“WW: 2),. i. x. A 4 . ‘4, i. . 451k. n 3% ”Nil. 4.. . (mu. .menu5wm a . « ...‘...:.,....q..u..t. 5 .. . “4‘, [)|?\ 1‘ , I in V: .. at. a... A» rum. t. Au‘ “m”: :1: ..v.\.. ., 3 1 :v.~u...-Aul.é,_: LIBRARY MiChigan State This is to certify that the U nive rsity dissertation entitled MOLECULARLY REINFORCED POLYMERS AND SELF ASSEMBLED NANOCOMPOSITES FOR SECONDARY LITHIUM BATTERIES presented by Fadi H Asfour has been accepted towards fulfillment of the requirements for the PhD. degree in CHEMISTRY ,fla 015% J ’/ /Major Professor’s Signature 2 2. .DéZ’éM/jflr" Zoo 3 Date MSU is an Affinnative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 chIRC/Datoouo.p65-p.15 — MOLECULARLY REINFORCED POLYMERS AND SELF ASSEMBLED NAN OCOMPOSITES FOR SECONDARY LITHIUM BATTERIES By Fadi H. Asfour AN ABSTRACT OF A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 2004 Dr. Gregory L. Baker ABSTRACT MOLECULARLY REINFORCED POLYMERS AND SELF ASSEMBLED NANOCOMPOSITES FOR SECONDARY LITHIUM BATTERIES By Fadi H. Asfour A series of soluble poly(p-phenylene)s (PPP) substituted with short poly(ethylene oxide) (PEO) side chains were synthesized. Characterization data for these polymers indicate that they take on hairy rod-like structures. As the side chain increases in length, the char- acteristics of these materials evolve from those of PPP to resembling those of PEO at long side chains. When the weight fraction of the tethered chains exceeds 80%, the thermal transitions of the polymers and the resulting ionic conductivities nearly match those of polyethylene oxide. Solid polymer electrolytes based on oligo(ethylene oxide)- substituted PPP and lithium perchlorate exhibit conductivities ranging from 10'6 to 5 x 104 S/cm at 30 °C, with the conductivity dependent on the length of the ethylene oxide chain attached to each ring. When the ethylene oxide chain is short, the solubility of Li- C104 is low, leading to undissolved salt and low conductivities. Lengthening the ethyl- ene oxide chains increases the solubility of LiClO4 and chain mobility, causing a more than two order of magnitude increase in the room temperature conductivity. The properties of composites based on modified silica nanoparticles in poly(ethylene ox- ide) (PEO) is also reported. Silica bound lithium sulfonimide salts were prepared through the synthesis of triethoxysilane, N-pentane trifluoromethane sulfonimide, subsequent at- tachment to the surface and formation of the lithium salt. The experimental results of PEO/modified silica composites revealed that an optimum conductivity was attained at a weight loading of 30 wt%. These composites exhibit ionic conductivities that are weakly dependent on temperature and are on the order of 106 S/cm at 30 °C. The lithium ion transport numbers were determined to be 0.86 i 0.03. To Mom and Dad, Ramzi and my loving wife Lucia iv ACKNOWLEDGMENTS I wish to acknowledge my parents, Juliana C. Ramos and Dr. Hani Shafiq Asfur, my brother Ramzi, my wife Lucia and my friends, Pantelis Trikalitis, Theodore Mertzimekis and Christopher Radano. To these people I owe many thanks and debts of gratitude for their constant support and encouragement. To my advisor Prof. Gregory L. Baker who has led me through this process by allowing me to err and learn from my mistakes with patience, along with his friendship, to him my deepest gratitude. To the research group past and present, I owe my thanks to for their words of wisdom and encouragement. I also thank my committee members, Prof. Babak Borhan, Prof. Aaron Odom and Prof. Mitch Smith for their guidance off and on the bas- ketball court. Finally I wish to thank Prof. Jim Dye, Prof. Greg Swain and our collabora- tors at North Carolina State University, Prof. Saacl Khan, Prof. Peter Fedkiw, Dr. Xiangwu Zhang and Dr. Jeff Yerian. Table of Contents List of Figures List of Schemes List of Tables List of Abbreviations 1 Introduction 1.1 Introduction 1.2 General 1.3 Ionic Conductivity 1.4 Ionic Conductivity Measurements — Impedance Spectroscopy 1.5 Electrolytes for Lithium Batteries 1.5.1 Small molecule electrolytes 1.6 Polymer Electrolytes 1.6.1 Poly(ethylene oxide) 1.7 Advances in PEO Based SPE vi xiv XV xvi 14 16 19 22 25 1.7.1 Gel polymer electrolytes 1.7.2 Filler based electrolytes 1.8 Proposed Solid Polymer Electrolyte Systems 1.8.1 1.8.2 1.8.3 Molecularly reinforced electrolytes Syntheses of rigid polymer systems Single ion solid polymer electrolytes 28 30 32 32 35 36 Synthesis of Substituted Poly(p-phenylene)s with Oligo(ethylene oxide) Side Chains 2.1 Introduction 2.2 Results and Discussion 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7 2.2.8 Monomer synthesis Polymerization Polymer characterization Differential scanning calorimetery (DSC) Thermogravimetric analysis (TGA) X-ray diffraction (XRD) Polarized light optical microscopy Conclusions vii 42 42 44 44 45 46 54 56 58 62 64 2.3 Experimental Lithium Ion Conductivity of Substituted Rigid Rod Solid Polymer Electrolytes 3.1 Introduction 3.2 Results and Discussion 3.2.1 Thermal properties 3.2.2 X-ray diffraction 3.2.3 AC impedance spectrscopy 3.3 Conclusions 3.4 Experimental Self-assembled Silica Nana-composite Polymer Electrolytes 4.1 Introduction 4.2 Results and Discussion 4.2.] Synthesis 4.2.2 Infrared spectroscopy 4.2.3 Thermogravimetric analysis 4.2.4 AC impedance spectroscopy 4.2.5 Transference number measurement viii 64 73 73 75 75 79 80 84 85 87 87 89 89 90 93 95 98 4.2.6 Differential scanning calorimetry 4.3 Conclusion 4.4 Future work 4.5 Experimental Appendix A Appendix B Appendix C Bibliography ix 99 101 102 104 109 113 127 139 Figure 1. Figure 2. Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. List of Figures Zn/Cu electrochemical cell. ........................................................................ 3 Ideal versus actual discharge curves for electrochemical cells ................... 3 Comparison of discharge performance of primary and secondary batteries. Primary batteries have a higher capacity and discharge faster than secondary batteries. ..................................................................................... 5 Potential ranges for various lithium compounds used as electrodes for lithium ion batteries. Reproduced by permission of The Electrochemical Society, Inc.2 ............................................................................................... 6 Anatomy of a high energy density lithium battery constructed of a lithiated metal-oxide cathode and a lithium intercalated graphite anode. The two electrodes are separated by electrolyte and an electrode separator. Current is collected at the electrode terminals that reside outside of the battery package. ...................................................................................................... 7 Potential profile in an AC impedance experiment for a resistor (top curve) and a capacitor (bottom curve) ................................................................. 12 The impedance is a vector comprised of two parts. The real component is resistance and the imaginary component is capacitance. 0 is the phase angle between current and potential. ........................................................ 13 Example of a typical Nyquist plot obtained from AC impedance spectroscopy .............................................................................................. l4 Planar zig-zag structure of PEO. ............................................................... 22 3D pictures of the side and top views of the 72 helix of PEO. ................. 23 X Figure 11. Figure 12. Figure 13. Figure 14. Figure 15. Figure 16. Figure 17. Figure 18. Figure 19. Figure 20. Figure 21. Arrhenius plot of the temperature-dependent conductivity of PEO-LiClO4 (OzLi = 6). ................................................................................................. 24 Schematic showing the segment-assisted diffusion of Li+ through the PEO matrix. The circles represent the ether oxygens of PEO. Reproduced with permission from reference 55. .................................................................. 25 Various polymer architectures designed to decrease crystallization of PEO in SPEs. ..................................................................................................... 26 Mesogenic polymers synthesized by Ingram and coworkers. Both 1 and 2 have predominantly poly(ethylene oxide) backbones. Structures 1 and 2 differ in that in 1, pendant mesogenic groups are attached to the chain via flexible alkyl spacers ................................................................................. 33 PPP(EO)X/y reported by Wegner and Meyer in reference 92. ................... 34 Extrapolation of dn/dc of PPP-EOm as a function of PPP weight fraction to obtain dn/dc of PPP-BOO ........................................................................... 50 Comparison of calculated values of dn/dc for PPP-E0m with experimentally obtained data. ................................................................... 50 The extinction coefficient (8) of PPP-E0m as a function of PPP weight fraction ...................................................................................................... 51 Plot of the ratio of molecular weights obtained by GPC to those obtained by LS as a function of the number of ethylene oxide units in the side chains of PPP-EOm polymers .................................................................... 53 DSC scans of PPP-EOm (7 < m > <45>); PEG-dm-2000 is shown for comparison. The data are second heating scans, taken at 10 °C/min after quenching the sample from 180 to ~100 °C. The arrows mark the Tg transitions. ................................................................................................. 55 The relationship between T8 and the weight fraction of PEO in the side chains. The dashed line is a least squares fit to the data. .......................... 56 xi Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. Figure 27. Figure 28. Figure 29. Figure 30. Figure 31. TGA scans of PPP-EOm polymers. Polymer samples were dried under vacuum at 100 °C overnight, equilibrated under N2 at 80 °C for 12 hrs, and then heated under N2 at 10 °/min. ...................................................... 57 Room temperature X-ray powder diffraction scans of PPP-E0m polymers. All scans were collected at a rate of 0.05 20/min over a range of 20 = 1- 50°. ............................................................................................................ 59 Fluorescence spectra of PPP-EOm and their corresponding monomers (inset). Data were collected at room temperature in chloroform using an excitation wavelength of 313 nm. ............................................................. 61 Solid-state Raman spectra of PPP-BOm polymers. Data shown were obtained at room temperature and are the average of 20 scans of 6 seconds in duration. ................................................................................................ 62 Cross polarized optical micrographs of PPP-E02 at 140 °C a) afier 5 minutes at 140 °C b) after shearing the sample and c) 1 minute afier shearing showing break-up of the elongated domains. ............................. 63 The relationship between TE and the length (m) of the ethylene oxide side chain in PPP-EOIm polymers. The data were acquired through MDSC. The line is a guide. ........................................................................................... 76 DSC scans of neat PPP-E0m (dashed) and LiClO4/PPP-E0m composites (solid). The scans shown are the first heating scans after quenching the sample from 180 °C. The samples were heated at 10 °C/min under He... 78 X-ray powder diffraction patterns of PPP-EOm/LiClO4 composites. The data were collected at room temperature at a rate of 0.05 20/min over a range of 20 = 1-50°. .................................................................................. 82 Temperature-dependent conductivities for selected PPP-EOn/LiClO4 composites. The data are derived from AC impedance data, with all samples having an OzLi ratio of 20 ........................................................... 83 A cross-section of the conductivity data at 30 °C, indicating a conductivity maximum at m = <16>. Decreases in o for m = <45> and PEO-dm-ZOOO reflect partial crystallization of the samples. ............................................ 84 xii Figure 32. Figure 33. Figure 34. Figure 35. Figure 36. Figure 37. Figure 38. Diffuse reflectance IR spectra of A200 silica nanoparticles, the sulfonimide—modified silica (A200-C5NHTt), and its lithiated form (A200-C5NTtLi). Data were obtained from samples prepared in a 2:1 ratio of sample to dry KBr ................................................................................. 92 IR spectra comparing the starting material (CSNHTt) and formation of the lithium salt (CSNTfLi). Data were collected as thin films sandwiched between NaCl plates. ................................................................................ 93 Thermogravimetric analysis of A200 and A200-C5NHTf measured in air from 30 °C to 800 °C at a rate of 10 °C/min. The weight loss at 700 °C corresponds to 36 % coverage of the silica nanoparticle surface. ............ 94 Conductivity of PEG-dm-SOO/AZOO-CSNTiLi composites. The data are derived from AC impedance measurements taken after equilibration for 20 min. ........................................................................................................... 97 Cross section of conductivity at 40 °C. The data are obtained from the Arrhenius conductivity plot. ..................................................................... 98 DSC scans of PEG-dm-SOO, 10, 20, and 40 wt% of PEG-dm-SOO/AZOO- CSNTfLi composite polymer electrolytes. Data were collected under He at a rate of 10 °C/min. ............................................................................. 100 Comparison of the heat of fusion for the various composite polymer electrolytes as a function of weight percentage of A200-C5NTfLi. The data were obtained by integrating the areas under the curve for the melting transition for each DSC scan ................................................................... 101 xiii Scheme 1. Scheme 11. Scheme III. Scheme IV. List of Schemes A proposed mechanism34 for the reduction of ethylene carbonate to form passivating layers at the solvent electrode interface. ................................ 18 Synthesis of monomers 4a — 4k ................................................................ 45 Synthesis of A200-C5NTfLi ..................................................................... 90 Proposed synthetic scheme for increasing the concentration of lithium ions per particle. ............................................................................................. 103 xiv Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. List of Tables Some common small molecule hosts for electrolytes and their properties. ................................................................................................................... 17 Structures of potential host polymers for use in solid polymer electrolyte lithium polymer batteries. ......................................................................... 21 Example of SPE polymers and their conductivities .................................. 27 Ionic conductivities of representative gel polymer electrolytes. Taken from reference 70 ............................................................................................... 29 Some examples of composite polymer electrolytes.7O .............................. 31 Examples of early polymer systems with immobilized anions designed to bring about single cation conductivity. ..................................................... 39 Properties of PPP-E0m polymers. a) Molecular weight data were obtained from GPC measurements in CHC13 using a light scattering and a refractive index detector. b) Tg and Tm values were obtained from MDSC and DSC measurements. c) The brackets < > indicate samples with polydisperse EO chains with average degrees of polymerization = m. ................................ 52 Properties of PPP-E0m polymers .............................................................. 77 Lithium concentrations of various composite polymer electrolytes and their measured ionic conductivities at 30 and 80 °C ................................. 97 An example of the saved data. The format is the complex impedance, phase angle, frequency. ........................................................................... 112 XV A200 A200-C5NHTf A200-C5NTfLi AC ATRP bp CPE DC DEA DRIF TS DSC EO fs FTIR GPC 10 133 IR 1 LS m Mn mp Mw MDSC n NMR PDA PDI PEG PEGDME PEG-dme-SOO PEG-dme-ZOOO PEO PPP-EOm PPP q 5 List of Abbreviations Aerosil® 200 triethoxysilyl N-pentane trifluoromethane sulfonimide bound to Aerosil® 200 lithium salt of A200-C5NHTf alternating current atom transfer radical polymerization boiling point composite polymer electrolyte direct current diethylamine diffuse reflectance infrared Fourier transform spectroscopy differential scanning calorimetry ethylene oxide fumed silica Fourier transform infrared gel permeation chromatography initial current steady state current infrared imaginary number light scattering multiplet number-average molecular weight melting point weight—average molecular weight modulated differential scanning calorimetry number of repeat units nuclear magnetic resonance photodiode array detector polydispersity index poly(ethylene glycol) poly(ethylene glycol) dimethyl ether poly(ethylene glycol) dimethyl ether ca. 500 g/mol poly(ethylene glycol) dimethyl ether ca 2000 g/mol poly(ethylene oxide) poly(p-phenylene) with ethylene oxide side chains of length m poly(p-phenylene) quartet singlet xvi SPE g TEA TGA + In UV-vis VI XRD Z, Z” solid polymer electrolyte crystallization temperature glass transition temperature triethylamine thermogravimetric analysis melting temperature triplet lithium ion transport number ultraviolet-visible virtual instrument degree of polymerization X-ray diffraction complex impedance real part of the complex impedance Z imaginary part of the complex impedance Z xvii Chapter 1 1 Introduction 1.1 Introduction The need for high energy density rechargeable lithium batteries for portable electronic devices and for electric vehicle propulsion has driven research on solid polymer electro- lytes. Rechargeable lithium batteries containing polymer electrolytes should have longer lifetimes, higher efficiencies and higher capacities than batteries based on liquid electro- lytes. A polymer based device is expected to be lightweight and flexible, two properties that are attractive in terms of design and manufacture. The purpose of this research is to examine two types of solid polymer electrolytes. The first is a comb-like polymer based on a rigid poly(p-phenylene) backbone substituted with flexible oligo(ethylene oxide) teeth. The second is a single ion conductor prepared from silica nanoparticles in poly(ethylene oxide). 1.2 General A battery is a device that transforms stored chemical energy into electrical energy. The fundamental design of a battery consists of two electrodes inserted in an ion conducting medium, usually a salt or a salt solution. Connecting a load between both electrodes completes the circuit, allowing electrons to flow from one electrode to the other, and con- sequently, chemical reactions at each electrode generate a flow of ions from one electrode to the other through the electrolyte. Electrons are generated at one electrode (anode) 1 through oxidation reactions, and consumed at the other electrode (cathode) through re- duction of species in the electrolyte. As shown in Figure 1 the anions move towards the anode and the cations move in the opposite direction towards the cathode. Their move- ment is dictated by the concentration gradient and electromotive field generated at the beginning of battery use. The driving force for the flow of electrons is the difference in the electrochemical poten- tials of the two redox reactions. Figure 1 is a simple representation of a Zn/Cu battery with a zinc anode and a copper cathode. The net electrochemical reaction is oxidation of Zno to Zn2+ at the anode, and reduction of Cu2+ to Cu0 (plating) at the cathode. The elec- tromotive force of this cell, defined in equation (1.1), is the difference between the stan- dard reduction potentials for the two half-cell reactions. Eccll = Eculhode _ Eanode (1-1) The relevant reduction potentials at 25 °C are: anode Zn2+ + 2e" —> Zn -0.76 V cathode Cu2+ + 26‘ —-) Cu 0.34V and thus Ea,” = 0.34—(—0.76) = 1.10V The free energy for the cell reaction (AG) is de- fined in equation (1.2) where F is F araday’s constant and n is the number of equivalents (electrons) involved in the reaction. AG = —nFEw,, (1.2) e- » LOAD Zn Cu Anode I Cw?“ 28' Cathode 2e- - + I l K (~N02 K -) Zn2+ Salt Bridge 211.90,, CuSO4 Figure 1. Zn/Cu electrochemical cell. II ideal Voltage actual fl Time Figure 2. Ideal versus actual discharge curves for electrochemical cells. When AG in equation (1.2) is negative (discharge conditions) the reaction is spontaneous. If the cell reaction was reversed then EC.” is negative and an external source of energy (charging) is needed to drive the chemical reaction. The theoretical capacity of a cell is the total charge in ampere-hours involved in the elec- trochemical reaction per gram of electrode materials. The theoretical energy density of a battery takes into account both voltage and theoretical capacity and is commonly ex- pressed in Watt-hours per gram (Wh/g) or Watt-hours per liter (Wh/L). The calculation of energy density is shown in equation (1.3). Energy density = V x Ah/g = Wh/g (1.3) In reality, the theoretical energy density is never realized. Figure 2 compares a typical experimental discharge curve1 with the ideal case. In an ideal system, the battery dis- charges at the theoretical voltage until the capacity is fully utilized (all active materials are depleted) at which point the voltage falls to zero. In a real cell, the voltage drops to some stable value at the onset of discharge after which the voltage drops to zero. Dis- charge at less than the theoretical voltage reflects an internal resistance in the cell, usually from polarization of active materials. Depending on the active materials and conditions of use, the practical capacity of a cell ranges from is 25-70% of the theoretical capacity. Batteries are classified as primary or secondary batteries. In a primary battery, the chemical changes that occur during the discharge process are irreversible, while those of secondary batteries are reversible. Thus, primary batteries are disposed of after discharge, while secondary batteries may be recharged. The difference in performance between a primary and a secondary cell can be illustrated as in Figure 3, where a high discharge rate is characteristic of secondary cells and high capacity of primary cells. high capacity primary battery Capacity \* high rate secondary battery ‘ Discharge current Figure 3. Comparison of discharge performance of primary and secondary batteries. Primary batteries have a higher capacity and discharge faster than secondary batteries. LI2_XCOyFezMn4_(y+z)08 x=2 ‘ 5 _ Li1-XC002 - 5 U an204 x=0. 5 I II x=1 4 F Li XV205 - - 4 S L x=1 m . > . . Li1N-Xin01'y02 a LIXTISZ 3 o 3 - electrolyte x=0. 65 7 g stability =1 a window I > electrolyte 2 ' decomposition ‘ 2 ------------------------------------------------------------------------------------------ Figure 4. Potential ranges for various lithium compounds used as electrodes for lith- ium ion batteries. Reproduced by permission of The Electrochemical Society, Inc.2 Lithium batteries currently are available as primary and secondary cells. The reduction potential of lithium (-3.04 V) and its low atomic mass make it an attractive element for high energy density batteries, which are critical for applications in portable electronic de- vices and for electric vehicle propulsion. In a lithium battery, the cathode is typically a metal oxide such as a manganese oxide or cobalt oxide while the anode is either lithium metal or lithium intercalated graphite. Such combinations lead to an electrochemical po- tential of 3 to 4 V. The chemical reactions that take place at the electrodes are oxidation of Li0 at the anode, and intercalation of Li+ into the channels and layers of the metal ox- ide cathode. When lithium intercalated graphite is used as the anode, the anode reaction is de-intercalation of Li+. The coupling of the de-intercalation/intercalation reactions yields what is described as a “rocking chair” cell, where ions are shunted from one anode to cathode. This design avoids the use of metallic lithium, which can lead to catastrophic failure of cells (Figure 5). A critical feature of the cell is the electrolyte, which must physically separate the cathode and the anode and provide a highly conductive medium for transport of lithium cations. Since most electrolytes are liquids, inert polymer screens are ofien inserted between the electrodes. Lithiated metal-oxide Cathode / Separtor + Electrolyte / LixC6 g Anode Electrode i’ Terminals // .5 Figure 5. Anatomy of a high energy density lithium battery constructed of a lithiated metal-oxide cathode and a lithium intercalated graphite anode. The two electrodes are separated by electrolyte and an electrode separator. Current is collected at the electrode terminals that reside outside of the battery package. A major problem that limits applications of lithium batteries is that most electrolytes de- compose at such high electrochemical potentials. New electrolytes are needed that have sufficient mechanically stability to separate the electrodes, possess high ionic conductivi- ties and are compatible with the high electrochemical potentials of lithium batteries. 1.3 Ionic Conductivity The ionic conductivity (0) of a material represents contributions from mobile cations and anions, and as shown in equation (1.4) can be described in terms of the number of charge carriers per unit volume (m), the carrier charge (qi) and the mobility of each ion relative to its average velocity in an applied field of unit strength (in) 0 = Zia-cm.- (1-4) The diffusivity of an ion depends on its size and mass, and thus it is not surprising that a small, light and electropositive metal such as Li+ is desirable for fast ion mobility. In ad- dition to the charge/size ratio, ion mobility is also highly temperature dependent due to electrostatic and dipolar interactions of ions with electrolytes. High ionic conductivity is a result of ions being able to diffuse rapidly through an electrolyte medium. Since is it is difficult for ions to move freely in a crystal lattice or a rigid solid, electro- lytes are rarely used below their melting (Tm) or glass transition (Tg) temperatures where ion mobility is hindered. The temperature dependence of 0 can be described by the V0- gel-Tammann-Fulcher (VTF) equation (1.5). 0'(T) : A T—%e(—Ea/kb(T-To)) (1.5) where the pre-exponential factor A is proportional to the number of charge carriers, E, is the apparent activation energy, kg, is the Boltzmann constant, and To is the temperature at which the conductivity is zero and is usually taken to be the glass transition temperature (Tg). Higher conductivity is attained as the larger the difference between To and the op- erational temperature T, increases. 1.4 Ionic Conductivity Measurements — Impedance Spectroscopy AC impedance spectroscopy is commonly used to measure the electrochemical properties of a material. The experiment involves sandwiching a material between two electrodes, applying a constant voltage bias across the two electrodes, and applying an alternating current as a function of frequency. The data collected are the complex impedance, the frequency, and the phase angle between current and applied bias. Information about dif- fusion, concentration of charged species and transference numbers can be extracted from the data. In addition, the data are commonly used to model electrochemical cells through an equivalent circuit analysis, which is a combination of capacitors and resistors arranged either in series or in parallel. The starting point for impedance spectroscopy is Ohm’s law, (equation (1.6)) which de- scribes the potential (E) as a fiinction of current (I) and resistance (R) E = [R (1.6) As written, equation (1.6) is valid for one circuit element: the ideal resistor. Impedance (Z) is a more general descriptor of resistance for a variety of circuit elements 15:12 (1.7) In AC impedance spectroscopy the potential is varied in a sinusoidal fashion and can be expressed as E = Esin(a)t) (1.8) It follows from Ohm’s law that current is then . E 1_R (1.9) -_ Esin(a)t) I———R (1.10) Where the magnitude of current I = 5;. In terms of impedance (Z) equation (1.10) can be writtenas 12:12 (1.11) E Z=—. 1.12 I ( ) Z=———§S‘“(“”) (1.13) —sin wt R ( ) In the case of a resistor Z = R. The charge of a capacitor (q) is defined as q = CE where C is the capacitance of the capacitor. Since current is the amount of charge passed per unit time . dq =— 1.14 1 dz ( ) Using the same treatment to define impedance for a resistor, the impedance for a capaci- tor is CE' (1.15) Q. II 10 $1:ng dt dt d —CE = CiEsinUut) dt dt 1: CwE cos(wt) and since cos(a)) = sin(a) +125) 1 = CwE sin(a)t + -;£) Using equation (1.12) the impedance is then E sin(a)t) z = CwEsin(a)t +125) 1 Z=-'— ij (1.16) (1.17) (1.18) (1.19) (1.20) (1.21) The complex number (1') takes into account the phase angle between current and potential in the complex plane as shown in the phasor diagrams of Figure 6. Most electrochemical cells can be represented by a combination of resistors and capacitors in which Z has two components, a resistive and a capacitive component. This means that Z is a complex value, the magnitude of which can be expressed as z =JR2+(-jZ)—1C—)2 11 (1.22) 1112 I/-\ n E = 2:: GD 1t/2 V a) V 0 It I 21: \ \ \ (0 \ A ,7 I I K \ \ \ I I \‘_I, 112 [11 x \x I a N Figure 6. Potential profile in an AC impedance experiment for a resistor (top curve) and a capacitor (bottom curve) Qualitatively impedance can be represented as shown in Figure 7. Mathematically, it is commonly expressed as Z = Z'—jZ" (1.23) where Z ' is the real component (Z '= Z cos(t9)) and Z" is the imaginary component (Z " = Z sin(19) ). Data are often plotted on a complex impedance diagram called a Ny- quist plot, where the vector quantity Z at each frequency is represented by a point in the complex plane, as shown in Figure 8. The real bulk resistance of the material (Rb) can be extracted from the Nyquist plot at the specific frequency where the capacitive compo- nent is zero and the applied bias and resulting current are in phase. This corresponds to the point where the plotted curve intercepts the real axis as shown in Figure 8. Conductivity (o) is directly related to Ohm’s law in that it is the reciprocal of resistivity (p) in Qcm as in equation (1.24) a = l/ p (1.24) and p=Rb(A/l) (1.25) where A (cmz) is the area and 1 (cm) is the thickness of the sample. Equation (1.24) be- comes a=l/RbA (1.26) The intercept Rb is then measured directly to calculate the ionic conductivity (6) of the electrolyte sample. -j/o)C ¥ 7 R Figure 7. The impedance is a vector comprised of two parts. The real component is resistance and the imaginary component is capacitance. 0 is the phase angle between cur- rent and potential. 13 8000 r (Fl/AR 6000 _ b ’5 :4 y“ 4000 ~ x N at at *¥** * y 1* # FIE 2‘ 2000 - . , Rb * z? 0 - 1 . 1 a 1 4_ 1 0 2000 4000 6000 8000 Z' (0) Figure 8. Example of a typical Nyquist plot obtained from AC impedance spectros- copy. 1.5 Electrolytes for Lithium Batteries Electrolytes for lithium ion batteries are prepared by dissolving lithium salts such as LiPF6, LiASFfi, LiBF4, LiPF3(C2F5)3, LiN(CF3SOz)2, LiCF3802 and LiClO4 in a polar or- ganic medium.2 While the choice of a particular salt is based on its thermal, chemical and electrochemical stability, a fundamental requirement is that the salt possesses a low lat- tice energy to ensure easy dissociation in solvents. Conversely, the solvent should have a high dielectric constant to overcome the lattice energy of the salt to favor dissolution of the ions. An ideal medium would dissolve a wide range of salts and be chemically, ther- mally and electrochemically stable throughout the voltage window of a lithium battery, typically 0 to 4.5 V vs. L10. The conductivity of an electrolyte at a given temperature de- 14 pends on the ability of the ions to move freely between the electrodes. This depends on ion size, the charge distribution of the anions, and the degree to which ions aggregate. Lithium salts with large charge delocalized anions have been shown to form electrolytes with the highest conductivities. For example lithium imide, LiN(CF3S02)2, with its large sulfonimide anion has a room temperature conductivity in poly(ethylene oxide) (PEO, Mr1 = 2000, an = 27) of 1.8 x 10‘5 S/cm vs. 6.6 x 10'6 S/cm for LiClO4.3 Many studies have been undertaken to understand the mechanism of lithium transport in electrolytes, and the interaction of lithium cations, anions and their ionic aggregates. These studies range from molecular simulations4'8 to experimental NMR,9'12 Ramans’w‘14 and impedance studies.'“8 In PEO, the fraction of dissociated ions has been shown to decrease at low temperature, and at high salt concentrations due to the formation of large immobile aggregate species. Under such conditions, it has been established that anions and aggregates are responsible for a large fraction of current in most electrolytes.l9 The fraction of current carried by a given species is the transference number. Anion and cation transference numbers, t. and t-, can be measured through a number of electro- chemical and diffusion based techniques.”23 Typical experimental values for lithium ion transference numbers are in the range of t. = 0.2-0.4, depending on temperature, solvent, the specific anion, and concentration.” Increasing transference numbers to near unity is important for minimizing polarization at electrode surfaces, which leads to overall im- provement in battery lifetime and efficiency, and not surprisingly, improvement in the lithium ion transference number has been the focus of many research projects. The gen- eral approach to the problem has been to immobilize anions by tethering the anion to a polymer or an inorganic particle. 15 In summary, an ideal electrolyte for secondary lithium batteries should be electrically in- sulating, have a high ionic conductivity (preferably a single ion conductor to minimize polarization effects at electrodes), be thermally stable, electrochemically stable through the working potential range and compatible with other cell components.24 1.5.1 Small molecule electrolytes Electrolytes in commercial lithium ion batteries are typically solutions of lithium salts in low molecular weight organic compounds. Their low viscosity leads to ionic conductivi- ties that range from 10'3 to 10’2 S/cm,25 about an order of magnitude lower than aqueous and alkaline electrolytes (10'2 to 0.1 S/cm), and their electrochemical stability results in good long-term cycling characteristics. Shown in Table 1 are common solvents used in electrolyte formulation and their physical proper- ties. Most electrolytes are binary or ternary combinations design for good conductivity 26’” For example, combining ethylene carbonate with and minimization of crystallinity. dimethyl carbonate in a 1:1 mixture yields a liquid with good dielectric properties at room temperature. 16 b Dielectric Conductivity Structure ($5) (08) Constant o (S/cm), (s @25°C) 1M LiAsF6 ft -3 ethylene carbonate O\_/0 36.4 248 89.78 6.97 x 10 O 0A0 -48.8 240 66.14 5.28 x 10'3 propylene carbonate \ ( ft \ / '2 dimethyl carbonate 0 O 4'6 91 3'12 1'1 x 10 O -43 203 39.0 1.01 x 10'2 y-butyrolactone 0 <07 109 66 775 29 10'2 tetrahydrofuran - ' 1' x Table 1. Some common small molecule hosts for electrolytes and their properties. These solvents are attractive due to their electrochemical stability, which can be as high as 5.1 V vs. Li/Li+ for carbonates and 4.0 V for ethers. They are also capable of rapidly forming stable passivating layers at the solvent electrode interface,29 which are important for the stabilization of the electrochemical cell and its longevity. These layers have been 17 extensively studied and are thought to be made of lithium carbonate and various alkox- ides from the decomposition of electrolyte at the electrode (Scheme I).30'32 However, since liquid electrolytes do not have sufficient mechanical stability to separate electrodes, polypropylene or polyethylene mesh separators must be incorporated into the design of a battery, adding to the cost and decreasing the energy density of the battery. A major dis- advantage of liquid electrolytes are safety concerns such as electrolyte leaking from im- properly sealed cells, pressure build up due to formation of volatile decomposition prod- ucts, and the flammability of organic solvents. Another complication with small molecule electrolytes is their ability to co-intercalate with lithium into the electrodes. Particularly common with propylene carbonate electrolytes, co-intercalation leads to electrode vol- ume changes, cracking, and eventually battery failure.33 Scheme 1. A proposed mechanism34 for the reduction of ethylene carbonate to form passivating layers at the solvent electrode interface. Li” yo Li“ 0 OX0 #1 ._/ -+ .... (‘3' ('3' / ,_jo > (CHZOCOZLi)2 + (32144 0A0 e‘ O .0 —>, - \__/ \ / X LI 0 L12CO3 + C2H4 l8 1.6 Polymer Electrolytes In 1973 Wright and coworkers first reported the ionic conductivity of PEO blended with alkali metal salts.35'37 M. Armand soon followed with his seminal work on the use of PEO as a solid electrolyte system.8 Their work initiated an explosion of research in polymer electrolytes. The advantages of a solid polymer electrolyte (SPE) are derived from their elastomeric strength and flexible mechanical properties. Polymers usually exhibit excellent adhesive properties that allow for almost complete electrode coverage. More importantly, polymers do not intercalate with lithium into the electrodes which enhances cell performance and overall lifetime of the battery. Further advantages are their low volatility, ease of handling and higher viscosities These proper- ties circumvent issues with leaking, pressure build-up due to evolution of gases, the need for an electrode separator and allow for the manufacture of thin films (~50 pm).38 For a polymer to be used as an SPE, the solvation energy of ions within a polymer must exceed the lattice energy of the ionic salt.39 For a polymer to be considered a suitable candidate for use in an SPE, it must meet the following criteria 1) be able to coordinate to ions, 2) have a low cohesive energy and 3) have a high degree of flexibility.”40 The abil- ity for a polymer to coordinate to ions is determined by the number of atoms within the polymer chain or pendant groups which possess available lone pairs such as oxygen or nitrogen. This ability in turn is manifested as a high enough dielectric constant to sepa- rate ions. A low cohesive energy in a polymer indicates a lack of intermolecular interac- tions such as hydrogen bonding. The last requirement is often expressed as the need for a 19 low T3. In line with these requirements, several polymers have been identified based on polyamines and polyethers; among them PEO is the most suitable candidate. It is impor- tant to note that under most circumstances, a high Tg for a polymer that sufficiently sol- vates ionic salts would render it unacceptable for use in solid polymer electrolytes. Listed in Table 2 are some of the important polymers that have been studied. 20 Polymer Structure poly(ethylene oxide) \(OV poly(propylene oxide) fO/ka/ poly(trimethylene oxide) \IO/Vh R‘ {R poly(s1loxanes) /IS"O’I/ '3 poly(phosphazenes) *sz}. R O poly(B-propiolactone) W I O R poly(acrylates) QT ofo/ O poly(ethylene succinate) ,(OWOW O poly(ethylene imine) Indy poly(N—propylaziridine) \E‘IV poly(alkylene sulfides) ISM Table 2. Structures of potential host polymers for use in solid polymer electrolyte lithium polymer batteries. 21 1.6.1 Poly(ethylene oxide) Poly(ethylene oxide) is the most successful host for polymer electrolytes. PEO is synthe- sized by cationic or anionic ring opening polymerization of ethylene oxide, which results in molecular weights ranging from 200 to 5 x 106 g/mol.22‘27““'42 The lower molecular weight oligomers are often termed polyethylene glycols. PEOs with number-average mo- lecular weight (Mn) <600 are viscous liquids, but at higher molecular weights they are waxy solids with a Tg near —65 °C and a Tm of about 65 °C. As shown in Figure 10, pure PEO crystallizes in a 72 helix, with seven ethylene oxide repeat units completing two turns in a fiber period of 19.3 A. A planar zigzag conformation (Figure 9), has also been observed for polymer crystallized under tension.43”“'4 Other structural features have been studied using X-ray analysis and various spectroscopic methods.24‘45'S3 At room tempera- ture, typical samples of PEO are semicrystalline with a degree of crystallinity of ~60%. Figure 9. Planar zig-zag structure of PEO. 22 Figure 10. 3D pictures of the side and top views of the 72 helix of PEO. Typical conductivity data for a PEO—LiClO4 electrolyte are plotted in Figure 11. De- pending on the concentration of dissolved salt, the conductivity usually ranges from 10‘3 to 104 S/cm at 100 °C, dropping to ~10'6 to 10'8 S/cm at room temperature. For a given temperature, a maximum in conductivity is observed at a particular salt concentration. Usually lithium salt concentrations are expressed as a the molar ratio of ether oxygens to lithium ions, and the concentration for maximum conductivity in PEO has been observed to be at OzLi = 20 or approximately 0.4M of L1CF3$O3 or LiClO4. This effect can be ex- plained by considering equation (1.4). At low salt concentrations, the number of charge carriers, n, is small leading to low values for 0'. As the salt concentration increases, the conductivity also increases. Molecular dynamics simulations show that lithium ions 23 complex to four or five ether oxygens of the PEO chain,54 and as a result, lithium cation mobility is related to the segmental motion of the PEO chain as shown in Figure 12. Thus complexation increases the T8 of the electrolyte, and decreases the carrier mobility (,u) leading to the observed maximum in 0'. 115-3 .- T 1 T T V'I'I A 1E-4 l T‘IUIYII' l 155 g I AAul o (S/cm) 1E-6 ffi Y '1 III] 1E—7 F E 2.4 25 26 27 28 29 3.0 31 32 33 34 1000a (K‘) Figure 11. Arrhenius plot of the temperature-dependent conductivity of PEO-LiClO4 (O:Li = 6). 24 Figure 12. Schematic showing the segment-assisted diffusion of Li+ through the PEO matrix. The circles represent the ether oxygens of PEO. Reproduced with permission from reference 55. 1.7 Advances in PEO Based SPE Despite PEO being an excellent solvent for alkali metal ions, SPEs based on pure PEO- metal salt complexes have poor ionic conductivities at ambient temperatures due to the partial crystallization of PEO. There are several well-explored strategies that reduce the crystallinity of PEO and enhance ionic conductivity at room temperatures, including the addition of solvents or plasticizers to high molecular weight PEO, the use of large anions that inhibit crystallization, and structural modification of PEG”58 Figure 13 is cartoon representation of several PEO architectures that have been investigated. 25 block copolymer —-—-———- random copolymer N branched copolymer network polymer .‘ Figure 13. Various polymer architectures designed to decrease crystallization of PEO in SPEs. These polymer architectures typically include PEO segments, which function as the con- ducting medium, linked to a highly flexible segment (very low Tg, < -60 °C) or a rigid segment (high T3). In both cases the added segment must be thermally, chemically and electrochemically stable. For example poly(phosphazene) (Tg = -70 °C)59 and poly(dimethyl siloxane) (T1?, = -123 °C)“) have been used as the backbone in comb poly- mers. Room temperature conductivities of 10'4 S/cm were observed with poly(methoxyethoxyethoxy phosphazene) (MEEP).61 Cross-linking62 or blending63 of MEEP with PEO resulted in a stable free standing material with no change in conductiv- 60.64 ity. Use of poly(methyl disiloxanes) and poly(ethylene)s65 resulted in conductivities 26 of 5 x 104 S/cm but at the expense of mechanical stability. In addition, the polysiloxane bonds are vulnerable to hydrolysis. Examples of other copolymers and their conductivi- ties are shown in Table 3. Copolymer Salt Conductivity Ref I —fsli-o—(CH20H0)4—l O,(CHZCH20)2CH3 +5;=~+ \(CHZCHZO)ZCH3 © © Ro\9 I ll Ro—Pe .P—OR / N \ R0 0R R: (CH2CH20)2CH3 L100: L100; LiSO3CF3 1.5 x10'4 S/cm 8 x 10'5 S/cm 7.5 x 10‘5 S/cm @55 °c Table 3. Example of SPE polymers and their conductivities. 60 66 67 1.7.1 Gel polymer electrolytes The preparation of gel electrolytes is much the same as for polymer electrolytes, except that the polymer is swollen in an organic electrolyte. In these two-component systems, the polymer, usually an insulator, provides the desired mechanical properties while ionic conductivity is governed by the organic electrolyte. Feuillade and Perche first demon- strated the use of an aprotic solution containing alkali metal salts as a polymer plasti- cizer.68 The resulting gels had ionic conductivities rivaling those of liquid electrolytes. Since then, a wide variety of combinations of polymeric hosts and liquid electrolytes have been described that exhibit dimensional stability and liquid-like ionic conductivities as shown in Table 4. An example of a commercialized gel electrolyte is Bellcore’s Plion technology, a copolymer of vinylidene difluoride and hexafluoropropylene (PVdF-HFP) swollen in an ethylene carbonate or dimethyl carbonate/LiPF6 electrolyte.69 Unlike other systems, the PVdF-HFP polymer is insoluble and relatively rigid and tends to form two phase systems with other solvents or polymers. 28 Polymer polymer polymer conductivity system host electrolyte (S/cm), 20°C plasticized . (PEO)3-LiClO4 _3 poly(ethylene ox1de) 10 linear PEO (ECzPC, 20 mol%) cross linked . (PEO)g-LiClO4 4 poly(ethylene ox1de) 8 x 10 PEO (PC, 50 mol%) _ PVdF-LiN(CF3S02)2 3 PVdF poly(vinylidene fluoride) 1.5 x 10' (EC2PC, 75 wt%) 01 eth lene 1 col PEGA/ PEGA p Y( y g y , 10'3 acrylate) (L1ClO4zPO, 1M) PEI poly(ethylene imine) PEI-LiCIO4 10'3 l (p h l PPTA- o - en ene PPTA p y p . y. (PCzECzLiBF4), 2.2 x 10'3 teraphthahmide) 25:25:08 mol%) ethylene glycol dimethacrylate EGDMA- 3 acrylates 2 x 10' (EGDMA) (LiClO4zPC, 1M) PAN PAN poly(acrylonitrile) (EC:PC:LiClO4), 10'3 38-33z21 :8 mol% Table 4. Ionic conductivities of representative gel polymer electrolytes. Taken from reference 70. 29 Other gel polymer electrolytes can be constructed from hosts such as poly(acrylonitrile) (PAN),71'73 poly(vinylidine fluoride) (PVdF),”‘15‘74 poly(methyl methacrylate) (PMMA),75‘77 cellulose ether polymers78 and PEO.9’19‘79 In all of these polymers a small organic molecule is responsible for high ionic conductivity, while the network polymer provides the mechanical support. Although the conductivities of gel electrolytes can ex- ceed 10'3 S/cm there are mechanical stability issues at elevated temperatures due to the large volume fraction of liquid electrolyte, which can be as high as 85%.80 1.7.2 Filler based electrolytes Improvements in the mechanical properties of amorphous polymers can be achieved through the addition of inert inorganic fillers. Weston and Steele described the addition of a-Ale3 to PEO/LiClO4 composites.“ The primary effect in PEO was inhibition of crys- tallization, which led to enhanced room temperature conductivity. Additional inert par- ticulate fillers such as ZrOz, TiOz, hydrophobic fumed silica, and fiber glass have been introduced into polymer electrolytes. The resulting composite polymer electrolytes show similar effects, decreased levels of polymer crystallinity.82 Examples of composite polymer electrolytes are listed in Table 5. 30 Conductivity Composite Polymer electrolyte (S/cm), 20 °C 1 l c m osites (0.56 Li2S — 0.19 8283 — 0.25 LiI) - 10.4 g ass 1” yme‘ 0 1’ ((PEO)6-LiN(CF3S02)2) (18:13 vol.%) gel polymer composites PAN-(PC:ECzLiAsF6)-zeolite 10'2 (833.35.323.13...) (PEO)3-LiBF4-alumina (10 wt%) 10'4 “anocompos‘tes PEGzoo-LiCF3SO3-silica, 20 wt% 1.5 x 10'3 (ceramic composites) Table 5. Some examples of composite polymer electrolytes.7O Raghavan and Khan investigated the rheological properties of hydrophobic fumed silica dispersed in poly(ethylene glycol) dimethyl ethers.83 They found that these composites have weak physical bonds between the silica particles and are shear sensitive. Shearing caused a sharp drop in the viscosity of the composite, which recovered with time. This thixotropic behavior can be an advantage when considering the processing of thin film electrolytes. In another study, Krawiec investigated the relationship between conductivity and the particle size of alumina fillers in PEO/LiBF4 composites.84 The ionic conductivity increased from 10‘5 S/cm for micrometer-sized A1203 to 104 S/cm for the nanometer- sized particles. These results were ascribed to Lewis acid sites on the particles interacting with the BE; anions. Other composites such as glass in PEO6-LiC(CF3S02)2 exhibited good mechanical stability and a conductivity of 104 S/cm at room temperature.85 More dramatic improvements in mechanical stabilities have been realized using polymerizable alkyl acrylate groups onto the surface of fumed silica nanoparticles. These enable forma- 31 tion of a permanent three dimensional network throughout a low molecular weight poly(ethylene oxide) matrix, leading to liquid-like ionic conductivities in dimensionally stable electrolytes.86 1.8 Proposed Solid Polymer Electrolyte Systems 1.8.1 Molecularly reinforced electrolytes Flexible polymers assume a coiled conformation in melts and solutions. For crystalliza- tion to occur from a polymer melt, the coils must disentangle to be included in the crystal lattice. Because the dynamics of polymer chains is typically slower than the crystalliza- tion rate, flexible polymers generally yield semicrystalline materials. In contrast, rod like polymers do not entangle, and instead the steric effects between rigid linear polymers lead to a parallel arrangement in which a transition to an order liquid crystalline system may occur.87 Such a transition may be induced by a flow process which forms macro- scopically oriented solutions or melts with order that is preserved in the solid state. Because of their shape, rigid-rod polymers cannot assume the wide range of conforma- tions characteristic of flexible polymers in melts or solutions. Not surprisingly, most such polymers have high melting points and are poorly soluble, making processing and study of these rigid polymers difficult.87 Consequently, dissolution into a solvent or melting the polymer requires the disruption of intermolecular interactions. However, the melting points and solubility of rod-like polymers can be depressed by one of the follow- ing: 1) Insertion of flexible comonomer units such as an n-alkylene chain that acts as a 32 spacer, 2) Inclusion of bent units of different size which leads to a disruption of the crys- tal structure, or 3) Addition of flexible side chain substituents to the rigid main chain. In the latter case, the side chains act as bound solvent, decreasing interactions between main chains which leads to a large increase in entropy when dissolving or melting the poly- rner.87-89 Solid polymer electrolyte materials of high mechanical strength have been obtained using rigid molecules. Ingram and coworkers reported the use of side chain liquid crystal poly- mer electrolytes based on rigid segments separated by hexa(ethylene oxide) as shown in O (OCHCH2)5‘}‘ 04}— O n O (OCHCH2)5+— I or O n 2 Figure 14.90 MeOHO(CH2)60 Figure 14. Mesogenic polymers synthesized by Ingram and coworkers. Both 1 and 2 have predominantly poly(ethylene oxide) backbones. Structures 1 and 2 differ in that in l, pendant mesogenic groups are attached to the chain via flexible alkyl spacers. 33 anal—- u—‘u‘m .‘-.. . . The T8 of 1 is 5 °C while that of 2 is -36 °C. As expected polymer 1 exhibits liquid crys- talline behavior and undergoes a smectic-isotropic transition at 44 °C while polymer 2 does not exhibit any transitions other than Tg. Upon addition of LiClO4 the Tg of both polymers increases by 20 °C and the clearing temperature of polymer 1 increases to 52 °C. Both polymers have similar conductivities (~10'7 S/cm at 30 °C), however the ionic conductivity of polymer 2 shows a greater temperature dependence. Wegner and Meyer reported SPEs based on rigid polymers shown Figure 14c.91'92 The poly(p-phenylene) main chain is highly rigid due to the 1,4 linkages of the aromatic rings. As shown in Figure 15 each repeating unit consists of a disubstituted aromatic ring and an unsubstituted phenyl ring that acts as a spacer. ”OR, n ODORy m Rx Rx=ICHZCH20)xCH3 25X 2 6 Ry=(CH2CH20)yCH3 25y 2 6 Figure 15. PPP(EO),(/y reported by Wegner and Meyer in reference 92. Through temperature dependant X-ray powder diffraction studies, they demonstrated that these polymers adopt a planar comb structure in which the PPP backbones are arranged in layers separated by the liquid ethylene oxide matrix. The resulting lamellar phases, de- scribed as a smectic B-type structure, remains unaltered to 150 °C, at which point a phase 34 transition leads to a lower order liquid-crystalline phase.55 When cast from solution, the PPP layers orient parallel to the film plane in small local domains. As mentioned previ- ously, the side chains act as bound solvent, and so it is not surprising that these polymers have low Tgs (~ -50 °C) which decrease as the side chains increase in length. The result- ing ionic conductivities are on the order or 104 S/cm at room temperature. 1.8.2 Syntheses of rigid polymer systems Controlled synthetic methods are necessary to produce ideal rigid polymer structures. Since any defects or side reactions can lead to a dramatic decrease in the desired proper- ties, the methods employed must be regiospecific and lead to stable products. As shown earlier, flexible links in a rigid rod backbone can dramatically alter the polymer charac- teristics. The solubilities of most rigid polymers decrease as the degree of polymerization increases. Traditionally, poor solubility has made it difficult to synthesize poly(p- phenylenes) (PPP). For PPP, the synthetic route must employ methods that allow for ex- clusive formation of the para aromatic linkages and high molecular weights. One synthetic route to these polymers utilized an electrochemical coupling of aromatic monomers to afford polyphenylenes on inert electrodes such as platinum. ”'94 The oxida- tive coupling and reduction of aryl halide monomers affords small quantities of polymer that are structurally irregular, containing a substantial mixture of ortho, meta and para aromatic linkages. While many catalytic methods such as Friedel-Crafls synthesis and polymerization of p-dibromobenzene with lithium quickly yield polyphenylenes, these materials are fraught with defects in the aromatic linkages affording nonlinear polymers. 35 Improvements were obtained through the use of the Ullmann homocoupling of 1,4- diiodobenzene using activated copper powder affording the correct 1,4 substitution pat- tern throughout the polymer chain. 95'97 However, the severe conditions for this reaction limited its scope of application to monomers that contained functional groups that can tolerate the prolonged reaction temperatures, > 200 °C. Milder conditions and greater chemoselectivity have been realized through the use of transition metal catalysts. 98404 These methods afford PPP in high yields and the para linkage with a high degree of fidel- ity. Specifically, palladium and nickel catalyzed aryl coupling reactions are tolerant of a wide variety of functional groups which allows for the synthesis of a wide range of sub- stituted PPPs. The catalytic polymerization involves three major steps: oxidative addition of the arylhalide onto the active metal center, a transmetallation step to add a second aryl group and eliminate the halides, and reductive elimination to form the aryl — aryl bond of the product. Due to the high efficiency and number of turnovers, these polymerizations can often be conducted at ~5 mol% catalyst loadings. However, these systems are sensi- tive due to the in-situ formation of zero valent metals. Thus impurities such as moisture, air and protic substances must be excluded from the reaction mixture. 1.8.3 Single ion solid polymer electrolytes Solid polymer electrolytes prepared by dissolving lithium salts in a host suffer from cell polarization and loss of conductivity at high concentrations. These effects contribute to decreased lifetime and efficiency of the electrochemical cell. It is widely accepted that 2.19.80 the formation of ion clusters are the cause of both phenomena. These aggregates of anions and cations form ionic crosslinks between polyether chains by strongly coordinat- 36 ing to the ether oxygens. Indeed the observed ionic conductivity is not simply due to the mobility of single ions, but includes contributions from triple and higher ionic aggregates with unequal numbers of positive and negative charges.105 Formation of aggregates at high salt concentrations reduces the number of charge carriers available to carry current and lowers efficiency. These drawbacks can be averted by a) immobilizing the anion within the electrolyte, b) increasing the size of the anion such that its mobility is signifi- cantly reduced, or c) adding Lewis acids, usually in the form of a ceramic filler that binds to anions and reduces their mobility. However immobilizing the anion causes an ap- 106'108 which reflects the participation proximately order of magnitude loss in conductivity, of the anion in the conduction mechanism. This is better understood through considera- tion of ion transference numbers, the fraction of the total ion current carried by a charged species. Mathematically the transference number is defined as z+D, t =l—t =_____ 1.27 + _ 2+0, —z_D_ ( ) where z,~ is the charge of the carrier and D0i is the diffusion constant associated with that species. In any given system EU, +t_ ] = 1. Typical experimental results for the lithium ion transference numbers in polymer electrolytes range from 0.2 to 0.4, and thus the ma- jority of the current is carried by anions during discharge of a battery. Since the anions do not intercalate into the electrodes, they deposit at the electrodes and form a space charge, thereby inducing polarization of the cell. Obviously, immobilization of the anions elimi- nates a large fraction of the ionic conductivity directly leading to the decreased ionic conductivity observed in single ion conductive polymer electrolytes. In many cases, how- ever, increased cell life and efficiency outweigh the loss in conductivity. 37 The earliest examples of anion immobilization in electrolytes was reported by McIntyre et al.,109 Tsuchida et a1.110 and Shriver et a1.”1 McIntyre reported ionic conductivities of poly(lithium 2-sulphethyl methacrylate) (PSEM-Li) and poly(lithium 2-(4- carboxyhexafluoro-butanoyl-oxy) ethyl methacrylate) (PCHFEM-Li) in comparison with dilithium hexafluoroglutarate (LiHFG). Tsuchida reported single ion conduction in poly(oligo(oxyethylene methacrylate)-co-(alkali-metal methacrylates), while Shriver in- vestigated poly(styrene sulphonates) to achieve single ion conduction. In all cases the ionic conductivities ranged from 1045 to 10'5 S/cm at 30 °C, however, Tsuchida was able to report lithium ion transport numbers of 0.99. Since then, there have been many inves- tigations aimed at achieving higher ionic conductivities (> 10'5 S/cm at room tempera- ture) in single ion conductor SPEs.l '2 38 Polymer name Polymer Structure poly(lithium 2-sulphethyl methacrylate) n poly(lithium 2-(4-carboxyhexafluoro-butanoyl- n 0 CF COO‘ L" oxy) ethyl methacrylate) O O/\/ Y( 2)3 ' O dilithium hexafluoroglutarate/PEG LiOOC—(CF2)3-C00Li 0 CU poly(oligo(oxyethylene methacrylate)-co- \ j/ 1'" o ()6 (lithium methacrylates) 0 My); n poly(sodium styrene sulphonate) SO3Na Table 6. Examples of early polymer systems with immobilized anions designed to bring about single cation conductivity. One class of electrolytes that has received a great deal of attention are those described as composite polymer electrolytes (CPEs) which incorporate inorganic or ceramic particu- lates in a polymer electrolyte. Aside from increased mechanical stability and ionic con- ductivity, the addition of such particles stabilizes the electrode/electrolyte interface. In- corporating a small molecule solvent as a plasticizer in a polymer electrolyte increases 39 the segmental motion of the host polymer and the addition of nanoparticles appears to do the same, giving rise to an straight line Arrhenius-type plotm'1M It is thought that the nanoparticles inhibit the reorganization of the polymer chains through Lewis acid — base interactions between the surface of the nanoparticles and the polymer chains.115 Scrosati et a1. suggested that the enhancement of the ionic conductivity and lithium ion transfer- ence number is due to the formation of pathways at the surfaces of the nanoparticles along which the lithium ions would follow.116 Furthermore, these CPEs can achieve lith- ium ion transference numbers that approach unity if the inorganic nanoparticles are nega- ‘7’120 reported the use of siloxyaluminate tively charged. For example, Shriver et a1.l polymers capable of achieving 10'5 S/cm at 25 °C and lithium ion transport numbers of 0.71. The material is comprised of a siloxyaluminate polymer main chain, with oligo(ethylene oxide) side chains. It is thought that due to the delocalization of anionic charge over the Si-O (p-d)rr system and the presence of substituents that the lithium cations would be inhibited from tightly pairing with the negatively charged backbone and ‘2' However, plots of conductivity versus 1000/T (K4) thus have a higher conductivity. revealed that these systems behave no differently that neat PEO/LiClO4 systems when the side chains were relatively long, 1000 and 2000 g/mol. The conductivity in these cases decreases in a linear fashion until crystallization occurs, at which point the conductivity dramatically reduces to 10'8 S/cm']. Crystallization was not observed for side chains with molecular weights of 400 or 600 g/mol. Fedkiw et a].108 reported recently lithium hectorite based CPEs in which they achieved lithium ion transference numbers of 0.98 and a conductivity of 2 x 10’4 S/cm'l at room 40 temperature. Hectorite belongs to a class of layered clays known as smectites and are characterized as a sandwich structure of cations trapped between negatively charged plates. These platelets are well known to disperse in water and other solvents to create a gelled system. By dispersing lithium hectorite in carbonates such as ethylene carbonate or propylene carbonate a mechanically stable system with high ionic conductivities was achieved. A similar approach would be to disperse modified inorganic fillers as a lithium salt in low molecular weight poly(ethylene oxide). By doing so, issues of electrolyte decomposition are avoided. Baker and Fedkiw et al. 122 demonstrated that incorporation of surface modi- fied silica nanoparticles into poly(ethylene glycol) dimethyl ether of 500 g/mol) (PEG- dm 500) forms CPEs with a high degree of mechanical stability, electrolyte — electrode interfacial stability, and high ionic conductivities on the order of > 10'3 S/cm at room temperature. In a similar fashion to the lithium hectorite CPE system, silica nanoparticles were modified by attaching a lithium sulfonamide salt onto the surface and dispersing in PEG-dm 500. The details of this study are examined in chapter 4. 41 Chapter 2 2 Synthesis of Substituted Poly(p-phenylene)s with Oligo(ethylene oxide) Side Chains 2.1 Introduction Our interests in novel solid polymer electrolytes have led us to consider ethylene oxide substituted poly(p-phenylene)s (PPPs), since the rigid backbone of PPP would endow an electrolyte with thermal and mechanical stability, and the pendant ethylene oxide seg- ments would support ionic conductivity. Although PPP itself is insoluble and unproc- essable, the addition of flexible side chains to PPP’s rigid backbone has proven to be an extremely effective procedure for obtaining tractable and fusible materials. Such materi- als ofien exhibit properties such as pronounced thermochromism and solvatochromism,123 124-128 129-132 thermotropic and lyotropic liquid crystallinity, and structural self-assembly. Because of the strict 1,4 connectivity of the benzene ring in PPP and the ability to vary the length of the oligo(ethylene oxide) side chains, the ethylene oxide substituted poly(p- phenylene)s are an ideal system for studying the structure property relationships of rigid macromolecules. If we view the rigid PPP backbone as a reinforcing element and the oligo(ethylene oxide) side chains as bound solvent, these materials can be regarded as ”92 “molecular composites of two components dispersed at the molecular 1evel.92'm'137 42 There have been several reports of the synthesis of poly(p-phenylene)s substituted with 1.138 ethylene oxide side chains. Most recently Lere-Porte et a used a Stille coupling to synthesize a thienylene-dialkoxyphenylene copolymer with diethylene glycol mono- 1139 used palladium-catalyzed coupling to prepare methyl ether side chains. Johansson et a a related system, oligo(ethylene oxide) substituted polythiophenes. These high molecular weight polymers were used for light emitting electrochemical cells. Wegner et a1.140 re- ported the synthesis of poly(p-phenylene)s with oxyethylene side chains via a Stille-like palladium mediated coupling reaction. These high molecular weight polymers self as- semble into lamellar sheets in which the rigid backbones are aligned with respect to each other. OXOWOTm Colon et aim first reported the use of Ni(0) to synthesize biaryls from aryl chlorides in quantitative yield at mild temperatures and in short reaction times. They extended this catalytic method to the polymerization of bis(aryl chlorides) to give poly(arylene ether I42 sulfone)s. Udea et al. used the same method to synthesize poly(arylene ether ke- 143 tone)s and poly(thiophene)s.I44 Ni(0)-catalyzed coupling reactions have also been used to prepare soluble substituted polyphenylenes. Sheares et al.145‘I46 synthesized substituted 147,148 poly(2,5-benzophenone)s, while Percec et al. reported the synthesis of various solu- ble aryl and alkyl substituted polyphenylenes. 43 Since esters are stable under the conditions of the Ni(0)-catalyzed coupling reaction, a series of ethylene oxide substituted poly(p-phenylene)s are readily available by the po- lymerization of the corresponding oligo(ethylene oxide) 2,5-dichlorobenzoates. The modular synthesis of the monomers and the ready access to their polymers via Ni(O) cata- lyzed coupling encouraged us to investigate a series of poly(p-phenylene)s (PPP—EOm) substituted with oligo(ethylene oxide) monomethyl ethers with various degrees of polym- erization (m). The purpose of this study is to understand the effect of side chain length on the mechanical, thermal and chemical properties of these polymers. Herein, we report the results of the synthesis and characterization of our monomers and their corresponding polymers. 2.2 Results and Discussion 2.2.1 Monomer synthesis The monomers were easily synthesized from 2,5-dichlorobenzoic acid. This modular route offers the flexibility of preparing a series of monomers by simply using various ethylene glycol monomethyl ethers. Only alcohols 3a-3d are commercially available in monodisperse form, and thus alcohols 3e-3h were synthesized using a modified version of a literature procedure.149 The monomers were purified by different methods depending upon their boiling points. Monomer 4a was recrystallized from methanol, and short side chain monomers 4b-4d with lower boiling points were vacuum distilled to yield clear oils. Since monomers with longer side chains thermally decomposed prior to boiling, monomers 4e-4h were purified by column chromatography. 44 Scheme 11. Synthesis of monomers 4a — 4k Cl 0 (3| 0 OH SOCI2 Cl CI Cl Cl 0 CH3(OCHZCH2)mOH 3(a-k) 0(CHZCH20)mCH3 NaH CI 4(a-k) a: m=0 91 "1:6 b: m=1 h: m=7 c m=2 i:m=12avg d: m=3 j: m=16avg e. m=4 k : m=45avg f : m=5 2.2.2 Polymerization The monomers were polymerized utilizing a nickel-catalyzed aryl coupling reaction to yield substituted poly(p-phenylene)s 5a-5k. Originally proposed as an inexpensive, high- yield route to substituted biphenyls tolerance to a variety of functional groups, it provided an efficient method for the polym- erization of dichloro aromatic compounds. polymers were ensured by starting with p-dichloro-substituted aryl monomers. The ac- 142,150,151 104,142,151 45 without terminating side reactions, and with p-Phenyl linkages in the resulting tive catalytic species, a Ni(0) complex, is generated in situ by the reduction of MCI; us- ing finely granulated zinc. Upon mixing of all reaction components, the polymerization solution becomes pink and develops into a dark red color shortly after heating due to the formation of Ni(O). Contamination of the polymerization by oxygen causes the develop— ing color to change to green within the first minutes of the reaction, indicating oxidation of active Ni(O) catalyst to form black NiO. Proton sources terminate the growing polymer chain and can be present as residual moisture in solvent, solvent decomposition products, and in monomers which may contain high boiling alcohols carried over from the synthe- sis of the monomer. Water was eliminated by vacuum distillation of solvent over 3A mo- lecular sieves prior to use, while residual alcohol in the monomers was removed by run- ning toluene solutions of the monomer monomers through a short pad of oven-dried silica gel inside the dry box, and/or through an azeotropic reflux of the monomer in benzene. 2.2.3 Polymer characterization Molecular weights obtained by size exclusion chromatography and calibrated relative to polystyrene standards may afford misleadingly high molecular weights due to the differ- ence in the hydrodynamic volumes (volume occupied by the polymer in a solvent) for chemically different polymers of the same molecular weight. For a random coil such as polystyrene and the linear, rod-like PPP-EOm system, the difference in hydrodynamic volumes is a consequence of the different end to end distances of the two polymers. For a random coil, the end to end distance is 4,112 where n is the number of repeat units and l is the length of each repeat unit, while that of a rigid linear rod is simply n1. Hence the 46 volume occupied by a rod-like polymer in solution is larger than that of a random coil of the same molecular weight.152 Light scattering provides an absolute measure of MW, the weight average molecular weight. Experimentally, the excess scattering, 19, (scattering corrected for scattering by solvent) of a polymer solution per unit volume at angle 0 is measured as a function of concentration and angle. The Raleigh ratio, defined as Ig/Io where IC) is the intensity of the incident light, can be expressed in terms of MW as shown in equation (2.1) R(6) = K *c(1+ cos2 6)Mw (2.1) where c is the concentration of the solute molecules (g/mL) and K* is a constant. K“ = 4n2(dn/dc)2n02/(NAAO4), where no is the refractive index of the solvent, NA is Avogadro’s number, 7&0 is the vacuum wavelength of the incident light, and dn/dc is the specific re- fractive index increment, which describes the change in refractive index of the polymer solution with solute concentration. The data from a light scattering experiment are usually analyzed using a Zimm plot where the quantity K*c/R(0)(l+cos20) is plotted as a function of both concentration and 0. Simultaneous extrapolation of the data to zero 0 and zero concentration gives l/Mw as the y-intercept. When light scattering is used as a detector in size exclusion chromatog- raphy, a concentration detector such as refractive index or UV-vis detector provides the concentration information as well as the molecular weight distribution. The molecular weight distribution can be analyzed to give the polydispersity index (PDI), Mw/Mn Since 47 M“, is directly measured by light scattering, Mn can be easily calculated. The definitions of Mn and MW are shown below in equations (2.2) and (2.3) Zn,M, 2c, 5 in. = 26./M. i 2:71,.M‘,2 ZciMi w: 'zni : '26!- I M, (2.2) M (2.3) The only polymer-specific quantity that must be known for light scattering is dn/dc, the specific refractive increment. Values of dn/dc for several polymers are reported in the literature, however since the PPP-E0m had not been synthesized previously, the value of dn/dc must be determined experimentally. dn/dc was measured for PPP-E0m where m = 0,1, 3, 5, <12> and <45>. The dn/dc values for the rest of the polymers in the series were calculated from the weighted average of PEO and PPP-E00 in the polymer, as shown in equation (2.4) dn dn dn _ : WA— + WB— (2.4) d C polymer dc A dc B O OCH3 A = B = (OCH2CH21m where component A is defined as PPP-E00 and component B is PEO. A plot of the ex- perimentally determined values of dn/dc as a function of the weight fraction of PPP-E00 48 in the polymer (Figure 16) yields a line where the left and right y-intercepts correspond to the dn/dc of PPP-E00 and PEO, respectively. The linearity of the data validates the use of equation (2.4). Figure 17 compares the experimental dn/dc values with those cal- culated using equation (2.4). As noted above, a UV-vis detector provides the concentration of polymer analyte in solu- tion for the calculation of M... The extinction coefficient (a) of each polymer can be ob- tained in the same manner as the specific refractive increment as shown in equation (2.5) 8A8 = WAEA + W883 (25) using the same definitions for A and B as in equation (2.4). Since 8 for ethylene oxide is zero in the visible part of the spectrum, a should simply reflect the weight fraction of the PPP component in PPP-EOm polymers. Values of s were measured for PPP-E0m where m = 0,1, 3, 5, <12> and <45>. A plot of s of the polymer versus the weight fraction of PPP in each polymer (Figure 18) was linear with an intercept of zero and the extinction coefficient of PPP as the slope. The 8 values for the rest of the polymers in the series were calculated from the weighted average of PEO and PPP-E00 in the polymer, using equation (2.5). 49 0.25 p t 1 0.20 : x" o I ,x'" u.’ i- I”’. g 0.15 r , x,” a h ,”’ 8 : 0.10 r x”. E : ‘."" 13 . ”,2’ 005 ': dn/depp= 0.203 : dn/cho = 0.059 000 4 1 1 1 1 1 1 1 14 1 1 J 1 l 1 1 1 1 L1 J L1_1‘ 0.0 0.2 0.4 0.6 0.8 1.0 wppp Figure 16. Extrapolation of dn/dc of PPP-E0m as a function of PPP weight fraction to obtain dn/dc of PPP-E00. 0.25 A - — Calculated 0.20 I . l A Experimental \ o A, 3 0.15 - ‘ 3 \ 8 \ E 0.10 - b\ c -‘L~~~_ -------- A 0'05 ' dn/dcpp, = 0.2031cc/ml dn/dceo = 0.0591cc/ml .1 lLAllllAlAh O‘OOEALJAIAAAAIAII 0 10 20 30 40 50 Ethylene Oxide Units (m) Figure 17. Comparison of calculated values of dn/dc for PPP-E0m with experimen- tally obtained data. 50 6.00E+04 =49702x 5.001904 » y2 ) R =0.9743 I,” I I ’5 400E+04 - t x .r ' x E r’ I I 3 3.001304 ~ ,I 8 1’ E x' w 2.00E+04 '- ’I” ’1‘ 1,005+04 » o/ I 0” 000E+00 1 1 1 1 l 1 1 1 1 L4 1 1 1 l 1 1 J 1 1 1 1 1 1 0.0 0.2 0.4 0.6 0.8 1.0 WPPP Figure 18. The extinction coefficient (a) of PPP-EOm as a function of PPP weight fraction 51 GPC' Light Scattering' DSC” m MW Xll PDI dn/dc Mw Xn PDI T: TIn 0 17,613 68 1.9 0.203 6,124 32 1.43 37 1 59,207 117 2.8 0.168 30,120 135 1.25 33 2 19,330 59 1.5 0.146 15,520 52 1.34 -5 3 65,795 94 2.6 0.132 48,790 107 1.71 -34 4 22,738 37 2.0 0.121 15,430 41 1.21 -29 5 23,008 29 2.2 0.114 18,570 35 1.48 -30 6 32,782 33 2.5 0.108 31,710 61 1.30 -44 7 20,324 22 2.1 0.103 17,870 29 1.40 -51 <12>c 29,079 21 2.1 0.088 25,770 22 1.73 -43 c 22,776 15 1.8 0.082 20,920 19 1.29 -54 22 <45>c 46,860 13 1.8 0.068 46,590 18 1.20 -53 49 Table 7. Properties of PPP-E0...n polymers. a) Molecular weight data were obtained from GPC measurements in CHC13 using a light scattering and a refractive index detec- tor. b) T3 and Tm values were obtained from MDSC and DSC measurements. c) The brackets < > indicate samples with polydisperse EO chains with average degrees of po- lymerization = m. 52 3.5 3.0 - l 2.5 I 2.0 1' 1.5 L GPCILS Mw 1.0 r o' -— - - .- - - __,. 0.5 - 0.0 L l l J l l l J l 0 5 1O 15 20 25 30 35 4O 45 50 Ethylene Oxide Units Figure 19. Plot of the ratio of molecular weights obtained by GPC to those obtained by LS as a function of the number of ethylene oxide units in the side chains of PPP-EOm polymers. Table 7 presents representative molecular weights obtained from size exclusion chroma- tography coupled with a light scattering detector. As noted previously, the GPC values obtained for PPP-EOm polymers with short side chains should be higher than those ob- tained from light scattering. However, both techniques begin to agree at approximately m = 6 where the ratio of the Mw values is almost 1. This suggests that in dilute solution, the polymers with long side chains behave as a random coil of poly(ethylene oxide) with a rigid core. These results agree with characterization data from similar PPP based poly- meTS.l53']54 53 2.2.4 Differential scanning calorimetery (DSC) Related PPP systems with short ethylene oxide chains are reported to show glass transi- ‘55 These poly- tion (Tg), melting (Tm) and in some cases crystallization (Tc) transitions. mers differ from the those of this study in that the PPP backbone consists of di- substituted aromatic rings spaced with unsubstituted phenyl rings, and possess polydis- perse oligo(ethylene oxide) chains. Shown in Figure 20 are DSC scans of the PPP-EOnn polymers with long side chains. No distinct thermal transitions were observed for poly- mers with m < 12 (5a-5h) using conventional DSC. However, modulated DSC did reveal glass transitions for the short side chain polymers which show a near-linear decrease in T8 from 40 to —60 °C as the side chain length increases (Figure 21). The polymer melting points show similar behavior, with the melting point of the long side chain polymers ap- proaching that of poly(ethylene oxide). Both the Tg and Tm data indicate the increasing “PEO character” of the polymers as m increases. 54 PEG-dm-2000 \ ’5 E E m=<45> ; \ .9 u. 46 m=<16> :“c’ \ m=<12> m=7 -100 -50 0 Temperature (°C) Figure 20. DSC scans of PPP-E0m (7 < m > <45>); PEG-dm-2000 is shown for comparison. The data are second heating scans, taken at 10 °C/min afier quenching the sample from 180 to ~100 °C. The arrows mark the Tg transitions. 55 60 40 I \\ ' PEG-dm-2000 20 ~ \ E 0 _ T \\.\ PPP-E07 13’ ‘20 ‘ PPP-E00 \\“ 40 - ' .9. . V -60 - \ o -80 , 1 . . . T 0.0 0.2 0.4 0.6 0.8 1.0 Weight fraction of ethylene oxide Figure 21. The relationship between Tg and the weight fraction of PEO in the side chains. The dashed line is a least squares fit to the data. 2.2.5 Thermogravimetric analysis (TGA) Thermogravimetric analysis reveals the dual nature of the PPP-EOm As seen in Figure 22 the PPP-EOm polymers exhibit similar thermal profiles, however, the onset of degrada- tion decreases as the side chains become longer. Since PPP degrades in inert atmospheres to a char that is stable to 800 °C)” the observed weight loss at T < 500 °C must corre- spond to the degradation of the PEO side chain. A simple model that assumes loss of the side chain and the ester group matches the weight loss data well. PEO is thought to de- grade to dioxanes through a radical process, and since PPP-EOm polymers with long side chains have a degradation profile that is nearly identical to that of PEO, they too may de- grade by the same mechanism. In contrast, PPP-EOm polymers with short side chains show enhanced stability, which may reflect the decreased number of sites to support 56 PEO—like degradation. Thus de-esterification and decarbonylation must be more impor— tant in the degradation of short side chain PPP-E0m polymers. 100 » \ 80 ~ .3 60 L H .C .9 m=1 0 g 40 ~ m=4 20 » m=6 PEG-dm-2000 O 1 . 1 . 0 200 400 600 800 Temperature (°C) Figure 22. TGA scans of PPP-E0m polymers. Polymer samples were dried under vacuum at 100 °C overnight, equilibrated under N2 at 80 °C for 12 hrs, and then heated under N2 at 10 °/min. 57 2.2.6 X-ray diffraction (XRD) Figure 23 shows X-ray diffraction data for the PPP-EOm polymers. Scans of PPP-E02 and PPP-E03 show a peak at approximately 20 = 5 (d = 15A) that disappears as the side chain increases in length. The same peak has been observed for PPPs disubstituted with triethyleneoxide monomethyl ether side chains, and has been assigned to the spacing be- tween the rigid polymer backbonesn‘w‘l4 The loss in structural periodicity reflects the transition from a PPP-like structure to one more characteristic of PEO. As the side chains lengthen, steric limitations to side-chain crystallization are eased and the chains eventu- ally organize in a structure characteristic of PEO. The scattering data for PPP-EO<45> are nearly identical to that of PEO. 58 IAJ PEG-dm-2000 h 1. JJ _EI:> P-E8<45> Relative Intensi 20 Figure 23. Room temperature X-ray powder diffraction scans of PPP-EOm polymers. All scans were collected at a rate of 0.05 20/min over a range of 20 = 1-50°. 59 Comb and brush structures can be proposed for the structures of these polymers. In the first, the backbone phenyl rings form a nearly flat ribbon (small dihedral angle between each ring), with side chains extended from the back bone to give an overall flat structure. The comb could be either single-sided (normal comb) or double—sided (teeth project from both sides of the backbone) The second model is akin to a polymer brush, in which the dihedral angle between rings is large and the side chains radiate from that center. We propose that the dominant solid state structure must be comb-like, while a brush architec- ture makes more sense for the structure in solution. The fluorescence data argue for a brush architecture in solution. Figure 24 shows that the fluorescence data for the PPP-E0m series are similar with only small shifts in 2mm. Thus the effective conjugation length is the same for each polymer, and the average dihedral angle between rings in the backbone must also be similar. If we compare the shift in 2mm for the polymers to that of a model repeat unit, methyl benzoate (398 nm), the similar 7km“ values argue for limited conjugation along the backbone and large dihedral angles. However, conjugation is manifested in the dramatically increased luminescence intensity for the polymers compared to the monomers. (The fluorescence intensities for the poly- mers and monomers are normalized such that the number of rings in each sample is con- stant). We also noted an increase in fluorescence intensity with increases in the length of the E0 side chain (inset to Figure 10). These effects are likely due to the side chains in- creasingly acting as the “solvent” for the PPP backbone in the fluorescence experiments. 60 Raman spectroscopy provides some insight into the solid state structure. Hernandez et a1. 1.158 used Raman spectroscopy to interrogate the structure of PPP poly- 157 and Lefrant et a mers. Figure 25 shows Raman spectra several PPP-E0m polymer films and PEG-dm- 2000. The prominent peak at 1608 cm'1 is due to ‘in plane’ asymmetric bending, which is sensitive to conjugation and hence the dihedral angle between adjacent rings. According to Pimenta, Mathews et al.'59 the Raman shift of the ‘planar’ form of PPP is 1601 cm], while the values calculated for helical PPP with 20° and 50° twists are 1630 and 1635 cm'1 respectively. The observed 1608 cm'1 transition is thus consistent with a nearly pla- nar PPP backbone. PPP—E0 monomers Relative |ntensrty - I // V/ /‘ ‘/ 300 350 400 450 500 550 600 Wavelength (nm) Figure 24. Fluorescence spectra of PPP-E0m and their corresponding monomers (in- set). Data were collected at room temperature in chloroform using an excitation wave- length of 313 nm. 61 PPP-E045 PPP-E0, PEG-2000 I 1 1 1 J 1800 1600 1400 Raman Shift (cm'I) Figure 25. Solid-state Raman spectra of PPP-E0m polymers. Data shown were ob- tained at room temperature and are the average of 20 scans of 6 seconds in duration. 2.2.7 Polarized light optical microscopy Rod-like polymers often exhibit liquid crystallinity. Preliminary tests show that some members of the PPP-E0m family are thermotropic and orient under shear. Examination of PPP-E02 at 140 °C under cross-polarizers (Figure 26) shows that the polymer melts into small circular domains that retain some degree of order. Upon shearing the sample, these domains elongate, maintain their alignment, and return to their original state after the re- lief of shear stress. This behavior disappears for m = 5-12. Instead, these polymers sim- 62 ply soften and show no birefringence. Samples with m > 12 are crystalline, and as ex- pected from the X-ray data, exhibit melting and recrystallization similar to PEO. Figure 26. Cross polarized optical micrographs of PPP-E02 at 140 °C a) after 5 min- utes at 140 °C b) afler shearing the sample and c) 1 minute after shearing showing break- up of the elongated domains. 63 2.2.8 Conclusions The Ni(O) mediated polymerization of ethylene oxide substituted 2,5-dichlorobenzoate monomers produces high molecular weight polyphenylenes with pendant ethylene oxide side chains. As shown by Tg, Tm, and X-ray data, the characteristics of these polymers evolve from PPP-like to that of PEO as the length of the side chains increase. With the exception of PPP-E00, PPP-E01, and PPP-EO<45>, all polymers are amorphous at room temperature and have Tgs below 0 °C. The Tgs of E00 and PPP-E0) are 37 and 33 °C, respectively, while PPP-EO<45> exhibits PEO-like crystallinity at room temperature. Simple changes in the length of the side chains enable tuning the physical properties of these polymers to match applications in display devices or as solid polymer electrolytes. Further investigation into the lithium ion conductivity of the PPP-EOm polymers will be described elsewhere. ‘60 2.3 Experimental Materials. Reagent grade nickel chloride hydrate was heated (250 °C) under vacuum to give a fine, orange powder with constant weight. Triphenylphosphine was recrystallized from ethanol (95%), dried over CaSO4, and after removing the ethanol, dried under vac- uum. Granulated zinc was purified by stirring in acetic acid, filtered, rinsed thoroughly with diethyl ether, and dried under vacuum. It was then crushed with a mortar and pestle to increase its surface area. The polymerization solvent, N,N-dimethyl formamide (DMF), was stirred over molecular sieves (3A, activated by heating ~300 °C under vac- uum), and distilled under vacuum. Poly(ethylene oxide) dimethyl ether (PEO-dm-2000, ca. 2000 g/mol) was obtained from Aldrich and was dried by azeotropic distillation of 64 benzene, followed by removal of the solvent in vacuo. Unless otherwise specified, all other materials and solvents were ACS reagent grade and were used as received from commercial suppliers without further purification. Characterization. Proton nuclear magnetic resonance ('H NMR) spectra were measured using a Varian Gemini-300 spectrometer at 300 MHz. All samples were run at room tem- perature in CDC13. Chemical shifls were calibrated using residual CHCl3 and are reported in ppm (5) relative to tetramethylsilane. Molecular weights of polymers were determined using a Wyatt Technologies miniDAWN light scattering detector, and a Waters 410 Dif- ferential Refractive Index detector. CHC13 was used as the eluting solvent at a flow rate of 1 mL/min. DSC data were obtained under helium using a Perkin-Elmer DSC 7 in- strument at a heating rate of 10 °C/min. The DSC 7 temperature was calibrated with an indium standard. The reported DSC results from polymer samples are the second heating scan, taken after flash cooling the sample from the melt to erase the thermal history. TGA were run under nitrogen and air at a heating rate of 10 °C/min on a Perkin-Elmer TGA 7 instrument. Elemental analyses were performed using a Perking-Elmer 2400 Se- ries II Analyzer. X-ray diffraction patterns were measured using a computer controlled Rigaku 2008 rotating anode diffractometer operating in reflective mode at 45kV/ 100 mA, with graphite monochromatized Cu (1(a) radiation. Fluorescence spectra were ob- tained at room temperature using a computer controlled Jobianon Spex Fluorolog-3 spectrometer utilizing a 450W Xe lamp at an excitation wavelength of 313 nm, and a slit width of 5.0 nm, scanning from 320 to 600 nm. Raman spectra were collected at room temperature using a computer controlled Chromex RAMAN 2000 spectrometer 65 (Chromex, Inc.) utilizing a diode-pumped, frequency doubled CW NszAG Laser (500 mW at 532 nm, Coherent), a Chromex 500is spectrometer (f/4, 600 grooves/mm holo- graphic grating), and a thermoelectrically cooled 1024 x 256 charge coupled device (CCD) detector by Andor Tech. Ltd. Typically, 20 6-second scans were collected for each spectrum with an incident power density of ca. 500 kW/cm2 (100 mW at the sample at 5 pm diameter spot size). Ethanol, 2-[2-[2-(2-hydroxyethoxy)ethoxy]ethoxy], 1-(4-methylbenzenesulfonate) Ts(OCH2CH2)4OH] (6) An aqueous solution of KOH (67.79 g, 1.21 mol, in 200 mL H20) was added to a solu- tion of tetra(ethylene glycol) (469.30 g, 2.42 mol) in THF (200 mL) at 0 °C. While stir- ring at 0 0C, a solution of tosyl chloride (114.93 g, 0.60 mol) was added dropwise. The reaction mixture was allowed to warm to room temperature and stir overnight. The reac- tion mixture was then dissolved in 1500 mL of water and extracted with chloroform (2 x 1500 mL). The combined organics were washed with saturated aqueous NaHCO; (2 x 1500 mL) and water (2 x 1500 mL), then dried over MgSO4, filtered and the solvent re- moved under reduced pressure. The crude mixture was then dissolved in methanol (1200 mL) and allowed to cool in a freezer overnight. The mixture was filtered and solvent re- moved to afford a yellow oil (163.0 g, 78 % yield) with spectral data that match those previously reproted.‘6' 'H NMR (300 MHz, CDC13) 8 2.4 (s, 3H, 3.4-3.8 (14 H), 4.1 (t, 2H), 7.3 (d, 2H), 7.8 (2H). 66 Ethanol, 2-[2-[2-[2-[(tetrahydro-2H-pyran-2-yl)oxy-]ethoxy]ethoxy]ethoxy]-, 1-(4- methylbenzenesulfonate) [Ts(OCH2CH2)4OTHP] (7). p-Toluenesulfonic acid (8.9 g, 0.05 mol) and 3,4-dihyrdo-2H-pyran (129.86 g, 1.55 mol) were added to a stirred solution of compound 6 (163.0 g, 0.46 mol) dissolved in dioxane (700 mL) and allowed to stir at room temperature overnight. The reaction mixture was then neutralized with triethylamine and concentrated under reduced pressure. The crude product was dissolved in chloroform (600 mL), washed with saturated aqueous NaCl (3 x 400 mL), dried over MgSO4, filtered and the solvent removed under reduced pressure to afford the crude product as a brown oil with spectral data that match those previously re- ported.162 The product was used without further purification. 1H NMR (300 MHz, CDC13) 8 2.4 (s, 3H), 3.4-3.9 (m, 14H), 4.1 (t, 2H), 4.6 (t, 1H), 7.3 (d, 2H), 7.8 (d, 2H). Alcohols 3e-3h: General Procedure. The appropriate alcohol (0.18 mol) was added as a THF solution (30 mL) to a stirred suspension of NaH (12.67 g, 0.34 mol) in THF (50 mL) at 0 °C, and stirred for an additional 30 min. A solution of compound 7 (70.83 g, 0.16 mol in THF 200 mL) was then added dropwise and the reaction mixture was heated to reflux for 6 hrs. Upon cooling, the reaction mixture was filtered and solvent removed un- der reduced pressure. The crude material was dissolved in a mixture of ethanol (400 mL) and 2N HCl (200 mL). The solution was refluxed overnight, and was then concentrated under reduced pressure to yield a dark brown oil, which was purified as indicated below. 2,5,8,1l-Tetraoxatetradecan—14-ol [CH3(OCH2CH2)4OH] (31). Reaction of methanol with 7 gave compound 3f as a clear colorless oil. Vacuum distillation (120 °C / 200 67 mtorr) yielded 8.63 g (25%) with spectral data that match those previously reported.92 1H NMR (300 MHz, CDC13) 8 3.35(s, 3H), 3.48-3.75(m, 16H). 2,5,8,11,14-Pentaoxahexadecan-l6-ol [CH3(OCH2CH2)5OH] (3g). Reaction of 2- methoxy ethanol with 7 gave compound 3g as a clear colorless oil, which was purified by column chromatography (silica gel) with a gradient eluent from hexane to chloro- form/methanol (9:1) to yield 6.00 g (14.5%) with spectral data that match those previ- ously reported.92 'H NMR (300 MHz, CDC13) 8 3.35(s, 3H), 3.48-3.75(m, 20H). 2,5,8,11,14,]7-Hexaoxanonadecan-l9-ol [CH3(OCH2CH2)60H] (3h). Reaction of 2- methoxy diethylene glycol with 7 gave compound 3b as a clear colorless oil, which was purified by column chromatography (silica gel) with a gradient eluent from hexane to chloroform/methanol (9:1) to yield 1.42g (15%) with spectral data that match those pre- viously reported.92 ‘H NMR (300 MHz, CDC13) 8 3.35(s, 3H), 3.48-3.75(m, 24H). 2,5,8,l1,14,]7,20-Hexaoxadocosan-22-01 [CH3(0CH2CH2)7OH] (3i). Reaction of 2- methoxy triethylene glycol with 7 gave compound 31 as a clear colorless oil, which was purified by column chromatography (silica gel) with a gradient eluent from hexane to chloroform/methanol (9: l) to yield 1.83g (23%) with spectral data that match those pre- viously reported.'63 ‘H NMR (300 MHz, CDC13) 8 3.35(s, 3H), 3.48-3.75 (m, 28H). Monomer synthesis: General procedure. The alcohol (3a-l, 37 mmol) and 2,5- dichlorobenzoyl chloride (5.16 g, 24.6 mmol) were dissolved in pyridine (50 mL) and 68 heated (SO-60 °C) overnight. Upon cooling, methylene chloride (100 mL) was added and the solution was washed three times with 50 mL of 1N HCl (aq) and once with 50 mL of H20. The combined organics were dried over magnesium sulfate, filtered, and the solvent removed under reduced pressure. The compounds were dried by refluxing in benzene us- ing a Dean-Starke trap and purified by recrystallization (4a), vacuum distillation (4b-4d) or column chromatography (4e-4l) inside a drybox. Methyl-2,5-dichlorobenzoate (4a). Compound 4a was purified by recrystallization from methanol to yield 4.74 g (94%). mp 35 °c (DSC) lit.164 (mp 38-39 °C). ‘H NMR (300 MHz, CDC13) 8 3.9(s, 3H), 7.4(d, 2H), 7.8(t, 1H). 3-0xybutyl-2,5-dichlorobenzoate (4b). Compound 4b was purified by vacuum distilla- tion at 105 °C/80 mtorr to yield 5.63 g (92%). 1H NMR (300 MHz, CDCI3) 5 3.4(s, 3H), 3.7(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C10H10C1203: C, 48.22; H, 4.05. Found: C, 48.26; H, 4.03. 3,6-Dioxyheptyl-2,5-dichlorobenzoate (4c). Compound 4c was purified by vacuum dis- tillation at 170 °C/ 130 mtorr to yield 5.98 g (83%). 1H NMR (300 MHz, CDC13) 5 3.4(s, 3H), 3.55(t, 2H), 3.7(t, 2H), 3.85(t, 2H), 4.5(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C12H14C1204I C, 49.17, H, 4..81 Found: C, 49.01, H, 4..50 3,6,9-Trioxydecyl—2,5-dichlorobenzoate (4d). Compound 4dl was purified by vacuum distillation at 195 °C/200 mtorr to yield 6.55 g (79%). 1H NMR (300 MHz, CDC13) 6 69 3.35(s, 3H), 3.5(t, 2H), 3.65(m, 6H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C14H13Cl205: C, 49.87; H, 5.38. Found: C, 49.75; H, 5.39. 3,6,9,12-Tetraoxytetradecyl—2,5-dichlorobenzoate (4e). Compound 4e was purified by colurrm chromatography (silica gel) with hexane/ethyl acetate (2: l) as the eluting solvent to yield 7.83g (54%). 1H NMR (300 MHz, CDC13) 5 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 10H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C16H22C1206 C 50.41; H 5.82. Found C 50.29; H 5.64. 3,6,9,12,15-Pentaoxyhexadecyl-2-5-dichlorobenzoate (41). Compound 4f was purified by column chromatography (neutral alumina) with hexane/ethyl acetate (2:1) to yield 3.39g (37%) 1H NMR (300 MHz, CDCI3) 5 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 14H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd for C13H26CI2O7 C 50.83; H 6.16. Found C 50.59; H 6.99. 3,6,9,12,15,18-Hexaoxynonadecyl-2-5-dichlorobenzoate (4g). Compound 4g was puri- fied by column chromatography (neutral alumina) with hexane/ethyl acetate (2:1) to yield 1.24g (55%) 1H NMR (300 MHz, CDC13) 5 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 18H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). ) Anal Calcd C20H30C1203 C 51.18; H 6.44. Found C5083; H 6.75. 3,6,9,12,15,18,21-Heptaoxyicosyl-2,5-dichlorobenzoate (4h). Compound 4h was puri- fied by column chromatography (neutral alumina) with hexane/ethyl acetate (2:1) to yield 70 1.83g (74%) ‘H NMR (300 MHz, CDC13) 8 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 22H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal Calcd C22H34Cl209 C 51.47; H 6.68. Found C 51.78; H 6.93 3,6,9,12,15,18,21,24,27,30,33,36-Dodecaoxyheptacontyl (average) -2,5- dichlorobenzoate (4i). Compound 4i was purified by column chromatography (silica gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 15.12 g (85%). 1H NMR (300 MHz, CDC13) 5 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 40H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C32H54Cl2014: C, 52.39; H, 7.42. Found: C, 51.15;H, 7.17. 3,6,9,12,15,18,21,24,27,30,33,36,39,42,45,48-Hexadeca-oxynonatetracontyl (average) -2,5-dichlorobenzoate (4j). Compound 4j was purified by column chromatography (sil- ica gel) with hexane/ethyl acetate (2:1) as the eluting solvent to yield 18.39 g (81%). 1H NMR (300 MHz, CDC13) 5 3.35(s, 3H), 3.5(t, 2H), 3.65(b, 58H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C40H70Cl20181 C, 52.80; H, 7.75. Found: C, 51.04;H, 7.15. 3,6,9,12,15,18,21,24,27,30,33,36,39,42,45,48,51,54,57,60,63,66,69,72,75,78,81,84,87,90, 93,96,99,102,105,108,111,114,]17,120,123,126,129,132,135-Pentatetraconta- oxyhexacontahectyl (average) -2,5-dichlorobenzoate (4k). Compound 4k was purified by column chromatography (silica gel) with hexane/ethyl acetate (2:1) as the eluting sol- vent to yield 32.07 g (60%). 1H NMR (300 MHz, CDC13) 5 3.35(s, 3H), 3.5(t, 2H), 71 3.65(b, l74H), 3.8(t, 2H), 4.45(t, 2H), 7.4(d, 2H), 7.8(t, 1H). Anal. Calcd. for C93H136C12047: C, 53.81; H, 8.57. Found: C, 53.80;H, 8.73. Nickel catalyzed polymerization, General procedure: Inside a helium atmosphere drybox, the monomer (1.8 mmol) was added to NiClz (23.3 mg, 0.18 mmol), 2,2’- bipyridine (28.1 mg, 0.18 mmol), triphenylphosphine (283.3 mg, 1.08 mmol), zinc pow- der (0.71 g, 1.08 mmol), and DMF (1.0 mL) and pre-heated for 30 min. at 80 °C in a 10 mL vial with a Teflon septum cap. The vial was then immersed in an oil bath set for 85 °C. The pink solution became red after 10 minutes, eventually turning deep red-brown after 30 minutes. The mixture was stirred overnight. Upon cooling, the solution was transferred into a flask containing 200 mL of 2N HCl/MeOH (1:1) and stirred until the excess zinc dissolved. The polymer was extracted from the HCl/MeOH mixture with me- thylene chloride (3 x 50 mL). The combined organic layers were dried over MgSO4, fil- tered, and the solvent removed under reduced pressure to yield a pale yellow film. The polymer was then purified via precipitation of a concentrated chloroform solution of the polymer into MeOH, filtered and dried. 72 Chapter 3 3 Lithium Ion Conductivity of Substituted Rigid Rod Solid Polymer Electrolytes 3.1 Introduction Solid polymer electrolytes (SPE) have been the focus of intense research because of their importance to applications such as lithium batteries and electrochromic devices. Since Fenton36 and Armand8 first reported the use of poly(ethylene oxide) (PEO) as a polymer electrolyte, polyether matrices have been the most studied due their ability to dissolve inorganic salts. In such systems the metal cations coordinate to the ether oxygens and thus their mobility is strongly coupled to the segmental motion of the polymer. In PEO, a semi-crystalline polymer, this limits ion transport to the amorphous regions of the poly- mer host. A number of approaches have been reported that limit crystallinity and increase the seg- mental chain motion and mechanical stability of polyether matrices. These include the use of graft and block copolymers, comb architectures, crosslinked systems and blending polyethers with plasticizing additives.S 8"36'165"67 To this end, Meyer and Wegner reported comb polymers that have rigid backbones and flexible oligo(ethylene oxide) side chains as teeth. In such structures the rigid backbone contributes thermal and mechanical stabil- ity, while the side chains provide a highly conductive medium for ionsglm’né‘I40 They reported good conductivities and provided proof that their system formed a lamellar structure composed of comb polymers. 73 Khan described the synthesis and characterization of a related system, a family of poly(thiophene) copolymers that contain thiophene units substituted with a single PEG- me-350 side chain (H(OCH2CH2),,OCH3, n~7).I68 Room temperature conductivities of 10°6 S/cm were reported for the homopolymer that has one PEG 350 chain per repeat unit. The rigid backbone comb polymers prepared to date generally contain polydisperse eth- ylene oxide units, and no systematic study of the effect of the side chain length on the thermal and conductive properties of these molecularly reinforced solid polymer electro- lytes has been reported. The poly(p-phenylene) system provides one of the best frameworks for studying the properties of rigid combs. The parent polymer lacks significant thermal transitions below 138 400 °C 169"") and is electrochemically stable. Attaching pendent ethylene oxide side . . . , . _ . ‘ 6‘ 4 chains to the I'lgld backbone provrdes a conductive medium for 1011 transport.”92 ‘3 l 0 9 ” 2 of two com- These “hairy rod” molecules87 can be regarded as “molecular composites ponents, a rigid polymer and PEO, that are dispersed at the molecular level.92"33'136 Steric interactions between substituents at the ortho positions increase the torsional angle be- tween the phenyl rings.‘3 7'17"”2 Twisting of adjacent rings out of planarity introduces an element of disorder and creates an environment in which the side chains have a larger spatial separation than for a planar conformation, thereby enhancing segmental chain mo- tion. However, in the limit of long oligo(ethylene oxide) side chains, the steric limits in- troduced by the twisting rings would be less important and the side chains would be ex- pected to associate and crystallize. Understanding the ionic conductivity of these co- polymers as a function of side chain length is the goal of this investigation. 74 We report herein the characterization of a series of poly(p-phenylene)s (PPP-EOm) substi- tuted with various length (m) ethylene side chains. We believe that a systematic study of the structure property relationships of these polymers can provide insight into how to bal- ancc high segmental motion with the need to maintain dimensional stability. 3.2 Results and Discussion 3.2.] Thermal properties The glass transitions of comb polymers depend on the rigidity of the polymer backbone and the length and stiffness of the teeth of the comb. For the PPP-E0". system, we ex— pected that for short teeth, the physical properties would be dominated by the PPP back- bone, but in the limit of very long teeth, the properties of the comb would converge to those of PEO. The most interesting material from a conductivity perspective would be when the teeth are long enough to have high segmental motion, but still too short to crys- tallize. DSC and MDSC were used to measure the Tgs and Tms of the PPP-150m polymers. As shown in Figure 27, the Tgs decreased as the lengths of the PEO segment increased, 75 eventually converging to the T8 of PEO for side chains of m = <45>. The glass transi- tions for polymers with short side chains were particularly weak, and could only be de- tected by MDSC. Crystallization was first observed at m = 16, which melts at 22 °C, and increasing m to 45 increased Tm to 42 °C. It is interesting to compare these data with PEO homopolymers of comparable molecular weights. For example, PEG-dm-SOO (m~10) has a melting point of 15 °C, while PEO-dm-ZOOO (m=<45>) melts at 49 °C. Ig- noring crystallization, polymers with the longest side chains should exhibit the highest conductivities because of their low T8. T9 (°C) Figure 27. 60 40 20 ’ L \ ’ \ AA - \\\‘ .‘firt---fi----r-l 10 20 30 4O Ethylene oxide units (m) The relationship between TS and the length (m) of the ethylene oxide side chain in PPP-E0m polymers. The data were acquired through MDSC. The line is a guide. 76 Neat Composite m Mn PDI Tg Tm Tg Tm 3 48,800 1.71 -34 - - - 7 17,900 1.40 -51 - - - <16> 20,900 1.29 -54 22 -41 - <45> 46,600 1.20 -53 49 -26 42 PEG-dm-ZOOO ~2000 - -60 66 -26 48 Table 8. Properties of PPP-E0nrl polymers 77 -- neat polymer — composite :1 '1 1' I I PEG-dm-2000 A / a _____ —:—=‘:‘."':_ ——————————————— \ _ E B E. a 2 '1 LI. PPP-EO<45> ,' . *5 fl: .3; ______ :7: ........... ,’ l _____ PPP-E0<16> ,4 l I l ‘ M , _______ v \ __________ l I \I 1 l J l l -75 -50 -25 0 25 50 75 Temperature (°C) Figure 28. DSC scans of neat PPP-ISOm (dashed) and LiClOanPP-EOm composites (solid). The scans shown are the first heating scans afler quenching the sample from 180 °C. The samples were heated at 10 °C/min under He. 78 Dissolution of salts in polyethers generally results in a decrease in crystallinity and a par- allel increase in Tg. The initially formed amorphous state is oflen a kinetic product; crys- tallization of PEO salt complexes is well-known.173 ‘179 The data for the polymer/LiClO4 complexes (O:Li = 20) are shown in Figure 28. The addition of LiClO4 depresses both the melting point and the crystallinity, as indicated by the decrease in AHfus. PPP- EO<16>/LiCIO4 is rendered amorphous, while the corresponding PPP-E045 complex shows a slight decrease in Tm, and a more substantial decrease in crystallinity. These re- sults predict that PPP-E016/LiClO4 should have the highest conductivity at room tem- perature. It is useful to note that the behavior of m = 45 and PEO-dm-ZOOO are qualita- tively the same, pointing out the transition in physical pr0perties from PPP-like to PEO as the side chains lengthen. The addition of LiClO4 to PPP-EO<16> and PPP-EO<45> leads to an increase in Tg. Increases in T8 are usually ascribed to transient cross-links, the coordi- nation of lithium cations with neighboring polyether chains. These ties between chains also inhibit crystallization by decreasing chain mobility. 3.2.2 X-ray diffraction The XRD data of the composites (Figure 29) reveal that the solubility of the lithium salt in PPP-E0m polymers depends on the length of the side chains. Despite all samples shown in Figure 29 having the same OzLi ratio, the XRD patterns for PPP-E03 and PPP- EO7 show evidence of undissolved LiClO4, while longer chains completely solubilize LiClO4. This effect is easily understood by considering the consequences of anchoring one end of a PEO oligomer to a rigid backbone. Ethylene oxide segments directly at- tached to the polymer have limited ability to coordinate effectively with ions due to the 79 inflexibility of PPP. Once the chains are sufficiently long, the effect of the PPP backbone is ameliorated and dissolution of LiClO4 is facile. We previously observed a similar ef- fect in (AB)n multiblock copolymers, where the A block is an exact length EO segment and the B block is an alkylene chain.65 The ionic conductivity in this system had low conductivities when the A block length was short, but it rapidly increased as the length of A increased. The dependence on A was traced to the need for two segments to solubilize each LiClO4 in short A blocks, but in polymers with longer A blocks, one segment was sufficient. 3.2.3 AC impedance spectrscopy Since the solubility of the lithium perchlorate salt in the PPP-BOm polymers increases with an increase of side chain length, it is clear that the conductivities of these solid polymer electrolytes should follow in similar fashion. Figure 30 shows the conductivity data for the composites, plotted as 0 vs. lOOO/(T-To) where T0 was set at -60 °C, the T8 of PEO. At 30 °C, the PEO and PPP-E03 composites both have poor conductivities, ~10"6 S/cm. The conductivity of PPP-E03 is low over the entire temperature range as a result of the poor solubility of LiClO4 in the polymer, while the low conductivity of the PEO complex reflects its crystalline nature. The conductivities of the other composites follow the trends predicted by the XRD data. The increases in conductivity seen for PPP-E07 and PPP-EO<16> parallel the increasing solubility of LiClO4 in the composites, while the appearance of diffraction peaks for PEO in PPP-EO<45> is consistent with a slight drop in conductivity relative to PPP-EO<16>. A slice through the conductivity data at 30 °C (Figure 31) shows a maximum conductivity of 3><10'5 S/cm for PPP-EO<16>. The activa- 8O tion energies for ion transport extracted from the conductivity data were 0.52 (PEO), 0.49 (PPP-EO<45>), 0.64 (PPP-EO<16>), 0.84 (PPP-E07) and 0.93 KJ/mol (PPP-E03 The conductivity of the PPP-E0m system currently is limited by crystallization of the side chains at m > 16. The most effective strategy for realizing further increases in conductiv- ity is to introduce defects or branching in the E0 side chains to inhibit crystallization. These data for PPP-E0m polymers can be compared to structurally related polymer elec- trolytes. Kim et a1. investigated the effects of PEO bock length and phenyl unit structure in alternating block copolymers of PEO with various length arylenes. In their system, E0 lengths of m<12 gave room temperature conductivities on the order of 10'8 to 10'6 S/cm.180 Polymers that combine the same structural elements in a comb architecture have higher conductivities due to the increased flexibility of the E0 teeth. Wegner et a1. re- ported that electrolytes prepared from PPPs substituted with mixtures of hexa- and penta(ethylene glycol) monomethyl ethers had room temperature conductivities of ~10'5 S/cm.92 Replacing the PPP backbone with a more flexible poly(methyl methacrylate) (PEGEEM), increases the room temperature conductivities to ~10'4 S/cm.'8"‘83 Going to an even more flexible PEO-like backbone with methoxy diethylene glycol teeth produced room temperature conductivities of ~3 x 10“1 S/cm.'84"86 81 v2 PEO VA m=<45> .6? 3 “A m=<16> 8 E E =7 E 0 a: M m=3 h l LiCIO4 0 10 20 30 40 50 20 Figure 29. X-ray powder diffraction patterns of PPP-EOm/LiClO4 composites. The data were collected at room temperature at a rate of 0.05 26)/min over a range of 20 = 1— 50°. 82 Temperature (°C) 80 70 60 50 40 30 1 e-3 E l I I i ‘r I = o . 9 a . . I . 16-4 E . U ' 6 7 : o I o u . . A O E 16-5 E . ‘ 2 ; V v b 1e-6 g A PEG-dm-ZOOO v i : I PPP'EO<45> V 5 O PPP‘EO<16> ‘ 19'7 F PPP-E0 V ‘= E ' 7 i * v PPP-E03 l 1e-8 ‘ P ‘ ‘ ' 6 7 8 9 1O 1 1 1 2 1000/(pro) (K) Figure 30. Temperature-dependent conductivities for selected PPP-EOm/LiClO4 com- posites. The data are derived from AC impedance data, with all samples having an OzLi ratio of 20. 83 104*E I o . o 10'5§ g o \ 6 £2. 10 b . o 107; . 10-8 AI%A-4+L#AALAAA#1+ 0.4 0.6 0.8 1.0 Weight Fraction of Ethylene Oxide Figure 31. A cross-section of the conductivity data at 30 °C, indicating a conductivity maximum at m = <16>. Decreases in o for m = <45> and PEO-dm-2000 reflect partial crystallization of the samples. 3.3 Conclusions Solid polymer electrolytes based on PPP substituted with oligo(ethylene oxide) side chains and lithium perchlorate exhibit conductivities ranging from 10'6 S/cm to 104 S/cm at 30 °C. The conductivity is dependent on the length of the E0 chain attached to each ring. When the E0 chain is short, the solubility of LiClO4 is low leading to undissolved salt and low conductivities. Lengthening the E0 chains increases the solubility of LiClO4 and chain mobility, causing a more than two order of magnitude increase in the room temperature conductivity when m = <16>. For longer side chains, crystallization begins 84 to suppress the ionic conductivity, an effect that can be minimize by using branched chains that inhibit crystallization. 3.4 Experimental Synthesis and characterization of the PPP-E0m polymers are reported elsewhere.187 Sol- vents used in the preparation of the composite electrolytes were vacuum distilled from molecular sieves and stored in a helium filled dry box. Poly(ethylene oxide) dimethyl ether (PEO-dm-ZOOO, ca. 2000 g/mol) was obtained from Aldrich and was dried by azeotropic distillation of benzene, followed by removal of the solvent in vacuo. All elec- trolytes were prepared in a helium filled dry box by mixing acetonitrile solutions of Li- C104 and the desired polymer. The samples were then concentrated under reduced pres- sure to afford a viscous solution which was directly cast into the sample holder used for impedance analysis. All composites had an OzLi ratio of 20. Only the oxygen atoms of the polyether chain were included in calculating the OzLi ratio; the carbonyl oxygen was not counted. AC impedance spectroscopy was conducted over a frequency range of 5 Hz to 13 MHz using an HP 4192A LF Impedance Analyzer, with an applied voltage of 10 mV. Conduc- tivity measurements were taken at 30, 40, 50, 60, 70 and 80 °C. Prior to each measure- ment, samples were held at temperature for at least 20 minutes under a flow of nitrogen. The sample cell was constructed of stainless steel disks separated by a Teflon collar. The sample films had a thickness of 0.053 cm and an area of 1.27 cmz. Differential Scanning Calorimetry (DSC) data were obtained under He using a Perkin-Elmer DSC 7 instrument 85 at a heating rate of 10 °C/min, with the temperature calibrated with an indium standard. Reported DSC results are second heating scans, taken afier flash cooling the samples from an isotropic melt. Modulated Differential Scanning Calorimetry (MDSC) data were obtained under N2 using a TA Instruments Q1000 fitted with a refrigerated cooling sys- tem. Samples were scanned from -70 to 200 °C at a scanning rate of 1 °C/min. X-ray powder diffraction patterns were obtained using a computer controlled Rigaku 2008 ro- tating anode diffractometer operating in reflective mode at 45 kV/ 100 mA, with graphite monochromatized Cu (1(a) radiation. 86 Chapter 4 4 Self-assembled Silica N ano-composite Polymer Electrolytes 4.1 Introduction Solid polymer electrolytes based on poly(ethylene oxide) (PEO) and lithium salts have been extensively studied due to their potential application in secondary lithium ion batter- ies.38’4°’80 Most electrolytes are prepared by dissolving a lithium salt such as LiPF6 or Li- am in PEO or a structural analog of PEO, however, the performance of such systems suffer from low lithium ion transference numbers (tL,-+) and cell polarization. The reason for the low tL,-+ values is due to coordination of lithium cations to the oxygen atoms of polyethers, and thus the ionic mobility becomes dominated by the more mobile anions.2‘80 Bruce and co-workers estimated that in such systems the tL,-+ can be as low as 0.1.188 A recent investigation by Scrosati and co-workers reported a tlf of 0.28 using a lithium di- cyanotriazolate salt, a typical value for PEO/salt systems.189 Higher transference numbers for Li+ may be attained by reducing the mobility of anions. Several strategies have been investigated, including tethering anions to the polymer host in polymer electrolytes, in- creasing the size of the anion, and addition of Lewis acids, usually in the form of ceramic particles, that complex with anions. In cases where the anions are fully immobilized, los- ing the contribution of anions in binary salt electrolytes depresses the conductivity to ~1/10‘h of its original conductivity. 106408 87 Early investigations of composite polymer electrolytes demonstrated the advantages of incorporating inorganic fillers into a PEO host matrix.81‘82’190 When the particle size of such fillers was small ( < 5 pm) they suppressed crystallization and substantially im- proved the conductivity.82 Incorporation of nanoparticles also were shown to enhance the mechanical and electrochemical properties of electrolytesm“92 For example, the addi- tion of silica nanoparticles to poly(ethylene glycol) dimethyl ether (PEG-dm)/LiClO4 sys- tems yielded electrolytes with conductivities of ~10'3 S/cm at 25 °C, and elastic moduli of > 106 dynes/cmz.122 However, the applications of particles described above do not re- solve the problems associated with low M values in polymer electrolytes. More recently, negatively charged inorganic fillers such as montrnorillonites and other silica-aluminates have been used as anions.‘°8"93"94 Since the anions are massive, their mobilities are minimal and tL,-+ was measured to be ~0.9. Another particle-based approach to immobili- zation of anions would be to tether them to the surface of nanoparticles. When incorpo- rated in a PEO matrix, such particles may serve multiple purposes, including providing 83,122 mechanical stability, improving the stability of the electrode-electrolyte inter- 195,196 face, and immobilization of the anions allowing the lithium cation to be the only mobile species. In this report, we explore the tethering of lithium trifluoromethane sulfonimide to fumed silica and the properties of electrolytes prepared by dispersing the silica in a low molecu- lar weight polyether. Our goal is to take advantage of the high surface area of fumed sil- ica to immobilize sufficient numbers of anions to support single ion conductivity by cations. 88 4.2 Results and Discussion 4.2.1 Synthesis The ease in which perfluoroalkyl sulfonimides can be synthesized'97 along with the good conductivity and cycling characteristics198 of lithium perfluoroalkyl sulfonimides identi- fies them as attractive candidates for anchoring to surfaces. Fumed silica is a suitable nanoparticle support because it has a high surface area (200 mz/g), and the silanol groups that decorate its surface can be readily functionalized in high yield by reaction with chloro- or alkoxysilanes. The synthetic sequence shown in Scheme I was used to anchor the salts to fumed silica. A five carbon tether simplifies the synthesis and places the salt a reasonable distance from the surface. The synthesis involves standard chemical transfor- mations, a Gabriel synthesis to generate the pentenyl amine, formation of the trifluoro- methylsulfonimide using triflic anhydride, and hydrosilylation of the resulting alkene with triethoxysilane using Karstedt’s catalyst. Each intermediate product was character- ized by standard spectroscopic methods. 89 Scheme 111. Synthesis of A200-C5NTfLi WOH i,ii.iii,iv : WNHZ v ,Tf (510)33qu vi H . WNfl'f H A ..... Li+ 83 vn.vm 8;Si\/\/\/N-‘SIS3 OH O 6’ CF3 A200 A200-C5NTfLi i) Eth, CHC13, p-toluene sulfonic acid, 98%; ii) potassium phthalimide, DMF, 80%; iii) hydrazine hydrate, EtOH, 60°C; iv) HCl, EtOH, reflux, 52%; v) (CF3SOz)2O, CHzClz, Eth, 58%; vi) (EtO)3SiH, Karstedt’s catalyst, benzene; vii) EtzNH, A; viii) BuLi, tolu- ene. 4.2.2 Infrared spectroscopy The imide was attached to the silica surface using diethylamine as a catalyst. DRIF TS IR spectra of the modified silica demonstrate successful anchoring of the sulfonimide (Figure 32). The IR spectrum of A200 exhibits a sharp peak at 3750 cm'1 corresponding to the free Si-OH stretching. Near-complete loss of the O-H band and the appearance of bands for the imide at 2800-3000 cm'1 (CH stretching) and at 1443 and 1369 cm'1 (sul- 90 fonyl asymmetric and symmetric stretching) confirm attachment of the imide. The ex- pected CF3 band was obscured by intense bands from the fumed silica. We also were un- able to observe the N-H stretch of the sulfonimide, a problem previously reported by oth- ers for similar systems. Jezorek et a1. characterized alkyl amines bound to silica surfaces and they found it difficult to observe the amine stretch.199 Only after the amine was chemically modified did evidence of the amine become apparent. In a similar manner, the existence of the amine is supported by the appearance of a broad peak at 1470 cm'1 after lithiation, which is attributed to the SO; asymmetric and symmetric stretches of the sulfonimide saltzoo’201 Conditions for deprotonation of the imide were developed by re- acting the N-pentenylsulfonimide with butyllithium in toluene. As shown in Figure 33, the N-H stretch at 3250 cm‘1 is lost upon reaction with methyllithium under these condi- tions. The ease in which the salt is formed assures the formation of the bound sulfonimide salt. 91 A200-C5NTfLi A200-C5NHTf A200 L 2000 i 1500 Wavenumbers (cm'1r 4000 A 3500 A 3000 i 2500 Figure 32. Diffuse reflectance IR spectra of A200 silica nanoparticles, the sulfonim- ide-modified silica (A200-C5NHTt), and its lithiated form (A200-C5NTfl.i). Data were obtained from samples prepared in a 2:1 ratio of sample to dry KBr. A C5NTfLi Relative Intensity % C5NHTf 4000 3000 2000 1000 Wavenumbers (cm‘) Figure 33. IR spectra comparing the starting material (C5NHTf) and formation of the lithium salt (C5NTfLi). Data were collected as thin films sandwiched between NaCl plates. 4.2.3 Thermogravimetric analysis The surface coverage of silanol groups on A200 is ~l.0 mmol/g. 202 Assuming the alkyl sulfonimide is attached to the silica through each of the alkoxy groups, complete cover- age of the nanoparticles with 0.33 mmol of imide (81 mg), would give modified silica with 7.9% of the mass as the bound imide. The extent of surface coverage can be probed by thermogravimetric analysis (TGA) experiments run in air. Assuming all decomposi- 93 tion products of the imide are volatile, the resulting weight loss should be at most 7.9%. The measured weight loss for the functionalized silica was 2.6 % corresponding to 36 % surface coverage, or 0.12 mmol/g silica (Figure 34) If each sulfonimide is converted to the corresponding lithium salt upon reaction with butyl lithium, then the maximum lith- ium carrier concentration in an electrolyte would be 0.12 mmol/g of silica. 100.0 99.0 ~ A200 33 g 98.0 - 2 a 4: .9 0 97.0 - 3 2.6 % weight loss as 36 % coverage 96.0 - A200-CSNHTf 95.0 I #4 1 l i l . 1 n l J l 1 l 1 l 0 100 200 300 400 500 600 700 800 Temperature (°C) Figure 34. Thermogravimetric analysis of A200 and A200-C5NHTf measured in air from 30 °C to 800 °C at a rate of 10 °C/min. The weight loss at 700 oC corresponds to 36 % coverage of the silica nanoparticle surface. 94 4.2.4 AC impedance spectroscopy Polymer composites of A200-C5NTfl.i and poly(ethylene oxide) dimethyl ether were prepared by mixing an acetonitrile solution of the polymer with a dispersion of silica in acetonitrile, removing the solvent under reduced pressure, and mechanically mixing the composite to obtain a paste. The weight fraction reported was based on the total mass of the composite, as opposed to the mass of the polymer, the protocol commonly used for polymer electrolytes. Composites with low silica loadings (10 — 25 wt%) are pastes while higher loadings (30 — 50 wt%) are powders that become pastes upon shearing. The con- ductivity data for the samples appear in Table 9 and in Figure 35. The conductivities at 30 °C for the 19, 24, 30 and 35 wt% composites of were ~104S S/cm. The conductivities of the 10, 15, and 50 wt% composites could not be determined due to the low concentra- tion of charge carriers (Li+) in the case of the first two samples. For samples > 45 wt%, the low conductivity stems from poor connectivity, i.e. the volume fraction of PEG-dru- 500 is too low to be continuous throughout the electrolyte. The conductivity of 40 wt% composites could only be determined at 80 °C and 70 °C due to the high resistance of the material, exceeding the operational limits of the impedance analyzer, again likely due to poor connectivity. Similar effects have been observed with R805/PEG—dm electrolytes (R805 is an octyl modified hydrophobic fumed silica) at comparable silica loadings.8(”122 Data for the R805/LiClO4/PEG-dm-500 system are shown in Figure 35. As expected, the conductivities of the A200-C5NTfLi composites were more than an order of magni- tude difference lower than the binary salt system. 95 Lithium concentrations in polyether electrolytes are ofien expressed as the ratio of ether oxygens to lithium ions. For electrolytes prepared from binary salts, the conductivity usu- ally increases with decreases in the OzLi ratio, reaching a maximum at ~20. For the cur- rent system, the OzLi ratios reported in Table 9 are an order of magnitude lower than the optimum OzLi ratio of 20. Shriver, et a1. previously reported that for an OzLi ratio of 26, the ionic conductivity of polysiloxane — trifluoromethyl sulfonamide polyelectrolytes was 1.3 x 10’6 S/cm at 25 °C.203 Similarly Tominaga, et al. reported conductivities of ~10'5 S/cm at room temperature in a poly(propylene oxide) based sulfonimide. Given that the OD ratio in the PEG-dm-SOO/AZOO-CSNTflJ composites exceeds 200, increases in car- rier concentration should substantially improve the conductivity, and thus we view the ionic conductivities observed for these samples as encouraging. One concern with the measured conductivities is a potential contribution to the conduc- tivity from residual silanols on the surface of the silica particle. These groups also should be lithiated under the conditions used to deprotonate the sulfonimide. To test for conduc- tivity due to the lithiated silanols, A200 was treated with butyllithium using the same pro- tocol for the synthesis of A200-C5NTfl.i. A 30 wt% composite prepared from the lithi- ated silica and PEG-dm-SOO was analyzed by impedance spectroscopy. The conductivity of the composite was < 10'8 S/cm, the detection limit of the AC impedance analyzer. Thus we conclude that the contribution from lithiated silanols can be neglected. 96 Sample (wt%) O:Ll ratio a (S/cm) 30 °C a (S/cm) 80 °C 19 710 1.2711106 2.49x10" 24 530 8.35x10'7 1.36:110'6 30 390 1.20x10’6 3.02x10'6 35 310 5.28x10'7 1.73x10’° 40 250 - 4.48x10'8 Table 9. Lithium concentrations of various composite polymer electrolytes and their measured ionic conductivities at 30 and 80 °C. Temperature (°C) 103 80 7O 60 50 40 30 10" 0 0 0 o o . 3 l 10'5 '5 l O 5 10'6 B ' ' V y 9 1. a, v a ‘0' -7 o R805/LiClO4/PEO 10 O 30 wt% j V V I 24 wt% 1 1 0.3 1 19 wt% .1 v 35 wt% 1 9 V 40 wt% 2 10' 7 l 1 1 L r . r 1 r . r 2.8 2.9 3.0 3.1 1 3.2 3.3 1000/T (K' ) Figure 35. Conductivity of PEG-dm-SOO/A200-C5NTflJ composites. The data are derived from AC impedance measurements taken after equilibration for 20 min. 97 : l r I. L . . .5 . 10 r A I E5 I 0 C \ a) r V b 10Jr : 1. T 108 . 1 . 1 A L . 1 1 r 15 20 25 3O 35 4O 45 A200-C5NTfLi weight fraction (%) Figure 36. Cross section of conductivity at 40 °C. The data are obtained from the Ar- rhenius conductivity plot. 4.2.5 Transference number measurement To support the claim that these composites are indeed single ion solid polymer electro- lytes, lithium ion transference numbers were measured according to the method described by Bruce and Vincent.23 The experiment involves the measurement of the bulk resistance of the material prior and after DC polarization. This provides the initial and the steady state interfacial resistance. The DC polarization step is conducted by application of a small bias, typically less than 10 mV, and the current decay to a steady state is monitored over time. This provides the initial and steady state current. The lithium ion transference number can then be determined using equation (4.1) 98 t _ I,(AV-1.R.) U‘ — I.(AV—ISRS) (4'1) where I, and 10 are the steady state and initial currents, and R, and R0 are the steady state and initial interfacial resistances respectively. According to Bruce and Vincent, the interfacial resistances of an ideal electrolyte system can be ignored if the concentration of the charge carriers is small enough to create a lin- ear concentration gradient between the electrodes. In addition the kinetics at the elec- trodes must be fast and the applied bias <10 mV. Since the lithium concentrations in the PEGDMESOO/AZOO-CSNTflJ composites are approximately one tenth of a typical elec- trolyte, we approximate this system as an ideal system, simplifying equation (4.1) to I S tLi‘ I (4.2) Using equation (4.2), tL,-+ was determined as described in the literature108 and was found to be 0.86 d: 0.03. In comparison, siloxyaluminate polymers containing ethylene oxide side chains exhibited tu+ of 0.71.121 Whereas, the tu+ for PEO-LiTFSI was found to be 0.60 :1: 0.03 for OzLi = 5.204 4.2.6 Differential scanning calorimetry Finally, the addition of fumed silica to PEO electrolytes is known to improve a variety of physical and electrochemical characteristics of electrolytes, including increasing their modulus and improving interfacial stability. We used DSC scans of the composites to 99 screen for particle-matrix interactions. In general, dissolution of salts in polyethers sup- presses crystallinity.'9"73’178 For the A200-C5Ntfl.i composites, the charged surface of the nanoparticles suggest that their interaction with the polymer matrix will be similar to that of salts in polyethers. Shown in Figure 37 are the DSC scans of 0, 10, 20 and 40 wt% PEG-dm-SOO/AZOO-CSNTflJ composites. The effects of the modified silica on the properties of PEG-dm-SOO were minor. Only small decreases in the melting point and AHqu temperature are seen over the entire composition range tested (Figure 38). I1 mW/mg PEG-dme 500 /\ 10 wt% /\ 20 Wt% fi/\ 40 wt% 1 n l 1 l -20 0 20 4O 60 Temperature (°C) Heat Flow (mW/mg) Figure 37. DSC scans of PEG-dm-500, 10, 20, and 40 wt% of PEG-dm-SOO/AZOO- C5NTfLi composite polymer electrolytes. Data were collected under He at a rate of 10 °C/min. 100 120 - O 110 - '5 O a 100* 3 :5 90» O 80 r 70)— . 0 10 20 30 40 A200-CSNTle weight fraction (%) Figure 38. Comparison of the heat of fusion for the various composite polymer elec— trolytes as a function of weight percentage of A200-C5NTfLi. The data were obtained by integrating the areas under the curve for the melting transition for each DSC scan. 4.3 Conclusion Lithium sulfonimide salts were successfully immobilized on the surface of nanoparticu- late fumed silica. Dispersion of the particles in PEG-dm-SOO gave polymer electrolytes with near unity lithium ion transference numbers. Composites containing 19-35 wt% sil- ica had conductivities of ~10'6 S/cm. The ionic conductivity of each complex was similar and exhibited a weak dependence on temperature. The conductivities of the 10, 15 and 50 wt% composites were <10'8 S/cm, presumably due to insufficient carriers in the case of 101 the first two composites, and discontinuity in the PEG-dm-SOO polymer phase for the later composite. Although the lithium concentrations are too low to be of immediate practical use, improved conductivities may be obtained by grafting branched polymers to the surface of the silica nanoparticles where each branch is terminated with a lithium sul- fonimide salt, thus increasing the lithium loading per nanoparticle. 4.4 Future work Grafting branched polymers onto the silica nanoparticles can increase the ionic conduc- tivity by increasing the lithium concentration within the composite polymer electrolytes. The termini of each branch would contain a lithium salt such as the lithium sulfonimide salt already used. The amount of lithium salt potentially delivered by each particle can be controlled by polymerizing a styrene derivative such as 4-vinylbenzoyl trifluoromethane- sulfonamide through atom transfer radical polymerization (ATRP). ATRP is a well es- tablished and widely used method to control the degree of polymerization in olefins. By controlling the reaction time and conditions one can optimize the lithium concentration per nanoparticle. Shown in Scheme IV is the proposed synthetic route by which this can be accomplished. 102 Scheme IV. per particle. Proposed synthetic scheme for increasing the concentration of lithium ions ,sozcr3 / C1 + O \S/ I 0/ Cl MeLr, Toluene 103 NH2802CF3. Et3N. Ergo 0 N,soch3 I CuClz, dNpry 4.5 Experimental 5-Bromopentene, triflouromethane sulfonyl anhydride, potassium phthalimide, butyllith- ium (1.6M in hexanes) (Aldrich) and Karstedt’s catalyst (Gelest) were used as received. Aerosil 200 fumed silica, a gift from Degussa, was dried at 100 °C for 2 hours under vac- uum before use. Triethyl amine was distilled over KOH pellets. Diethyl ether was dried by stirring over KOH pellets for 1 hr, and then was filtered three times through a pad of dried silica gel (Davisil, grade 633, 100-425 mesh (Aldrich). Toluene was distilled from sodium benzophenone ketyl. Poly(ethylene glycol) dimethyl ether (PEG-dm-SOO, ca. 500 g/mol) (Aldrich) was dissolved in dry diethyl ether, stirred over KOH pellets, and filtered four times through activated basic alumina. Diethyl ether, toluene and PEG-dm-SOO were degassed immediately after purification and transferred into a helium filled drybox. IH NMR spectra were measured using a Varian Gemini-300 spectrometer at 300 MHz. All samples were run at room temperature in CDC13. Chemical shifts were calibrated us- ing residual CHC13 and are reported in ppm (5) relative to tetramethylsilane. 19F NMR spectra were measured in CDCl3 using a Varian Inova-300 spectrometer at 282.2 MHz. The chemical shifts are reported in ppm (5) relative to or,or,or-trifluoromethylbenzene. 1n- frared spectra of 1 and its lithium salt were obtained with a Mattson Galaxy Series FTIR 3000 as thin films sandwiched between NaCl plates. Diffuse reflectance infrared spectra of modified silicas were obtained from a computer-controlled Nicolet Prote’gé 460 equipped with a DRIFTS (diffuse reflectance infrared Fourier transform spectroscopy) Auxiliary experiment module. The samples were powders mounted on sample stubs and were mixed with potassium bromide in a 2:1 ratio of sample to salt. 104 All manipulations of the polymers and electrolytes were carried out in a helium drybox. AC impedance data were obtained from an HP 4192A LF Impedance Analyzer scanning from 5H2 to l3MHz with an applied voltage of 10 mV that was controlled by an in-house designed LabView application. All impedance samples have a thickness of 0.053 cm and an area of 1.27 cmz. Data were taken at 30, 40, 50, 60, 70 and 80 °C, with the samples equilibrated at each temperature for at least twenty minutes prior to measurement. The sample cell was constructed of stainless steel disks separated by a Teflon collar. Differ- ential scanning calorimetry (DSC) measurements were run under helium at a heating rate of 10 °C/min using a Perkin Elmer DSC 7 calibrated with indium. The reported DSC curves are second heating scans taken after an initial heating scan to erase the thermal history, and a fast quench to —100 °C. Thermogravimetric analyses (TGA) were run in air at a heating rate of 10 °C/min on a Perkin-Elmer TGA 7 instrument. Synthesis of N-pentenyl triflouromethane sulfonimide, C5NHTf (1). Trifluoro- methane sulfonyl anhydride (6.49 g, 23 mmol) was added dropwise to a vigorously stirred solution of N-pentenyl amine205 (2.00 g, 23 mmol), triethyl amine (2.56 g, 25 mmol) and freshly distilled dichloromethane (25 mL) cooled to -70 °C. After the addi- tion was complete, the reaction mixture was allowed to warm to room temperature and stir for 2 hrs under nitrogen. The solvent was removed in vacuo to afford a yellow oil that was redissolved in 4N NaOH and washed with dichloromethane (2 x 50 mL). The product was then neutralized with 4N HCl and extracted with dichloromethane (8 x 50 mL). The combined organic layers were dried over MgSO4, filtered, and the solvent re- moved under reduced pressure to afford a light yellow oil (2.5 g, 50 % yield) with spec- 105 tral data that match those reported in the literature.206 1H NMR 5 1.7 (q, 2H), 2.1 (q, 2H), 3.25 (q, 2H), 4.9 (br, 1H), 5.1 (m, 2H), 5.8 (m, 1H). Synthesis of triethoxysilane, N-pentane trifluoromethane sulfonimide (2). Triethox- ysilane (0.76 g, 4.6 mmol), compound 1 (1.0 g, 4.6 mmol) and Karstedt’s catalyst (30 uL) were dissolved in 1 mL of benzene and stirred in air overnight. The volatiles were removed under vacuum, and the residue was dissolved in diethyl ether and stirred over decolorizing carbon. Following filtration through a glass microfiber filter, the solvent was removed under vacuum to afford a mixture of compound 2, and divinyl silane (cata- lyst) as a viscous yellow oil. The crude product was used without further purification. 1H NMR 5 0.5 (t 2H), 1.1 (t, 9H), 1.4 (m, 4H), 1.6 (q, 2H), 3.2 (q, 2H), 3.8 (q, 6H). ”P NMR (282.2 MHz, C6D6 5) -724. Modification of A200 fumed silica, A200-C5NHTf (3). Diethyl amine (44 uL, 0.43 mol) and A200 fumed silica were stirred in 50 mL of toluene for 30 minutes under nitro- gen. Compound 2 was then added to the mixture and stirring was continued at room temperature overnight. The solution was filtered and the residue was washed alternately with toluene (3 x 100 mL) and diethyl ether (3 x 100 mL) then finally rinsed with diethyl ether (6 x 100 mL). The residue (4.05 g) was further dried under vacuum at 50-60 °C overnight. IR 3263, 2989, 2932, 2893, 1989, 1880, 1688, 1447, 1376 cm". Lithiation of A200-C5NHTf, A200-C5NTfLi (4). Butyllithium (2 mL, 1.6 M in hex- anes) was added dropwise to a dispersion of 3 (1.0 g) in 50 mL of freshly distilled tolu- 106 ene. After stirring overnight at ambient temperature, the solution was filtered under an inert atmosphere and the residue was washed alternately with toluene (3 x 50 mL) and diethyl ether (3 x 50 mL). The residue (0.74 g) was further dried under vacuum at 50-60 °C overnight. IR 3681, 2967, 2927, 2876 cm". Preparation of PEO/A200-C5NTfLi composites — general procedure. All electrolytes were prepared in a helium filled dry box. A specific example, the preparation of a 30 wt% composite, is described below. The modified silica nanoparticles (0.0162 g) were com- bined with a 1 mL solution of PEG-dm-SOO (0.0385 g) in acetonitrile. After mixing, the sample was concentrated under reduced pressure to afford a dry powder that was further dried under vacuum at 56 °C. Upon shearing the electrolyte samples became viscous pastes that were then loaded into the sample cell. 107 Appendices 108 5 Appendix A AC impedance spectroscopy LabView VI Introduction LabView is a program from National Instruments that allows a user to design a custom program to interface with a device to send and collect information as well as automate procedures. Examples of such programs can be found in pilot plants. These programs are referred to as virtual instruments (VI), of which there are two com- ponents. The first is the user interface, referred to as the front panel, and the second is the programming, referred to as the wire diagram. Functions for communications, loops, boo- lean statements and the like are placed in this diagram and connections that symbolize data flow are drawn in as wires between these components. LabView offers a myriad of functions from the basic while loop statements to advanced firnctions capable of image analysis. In the case of AC impedance spectroscopy the device used to collect the data is an HP 4192A LF analyzer capable of measuring the complex impedance and the phase angle between the applied bias and the measured current through the sample. To ease data col- lection and automate the process, a program was designed to interface with the HP 4192A and collect data at frequencies specified by the user. 109 Description This is program is designed to interface with the HP 4192A LF Analyzer for impedance spectroscopy. It will initiate communications, setup the instrument to read the complex impedance and the phase angle. Data from the instrument is collected and stored in a user specified file after each measurement point and displayed at the same time. Algorithm Start a while loop to keep the application going. 0 End when the boolean to exit is switched to true. 0 Continue while the boolean to exit is false. As long as the while loop is running, the program is in a ready state to acquire data. 0 If boolean is false, stay in ready state 0 If boolean is true continue. Initiate communications with the HP 4192A through the GPIB interface. Send experiment parameters to the device. These are the applied bias and the data to collect. In this case the data are the complex impedance and the phase angle be- tween the applied bias and the current. Collect the data from the instrument at each frequency, record the data to a file and display the data in a bode plot. The data collection ends after the impedance at the last frequency is collected. Return to ready state. 110 Implementation The algorithm is implemented by starting with three control structures, a while loop and two boolean statements. The while loop runs continuously as long as the boolean state- ment is set to false, if it becomes true then the program will end and close the application. Embedded in the boolean statement is another boolean to control data acquisition. When it is true, the embedded functions initiate communications with the instrument, sets up the instrument and begin the data acquisition. Upon completion the boolean is reset to false, in which the application is in ready mode. Front panel The applied bias, starting frequency, ending frequency and acquisition interval are all set by the user. The default settings are 0.01 V, 5.00E-3 KHz, 1.30E+3 KHz, and 1.00 s re- spectively. Two graphs on the right will display the data as it is collected from the HP 4192A LF Analyzer. Acquisition is commenced by pressing on the START button. Should the impedance of the material be greater than or smaller than the detection limits of the instrument, the INSTRUMENT FLOW light will turn red. The application is closed by pressing on the STOP button. After pressing the acquisition button, a dialogue will popup and ask the user for the file- name and its location, after which, it will commence with data collection. The data is saved in a comma separated format in which the first entry is the complex impedance, the second the phase angle and the third, the frequency all separated by a comma. Each line in the file is a different measurement at that frequency as shown in Table 10. 111 Line # Data: l l 80000.000000,-1 50000000005000 2 l 70000.000000,- 1 3 .000000,0.005 903 3 170000.000000,-1 l .000000,0.0073 79 4 170000000000,-10.000000,0.009223 5 l70000.000000,-8.000000,0.0l 1529 Table 10. An example of the saved data. The format is the complex impedance, phase angle, frequency. Data manipulation The user is encouraged to save the file as *.csv file in order to facilitate opening the file in Excel. The real and imaginary components of the complex impedance can then be cal- culated as follows: real _ ._ .75. Z Z —|Z|+cos(6180) n - 7T Zimg = Z = IZI —Sln(61—8—0) Due to the fact that Excel and SigmaPlot calculate cosine, sine and tangents in radians, 7t/180 is inserted to convert the phase angle from degrees to radians. A Nyquist plot can be created using 2’ as the x-axis and Z” as the y-axis. A macro was written for Sig- maPlot to automate the process. It may be accessed through the Toolbox menu and then impedance analysis. New dialogues will popup asking the user to input the files to ana— lyze. After pressing the OK button, the macro will proceed to analyze each file, create a new worksheet with the original and calculated data and a nyquist plot for each data set. The text of the macro is included in Appendix B. 112 6 Appendix B Impedance Analysis Macro for SigmaPlot Option Explicit Dim SelectedFiles$(),SaveFileS Dim i, Index, ReportIndex As Integer Function FlagOn(flag As Long) FlagOn == flag Ch: FLAG_SET_BIT ' Use tx> set option flag bits on, leaving others unchanged End Function Function FlagOff(f1ag As Long) FlagOff = flag Or FLAG_CLEAR_BIT ' Use to set option flag bits off, leaving others unchanged End Function Sub Main 'Macro edited 08/19/01 Fadi Asfour 'For the purpose of batch analysis of impedance data col- lected from HP 4192A LF Impedance Analyzer 'the format of the data is assumed to be in comma delimited form and the extension of the files 'are .csv 'Macro created 01/19/2000 John Kuo 'This macro is an example of automating a batch process. You can create a 'list of any number of Excel files, then import a specified block of data from 'each file. The data from each Excel file is imported into a separate 'worksheet. You can then either plot and/or curve fit the first two columns 'of imported data, and save the results to a specified notebook. 'Note that only the data for the first sheet is currently imported, and there 'is no way to specify a different sheet. This import prop- erty will be forthcoming |****‘k*********************‘k************ 113 '* Set default save file; edit this to * '* change the default save path * I*************************************** SaveFile = path + "\" + "BatchFile.jnb" 'Create Graph Type list Dim GraphTypes$() ReDim GraphTypes(l) GraphTypes(0)="Simple Scatter Plot" GraphTypes(1)="Simple Bar Chart" 'Initialize file list ReDim SelectedFiles$(O) SelectedFiles(0)=Empty 'Initialize file name list 'Dim FileNames$(0) ' FileNames(0)=Empty 'Dim DataDir As String ' DataDir = path + "\" 'Dim Equations$() 'Dim FitLibraryS ‘Defines the equation source. Edit to use a different fit library 'FitLibrary = "Standard.jfl" 'Open the fit library 'Notebooks.Open(path + "\" + FitLibrary, ".jfl") 'Dim FitFile As Object 'Set FitFile = Notebooks(path + "\" + FitLibrary) 'FitFile.Visible=False 'Populate equation list with all equation items in fit li— brary 'i=0 'Index = 0 'For i = 0 To FitFile.NotebookItems.Count - 1 ' If FitFile.NotebookItems(i).ItemType = 6 Then ' ReDim Preserve Equations$(Index) ' Equations(Index) = FitFile.NotebookItems(i).Name ' Index = Index + 1 'End If 'Next i i=-1 MacroDialog: 114 Begin Dialog ‘UserDialog' 470,546,"Batch Process Excel Files",.DialogFunc ' %GRID:10,7,1,1 ’browsebutton for data path 'PushButton 10,21,90,21,"Browse...",.BrowseButtonl 'TextBox 120,21,330,21,.DataPath CheckBox 50,63,140,14,"Sing1e—step mode",.stepmode PushButton 240,63,90,21,"Add &File...",.AddFile 'Click Add File to add excel files to the list PushButton 350,63,90,21,"Delete File",.DeleteFile Text lO,84,90,14,"&Data files:",.Text1 ListBox 10,105,450,126,SelectedFiles(),.Files OKButton 270,518,90,21 CancelButton 370,518,90,21 PushButton 10,518,90,21,"Help",.Help GroupBox 10,245,450,70,"Import Range",.Range Text 20,259,80,21,"&Start column",.Text2 TextBox 110,259,60,21,.startcol Text 200,259,80,14,"&End column",.Text4 TextBox 290,259,60,21,.endcol Text 20,287,90,l4,"Start &row",.Text3 TextBox 110,287,60,21,.startrow Text 200,287,60,l4,"End ro&w",.Text5 TextBox 290,287,60,21,.endrow GroupBox 10,329,450,105,"Process",.GroupBoxl CheckBox 20,350,80,14,"&Plot data",.PlotData Text 110,350,20,14,"&as:",.Text6 DropListBox 180,350,260,63,GraphTypes(),.GraphList 'CheckBox 20,287,110,14,"&Curve fit data",.FitData 'Text 136,287,90,l4,"&using:",.Text7 'DropListBox 180,284,260,154,Equations(),.FitList Text 20,399,420,28,"Note: columns 4 a; 5 will be created from imported data and will be plotted",.Text8 Text 10,462,120,14,"Sa&ve notebook to:",.Text9 TextBox 10,483,450,21,.SavePath PushButton 370,455,90,21,"Browse...",.BrowseButton End Dialog Dim dlg As UserDialog 'Default settings 'SaveFile dlg.SavePath = SaveFile I11'******************************************** 115 '* Change this setting to change the default * '* location of the source data block * 1********************************************* dlg.startcol = "l" dlg.startrow = "1" dlg.endcol = "3" dlg.endrow = "512" I******************************************** '* You can also change whether the data are * '* plotted or fitted by default * I***********************~k***************~k**** dlg.PlotData O 'dlg.FitData = 1 '********************************************* '* Sets the default Graph Type * '* 0=Simple Scatter Plot, l=Simple Bar Chart * I********************************************* dlg.GraphList = 0 1********************************************* '* Sets the default Fit Equation * '*****‘k‘k‘k************************************* 'dlg.FitList = 22 '4 parameter Logistic 'See the end of the file for a list of all built—in equa- tions by number 'These numbers only apply to the factory default Stan— dard.jfl library Select Case Dialogldlg) Case 0 'Handles Cancel button GoTo Finish End Select 'Error if no Excel files picked If SelectedFiles(0) = Empty Then MsgBox "You have not selected any Excel Files",vixclamation,"No Files Selected" GoTo MacroDialog End If Dim CurrentNotebook Set CurrentNotebook = Notebooks.Add 'Iterate through each selected Excel file 'You can change the extension to import files of different types Index = 0 ReportIndex = 0 116 For Index = 0 To UBound(SelectedFiles) CurrentNotebook.CurrentDataItem.Open CurrentNote- book.CurrentItem.Import(SelectedFiles(Index), 0, 0, CLng(dlg.startcol)-1, CLng(dlg.startrow)-l, CLng(dlg.endcol)-l, CLng(dlg.endrow)—1, ".CSV") 'Name the data sheet according to its file name CurrentNotebook.CurrentItem.Name = GetFile- Name(SelectedFiles(Index)) If dlg.stepmode = 1 Then MsgBox("The data is imported from the Excel Worksheet...",vbInformation,"SigmaPlot") 'Calculate ReIZI and Imngl Dim SPTransform As Object Set SPTransform = ActiveDocument.NotebookItems.Add(9) SPTransform.Open 'note must specify use trig functions in units of radians; data is in units of degrees '0 = Radians, 1: Degrees, 2 = Grads SPTransform.TrigUnit = 0 SPTransform.Text = "pi=3.l415926" + vbCrLf + _ "z=col(l)" + vbCrLf + _ "w=col(2)" + vbCrLf + _ "col(4)=z*COS(w*pi/180)" + vbCrLf + _ "col(5)=-z*SIN(w*pi/180)" + vbCrLf 'use + vbCrLf + _ to create a line break I***************************************************** '* Debug transform code; this opens the transform in * '* the transforms dialog for viewing and editing * I**************~k************************************** 'SPTransform.RunEditor SPTransform.Execute SPTransform.Close(False) 'Plot the graph If d1g.PlotData = 1 Then Dim SPPage Set SPPage = CurrentNotebook.NotebookItems.Add(2) 'Creates graph page Dim PlottedColumns(l) As Variant PlottedColumns(0) = 3 PlottedColumns(l) 4 117 SPPage.CreateWizardGraph("Scatter Plot", "Simple Scat- ter","XY Pair",PlottedColumns) 'select current graph as CurrentPageItem Dim GPage Set GPage = ActiveDocument.CurrentPageItem GPage.GraphPages(0).Graphs(0).SelectObject 'Sets the X-Axis title To "Re IZI" And the Y-Axis ti- tle To "Img IZI". GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SLA_SELECTDIM, l) GPage.GraphPages(0).CurrentPageObject(GPT_AXIS).NameOb ject.SetObjectCurrent GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_RTFNAME, "{\rtfl\ansi0{\colortbl\red0\green0\blue0;}\deff0{\fonttbl\ f0\fnil Arial;}l\31240\slmult0\f0\cf0\up0\fsl6\i0\b0\ulO\ql Re Ilel") GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SLA_SELECTDIM, 2) GPage.GraphPages(0).CurrentPageObject(GPT_AXIS).NameOb ject.SetObjectCurrent GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_RTFNAME, "{\rtfl\ansi0{\colortbl\red0\green0\blue0;}\deff0{\fonttbl\ f0\fnil Arial;}‘}") 'Set the tick orientation inward on all sides GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SELECTLINE, 2) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_TICSIZE, 50) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SEA_THICKNESS, 10) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SUBlOPTIONS, &H0001515C&) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SUBZOPTIONS, &H0000011c&) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SELECTLINE, 2) GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SLA_SELECTDIM, l) GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SLA_SELECTDIM, 1) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SELECTTIC, 1) 118 GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SELECTLINE, 2) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_TICSIZE, 50) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SEA_THICKNESS, 10) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SUBlOPTIONS, &H0001515C&) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SUBZOPTIONS, &H0000011c&) GPage.SetCurrentObjectAttribute(GPM_SETAXISATTR, SAA_SELECTLINE, 2) 'Set the symbols in the graph to small size 0.060in circles GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SSA_OPTIONS, &H00000201&) GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SSA_EDGETHICKNESS, 10) GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SSA_SIZE, 60) GPage.SetCurrentObjectAttribute(GPM_SETPLOTATTR, SSA_SIZEREPEAT, 2) 'Dim ColumnsPerPlot() 'ReDim ColumnsPerPlot(2, 1) 'ColumnsPerPlot(0, 0) = 0 'ColumnsPerPlot(l, 0) = 0 'ColumnsPerPlot(2, 0) = 31999999 'ColumnsPerPlot(O, 1) = 1 'ColumnsPerPlot(l, 1) = 0 'ColumnsPerPlot(2, 1) 31999999 'Dim PlotColumnCountArray() 'ReDim PlotColumnCountArray(O) 'PlotColumnCountArray(0) = 2 'Select Case dlg.GraphList ' Case 0 'Simple Scatter Plot ' SPPage.CreateWizardGraph("Scatter Plot", "Simple Scatter", "XY Pair", ColumnsPerPlot, PlotCol- umnCountArray) ' Case 1 'Simple Bar Chart ' SPPage.CreateWizardGraph("Vertical Bar Chart", "Simple lknflfl "XY Pair", ColumnsPerPlot, PlotCol- umnCountArray) ‘End Select 'SPPage.GraphPages(0).Graphs(0).Plots(0).SelectObject 'Curve needs to be selected in order to plot curve fit 119 SPPage.Open If’ dlg.stepmode == 1 TNKNI MsgBox("The data ii; plot- ted...",vbInformation,"SigmaPlot") End If 'Fit the data; modify the fit options to suit your needs 'On Error GoTo FitFailed 'If d1g.FitData = 1 Then 'Dim FitEquationS 'FitEquation = Equations(dlg.FitList) 'Dim FitObject As Object 'Set FitObject = FitFile.NotebookItems(FitEquation) 'FitObject.Open 'FitObject.DatasetType = CF_XYPAIR 'FitObject.Variable("x") = "col(l)" 'FitObject.Variable("y") = "col(2)" 'FitObject.Run 'FitObject.OutputReport = True 'FitObject.OutputEquation = False 'FitObject.ResidualsColumn = —1 'FitObject.PredictedColumn = —l 'FitObject.ParametersColumn = -1 'FitObject.OutputGraph = False 'FitObject.OutputAddPlot = True 'FitObject.ExtendFitToAxes = True 'FitObject.AddPlotGraphIndex = 0 'FitObject.XColumn = -1 'FitObject.YColumn = -l 'FitObject.ZColumn = -2 'FitObject.Finish 'If dlg.stepmode = 1 Then MsgBox("The data is curve fitted and a report is gener— ated...",vbInformation,"SigmaPlot") 'End If Wait 1 'Close the document windows CurrentNotebook.CurrentDataItem.Close(True) 'If dlg.FitData = 1 Then CurrentNote- book.NotebookItems("Report " + CStr(ReportIndex + 1)).Close(True) 'GoTo Skip 'GoTo Finish 'FitFailed: 'MsgBox("Error(s) have occurred in fitting your data.",vixclamation,"SigmaPlot") 120 'ReportIndex = ReportIndex - 1 'Skip: 'If dlg.PlotData = 1 Then CurrentNote- book.NotebookItems("Graph Page " + CStr(Index + 1)).Close(True) 'If dlg.stepmode == 1 Then MsgBox("The results windows are closed...",vbInformation,"SigmaPlot") 'Create a new worksheet for the next file If Index <> UBound(SelectedFiles) Then CurrentNotebook.NotebookItems.Add(1) If dlg.stepmode ==Il Then MsgBox("A worksheet is cre- ated for the next Excel work- sheet...",vbInformation,"SigmaPlot") End If ReportIndex = ReportIndex + 1 Next Index 'FitFile.Close(False) 'Close standard.jf1 'Save the file If dlg.stepmode == 1 Then MsgBox("You are prompted to save the file...",vbInformation,"SigmaPlot") CurrentNotebook.SaveAs(SaveFile) Finish: End Sub Rem See DialogFunc help topic for more information. Private Function. DialogFunc(DlgItem$, .Action%, SuppValue&) As Boolean Select Case Action% Case 1 ' Dialog box initialization DlgEnable ("DeleteFile",False) Case 2 ' Value changing or button pressed Select Case DlgItem$ Case "CancelButton" DlgEnd 1000 'Handles Cancel button from file dia- log Case "Help" Dim ObjectHelp, HelpID As Variant ObjectHelp = Path + "\SPW.CHM" HelpID = 70200 ' Help ID number for this topic in SPW.CHM Help(ObjectHelp,HelpID) DialogFunc = True 'do not exit the dialog Case "AddFile" 'Adds file to list Dim SelectedFileS 121 SelectedFile == GetFilePath (,"csv",,"Select comma delimited File")'You can change the extension to im- port files of different types If SelectedFile <> "" Then i=i+1 ReDim Preserve SelectedFiles$(i) SelectedFiles(i) = SelectedFile DlgListBoxArray "Files",SelectedFiles End If DialogFunc = True 'do not exit the dialog Case "DeleteFile" 'Removes files from list SelectedFiles(DlgValue("Files"))=Empty If DlgValue("Files") <1 UBound(SelectedFiles) Then 'Re—indexes array if index removed from middle of ar- ray For Index = DlgValue("Files") To UB- ound(SelectedFiles)-1 If SelectedFiles(Index)=Empty Then Selected— Files(Index)=SelectedFiles(Index+1) SelectedFiles(Index+l)=Empty End If Next Index End If If i >= 1 Then i=i-l If i <= 0 Then i=0 DlgListBoxArray "Files",SelectedFiles ReDim Preserve SelectedFiles$(i) If i = 0 Then i = -l DlgEnable("DeleteFile",False) DialogFunc = True 'do not exit the dialog Case "BrowseButton" 'Set save path SaveFile == GetFilePath(,"JNB",,"Select Note- book File",l) If SaveFile <> "" Then DlgText("SavePath",SaveFile) DialogFunc = True 'do not exit the dialog Case "Files" 'Enables Delete button if ea file is se— lected If DlgValue("Files") <> —1 Then DlgEnable("DeleteFile",True) End Select Case 4 Select Case DlgItemS 122 Case "Files" 'Enables Delete button if a file is selected If DlgValue("Files") <> -1 DlgEnable("DeleteFile",True) End Select Case 5 DialogFunc = True End Select End Function ’Author: Maria Rapini 'Source: http://www.freevbcode.com/ShowCode.Asp?ID=1638 'Modified 8/19/01 Fadi Asfour Public Function GetFileName(flname As String) As String 'Get the filename without the path or extension. 'Input Values: ' flname - path and filename of file. 'Return Value: ' GetFileName — name of file with extension. Dim posn As Integer, i As Integer Dim fName As String Then posn = 0 'find the position of the last "\" character in file- name For i = 1 To Len(flname) If (Mid(flname, i, l) = "\") Then posn = 1 Next i 'get filename without path fName = Right(flname, Len(flname) — posn) 'get filename without extension 'posn = InStr(fName, ".") ' If posn <> 0 Then ' fName = Left(fName, posn - l) ' End If GetFileName = fName End Function 'Standard.jfl equation list 'Use the following values to set the default equation num— ber: 'Polynomial ' 0' Linear 123 I lI I 2I I 3I I 4| I SI I 6I I 7I I 8I I 9' I lOI I lll I 12I I 13I I l4I I lSI I 16I I 17I 'Sigmoidal I l8I I 19' I 20I I 21I I 22I I 23I I 24I I 25! I 26I I 27! I 28I I 29I I 3OI 'Exponenti I 31I I 32I I 33I I 34I I 35I I 36I I 3’7I I 38I 'Exponenti I 39! I 4OI I 41I I 42I I 43I Quadratic Cubic Inverse First Order Inverse Second Order Inverse Second Order Gaussian, 3 Parameter Gaussian, 4 Parameter Modified Gaussian, 4 Parameter Modified Gaussian, 5 Parameter Lorentzian, 3 Parameter Lorentzian, 4 Parameter Pseudo-Voigt, 4 Parameter Pseudo-Voigt, 5 Parameter Log Normal, 3 Parameter Log Normal, 4 Parameter Weibull, 4 Parameter Weibull, 5 Parameter Sigmoid, 3 Parameter Sigmoid, 4 Parameter Sigmoid, 5 Parameter Logistic, 3 Parameter Logistic, 4 Parameter Weibull, 4 Parameter Weibull, 5 Parameter Gompertz, 3 Parameter Gompertz, 4 Parameter Hill, 3 Parameter Hill, 4 Parameter Chapman, 3 Parameter Chapman, 4 Parameter a1 Decay Single, 2 Parameter Single, 3 Parameter Double, 4 Parameter Double, 5 Parameter Triple, 6 Parameter Triple, 7 Parameter Modified Single, 3 Parameter Exponential Linear Combination al Rise To Maximum Single, 2 Parameter Single, 3 Parameter Double, 4 Parameter Double, 5 Parameter Simple Exponent, 2 Parameter 124 44' Simple Exponent, 3 Parameter 'Exponential Growth 'Hyperbola 45' 46' 47' 48' 49' 50' 51' 52' 53' 54' 55' 56' 57' 58' 59' 60' 61' 62' 63' 64' 65' Waveform 66' 67' 68' 69' 70' 71' 72' 73' 74' Power 75' 76' 77' 78' 79' 80' 81' 82' Rational 83' 84' 85' Single, 1 Parameter Single, 2 Parameter Single, 3 Parameter Double, 4 Parameter Double, 5 Parameter Modified Single, 1 Parameter Modified Single, 2 Parameter Stirling Model Simple Exponent, 2 Parameter Simple Exponent, 3 Parameter Modified Simple Exponent, 2 Parameter Single Rectangular, 2 Parameter Single Rectangular i, 3 Parameter Single Rectangular II, 3 Parameter Double Rectangular, 4 Parameter Double Rectangular, 5 Parameter Hyperbolic Decay, 2 Parameter Hyperbolic Decay, 3 Parameter Modified Hyperbola i Modified Hyperbola II Modified Hyperbola III Sine, 3 Parameter Sine, 4 Parameter Sine Squared, 3 Parameter Sine Squared, 4 Parameter Damped Sine, 4 Parameter Damped Sine, 5 Parameter Modified Sine Modified Sine Squared Modified Damped Sine 2 Parameter 3 Parameter Pareto Function Symmetric, 3 Parameter Symmetric, 4 Parameter 2 Parameter Modified i 2 Parameter Modified II Modified Pareto Function 1 Parameter i 1 Parameter II 2 Parameter i 125 I 86I I 8'7I I 88I I 89I I 90I I 91' I 92I I 93I I 94I I 95! I 96I I 97I I 98I 'Logarithm I 99I ' 100' ' 101' ' 102' ' 103' ' 104' ' 105' ' 106' ' 107' ' 108' Parameter II Parameter i Parameter II Parameter III Parameter IV Parameter Parameter Parameter Parameter Parameter 9 Parameter 10 Parameter ll Parameter ooqmmbwwwwrx) 2 Parameter i 2 Parameter II 2 Parameter III 3 Parameter 2nd Order 3rd Order Plane Paraboloid Gaussian Lorentzian 'User—Defined ' 109' Untitled 'Standard Curves ' 110' ' 111' Linear Curve Four Parameter Logistic Curve 126 7 Appendix C Transference number measurements based on Bruce and Vincent papers (J. Electroanal. Chem, 1987, 225, 1-17) (Solid State Ionics, 1992, 53-56, 1087-1094) A transference number is the fraction of current carried by a charged species relative to all mobile species in an electrolyte. To quantify transference numbers, it is necessary to measure the amount of current for each charged species. Bruce and Vincent provide a relatively simple technique to conduct such measurements. First, recall that Onsager’s equations relates the current flow of each species to the ap- propriate electrochemical potential gradient __o.. 427. _a.- cm. ledex _ _ (7.1) 0'__ dp_ o;_ d,u+ ' F dx F dx where (7++ , 0'___ = free cation & anion conductivities 0' _ = cation - anion interaction + F = F araday's constant )7, , fl = chemical potential This means that the current carried by a charged species is dependent on the chemical gradient across the sample and its interactions with other ions. 127 Second, it must be understood that an ideal electrolyte is one where 1) ion-ion interac- tions may be neglected, 2) charge transfer processes are effectively infinitely fast, 3) con- vection does not contribute to ion transport within the electrolyte and 4) the concentration of the electrolyte remains sufficiently high such that the double layers at the electrodes can be neglected. There are three states that need to be considered. The first is the initial state in which a bias has just been applied across the electrolyte. The second is at some intermediate state after t = 0 in which a concentration gradient begins to form. The third is when a steady state condition has been reached. Initial State When a DC potential of AV is applied across the electrolyte the potential drop (21¢) is A¢ = AV (7.2) and the initial current is given as [O = F(u: + u: )coAV = 023V (7.3) where u: and u: are the cationic and anionic mobilities respectively in an ideal electro- lyte. Since mobilities are a function of diffusion one can use the Nemst-Einstein rela- tionship to redefine the initial current as _ F2(Dj + D:)c° 1 0 RT AV (7.4) where D: and D: are the cationic and anionic diffusions in an ideal electrolyte respec- tively. The concentration of ions at this initial state is homogenously distributed and can 128 be represented by a plot of the concentration of an ion where c, is the concentration of species i and x is the distance between the electrodes. 1 l l 1 0O 0.2 0.4 0.6 0.8 l X Intermediate state At t > 0 cations arriving at the cathode are consumed while an equivalent number are pro- duced at the anode. This creates a salt concentration gradient in the vicinity of each elec- trode 0 0.2 0.4 0.6 0.8 1 X The change in concentration at the electrode surfaces causes each electrode to develop a potential difference, where 129 Cathode :> negative (low M+, high 6‘) Anode :> positive (high MI, low e") The sum of the potential developed is AE=£Zm[$J (73 F c which is the Nemst equation. This means that the potential drop across the electrolyte is A¢=AV—AE (7m Note that ion diffusion opposes the migration current and the migration currents of the anions and cations decreases with time as AE increases. Ion transport is now influenced by diffusion due to the presence of the salt concentration gradients. In polymer electro- lytes the diffusion layers will continue to grow until they merge and a steady state condi- tion is established. Steady state In the steady state condition the electrolyte has reached an equilibrium where o Electro-neutrality requires that 6+ = c_ = c 0 Ion migration and diffusion must be equal and opposite llfl=|IT| and (I 1. m! A; In this condition a chemical gradient has been established between the electrodes 130 1 l l 0 0.2 0.4 0.6 0.8 1 X This implies no further mobility of ions, which means that the electrochemical potential for the anions throughout the electrolyte is d;7_=Rlenc+Fd¢=0 (7.7) where the electrical and chemical potential gradients are related as etfldlnaflic. dx-F dx-chx (7'8) Recall that in the steady state the current due to the cations (1+) is L=H+U 0% Consider the linear mass transfer or flux (J) of species i —J,=D,$+fl£Dic,-€£ (7.10) dx RT dx where z,- is the charge of the species, which in most cases is l. D,- is the diffusion and c,- is the concentration of species i. Since the mass transfer of species i is based on diffusion and migration then the flux equation can be written as 1. I." 1;" + l _ l PEI-El FA (7.11) 131 where A is the area through which the ions flow through. For simplification set A = 1. The following equations are obtained by setting equation (7.10) equal to (7.11) I," = FD, 53" (7.12) dx 2 1;" = F 0' elm-(Q (7.13) RT dx Equation 7.8 can now be written as 2 o I, =FD:£+ F D“ ca (7.14) dx RT dx . . 61¢ . . . . Substrtutrng — from equation (7.8) rnto equation (7.14) redefines 1+ in terms of a con- centration gradient and equation (7.14) becomes 2 o I+=FD:—d—C-+ FD‘ c flic- (7.15) dx RT Fc dx 1+ = 217D: _c_l_c_ (7.16) dx This implies that under steady state conditions I, = 211’ = 213' (7.17) so 15 = 1;" (7.18) The current due to the mass transfer of cations can also be written in terms of the anions as If = {23]13 (7.19) 132 0° 1;" =[Di )1? (7.20) Since the current flow is constant throughout the electrolyte, then, Dj/D: is constant. This would mean that the concentration gradient is constant across the electrolyte, so £=£=c —c (7.21) where C, and Cc are the concentrations of ions at the anode and cathode respectively. Since Ax =1 then Ac = ca —C<‘ so the concentration at any given point is c=Acx+cc =(ca-cc)x+cc (7.22) SO C, + c = 2c, (7.23) The potential drop across the electrolyte in the steady state becomes 9:1; 5T0 1 (7.24) dx FD+ c @1213: 1:70 l (7.25) dx FD+ (ca—cc)x+cc ¢ —¢ -A¢-l”’l— RT II dx (726) u ‘2 i LFZD: 0(ca-cc)x+cc . RT 1 c A 21’". , ln—“— 7.27 ¢ + [FZD+]Ca—Cc Cc, ( ) Since m d . dc . 1+ =1+ =FD+E=FD+ (ca—cc) (7.28) 133 then . RT 1 c A = FD —- , , ln—“— 7.29 ¢ I: +(Ca CL)][F2D+]Ca—cc Cc ( ) A¢=fllnC—"=AE (7.30) F cc SO A¢=AV—AE=AV-A¢ (7.31) AV = 2A¢ = zflrni (7.32) F c For any small value of AV , the equation 7.15 can be written as 1, = 21?ng = ZFD: (c, —c,) (7.33) Since current and voltage are the only variables that can be measured experimentally a . . . . . . c relatronshrp between the concentration gradient (ca —cc)1n equatron (7.33) and ln-F- c C from equation 7.31 must be established. To do this let ; = c) (7.34) then coln—“—= " “ln%"—=cc[l+%)ln(l+{) (7.35) 2 3 4 in which ln(1+§)can be expanded as ({—%+%——%~+m] when (<1. Equation 7.34 can be written as 134 n—“= 2 c C C 2 3 4 —c"+c‘l 6 [1+5 L+3C_ For small values of C then equation (7.36) simplifies to c +c c —£——Cln-i=ccC =60 —cc 2 c C This implies that the current is I, = 2m: (c, —c,) 1, = 21?ch0 1nc—a C C Remember that A¢ = é: 1n fi which means that c C 1nc_a-_-A¢_F_ c RT (' current is then , . 1,='——F D+c° 2A¢ RT 2 9 1+ :mAV RT which simplifies to I, = Fu:c0AV This is important since equation (7.3) is 10 = F(u: +u:)c0AV = aAV 135 12 12 40 (7.36) (7.37) (7.38) (7.39) (7.40) (7.41) (7.42) (7.43) Recall that the transference number is defined as _ Iziluic, t, — — zlziluic, i (7.44) Remembering that z, = 1 and the concentrations are equal then equation 7.38 simplifies to t. = i (7.45) Rewriting equations (7.3) and (7.43) the following equation are obtained F (u+ + u_ )00 1 AV = 3 (7.47) F u,c0 Since AV is constant, then the ratio of the cationic steady state current to the initial cur- rent can be written as _I_+_ = . Fu+c.AV° (7.48) IO Fu+C.AV + Fu_c.AV which simplifies to I . —+— = ,”+ ., =1, (7.49) 10 u+ + u_ This means that a chronoamperometry experiment with an applied volatage of S 10 mV is needed to determine t+. 136 Conditions of use The following conditions must be met in order to use the above equation a) The applied potential must be S 10 mV b) The ratio of c0 Incl/(ca —cc) must be = 1 i 1% cc c) The concentration of salt must be very small. In terms of lithium ion polymer electrolytes the OzLi ratio <<<<<< 20!! Otherwise for OzLi = 20 then the calculation of lithium ion transference numbers is de- termined through the following equation :51 ———AV_I°R° (7.50) L" 10 AV—IfiR” where the subscript ss denotes steady state conditions of the cation. This determination still involves a chronoamperometry step with an applied bias of S 10 mV. However, it is conducted after the measurement of the initial interface resistance (R0) and prior to meas- urement of the final interfacial resistance (Rm ). Both resistances are determined through AC impedance spectroscopy. 137 Bibliography 138 .11. .. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) 8 Bibliography Linden, D. Handbook of Batteries; 2nd ed.; McGraw-Hill Inc.: New York, 1995 Schalkwijk, W. A. V.; Scrosati, B., Eds. Advances in Lithium-Ion Batteries; Klu- wer Academic / Plenum Publishers: New York, 2002. Carvalho, L. M.; Guegan, P.; Cheradame, 11.; Gomes, A. S. Eur. Polym. J., 2000, 36, 401-409. Aabloo, A.; Thomas, J. Solid State Ionics, 2001, 143, 83-87. Borodin, 0.; Smith, G. D. Macromolecules, 1998, 31, 8396-8406. Lemaitre-Auger, F.; Prud'homme, J. Electrochim. Acta, 2001, 46, 1359-1367. Song, L.; Evans, J. W. J. Electrochem. Soc., 2000, 14 7, 2086-2095. Armand, M. B. Fast Ion Transport in Solids; Gool, W. V., Ed.; North-Holland: Amsterdam, 1973, pp 665-673. Aihara, Y.; Appetecchi, G. B.; Scrosati, B.; Hayamizuc, K. Phys. Chem. Chem. Phys, 2002, 4, 3443-3447. Bayoudh, S.; Parizel, N.; Reibel, L. Polym. Int., 2000, 49, 703-711. Kataoka, H.; Saito, Y.; Sakai, T.; Deki, S.; Ikeda, T. J. Phys. Chem. B, 2001, 105, 2546-2550. Mello, N. C.; Bonagamba, T. J.; Panepucci, H.; Dahmouche, K.; Judeinstcin, P.; Aegerter, M. A. Macromolecules, 2000, 33, 1280-1288. Brissot, C.; Rosso, M.; Chazalviel, J. N.; Lascaud, S. .1. Power Sources, 2001, 94, 212-218. 139 (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) Capiglia, C.; Imanishi, N.; Takeda, Y.; Henderson, W. A.; Passerinib, S. J. Elec- trochem. Soc., 2003, 150, A525-A531. Song, M.-K.; Kim, Y.-T.; Kim, Y. T.; Cho, B. W.; Popov, B. N.; Rhee, H.-W. J. Electrochem. Soc., 2003, 150, A439-A444. Kato, Y.; Suwa, K.; Ikuta, H.; Uchimoto, Y.; Wakihara, M.; Yokoyama, S.; Yabe, T.; Yamamoto, M. J. Mater. Chem, 2003, 13, 280-285. Gnanaraj, J. S.; Levi, M. D.; Gofer, Y.; Aurbach, D.; Schmidt, M. J. Electrochem. Soc., 2003, 150, A445-A454. Yamaguchi, H.; Takahashi, H.; Kato, M.; Arai, J. J. Electrochem. Soc., 2003, 150, A3l2-A315. Gray, F. M. Solid Polymer Electralytes--Fundamentals and Technological Appli- cations; VCH publishers, Inc.: New York, 1991; pp 45-53. Christie, L.; Christie, A. M.; Vincent, C. A. Electrochem. Solid State Lett., 1999, 2, 187-188. Gauthier, M.; Belanger, A.; lKapfer, K.; Vassort, G.; Armand, M. Polymer Elec- trolyte Reviews; MacCallum, J. R., Vincent, C. A., Eds.; Elsevier: London, 1989; Vol. 2, pp 285-332. Vincent, C. A. Prog. Solid State Chem, 1987, 17, 145-261. Bruce, P. G.; Vincent, C. A. J. Electroanal. Chem, 1987, 225, 1-17. Chandrasekhar, V. Adv. Poly. Sci, 1998, 135, 139-205. Scrosati, B. Electrochim. Acta, 2000, 45, 2461-2466. Chen, H. P.; Fergus, J. W.; Jang, B. Z. J. Electrochem. Soc., 2000, 147, 399-406. Klassen, B.; Aroca, R.; Nazri, M.; Nazri, G. A. J. Chem. Phys. B., 1998, 102, 4795-4801. 140 (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) Ding, M. S.; Xu, K.; low, T. R. J. Electrochem. Soc., 2000, 14 7, 1688-1694. Peled, E.; Golodnitsky, D.; Menachem, C.; Bar-Tow, D. J. Electrochem. Soc., 1998, 145, 3482-3486. Aurbach, D.; Ein-Eli, Y. J. Electrochem. Soc., 1995, 142, 1746-1752. Chu, A.; Josefowicz, J.; Farrington, G. J. Electrochem. Soc., 1997, 144, 4161. Ogumi, Z.; Inaba, M. Bull. Chem. Soc. Jpn., 1998, 71, 721. Blomgren, G. E. J. Power Sources, 1999, 81-82, 1 12-118. Aurbach, D.; Ein-Ely, Y.; Zaban, A. J. Electrochem. Soc., 1994, 141, L1-L3. Wright, P. V. J. Polym. Sci. Part B: Polym. Phys, 1976, 14, 955-957. Fenton, D. B; Parker, J. M.; Wright, P. V. Polymer, 1973, 14, 589-589. Wright, P. V. Brit. Polym. J., 1975, 7, 319-327. MacCallum, J. R.; Vincent, C. A. Polymer Electrolyte Reviews - 2; Elsevier Sci- ence Publishers Ltd.: New York, 1989 Wagner, A.; Kliem, H. J. Appl. Phys, 2002, 91, 6638-6649. MacCallum, .1. R.; Vincent, C. A. Polymer Electrolyte Reviews-1; Elsevier Ap- plied Science: New York, 1987 Staudinger, H.; Lohmann, H. Ann. Chim., 1933, 505, 41. Hill, F. N.; Bailey, F. E.; Fitzpatrick, J. T. Ind. Eng. Chem, 1958, 50, 5. Tashiro, K.; Tadokoro, H. Reports Prog. Poly. Phys. Jpn, 1978, 2], 417-420. 141 (44) (45) (46) (47) (48) (49) (50) (51) (52) (53) (54) (55) (56) (57) (58) Takahashi, Y.; Sumita, 1.; Tadokoro, H. J. Polym. Sci. Part B: Polym. Phys, 1973, 11, 2113-2122. Tadokoro, H.; Chatani, Y.; Yoshihara, T.; Tahara, S.; Murahashi, S. Makromol. Chem, 1964, 73, 109-127. Bailey, F. E., Jr.; Koleske, J. V. Poly(ethylene oxide); Academic: New York, N.Y., 1976; p 174. Yoshihara, T.; Tadokoro, H.; Murahashi, S. J. Chem. Phys, 1964, 11, 2902-2904. Koenig, J. L.; Angood, A. C. J. Polym. Sci, PartA: Polym. Chem, 1970, 3, 1787- 1796. Liu, K.-J.; Parsons, J. L. Macromolecules, 1969, 2, 529-533. Machida, K.; Miyazawa, T. Spectr. Acta, 1964, 20, 1865-1873. Matsuura, H.; Fukuhara, K. J. Mol. Struct, 1985, 126, 251-260. Matsuura, H.; Miyazawa, T. Spectr. A cta, 1967, 23A, 2433-2447. Matsuura, H.; Fukuhara, K. J. Polym. Sci. Part B: Polym. Phys, 1986, 24, 1383- 1400. Mullerplathe, F.; Vangunsteren, W. F. J. Chem. Phys, 1995, 103, 4745-4756. Meyer, W. H. Adv. Mater., 1998, 10, 439-448. Scrosati, B. Chem. Rec., 2001, 1, 173-181. Souquct, J. L.; Duclot, M. Solid State Ionics, 2002, 148, 375-379. Wright, P. V. MRS Bull, 2002, 27, 597-602. 142 (59) (60) (61) (62) (63) (64) (65) (66) (67) (68) (69) (70) (71) (72) Chandrasekhar, V.; Krishnan, V.; Athimoolam, A.; Nagendran, S. Curr. Sci, 2000, 78, 464-472. Nagaoka, K.; Naruse, H.; Shinohara, 1.; Watanabe, M. J. Polym. Sci. Part A: Po- lym. Chem, 1984, 22, 659-663. Blonsky, P. M.; Shriver, D. F .; Austin, P.; Allcock, H. R. J. Amer. Chem. Soc., 1984, 106, 6854-6855. Shriver, D. F. .1. Electrochem. Soc., 1987, I34, 270. Abraham, K. M.; Alamgir, M. Chem. Mater., 1991, 3, 339-348. Spindler, R.; Shriver, D. F. J. Amer. Chem. Soc., 1988, 110, 3036-3043. Qiao, J.; Chen, Y. Y.; Baker, G. L. Chem. Mater., 1999, 11, 2542-2547. Hall, P. G.; Davies, G. R.; Ward, 1. M.; McIntyre, J. E. Polym. Commun, 1986, 27, 100-102. Inoue, K.; Nishikawa, Y.; Tanigaki, T. Macromolecules, 1991, 24, 3464-3465. Feuillade, G.; Perche, P. J. Appl. Electrochem., 1975, 5, 63-69. Tarascon, J. M.; Gozdz, A. S.; Schmutz, C.; Shokoohi, F .; Warren, P. C. Solid State Ionics, 1996, 86-8, 49-54. Dias, F. B.; Plomp, L.; Veldhuis, J. B. J. J. Power Sources, 2000, 88, 169-191. Lee, K. 11.; Park, J. K.; Kim, W. J. J. Polym. Sci. Part B: Polym. Phys, 1999, 37, 247-252. Choe, H. 8.; Carroll, B. G.; Pasquariello, D. M.; Abraham, K. M. Chem. Mater., 1997, 9, 369-379. 143 (73) (74) (75) (76) (77) (78) (79) (80) (81) (82) (83) (84) (85) (86) (87) Forsyth, M.; MacFarlane, D. R.; Hill, A. J. Electrochim. Acta, 2000, 45, 1243- 1247. Abbrent, S.; Plestil, J.; Hlavata, D.; Lindgren, J .; Tegenfeldt, J .; Wendsjo, A. Polymer, 2001, 42, 1407-1416. Hou, X. P.; Siow, K. S. Polymer, 2000, 41, 8689-8696. Rajendran, 8.; Uma, T. Bull. Mat. Sci, 2000, 23, 27-29. Rajendran, S.; Mahendran, O.; Kannan, R. J. Phys. Chem. Solids, 2002, 63, 303- 307. Yue, Z.; McEwen, I. J .; Cowie, J. M. G. J. Mater. Chem, 2002, 12, 2281-2285. Aihara, Y.; Kuratomi, J.; Bando, T.; Iguchi, T.; Yoshida, H.; Ono, T.; Kuwana, K. J. Power Sources, 2003, 114, 96-104. Gray, F .; Armand, M. Handbook of Battery Materials; Wiley-VCH: New York, 1999, pp 499-523. Weston, J. E.; Steele, B. C. H. Solid State Ionics, 1982, 7, 75. Capuano, F .; Croce, F .; Scrosati, B. J. Electrochem. Soc., 1991, 138, 1918-1922. Raghavan, S. R.; Riley, M. W.; Fedkiw, P. S.; Khan, S. A. Chem. Mater., 1998, 10, 244-251. Krawiec, W.; Scanlon, L. G.; Fellner, J. P.; Vaia, R. A.; Giannelis, E. P. J. Power Sources, 1995, 54, 310-315. Cho, J. P.; Liu, M. L. Electrochim. Acta, 1997, 42, 1481-1488. Hou, J.; Baker, G. L. Chem. Mater., 1998, 10, 3311-3318. Ballauff, M. Angew. Chem. Int. Ed. Engl, 1989, 28, 253-267. 144 (88) (89) (90) (91) (92) (93) (94) (95) (96) (97) (98) (99) (100) (101) (102) Lenz, R. W. Faraday Discuss, 1985, 79, 21-32. Ballauff, M. Macromolecules, 1986, 19, 1366-1374. McHattie, G. S.; Imrie, C. T.; Ingram, M. D. Electrochim. Acta, 1998, 43, 1151- 1154. Baum, P.; Meyer, W. H.; Wegner, G. Polymer, 2000, 41, 965-973. Lauter, U.; Meyer, W. H.; Wegner, G. Macromolecules, 1997, 30, 2092-2101. Amautov, S. A.; Gavrilenko, V. V. Synth. Met, 1997, 84, 283-284. Shimura, T.; Funaki, H.; Nishihara, H.; Aramaki, K.; Ohsawa, T.; Yoshino, K. Chem. Lett., 1992, 457-460. Bringmann, G.; Walter, R.; Weirich, R. Angew. Chem. Int. Ed. Engl, 1990, 29, 977-991. Fanta, P. E. Synthesis, 1974, 9-21. Wirth, H. O.; Herrmann, F. U.; Hem, W. Makromol. Chem, 1964, 80, 120-140. Percec, V.; Hill, D. H. Step-Growth Polymers For High-Performance Materials; American Chemical Society: Washington, DC, 1996; Vol. 624, pp 2-56. Percec, V.; Pugh, C.; Cramer, E.; Okita, S.; Weiss, R. Makromol. Chem, 1992, 54-5, 113-150. Durairaj, K. Curr. Sci, 1994, 66, 833-838. Goodson, F. E.; Wallow, T. 1.; Novak, B. M. J. Amer. Chem. Soc., 1997, 119, 12441-12453. Saa, J. M.; Martorell, G. J. Org. Chem, 1993, 58, 1963-1966. 145 (103) (104) (105) (106) (107) (108) (109) (110) (111) (112) (113) (114) (115) (116) Rehahn, M.; Schluter, A. D.; Wegner, G.; Feast, W. J. Polymer, 1989, 30, 1060- 1062. Kwiatkowski, G. T.; Colon, 1.; Elhibri, M. J .; Matzner, M. Macromol. Symp., 1992, 54-5, 199-224. Maccallum, J. R.; Tomlin, A. S.; Vincent, C. A. Eur. Polym. J., 1986, 22, 787- 791. Doyle, M.; Newman, J. Electrochim. Acta, 1995, 40, 2191-2196. Doyle, M.; Fuller, T. F.; Newman, J. Electrochim. Acta, 1994, 39, 2073-2081. Riley, M.; Fedkiw, P. S.; Khan, S. A. J. Electrochem. Soc., 2002, 149, A667- A674. Bannister, D. J.; Davies, G. R.; Ward, 1. M.; McIntyre, J. E. Polymer, 1984, 25, 1291-1296. Tsuchida, E.; Kobayashi, N.; Ohno, H. Macromolecules, 1988, 21 , 96-100. Hardy, L. C.; Shriver, D. F. J. Amer. Chem. Soc., 1985, 107, 3823-3828. Jacob, M. M. E.; Hackett, E.; Giannelis, E. P. J. Mater. Chem, 2003, 13, 1-5. Ratner, M. A.; Johansson, P.; Shriver, D. F. MRS Bull, 2000, 25, 31-37. Scrosati, B.; Vincent, C. A. MRS Bull, 2000, 25, 28-30. Wieczorek, W.; Florjanczyk, 2.; Stevens, J. R. Electrochim. Acta, 1995, 40, 2251- 2258. Croce, F .; Curini, R.; Martinelli, A.; Persi, L.; Ronci, F .; Scrosati, B.; Caminiti, R. J. Phys. Chem. B, 1999, 103, 10632-10638. 146 (117) (118) (119) (120) (121) (122) (123) (124) (125) (126) (127) (128) (129) (130) (131) Fujinami, T.; Tokimun, A.; Mehta, M. A.; Shriver, D. F .; Rawsky, G. C. Chem. Mater., 1997, 9, 2236-2239. Doan, K. E.; Ratner, M. A.; Shriver, D. F. Chem. Mater., 1991, 3, 418-423. Onishi, K.; Matsumoto, M.; Nakacho, Y.; Shigehara, K. Chem. Mater., 1996, 8, 469-472. Rawsky, G. C.; Fujinami, T.; Shriver, D. F. Chem. Mater., 1994, 6, 2208-2209. Fujinami, T.; Tokimune, A.; Mehta, M. A.; Shriver, D. F .; Rawsky, G. C. Chem. Mater., 1997, 9, 2236-2239. Walls, H. J.; Zhou, J.; Yerian, J. A.; Fedkiw, P. S.; Khan, S. A.; Stowe, M. K.; Baker, G. L. J. Power Sources, 2000, 89, 156-162. Levesque, I.; Leclerc, M. J. Chem. Soc., Chem. Commun, 1995, 2293-2294. Chang, 5.; Han, C. D. Macromolecules, 1997, 30, 1670-1684. Steuer, M.; Ballauff, M. J. Polym. Sci. Pol. Chem, 1993, 31, 1609-1619. Majnusz, J .; Catala, J. M.; Lenz, R. W. Eur. Polym. J., 1983, 19, 1043-1046. Stern, R.; Ballauff, M.; Lieser, G.; Wegner, G. Polymer, 1991, 32, 2096-2105. Majnusz, J.; Lenz, R. W. Eur. Polym. J., 1989, 25, 847-855. Steiner, U. B.; Caseri, W. R.; Suter, U. W.; Rehahn, M.; Rau, I. U. Langmuir, 1994, 10, 1164-1170. Delnoye, D. A. P.; Sijbesma, R. P.; Vekemans, J. A. J. M.; Meijer, E. W. J. Amer. Chem. Soc., 1996, 118, 8717-8718. Kato, T.; Kihara, H.; Uryu, T.; F ujishima, A.; Frechet, J. M. J. Macromolecules, 1992, 25, 6836-6841. 147 (132) (133) (134) (135) (136) (137) (138) (139) (140) (141) (142) (143) (144) Moriya, K.; Mizusaki, H.; Kato, M.; Suzuki, T.; Yano, S.; Kajiwara, M.; Tashiro, K. Chem. Mater., 1997, 9, 255-263. Song, H. H.; Dotrong, M.; Price, G. E.; Dotrong, M. H.; Vakil, U. M.; Santhosh, U.; Evers, R. C. Polymer, 1994, 35, 675-680. Ciferri, A. Polym. Eng. Sci, 1994, 34, 377-378. Krause, S. J.; Haddock, T. B.; Price, G. B; Adams, W. W. Polymer, 1988, 29, 195-206. Meyer, W. H. Adv. Mater., 1998, 10, 439-448. Eisenbach, C. D.; Hofmann, J.; Goeldel, A.; Noolandi, J .; Shi, A. C. Macromole- cules, 1999, 32, 1463-1470. Bouachrine, M.; Lere-Porte, J .-P.; Moreau, J. J. E.; Spirau, F. 8.; Augusta da Silva, R.; Lmimouni, K.; Ouchani, L.; Dufour, C. Synth. Met, 2002, 126, 241- 244. Johansson, T.; Mammo, W.; Andersson, M. R.; Inganas, 0. Chem. Mater., 1999, 11, 3133-3139. Lauter, U.; Meyer, W. H.; Enkelmann, V.; Wegner, G. Macromol. Chem. Phys, 1998, 199, 2129-2140. Colon, 1.; Kelsey, D. R. J. Org. Chem, 1986, 51, 2627. Colon, 1.; Kwiatkowski, G. T. J. Polym. Sci, Part A: Polym. Chem, 1990, 28, 367-383. Ueda, M.; Ichikawa, F. Macromolecules, 1990, 23, 926-930. Ueda, M.; Miyaji, Y.; Ito, T.; Oba, Y.; Sone, T. Macromolecules, 1991, 24, 2694- 2697. 148 (145) (146) (147) (148) (149) (150) (151) (152) (153) (154) (155) (156) (157) Pasquale, A. J.; Sheares, V. V. J. Polym. Sci, Part A: Polym. Chem, 1998, 36, 2611-2618. Luzinov, 1.; Julthongpiput, D.; Bloom, P. D.; Sheares, V. V.; Tsukruk, V. V. Mac- romol. Symp., 2001, 167, 227-242. Percec, V.; Zhao, M.; Bae, J.-Y.; Hill, D. H. Macromolecules, 1996, 29, 3727- 3735. Percec, V.; Asandei, A. D.; Hill, D. H.; Crawford, D. Macromolecules, 1999, 32, 2597-2604. Chen, Y.; Baker, G. L. J. Org. Chem, 1999, 64, 6870-6873. Kwiatkowski, G. T.; Matzner, M.; Colon, I. J. Macromol. Sci, Pure Appl. Chem, 1997, A34, 1945-1975. Kwiatkowski, G. T.; Colon, 1. Contemporary Topics in Polymer Science; Pearce, E. M., Schaefgen, J. R., Eds; Plenum Press: New York, 1992; Vol. 7, pp 57-74. Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, 1953 Petekidis, G.; Vlassopoulos, D.; Fytas, G.; Kountourakis, N.; Kumar, S. Macro- molecules, 1997, 30, 919-931. Goodson, F. E.; Wallow, T. 1.; Novak, B. M. Macromolecules, 1998, 31, 2047- 2056. Bouachrine, M.; Lere-Porte, J .-P.; Moreau, J. J. E.; Spirau, F. 8.; Augusta da Silva, R.; Lmimouni, K.; Ouchani, L.; Dufour, C. Synthetic Metals, 2002, 126, 241-244. Dubois, M.; Naji, A.; Buisson, J. P.; Humbert, B.; Grivei, E.; Billaud, D. Carbon, 2000, 38,1411-1417. Hernandez, V.; Soto, J.; Navarrete, J. T. L. Synth. Met, 1993, 57, 4461-4466. 149 (158) (159) (160) (161) (162) (163) (164) (165) (166) (167) (168) (169) (170) (171) Louarn, G.; Athouel, L.; Froyer, G.; Buisson, J. P.; Lefrant, S. Synth. Met, 1993, 5 7, 4762-4767. Marucci, A.; Pimenta, M. A.; Brown, S. D. M.; Matthews, M. J .; Dresselhaus, M. S.; Endo, M. J. Mater. Res, 1999, 14, 3447-3454. Asfour, F. 11.; Ruud, C. J .; Baker, G. L., submitted for publication in Chem. Ma- ter. Janssen, R. G.; Utley, J. H. P.; Carre, E.; Simon, E.; Schirmer, H. J. Chem. Soc., Perkin Trans. 2, 2001, 1573-1584. Asakawa, M.; Ashton, P. R.; Balzani, V.; Boyd, S. E.; Credi, A.; Mattersteig, G.; Menzer, S.; Montalti, M.; Raymo, F. M.; Ruffilli, C.; Stoddart, J. F .; Venturi, M.; Williams, D. J. Eur. J. Org. Chem, 1999, 985-994. Delamarche, E.; Donzel, C.; Kamounah, F. 8.; Wolf, H.; Geissler, M.; Stutz, R.; Schmidt-Winkel, P.; Michel, B.; Mathieu, H. J .; Schaumburg, K. Langmuir, 2003, 19, 8749-8758. Leaper, J. M. F.; Raman, H. P. J. US Patent 3 703 546, 1972. Cowie, J. M. G. Polym. Int, 1998, 47, 20-27. Scrosati, B. Polym Int, 1998, 4 7, 50-55. Murata, K.; Izuchi, S.; Yoshihisa, Y. Electrochim. Acta, 2000, 45, 1501-1508. Burke, L. M., Jr.; Khan, 1. M. Macromol. Chem. Phys, 2000, 201, 2228-2233. Wang, Y. C.; Quirk, R. P. Macromolecules, 1995, 28, 3495-3501. Philips, R. W.; Sheares, V. V.; Samulski, E. T.; Desimone, J. M. Macromolecules, 1994, 27, 2354-2356. Baker, K. N.; Fratini, A. V.; Resch, T.; Knachel, H. C.; Adams, W. W.; Socci, E. P.; Farmer, B. L. Polymer, 1993, 34, 1571-1587. 150 (172) (173) (174) (175) (176) (177) (178) (179) (180) (181) (182) (183) (184) (185) Pasco, S. T.; Baker, G. L. Synth. Met, 1997, 84, 275-276. Kim, Y. W.; Lee, W.; Choi, B. K. Electrochim. Acta, 2000, 45, 1473-1477. Nafshun, R. L.; Lerner, M. M.; Hamel, N. N.; Nixon, P. G.; Gard, G. L. J. Elec- trochem. Soc., 1995, 142, L153-L155. Dygas, J. R.; Misztal-Faraj, B.; Florjanczyk, Z.; Krok, F .; Marzantowicz, M.; Zy- gadlo-Monikowska, E. Solid State Ionics, 2003, 15 7, 249-256. Wendsjoe, A.; Yang, H. Int. Symp. Polym. Electrolytes, 2nd, 1990, 225-237. Wright, P. V. J. Mater. Chem, 1995, 5, 1275-1283. del Rio, C.; Wright, P. V. J. Mater. Chem, 1996, 6, 947-952. Gray, F. M. Solid Polymer Electrolytes--Fundamentals and Technological Appli- cations; VCH publishers, Inc.: New York, 1991; p 245. Kim, J.; Park, M. K.; Bae, J. Y.; Ahn, J. H. Electrochem. Commun, 2001, 3, 643- 648. Hou, W. H.; Chen, C. Y.; Wang, C. C.; Huang, Y. H. Electrochim. Acta, 2003, 48, 679-690. Nishimoto, A.; Agehara, K.; Furuya, N.; Watanabe, T.; Watanabe, M. Macro- molecules, 1999, 32, 1541-1548. Gnanaraj, J. S.; Karekar, R. N.; Skaria, S.; Rajan, C. R.; Ponrathnam, S. Polymer, 1997, 38, 3709-3712. Kerr, J. B.; Liu, G.; Curtiss, L. A.; Redfem, P. C. Electrochim. Acta, 2003, 48, 2305-2309. Watanabe, M.; Endo, T.; Nishimoto, A.; Miura, K.; Yanagida, M. J. Power Sources, 1999, 81 -82, 786-789. 151 (186) (187) (188) (189) (190) (191) (192) (193) (194) (195) (196) (197) (198) Kerr, J. B.; Sloop, S. E.; Liu, G.; Han, Y. B.; Hou, J.; Wang, S. J. Power Sources, 2002, 110, 389-400. Asfour, F. H.; Ruud, C. J.; Baker, G. L., submitted for publication in Macromole- cules. Bruce, P. G.; Hardgrave, M. T.; Vincent, C. A. Solid State Ionics, 1992, 53-56, 1087-1094. Egashira, M.; Scrosati, B.; Armand, M.; Berangcr, S.; Michot, C. Electrochem. Solid State Lett., 2003, 6, A71-A73. Wieczorek, W.; Such, K.; Przyluski, J .; F lorianczyk, Z. Synth. Met, 1991, 45, 373-383. Croce, F.; Scrosati, B. Advanced Membrane Technology; New York Academy of Sciences: New York, 2003; Vol. 984, pp 194-207. Croce, F .; Appetecchi, G. B.; Persi, L.; Scrosati, B. Nature, 1998, 394, 456-458. Matsushita, K.; Shimazaki, Y.; Mehta, M. A.; Fujinami, T. Solid State Ionics, 2000, 133, 295-301. Chen, H.-W.; Chang, F.-C. Polymer, 2001, 42, 9763-9769. Li, Y.; Fedkiw, P. S.; Khan, S. A. Electrochim. Acta, 2002, 47, 3853-3861. Zhou, J.; Fedkiw, P. S.; Khan, S. A. J. Electrochem. Soc., 2002, 149, A1121- A1126. Lascaud, S.; Perrier, M.; Vallee, A.; Besner, S.; Prudhomme, J.; Armand, M. Macromolecules, 1994, 27, 7469-7477. Krause, L. J .; Lamanna, W.; Summerfield, J.; Engle, M.; Korba, G.; Loch, R.; Atanasoski, R. J. Power Sources, 1997, 68, 320-325. 152 (199) (200) (201) (202) (203) (204) (205) (206) Fulcher, C.; Crowell, M. A.; Bayliss, R.; Holland, K. B.; Jezorek, J. R. Anal. Chim. Acta, 1981, 129, 29-47. Roesky, H. W. Angew. Chem. Int. Ed., 1967, 6, 711. Glemser, O.; Roesky, H. W.; Heinze, P. R. Angew. Chem. Int. Ed., 1967, 6, 710- 711. Leonardelli, S.; Facchini, L.; Fretigny, C.; Tougne, P.; Legrand, A. P. J. Amer. Chem. Soc., 1992, 114, 6412-6418. Siska, D. P.; Shriver, D. F. Chem. Mater., 2001, 13, 4698-4700. Edman, L.; Doeff, M. M.; Ferry, A.; Kerr, J.; De Jonghe, L. C. J. Phys. Chem. B, 2000, 104, 3476-3480. Gagne, M. R.; Stem, C. L.; Marks, T. J. J. Amer. Chem. Soc., 1992, 114, 275-294. Larock, R. C.; Yang, H.; Weinreb, S. M.; Herr, R. J. J. Org. Chem, 1994, 59, 4172-4178. 153