter 4, particle size and

PAGE 17

17 size distribution effect s on electrochemical properties of LFP materials synthesized by wet vibratory milling so called soli d state method with and without surfact ant addition were discussed in C hapter 5. Chapter 6 presented the effect of metal coating used as a catalyst for transformation of disordered carbon to graphitic carbon. H e rein, Cu nano flakes were employed Chapter 7 is simi lar work to C hapter 6 in the respect of using catalysis for improving the electrical conductivity of LFP materials, however the addition of metal oxide (herein, ZnO nano powder ) was introduced to improve the e lectrochemical rate capabilities. Lastly, in C h apter 8 summaries of all results and discussions were presented

PAGE 18

18 CHAPTER 2 B ACKGROUND AND LITERA TURE REVIEW 2. 1 Fundamentals o f Battery 2. 1 .1 Oxidation a nd R eduction The essence of e lectrochemistry is that some species are gaining electrons and others are l osing electrons when participating in reaction. As a result of that, electric current s flow according to the electron energy level difference For instance, there are two species A (IV) and B (II). When they are mixed, A (IV) is changed to A(III) by g aining electron while B(II) is changed to B (III) by losing electron like following : 6 8 a A (IV) + n e aA (III) (2 1) b B (II) bB (III) + n e (2 2) For the whole reaction, charge compensation rule is well observed like following: aA (IV) + bB (II) aA (III) + bB (III) (2 3) So, A (IV) specie is an oxidizing agent and B (II) specie is a reducing agent as shown in Fig 2 1. From the Thermodynamics 2 nd law, the relation between cell potential (E) and gibbs free energy ( G) is like following: (2 4) w here n is the number of electrons and F is the faraday constant (96500 C). The unit of E and F are V and C mol 1 respectively. Thus, G is represented as J mol 1 From the equation 2 4, cell potential can be repres ented like following equation 2 5 in terms of a potential ( ) and activity ( P ): (2 5 ) where R is the gas constant (8.314 JK 1 mol 1 ) and T is temperature (K). The term of (A(IV,III) ) (B(III,II) ) can be substituted with the standard potential (E o )

PAGE 19

19 Thus, equation 2 5 can be simply arranged like following equation 2 6 which is the Nernst e quation : (2 6 ) 2. 1 .2 Major C haracteristics There are three major characteristics in energy conversion and storage devices : 1. O perating voltage which is related to power output 2. Current related to charge rate 3. Capacity related to how long it can last 2 1 2 .1 Cell v oltage The cell voltage can be obtained from the difference between each electrode potential in the cell. Note that c ell voltage can be obtained from the Nernst equation. For the detailed information, let s look into the following reaction of the firstly designed battery, Daniel cell : 6 7 9 Anode: Zn Zn 2+ + 2e E o = 0.76 V (2 7) Cathode: Cu 2+ + 2e Cu E o = 0.34 V (2 8) ----------------------------------------------------------------------Cell: Cu 2+ + Zn Zn 2+ + Cu E o = 1.1 V (2 9) The oxidation reaction occurred in Zn electrode and reduction reaction in Cu electrode. Thus Zn and Cu electrode s are anode and cathode respectively. As a result, total cell voltage of 1.1V can be obtained. However, the output cell voltage is a little lower than the theoretical value due to the internal resistance caused by activation polarization and concentration p olarization Activation polarization is caused by inhibition of the passage of potential determining ions (Herein, Zn 2+ and Cu 2+ ) at the

PAGE 20

20 interface between electrode and electrolyte. Concentration polarization is caused by the inhibition of transport process due to concentration difference. 2 1 2 2 Energy density Energy of a cell can be obtained by the relation of cell voltage times electricity equivalent to 1 mol (96500C or 26.8 Ah) like following : 6 8 Energy of a cell [ W h ] = Cell voltage [ V ] x E lectricity [ Ah ] (2 10 ) Energy density can be expressed in unit of [WhKg 1 ] or [Whl 1 ]. Thus the energy of a cell calculated from the equation 2 10 divided by a weight or volume of active materials can yield the energy density. For the comparison study of the various battery cells, Fig. 2 2 shows the comparison of various battery cells in terms of energy density. Li metal is the ideal case, however, during the cycle of charge and discharge reaction there is a dendrite growth of Li metal by an adsorption on the Li metal electrode, which causes a short circuit. Thus, it is unsafe material. Beyond Li metal, Li ion battery is the most ideal application fo r a battery use. 2 1 2 3 Theoretical capacity T he actual output capacity is less than the theoretical capacity because the t heoretical capacity can be obtained only at an ideal state. Simply the theoretical capacity can be calculated by the electricity equivalent to 1 mol divided by molecular weight of active material like follow ing : 6 8 (2 11) 2. 2 Rechargeable Batteries There are various types of rechargeable batteries such as Lead acid battery, Nickel Metal Hydride (Ni MH) and Lithium ion batteries. Lead acid battery is composed

PAGE 21

21 of active materials such as lead and lead dioxide, sulfuric acid electrolyte and gl ass fib er separator Lead sulfate produced from the electrode reactions is an insulator and hinders the available energy by build up on the electrode Ni MH battery is composed of active materials such as nickel oxyhydroxide and hydrogen absorbing alloy N i MH replaced the Ni Cd battery and ha s two times the capacity of Ni Cd when both Ni MH and Ni Cd are manufactured in the same size. Its working voltage is too low. Lithium ion battery is composed of graphite anode, various cathode materials such as layered, spinel and olivine structured materials, polymer separator and electrolytes. T he detailed information about the LIBs will be discussed in section 2.3. C haracteristic c omparisons of these three type s of batteries are listed in the following Table 2 1 Table 2 1 C omparison s of three typical rechargeable batter ies 10 Items L IB Ni MH Lead acid Working voltage (V) 3.7 1.2 2.0 Gravimetric energy density (Whkg 1 ) 130 200 60 90 30 40 Volumetric energy density (Whl 1 ) 340 400 200 250 130 180 Cycle life (cycles) 500 400 300 Capacity self discharge rate (% per month) 5% 30% 10% Energy efficiency (C discharge /C charge ) 99% 70% 75% Reliability High Low High Lead acid B attery ha s been used as power suppliers for the EVs and HEVs Among these three different products and technologies, LIB is the best power supplier for EV and HEV applications, because of its long cycle life high operating voltage and high gravimetric and volumetric energy density. Even though Lead acid Battery is highly reliable, it is an outdate d technology which is not proper for the application of EV s because of its low gravimetric and volumetric energy density due to the large scale and

PAGE 22

22 heavy weight Furthermore, toxic property by using the materials such as sulfuric acid (H 2 SO 4 ) and lead (Pb) even hinders lead acid battery for using green energy application According to the degree of both g ravimetric energy density ( Whkg 1 ) and volumetric energy density ( Whl 1 ) the candidate electrode materials can be evaluated for the applications in the LIB s 11 13 For example, the ma terial with high energy density is required for a light weight system like portable devices and the material high volumetric energy density is for the compact system. 2. 3 Lithium Ion Batteries Since the development of l ithium ion rechargeable batteries (LIB) s of LiCoO 2 (LCO)/C system in 1991 by Sony Corporation 14 LIBs have been widely u sed as a power source in various por table electronic devices such as mobile phones, laptop computers, digital cameras and recently even in electrical vehicles (EV) and hybrid electrical vehicles (HEV) due to its high energy density, long cycle life and excellent safety There ha ve been many reports why they receive the great attention at both fund a mental and application 11 15 24 There are four components which consist of LIBs. Those are lithium ion intercalation positive electrode (i.e. generally lithium transition metal oxide s such as layered structured LiMO 2 (M = Mn, Co, Ni), spinel structured LiMn 2 O 4 and olivine structured LiMPO 4 (M = Fe Mn ) ) negative electrode (generally layered structured graphite), Li salt containing electrolyte (i.e. LiPF 6 in EC( ethylene carbonate ) DEC( diethylcarbonate ) or EC( ethylene carbonate ) DMC (dimethylcarbonate) ) for a medium phase between electrode s and separator made of polymer for separating electrodes perfectly and passing through back and forth for only Li + ion as shown in Fig. 2 3 25 Actually, the capacities of cathode materials are too low compared t o th ose of

PAGE 23

23 anode materials, unsafe in overcharge state and very expensive. So, intensive studies should be performed in order to replace them. The requirements of cathode material candidates for charging Li + ion batteries are like following : 15 1. Transition metal which is capable of oxidation and reduction. 2. No structural change when intercalated to cathode material 3. Material which can deliver high capacity and generate high voltage 4. Material which can lead fast Li + ion intercalation resulting in fast charging 5. High electrical conductivity for a capable of transferring electrons 6. High stability at an ambient temperature 7. Cost effective material 8. Environment friendly material There are three typical types of lithium ion battery cathode as mentioned above They are lithium cobalt oxide (LiCoO 2 ) which is a principal cathode material lithium manganese oxide ( LiMnO 2 ) which has a layered structure LiMnO 2 another different chemical formula of lithium manganese oxide ( LiMn 2 O 4 ) which is a spinel structure and lithium iron phosphate ( LiFePO 4 ) with olivine structure Even t hough LiCoO 2 (LCO) has been widely used as a principal cathod e material for commercial LIB s there has been a continuous effort to replace it with other novel cathode materials because of certain drawbacks such as toxicity, high cost of the rare mineral resource Co an d unreliability of explosiveness at high temperature s LiMnO 2 which is cost effective, safe and higher capacity than that of spinel l ithium manganese oxide h as been studied at the high rate. However, its poor cycle performance at high temperature is the weakness to be solved Among the candidates to replace LCO, LiFePO 4 (LFP) is the most attractive cathode material because of its high theoretical capacity (170 mAhg 1 ), structural stability, low cost and eco friendly characteristic 5 Despite these advantages a major disadvantage of LFP is its poor rate performance mainly caused by intrinsically low

PAGE 24

24 electrical conductivity (10 9 to 10 10 Scm 1 ) at room temperature compared to LiCoO 2 (10 3 Scm 1 ) and LiMn 2 O 4 (10 5 Scm 1 ) which results from very slow lithium ion diffusion rate 26 30 Another challenge is the easy oxidation of LFP from Fe 2+ to Fe 3+ during synthesis process. In terms of high cost, toxicity, poor safety and chemical instability t he conventional LiCoO 2 Li Mn O 2 Li Ni O 2 and LiMn 2 O 4 cathode materials are not appropriate as electric supplier s for EV s and other applications As an a lterna tive cathode candidate material for LIBs lithium iron phosphate is the best promising cathode material for LIBs by comparing with other cathode candidate materials listed in the Table 2 2 LFP with olivine structure shows a high lithium intercalation voltage ( 3.5 V vs Li metal ) as shown in Fig. 2 4 high theoretical capacity (170 mAhg 1 ) and an excellent chemical stability with common ly used organic electrolyte s of LiPF 6 in EC:DEC or EC:DMC 31 Lithium iron phosphate was developed as a promising cathode material for rechargeable lithium ion batteries in 1997 by John Goodenough 5 However, t he LiFePO 4 could not be used as a commercial cathode material because it has critical drawbacks like intrinsically too low conductivity as mentioned above. Furthermore, LFP shows two phase (LiFePO 4 and FePO 4 ) reactions. It means that there is only one oxidation reaction like Fe 2 + or Fe 3+ during charge/discharge cycling exhibiting a poor conducting property. Thus LFP could not be recognized as a promising cathode electrode candidate for LIBs However, thanks to the many efforts fo r improving the drawbacks which LFP has, p resently it exhibited an approximate 90% of its theoretical capacity ( > 16 0 mAhg 1 ) at a rate of 0. 2 C showing high rate performance Therefore,

PAGE 25

25 LFP can be considered as a promising cathode candidate for the next generation of LIBs 32 More detailed discussion a bout crystal structure and current efforts for improving electrochemical properties of LFP will be introduced in the following section. Table 2 2 Comparison data among various Lithium base batteries 10 33 Battery LiFePO 4 LiCoO 2 LiMn 2 O 4 Li(NiCo)O 2 Safety Safest Not Stable Acceptable Not Stable Environmental Concern Most Enviro friendly Very Dangerous Acceptable Very Dangerous Cycle life Best/Excellent Acceptable Acceptable Acceptable Power/Weight Density Acceptable Good Acceptable Best Long Term Cost Most Economic/Excellent High Acceptable High Temperature Range Excellent ( 20 to 70 o C) Decay beyond ( 20 to 55 o C) Decay extremely fast over 50 o C 20 to 55 o C 2. 4 Current State o f Arts 2. 4.1 Crystal Structure o f L ithium Iron Phosphate Li, Fe and P atoms in LFP are located in 4a, 4c and 4c sites respectively forming LiO 6 FeO 6 octahedra and PO 4 tetrahedra as shown in Fig. 2 5 (a). In LFP olivine structure ( S.G. (62) Pnma ) LFP becomes very structurally stable due to t he strong PO 4 covalent bonding. Furthermore, the Fe 2+/3+ redox couple leads to maintain frame crystal structure without any changes and chemical stabilities than M 3+/4+ redox couples during charge and discharge reactions resulting in good safety features u nlike layered LiMO 2 (M = Mn Co, Ni ) oxides When all of the lithium atoms are extracted from olivine L FP during charge reaction FePO 4 (FP) can be formed with a same orthorhombic structure ( S.G. (62), Pbnm ) As a counter reaction, LFP can be formed during discharge reaction. Thus, two phases of FePO 4 (heterosite ) or LiFePO 4 (triphylite) are forming repeatedly

PAGE 26

26 during charge/discharge cycling. No structural change during cycling led LFP to be a promising cathode ca n didate material. However, LFP shows poor rate performance and limited capacity of x=0.6 in Li x FePO 4 when synthesized by solid state method because LFP has an intrinsically poor electrical conductivity than those of layered and spinel structured cathode candidates as discussed in section 2.3. The materials synthesized with ca rbon used as a conducting agent at a low temperature (conducting material), nearly theoretical capacity was obtained during charge and discharge cycling 34 35 Howe ver, a s compared with other cathode materials (LiMO 2 M= Mn Co, Ni) with layered structure, The Li O 6 octahedras are located to form the linear chains of the alternat ing planes and the no continuous network of edge shared FeO 6 octahedras occupy the zigzag chains which limits the Li atom intercalation, while MO 6 octahedras in layered structured materials have the continuous network as shown in Fig. 2 5 (b). Among the lithium metal oxides of layered structure d materials including spinel structure oxygen arrays in a cubic close packed system provid es more space for Li + ion motion and might contribute to electronic conductivity These materials are good Li + ion conductors as a result of that they show inherent ly good electronic conductivity from the mixed valent cations in the cathode materials For instance, Mn 3+ / 4+ in Li 1 x Mn 2 O 4 and Co 3+ / 4 + ions in Li 1 x CoO 2 ( 0
PAGE 27

27 1 0 ) surface of FePO 4 indicated the low est energy. This result means that Li + ions extraction from LFP cathode material prefers through two low energy surfaces ( 0 1 0 ) and ( 2 0 1 ) during the charge reaction and Li + ions insertion occurs preferably through the surface (0 1 0) during discharge reaction. 36 The lithiation through one channel can limit the Li + ion intercalation. As a result, LFP cathode delivers relatively small current densiti es due to Li + ion insertion reaction through one channel 37 39 Furthermore, LFP is a poor conductor because of only one oxidation state of Fe 2+ or Fe 3+ d uring the electrochemical charge and discharge r eactions as mentioned in section 2.3. Therefore, the LFP cathode materials with olivine structure show poor rate performance and low Li + ion diffusion rate This can be the critical impediments for wide application s of LFP as a promising cathode candidate in LIBs 16 35 40 41 2. 4.1 Current Efforts f or I mprove d E lectrochemical P roperties M any efforts have been tried to improve the electrochemical properties of LFP There have been three ways for obtaining enhanced electrochemical performances Among them, the first way is doping method with certain elements with higher valences making defec ts such as vacanc ies to improve the intrinsic ally poor conductivity The second way is coating method with conducting materials such as C, Ag, Cu and etc. By using them, s urface modification of the LFP can be achieved to reduce the interfacial resistance b etween electrode and electrolyte Finally, the last thing is size reduction to nano scale resulting in decrease of Li + ion diffusion path by using various synthesis methods. 2 4 .1.1 Doping methods For the doping methods, three sites such as Li, Fe and O sites are available with doped atoms. For a doping methods using diverse atoms in Li sites, Chung et al. have

PAGE 28

28 reported that controlled cation non stoichiometry combined with solid solution doping by metal ions such as Nb 5+ Ti 4+ Zr 4+ Al 3+ and Mg 2+ supervalent to Li + ion increase d the electronic conductivity of LFP by a factor of ~10 8 26 The controlled cations doping methods with a rare earth material of La 3+ in Li sites were reported by several battery researchers and also increased the electronic conductivity of LFP resulting in improved electrochemical performance 42 44 Li et al. reported that t he s ubmicron sized sample of 3 at % Ti doped LFP synthesized by co precipitation and normal temperature r e duction method show ed the most impressive 100 cycling performance and discharge capacity of 150.1 mAhg 1 at 0.1C rate. 45 Yang et al. reported that small size (100 200 nm) of Li 0.98 Cu 0.01 FePO 4 synthesized by improved co precipitation allow intercalation and de intercalation of lithium ions to occur with ease during the charge discharge processes and exhibit high discharge capacity of 154.5 mAhg 1 at 0.1C rate 46 Yin et al. reported single phase Li 0.97 Na 0.03 FePO 4 /C samples are synthesized by in situ polymerization restriction carbo thermal reduction method showed the smaller charge transfer resistance than other samples, the highest reversible capacity of 158 mAhg 1 at 0.1C rate and the improved rate capabilit y. It might be attributed to the larger lattice constants in both a and c resulting in enough space for Li + ion motion when doped with Na + ions. 47 Zhang et al. reported Li 0.99 Nd 0.01 FePO 4 /C cathode composite with t he olivine structure prepared by a novel solid state reaction method at 750 C without using inert gas. 48 The main advantage of this method is that the oxidation of Fe 2+ to Fe 3+ could be avoided without inert gas. The Nd doping could increase the number of vacancy in Li sites resulting in enhanc ing the intrinsic electronic conductivity of LFP like following reaction: (2 12 )

PAGE 29

29 Ying et al. reported t he high tap density 1.8 gcm 3 of the spherical Li 0.97 Cr 0.01 FePO 4 /C powders synthesized by carbothermal reduction method showed 152 mAhg 1 at 0.1C rate. This is attributed to the high density of cathode materials due to the spherical shape of powders. However, there are some reports against these electrochemically improved results by cation doping in Li sites. 49 Ouyang et al. reported that the enormously enhanced electronic conductivity through Li site doping does not improve the electrochemical performance as expected for LFP cathode material. The results show that the heavy Cr ions block ed the one dimensional diffusion pathway for the Li + ion migration from the Monte Carlo simulation. The capacity loss is attributed to the amount of Cr ion doping. It has still been controversial that the conductivity is improved by employing supervalent doping. 50 Wagemak er et al. reported that cation doped LFP (Li 1 xy D xy + FePO 4 ; D = Zr, Nb, Cr) synthesized by a solid state method at 600 C exhibited only 0.3% increase in the size of the lithium channels which is not expected to better the Li ion mobility. Furthermore, the location of the immobile cation dopant s in Li sites may hinder Li + ion diffusion 51 Likewise, doping in Fe sites with small amount of metals such as Cr Mn, Co Ni Cu and Zn also resulted in enhanced electrochemical performances. It is mainly attributed to the enhanc ement of the electronic conductivity. In addition to this the interaction between Li and O can be changed by doping in Fe sites. For the reports about this, Abbate et al. carried out t he Fe 2p X ray absorption spectroscopy (XAS) analysis. The results are that the Fe 2+ high spin state in pristine LFP and the electronic structure of these ions were not significantly affected by the Ti Al and Cu doping while t he O 1s XAS spectra show ed extra absorption intensity 52 Yang et al. reported

PAGE 30

30 that LiFe 0.95 M 0.05 PO 4 ( M = Mg 2+ Ni 2+ Al 3+ V 3+ ) prepared by the solution method showed t he synergetic effect of the supervalent doping and lattice expansion which led improved electrochemical performance of increase in discharg e capacity and reducing capacity fade. 53 In the same way, Liu et al. reported that LiZn 0.01 Fe 0.99 PO 4 synthesized by solid state route do not destroy the lattice structure of LFP and enlarge s the lattice volume. During Li + ion intercalation process of lithium ions, the doped zinc atoms protect the L FP structure for the movement of Li + ions Consequently, the conductivity and li thium ion diffusion coefficient are enhanced by Zn doping 54 As a result, Wang et al. could obtain the improvement of Li + ion diffusion coefficient and electronic conductivity with LiFe 1 x M x PO 4 (M = Ni, Co, Mg) 55 Shin et al. reported that LiFe 0.97 Cr 0.03 PO 4 /C synthesized by a mechano chemical process showed an excellent discharge rate performance. The Cr doping facilitates the phase transformation between LiFePO 4 and FePO 4 during cycling, and conductivity improvement by carbon coating. 56 Li et al. reported o livine composites LiFe 0.95 Mn 0.05 PO 4 synthesized by mechano chemical activation method using C 2 H 2 O 4 as the chelating reagent showed the discharge capacities of 155.6 mAhg 1 and 102.9 mAhg 1 at 1 C and 10 C rate respectively. 57 Yoon et al. reported that t he Fe and Co K edge XAS results of Li Fe 0.5 Co 0.5 PO 4 system showed that the major charge compensation at the metal sites during charge is achieved by the oxidation of Fe 2+ ions at lower potential plateau ( ~3 .6 V) and the oxidation of Co 2+ ions at higher potential plateau ( ~ 5.0 V). After Co doping, strong P O bonds become less covalent during delithiation process due to the increased covalency of Fe 3+ O bond s. 58 Zhang et al. reported that Ni doped LFP /C prepared by the solid state method using low cost asphalt

PAGE 31

31 as a carbon source and a reduction agent exhibit ed better electrochemical performances than the bare material 59 Chang et al. reported that the s toichiometric LiFe 0.98 Cu 0.02 PO 4 /C cathode materials with a high tap density of 1.98 g cm 3 synthesized by a solid state method showed the excellent discharge capaci ty of about 150 mAhg 1 at a rate of 0.1 C 60 Finally, the anion doping using F and Cl in O sites have been tried to improve electrochemical properties. Liao et al. reported that LiFe(PO 4 ) 0.9 F 0.3 /C materials synthesized by a ball milling method with the addition of LiF to the starting materials of LFP /C showed the most attractive high rate performance. 61 Yang et al. reported o livine type LiFe(PO 4 ) 0.97 Cl 0.09 /C samples prepared by low temperature solid state method showed increase d conductivity and high rate performance. 62 2 4 .1. 2 Coating methods Besides doping methods, surface modification by conducting agents such as Ag, Cu polymer and carbon has been designed to improve the electrochemical properties of bare LFP. Moreover, metal oxide also can be used as coating material for surface modification resulting in improved electrochemical performances. How is this possible? Carbon coating provides the electrons paths and compresses the particle growth as shown in Fig. 2 6 Thus, in case of homogeneous carbon coating on LFP particle s electrons can flow well without blockage and small particle size can help Li + ion diffusion. As a result, electrochem ical properties can be enhanced. The optimization of coating carbon thickness can also significantly influence on the electrochemical performances of LFP. Cho et al reported that LFP/C composite materials were prepared by t wo types of carbon source of 20 wt. % polystyrene (PS) and 50 wt. % malonic acid used as a carbon vapor source 63 The carbon thick ness is

PAGE 32

32 related to the carbon contents. That is, with the increase of carbon content the thickness of carbon film increases. Among the samples, LFP/C composite materials with 2.28 wt. % of carbon contents and thickness of 4 8 nm showed the best electrochem ical performance. In addition, the structure of carbon is also key point for improving electrochemical performance of LFP. Wilcox et al. reported that carbon structural factors such as sp 2 /sp 3 and disordered/graphene ( D/G ) ratio influenced enormously the conductivity and rate performance strongly. 64 S ignificant lower ing of the D/G ratio and a concomitant rise in the sp 2 /sp 3 ratio of the carbon resulted in improved electrochemical performances by a ddition of both iron nitrate and PA Croce et al. reported that the addition of very small amount of 1 wt. % of Cu or Ag to LFP prepared by sol gel method showed the discharge capacity improvement of 140 mAhg 1 at 0.2C rate. The addition of Ag or Cu d id n t influence the LFP structure and considerably improve d conductivity. 65 Huang et al. reported that 16 wt. % of polymer polypyrole(PPY) coated LFP and 7 wt. % of polymer p olyaniline(PANI) co coated LFP showed discharge capacity of 145 mAhg 1 and 140 mAhg 1 respectively. 66 Liu et al. reported that LFP/C composite were prepared by a high energy ball milling combined with spray drying method Even though LFP/C composite exhibited the low tap density, t his material delivers an improved tap density of 1.3 gcm 3 and high electronic conductivity of 10 2 to 10 3 Scm 1 The discharge capacities are 109 mAh g 1 at 6.5C rate and 94 mAhg 1 at 11C rate respectively 67 By using metal oxide for a surface modification, there have been many achievements in obtaining enhanced electrochemical properties. Son et al. reported that

PAGE 33

33 nano crystallized Al 2 O 3 coated LFP synthesized by a novel dry coating method exhibited an improvement in both interfacial resistance and cathode polarization. This behavior may be attributed to the prohibition of structural change by Al 2 O 3 coating. 68 ZrO 2 coated LFP powders prepared by chemical precipitation method by Liu et al. 69 and CeO 2 coat ed LFP powders synthesized by sol gel method by Yao et al. 70 showed improved electrochemical properties due to the decrease in electrode polarization and interfacial charge resistance. 2 4 .1. 3 Particle s ize reduction P a rticle size and size distribution are very important factor for Li + ion diffusion path. In this section, particle size effect on electrochemical properties will be discussed. Mcneil et al. reported that two types of LFP/C with particle size of ~10 m and 0.5 ~ 1 m are compared to study the size effect on electrochemical properties. LFP/C with 0.5~1 m showed the discharge capacity of 129 mAhg 1 while LFP/C with ~10 m exhibited 79 mAhg 1 71 Delacou rt et al. reported that inspite of no addition of carbon to LFP, LFP with a particle size of 100 200 nm showed the improved discharge capacity of 147 mAhg 1 72 Kim et al. reported that LFP/C composite with a particle size of 20 nm showed the high rate performance of 163 mAhg 1 at 1C rate. Even though particle size reduction improved the electrochemical performance, it doesn t gu ara ntee that unlimited capacity and rate performance can be obtained as the particle size decreases to few nano meters 37 For example, Zhang et al. reported that some bare LFP and LFP/C composite materials with a larger particle size of 150 nm showed higher discharge capacity than the materials with smaller particle of ~ 100 nm. 73

PAGE 34

34 2 4 .1. 4 Synthesis m ethods Various synthesis techniques has been used to prepare LFP with a variety of particle morphology and size distributions. Fig. 2 7 shows the SEM and TEM images of particle morphology prepared by various synthesis methods. Yan et al. reported that LFP/C cathode material synthesized by a solid state reaction using ci tric acid as a carbon source showed an improved electric conductivity of by eight orders of magnitude from 10 9 Scm 1 The spherical type LFP/C particles (Fig. 2 7 (a)) deliver ed an initial discharge capacity of 128mAhg 1 at 4C 74 Nien et al. reported that polystyrene (PS) containing L FP/C composite (Fig. 2 7 (b)) prepared by co precipitation method showed a lower I D /I G and sp 3 /sp 2 peak ratio indicating more highly graphite like carbon formation during polymer pyrolysis resulting in better electrochemical performance. 75 Kim et al. reported that t he porous phase pure LFP/C composite particles with a few nanomete rs thick layer of carbon synthesized by sol gel method showed porous particle morphology as shown in Fig. 2 7 (c) 76 Zhou et al. reported that t he LFP/C prepare d by co precipitation and microwave processing showed high surface area of 98.3 m 2 g 1 and porous carbon morphology as shown in Fig. 2 7 (d) 77 Jin et al. reported that LFP cathode materials with 5wt. % carbon black using solid polymer electrolyte (SPE) prepared by hydrothermal methods showed more uniform and round in shape with increasing carbon black content as shown in Fig. 2 7 (e) 78 Muraliganth et al. reported that bundles of LFP nano rod s with an average size of 40 6 nm synthesized by a rapid microwave sintering followed by s olvo thermal metho d within 5 minutes at temperatures as low as 300 o C didn t require any post annealing in reducing gas atmospheres as shown in Fig. 2 7 (f). 79 The result was that d ue to th e high surface area of nano rod,

PAGE 35

35 multi walled carbon nanotube (MW CNT) coating on it could be successfully achieved for improved electrochemical performances. 2. 5 Solid State Methods The large sized particles are transformed into small sized ones by mechanical forces following the processes such as crushing, grinding and milling. This is comminution process. Traditionally, powders of size range of > 1 mm were obtained by the size reduction process using jaw, gyratory, cone, compound crusher and shaft impactor Even though there is a harmful effect of media contamination on synthesized powders when milling was used for a long time, m illing process is the most common method to obtain th e size reduction Thu s, it has been still used for synthesis of powders in industrial. In the milling process es of compressing and shearing there are mechanical stresses which can lead defo r mation at the contact point or face between particles and between particles and mill ing media Through the given mechanical energy from milling, new surfaces are created and there is a shape a nd size change in the particle s There are various types of b all mills For instance, those are vibration, tumbling and agitation mill. Among them, vibratory milling used in this work will be discussed because it has the greatest impact energy among three mills Fig. 2 8 shows the model of vibratory milling machine and schematic of crushing process between particles and milling media. Vibratory milling machine is composed of drum, which is vibrating upward and downward and milling jar which is placed in the drum. By using milling jar filled with milling media of yttria stabilized zirconia (YSZ) an energy efficient and effective millin g process in all direction is expected to distribute the particles well resulting in preventing segregation of particles when using wet milling. The shape of media is hexahedron and

PAGE 36

36 should fill at least over 8 0% of milling jar for an efficient milling Cur rently, by adding some polymeric additives more uniform size distribution and less agglomeration of particles can be obtained 80 2. 6 Electrochemical Characterizations 2. 6. 1 Potential Intermittent Titration T echnique (PITT) Potentiostatic intermittent titration technique (PITT) is experimentally used to measure the chemical diffusion coefficient of Li + ion during charge or discharge reaction of the lithium ion cell by applying constant current ( I ) as a function of time Note that this technique application started based on the supposition of that the permeability are equal at both boundaries because there ar e still many limitations such as insufficient information to develop the property difference at both boundaries by mathematical approach. PITT measurements are derived from 81 (2 13 ) where t is time, x is the distance, C is the concentration of Li + ion and D Li is the Li + ion chemical diffusion coefficient. The boundary conditions are categorized into three conditions like initial, semi infinite and electrode surface boundary conditions like following s: 81 82 (a) Initial conditions Li + ion is uniformly distributed throughout the solution, C is equal to C o ( initial concentration) for all x region like following. (2 1 4 ) (b) Semi infinite boundary conditions

PAGE 37

37 When the distance is enough large from the electrode wall, there is no alteration process of solution at the wall of the cell, thus the concentration has a constant value. For thin layer cell, if the distance to the wall of the cell is l x should be l like following: (2 1 5 ) (c) Electrode surface boundary conditions The last condition is related to the concentration gradient at the electrode surface I f the current is quantity the boundary condition is like following : (2 1 6 ) where C s is the concentration at the interface between electrode and electrolyte From these boundary conditions, the time dependent current I(t) is related to the concentration gradient at the interface between electrode and electrolyte it can be represented like following: 81 82 (2 17 ) where z i is the charge number of the electro active specie i F is faraday s constant (96500 C) and A is the cross sectional area of the electrode As a result, t he current response can be expressed as a function of time In case of the condition of t << L 2 /D ( L: characteristic length of electrode ) which indicates initial stage of diffusion the chemical diffusion coefficient D Li can be obtained from relation of I vs. t 1/2 like following Cottrell equation : 81 83 (2 1 8 ) If the condition is t>>L 2 /D which is the long time approximation the chemical diffusion coefficient can be determined from the relation of I vs e t which can be simply converted

PAGE 38

38 to the relation of ln( I(t) ) vs. t by applying natural logarithm followed by differential like following : (2 1 9 ) 2. 6.2 Alternative Current I mpedance (ACI) Alternating current impedance (ACI) spectroscopy is a widely used technique for the characterization of various electrochemical reactions at the interface between electrode and electrolyte An electrochemical cell can be understood acco rding to the equivalent circuit which is composed of el ectrolyte resistance (R s ), constant phase element (CPE) for an electrical double layer capacitance (C d ) and faradic impedance (Z f ) which is divided into Warburg impedance (Z w ) and charge transfer resistance (R ct ) as shown in Fig. 2 9 The measured impedance data, which is called as Nyquist plot, is plot ted on X axis of real impedance (Z re ) and Y axis of imaginary impedance ( Z im ) as shown in Fig. 2 10 The nyquist plot is divided into charge transfer control region at a high mid dle fr equency ( ) range and mass transfer control region at low frequency range. The impedance is represented like following two equations 2 20 and 2 21 : 81 (2 20 ) (2 21 ) At low frequency region, as 0, the impedance can be rep res ented as like following two equations: (2 22 ) (2 23 )

PAGE 39

39 At high frequency region, Warburg impedance can be ignored compared to R ct That is, as 0, t he impedance can be represented as like following two equation s: (2 2 4 ) (2 2 5 ) By combining above two equations 2 2 4 and 2 2 5 the following 2 26 is obtained: (2 2 6 ) Thus, circular plot is obtained, which is a center of (Z re =R + R ct /2, Z im =0) and a radius of R ct /2. The line with a slope at low frequency range is related to the diffusion of Li + ion in the solid electrodes, semi circle at mid frequency range for charge transfer process and semi circle intercept at high frequency limit for electrolyte resistance. The ratio of voltage and current based on Ohm s law determines the Impedance at a spe cific frequency ( )

PAGE 40

40 Figure 2 1 Schematic of reaction between A(IV) and B(II) with electron flow through external circuit (only X can put through ( ) ion selective membrane) 6 Figure 2 2 Energy density comparison of various batteries 32

PAGE 41

41 Figure 2 3 Charge and discharge diagram of LiFePO 4 /graphite battery 12 16 Figure 2 4 Cycling behavior at 55 o C of an optimized LFP/C composite electrode at a rate of C/10 32

PAGE 42

42 A B Figure 2 5 Crystal structure c omparison A ) Comparison of c rystal structure s of LiFePO 4 (Triphylite) and FePO 4 ( heterosite) B ) C omparison of layered structured materials and olivine structure 84

PAGE 43

43 Figure 2 6 Schematic of comparison of partial and homogeneous carbon coating as electrons path 85

PAGE 44

44 Figure 2 7 Particle morphology comparison according to the various synthesis method s. A ) S olid state method. 74 B ) C o precipitation method. 75 C ) S ol gel method. 76 D ) M icrowave method. 77 E ) H ydrothermal method. 78 F ) S olvothermal microwave method 79

PAGE 45

45 Figure 2 8 Vibratory milling and schematic of milling process 80

PAGE 46

46 Figure 2 9 Equivalent electrical circuit of an electrochemical cell for ACI 81 Figure 2 10 Nyquist plot for an electrochemical cell 81

PAGE 47

47 CHAPTER 3 EXPERIMENTAL PROCEDU RES 3. 1 Powder S ynthesis In this section, the detail ed information of experimental procedures that were used to synthesize phase pure powder samples for different characterizations such as XRD ,SEM, TEM, XPS, Raman, specific surface area and density measurement will be discussed. 3. 1 .1 Material S election Fig. 3 1 shows the conductivity as a reference of Cu and price per pound of various materials. From these data, Cu material was selected as a coating material for the direct metal coating on LFP in Chapter 6 because Cu is the best coating materials among all materi al candidates in terms of conductivity and cost. For the stability with no structural and chemical formula change at high temperature, ZnO was chosen as a metal oxide coating on LFP resulting in forming composite with LFP in Chapter 7 expecting high conduc tivity and cost effectiveness. 3. 1 2 Powder P rocessing Olivine LiFePO 4 powder was manufactured by using wet ball milling called solid state mechanical method. Precursors such as Li 2 CO 3 for Li source, FeC 2 O 4 2H 2 O for Fe source and NH 4 H 2 PO 4 for PO 4 source were mixed in an anhydrous ethanol with a molar ratio of 0.5:1:1. Then, the well mixed precursors were ball milled with yttria stabilized zirconia (YSZ) media for 1day. After wet ball milling, the form of slurry was obtained. It is dried in a dryin g oven at 70 o C for 16h.

PAGE 48

48 3. 1 3 Conventional S intering The precursor powders synthesized using solid state method which is wet vibratory ball milling, were firstly calcined at 350 o C for 4 h in order to decompose oxalate, carbonate and ammonium followed by sintering at the range of 600 ~ 7 50 o C for 10 h under a reduced atmosphere of 5 % H 2 in Ar The ramp rate of heating and cooling was 2 o C/min 3. 2 Sample P reparation f or C haracterization After sintering process of crystallized LFP powders with phase pure they were processed for different characterizations such as XRD, SEM, TEM, XPS, Raman, particle size distribution measurement and specific surface area measurement In this section vario us sample preparation techniques for the different characterization analysis will be introduced 3. 2 1 X R ay D iffraction X ray diffraction (XRD) was performed on the bare LFP and C and metal/metal oxide coated LFP samples to study the crystalline phase and parameter s related to crystal structure such as crystallite size, full width at half maximum (FWHM) and lattice parameters X ray diffraction was performed in the 2 range of 10 ~ 70 o using X ray diffractometers (APD 3720, Philips and X pert powder PANalytical ) on the sintered fine particles of LFP 3. 2 2 Scanning E lectron M icroscopy Field emitting s canning electron microscope ( FE SEM 6335F JEOL ) was used to characterize the microstructur e and surface morphology. A conductive gold was coated on the particle sample in order to reduce surface charge resulting in better image. SEM analysis was done at the operating voltage of 15 kV and working distance of 15 mm

PAGE 49

49 E nergy dispersive X ray (EDX) linked to SEM was analyzed to identify the elements of the sam ples based on the energy. 3. 2 3 Transmission E lectron M icroscopy Transmission electron microscopy (TEM, JEOL 2010F TEM ) was used to examine the nanoscale microstructure of the particles at 200 kV having an ultra thin Be window for EDX, attached with a Link Analyzer. After sonication for 1 h the powder sample was mounted on the Cu TEM sample holder with a circle ring shape and then TEM analysis was performed. 3. 2 4 X R ay Photoelectron S pectroscopy XPS (Perkin Elmer PHI 5100 ESCA) will be used to obtain elemental compositional information from the surface of powder samples This technique is capable of leading both general surface qualitative chemical state (compositional analysis) and q uantitative analysis due to its high surface sensitivity. X ray source is monochromatized Mg K with energy of 1253.6 eV with a binding energy range of 0 ~ 1000 eV. The binding energy was tuned based on C 1s peak (284.5 eV) used as a reference 3. 2 5 Raman S pectroscopy Raman spectroscopy will be used to study the status ( Disordered/ Graphene ) of C in the powder sample by analyzing the I D /I G intensity ratio. I nelastic scattering and R aman scattering of monoc hromatic light from a laser in the visibl e light range have an influence on it Raman analysis in this work was performed with a laser wave length of 532 nm.

PAGE 50

50 3. 2 6 Specific S urface A rea a nd D ensity M easurements The surface area was characterized by using Quantachrome NOVA 1200 which enables to measure N 2 gas sorption on particle surfaces Prior to gas adsorption, the powder sample is degassed and dried in a vacuum The specific surface area (SSA) m easurements use multipoint Brunauer Emmett Teller (BET) method Prior to measuring SSA, the density of the sample should be known. The Quantachrome Ultrapyc 1000 Gas Pycnometer enables to measure a density of powder sample The apparent density of the material is calculated by the volume obtained from the following relation when different pressures were applied: P 1 x V 1 = P 2 x V 2 (3 1) 3. 2 7 Particle S ize a nd Si ze D istribution M easurements The particle size and size distribution was measured using Coulter LS 13 320 from 40 nm to 2,000 microns by laser light scattering (LS). The principle is that light intensity scattered from the particles which are passing through the detector can be variable according to the particle size shape, scattering angle, wavelength, a nd the material whose light refractive index is variable Prior to measuring particle size and size distribution powder samples were sonicated in d e ionized (DI) water for 1 h in order to well disperse the powders in the solution for the precise particle size and size distribution 3. 3 Electrochemical M easurements 3. 3 1 Cathode P reparation a nd G alvanostatic M easurement All cathodes used in this work were prepared by mixing 85 wt. % of the active mat erial of LFP with 1 5 wt % Acetylene Carbon Black (TIMCAL) and 5 wt % poly vinylidene fluoride (PVDF) in N methyl pyrrolidone (NMP) solution. The slurry was

PAGE 51

51 poured and casted using a doctor blade on an Al foil which is used as a current collector and substrate Then, the cast Al sheets are dried in a vacuum oven at 120 C for 12 hr The electrode s were punched into disc shape with 1.4 cm diameter and were put in an argon filled glove box ( Brown, H 2 O level < 0.3 ppm) for the cell assembling The active cathode material loaded on Al current collector is approximately 0.7 mgcm 2 All electrochemical charge and discharge tests were done using a 2016 (20 mm: diameter, 1.6 mm: thickness) coin type cell. Fig. 3 2 show the schematic of the electrochemical measurement employing 2016 coin type cell used in this work Disc shaped l ithium metal was used as both reference and counter electrode and 1M LiPF 6 in ethylene carbonate: dimethyl carbonate (EC:DMC v/v = 1:1 ) solution (Novolyte) was used as an electrolyte. Polypropylene ( C480, Celgard Inc ) was used as a separator. Once coin cells were assembled in an Ar filled glovebox for the stabilization after 3 ~ 9 hr t he coin cells were tested galvanostatically on an Arbin battery tester in the voltage range of 2.5 ~ 4.0 V and at various C rates such as C/10, C/5, C/2, 1C, 2C, 5C, 10C for the rate performance test The 1C of charge and discharge rate was estimated by the weight of electrode with a theoretical capacity of LFP ( 170 mAhg 1 ). The cycle test s w ere performed by 50 th charge and discharge cycles at 0.1C or 0.2C rate. 3. 3 2 Experimental Setu p f or A lternative C urrent I mpedance After cell was assembled in Ar filled glove box, it was tested galvanostatically for 3 cycles at a discharge terminated voltage of 2.5 V, which indicates that Li + ion s are inserted into cathode material resulting in formation of LiFePO 4 Then, alternative current impedance (ACI) was measured with a voltage amplitude of 5 mV and frequency range of 100 mHz ~ 100 KHz. The charge transfer resistance of the cell was

PAGE 52

52 determined based on the eq uivalent circuit. The schematic of ACI measurement using an electrochemical workstation (CHI660 ) is shown in Fig. 3 3 3. 3 3 Electrical Conductivity M easurement In order to measure the conductivity, 0.1 g samples of the precursor powders were pressed into disk shaped pellets at a pressure of 8000 psi for 10 min with a 10 mm diameter punch and die. The ethanol was used as a lubricant, thus for an evaporation of eth anol after pressing the pellets are dried in the oven until no residual ethanol is left followed by sintering at 650 o C for 10h under reduced atmosphere The ramp rate for heating and cooling is 2 o C/ min Prior to conductivity measurements, both faces of disc shaped pellets were sputtered by gold. Conductivity measurements were performed with two point and four point d.c. methods using a Keithley Model 2001 Digital Multimeter was used.

PAGE 53

53 A B Figure 3 1 Mat erials comparison 86 A ) C onductivity comparison as a reference of Cu B ) Comparison of materials in price per pound

PAGE 54

54 Figure 3 2 Schematic of electrochemical measurement using 2016 coin type cell Figure 3 3 Schematic of Alternative Current Impedance (AC I) measurement experimental set up

PAGE 55

55 CHAPTER 4 OPTIMIZATION OF BARE ACTIVE MATERIAL AS A CATHODE The mechanism of solid state mechanical milling or alloying (MA) using solid liquid reaction is to synthesize product by mechanical forces The MA process can be divided into five stages: i) starting period, ii) welding period iii) equiaxed particle forma tion period i v ) random welding period and v ) steady state processing. To achieve a homogeneous particle size distribution as well as an appropriate hardness is dependent on t h e onset of steady state processing 87 By using wet vibratory ball milling, three dimensional milling is beneficial to minimize the segregation of particles. The newly formed out p roduct was obtained from the repeated process es of peel ing and break ing resulting in fine particle formation during the collision process among the balls in all directions There is a fast solid liquid reaction on the peeled off region on the surface resul ting in formation of new layer. Once the solid material participated in the solid liquid reaction is completely exhausted the formation process of output product is ended up Furthermore, through the solid liquid reaction milling it is possible to make nanometer sized particles In this C hapter, i n order to study the electrochemical properties of LFP cathode bare LFP powders were prepared by vibratory wet ball milling of Li 2 CO 3 FeC 2 O 4 2H 2 O and NH 4 H 2 PO 4 followed by heat treatment at different temperatu res ranging from 600 to 750 o C. It has been suggested that LiFePO 4 (LFP) with a poor crystalline is obtained if the calcination temperature is too low, while if it is too high it might result in larger particle sizes. Thus, optimizing the calcination t emp erature for picking up the best electrochemical performance delivering LFP is the key point The preparation conditions

PAGE 56

56 are variable dependent on the calcination temperature to investigate the temperature effect on LFP. 4 1 Experimental 4. 1 .1 Synthesis o f Active M aterial Solid state reaction method was used to synthesize LFP particles by wet vibratory ball milling stoichiometric amounts of lithium carbonate ( Li 2 CO 3 >98%, Alfa Aesar), iron(II) oxalate ( FeC 2 O 4 2H 2 O, >99%, Alfa Aesar), ammonium di hydrogenophosphate ( NH 4 H 2 PO 4 >98%, ACROS) in 200 ml anhydrous ethanol followed by a two step heat treatment process. The precursors were milled for 24 h thereafter the mixture was rinsed with ethanol 3 times using centrifuge and filtered. Fig 4 1 shows the schematic of the synthesis process. T he precursor mixture after rinsing was dried at 50 C for 16 h in a drying oven. The dried powders were subjected to a two step heat treatment process. It was first heated to 350 o C for 4 h to decompose the carbonate, oxalate, and ammonium mixture of the starting materials followed by heat treatment at the range of 600~750 o C for 10 h to crystallize LFP. All heat treatments were done under reduced atmosphere formed by continuously flowing 5% H 2 in Ar to p revent the oxidation of Fe from Fe 2+ to Fe 3+ The ramp rate for heating/cooling was 2 K min 1 at each step. Prior to heating, the furnace was purged with 5% H 2 in Ar gas for 20 min. After heat treatment the particles were manually grinded for characterizat ion and electrochemical measurements. Samples prepared by solid state reaction and subsequent calcination process according to the sintering temperature are denoted as LFP600, LFP650, LFP700 and LFP750.

PAGE 57

57 4 1 2 Material C haracterization Phase analysis and crystallinity of the heat treated particle was measured using powder X ray diffractometer (XRD, Philips APD 3720). XRD was done using Cu K radiation source ( = 1.5406 ) with a step size of 0.02 o a scan rate of 0.05 deg/sec and a 2 range of 10 70 o Latt ice parameters were determined by Rietveld refinement using Fullprof program. Crystallite size was calculated using the Scherrer s equation with average full width at half maximum (FWHM) of (111), (211), (301), (311), (121) peaks. Field emission scanning e lectron microscopy (FE SEM, JEOL FEG SEM 6335) was used to observe the morphology of the particles. P article size distribution (PSD) was measured using Coulter L ight S cattering (LS 13320 ) which can measure particle size >0.04 m. The particles were sonicate d for 1 h before SEM and particle size measurements. Density was measured three times using Pycnometry ( Quantachrome Ultrapyc 1000 Gas Pycnometer ). The specific surface area was measured using Brunauer Emmett Teller (BET, Nova 1200) method 4. 1 3 Electrochemical C haracterization The electrode s were prepared by coating slurries of active material (80 wt %), Acetylene Carbon Black (1 5 wt .% ) and polyvinylidine fluoride (PVdF, 5 wt %) dissolved in N methyl pyr r olidi n one (NMP) using doctor blade on alum inum foil as a current collector. Prior to coating, the slurry was stirred in a 20 ml vial for 24 h using magnetic bar and stirrer. After coating, the electrodes were dried for 4 h at 120 C in low pressure (200 mTorr) atmosphere and pressed The electrode material (1 mg) was loaded on the disc shape d (1 4 mm in diameter 7 m thick ) current collector Coin type test cells (2016) were assembled in an argon filled glove box in which H 2 O level was automatically maintained below 0.1 ppm. Celgard 400 (Celgard In c.) was used as a

PAGE 58

58 separator, 1M LiPF 6 in ethylene carbonate (EC)/ dimethyl carbonate (DMC) (1:1 v/v ) as an electrolyte, and Li foil as a reference and counter electrode. Discharge (Li insertion)/ charge (Li extraction) to/from LFP cathode were carried out galvanostati sti cally using battery cycler (Arbin Instrument) within a voltage window of 2.5 4. 0 V (vs. Li/Li + ) at C/10 (C/10 = 17 mA/g) rate 4. 2 Results a nd Discussion 4. 2 1 Crystal S tructure a nd P article M orphology XRD patterns of samples prepared by vibratory wet ball milling process are shown in Fig. 4 2. All LFP samples prepared under different calcining temperature of 600, 650, 700 and 750 o C show ordered orthorhombic olivine ( S.G. (62) Pnma ) crystal structure (JCPDS card No. 81 1173). The calcination temperature significantly influences on the crystallization of LFP from the gradually increased intensity of diffraction peaks along with the increase of temperature For all of the particles there are no peaks corresponding to impurity p hases such as Fe 2 P or Fe 2 O 3 formed by Fe 3+ In order to further study on the crystal structure, Rietveld refinement using Fullprof program was performed as shown in Fig 4 3. The refinements for all samples are well fitted with the original XRD patterns. According to the Rietveld refinement, lattice constants for all samples are tabulated in Table 4 1. The detailed refinement data including structure factor, Bragg position a nd etc. for all samples are well arranged in Appendix. LFP samples in the calcined temperature range of 600 ~ 700 o C look very close to each other. But in the case of LFP calcined at 750 o C, the lattice constants are a bit higher than other samples resulti ng in relative high volume. The crystallite size ( D ) was calculated using Scherrer s equation D=K / cos where D is the crystallite size, K is the shape factor (0.89) ray wavelength and

PAGE 59

59 Bragg angle The calculated mean crystallite sizes for all samples are also tabulated in Table 4 1. Among all of the samples, LFP calcined at 600 o C shows the smallest crystallite size, which is related to short diffusion paths for Li + ions. Fig. 4 4 shows the SEM images (X20000) of LFP calcined at 600~750 o C. The LFP particles show significant agglomeration during the calcination process. The results clearly show that larger average particle sizes are obtained with the increase of calcination temperature which is consistent with the XRD analysis. The benefit for the s mall particle size is an easy penetration of the electrolyte through an active material resulting in a short pathway for Li + diffusion 63 There is a clear difference in terms of particle size between LFP700 and LFP750. The agglomeration and large size particles can be an impediment for Li + ion diffusion in the cath ode material resulting in p oor electrochemical performance 88 Table 4 1 Comparisons of lattice parameters and crystallite size of all samples Samples a () b ( ) c ( ) Vol. ( 3 ) D (nm) LFP600 10.309(5) 5.996(3) 4.682(0) 289.4 33.5 0.9 LFP650 10.308(5) 5.996(6) 4.682(2) 289.4 35.6 0.9 LFP700 10.309(7) 5.995(9) 4.682(1) 289.4 39.3 1.4 LFP750 10.298(3) 5.998(3) 4.677(0) 288.4 41.6 0.9 To get the exact particle size for all samples, particle size distribution using LS13323 Coulter was done in terms of particle size versus differential volume (%) and differential number (%). As confirmed with SEM analysis, the mean particle sizes (MPS) in the range of 80~150 nm for all samples (Fig. 4 5 (b)) were shown to increase with the increase of calcination temperature. The sample LFP750 shows wide particle size distribution and much amount of large sized particles (>10 m) as shown in Fig. 4 5 (a). Particle size data for all samples is listed in Table 4 2.

PAGE 60

60 Fig. 4 6 shows the relationship between mean particl e size (MPS) and specific surface area (SSA) can be well understood SSA for LFP600, LFP650, LFP700 and LFP750 is 13.8 0.4, 13.1 0.2, 11.6 0.3 and 3.8 0.3 m 2 g 1 respectively. With the increase of calcination temperature of LFP particles, the MPS in creased and SSA decreased, especially sharp drop was shown in LFP750 due to the rapid growth of large particle size. This result indicates that larger particle size has smaller surface area which can lead to block the penetration of electrolyte and acted a s an impediment for Li + ion diffusion in the active material. Table 4 2 Particle size data for volume of all samples Samples D 10 D 90 LFP600 1.73 3.03 0.29 6.54 LFP650 1.92 3.27 0.32 7.87 LFP700 3.21 3.61 0.54 18.79 LFP750 5.03 3.54 0.72 19.54 4. 3 2 Electrochemical C haracteristics The initial discharge profiles of all L FP electrode samples synthesized at different temperatures for 10 h with in a voltage window of 2. 5~ 4. 0 V at C /10 rate and rate capabilities at various C rates (0.05 1C) are shown in Fig. 4 7 All cathode samples synthesized at different temperatures show a plateau with continuous slope rather than flat plateau between 3.3 and 3. 0 V Up to 700 o C, LFP cathodes sho w similar voltage profile but LFP750 shows poor discharge profile. It is attributed to the large particle size distribution which was confirmed in SEM and particle size distribution analysis. LFP600 delivers a discharge capacity of 1 24 6 mAhg 1 Among all the LFP samples, LFP650 exhibits the highest discharge capacity of 1 32 .8 mAhg 1 which is about 80 % of the theoretical capacity of LFP The sample s prepared at 7 0 0 C and 750 o C show discharge capacity of only 1 27 .5 mAh g 1 and 99.8 mAhg 1 respectively It is likely that

PAGE 61

61 Li + ion extraction and insertion from/to LFP cathode are significantly influenced by the growth of particles. LFP750 cathode sample showed the lowest discharge capacity among all samples and poor rate performance s This can be attributed to the small space for Li + ion motion in LFP structure as confirmed from structural analysis. With the increase of C rate, all electrodes show a rapid drop in capacity due to the electrode polarization which impedes Li + ion intercalation through LFP electrod es during cycling. Among the LFP electrodes, LFP650 shows the best initial discharge and rate performance. With the increase of cycle number s all electrode samples show a continuous decrease in capacity in Fig. 4 8. The increased polarization of LFP cathode material during charge/discharge reaction also limits the cycle performance even though the C rate is low like C/10. Initially one Li + ion per unit of LiFePO 4 can be extracted from LFP cathode and inserted in to the anode during the charge process, accompanying charge compensation by the oxidation reaction from Fe 2+ to Fe 3+ in the LFP cathode as shown in Fig. 4 9(a). Cyclic voltammetry (CV) test was performed to identify the characteristics of the redox(reduction and oxidation) reactions of LFP650 el ectrode in Li ion cell (Fig. 4 9 (b)) There is o nly a pair of anodic and cathodic peaks ascribed to the Fe 2+ /Fe 3+ redox reaction accompanying Li + ion extraction and insertion from/to the LFP cathode 89 The anodic peak at 3.58V indicates the oxidation o f Fe 2+ to Fe 3+ while the cathodic peak at 3.29V represents the reduction of Fe 3+ to Fe 2+ This redox reaction means typical two phase reaction between FePO 4 and LiFePO 4 like following eq. ( 4 1): LiFePO 4 FePO 4 + Li + + e ( 4 1)

PAGE 62

62 (Forward: charge reaction, Backward: discharge reaction) 4. 3 Summary LFP powders were successfully synthesized via solid state milling of Li 2 CO 3 FeC 2 O 4 2H 2 O and NH 4 H 2 PO 4 followed by calcination process in the range of 600~750 o C. No impurity was detected and all samples prepared at different calcination temperatures indicated LFP with orthorhombic olivine structure. With the increase of calcination temperature, large sized particles were formed, which can be acted as an impedime nt for Li + ion diffusion resulting in poor electrochemical performance. The space for Li + ion motion in LFP structure is also influencing factor on electrochemical performance Moreover, the electrode polarization led a high resistance for Li + ion diffusion between electrode and electrolyte. Therefore, to success in suppressing a formation of an electrode polarization can lead better electrochemical performance. In this research, LFP calcined at 650 o C exhibited the best electrochemical perform ance among other samples. However, bare LFP still has its own weakness in terms of intrinsically low conductivity. In this work, reducing particle size and surface morphology control by coating with conductive agents like carbon, metal and metal oxide for an enhanced electrochemical performance wou ld be introduced from the next C hapter.

PAGE 63

63 Figure 4 1 Schematic diagram of LFP synthesis process

PAGE 64

64 Figure 4 2 XRD patterns of LFP calcined at from 600~750 o C a fter vibratory ball milling

PAGE 65

65 A B

PAGE 66

66 C D Figure 4 3 Rietveld refinement s using Fullprof program of LFP A ) 600 o C B ) 650 o C C ) 700 o C D )750 o C

PAGE 67

67 Figure 4 4 Typical FE SEM images of LFP A ) 600 o C B ) 650 o C C ) 700 o C D ) 750 o C

PAGE 68

68 A B Figure 4 5 Particle size distribution (PSD) A ) N umber percent B ) V olume percent

PAGE 69

69 Figure 4 6 Comparison of mean particle size distribution (MPSD) of volume percent and specific surface area (SSA) according to the calcination temperature

PAGE 70

70 A B Figure 4 7 Electrochemical properties A ) I nitial discharge capacity comparison at C/10 B ) R ate performance comparison of LFP calcined at 600 750 o C

PAGE 71

71 Figure 4 8 Cycle performance comparison between LFP w/o and w/ surfactant at C/10

PAGE 72

72 A B Figure 4 9 Structure transformation during charge and discharge. A ) O livine structure of LiFePO 4 and heterosite structure of FePO 4 84 B ) C ycl ic voltammogram of LFP 650 electrode between 2.0 and 4.5V at the scan rate of 0.05 mVs 1

PAGE 73

73 CHAPTER 5 ELECTROCHEMICAL PERF ORMANCE OF SURFACTAN T PROCESSED LFP AS A CATHODE MATERIAL processed LiFePO 4 as a cathode material for lithium Jungbae Lee, Purushottam Kumar, Gwangwon Lee, Brij M. Moudg il, Rajiv K. Singh.; Ionics, 2012 (http://dx.doi.org/10.1007/s11581 012 0830 9). Reproduced by permission of Springer. Particle size, size distribution and morphology have been shown to greatly influence on the capacity of LFP cathode 90 Typically smaller particle size improves the electrical conductivity and also helps in Lithium ion diffusion 91 Different techniques have been used to control the size and morphology of LFP particles either during synthesis or post synthesis using an additional processing step e.g. ball milling. Milling is one of the most common techniques to reduce the size of particles prepared by different sy nthesis routes. Micro size particle has been reduced to <100 nm using ball milling by controlling the milling time. Solid state reaction, which is widely used for LFP synthesis, uses ball milling to reduce the particle size. However, it is well known that the size distribution of particles obtained after ball milling can be very wide 92 In this work, we have studied the effect of addition of surfactant during ball milling of LFP precursor on LFP parti cle size, morphology, and the resultant effect on electrochemical performance. In LFP synthesis, surfactants have been used by Porcher et al. to effectively disperse carbon black on LFP particles to increase conductivity. 93 Choi et al. used lauric acid to create porous LFP particle for improved high rate cyclic performance. 94 Ferrari et al. used P olyvinylpyrrolidone ( PVP ) a kind of surfactant, as a carbon source and also to control the growth, distribution and orientation of LFP

PAGE 74

74 particles during hydrothermal synthesis. 95 Yang et al. used PVP to obtain dumbbell like LFP particle morpho logy. 96 Though surfactants are known to improve milling efficiency in ceramic processing, it has not been used with LFP precursor material to improve the particle size distribution and morphology. In this study, we observed the addit ion of surfactant reduced the size of LFP particles after ball milling and also led to substantial increase in the surface area, which greatly improved cyclic performance of LFP electrode. Two different LFP samples were prepared by ball milling precursor m aterials in presence and absence of surfactant followed by calcination process. Surfactant used in this study is Avanel S 150 which is an amphiphillic molecule with a negatively charged head and long carbon chain. 5 1 Experimental 5 1 .1 Synthesis o f S urfactant P rocessed A ctive M aterial Solid state reaction method was used to synthesize LFP particles by wet vibratory ball milling stoichiometric amounts of lithium carbonate ( Li 2 CO 3 >98%, Alfa Aesar), iron(II) oxalate ( FeC 2 O 4 2H 2 O, >99%, Alfa Aesar), ammonium di hydrogenophosphate ( NH 4 H 2 PO 4 >98%, ACROS) and 0.1 vol.% of surfactant in 200 ml anhydrous ethanol followed by a two step heat treatment process. Avanel S 150, which is an anionic surfactant was used to modify the milling process. Fig. 5 1 sho ws the chemical structure of Avanel S 150. Surfactant was added to precursors during ball milling. The precursors were milled for 24 h thereafter the mixture was rinsed with ethanol 3 times using centrifuge and filtered to wash out the surfactant from the mixture. Fig. 5 2 shows the schematic of the synthesis process. T he precursor mixture after rinsing was dried at 50 C for 16 h in a drying oven. The dried powders were subjected to a two step heat treatment process. It was first heated to 350 o C for 4 h t o decompose

PAGE 75

75 the carbonate, oxalate, and ammonium mixture of the starting materials followed by heat treatment at 650 o C for 10 h to crystallize LFP. All heat treatments were done under reduced atmosphere formed by continuously flowing 5% H 2 in Ar to preven t the oxidation of Fe from Fe 2+ to Fe 3+ The ramp rate for heating/cooling was 2 K min 1 at each step. Prior to heating, the furnace was purged with 5% H 2 in Ar gas for 20 min. After heat treatment the particles were manually grinded for characterization and electrochemical measurements. 5 1 2 Material C haracterization Phase analysis and crystallinity of the heat treated particle was measured using powder X ray diffractometer (XRD, Philips APD 3720). XRD was done using Cu K radiation source ( = 1.5406 ) with a step size of 0.02 o a scan rate of 0.05 deg/sec and a 2 range of 10 70 o Lattice parameters were determined by Rietveld refinement using Fullprof program. Crystallite size was calculated using the Scherrer s equation with average full width at hal f maximum (FWHM) of (111), (211), (301), (311), (121) peaks. Field emission scanning electron microscopy (FE SEM, JEOL FEG SEM 6335) was used to observe the morphology of the particles. P article size distribution (PSD) was measured using Coulter L ight S cat tering (LS 13320 ) which can measure particle size >0.04 m. The particles were sonicated for 1 h before SEM and particle size measurements. Density was measured 3 times using Pycnometry ( Quantachrome Ultrapyc 1000 Gas Pycnometer ). The specific surface area of the particles was measured 3 times using Brunauer Emmett Teller (BET, Nova 1200) method From the thermal gravimetric analysis (TGA), the carbon amount in surfactant processed LFP sample was approximately 0.4 wt. % as compared with a standard of bare LF P

PAGE 76

76 5 1 3 Electrochemical C haracterization The electrode s were prepared by well mixed coating slurries using a magnetic bar. Coating slurry consists of active material (80 wt%), Acetylene Carbon Black (1 5 wt%) and polyvinylidine fluoride (PVdF, 5 wt%) dissolved in N methyl pyrrolidinone (NMP) using doctor blade on aluminum foil as a current collector. After coating, the electrodes were dried for 4 h at 120 C in low pressure (200 mTorr) atmosphere and pres sed The electrode material (1 mg) was loaded on the disc shape (1 4 mm in diameter 7 m thick ). Coin type test cells (2016) were assembled in an argon filled glove box in which H 2 O level was automatically maintained below 0.1 ppm. Celgard 400 was used as a separator, 1M LiPF 6 in ethylene carbonate (EC)/ dimethyl carbonate (DMC) (1:1 v/v ) as an electrolyte, and Li foil as a counter electrode. Discharge (Li insertion)/ charge (Li extraction) to/from LFP cathode were carried out galvanostatistically using bat tery tester (Arbin Instrument) within a voltage window of 2.5 4. 0 V (vs. Li/Li + ) at C/10 (1C = 170 mA/g) rate The electrochemical i mpedance spectroscopy (EIS) was performed to compare the conductivity at the amplitude of 5 mV and at the frequency from 100 KHz to 0.01 Hz. 5 2 Results a nd Discussion 5 2 .1 Phase A nalysis o f P repared P articles XRD pattern of powder after wet vibratory ball milling along with that of different constituents namely Li 2 CO 3 FeC 2 O 4 and NH 4 H 2 PO 4 is shown in Fig. 5 3. Prior to ball milling, the individual constituents show highly crystalline peaks. XRD pattern of powder after ball milling shows significant increase in FWHM for FeC 2 O 4 and NH 4 H 2 PO 4 whereas surprisingly there is no peak visible for Li 2 CO 3 Li 2 CO 3 appea rs to have completely amorphised after ball milling. Despite significant amorphisation, the powders

PAGE 77

77 of FeC 2 O 4 and NH 4 H 2 PO 4 still have structures of individual component. Amorphisation plays an important role in solid state reaction by increasing the diffus ion and reaction kinetics of various ions. It is reported that NH 4 H 2 PO 4 starts decomposition above 145 o C, whereas FeC 2 O 4 decomposes to FeO between 320 to 450 o C 97 Though Li 2 CO 3 decomposes at high temperature (>800 o C) the amorphisation of Li 2 CO 3 during ball milling, could aid in the solid state reaction. 98 Therefore, a two step heat treatment leads to decomposition and reaction of precursors at 350 o C followed by a high annealing temperature at 6 50 o C for improving the crystallinity. XRD patterns of different LFP powder samples after vibratory ball milling and subsequent heat treatment is shown in Fig. 5 4 The particles were prepared by milling the precursors Li 2 CO 3 FeC 2 O 4 and NH 4 H 2 PO 4 with and without surfactant in anhydrous ethanol followed by drying and heat treatment High temperature annealing led to formation of LiFePO 4 particles with ordered orthorhombic olivine crystal structure (S.G. (62), Pnma JCPDS# 81 1173). Though formatio n of Fe 2 P phase has been reported at annealing temperature above 600 o C by Xu et al., 99 in our work XRD does not show formation of any other contaminant phase Prior to selecting 650 o C annealing temperature, the milled LFP particles were annealed at different temperatures till 750 o C. No Fe 2 P phase was observed till 750 o C annealing under the reduced atmosphere. Annealing temperature of 650 o C was selected to remove the possibility of Fe 2 P phase formation. The crystal lattice constants and crystallite sizes by Rietveld refinement using Fullprof program as shown in Fig. 5 5 are tabulate d in Table 5 1 The matching of the lattice constants indicates absence of any effect of surfactant on the crystal structure, which was expected since the surfactants were washed away after ball milling. The

PAGE 78

78 ray The mean crystallite size of LFP when precursor was milled with surfactant is smaller than that of LFP milled in the absence of surfactant. Table 5 1 Comparisons of lattice p arameters and crystallite size Samples a ( ) b ( ) c ( ) Volume ( 3 ) D (nm) LFP w/o surfactant 10.308(5) 5.996(6) 4.682(2) 289.4 35.62 0.91 LFP w/ surfactant 10.311(6) 5.996(2) 4.682(4) 289.5 31.71 0.95 5 2 2 Particle S ize D istribution a nd S pecific S urface A rea Ball milling is commonly used in solid state reaction synthesis and other methods e.g. hydrothermal etc. to homogenize the precursor and also reduce the size of LFP particles. Typically a wide multimodal distribution of particle size with irregular morphology is obtained after ball milling. 93 Fig 5 6 shows the SEM images of LFP particles obtained after heat treatment when precursor was milled (a) in the absence of surfactants (b) with 0.1 vol .% anionic Avanel S size, irregular shaped particles when milling was done in the absence of surfactant. Similar morphology has been reported for ball milled particles in the literature 100 H owever, the spherical particle morphology with no irregular oversize particle was observed when surfactant was used in the milling mixture. Fig 5 7 shows the differential and cumulative volume weighted particle size distribution of ball milled particles w ith and without surfactants. The multimodal distribution seen in the figure is not uncommon in ball milling of particles. It should be noted that since the particles were formed by annealing at high temperature, there would be severe agglomeration of parti cles. These agglomerates formed by fusion of sub micron particles, as clearly seen in the SEM

PAGE 79

79 image, cannot be broken by sonication. Despite severe agglomeration, t wo effects of surfactant can be seen in the Fig. 5 7 : (i) total removal of oversize particle (>2 um in size) and (ii) increase in the volume fraction of smaller size particles. The oversize particles, which was close to 8 vol.% in LFP is completely converted to smaller size particles when milling was carried out in presence of surfactants (Fig. 5 7 (b)) Fig 5 7 ( a ) shows the differential number fraction of ball milled particles. The number fraction is unimodal and shifted to smaller particle size, which is typical of number weighted distribution curves. Despite relatively smaller in number, larg e size particles of the distribution contribute a much larger volume fraction. The number distribution shows a slightly smaller median size when surfactant was used. The agglomeration of particles due to partial fusion during heat treatment makes the parti cle size distribution measurement misleading. The effect of surfactant on the particle size and morphology can be gauged more accurately by measuring the specific surface area of the particles. The specific surface areas and density of the two different po wder samples are listed in Table 5 2. Table 5 2 Specific surface area (SSA) and density of samples Samples 3 ) SSA (m 2 g 1 ) LFP w/o surfactant 3.42 0.01 13.03 0.24 LFP w/ surfactant 3.12 0.02 23.98 0.27 The LFP sample milled with surfactant shows higher specific surface area of 2 3 .9 8 0. 27 m 2 g 1 with a low density of 3.12 0. 02 gcm 3 compared to 13.03 0.24 m 2 g 1 and 3.42 0.01 gcm 3 for LFP powder processed in the absence of any surfactant. The specific surface area and density measurement were repeated multiple times. Particles prepared by milling precursors with surfactant shows 1.8 times higher surface area. The increase in surface area clearly shows the enhancement in milling efficiency.

PAGE 80

80 LFP processed with surfactant is expected to show a better electrochemical performance than LFP processed without surfactant in terms of electrochemically reacting surface area and diffusion length for Li + ion. Addition of surfactant to inc rease the milling efficiency in a wet milling process has been well known in metallurgical and mineral processing. This phenomenon is known as Rehbinder effect. 101 Though the mechanism is not yet clearly understood, several factors are believed to result in enhanced milling efficiency e.g. lowering of interfacial energies of freshly cleaved surfaces, repulsive/attractive capillary forces induced by adsorption of sur factants, lowering of interparticle friction and dewetting of adsorbed surfactants at edges to generate spherical particles of homogeneous size distribution. 102 5 2 3 Electrochemical C haracteristics In order to study the electrochemical properties of LFP cathode materials prepared by vibratory ball milling with and without surfactant, coin cells were fabricated and subjected to charge/discharge test at C/10 rate as shown in Fig. 5 8 The figure shows the charge/discharge curves for the 1 st and 50 th cycles. The discharge capacity of both electrodes was similar for the 1 st cycle. However, after 50 th cycle the capacity for electrode prepared by LFP processed without surfactant decreased by almost 22%. The charge/discharge plateaus of both electrodes are flat on the 1 st cycle, but at 50 th cycle there is a continuous slope in the discharge curve of electrode prepared by LFP processed without surfactant. The flat charge/discharge plateaus were 3.49/3.36 V for electrode prepared by surfactant processed LFP and 3.54/3.27 V for electrode prepared by LFP processed without surfactant Another stark difference in the charge/discharge curves, is the voltage difference between charge/discharge plateaus. The voltage

PAGE 81

81 difference is much smaller for the electrode prepared by surfactant processed LFP. The voltage difference increases with cyc ling, but proportionally the increase is much higher for the electrode prepared by LFP processed without surfactant. The typical voltage difference between charge/discharge plateau of 1 st and 50 th cycle for electrode prepared by surfactant processed LFP wa s 0.13 V and 0.33 V respectively, whereas it was 0.27 V and 0.59 V for electrode prepared by LFP processed without surfactant. With better size uniformity, the kinetics of the LFP electrode has improved resulting in a lower degree of polarization and bette r electrochemical reversibility. The capacity of both electrodes for 50 charge/discharge cycles at C/10 rate is shown in Fig. 5 9 Electrodes prepared by LFP processed without surfactant showed a maximum discharge capacity of 134.6 mAhg 1 whereas electrodes prepared by surfactant processed LFP showed 132.6 mAhg 1 within the first 5 cycle. It typically takes few cycles for the cells to stabilize because of incomplete wetting of electrodes with electrolyte. With the increase in number of cycl es, the charge/discharge capacities of electrodes prepared by LFP processed without surfactant decreased gradually, while those of electrode prepared by surfactant processed LFP remained relatively stable. After 50 cycle, the discharge capacity for electro de prepared by LFP processed without surfactant reduced from 134.2 mAhg 1 to 104.6 mAhg 1 (78% discharge capacity), whereas electrode prepared by surfactant processed LFP from 131.7 mAhg 1 to 127.3 mAhg 1 (97% discharge capacity). The lower voltage differe nce, relatively flat plateau, and improved cyclic behavior for electrodes prepared by surfactant processed LFP indicate a lower degree of polarization. Flat plateaus and low voltage difference have been typically observed for carbon coated or metal oxide c oated LFP particles 103 104

PAGE 82

82 Typically in these studies carbon or metal oxide contents varied from 1 10 wt. %. The added amount of surfactant during ball milling in our study was only 0.1 vol %, moreover the surfactants was washed away after rinsing. Trace amount of carbon impurity has been shown not to effect on the electrochemical properties. Electrochemical impedance spectra (EIS) measurements were performed to determine the charge transfer resistance for the electrodes. Fig. 5 10 shows EIS data of different LFP electrodes. The ny quist plot for a coin cell typically comprises of semicircle at middle high frequency region and linear at low frequency region. EIS can be understood well based on the equivalent circuit (inset in Fig. 5 10 ) with ohmic resistance (R s ), constant phase elem ent (CPE) which represents a capacitance of double layer (C d ), charge transfer resistance (R ct ), and Warburg impedance (Z w ). Ohmic resistance was very similar for both materials because same electrolyte was used. Charge transfer resistance of electrode pre pared by surfactant processed LFP (22 9 .6 ) was smaller than that of electrode prepared by LFP processed without surfactant (259.2 ) (Table 5 3). This charge transfer resistance originates from the electrochemical interfacial reaction between the electrod e and electrolyte. The degree of reversibility of electrode can be parameterized by calculation of exchange current density ( i o ) using the following: i o = RT/nFR ct (5 1) where R is the gas constant (8.314 Jmol 1 K 1 ), T is the temperature (298.5 K), F is the Faraday s constant (96500 Cmol 1 ) and n is the number of electrons transferred per molecule during intercalation of Li + ion (1 for LiFePO 4 ). The exchange current density of electrode prepared by surfactant processed LFP is higher than that of electrode prepared by LFP processed without surfactant, indicating more reversible

PAGE 83

83 electrochemical reaction. The small sized particles with less agglom erates showed easy throughput of the electrolyte and made Li + ion diffusion path small er in the cathode material 105 The well dispersed and narrow particle size distribution of surfactant processed LFP compared to LFP processed without surfactant also mak es better contact with electrolyte resulting in lower charge transfer resistance. Li + ions can insert and extract to/from electrode prepared by surfactant processed LFP with less difficulty compared to electrode prepared by LFP processed without surfactant Table 5 3 EIS parameters of the samples Samples R s R ct i o 2 ) LFP w/o surfactant 7.7 259.2 6.45 x 10 2 LFP w/ surfactant 7.8 2 29 6 7. 39 x 10 2 The addition of surfactant during vibratory ball milling of precursor materials led to formation of uniform sized smaller LFP particles. It also lead to removal of over size (>2um) particles. Much smaller particle size and relatively narrow size distributi on was observed by SEM and particle size measurement. Enhanced morphological control significantly improved the cyclic performance of the LFP electrode by lowering the degree of polarization during cycling and enhancing the Li + ion diffusion. 5 3 Summary In this study, LFP samples with and without surfactant s were successfully prepared by vibratory milling followed by thermal heat treatment. The addition of surfactant during ball milling of LFP precursor led to uniform sized smaller particles with less agg lomerates and higher specific surface area. The surface area increased by approximately 2 fold upon milling in the presence of surfactant. EIS measurement showed low charge transfer resistance in electrodes formed from these particles, which led to signifi cantly enhanced cyclic performance s Both less agglomeration between

PAGE 84

84 LFP particles and small sized particles led to an easy throughput of electrolyte resulting in low resistance between electrode and electrolyte. After 50 cycles, the discharge capacity rem ained 97% compared to 78% for conventionally prepared LFP electrode.

PAGE 85

85 Figure 5 1 Structural Formula of Avanel S 150 anionic surfactant Figure 5 2 Schematic diagram of surfactant processed LFP synthesis

PAGE 86

86 Figure 5 3 XRD patterns of LFP precursor after vibratory ball milling and before heat treatment

PAGE 87

87 Figure 5 4 XRD pattern comparison of LFP w/ and w/o surfactant prepared by solid state method

PAGE 88

88 A B Figure 5 5 Rietveld refinement using Fullprof program of LFP A ) LFP w / o surfactant. B ) LFP w/ surfactant

PAGE 89

89 A B Figure 5 6 Typical FE SEM images A) LFP w/o surfactant B ) LFP w/ surfactant

PAGE 90

90 A B Figure 5 7 Particle size distribution A ) N umber percent B ) V olume percent

PAGE 91

91 Figure 5 8 Initial and 50 th charge/discharge capacity comparison between LFP w/ and w/o surfactant at C/10

PAGE 92

92 Figure 5 9 Cycle performance comparison between LFP w/o and w/ surfactant at C /10

PAGE 93

93 Figure 5 10 EIS spectra of the LFP prepared by vibratory ball milling w/o and w/ surfactant in the frequency range between 100 kHz and 10 mHz

PAGE 94

94 CHAPTER 6 ELECTROCHEMICAL ENHA NCEMENT OF LFP AS A CATHODE MATERIAL BY INCORPORATING CU NAN O FLAKES This Chapter is part of 4 as a cathode material by incorporating Cu flakes for Lithium ion rechargeable battery" by Jungbae Lee, Purushottam Kumar, Brij M. Moudgil, Rajiv K. Singh.; Solid State Ionics, 2012 (http://dx.doi.org/10.10 16/j.ssi.2012.10.015). Reproduced by permission of Elsevier In order to get a pure phase LFP various synthesis routes have been used e.g. hydrothermal, solid state, sol gel, microwave heating method. 106 109 There have also been numerous efforts for enhancing the capacity by improving the electrical conductivity using carbon coating, metal oxide coating, doping and particle size reduction for short diffusion leng th. 93 103 110 116 Capacity in the range of ~ 1 6 0 mAhg 1 has been regularly reported using these techniques. How ever, these techniques have faced limitations in terms of either capacity, rate performance, cyclic performance or scalability. In this work, Cu incorporated LFP composites were prepared by vibratory wet ball milling of LFP particles synthesized by solid s tate method, high surface area Cu flakes and Polyethylene glycol (PEG). Incorporation of metal directly into LFP has several advantages for enhancing overall electronic conductivity. However, its amount should be small in order to avoid the side reaction w ith electrolytes at high operating voltage. Cu flakes and carbon incorporated LFP particles showed high capacity, excellent rate and cyclic performance. Cui et al. have previously used chemically precipitated CuO to repair incomplete carbon network in LFP particles. 117 The CuO/C co coated LFP composites demonstrated modest enhancement in cyclic performance

PAGE 95

95 and capacity. The existence of CuO was not detected by XRD pattern of LFP particle in their work. In this study, 1 5 wt. % high surface area Cu flakes along with PEG was directly dispersed in LFP particles using ball milling. L FP with Cu flakes and carbon not only improve d the electrical conductivity by lowering interfacial resistance between the electrode and the electrolyte but also prevent the growth of crystallite size resulting in short lithium ion diffusion path. 6 1 Experimental 6 1 .1 Synthesis o f Cu N ano F lake I ncorporated A ctive M aterial Fig. 6 1 shows the schematic of the synthesis method for metal and carbon composite with LFP particles. T wo steps were employed in this work The first step was for synthesis of bare LFP whereas t he LFP/Me tal /Carbon co mposite was prepared through the second step. In the first step, solid state reaction using vibratory wet ball milling was performed with stoichiometric am ount s of lithium carbonate ( Li 2 CO 3 >98%, Alfa Aesar), iron(II) oxalate ( FeC 2 O 4 2H 2 O, >99%, Alfa Aesar), ammonium di hydrogenophosphate ( NH 4 H 2 PO 4 >98%, ACROS) in anhydrous ethanol The precursors were milled for 24 h thereafter the mixture was rinsed with ethanol 3 times using centrifuge and filtered to remove contaminants from the mixture. T he precursor mixture after rinsing was dried at 50 C for 16 h in a drying oven. The dried powders were subjected to a two step heat treatment process. It was first he ated to 350 o C for 4 h to decompose the carbonate, oxalate, and ammonium mixture of the starting materials followed by heat treatment at 650 o C for 10 h to crystallize LFP. All heat treatments were done under reduced atmosphere formed by continuously flowi ng 5% H 2 in Ar to prevent the oxidation of Fe from Fe 2+ to Fe 3+ The ramp rate for heating/cooling was 2 K min 1 at each step. Prior to heating, the furnace was purged with 5% H 2 in Ar gas for

PAGE 96

96 20 min. At the second step, Cu nano flakes incorporated LFP com posite was prepared by using vibratory milling with the mixture of 1 g of LFP, 1 0 wt. % of Polyethylene glycol (PEG, M.W. = 1450 gmol 1 ACROS) and Cu flake s (1, 3, 5 wt. %). T he obtained mixture was washed and filtered again with ethanol followed by dryin g at 50 o C for 16 h. After heating at 600 o C for 2 h under flow of 5% H 2 in Ar gas, the Cu flakes and carbon incorporated LFP composite was obtained. 6 1 2 Material C haracterization LFP particles were characterized by powder X ray diffractometer (XRD, Philips 1730) using Cu K radiation source ( = 1.5406 ) for the crystal structure and the crystallite size Average crystallite size was calculated using the Scherrer s equation with full width at half maximum (FWHM) of (111), (211), (301), (311), (121) peaks. Field e mission s canning e lectron m icroscopy ( FE SEM JEOL FEG SEM 6335) was used to characterize the morphology and dispersion of LFP particles. Density was measured 3 times using Pycnometry ( Quantachrome Ultrapyc 1000 Gas Pycnometer ). The specific surface area was measured using Brunauer Emmett Teller (BET, Nova 1200) method X ray photoelectron spectroscopy (XPS, Kratos Axis spectrometer) using monochromatic Mg K (1253.6 eV) radiation was used to analyze the chemical bonding energi es of the samples. Raman spectroscopy ( Horiba Aramis Micro Raman ) with a laser wavelength of 532 nm was done to determine the I D /I G band intensity ratio of carbon in the LFP particles. Thermal gravimetric analysis ( Mettler Toledo TGA/SDTA) w as performed fo r the thermal analysis and carbon amount estimation The analysis was done in nitrogen inert gas between r oom temperature and 1 0 73 K at a heating ramp of 5 Kmin 1 with a 15 mg.

PAGE 97

97 6 1 3 Electrochemical C haracterization The electrode s were prepared by coating slurries of LFP active material (80 wt %) Acetylene Carbon Black (1 5 wt % ) and polyvinylidine fluoride (PVdF, 5 wt %) dissolved in N methyl pyr r olidi n one (NMP) using a doctor blade on aluminum foil as a current collector. After coating, the electrodes were dried for 4 h at 120 C in low pressure (200 mTorr) atmosphere and pressed The electrode material (1 1.2 mg) was loaded on the disc shape (1 4 mm in diameter 7 m thi ck ). Coin type test cells (2016) were assembled in an argon filled glove box in which H 2 O level was automatically maintained below 0.1 ppm. Celgard 400 (Celgard Inc.) was used as a separator, 1M LiPF 6 in ethylene carbonate (EC)/ dimethyl carbonate (DMC) (1 :1 v/v ) as an electrolyte, and Li foil as a counter electrode. Discharge (Li insertion)/ charge (Li extraction) to/from LFP cathode were carried out galvanostati sti cally using battery tester (Arbin Instrument) within a voltage window of 2.5 4. 0 V (vs. Li/Li + ) at C/10 (C/10 = 17 mAg 1 ) rate The d.c. electrical conductivity measurements using Precision Semiconductor Parameter Analyzer (Agilent, 4156C, USA) were made by a direct V I method on disk shaped sintered pellet samples. The densified pellet samples we re prepared by uni axial (180 MPa) and isostatic pressing (200 MPa) followed by sintering and then gold electrodes were sputtered on both opposing faces of the masked discs of ~7 mm diameter and ~1.6 mm thickness. The electrochemical i mpedance spectroscop y (EIS) was performed to compare the conductivity at the amplitude of 5 mV and at the frequency from 100 KHz to 100 mHz. The potentiostatic intermittent titration technique (PITT) was performed to measure the lithium ion chemical diffusion coefficient by a pplying potential step of 10mV and recording the current as a function of time between 3.4 and 3.6 V.

PAGE 98

98 6 2 Results a nd Discussion 6 2 1 Thermal A nalysis In order to determine the decomposition and the C content in final PEG based composites TGA analys is was performed Fig. 6 2 shows the TGA curves of the PEG, bare LFP and LFP/PEG precursor with 1 0 wt. % PEG obtained after ball milling There w ere three stage s of weight loss process in Fig. 6 2. Stage I below 473 K is due to the loss of H 2 O from the precursor. The second stage II ranged from 473 to 62 0 K is attributed to the decomposition of oxalate and ammonium The third stage III ranged from 62 0 to 673 K indicates the decomposition of PEG, which can be confirmed from the thermal analysis of PEG The weight loss of the precursor for LFP/PEG composite in the third stage was about 1 4 wt. % due to the decomposition of PEG resulting in form ing C, CO, CO 2 118 Finally, the crystallization process of olivine LFP occurred a bove 720 K. 6 2 2 Crystal S tructure a nd P article M orphology XRD patterns of samples prepared with and without Cu flakes are shown in Fig. 6 3 All LFP samples prepared under different conditions show ordered orthorhombic olivine ( S.G. (62) Pnma ) crystal structure (JCPDS card No. 81 1173). For particles with 1~5 wt. % of Cu XRD pattern show s a new phase corresponding to CuO (See Fig. 6 3 (a)) A relatively more intense (311) peak was seen from the XRD pattern of Cu flakes incorporated LFP composite than that of LFP/PEG, which can be attributed to the overlap of the (111) reflection of CuO phase and the (311) reflection of olivine LFP. The FWHM of LFP/PEG/Cu and LFP/PEG was larger than that of bare LFP particles. It has been reported that carbon coating prevents the growth of crystallite size during calcination process resulting in higher FWH M 88 CuO and carbon composite s showed

PAGE 99

99 similar effect on FWHM. The crystallite size ( D ) was calculated using Scherrer s equation D=K / cos where D is the crystallite size, K is the shape factor (0.89) is the X ray wavelength and The calculated mean crystallite sizes for all samples are tabulated in Table 6 1. Fig. 6 4 shows the SEM image of Cu flakes used during ball milling. The Cu flakes had very high aspect ratio (~10 m length, 50 nm thickness). The high aspect ratio provides high surface area for well dispersion of Cu flakes around LFP particles. Li et al. reported that the plane to point contact mode with high contact area is the best for electric contact among three types of point to point, line to point and plane to point contact between LFP particles and coati ng materials 119 The Plane to point contact can be formed between LFP particle s and Cu flakes Thus, the enhanced electric contact may significantly enhance the electrochemical performance of LFP materials owing to the superior ele ctric contact. The Cu flakes after milling and heat treatment transformed to CuO even though calcination process was done under a reduced atmosphere as seen from the CuO peaks in XRD. Table 6 1 Comparisons of crystallite size of all samples Samples D (nm) LFP 35.6 0.9 LFP/PEG 33.2 0.7 LFP/PEG/CuO (1 wt.%) 32.3 0.8 LFP/PEG/CuO (3 wt.%) 31.3 0.8 LFP/PEG/CuO (5 wt.%) 31.7 0.9 SEM images of (a) bare LFP (b) LFP/PEG and (c) Cu flakes and PEG incorporated LFP are shown in Fig. 6 5 The LFP particles without PEG and Cu shows agglomeration during the calcination process. Though LFP particles prepared with PEG and PEG/Cu also show agglomeration during calcinations the extent of agglomeration

PAGE 100

100 observed in these pa rticles over several SEM images appeared less than that in bare LFP particles. Coating of LFP particles, e.g. metal oxide coating, carbon coating has been generally observed to prevent agglomeration by forming a diffusion barrier layer around LFP particles In addition, PEG is also a dispersant used in several earlier works to disperse LFP particles. 104 105 120 The SEM image of Cu flakes incorporated LFP composite after milling and calcin ations does not show Cu flakes, which were homogeneously dispersed during milling. However, the peaks corresponding to Cu was detected through the EDX spectrum (Fig. 6 5 (d)). The specific surface areas of all samples are listed in Table 6 2. Cu flakes incorporated LFP and LFP/PEG showed much higher surface area compared to bare LFP particles. Each sample was measured several times. Cu flakes incorporated LFP composite showed >2 fold increase in surface area. High surface area can be due to smaller size, porous surface or less agglomeration of particles. Table 6 2 Specific surface area(SSA) of samples through BET measurement Samples SSA (m 2 /g) LFP 13.1 0.2 LFP/PEG 22.2 0.3 LFP/PEG/CuO (1 wt.%) 27.5 0.5 LFP/PEG/CuO (3 wt.%) 47.9 0.6 LFP/PEG/CuO (5 wt.%) 42.5 0.4 X ray photoelectron spectroscopy (XPS) measurement was performed to confirm oxidation states of Fe and Cu base d on the binding energy of C 1s (284.5eV). In Fe 2 p spectrum, there are two peaks corresponding to Fe 2p 3/2 (710. 1 eV) and Fe 2p 1 /2 (72 3 3 eV) indicating that all Fe existed in Fe 2+ state as shown in Fig. 6 6 (a) 121 For the Cu 2p spectrum of LFP/PEG/Cu there are four peaks two main peaks corres ponding to Cu 2p 1/2 (933.4 eV) and Cu 2p 3/2 (953.4 eV) respectively and two other satellite peaks as

PAGE 101

101 shown in Fig. 6 6 (b) T he binding energy difference between Cu 2p 1/2 and Cu 2p 3/2 was 20.0 eV which is consistent with the reported data. 122 The r esults of XRD and XPS data suggest that Cu nano flakes were completely converted to CuO after the calcination process. 6 2 3 Carbon S tructural A nalysis (I D /I G ) Fig. 6 7 shows the Raman spectroscopy of the bare LFP, LFP/P EG, and 3wt % Cu flakes incorporated LFP Raman spectroscopy is an important method to investigate the property of dispersed carbon. In the first order Raman spectra, strong and sharp peaks around 1590 and 1345 cm 1 were observed, which was ascribed to the graphite like G band and amorphous carbonaceous D band of residual carbon respectively. Doeff et al. and Hu et al. reported that the both D and G bands can be deconvoluted into four peaks at around 11 90 13 50 15 20 and 15 90 cm 1 for a precise fitting a s shown in the inset s of Fig 6 6. 123 124 F our Gaussian bands are satisfactorily fitted with minimum fitting error. The bands at 119 0 and 15 20 cm 1 are assigned to sp 3 type carbon band which are often observed in highly amorphous carbonaceous materials The sp 2 hybridization similar to that in graphite, contributes to electronic conductivity. The ratio of D/G ( I D / I G ) band intensity is lower for Cu nano flakes incorpora ted LFP composite indicating higher amounts of graphitic carbon which could lead to higher conductivity of LFP electrode. The increase of graphite like G band intensity is attributed to the catalytic effect of Cu nano flakes. T he activation energy barrier for this transformation is lowered by catalytically active metal of Cu nano flakes by diminishing the energy of the system 125

PAGE 102

102 6 2 4 Electrochemical C haracteristics PITT method was used to measure the lithium ion chemical diffusion coefficients based on Fick s law. Fig. 6 8 shows the transient current ( I t ) change with a function of time ( t ) at a potential step from 3.44 to 3.45 V for the 3wt. % Cu flakes incorporated LFP composite electrode. The relation between transient current ( I t ) and time ( t ) at each potential step is represented by following equation: 82 126 127 ( 6 1) where F is the Faraday constant (96500 C), A is the surface area of the electrode, C s and C o are concentration at the surface at time t and t = 0 respectively, and L (cm) is the characteristic length of the electrode material. The slope at the linear region of the l n ( I ) vs. t plot as shown in the inset of Fig. 6 8 is used for calculating t he Li using the following equation: 82 126 127 ( 6 2) The Li values of 3wt. % Cu incorporated LFP composites are in the range of 10 1 5 to 10 1 3 cm 2 s 1 while those of LFP and LFP/PEG are in the range of 10 1 5 to 10 1 4 cm 2 s 1 (Fig. 6 9 ). In the range of 3.44 3.49 V, where LFP shows two phase coexistence region, Li values are lower in all of the samples. Note that the observed difference of Li values (one order) among these samples is not considered significant. 127 The addition of carbon and Cu flakes has little effect on lithium ion diffusion coefficient Fig. 6 10 (a) shows the initial discharge capacity of all samples at C/10 rate. The capacity of LFP/PEG/Cu (3wt. %) is higher than LFP/PEG and bare LFP. LFP /PEG/Cu (3wt. %) showed the capacity of 157.6 mAhg 1 which is 94% of the theoretical capacity of 170 mAhg 1 Typically, capacity in the range of 140 155 mAhg 1 has been regularly

PAGE 103

103 reported for doped LFP, coated LFP and nano sized LFP. 113 128 129 The higher capacities has generally been reported at very slow charge/discharge rate e.g. C/35. 104 Bare LFP material has typically shown capacity in the range of 80 120 mAhg 1 Cui et al. reported th e discharge capacities (142 mAh g 1 at 0.1C, 125 mAh 1 at 1C) of LFP/CuO/Carbon prepared by chemical precipitation method for the CuO (1.7 wt. %) and carbon co coating, which was lower than our experiment data (161.7 mAhg 1 at 0.1C, 149.1 mAhg 1 at 1C) for solid state method synthesized LFP with Cu flakes for CuO formation. 117 The enhanced electrochemical capacity can be attributed to the increase in the sp 2 bonded (G band) carbon which increases the electronic conductivity of cathode. Fig. 6 10 (b) shows both the cyclic and rate performance of all samples. LFP/PEG/Cu (3wt. %) shows excellent cyclic behavior. Even after 50 cycles the drop in capacity was less than 1% of initial capacity. LFP/PEG/Cu (3 wt.%) also show excellent rate performance till 1C charge/discharge rate. Though the capacity reduces to 130 mAhg 1 at 2C rate, the performances of LFP/PEG and LFP are even worse. The capacity of LFP/PEG and LFP drops to 85 mAhg 1 (40% decrease) and almost 0 mAhg 1 respectively. In contrast to bare LFP and LFP/PEG, LFP/PEG/Cu (3wt. %) composite at 2C rate showed less than 25% degradation in capacity. The performance of LFP/PEG/Cu (1wt. %) composite at low charge/discharge rates is similar to that of LFP/PEG/Cu (3wt. %) composite, but at higher rates the performance drops drastically. The LFP/PEG/Cu (5wt. %) composite electr ode performs similar to bare LFP. Over the concentration of 3wt. % of Cu, the lower discharge capacit y of LFP/ PEG /C u (5wt. %) composite electrode can be mainly attributed to the thicker coating layer (herein, Cu

PAGE 104

104 flake), which is more resistive for Li + ion transport rather than the electrochemical enhancement. 130 Apart from the enhancement in capacity of electrodes, it can be seen that the discharge curve of bare LFP has a continuous slope whereas that of other electrodes have flat plateaus. The flat plateaus suggest lower degree of polarization and better electrochemical reversibility. The cyclic performance of the LFP/PEG/Cu (3wt. %) composite shows a good cycling stability. After 50 cycles at C/10 rate, the discharge capacity retention for the bare LFP, LFP/PEG and LFP/PE G/Cu (3wt. %) composites were 104.6, 138.2 and 160.4 mAhg 1 respectively as shown in Fig. 6 11 The LFP/PEG/Cu (3wt. %) composite exhibits a highly stable electrochemical performance in terms of cycling and capacity than those of the LFP and LFP/PEG. The disk shaped sintered pellet samples were prepared for the electrical conductivity measurements by using direct d.c. method for obtaining resistance (R). The 6 3: (6 3) where t is thickness and A is the area of the disk shaped sample. LFP/PEG/Cu (3wt. %) composite showed 4 orders of magnitude higher electronic conductivity of 1.14 x 10 4 Scm 1 than that of bare LFP (4.41 x 10 8 Scm 1 ). LFP/PEG sample showed a conductivity of 5.48 x 10 6 Scm 1. It is clear that the increased electronic conductivity between bare LFP and LFP/PEG samples was attributed to the carbon content and the highest value of electronic conductivity for LFP/PEG/Cu (3wt. %) com posite resulted from CuO/C co coating. Even though the carbon content is similar in both LFP/PEG and

PAGE 105

105 LFP/PEG/Cu (3wt. %) composite, the electronic conductivity difference between two samples is 2 orders of magnitude. The impedance spectra of bare LFP, LFP/PEG, and LFP/PEG/Cu (3wt. %) electrodes at 3 rd cycle are compared in Fig. 6 1 2 All the EIS measurements were carried out at the terminal voltage of 2.5 V, i.e., at the fully discharged state, at 298 K The EIS data can be classified into middle high f requency (Hz) region corresponding to the charge transfer resistance ( ) for Li + ion migration through /at the solid electrolyte interface (SEI) film formed on the surface of the electrode and linear region at low frequency. EIS data can be understood well based on the equivalent circuit (inset in Fig. 11) with ohmic resistance (R s ), constant phase element (CPE) which represents a capacitance of double layer (C d ), charge transfer resistance (R ct ), and Warburg impedance (Z w ). The charge transfer resistance ( 100.7 ) for the LFP/PEG/Cu (3wt. %) was the lowest among those of the bare LFP (259.2 ) and LFP/PEG (225.5 ). The low charge transfer resistance can be attributed to the CuO and carbon co coating on the LFP particles which enables the interfacial resistance to be lower. The increase in graphite like G band intensity, lowering of charge transfer resistance, no significant change in Li ion diffusion rate and drop in capacity for high Cu content (3wt. %) LFP indicates that the primary mechanism for im proved capacity, rate and cyclic performance is increase in the electronic conductivity of cathode by addition of Cu flakes. 6 3 Summary Addition of Cu nano flakes with very high aspect ratio (~10 m length, ~50 nm thickness) during ball milling of LFP was found to be highly effective in improving the capacity, cyclic and rate performance. LFP particles prepared by solid state reaction

PAGE 106

106 method were ball milled with PEG and Cu nano flakes to form LFP cathode materials. 3wt. % Cu incorporated LFP composite sho wed a maximum discharge capacity of 161 mAh g 1 which is 95 % of theoretical capacity. In addition, there was negligible drop in capacity after 50 charge/discharge cycles at C/10 rate whereas at a high rate of 2C the capacity decreased by less than 25 % co mpared to 60 % for electrodes fabricated using LFP/PEG and almost 0 % for the bare LFP electrode. These enhanced properties were attributed to the Cu flakes used as a catalyst for the transformation from amorphous carbon to graph ite like carbon resulting i n low charge transfer resistance in LFP/PEG/Cu composite.

PAGE 107

107 Figure 6 1 Schematic diagram of LFP and Cu incorporated LFP synthesis Figure 6 2 TGA curves of the PEG, bare LFP and LFP/PEG precursor with 1 0 wt. %

PAGE 108

108 A B Figure 6 3 XRD analysis c omparison A ) XRD pattern s of all samples. B ) FWHM of (311) of all samples

PAGE 109

109 Figure 6 4 FE SEM image of Cu nano flake

PAGE 110

110 A B

PAGE 111

111 C D Figure 6 5 FE SEM images (The insets indicate magnified image(x100,000) A ) LFP B ) LFP/PEG C ) LFP/PEG/Cu D ) EDX spectrum of LFP/PEG/Cu composite

PAGE 112

112 A B Figure 6 6 XPS spectra A ) Fe 2p of bare LFP, LFP/PEG and LFP/PEG/Cu B ) Cu 2p of LFP/PEG/Cu

PAGE 113

113 Figure 6 7 Raman spectra of LFP, LFP/PEG, and LFP/PEG/Cu (The insets indicate resolved raman spectra using Gaussian distribution fitting)

PAGE 114

11 4 Figure 6 8 PITT measurement data of LFP/PEG/Cu (3wt. %) between 3.44 and 3.45V (The inset indicates ln ( I t /A) vs. t (s))

PAGE 115

115 Figure 6 9 Chemical diffusion coefficient comparisons of LFP, LFP/PE G, and LFP/PEG/Cu using PITT method between 3.4 and 3.58V

PAGE 116

116 A B Figure 6 10 Electrochemical properties A ) I niti al discharge comparison at C/10 B ) R at e performance of LFP, LFP/PEG, LFP/PEG/Cu comparison at different C rates

PAGE 117

117 Figure 6 11 Cycle performance comparison between bare LFP and LFP/PEG/Cu discharge capacity at C/10

PAGE 118

118 Figure 6 12 Electrochemical Impedance Spectra of bare LFP and LFP/PEG/Cu

PAGE 119

119 CHAPTER 7 ZINC OXIDE INCORPORA TED LFP FOR HIGH RAT E ELECTROCHEMICAL PERFORMANCE 4 for high rate electrochemical performance in lithium ion rechargeable batteries" by Jungbae Lee, Purushottam Kumar, Jinhyung Lee, Brij M. Moudgil, Rajiv K. Singh.; Journal of Alloys and Compounds, 2012 (http://dx.doi.org/10.1016/j.jallcom.2012.10.092). Reproduced by permission of Elsevier Surface modification by various metal oxide coatings on LFP surface with Z rO 2 TiO 2 CuO, Al 2 O 3 and CeO 2 reduced the impedance of LFP cathode and enhanced the electrochemical performance 68 103 117 131 132 Singhal et al. reported ZnO coating on LiMn 1.5 Ni 0 .5 O 4 prepared by sol gel method and annealed at 850 C for 15 h to improve the electrochemical performance of the initial discharge capacity of about 14 6 mAhg 1 and 97% of t he discharge capacity retention after 50 cycles 133 Cui et al. reported enhanced properties of ZnO and carbon co coated LFP particles made by sol gel and freeze drying process. 112 The ZnO/C co coated LFP composites showed high exchange current density ( i o ) which enhanced the electrochemical performance in terms of rate performance and capacity. Though the amount of ZnO in LFP cathode was not detected, the absence of ZnO peak after annealing in the XRD diffraction pattern indicates a lower concentration or poor crystallininity of ZnO. Other work with ZnO coated LFP parti cles showed limited enhancement in electrochemical properties. 134 In this work, LFP/ZnO/Carbon composites were prepared by vibratory wet ball milling of ZnO, Polyethylene glycol (PEG) and LFP particles synthesized by solid state method. The subsequent heat treatment led to formation of ZnO/C co coated LFP particles which showed high capacity, excellent rate and cyclic performance. ZnO (1 4

PAGE 120

120 wt. %) in th is study along with PEG was directly dispersed in LFP particles using ball milling. L FP with ZnO and carbon not only improve d the electrical conductivity by lowering interfacial resistance between the electrode and the electrolyte but also prevent ed the gr owth of crystallite size resulting in short lithium ion diffusion path. 7 1 Experimental 7 1 .1 Synthesis o f Z inc Oxide I ncorporated A ctive M aterial Fig. 7 1 shows the schematic of the synthesis method for metal oxide (MO) and carbon composite with LFP particles. T wo steps were employed in this work The first step was for synthesis of bare LFP whereas t he LFP/MO/C co mposite was prepared through the second st ep. In the first step, solid state reaction using vibratory wet ball milling was performed with stoichiometric amounts of lithium carbonate ( Li 2 CO 3 >98%, Alfa Aesar), iron(II) oxalate ( FeC 2 O 4 2H 2 O, >99%, Alfa Aesar), and ammonium di hydrogenophosphate ( NH 4 H 2 PO 4 >98%, ACROS) in anhydrous ethanol The precursors were milled for 24 h thereafter the mixture was rinsed with ethanol three times using centrifuge and filtered to remove contaminants from the mixture. T he precursor mixture after rinsing was dried at 50 C for 16 h in a drying oven. The dried powders were subjected to a two step heat treatment process. It was first heated to 350 o C for 4 h to decompose the carbonate, oxalate, and ammonium mixture of the starting materials followed by heat treatment at 650 o C for 10 h to crystallize LFP. All heat treatments were done under reduced atmosphere formed by continuously flowing 5% H 2 in Ar to prevent the oxidation of Fe from Fe 2+ to Fe 3+ The ramp rate for heating/cooling was 2 K min 1 at each step. Prior t o heating, the furnace was purged with 5% H 2 in Ar gas for 20 min. At the second step, LFP /ZnO/Carbon composite was prepared by vibratory ball milling a mixture of 1 g of LFP, 1 0 wt. % of Polyethylene

PAGE 121

121 glycol (PEG, M.W. = 1450 gmol 1 ACROS) and 1 4wt. % of <100 nm sized Zinc oxide (ZnO, >98%, Nano Tek) for 24 hours. T he obtained mixture was washed and filtered again with ethanol followed by drying at 50 o C for 16 h. After heating at 600 o C for 2 h under flow of 5% H 2 in Ar gas, the LFP/ZnO/Carbon composite was obtained. 7 1 2 Mater ial C haracterization LFP particles were characterized by powder X ray diffractometer (XRD, PANalytical X pert powder) using Cu K radiation source ( = 1.5406 ) for the crystal structure and the crystallite size Average crystallite size was calculated using the Scherrer s equation with full width at half maximum (FWHM) of (111), (211), (301), (311), (121) peaks. Field e mission s canning e lectron m icroscopy ( FE SEM JEOL FEG SEM 6335) was used to characterize the morphology and dispersion of LFP particles. Transmission electron microscopy (TEM, JEOL 2010F TEM ) was used to examine the nanoscale microstructure of the particles. Density was measured three times using Pycnometry ( Quantachrome Ultrapyc 1000 Gas Pycnometer ). The s pecific surface area was measured using Brunauer Emmett Teller (BET, Nova 1200) method X ray photoelectron spectroscopy (XPS, Kratos Axis spectrometer) using monochromatic Mg K (1253.6 eV) radiation was used to analyze the chemical bonding energies of the samples. Raman spectroscopy ( Horiba Aramis Micro Raman ) with a laser wavelength of 532 nm was done to determine the I D /I G band intensity ratio of carbon in the LFP particles. 7 1 3 Electrochemical C haracterization The electrode s were prepared by coating slurries of active material (80 wt %), Acetylene Carbon Black (1 5 wt .% ) and polyvinylidine fluoride (PVdF, 5 wt %) dissolved in N methyl pyr r olidi n one (NMP) using doctor blade o n aluminum foil as a current

PAGE 122

122 collector. After coating, the electrodes were dried for 4 h at 120 C in low pressure (200 mTorr) atmosphere and pressed The electrode material (1 mg) was loaded on the disc shape (1 4 mm in diameter 7 m thick ). Coin type test cells (2016) were assembled in an argon filled glove box in which H 2 O level was automatically maintained below 0.1 ppm. Celgard 400 (Celgard Inc.) was used as a separator, 1M LiPF 6 in ethylene carbonate (EC)/ dimethyl carbonate (DMC) (1:1 v/v ) as an electrolyte, and Li foil as a counter electrode. Discharge (Li insertion)/ charge (Li extraction) to/from LFP cathode were carried out galvanostatically using battery tester (Arbin Instrument) within a voltage window of 2.5 4. 0 V (vs. Li/Li + ) at various C rates from C/10 to 10C (C/10 = 17 mAg 1 ) rate The electrochemical i mpedance spectroscopy (EIS) was performed to compare the conductivity at the amplitude of 5 mV and at the frequency from 100 KHz to 100 mHz. The potentiostatic intermittent titr ation technique (PITT) was performed to measure the lithium ion chemical diffusion coefficient by applying potential step of 10mV and recording the current as a function of time between 3.4 and 3.6 V. 7 2 Results a nd Discussion 7. 2 .1 Crystal S tructure a nd P article M orphology XRD patterns of samples prepared with and without ZnO is shown in Fig. 7 2. All LFP samples prepared under different conditions show ordered orthorhombic olivine ( S.G. (62) Pnma ) crystal structure (JCPDS card No. 81 1173). For particle s with 1 4 wt. % of ZnO which are the LFP/ PEG/ ZnO composite s ample s there are no peaks corresponding to ZnO and carbon which is attributed to both small amount of ZnO and carbon or existence as an amorphous phase or poor crystalline form The FWHM of (311) of LFP/PEG/ZnO and LFP/PEG was larger than that of bare LFP particles (See Table 7 1). It has been reported that carbon coating prevents the growth of crystallite

PAGE 123

123 size during calcination process resulting in higher FWHM 88 ZnO and carbon composite showed a similar effect on FWH M. The crystallite size ( D ) was calculated using Scherrer s equation D=K / cos where D is the crystallite size, K is the shape factor (0.89) ray wavelength and The calculated mean crystallite sizes for all samples are tabulated in Table 7 1. Among all of the samples, LFP/PEG/ZnO (2wt. %) showed the smallest crystallite size. Table 7 1 Comparisons of FWHM and crystallite size of all samples Samples FWHM of (311) D (nm) LFP 0.21 40.8 0.8 LFP/PEG 0.22 37.8 0.3 LFP/PEG/ZnO (1 wt.%) 0.25 32.7 0.4 LFP/PEG/ZnO (2 wt.%) 0.27 30.4 0.4 LFP/PEG/ZnO (4 wt.%) 0.24 33.9 0.9 Fig. 7 3 shows the SEM image of ZnO nano powder (50~200 nm). Fig. 7 4 shows the SEM images (X40000) of (a) bare LFP (b) LFP/PEG and (c) LFP/PEG/ZnO. The LFP particles without PEG and ZnO show significant agglomeration during the calcination process. Though LFP particles prepared with PEG and PEG/ZnO also show agglomeration during calcinations process, the extent of agglomeration observed in these particles over several SEM images appeared less than that in bare LFP particles. Coating of LFP particles, e.g. metal oxide coating, carbon coating has been generally observed to p revent agglomeration by forming a diffusion barrier layer around LFP particles. In addition, PEG is also a dispersant used in several earlier works to disperse LFP particles and was observed to prohibit LFP particles from agglomeration in our previous work 104 105 120 Fig. 7 4(d) shows the EDX spectrum of LFP/PEG/ZnO particles, showing different elements including Zn. The peak corresponding to Al is ascribed to the Al holder. The specific surface areas of all samples are listed in Table 7 2.

PAGE 124

124 Each sample was measured several times. LFP/PEG/ZnO (2wt. %) composites showed the highest specific surface area of 52.1 0.6 m 2 g 1 among the samples. High surface area can be achieved due to small particle size, porous and rough surface caused by adsorption of nano sized coating material on LFP particles and less agglomeration of particles. SEM images show a less agglomerated and smal l particle size distribution for LFP/PEG/ZnO particles. From the specific surface area measurements, it can be observed that metal oxide and carbon co coating was highly effective for an increase in specific surface area by inhibiting the particle growth a nd agglomeration between LFP particles. Table 7 2 Specific surface area(SSA) of samples through BET measurement Samples SSA (m 2 g 1 ) LFP 13.1 0.2 LFP/PEG 22.2 0.3 LFP/PEG/ZnO (1wt.%) 42.8 0.8 LFP/PEG/ZnO (2wt.%) 52.1 0.6 LFP/PEG/ZnO (4wt.%) 36.2 0.6 In order to clearly confirm the presence of ZnO on the LFP particles, TEM image and EDX mapping of LFP/PEG/ZnO (2wt. %) composite s were done (Fig. 7 5). The EDX mapping shown in Fig. 5(b) for each element is from the rectangular area in TEM image in Fig. 7 5(a). Through the EDX mapping, it can be observed that cabon is homogeneously coated on LFP particles, whereas Zn segregation shows the existence of ZnO as separate particles. With the increase of ZnO contents from 1 wt. % to 4 wt. %, further degree of segregation can be expected X ray photoelectron spectroscopy (XPS) measurement s w ere performed to confirm the oxidation states of Fe and Zn base d on the binding energy of C 1s (284.5eV). In Fe 2 p spectrum, there are two peaks corresponding to Fe 2p 3/2 (710. 1 eV)

PAGE 125

125 and Fe 2p 1 /2 (72 3 3 eV) indicating that all Fe existed in Fe 2+ state as shown in Fig. 7 6(a) 121 For the Zn 2p spectrum of LFP/PEG/ Zn O there are two peaks one peak is corresponding to Zn 2p 1/2 ( 1044.8 eV) and the other peak is to Zn 2p 3/2 ( 1021.7 eV) respectively (Fig. 7 6(b)) T he binding energy difference between Zn 2p 1/2 and Zn 2p 3/2 w as 2 3 1 eV which is consistent with the reported data. 122 The results of XRD and XPS data suggest that there is no change in the valence (+2) of Zn after the calcination process. 7. 2 2 Carbon S tructural A nalysis (I D /I G ) With conductive metal oxides, carbon is also used to increase the electronic conductivity. Herein, the form of carbon and its bonding characteristic is a very important factor for the extent of increase in electronic conductivity. Fig. 7 7 shows the Raman spectroscopy of the bare LFP, LFP/P EG, and LFP/PEG/ Zn O In the first order Raman spectra, strong and sharp peaks around 1590 and 13 50 cm 1 were observed, which was ascribed to the graphite like G band and amorphous carbonaceous D band of residual carbon respectively. Doeff et al. sugge sted that both D and G bands, which are assigned to sp 2 type carbon can be deconvoluted into four peaks for a precise fitting as shown in the insets of Fig. 7. 123 124 Four peaks are satisfactorily fitted with minimum fitting error. The other two peaks at ~1200 and ~1510 cm 1 except both D and G bands are assigned to sp 3 type carbon band. The sp 2 hybridization similar to that in graphite, contributes to electronic condu ctivity. Two characteristic band ratios of I D /I G (Band intensity ratio) and A sp3 /A sp2 (Band area integrated ratio) were evaluated for resolving Raman spectra. The ratios of I D /I G (0.91~0.93) and A sp3 /A sp2 (0.36~0.63) for LFP/PEG/ZnO composites are lower th an those of bare LFP ( I D /I G = 0.98, A sp3 /A sp2 = 0.72) and LFP/PEG ( I D /I G = 0.97, A sp3 /A sp2 = 0.69) as shown in Table 7 3, which

PAGE 126

126 indicates higher amount of graphitic carbon in LFP/PEG/ZnO composites. The sp 2 hybridization similar to that in graphite, contributes to electronic conductivity. The ratio of D/G ( I D / I G = 0.91~0.93) band intensity is lower for LFP/PEG/ZnO com posite indicating higher amount of graphitic carbon which could lead to higher conductivity of LFP electrode. Table 7 3 Raman spectra parameters of samples Samples Peak (cm 1 ) Intensity (a.u.) I D /I G Area (a.u.) A sp3 /A sp2 LFP sp 2 1351 (D) 135 0.98 11716 0.72 1602 (G) 137 5311 sp 3 1220 7869 1530 4421 LFP/PEG sp 2 1343 (D) 83 0.97 7939 0.69 1599 (G) 85 3461 sp 3 1227 5600 1515 2317 LFP/PEG/ZnO (1 wt. %) sp 2 1356 (D) 162 0.92 17410 0.58 1596 (G) 176 8514 sp 3 1244 11717 1500 3508 LFP/PEG/ZnO (2 wt. %) sp 2 1358 (D) 172 0.91 21970 0.36 1598 (G) 188 8682 sp 3 1226 7963 1532 3134 LFP/PEG/ZnO (4 wt. %) sp 2 1353 (D) 80 0.93 7599 0.63 1598 (G) 86 4022 sp 3 1233 5713 1496 1807 In the previous Chapter 6 Cu nano flakes were effectively used as a catalyst to increase the amount of graphitic carbon in carbon from PEG during calcination process and were completely converted to CuO Likewise, herein the increase of graphite like G band intensity is mainly attributed to the catalytic effect of metal oxide (Herein, ZnO) for

PAGE 127

127 the transform ation from disordered amorphous carbon to graphit e like carbon and the use of catalyst is expected to result in increase in electronic conductivity. 7. 2 3 Electrochemical C haracteristics PITT measurements were performed to calculate lithium ion chemical di ffusion coefficients in the range of voltage from 3.40 to 3.58 V with a voltage step of 0.01 V in the first charging cycle of the LFP/PEG/ZnO (2wt. %) composite electrode in Fig. 7 8(a). The PITT profile shows a behavior consistent with the voltage plateau of the first charge cycle. The relation between transient current ( I t ) and time ( t ) at each potential step is represented by following equation based on Fick s law: 82 126 127 ( 7 1) where F is the Faraday constant (96500 C), A is the surface area of the electrode, C s and C o are concentration at the surface at time t and t = 0 respectively, and L (cm) is the characteristic length of the electrode material. The slope at the linear region of the l n ( I ) vs. t p lot as shown in the inset of Fig. 7 8(b) was used for calculating t he Li using the following equation : 82 126 127 ( 7 2) The Li values of LFP/PEG/ZnO (2wt. %) composites are in the range of 10 15 to 10 1 3 cm 2 s 1 while those of LFP and LFP/PEG are in the range of 10 1 5 to 10 1 4 cm 2 s 1 (Fig. 7 9). Note that the observed difference of Li values (one order) among these samples is not considered significant 127 Their kinetic behaviors look very similar specially in the voltage range of 3.46 to 3.51 V, where two phases (FePO 4 and LiFePO 4 ) co exist. Therefore ZnO/C co coating methods seem to have a limited influence on the lithium ion chemical diffusion coefficients in LFP cathodes.

PAGE 128

128 The Initial discharge capacities of all samples are shown in Fig. 7 10 (a) at 0.1C rate. With the increase of ZnO contents up to 4 wt. %, they don t show continuo us increased capacity but the electrochemical performances of electrodes at the 2wt. % of ZnO concentration are the best. The initial discharge capacity of LFP/PEG/ZnO (2wt.%) at C/10 rate is 151.5 mAhg 1 which is the highest among all the samples. The dis charge curve of bare LFP electrodes show a continuous slope, on the other hand LFP/PEG and LFP/PEG/ZnO composite electrodes show very flat plateaus. Both the voltage plateaus of LFP/PEG/ZnO (2wt. %) for Li extraction (charge) and insertion (discharge) reac tion from/to LFP cathode material are the lowest and highest respectively among all the electrodes and are very flat, which means LFP/PEG/ZnO (2wt. %) composite electrodes have the lowest degree of polarization resulting in an excellent electrochemical rev ersibility. Over the concentration of 2wt. % of ZnO, the lower charge/discharge capacities and higher voltage gap between charge/discharge plateau of LFP/PEG/ZnO (4wt. %) composite electrodes can be mainly attributed to the blocking of Li + ion transport th rough thicker coating layer (herein, ZnO), which led to more polarization rather than the electrochemical enhancement from the increased electronic conductivity 130 Fig. 7 10 (b) shows the discharge profiles of LFP/PEG/ZnO (2wt. %) composite electrodes at different C rates ra nging from C/10 to 10C. With the increase of C rates up to 1C, even though there is only a small difference in discharge capacities around 158 mAhg 1 the plateau voltages are lowered continuously, which is attributed to the increased degree of polarizatio n in the electrode. Above 5C discharge rate, the plateau voltages significantly decreased to around 2.9 V reducing the discharge capacities to 145.7 mAhg 1 at 5C and 109.3 mAhg 1 at 10C respectively.

PAGE 129

129 The maximum discharge capacity of LFP/PEG/ZnO (2wt. %) reached to 158.9 mAhg 1 which is 93% of the theoretical capacity of 170 mAhg 1 as shown in Fig. 7 11(a). With the increase of C rate up to 5C, LFP/PEG/ZnO (2wt. %) composite electrodes showed excellent rate performances of 92% capacity retention of the initial maximum capacity and 69% at 10C rate, while LFP and LFP/PEG show poor rate performances. The capacity of LFP/PEG and LFP drops to 85 mAh/g (40% decrease) and almost 0 mAh/g respectively at 2C rate. In contrast to bare LFP and LFP/PEG electr odes, LFP/PEG/ZnO (2wt. %) composite electrodes at 2C rate showed less than 5% degradation in capacity. The performances of LFP/PEG/ZnO (1wt. %) composite electrodes at low C/5 rate are close to those of LFP/PEG/ZnO (2wt. %) composite electrodes, but at hi gher C rates the performance drops drastically. The degradation of electrochemical rate performance of LFP/PEG/ZnO (4wt. %) composite electrodes at high charge/discharge rates are shown, which is attributed to the blocking of Li + ion motion by relatively t hick ZnO coating film compared to LFP/PEG/ZnO (2wt. %) composite. The enhanced electrochemical capacity can be attributed to the increase in the sp 2 bonded (G band) carbon which increases the electronic conductivity of the cathode. The cyclic performance o f the LFP/PEG/ZnO (2wt. %) composite shows a good cycling stability. After 50 cycles at C/10 rate, the discharge capacity retention for the bare LFP, LFP/PEG and LFP/PEG/ZnO (2wt. %) composites were 104.6, 138.2 and 154.68 mAhg 1 respectively as shown in F ig. 7 11(b). The LFP/PEG/ZnO (2wt. %) composite electrodes exhibit a highly stable electrochemical performance in terms of cycling and capacity than those of the LFP and LFP/PEG.

PAGE 130

130 The impedance spectra of bare LFP, LFP/PEG, and LFP/PEG/ZnO (1 4 wt. %) elect rodes are compared in Fig. 7 12 All the EIS measurements were carried out after 3 cycles at the terminal voltage of 2.5 V, i.e., at the fully discharged state, at 298 K The EIS data can be classified into middle high frequency (Hz) region corresponding t o the charge transfer resistance ( ) for Li + ion migration through /at the solid electrolyte interface (SEI) film formed on the surface of the electrode during cycles and linear region at low frequency. EIS data can be understood well based on the equivalent circuit (Left hand side inset in Fig. 7 12) with ohmic resistance (R s ), constant phase element (CPE) which represents a capacitance of double layer (C d ), charge tra nsfer resistance (R ct ), and Warburg impedance (Z w ). LFP/PEG/ZnO (2wt.%) composite (Right hand side inset in Fig. 7 12) exhibited the lowest charge transfer resistance (82.8 ) compared to bare LFP (259.2 ) and LFP/PEG (225.5 ) as shown in Table 7 3 whic h is also confirmed from the shallowest voltage plateau gap between charge / discharge plateaus for the LFP/PEG/ZnO (2wt.%) composite electrodes. The low charge transfer resistance can be attributed to the ZnO and carbon co coating on the LFP particles which enables the interfacial resistance to be lower. The degree of reversibility of electrode can be parameterized by calculation of exchange current density ( i o ) using the equation: ( 7 3) where R is the gas constant (8.314 Jmol 1 K 1 ), T is the temperature (298.5 K), and n is the number of electrons transferred per molecule during intercalation of Li + ion (1 for LiFePO 4 ). The exchange current density of LFP/PEG/ZnO (2wt. %) composite in Table 7 3 is the highest among all samples, w hich led to better electrochemical reversibility.

PAGE 131

131 In addition, the ZnO and carbon co coating on the LFP surface can also provide a protective layer for LFP particles to prevent them from direct contact with the acidic electrolyte. 135 From the results of the increase in graphite like G band intensity, lowering of charge transfer resistance and no significant change in Li ion diffusion rate, the primary mechanism for the improved capacity, rate and cyclic performance is increase in the electronic conductivity of cathode by addition of ZnO nano powders. Table 7 4 EIS parameters of the samples Samples R s R ct i o ( 2 ) LFP 7.7 259.2 6.45 x 10 5 LFP/PEG 5.8 225.5 7.41 x 10 5 LFP/PEG/ZnO (1 wt. %) 6.5 137.4 1.22 x 10 4 LFP/PEG/ZnO (2 wt. %) 6.4 82.8 2.02 x 10 4 LFP/PEG/ZnO (4 wt. %) 7.1 190.5 8.77 x 10 5 7 3 Summar y Addition of ZnO nano powders during ball milling of LFP was found to be highly effective in improving the capacity, cyclic and rate performance. LFP particles prepared by solid state reaction method were ball milled with PEG based ZnO nano powders to form ZnO/C co coated LFP particles. LFP/PEG/ZnO (2 wt. %) composite electrode showed a maximum discharge capacity of 158.9 mAhg 1 which is 93 % of theoretical capacity at C/10 and excellent high rate performances of 145.7 mAhg 1 at 5C and 109.3 mAhg 1 at 10C respectively. In addition, there was negligible drop in capacity for LFP/PEG/ZnO (2wt. %) composite electrode after 50 charge/ discharge cycles at C/10 rate. These enhanced properties were attributed to the ZnO nano coating on the LFP surface used as both a catalyst for the transformation from amorphous carbon to graph ite like carbon resulting in low charge transfer resistance in LFP/PEG/ZnO composite and a protective layer for LFP core particles.

PAGE 132

132 Figure 7 1 S chematic diagram of LFP and ZnO coated LFP synthesis Figure 7 2 XRD patterns of LFP, LFP/PEG and LFP/PEG /ZnO

PAGE 133

133 Figure 7 3 FE SEM image of ZnO

PAGE 134

134 A B

PAGE 135

135 C D Figure 7 4 FE SEM images and EDX spectrum A ) LFP B ) LFP/PEG C ) LFP/PEG/ZnO D) EDX spectrum of LFP/PEG/ZnO composite

PAGE 136

136 A B Figure 7 5 Surface analysis A ) HR TEM image B ) E lemental EDX mapping for a rectangular area of LFP/PEG/ZnO composite.

PAGE 137

137 A B Figure 7 6 XPS spectra A) Fe 2p of bare LFP, LFP/PEG and LFP/PEG/ZnO B ) Zn 2p of LFP/PEG/ZnO

PAGE 138

138 Figure 7 7 Raman spectra of LFP, LFP/PEG, and LFP/PEG/ZnO

PAGE 139

139 A B Figure 7 8 Diffusion coefficient measurements A ) PITT method of LFP/PEG/ZnO (2wt. %) in the first charge cycle test in the voltage range of 3.40 to 3.58V B ) I t vs. t plot between 3.45 and 3.46V. (The inset indicates ln ( I t ) vs. t )

PAGE 140

140 Figure 7 9 Chemical diffusion coefficient comparison s of LFP, LFP/PEG, and LFP/PEG//ZnO using PITT method between 3.4 and 3.58V

PAGE 141

141 A B Figure 7 10 Electrochemical voltage profiles A) Initial charge/discharge capacity comparison of all samples at C/10 B ) V oltage profile of discharge capacity of LFP/PEG/ZnO (2wt. %) at different C rates (0.1~10C)

PAGE 142

142 A B Figure 7 11 Electrochemical test comparison A) R ate performance comparison of all samples at different C rates (0.1~10C) B ) C ycle performance comparison of LFP, LFP/PEG and LFP/PEG/ZnO discharge capacity at C/10

PAGE 143

143 Figure 7 12 Electrochemical Impedance Spectra of bare LFP, LFP/PEG and LFP/PEG/ZnO (The insets indicate an equivalent circuit for left hand si de and LFP/PEG/ZnO (2wt.%) for right hand side)

PAGE 144

144 CHAPTER 8 CONCLUSIONS AND FUTU RE WORK 8.1 Conclusions The electrochemical properties of LFP cathode materials prepared by solid state method have been investigated in the aspects of surfactant processing effect on LFP particle size and size distribution, surface modification by metal and metal oxide coating with carbon and catalytic effect of added metal and metal oxide. From the CV test, it was confirmed that LFP has two phase reaction during charge and d ischarge reactions. For the optimization of calcination temperature of bare LFP, calcination process was performed in the range of 600~750 o C under a reduced atmosphere of 5% H 2 in Ar. With the increase of calcination temperature, both crystallite and particle sizes increased on the contrary, specific surface areas were decreased. There is no change in lattice parameters up to 700 o C, there was a decrease in volume at 750 o C. For the initial discharge capacities, LFP600 LFP650 and LFP700 cathode materials showed similar discharge capacities except LFP750. This can be well explained using results of the structural parameters. LFP750 cathode sample has smaller space for Li + ion motion than other sample electrodes as a res ult of that, LFP750 sample showed the limited electrochemical performance. From the aspect of particle size, LFP750 is the largest sized particles among all samples which can also be a limitation for the Li + ion diffusion. For the optimization of bare LFP, LFP650 has the high crystallinity, proper surface area for the electrochemical reaction between electrode and electrolyte and enough space in the structure for Li + ion motion. However, the severe agglomeration between LFP particles still has been found. T his impedes throughput of the electrolyte to the electrode material. As a result,

PAGE 145

145 interfacial resistance between electrode and electrolyte can increase, which affects the charge transfer of Li + ions. Thus, small amount of anionic surfactant which can also be used as dispersant was employed during wet vibratory ball milling process Surfactant processed LFP particles showed that there is no influence on LFP olivine structure Both crystallite size and particle size decreased and specific surface area increas ed after surfactant processing. Above all, LFP with surfactant processing exhibited less agglomeration than LFP without surfactant processing E ven though the residual carbon is very small amount, it helped the interfacial resistance to decrease. After sur factant processing, t he improved cycle performance at 0.1C rate up to 50 th cycle and the electrochemical reversibility are attributed to the decrease in charge transfer resistance from less agglomeration between LFP particles, which enhanced an easy throug hput of the electrolyte, and residual carbon amount. Based on the results and discussions of surfactant processing, dispersant or surfactant should be used for less agglomeration between particles. Thus, non ionic surfactant and dispersant PEG and metal (Cu nano flakes) were employed together in order to improve conductivity. Cu nano flake s has a high surface area and plane to point contact was formed between LFP particles and Cu nano flakes. Among three types of electric contact modes, plan e to point contact is the best contact mode. Thus, the decrease in charge transfer resistance can be attributed to this contact mode. Moreover, after calcination process PEG was transformed to higher amount of graphitic carbon when Cu nano flakes were pres ent together from the Raman spectroscopy analysis ( I D /I G ). The lower degree of ratio of I D /I G is related to an enhanced conductivity. Thus, Cu nano flakes acted as a catalyst for a carbon structural transformation from

PAGE 146

146 disordered carbon to graphitic carbon By addition of direct metal Cu nano flakes to LFP materials both surface modification s through Cu nano flakes coating on the LFP particle surface and an increased conductivity from catalytic properties greatly improved the initial charge and discharge capacity of LFP/PEG/Cu composite electrode samples, resulting in the excellent electrochemical cycle and high rate performances while bare LFP electrode materials showed poor rate performances and limited charge and discharge capacities. Likewise, metal o xide ( ZnO nano powder ) was used as a catalyst with PEG in the third experimental procedure. Similar principle to metal coating method was applied to improve electrochemical performances. However, using metal oxide instead of metal can reduce the cost of ma terial if the potential as a catalyst is same. The ZnO nano powders with carbon were homogeneously coated on LFP particles. After calcination process, lower degree of I D /I G ratio similar to that of Cu nano flake coating case was obtained. These enhanced pr operties led to high rate performance up to 10 C rate. Interestingly, the behaviors of Li + ion diffusion coefficient measurements using PITT techniques for Cu nano flakes and ZnO nano powders are similar to that of bare LFP in two phase (LFP, FP) co existing region. That is, even though metal or metal oxide coating improved the electrochemical performances of LFP materials, small differences in Li + ion diffusion coefficients indicate that the contribution of ionic conductivi ty on an improved electrical conductivity is little. One order difference in Li + ion diffusion coefficients between bare LFP materials and Metal/C or Metal oxide/C coated LFP was shown, which can be attributed to the decrease in both crystallite size and p article size related to the Li + ion diffusion path The main reason for the improved cycle and rate

PAGE 147

147 performance is attributed to the increase in electronic conductivity from the lower degree of I D /I G ratio and improved electric contact by homogeneous ZnO/C co coating co coating method 8.2 Future Work Electronic conductivity measurement of active materials is required because it is informative even though it is not critical. Thus, disk shaped pellets of all active materials used in this wo rk are required to be manufactured for measuring an electronic conductivity. I n Chapter 5 the effect of surfactant processing during LFP synthesis was investigated. To extend this work, the effect of different concentration of surfactant on LFP electrochemical performance is required to investigate for the completion of the work using surfactant. Likewise, the optimization of concentration of PEG employed as dispersant and surfactant in Chapter 6 and 7 is another work to investigating the effect of PEG concentration on the electrochemical performance. In addition, i t was clear that the catalyst greatly influenced on the I D /I G ratio for high amount of graphitic carbon related to high conductivity and the addition of metal or metal oxide act ed as a catalyst for carbon structural change from disordered to graphitic carbon. Thus, it is required to find different candidate materials which can be used as a proper catalyst of carbon. In addition, designing optimized different synthesis methods based on th e properties of catalyst candidate material s is required to obtain an energy efficient cathode development.

PAGE 148

148 LIST OF REFERENCES 1. U.S. Energy Information Administration, Sep. 2012 Available: http://www.eia.gov/ 2. U.S. Energy Information Administration (Renewables from Executive Summary), Jun. 2012 Available: http://www.eia.gov/forecasts/aeo/source_renewabl e_all.cfm#renewable 3. Hybrid boom to profit battery firms, Aug. 2009 Available: http://koreajoongangdaily.joinsmsn.com/news/article/article.aspx?aid=2908287 4. U. S. Energy Information Administration (Economics of Plug In Hybrid Electric Vehicles), AEO2009 Available: http://www.eia.gov/oiaf/aeo/otheranalysis/aeo_2009analysispap ers/ephev.html 5. A. Padhi, K. Nanjundaswamy, and J. B. Goodenough, Journal of the Electrochemical Society, 1997, 144 1188 1194. 6. D. B. Hibbert, Introduction to electrochemistry MACMILLAN Press Ltd.: Houndmills,Basingstoke, Hampshire, 1993. 7. W. Sc hmickler, and E. Santos, Interfacial electrochemistry Springer: New York, 2010. 8. R. Holze, Experimental electrochemistry: a laboratory textbook Wiley Vch: Weinheim, 2009. 9. J. S. Newman, Electrochemical systems Prentice Hall, Inc.: Englewood Cliffs, New Jersey 1973. 10. G. M. Ehrlich, Handbook of Batteries McGraw Hill: NY and London, 2002. 11. J. Bates, N. Dudney, B. Neudecker, A. Ueda, and C. Evans, Solid State Ionics, 2000, 135 33 45. 12. M. Armand, and J. M. Tarascon, Nature, 2008, 451 652 657. 13. J. Tollefson, Nature, 2008, 456 436 440. 14. K. Ozawa, Solid State Ionics, 1994, 69 212 221. 15. M. Wakihara, and O. Yamamoto, Lithium ion batteries: fundamentals and performance Wiley Vch: 2008. 16. M. S. Whittingham, Chemical Reviews Columbus, 2004, 104 4271 4302. 17. P. B. Balbuena, and Y. Wang, Lithium ion batteries World Scientific: 2004.

PAGE 149

149 18. W. A. Van Schalkwijk, and B. Scrosati, Advances in lithium ion batteries Springer: 2002. 19. M. Thackeray, Journal of th e Electrochemical Society, 1995, 142 2558 2563. 20. M. Wakihara, Materials Science and Engineering: R: Reports, 2001, 33 109 134. 21. P. Arora, R. E. White, and M. Doyle, Journal of the Electrochemical Society, 1998, 145 3647 3667. 22. B. Ammundsen, and J. Paulsen, Advanced Materials, 2001, 13 943 956. 23. P. G. Bruce, B. Scrosati, and J. M. Tarascon, Angewandte Chemie International Edition, 2008, 47 2930 2946. 24. J. M. Tarascon, Philosophical Transactions of the Royal Society A: Ma thematical, Physical and Engineering Sciences, 2010, 368 3227 3241. 25. A. S. Arico, P. Bruce, B. Scrosati, J. M. Tarascon, and W. Van Schalkwijk, Nature materials, 2005, 4 366 377. 26. S. Y. Chung, J. T. Bloking, and Y. M. Chiang, Nature materials, 20 02, 1 123 128. 27. Solid State Ionics, 1989, 36 53 58. 28. Y. Shimakawa, T. Numata, and J. Tabuchi, Journal of Solid State Chemistry, 1997, 131 138 143. 29. C. M. Burba, and R. Frech, Journal of Power Sources, 20 07, 172 870 876. 30. A. Andersson, and J. O. Thomas, Journal of Power Sources, 2001, 97 498 502. 31. A. Yamada, S. C. Chung, and K. Hinokuma, Journal of the Electrochemical Society, 2001, 148 A224 A229. 32. J. M. Tarascon, and M. Armand, Nature, 2001, 414 359 367. 33. C. G. Zoski, Handbook of electrochemistry Elsevier Science: Amsterdam, 2007. 34. H. Huang, S. C. Yin, and L. Nazar, Electrochemical and solid state letters, 2001, 4 A170 A172. 35. P. P. Prosini, M. Carewska, S. Scaccia, P. Wisn iewski, S. Passerini, and M. Pasquali, Journal of the Electrochemical Society, 2002, 149 A886 A890.

PAGE 150

150 36. L. Wang, F. Zhou, Y. Meng, and G. Ceder, Physical Review B, 2007, 76 165435. 37. D. H. Kim, and J. Kim, Electrochemical and solid state letters, 200 6, 9 A439 A442. 38. R. Santoro, and R. Newnham, Acta Crystallographica, 1967, 22 344 347. 39. C. A. J. Fisher, and M. S. Islam, J. Mater. Chem., 2008, 18 1209 1215. 40. L. Laffont, C. Delacourt, P. Gibot, M. Y. Wu, P. Kooyman, C. Masquelier, and J. M. Tarascon, Chemistry of Materials, 2006, 18 5520 5529. 41. D. MacNeil, Z. Lu, and J. Dahn, Journal of the Electrochemical Society, 2002, 149 A1332 A1336. 42. S. LUO, Y. T IAN, H. LI, K. SHI, Z. TANG, and Z. ZHANG, Journal of Rare Earths, 2010, 28 439 442. 43. Y. Cui, M. Wang, R. Guo, and Z. Xu, Rare metals, 2009, 28 127 131. 44. D. Li, Y. Huang, D. Jia, Z. Guo, and S. J. Bao, Journal of Solid State Electrochemistry, 2010, 14 889 895. 45. L. Li, X. Li, Z. Wang, L. Wu, J. Zheng, and H. Guo, Journal of Physics and Chemistry of Solids, 2009, 70 238 242. 46. R. Yang, X. Song, M. Zhao, and F. Wang, Journal of Alloys and Compounds, 2009, 468 365 369. 47. X. Yin, K. Hua ng, S. Liu, H. Wang, and H. Wang, Journal of Power Sources, 2010, 195 4308 4312. 48. Q. Zhang, S. Wang, Z. Zhou, G. Ma, W. Jiang, X. Guo, and S. Zhao, Solid State Ionics, 2011, 191 40 44. 49. J. Ying, M. Lei, C. Jiang, C. Wan, X. He, J. Li, L. Wang, an d J. Ren, Journal of Power Sources, 2006, 158 543 549. 50. C. Y. Ouyang, S. Q. Shi, Z. X. Wang, H. Li, X. J. Huang, and L. Q. Chen, Journal of Physics: Condensed Matter, 2004, 16 2265 2272. 51. Hecht, F. M. Mulder, and L. F. Nazar, Chemistry of Materials, 2008, 20 6313 6315. 52. M. Abbate, S. Lala, L. Montoro, and J. Rosolen, Electrochemical and solid state letters, 2005, 8 A288 A290.

PAGE 151

151 53. M. R. Yang and W. H. Ke, Journal of the Electrochemical Society, 2008, 155 A729 A732. 54. H. Liu, Q. Cao, L. J. Fu, C. Li, Y. P. Wu, and H. Q. Wu, Electrochemistry communications, 2006, 8 1553 1557. 55. D. Wang, H. Li, S. Shi, X. Huang, and L. Chen, Electrochim ica Acta, 2005, 50 2955 2958. 56. H. C. Shin, S. B. Park, H. Jang, K. Y. Chung, W. I. Cho, C. S. Kim, and B. W. Cho, Electrochimica Acta, 2008, 53 7946 7951. 57. C. Li, N. Hua, C. Wang, X. Kang, W. Tuerdi, and Y. Han, Journal of Alloys and Compounds, 2 011, 509 1897 1900. 58. W. S. Yoon, K. Y. Chung, K. W. Nam, J. McBreen, D. Wang, X. Huang, H. Li, L. Chen, and X. Q. Yang, Journal of Power Sources, 2008, 183 427 430. 59. W. Zhang, Y. Hu, X. Tao, H. Huang, Y. Gan, and C. Wang, Journal of Physics and Chemistry of Solids, 2010, 71 1196 1200. 60. Z. R. Chang, H. J. Lv, H. Tang, X. Z. Yuan, and H. Wang, Journal of Alloys and Compounds, 2010, 501 14 17. 61. X. Z. Liao, Y. S. He, Z. F. Ma, X. M. Zhang, and L. Wang, Journal of Powe r Sources, 2007, 174 720 725. 62. L. Yang, L. Jiao, Y. Miao, and H. Yuan, Journal of Solid State Electrochemistry, 2008, 13 1541 1544. 63. Y. Cho, G. Fey, and H. Kao, Journal of Power Sources, 2009, 189 256 262. 64. J. D. Wilcox, M. M. Doeff, M. Marc inek, and R. Kostecki, Journal of the Electrochemical Society, 2007, 154 A389 A395. 65. Electrochemical and solid state letters, 2002, 5 A47. 66. Y. H. Huang, and J. B. Goode nough, Chemistry of Materials, 2008, 20 7237 7241. 67. J. Liu, J. Wang, X. Yan, X. Zhang, G. Yang, A. F. Jalbout, and R. Wang, Electrochimica Acta, 2009, 54 5656 5659. 68. J. T. Son, Journal of the Korean Electrochemical Society, 2010, 13 246 250.

PAGE 152

152 69. H. Liu, G. X. Wang, D. Wexler, J. Z. Wang, and H. K. Liu, Electrochemistry communications, 2008, 10 165 169. 70. J. Yao, F. Wu, X. Qiu, N. Li, and Y. Su, Electrochimica Acta, 2011, 56 5587 5592. 71. D. MacNeil, L. Devigne, C. Michot, I. Rodrigues, G. Liang, and M. Gauthier, Journal of the Electrochemical Society, 2010, 157 A463 A468. 72. C. Delacourt, P. Poizot, S. Levasseur, and C. Masquelier, Electrochemical and solid state letters, 2006, 9 A352 A355. 73. W. J. Zhang, Journal of Power Sources, 2011, 196 2962 2970. 74. X. Yan, G. Yang, J. Liu, Y. Ge, H. Xie, X. Pan, and R. Wang, Electrochimica Acta, 2009, 54 5770 5774. 75. Y. H. Nien, J. R. Carey, and J. S. Chen, Journal of Power Sources, 2009, 193 822 827. 76. J. K. Kim, J. W. Choi, G. S. Chauhan, J. H. Ahn, G. C. Hwang, J. B. Choi, and H. J. Ahn, Electrochimica Acta, 2008, 53 8258 8264. 77. W. Zhou, W. He, Z. Li, H. Zhao, and S. Yan, Journal of Solid State Electrochemistry, 2008, 13 1819 1823. 78. E. M. Jin, B. Jin, D. K. Jun, K. H. Park, H. B. Gu, and K. W. Kim, Journal of Power Sources, 2008, 178 801 806. 79. T. Muraliganth, A. V. Murugan, and A. Manthiram, Journal of Materials Chemistry, 2008, 18 5661 5668 80. M. N. Rahaman, Ceramic processi ng CRC: Boca Raton, 2007. 81. A. J. Bard, and L. R. Faulkner, Electrochemical methods: fundamentals and applications Wiley: New York, 1980. 82. C. J. Wen, B. Boukamp, R. Huggins, and W. Weppner, Journal of the Electrochemical Society, 1979, 126 2258 2 266. 83. A. Churikov, A. Ivanishchev, I. Ivanishcheva, V. Sycheva, N. Khasanova, and E. Antipov, Electrochimica Acta, 2010, 55 2939 2950. 84. G. A. Nazri, and G. Pistoia, Lithium batteries: science and technology Kluwer Academic: Norwell, 2003.

PAGE 153

153 85. V. Palomares, I. Ruiz de Larramendi, J. Alonso, M. Bengoechea, A. Goi, O. Miguel, and T. Rojo, Applied Surface Science, 2010, 256 2563 2568. 86. P. F. Kerr, and C. K. Cabeen, Economic Geology, 1925, 20 729 737. 87. J. Benjamin, and T. Volin, Metallurgical and Materials Transactions B, 1974, 5 1929 1934. 88. Y. Dong, Y. Zhao, Y. Chen, Z. He, and Q. Kuang, Materials chemistry and physics, 2009, 115 245 250. 89. G. Fey, T. Lu, F. Wu, and W. Li, Journal of Solid State Electrochemistry, 2008, 1 2 825 833. 90. G. Fey, Y. Chen, and H. Kao, Journal of Power Sources, 2009, 189 169 178. 91. R. Malik, D. Burch, M. Bazant, and G. Ceder, Nano letters, 2010, 10 4123 4127. 92. C. Li, N. Hua, C. Wang, X. Kang, T. Wumair, and Y. Han, Journal of Solid S tate Electrochemistry, 2011, 15 1971 1976. 93. W. Porcher, B. Lestriez, S. Jouanneau, and D. Guyomard, Journal of Power Sources, 2010, 195 2835 2843. 94. D. Choi, and P. N. Kumta, Journal of Power Sources, 2007, 163 1064 1069. 95. S. Ferrari, R. L. L avall, D. Capsoni, E. Quartarone, A. Magistris, P. Mustarelli, and P. Canton, The Journal of Physical Chemistry C, 2010, 114 12598 12603. 96. H. Yang, X. L. Wu, M. H. Cao, and Y. G. Guo, The Journal of Physical Chemistry C, 2009, 113 3345 3351. 97. K. Rane, A. Nikumbh, and A. Mukhedkar, Journal of Materials Science, 1981, 16 2387 2397. 98. N. Pereira, C. Matthias, K. Bell, F. Badway, I. Plitz, J. Al Sharab, F. Cosandey, P. Shah, N. Isaacs, and G. Amatucci, Journal of the Electrochemical Society, 20 05, 152 A114 A125. 99. Y. Xu, Y. Lu, L. Yan, Z. Yang, and R. Yang, Journal of Power Sources, 2006, 160 570 576. 100. H. C. Kang, D. K. Jun, B. Jin, E. M. Jin, K. H. Park, H. B. Gu, and K. W. Kim, Journal of Power Sources, 2008, 179 340 346. 101. P. Rehbinder, and E. Shchukin, Progress in Surface Science, 1972, 3 97 188.

PAGE 154

154 102. W. Kaczmarek, and B. Ninham, Materials chemistry and physics, 1995, 40 21 29. 103. H. Liu, G. Wang, D. Wexler, J. Wang, and H. Liu, Electrochemistry communications, 2008, 1 0 165 169. 104. Y. Liu, C. Cao, and J. Li, Electrochimica Acta, 2010, 55 3921 3926. 105. A. Fedorkov, A. Nacher Alejos, P. Gmez Kaniansky, Electrochimica Acta, 2010, 55 943 947. 106. S. Yang, P. Y. Zavalij, and M. Stanl ey Whittingham, Electrochemistry communications, 2001, 3 505 508. 107. C. Mi, X. Zhao, G. Cao, and J. Tu, Journal of the Electrochemical Society, 2005, 152 A483 A487. 108. M. Sanchez, G. Brito, M. Fantini, G. Goya, and J. Matos, Solid State Ionics, 2006, 177 497 500. 109. L. Wang, Z. Zhang, and K. Zhang, Journal of Power Sources, 2007, 167 200 205. 110. A. V. Murugan, T. Muraliganth, and A. Manthiram, Electrochemistry communications, 2008, 10 903 906. 111. X. L. Wu, L. Y. Jiang, F. F. Cao, Y. G Guo, and L. J. Wan, Advanced Materials, 2009, 21 2710 2714. 112. Y. Cui, X. Zhao, and R. Guo, Materials Research Bulletin, 2010, 45 844 849. 113. Z. Liu, S. W. Tay, L. Hong, and J. Y. Lee, Journal of Solid State Electrochemistry, 2011, 15 205 209. 114. C. Hu, H. Yi, H. Fang, B. Yang, Y. Yao, W. Ma, and Y. Dai, Int. J. Electrochem. Sci, 2010, 5 1457 1463. 115. M. R. Roberts, G. Vitins, and J. R. Owen, Journal of Power Sources, 2008, 179 754 762. 116. C. L. R. Amin, J. Maier, Physical Chemistry and Chemical Physics, 2008, 10 754 762. 117. Y. Cui, X. Zhao, and R. Guo, Journal of Alloys and Compounds, 2010, 490 236 240.

PAGE 155

155 118. H. Liu, and D. Tang, Solid State Ionics, 2008, 179 1897 1901. 119. X. Li, Y. Zhang, H. Song, K. Du, H. Wang, H. Li, and J. Huang, International Journal of Electrochemical Science, 2012, 7 7111 7120. 120. B. Huang, X. Zheng, X. Fan, G. Song, and M. Lu, Electrochimica Acta, 2011, 56 4865 4868. 121. S. T. Myung, S. Komaba, N. Hirosaki, H. Yashiro, and N Kumagai, Electrochimica Acta, 2004, 49 4213 4222. 122. J. Moulder, W. Stickle, P. Sobol, and K. Bomben, Handbook of X ray Photoelectron Spectroscopy Perkin Elmer: Eden Prairie, MN, 1992. 123. M. M. Doeff, Y. Hu, F. McLarnon, and R. Kostecki, Electroc hemical and solid state letters, 2003, 6 A207 A209. 124. Y. Hu, M. M. Doeff, R. Kostecki, and R. Finones, Journal of the Electrochemical Society, 2004, 151 A1279 A1285. 125. Manzo, C. Pham Huu, and F. Banhart, ACS Nano, 2011, 5 1529 1 534. 126. W. Weppner, and R. A. Huggins, Journal of the Electrochemical Society, 1977, 124 1569 1578. 127. J. Xie, N. Imanishi, T. Zhang, A. Hirano, Y. Takeda, and O. Yamamoto, Electrochimica Acta, 2009, 54 4631 4637. 128. J. Xu, G. Chen, H. J. Li, and Z. S. Lv, Journal of Applied Electrochemistry, 2010, 40 575 580. 129. Y. Wang, B. Sun, J. Park, W. S. Kim, H. S. Kim, and G. Wang, Journal of Alloys and Compounds, 2011, 509 1040 1044. 130. X. Rui, D. Sim, C. Xu, W. Liu, H. Tan, K. Wong, H. H. Hng, T. M. Lim, and Q. Yan, Rsc Advances, 2012, 2 1174 1180. 131. H. H. Chang, C. C. Chang, C. Y. Su, H. C. Wu, M. H. Yang, and N. L. Wu, Journal of Power Sources, 2008, 185 466 472. 132. J. Yao, F. Wu, X. Qiu, N. Li, and Y. Su, Electrochimica Acta, 2011, 56 5587 5592. 133. R. Singhal, M. S. Tomar, J. G. Burgos, and R. S. Katiyar, Journal of Power Sources, 2008, 183 334 338.

PAGE 156

156 134. B. Len, C. P. Vicente, J. Tirado, P. Biensan, and C. Tessier, Journ al of the Electrochemical Society, 2008, 155 A211 A216. 135. Z. Wang, X. Huang, and L. Chen, Journal of the Electrochemical Society, 2003, 150 A199 A208.

PAGE 157

BIOGRAPHICAL SKETCH Jungbae Lee was born in 19 78 in Incheon South Korea He attended the Yonsei University obtaining his b achelor s degree (B.S.) in M etallurgy Engineering in Feb ruary of 20 05 Then, he joined in Batteries and Electrolysis Lab. at Seoul National University as a graduate research student. During his graduate st udies for two years he published two peer reviewed journal articles under advising of Dr. Sohn Hun Joon This led him to be active in Electrochemistry. He got a m aster s degree (M.S.) in Materials Science and Engineering at Seoul National University in Fe b ruary of 2007. For a short time, he worked for Hyundai Steel Company as a research scientist after graduation. However, he had a yearning of studying and researching electrochemistry more deeply, so he decided to transfer to Korea Institute Science and Te chnology (KIST) for gaining precious experience and valuable knowledge on Fuel Cell. Then, he was admitted to Materials Science and Engineering of University of Florida in Aug. of 2008. Since he joined in Dr. Singh s group, he has tried to publish three mo re peer reviewed articles under an advising of Dr. Singh Now he plans to work in the area of research and development after completion of his doctora te (Ph.D). Future is another opportunity. He is about to step up for his objective of becoming a Materials Science E ngineer.