AMINOPHENYL DOUBLE DECKER SILSESQUIOXANES : SPECTROSCOPIC ELUCIDATION, PHYSICAL AND THERMAL CHARACTERIZATION, AND THEIR APPLICATIONS By Beth Whitney Schoen A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemical Engineering Ð Doctor of Philosophy 2013 ABSTRACT AMINOPHENYL DOUBLE DECKER SILSESQUIOXANES: SPECTROSCOPIC ELUCIDATION, PHYSICAL AND THERMAL CHARACTERIZATION, AND THEIR APPLICATIONS By Beth Whitney Schoen The incorporation of cage -like silsesquioxanes (SQ) to form polymers has demonstrated property enhancements in areas such as: t hermal and mechanical characteristics, flame retarda nce, dielectric properties , and oxidative resistance. However , with most hybrid polymers investigated, the attached SQs are pendant with respect to the polymer bac kbone. A recently developed class of these nano -structured, cage -like silsesquioxanes, formally known as double decker silsesquioxanes (DDSQ) , offers the opportunity to form hybrid polymers with SQ cages as a part of the polymer backbone. However , during the capping reaction, these functionalized DDSQs generate cis and trans isomers with respect to the 3D Si -O core. Therefore, it is logical to characterize properties , which will allow for optimization of capping reaction parameters, particularly if one isomer is favored over the other. Moreover, these characteristics are also relevant when reacting or incorporating these isomers, or mixtures thereof, with other molecules to form novel materials. In this dissertation , three aminophenyl DDSQ s were synthesized . More specifically, two meta- aminophenyl DDSQ s, which were differentiated according to the moiety attached to the D -Si (methyl or cyclohexyl), and one para -aminophenyl DDSQ with a methyl moiety were used . Chemical, p hysical , and thermal charac teristics were evaluated for individual isomers a s well as binary mixtures of different cis/trans ratios . The 1H N MR spectra of the cis and trans isomers of these DDSQ had not previously been assigned to a degree that allowed for quantification , which was necessary for these studies . Thus, 1HÐ29Si HMBC correlations were applied to facilitate 1H spectral assignment s and also to confirm previous 29Si assignments. Using 1H NMR not only saves time and material over 29Si NMR, but also provides a more accurate quantification, thus allowing for the ratio of cis and trans isomers present in each compound to be determined . Solubility behavior was investigated for the individual isomers, such that separation techniques could be deve loped. It was found that cis isomers were 33 times more soluble than trans isomers for para -aminophenyl (R = methyl), and 22 times more soluble for the analogous meta-aminophenyl in a solution of THF and hexanes. For a more sterically hindered meta -aminoph enyl (R = cyclohexyl), cis isomers were only 3.5 times more solubl e, and the overall solubility was also the lowest. Phase diagrams representing solid -liquid melt equilibria of the binary cis/trans mixtures were developed . Single crystal x -ray diffraction data of isolated isomers helped to interpret the phase behavior . Both compounds with a methyl moiety exhibited eutectic phase diagrams. Their trans isomers were higher melting and exhibited larger packing density . Changing from para - to meta-aminophenyl shifted the solid -liquid equilibrium further from i deal, with decreased activity coefficients. Cyclohexyl DDSQ exhibited an isomorphic phase diagram (a complete solid -solution) , attributed to cyclohexyl being more similar in size to the phenyl moieties. A specific application that was chosen for the se DDSQ was high performance thermosetting oligoimide s. They displayed advantages over their organic counterparts in areas such as the liquid to solid transition and viscosity, whi ch improves the process ing window . Additionally, they increased the overall degree of cross -linking leading to improved oxidative stability at elevated temperatures. iv This doctoral dissertation is dedicated to my Father , Howard Gilbert Schoen (05.14.1943-11.27.1999). His love for science was instilled in me. May his memory be a blessing to all. !"#$% &'&#"( v ACKNOWLEDGMENTS The completion of my doctoral dissertation would not have been possible without the help and guidance of so many individuals. First, and foremost I would like to thank my advisor, Professor Andre Lee. When I began working for him, just four years ago, I was not the researcher that I have become today. Throughout these years, he has sought to create an environment of independence and critical -thinking, which has allowed me to take ownership of my research and pursue experiments that I fin d to be important . I would also like to thank each person on my guidance committee individual ly, beginning with Professor Melissa Baumann, who was my co -advisor for 3.5 years of my dissertation. She was always there to push me forward and provide addition al insight on research problems. She also introduced me to other facets of research, encouraging me to expand my area of expertise and think outside the box. Next, I would like to thank my committee member, Professor Robert Maleczka Jr. , for his incessant encouragement in organic chemistry. Without his help and guidance, I would not have had access to so many organic chemists, chemistry facilities, and learning opportunities. I like to think of his class that I took as one of the turning point s in my Ph.D s tudies. It encouraged me to continue taking and succeeding in the organic chemistry doctoral courses, although they were not easy and I was already finished with my course requirements, throughout my Ph.D studies. Additionally, I cannot express my gratitud e towards him allowing me to attend and take part in presenting at his weekly group meetings. His encouragement for all that I have mentioned contributed greatly towards the understanding necessary for the synthetic aspect of my dissertation , and opened do ors to observe research outside my area that would not have been available otherwise. vi I would also like to thank my committee member, Professor Carl Lira, for his persistent reassurance and instruction with MatLab and computer programming. He has spent en dless hours with me teaching me how to use MatLab for modeling purposes and even more time going over the results for our recently submitted paper for publication. As a teaching assistant for him in thermodynamics, I was able to further expand my thermodyn amics knowledge and apply this to my dissertation. I appreciate all the time he took out of his already busy schedule for me and how he was able to make very difficult problems seem easy. His contributions to the thermodynamics portion of my dissertation a re innumerable . My final committee member is Professor Dennis Miller. I would like to thank him for his positive reinforcement and encouragement throughout my Ph.D studies. His door was always open for any questions that I had. I also learned a great deal from him as a teaching assistant, and would seek advice from him regarding my teaching experiences in the last four years. Outside of academia, Professor Miller was also available and I would like to express my deepest gratitude towards him for this. Speci fically, I would like to thank him for playing volleyball with me and several other graduate st udents throughout my time at Michigan State University; we could not have won the intramural championship without you! I would also like to thank Daniel Holmes, Ph.D for the countless hours he spent teaching me about Nuclear Magnetic Resonance and his assistance in interpreting my results. Basically, I could not have completed my chapter 2 without his he lp, and greatly appreciate al l the time he spent reviewing my paper for our publication. I would also like to thank him for his help with my cover letters for my post -doctoral applications, and any other writing assignments that I requested. vii I would also like to express gratitude for the numerous hours that Richard Staples, Ph.D spent with me teaching me the ins and outs of single crystal X -ray Diffraction. His assistance with growing crystals and interpreting the results has contributed tremendously to my chapter 4 and the current article we are working on for publi cation. Additionally, he was always available for my myriad of questions and helped me gain an understanding for the structural implications of my molecules. I would also like to thank Professor Aaron Odom and his research group for allowing me to use spa ce in his laboratory and graduate student office to complete my dissertation studies. This helped me tremendously to finish my synthetic work and was especially beneficial to have access to his group members when I had questions regarding specific syntheti c procedures. Additionally, I would like to thank Tim Haddad (Ph.D, AFRL), Air Force Research Labs (Edwards, Air Force Base), and my funders the Air Force Office of Scientific Research (grant number FA9550 -08-1-0213. I would not have developed my foundati on in organic chemistry had it not been for the countless hours that Tim spent with me teaching me the basics of the synthetic procedures for my reactions. I would also like to thank several other people; including Gina Comiskey for taking time out of her busy schedule in the final two year s of her dissertation to teach me how to synthesize polymers. Also, I would to thank Brad Seurer for introducing me to my project and teaching me the basics during his time as a postdoctoral researcher in my research grou p. Additionally, I would like to thank several other friends and undergraduate research assistants that at some point helped me with various aspects of my dissertation including: thermodynamics, MatLab, reviewing chapters, and learning experimental procedu res. These people are: Annie Lown, Arati Santhanakris hnan, David Olson, Nate Leonard , Bria Kamdem, and Jake Finkbiner . I would like viii to thank all my other friends and my research group for their continuous support and encouragement throughout this time peri od, and for lack of space I cannot mention all your names, but I am sure you know who you are. Last, but certainly not least, I would like to thank my family, specifically my sisters (Jill, Leah, and Marly) and my mother (Shelley Schoen) for their unremitt ing support and encouragement throughout my Ph.D. I thank them all for und erstanding why I could not come on all the family vacations and home for the holidays, especially during my last year of my doctoral studies. Most of all, I would like to thank my mo ther for just being herself. Without her I would not have gotten as far as I have today. Throughout my life, she has always been there for me and supported me in everything I endeavor to achieve. ix TABLE OF CONTENTS LIST OF TABLES ..................................................................................................................... xiii LIST OF FIG URES ................................................................................................................... xvi LIST OF SCHEMES ............................................................................................................... xxiii KEY TO SYMBOLS AND ABBREVIATIONS ................................................................... xxiv CHAPTER 1. INTRODUCTION .................................................................................................1 1. Introduction ....................................................................................................................2 1.1 Overall Goal .................................................................................................................2 1.2 Motivation .....................................................................................................................2 1.3 Fundamental Idea and Specific Aims ........................................................................4 REFERENCES ...................................................................................................................6 CHAPTER 2. BACKGROUND INFORMATION .....................................................................9 2. Background information .............................................................................................10 2.1 Silsesquioxanes ...........................................................................................................11 2.1.1 Monofunctionalized ......................................................................................12 2.1.1.1 Corner -capped SQ ......................................................................................12 2.1.1.2 Side -capped SQs ........................................................................................15 2.1.2 Difunctionalized SQs ....................................................................................16 2.1.2.1 Double decker silsesquioxanes (DDSQ) ....................................................19 2.2 Spectroscopic Elucidation .........................................................................................30 2.2.1 29Si NMR ......................................................................................................30 2.2.2 1H NMR .......................................................................................................32 2.3 Isomer Separation ......................................................................................................35 2.3.1 Thermodynamic Modeling of SLE ...............................................................35 2.4 Structural Characterization ......................................................................................42 2.4.1 Melting beha vior and phase diagrams of mixed isomers ..............................46 2.5 Selected Applic ations .................................................................................................52 2.5.1. Thermosets ...................................................................................................52 2.5.2 Thermoplastics ..............................................................................................57 2.5.3 Ionic Liquid ...................................................................................................64 REFERENCES .................................................................................................................67 CHAPTER 3. SYNTHESIS OF MONOMERS ........................................................................76 3. Synthesis of monomer s ................................................................................................77 3.1 Introduction ................................................................................................................77 3.2 Solvents and Reagents ...............................................................................................79 3.3 NMR spectroscopy .....................................................................................................80 3.4 Synthesis .....................................................................................................................80 3.4.1 Compound 3a ................................................................................................81 x 3.4.2 Compound 3b ...............................................................................................82 3.4.3 Compound 3c ................................................................................................84 3.4.4 Compound 3d (mixture of 3a and 3c) ...........................................................87 3.5 Yield analysis of synthetic procedures .....................................................................88 3.6 Concluding remarks on synthesis .............................................................................89 NOTES ..............................................................................................................................90 REFERENCES .................................................................................................................92 CHAPTER 4. IDENTIFICATION AND QUANTIFICATION OF cis AND trans 3a-c USING 1H-29Si gHMBC NMR ...................................................................................................95 4. Identification and quantification of cis and trans 3a-c using 1H-29Si gHMBC NMR ..................................................................................................................................96 4.1 Introduction ................................................................................................................96 4.2 Spectroscopic characterization methods ..................................................................97 4.3 Identification and quantification analysis ...............................................................98 4.3.1 Compound 3a ..............................................................................................101 4.3.2 Compound 3b .............................................................................................103 4.3.3 Compound 3c ..............................................................................................103 4.4 Concluding remarks on spectroscopy ....................................................................108 NOTES ............................................................................................................................109 REFERENCES ...............................................................................................................111 CHAPTER 5. CHARACTERIZATION OF THE SOLUBILITY BEHAVIOR OF cis and trans ISOMERS IN NANO-STRUCTURED DOUBLE DECKER SILSESQUIOXANES ................................................................................................................113 5. Characterization of the solubility behavior of cis and trans isomers in nano -structured double decker silsesquioxanes ....................................................................114 5.1 Introduction ..............................................................................................................114 5.2 Materials and Methods ............................................................................................115 5.2.1. Solvents and Reagents ...............................................................................115 5.2.2. Isomer separation .......................................................................................115 5.2.3 Chromatographic Purification .....................................................................117 5.2.4 NMR spectroscopy ......................................................................................118 5.2.5 Differential scanning Calorimetry ..............................................................118 5.2.6 Modeling .....................................................................................................118 5.3 Results and Discussion .............................................................................................120 5.3.1 Separation of cis and trans 3a-c ..................................................................120 5.3.1.1. Compound 3a ..............................................................................122 5.3.1.2. Compound 3b .............................................................................124 5.3.1.3. Compound 3c ..............................................................................125 5.3.2 Chromatography Results .............................................................................127 5.3.3 DSC Results ................................................................................................127 5.3.4 Results of Model Fitting .............................................................................128 5.4 Concluding remarks on isomer separations ..........................................................135 NOTES ............................................................................................................................137 REFERENCES ...............................................................................................................139 xi CHAPTER 6. PHASE BEHAVIOR FOR cis AND trans ISOMERS OF THREE AMINOPHENYL DOUBLE DECKER SILSESQUIOXANES ............................................141 6. Phase behavior for cis and trans isomers of three aminophenyl double decker silsesquioxanes ................................................................................................................142 6.1 Introduction ..............................................................................................................142 6.2 Materials and Methods ............................................................................................144 6.2.1 Solvents and Reagents ................................................................................144 6.2.2 Cis/trans isomers sample preparation .........................................................144 6.2.3 Characterization ..........................................................................................145 6.2.3.1 NMR spectroscopy .......................................................................145 6.2.3.2 Differential scanning Calorimetry ...............................................146 6.2.3.3 Single crystal X -ray diffraction ...................................................146 6.3 Results .......................................................................................................................147 6.3.1 Thermal behavior of isolated cis and trans isomers ...................................148 6.3.2 Structural results .........................................................................................151 6.3.2.1 Compound 3c ...............................................................................151 6.3.2.2 Compound 3a ...............................................................................152 6.3.2.3 Compound 3b ..............................................................................153 6.3.2.4 Structural symmetry .....................................................................154 6.3.3 Melting behaviors of binary cis/trans mixtures ..........................................154 6.3.3.1 Data analysis ................................................................................154 6.3.4 Determination of the solid + liquid phase diagrams of cis/trans binary mixtures ................................................................................................................156 6.3.4.1 Phase diagram of binary cis/trans mixtures for compound 3c ....156 6.3.4.2 Phase diagram of binary cis/trans mixtures for compound 3a ....159 6.3.4.3 Calculated binary interaction pa rameters of eutectic -type model fitting ........................................................................................................162 6.3.4.4 Phase diagram of binary cis/trans mixtures for compound 3b ....162 6.3.5 Solidification behavior ................................................................................166 6.4 Discussion .................................................................................................................170 6.4.1 Melting Behavior ........................................................................................170 6.4.1.1 Trans vs. cis isomers ....................................................................170 6.4.1.2 Para vs. met a ................................................................................171 6.4.1.3 Methyl vs. cyclohexyl ..................................................................172 6.4.2 Solidification behavior ................................................................................172 6.4.2.1 Trans isomers ...............................................................................173 6.4.2.3 Cis isomers ...................................................................................173 6.4.3 Crystal structures ........................................................................................175 6.4.4 Same molecular symmetry, different molecular structu res ........................176 6.4.5 Melting behavior for Incongruent -type phase diagrams (compounds 3c and 3a) ........................................................................................................................176 6.4.6 Melting behavior for isomorphous solid solution -type phase diagram (compound 3b) .....................................................................................................177 6.4.7 Solidification of binary cis/trans mixtures .................................................178 6.5 Concluding remarks on solid -liquid thermal equilibria ......................................180 REFERENCES ...............................................................................................................182! xii CHAPTER 7. APPLICATIONS ...............................................................................................186 7. Applications ................................................................................................................187 7.1 Polyimide thermoset ................................................................................................187 7.1.1 Introduction .................................................................................................187 7.1.2 Solvents and reagents ..................................................................................187 7.1.3 Nuclear magnetic resonance .......................................................................187 7.1.4 Synthesis .....................................................................................................188 7.1.4.1 Compound 5a ...............................................................................190 7.1.4.2 Compound 5c ...............................................................................191 7.1.4.3 Compound 5d ..............................................................................192 7.1.5 Thermal behavior ........................................................................................193 7.1.6 Viscosity measurements ..............................................................................193 7.1.7 Results and discussion ................................................................................194 7.1.7.1 Thermal analysis ..........................................................................194 7.1.7.2 Viscosity ......................................................................................196 7.1.8 Concluding remarks on polyimide thermosets ...........................................198 7.2 Polyaramid thermoplastics .....................................................................................200 7.2.1 Introduction .................................................................................................200 7.2.2 Solvents and reagents ..................................................................................200 7.2.3 Nuclear magnetic resonance .......................................................................201 7.2.4 Synthesis .....................................................................................................201 7.2.4.1 Silylated MPDA ...........................................................................202 7.2.4.2 DDSQ -like Polyaramide ..............................................................202 7.2.5 Molecular weight ........................................................................................205 7.2.5.1 Gel permeation chromatography ..................................................205 7.2.5.2 Viscosity measurements ...............................................................206 7.2.6 Degradation analysis ...................................................................................208 7.2.7 Glass transition and melting temperatures ..................................................211 7.2.8 Films ...........................................................................................................211 7.2.9 Concluding remarks on DDSQ -based Nomex ............................................213 7.3 Ionic Liquid ..............................................................................................................214 7.3.1 Introduction .................................................................................................214 7.3.2 Solv ents and reagents ..................................................................................214 7.3.3 Synthesis of ionic liquid (6a) .....................................................................214 7.2.4 Mass spectroscopy ......................................................................................217 7.2.5 Concluding remarks on DDSQ -derived ionic liquids .................................217 NOTES ............................................................................................................................219 REFERENCES ...............................................................................................................221 APPENDICES ............................................................................................................................223 APPENDIX A. SYNTHESIS OF MONOMERS ........................................................224 APPENDIX B. SPECTROSCOPY ...............................................................................235 APPENDIX C. SOLUBILITY BEHAVIOR ...............................................................250 APPENDIX D. CRYSTAL STRUCTURES AND PHASE DIAGRAMS .................266 APPENDIX E. APPLICATIONS .................................................................................284 APPENDIX F. FUTURE WORK .................................................................................293 xiii LIST OF TABLES Table 2-1. Thermal -mechanical data for examples of monofunctionalized, corner -capped SQs. ................................................................................................................................................13 Table 2 -2. Properties of materials CB and CC. 2 ............................................................................24 Table 2 -3.62 Thermal properties of polymer CI with different wt % of CH .................................27 Table 2 -4.3 Thermal and mechanical properties of polymers DA .................................................28 Table 2 -5. Tm of meta - vs. para - isomers ......................................................................................42 Table 2 -6. Melting transitions based on meta - and para - substituted benzenes .............................43 Table 2 -7. Tm of cis/trans isomers ................................................................................................43 Table 2 -8. Tm of increasing MW ...................................................................................................44 Table 2 -9. Tm for altered moiety ...................................................................................................45 Table 2 -10.114 Complex viscosity .................................................................................................54 Table 2 -11.120,121 Complex viscosi ty according to MW of PI -A ................................................55 Table 2 -12.114 Viscosity stability of PETI -3K and PE -3F ............................................................56 Table 2 -13.114 Viscosity stability of PE 6F ...................................................................................56 Table 2 -14.120,121 Viscosity stability for samples PI -A held at 280 ¡C for 2 hours .....................57 Table 2 -15. !inh of synthetic Nomex ............................................................................................60 Table 2 -16.122 Properties of PPD -T and MPD -I ...........................................................................62 Table 3 -1. Yield analysis of the products of the reactions involved for compounds 3a-c ............89 Table 4 -1. 29Si and 1H resonances of cis/trans 3a ......................................................................103 Table 4 -2. 29Si and 1H resonances of cis/trans 3c ......................................................................105 xiv Table 4 -3. Integrated values of 1H NMR spectra from various mixtures of 3 ............................106 Table 4 -4. Integrated values of 1H NMR spectra vs. 29Si NMR spectra from 3a mixtures ........107 Table 5 -1. Isomers were obtained from fractional crystallization/solubility experiments; ppt1 (step 5, Figure 3 -1), and ppt 2 (step 8, Figure 3 -1). Their purity was determined by 1H NMR spectroscopy, experimental data was determined from material recovered Figure 3 -1 for compounds (a) 3a, (b) 3b, and (c) 3c. ..........................................................................................121 Table 5 -2. Melting temperature (T m) and the heat of fusion ( !Hm) for compounds 3a-c as determined by differential scanning Calorimetry, from T = 40 ¡C Ð350 ¡C with a heating rate of 10 ¡C/min .....................................................................................................................................127 Table 5 -3. % Relative error (% RE) determined by a +/ - 5% confidence interval in 1H NMR measurements ...............................................................................................................................131 Table 5 -4. Binary interaction coefficients for 3a, 3b, and 3c ......................................................133 Table 5 -5. Result of Schrıder -van Laar equation (RHS) at room temperature and the corresponding solubility limits based on ideal solution assumptions ( " = 1.00) for compounds 3a, 3b, and 3c .....................................................................................................................................133 Table 6 -1. Composition, in % trans isomer, of the eluent fraction obtained in column chromatography for compound 3c ...............................................................................................145 Table 6 -2. Solvents used for crystallization of compounds 3a, 3b, and 3c .................................147 Table 6-3. Melting temperature and the heat of fusion for compounds 3a-c as determined by differential scanning Calorimetry, from T = 40 ¡C Ð350 ¡C with a heating rate of 10 ¡C/min ...148 Table 6 -4. Binary interaction parameters of compounds 3c and 3a ............................................162 Table 6 -5. Experimental melting temperatures (T experimental ) determined from DSC apparatus vs. calculated melting temperatures (T calculated ) determined from an ideal solution to the Schrıder -van Laar equation for an isomorphous solution (equation 4 -2) with an ideal approximation, where " = 1.00 of compound 3b .........................................................................166 Table 7 -1. DSC data for polyimide thermosets ...........................................................................194 Table 7 -2. Complex viscosity for compound 5d .........................................................................198 Table 7 -3. Polymer experiments ..................................................................................................203 Table 7 -4. Additional polymer reactions .....................................................................................203 xv Table 7 -5. Mn for select polymers that are soluble in DMF ........................................................205 Table 7 -6. Intrinsic viscosity measu remen ts of selected polymers in NMP ................................208 Table 7 -7. Onset of degradation (T d) of Nomex polymers ..........................................................209 Table 7 -8. Tg and T m of selected Nomex polymers ....................................................................211 Table A -1. Deprotection analysis ............................................................................................... 225 Table A -2. Parameters varied in an attempt to improve the yield of the dichlorocyclohexylsilane to Ph 8tetrasilanol coupling reaction (Compound 3b) ..................................................................228 Table B -1. 29Si T1 analysis; peaks can be seen in FigureII -3a ...................................................238 Table B -2. 1H T1 analysis; peaks can be seen in FigureII -3b .....................................................239 Table C -1. Total mass balance of the cis and trans isomers for compounds 3a, 3b, and 3c .......252 Table C -2. The affect of the isomeric ratio in the SM on the MB of the cis and trans isomers for compound 3c ................................................................................................................................253 Table C -3 Retardation factor determined from equation III -2, using TLC plates with spots of compounds 3a-c dissolved in dichloromethane ...........................................................................254 Table D -1. Crystallographic data of compounds 3c ....................................................................267 Table D -2. Crystallographic data of compounds 3a ....................................................................268 Table D -3. Cryst allographic data of compounds 3b ....................................................................269 xvi LIST OF FIGURES Figure 1 -1 Double Decker Sil sesquioxane (DDSQ). For interpretation of the references to color in this and all other figures, the reader is referred the reader is referred to the electronic version of this dissertation ............................................................................................................................2 Figure 1-2. Tran s and cis isomers of DDSQ ....................................................................................3 Figure 2 -1. Examples of (a) silsesquioxanes, (b) silsesquioxane isomers, and (c) a silsesquioxane application. .....................................................................................................................................10 Figure 2 -2. Cage -like silsesquioxane .............................................................................................11 Figure 2 -3. Selected examples of monofunctionalized, corner -capped SQs .................................14 Figure 2 -4. Example of a difunctionalized SQ from a disilanol ....................................................17 Figure 2 -5. Isomers of difunctionalized SQs .................................................................................18 Figure 2 -6. Twisted SQ polymers ..................................................................................................19 Figure 2 -7. Fully condensed DDSQ structures ..............................................................................20 Figure 2 -8. Poly mer AA from DDSQ A ........................................................................................20 Figure 2 -9. Polymer AB from DDSQ A ........................................................................................21 Figure 2 -10. Polymer BA from DDSQ B ......................................................................................22 Figure 2 -11. Material applications from fully condsensed C ........................................................23 Figure 2 -12. Structure CD and polymer CE from DDSQ C ..........................................................24 Figure 2 -13. Polymer CF from DDSQ C .......................................................................................25 Figure 2 -14. Co-polymer CG from DDSQ C .................................................................................26 Figure 2 -15. Compound CH and Polymer CI from DDSQ C ........................................................26 Figure 2 -16. Polymer DA from DDSQ D ......................................................................................28 Figure 2 -17. Polymer DC from DDSQ D ......................................................................................29 Figure 2-18.25 29Si NMR spectra of (a) cis and trans , (b) majority trans , and (c) majority cis DDSQ D .........................................................................................................................................30 xvii Figure 2 -19. DDSQ silicon atom labels .........................................................................................31 Figure 2 -20. Representation of a hetereonuclear spectrum ...........................................................33 Figure 2 -21. 2J-coupled Si -H atoms (red, top), and 3J-cou pled Si -H atoms (blue, bottom ) .........33 Figure 2 -22. 1H-1H COSY spectrum. Correlation peaks that are not on the diagonal represent J -coupled peaks .................................................................................................................................34 Figure 2 -23.90 Interaction energy of the molecules in the Wilson activity coefficient model, a central molecule of type 1 (left), and a cen tral molecules of type 2 (rig ht) ...................................38 Figure 2 -24. Flow chart for determining calculated xl and !l ........................................................41 Figure 2 -25. Meta - and para - anisylpinacolone .............................................................................45 Figure 2 -26.107 Binary phase diagram of solids A and B .............................................................46 Figure 2 -27. Phase diagram representing two systems of co -crystallization, congruent melting system (left), and incongruent melting system (right ....................................................................48 Figure 2 -28. Phase diagram of an isomorphous, solid -solution melting system ...........................49 Figure 2 -29. An example of a phenylethynyl group ......................................................................52 Figure 2 -30.119 Oligoimide structures used to reduce viscosity ...................................................53 Figure 2 -31.25 Phenylethynyl end -capped SQs .............................................................................58 Figure 2 -32. Kevlar (PPD -T) and Nomex (MPD -I) polymer structures ........................................59 Figure 2 -33.137 Stereoisomers of a polyamide ..............................................................................63 Figure 2 -34.138 Diffusion of water into the chain of a polyaramid ...............................................63 Figure 3 -1. Trans (left) and cis (right) isomers of DDSQ(X)(R ) ..................................................77 Figure 3 -2. Cis and trans isomers of compounds 3a-d ..................................................................78 Figure 3 -3. 1H NMR spectrum of ( N-trimethylsilyl)2 -aniline -3-(dichloromethylsilane ) .............81 Figure 3 -4. NMR spectra of compound 3a (a) 29Si and (b) 1H .....................................................82 xviii Figure 3 -5. 1H NMR spectrum of ( N-trimethylsilyl)2 -aniline -3-(cyclohexyl dichlorosilane ) ......83 Figure 3 -6. NMR spectra of compound 3b (a) 29Si and (b) 1H ....................................................84 Figure 3 -7. 1H NMR spectrum of ( N-trimethylsi lyl)2 -aniline -4-(cyclohecyl dichlorosilane ........85 Figure 3-8. NMR spectra of compound 3c (a) 29Si and (b) 1H .....................................................86 Figure 3 -9. NMR spectra of compound 3d (a) 29Si and (b) 1H ....................................................88 Figure 4 -1. Cis and trans isomers of (a) 3a, (b) 3b, (c) 3c ............................................................96 Figure 4 -2. 29Si NMR spectra of (a) cis/trans 3a, (b) majority trans 3a, (c) majority cis 3a, (d) cis/trans 3b, and (e) cis/trans 3c ....................................................................................................99 Figure 4 -3. 1H NMR spectrum of (a) cis/trans 3a, (b) majority trans 3a, (c) majority cis 3a, (d) cis/trans 3b, (e) cis/trans 3c, (f) majority trans 3c, (g) majority cis 3c .......................................100 Figure 4 -4. 1H- 29Si gHMBC connectivity of (a) cis/trans 3a and (b) cis/trans 3c ....................102 Figure 5 -1. Flow chart describing the fractional crystallization/isomer separation procedure of trans and cis isomers of compounds 3a, 3b, and 3c ....................................................................116 Figure 5 -2. Experimental, circle ( !) = cis, square ( ") = trans , and modeled, solid line ( !) = cis, dashed line ( ---) = trans , solubility limits in a Hexanes to THF (XH : XT) solvent solution for isomers of compound 3a ..............................................................................................................123 Figure 5 -3. Experimental, circle ( !) = cis, square ( ") = trans , and modeled, solid line ( !) = cis, dashed line ( ---) = trans , solubility limits in a Hexanes to THF (XH : XT) solvent solution for isomers of compound 3b ..............................................................................................................125 Figure 5 -4. Experimental, circle (!) = cis, square ( ") = trans , and modeled, solid line ( !) = cis, dashed line ( ---) = trans , solubility limits in a Hexanes to THF (XH : XT) solvent solution for isomers of compound 3c ..............................................................................................................126 Figure 5 -5. Example of a melting endotherm ( trans 3a) showing the melting temperature and heat of fusion as determined by differential scanning Calorimetry, from T = 40 ¡C Ð350 ¡C with a heating rate of 10 ¡C/min ..........................................................................................................128 Figure 5 -6. Modeled solubility limits in a Hexanes to THF (XH : XT) solvent solution for the cis and trans isomers of compounds 3a, 3b, and 3c. The three upper curves represent cis and the lower three curves represent trans ...............................................................................................130 xix Figure 5 -7. Activity coefficients of the solute vs. the mole fraction of (a) anti -solvent (hexanes) and (b) solvent (THF), when each compound is considered a binary. When the natural logarithm of the activity coefficient is larger than zero, the interaction is between DDSQ and hexanes. When the natural logarithm of the activity coefficient is below zero, the interaction is between DDSQ and THF ...........................................................................................................................132 Figure 6 -1. Melting endotherms of trans and cis isomers for compounds 3a-c; from T = 225 ¡C Ð325 ¡C with a heating rate of 10 ¡C/min. All isomers have the same scale bar as (a) unless otherwise noted ............................................................................................................................149 Figure 6 -2. Cooling exotherms of trans and cis isomers for compounds 3a-c; from T = 120 ¡C Ð290 ¡C with a cooling rate of 5 ¡C/min. All isomers have the same scale bar as (a ) ..................150 Figure 6 -3. Second heating of cis 3a, which demonstrates a glass transition (T g), and a recrystallization (T r); from T = 45 ¡C Ð200 ¡C with a heating rate of 10 ¡C/min .......................152 Figure 6 -4. Hydrogen bonding (dotted line) in the crystal lattice of cis 3a; red = O, blue = Si, black = C, light blue = N, and pink = H. Phenyl moieties and other HÕs not shown for simplicity ......................................................................................................................................153 Figure 6 -5. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3c, plotted as melting temperature (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental T m measured with DSC a pparatus, star ( ) = experimental T p, starburst ( !) = experimental T c, triangle ( !) = experimental T E, circle ( ") = experimental T m-E, solid line ( Ñ) = calculated T m from NRTL model (equation 4 -2), and dash dot ( úøúø) = ideal Tm, from Schrıder -van Laar equation (equation 4 -1) where ! = 1.00. .......................................157 Figure 6 -6. Melting behavior of binary cis/trans mixtures for compound 3c; T = 225 ¡C Ð325 ¡C with a heating rate of 10 ¡C/min. All thermal traces have the same scale bar as 100 % trans unless otherwise noted .................................................................................................................158 Figure 6 -7. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3a, plotted as melting temperature (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental T m measured with DSC apparatus, ex (x) = experimental T p, triangle (!) = experimental T E, circle ( ") = experimental T m-E, solid line ( Ñ) = calculated T m from NRTL model (equation 4 -2), and dash dot ( úøúø) = ideal T m, from Schrıder -van Laar equation (equation 4 -1) where ! = 1.00 ......................................................................................................160 Figure 6-8. Melting behavior of binary cis/trans mixtures for compound 3a; T = 225 ¡C Ð325 ¡C with a heating rate of 10 ¡C/min. All thermal traces have the same scale bar as 80 % trans unless otherwise noted ............................................................................................................................161 xx Figure 6 -9. Activity coefficients ( !) of isomer ( i) vs. the mole fraction of ( xtrans ) for compounds (a) 3c and (b) 3a; left hand side i = cis isomers, and right hand side, i = trans isomer; square ( !) = experimental xi, and solid line ( Ñ) = calculated xi ..................................................................163 Figure 6 -10. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3b, plotted as melting temperature (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental T m measured with DSC apparatus, solid line ( Ñ) = liquidus line, dashed line ( ---) = solidis line. Solidus and liquidus lines are determined from an ideal solution to the Schrıder -van Laar equation for an isomorphous solution (equation 4 -3) with an ideal approximation, where ! = 1.00 ....................................................................................................164 Figure 6 -11. Melting behavior of binary cis/trans mixtures for compound 3b; T = 250 ¡C Ð300 ¡C with a heating rate of 10 ¡C/min. All thermal traces have the same scale bar as 100 % trans unless otherwise noted .................................................................................................................165 Figure 6 -12. Solidification behavior of binary cis/trans mixtures for compound 3c; T = 270 ¡C Ð120 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as the trace immediately above it unless otherwise noted ..............................................................................167 Figure 6 -13. Solidification behavior of binary cis/trans mixtures for compound 3a; T = 300 ¡C Ð170 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as the trace immediately above it unless otherwise noted ..............................................................................168 Figure 6 -14. Solidification behavior of binary cis/trans mixtures for compound 3b; T = 230 ¡C Ð180 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as 100 % trans .............................................................................................................................................169 Figure 6 -15. Undercooling (T m-Tc) of binary cis/trans mixtures vs. xtrans for compounds (a) 3c (b) 3a, and (c) 3b .........................................................................................................................179 Figure 7 -1. Compoun ds (a) 5a- all meta, (b) 5c- all para, (c) 5d- one side meta and one side para. All compounds also have cis and trans isomers about the SQ cage with respect to the D - Silicon, as seen in the previous chapters for compounds 3a, 3b, and 3d ..................................................189 Figure 7 -2. NMR spectra of compound 5a (a) 29Si and (b) 1H, the resonances at "H 3.5 and 1.9 ppm are residual methanol ...........................................................................................................190 Figure 7 -3. NMR spectra of compound 5c (a) 29Si and (b) 1H ...................................................191 Figure 7 -4. NMR spectra of compound 5d (a) 29Si and (b) 1H ..................................................192 Figure 7 -5. DSC curves of (a) 5a, (b) 5c, (c) trans 5c, (d) cis 5c, and (d) 5d, from 30 to 500 ¡C with a constant heating rate of 10 ¡C / min ..................................................................................195 xxi Figure 7 -6. Complex viscosity of two batches of compound 5d; filled square ( !) = batch 1, and unfilled square ( ) = batch 2 ......................................................................................................197 Figure 7 -7. 1H NMR spectrum of silylated MPDA .....................................................................202 Figure 7-8. 1H NMR spectra of (a) 0 mole % DDSQ (Nomex), (b) 50 mole % DDSQ, and (c) 100 mole % DDSQ ......................................................................................................................204 Figure 7 -9. Extrapolation to intrinsic viscosity for 1.1 mass % DDSQ .......................................207 Figure 7 -10. TGA analysis of select polymers, (a) degradation (b) derivative of the degradation ...................................................................................................................................210 Figure 7 -11. Films made from select polymers (a) slow evaporation, (b) vacuum oven, and (c) UV lamp .......................................................................................................................................212 Figure 7 -12. 1H NMR of (a) 6a and (b) 7a; resonances at 3.72 and 1.83 ppm are residual THF ..............................................................................................................................................216 Figure 7 -13. Product without enough MeI ...................................................................................216 Figure A -1. Deprotection analysis, while investigating the concentrations of (a) DDSQ and (b) acetic acid, and (c) the ratio of X DEE : X MeOH vs. % yield .......................................................226 Figure A -2. Comp ound 3b impurities ..........................................................................................228 Figure B -1. Europium complex, where B: is the aminophenyl group .........................................236 Figure B -2. 1H NMR spectra of increasing content of europium; reported as Compound 3c : Eu(FOD)3; (a) 1 : 0, (b) 10 : 1, (c) 4 : 1, (d) 2 : 1, (b) 1 : 1 .........................................................237 Figure B -3. Spectra of compound 3a (a) 29Si NMR spectrum and (b) 1H NMR spectrum ........240 Figure B -4. N-silylated amines (a) 1H NMR spectrum and (b) 29Si NMR spectrum .................241 Figure B -5. 1H NMR spectrum of compound 3d ........................................................................242 Figure B -6. 29Si NMR spectrum of compoun d 3d; (a) D -group silicon atoms, (b) T -group silicon atoms, and (c) whole spectrum ....................................................................................................244 Figure B -7. LCMS spectra of compound 3d (top), 3c (middle), and 3a (bottom ) ......................247 Figure D -1. Crystal structures for compound 3c .........................................................................270 xxii Figure D -2. Crystal structures for compound 3a .........................................................................271 Figure D -3. Crystal structures for compound 3b .........................................................................272 Figure E -1. 29Si NMR spectrum of trans 5c that is not a pure product ......................................286 Figure E -2. Anticipated impurity in the synthesis of 5 ................................................................286 Figure E -3. 29Si NMR spectru m of compound 5d .....................................................................288 Figure E -4. 1H NMR spectrum of silylated aniline .....................................................................289 Figure F -1. Compound (a) 3a-all meta -structure, (b) 3c-all para -structure, and (c) 3d-one side meta- and one side para -structure. ...............................................................................................294 Figure F -2. DDSQ(p ropylamine)(Me). ........................................................................................295 xxiii LIST OF SCHEMES Scheme 2 -1. Synthetic Scheme for corner capping a trisilanol SQ; LG can be a variety of leaving groups including: Cl, OMe, OEt; R can be: Phenyl, cyclohexyl, cyclopentyl, isobutyl, isooctyl ...........................................................................................................................................12 Scheme 2 -2. Synthetic pathway to a side -capped SQ through the reaction of disilanol and silane with two LGs ..................................................................................................................................16 Scheme 2 -3. Synthetic schemes for difunctionalized SQs .............................................................16 Scheme 2 -4. Nomex synthesis using (a) DA with PCL, and (b) N -silylated DA with PCL .........59 Scheme 2 -5. Mechanism for reaction of N -silylated DA and PCL ...............................................60 Scheme 2 -6. Synthesis of SQ ionic liquids ....................................................................................65 Scheme 3 -1. Synthesis of DDSQ(AP)(R); X = Cl/Br, R = Me/Cy ................................................80 Scheme 7 -1. Synthesis of compound 5 ........................................................................................188 Scheme 7 -2. DDSQ -based Nomex scheme .................................................................................201 Scheme 7 -3. Synthesis of IL, 6a ..................................................................................................215 Scheme A -1. Deprotection of aminophenyl groups .....................................................................225 Scheme E -1. Model reaction for Nomex .....................................................................................289 xxiv KEY TO SYMBOLS AND ABBREVIATIONS Common Acronyms SQ silsesquioxane sil silicon sesqui each silicon atom is bound to an average of 1.5 oxygen atoms ox oxygen ane each silicon atom is bound to one hydrocarbon group DDSQ double decker silsesquioxane SLE solid -liquid equilibrium LLE liquid -liquid equilibrium MW molecular weight NRTL non-random two liquid HPM high performance material PEPA phenylethynylphthalic anhydride PEPI phenylethynylphthalic imide PETI phenylethynyl -terminated polyimide oligomer Mmv minimum viscosity PPD-T poly(para -phenylene terephthalamide); Kevlar¨ MPD-I poly(meta -phenylene isophthalamide); Nomex¨ MDA meta-diamines PDA para -diamines PCL phthaloyl chloride xxv H-bond hydrogen bond IL ionic liquid Common Apparatuses DSC differential scanning calorimetry XRD x-ray diffraction NMR nuclear magnetic resonance TGA thermal gravi metric analysis GPC gel permeation chromatography Common Symbols Tg glass transition temperature Tm crystalline melting transition temperature Td degradation temperature Tr recrystallization temperature upon heating Tc crystallization temperature upon cooling TE eutectic temperature Tm-E meta-stable eutectic temperature TP peritectic temperature ! activity coefficient "Cp heat capacity "Hm heat of fusion; enthalpy of the melt transition "Sm entropy of fusion xxvi !Hc enthalpy of the crystallization transition 1 !"#$%&'() ! "#$%&'()$"&# ! Keywords double decker silsesquioxane, cis/trans isomers 2 !" #$%&'()*%+'$ ,!"!,-./&011,2 '01 The goal of this Ph.D. dissertation was to develop a fundamental understanding of a class of double decker silsesquioxane (DDSQ). This understanding was obtained through examining how configurational and structural modifications to these macromolecules influence chemical and physical characteristics . An appropriate st ructure to property relationship was established for the selected D DSQ (Figure 1-1) and was ultimately applied to create novel engineering designs. Figure 1-1. Double Decker Sil sesquioxane (DDSQ). For interpretation of the r eferences to color in this and all other figures, the reader is referred the reader is referred to the electronic version of this dissertation. 1.2 Motivation Hybrid inorganic/organic nano -structures play a significant role in the advancement of new technology through the development of a large variety of new materials, such as : polymers, 1 sensors, 2 fuel cells, 3 semi-conductors, 4 catalysts, 5 and biological materials. 6 Furthermore , they provide a synergistic combination of the flexibility and reactivity of organic networks with the excellent thermal and mechanical properties of inorganics. Three -dimensional (3D) nano -structures such as fullerenes, 7 oxo/alkoxide transition metal clusters, 8 and cage -like SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiRSiRNH2H2N 3 silsesquioxanes 9 increase the surface area of the hybrid segments when incorporat ed into any of these materials, which can enhance these hybrid affects. Of the se 3D architectures, cage -like silsesquioxanes are the easiest to functionalize and purify, do not contain toxic transition metals, and display mono -dispersed physical and chemic al characteristics. 9 Superior properties over their organic polymer s have been displayed by these cage -like silsesquioxanes in areas such as: solubility, 10-12 thermal and mechanical properties, 13-15 flame retardance, 16-18 dielectric properties ,19-21 and oxidative resistance 22-24. However, with most hybrid polymers investigated, the attached SQs are pendant with respect to the polymer backbone. A recently developed class of these nano -structured cage -like silsesquioxanes, for mally known as double decker silsesquioxanes (DDSQ) offers the opportunity to form hybrid polymers with SQ cages as a part of the polymer backbone. 25 During the capping reaction, these functionalized DDSQs generate cis and trans isomers with respect to the 3D Si -O core (Figure 1-2). Therefore, it is logical to Figure 1-2. Tran s and cis isomers of DDSQ. 4 derive explicit understandings for the characteristics of the individual isomers, which will allow for optimization of capping reaction parameters, particularly if one isomer is favored over the other. Moreover, these characteristics are also relevant when reacting or incorporating these isomers, or mixtures thereof, with other molecules to form novel materials. Consequently , a structure/property relationship becomes essential . Therefore, the motivation of this the sis was to develop an appropriate structure/property relationship for these DDSQ , thus making it possible to systematically enhance material properties from a bottom -up approach. The particular reactive moiety (X) in Figure 2 of thes e DDSQ not only has a significant impact on the characteristics of the molecule itself, but it also limits the applications. An aminophenyl moiety (Figure 1-1) was se lected for this dissertation as the specific reactive moiety for the following reasons: 1. The aminophenyl moiety can be thought of as a reactive version of the phenyl moieties attached to each silicon atom. 2. The amine is particularly reactive and has been used in the development for a variety of applications of engineering importance in areas such as: high performance polymers, 26 ionic -liquids, 19 photoc hemistry, 27 and catalysis 28. !"#$%&'()*+',)-$.(+)$)'($/0+12321$42*5 The fundamental idea of this dissertation was to use a thermally stable, functionalized double -decker silsesquioxane (DDSQ) that can be chemically modified , and develop a structure/property relationship based upon the altered properties of the modified structu re. This study has address ed fundamental questions regarding how the isomers of functionalized DDSQ monomers impact their chemical, physical, and structural properties, thus allowing for the logical 5 introduction of regio - and diastereo -isomeric functionali ty, providing for novel e ngineering designs . Specific aims of this project include: 1. Careful design and analy sis of the synthetic procedures of selected DDSQ. 2. Spectroscopic elucidation of the stereoisomers of each DDSQ for quantification . 3. Determinin g solubility limits and investigating solubility behavior of the individual isomers , such that concise separation techniques could be developed . 4. Development of phase diagrams representing the solid -liquid melt equilibria of the individual isomers and binary mixtures . 5. Exploring material applications based on the DDSQ from this work. Quantifiable structure -property -performance relationships of DDSQ molecules were developed based on the specific regio - and diastereo -isomers . This res earch endeavor ed to advance the general understanding of how introducing isomerism influences three -dimensional macromolecular assemblie s: including fullerenes, carboranes, and oxo/alkoxide transition metal clusters . After this work was completed, a fundamental structu re/property relationship was developed from this bottom -up approach to chemical engineering and materials scien ce. 6 !!!!! "#$#"#%&#' ! 7 !"#"!"$%"& ' (1) Allcock, H. R. Advanced Materials 1994, 6, 106. (2) Wang, S.; Kang, Y.; Wang, L.; Zhang, H.; Wang, Y.; Wang, Y. Sensors and Actuators B -Chemical 2013, 182, 467. (3) Laberty -Robert, C.; Valle, K.; Pereira, F.; Sanchez, C. Chemical Society Reviews 2011, 40, 961. (4) Zhao, L.; Lin, Z. Advanced Materials 2012, 24, 4353. (5) Diaz, U.; Brunel, D.; Corma, A. Chemical Society Reviews 2013, 42, 4083. (6) Kellermeier, M.; Coelfen, H.; Manuel Garcia -Ruiz, J. European Journal of Inorganic Chemistry 2012, 5123. (7) Sergio, M.; Behzadi, H.; Otto, A.; van der Spoel, D. Environmental Chemistry Letters 2013, 11, 105. (8) Turova, N. Y. Uspekhi Khimii 2004, 73, 1131. (9) Hartmann -Thompson, C.; SpringerLink (Online service) In Advances in Silicon Science, ; Springer Neth erlands,: Dordrecht, 2011, p XXVII. (10) Gnanasekaran, D.; Reddy, B. S. R. Polymer Composites 2012, 33, 1197. (11) Guenthner, A. J.; Lamison, K. R.; Lubin, L. M.; Haddad, T. S.; Mabry, J. M. Industrial & Engineering Chemistry Research 2012, 51, 12282. (12) Rizvi, S. B.; Yildirimer, L.; Ghaderi, S.; Ramesh, B.; Seifalian, A. M.; Keshtgar, M. International journal of nanomedicine 2012, 7, 3915. (13) Yang, B.; Li, M.; Wu, Y.; Wan, X. Polymers & Polymer Composites 2013, 21, 37. (14) Wu, S.; Hayakawa, T.; Kakimoto, M.; Oikawa, H. Macromolecules 2008, 41, 3481. (15) Wu, J.; Mather, P. T. Polymer Reviews 2009, 49, 25. (16) Fan, H.; Yang, R. Industrial & Engineering Chemistry Research 2013, 52, 2493. (17) Rakesh, S.; Dharan, C. P. S.; Selladurai, M.; Sudha, V.; Sundararajan, P. R.; Sarojadevi, M. High Performance Polymers 2013, 25, 87. 8 (18) Vahabi, H.; Eterradossi, O.; Ferry, L.; Longuet, C.; Sonnier, R.; Lopez -Cuesta, J. M. European Polymer Journal 2013, 49, 319. (19) Cardiano, P.; Lazzara, G.; Manickam, S.; Mineo, P.; Milioto, S.; Lo Schiavo, S. European Journal of Inorganic Chemistry 2012, 5668. (20) Geng, Z.; Ba, J.; Zhang, S.; Luan, J.; Jiang, X.; Huo, P.; Wang, G. Journal of Materials Chemistry 2012, 22, 23534. (21) Ke, F.; Zhang, C.; Guang, S.; Xu, H. Journal of Applied Polymer Science 2013, 127, 2628. (22) Vila Ramirez, N.; Sanchez -Soto, M. Polymer Composites 2012, 33, 1707. (23) Jin, L.; Ishida, H. Polymer Composites 2011, 32, 1164. (24) Blanco, I.; Abate, L.; Bottino, F. A.; Bottino, P. Polymer Degradation and Stability 2012, 97, 849. (25) Takashi, K.; Takashi, K.; Masaya, I.; Kazuhiro, Y.; Yasuhiro, Y. Japan, 2006. (26) Garcia, J. M.; Garcia, F. C.; Serna, F.; de la Pena, J. L. Progress in Polymer Science 2010, 35, 623. (27) Slegt, M.; Overkleeft, H. S.; Lodder, G. European Journal of Organic Chemistry 2007, 5364. (28) Wang, L.; Du, W.; Wu, Y.; Xu, R.; Yu, D. Journal of Applied Polymer Science 2012, 126, 150. 9 !"#$%&'() ! "#$%&'()*+!,*-('.#/,(* ! Keywords double decker silsesquioxane, NMR, cis/trans isomers, separations, phase diagrams, eutectic, isomorphous, polyamide, polyimide 10 !" #$%&'()*+,-.+/)(0$12)+ - This section includes information on the following topics: section 2.1 outlines general information on silsesquioxane (SQ) macromolecules. Section 2 .2 discusses the necessity of spectroscopic elucidation using 1H-NMR spectra . Section 2 .3 examines information relevant to isomer separation techniques. Section 2 .4 explores the impact of molecular structure on melting temperatures, melting enthalpies, solidification, crystallization, and solubility. Finally, section 2.5 highlights the motivation for the selected material applications. 2.1 Silsesquioxanes a) 2.2 Spec troscopic Elucidation 2.3 Isomer Separation 2.4 Structural Characterizaion b) 2.5 Selected Applications c) Figure 2-1. Examples of (a) silsesquioxane s, (b) silsesquioxane isomers , and (c) a silsesquioxane application. SiOSiOSiSiOSiOSiOSiOSiOXRRRRRRROOOOOSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOXSiRSiRX 11 !"#$%&'()(*+&,-./)( $ Fully condensed cage -like SQs with different functionalities have revealed superior performance over their organic counterparts in areas such as : thermal and mechanical properties, 1-4 flame retardance, 5-13 solubility, 14-16 oxidation resistance, 17-20 and dielectric properties. 21-23 These benefits are mostly attributed to molecular level reinforcement, and the ceramic -like properties of the inorganic cage. 24 SQs are comprised of covalently bonded oxygen to silicon atoms in a three to two (sesqui) ratio. Additionally, e ach silicon atom is covalently bonded to an organic peripheral group (R), which allows these molecules to interact with themselves, or other organics in the medium (Figure 2-2).25 Addi tionally, one or more of these peripheral groups can be modified to make the molecule reactive. Figure 2-2. Cage -like silsesquioxane. It has been demonstrate d that the incorporation of SQs into material sy stems decouples strength from toughness, while simultaneously provi ding the same or better thermal and mechanical properties , as a result of the inherent flexibility of the Si -O cage .26,27 This feature has been exp loited to decrease stiffness i n brittle polymers. Furthermore, SQs are easy to functio nalize, and incorporate in to material systems. The combination of all these fe atures distinguishes SQs from other three -dimensional macromolecular assemblies , which tend to be 12 stiff and/or difficult to functionalize , including: fullerenes, carboranes, and oxo/alkoxide transition metal clusters. An entire background in the field of SQs would require a published volume, and is beyond the scope of this dissertation. Therefo re, several SQs have been selected to demonstrate advances in this field and are presented in the following sections. For a more detailed review, the reader is encouraged to refer to a recently published volume on cage -like SQs. 28 !"#"#$%&'&()'*+,&'-.,/01 $!"#"#"#$2&3'03 4*-5501$67 $ Corner capping is t he most common m ethod for modifying one of the peripheral groups on a cage -like SQ . Corner capping is accomplished through reacting a trisilan ol SQ with a trichlorosilane, dimethoxysilane, diethoxysilane, or a silane with two similar leaving group s (Scheme 2-1).29 In a corner -capped SQ, an R-moiety is replace d by an X- (or reactive) moiety. These structures are also known as a monofunctionalized SQ, since only one R -moiety is replaced . Specific R-moieties that have been functionalized by corner capping include: Scheme 2-1. Synthetic Scheme for corner capping a trisilanol SQ; LG can be a variety of leaving groups including: C l, OMe, OEt; R can be: Phenyl, cyclohexyl, cyclopentyl, isobutyl, isooctyl . SiOSiOSiOHSiSiOSiOSiORRRRRRROOOOHOHO+SiLGLGLGXtrisilanol SQSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiXX = H, Cl, OH,nitriles, amines, isocyanates, styryls, olefins, acrylics, epoxides, norbynyls, bisphenols, acid chlorides, alcohols, acids 13 hydride, chloride, hydroxide, nitriles, amines, isocyanates, styryls, olefins, acrylics, epoxides, norbornyls, bisphenols, acid chlorides, alcohols, acids. 29 Specific functionalized corner -capped SQs have been used in many applications (Table 2-1, Figure 2-3). Particularly , the addition of corner -capped SQs into organic monomers and polymers has realized an increase in glass transition (Tg), melting (Tm), and decomposition (T d) temperature s, and increases in th e shear storage modulus (E !) (Table 2-1, Figure 2-3). It has been suggested that the bulky, inorganic SQ cages retard segmental motion of the polymers through interchain interaction. 29 Table 2-1. Thermal -mechanical data for examples of monofunctionalized, corner -capped SQs. Compound Mol % SQ Tg (¡C) Tm (¡C) Td (¡C) E' (Pa) Homopolymer A30 - 396 - 445 - Copolymer B30 9 132 - 402 > 1000* Organic Polymer poly(4 -methylstyrene) - 116 - - 14.45* Homopolymer C31 - None + None + 388 - Organic Polymer PMMA - 43-163 - 200 - Copolymer D32 7.7 81 - - - Organic Polymer polynorbornene - 52.3 - - - Copolymer E33 0.7 - 125 410^ - Organic Polymer polyethylene - - 132 313 - Monomer F34 - None 270 - - Organic Monomer propyl amine - None - 83 - - Copolymer F34 2.5 80 ~ 200 - > 300# Organic Polymer Nylon 6 - 65 ~ 200 - 225 Copolymer G35 20 -45 - 355** - Organic Polymer poly(ethylene imine) - - 51 - 355 - Copolymer H36 3.2 245.3 - - - Organic Polymer PHS-PVP - 194.9 - - - * Frequency = 0.1 rad/s and T = 180 ¡C. + None observed between 0 Ð 400 ¡C. ^ 30 % weight loss under air. #Frequency = 1 Hz and T = 100 ¡C. ** Although Td is similar, the rate for the SQ containing polymer is much lower. ^^PHS-PVP: poly(hydroxystyrene -co -vinylpyrrolidone -co isobutylstyryl. For additional examples please see the following reference: 28 14 R = c-C6H11 R = c-C6H11 R = c-C6H11 R = c-C6H11 x = 0.91; y = 0.09 mA30 B30 C31 D32 R = ethyl R = isobutyl R = CH2CH2CF3 E33 F F34 G37 R = isobutyl R = isobutyl R = hepta(isooctyl) H36 I I27 J21 Figure 2-3. Selected examples of monofunctionalized, corner -capped SQs. SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSi**nSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSi*y*xnSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiOO**CH3nSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSi*SiOSiOSiOSiSiOSiOSiORRRRRROOOOOOSiRn*mSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiNH2SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiOOSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSi**HOOSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiHOOHSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiN+I 15 In addition to have a higher decomposition temperature, polymers with corner -capped SQs also exhibit a decrease in decomposition rate. Upon degradation, the organic groups of SQs initia lly degrade rapidly, but do not volatize (Table 2-1, Figure 2-3). Instead, they form an effective char -like coating on top of the still intact Si -O cage, providing extra protection .38 This is specifical ly useful in increasing the limiting oxygen intake (LOI) , or the volume fraction of oxygen necessary for a material to sustain combustion, 39 for flame retardant applications. SQs have also been incorporated into polyurethane (PU) polymers (e.g. I , Figure 2-3) for use in vascular prostheses. 27 These SQ -PUs have demonstrated increased endothelial cell adhesion, proliferation, and growth, wh ile displaying no signifi cant increase in toxicity. Moreover, these SQs provide flexibility and further stabilize the prostheses. Additional biomedical applications include, but are not limited to: SQs for drug delivery 40,41 and scaffolds for tissue engineering. 42-47 Furthermore , SQs have low dielectric constants . This feature has been exploited in applications relating to ele ctronics and energy .28 Recently , these latter properties have been advantageous in the design of a corner -capped SQ for potential use in fuel cell s, which displays characteristics of an ionic liquid (e.g. J, Figure 2-3).21 !"#"#"!$%&'( )*+,,('$%-. $Another form of mono functionalized cage -like SQs are side -capped SQs. These are synthesized through the reaction of a disilanol and a dichlorosilane ( Scheme 2-2). Although the potential of side -capped SQs have been reported in the literature, there h ave not been significant applications based on this synthetic method. 28 16 Scheme 2-2. Synthetic pathway to a side-capped SQ through the reaction of disilanol and silane with two LGs. !"#"!$%&'()*+&,)-.&/01$234 $ Difunctionalized SQs have appeared more recently in the literature. 2,3,48 -54 There are three major silanol starti ng materials that have been used to synthesize difunctional SQs , including: two types of disilanol SQs, 48,54 and one type of tetrasilanol 25 (Scheme 2-3). Scheme 2-3. Synthetic schemes for difunctionalized SQs. SiLGR'LGXdisilanol SQSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOSiROHOHSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOSiROO+SiR'Xdisilanol SQSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOSiROHOHSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOSiROOBF3OEt2FFR' LiSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOSiROOR'R'HBF 4OMe2R = i-Pr, i-Bu, c-C5H9, c-C 6H11twisted disilanol SQSiOSiOSiSiSiOSiOSiRRRRRRROOOOOOSiORHOHOSiOSiOSiSiSiOSiOSiRRRRRRROOOOOOSiORR'R'+R = i-Pr, i-Bu, c-C 5H9, c-C 6H11R'Si-X X = Cl or NMe2SiOSiOSiOHSiOSiHOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOHHOSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOXSiR'SiR'X+R'ClClXtetrasilanol DDSQ 17 Difunctional SQs are unique for their ability to become part of the main chain of a polymer, and have been called ÒbeadsÓ on a chain when polymerized, 54 whereas monofunctional SQs are pendant to the backbone of a polymer chain. Difunctional SQs synthesized from disilanol SQ have mostly been used for space application s ( Figure 2-4).28 These SQs have been incorporated into the polyimide Kapton ¨ to provide additional protection in lower Earth orbit to atomic oxygen (AO) for spacecraft applications. Polymers based on SQs have demonstrated 10 times higher durability than neat Kapton ¨, which has the highest resistance of conventional polymers towards active AO. Similarly to thermal degradation, when th ese SQs are exposed to AO their organic groups degrade and a silica (SiO 2) layer is preserved, providing protection from degradation to the degrade and a silica (SiO 2) layer is preserved, providing protection from degradation to the underlying polymer. 50 The erosion yield of the SQ -Kapton ¨ is as low as ~ 0.01 that of neat Kapton ¨, depending on the weight % of SQ in the polymer. Figure 2-4. Example of a difunctionalized SQ from a disilanol. The remaining two difu nctionalized SQs are uniqu e when comp ared to all other SQs mentioned so far. They are the only SQs capable of having isomers about the Si -O cage (Figure 18 2-5). The f irst is called Òtwisted SQÓ since the cage itself is twisted when compared to the other disilanol, and t here is an extra oxygen atom linking two s ilicon ato ms that were not previously bounded to one another . The second is forma lly known as double decker silsesquioxane (DDSQ) Figure 2-5. Isomers of difunctionalized SQs. because it is comprised of two ÒdecksÓ of silsesquioxanes stacked on top of one another .55 Twisted SQs are difficult to synthesize in large quantities. However, there have been a few functionalized twisted SQ polymers that have been synthesized ( Figure 2-6). The Td was significantly increased for A, T d = 490 ¡C, as compared to poly(di methyl)siloxane where the Td = 350 ¡C.56 Twisted S Qs have mostly been synthesized to contain organometallic, or inorganic segments (e.g. B). This dem onstrates the first rational synthetic construction of SQ polymers to linearly contain tran sition metals in the structure, and proved that air/moisture stable materials could be designed. However, due to the difficulty of scale -up for twisted SQs, research in this area is not prevalent. 19 A B Figure 2-6. Twisted SQ polymers. !"#"!"#$%&'()*$+*,-*.$/0)/*/1'0&234*/$5%%678 $DDSQs can be easily synthesized in large quantities, incorporated into a polymer back bone, and have cis and trans isomers about the Si -O cage . These attributes make it attractive for material applications where rational control of properties is desired . Several fully condensed DDSQs have been synthesized ( Figure 2-7) according to the previous schematic ( Scheme 2-3). Currently, only tetrasilanol with phenyl (Ph) moieties is synthetically available. DDSQs have also been synthesized to encapsulate metal nanoparticles similar to twisted SQs s tructures; however, that is beyond the scope of this dissertation and interested persons are encouraged to consult the literature for information regarding these materials. 57,58 These DDSQ structures (A -D) were fur ther reacted in order to incorporate into co -polymer systems ( Figure 2-8 through Figure 2-17). Compound A ( Figure 2-7) was used to synthesize polymer AA ( Figure 2-8) as a potential substrate for lightweight, microelectronic devices. 59 However, in order for this application to be feasible, the hydrophobic surface of films SiOSiOSiSiSiOSiOSiRRRRRRROOOOOOSiOOROR = c-C 6H11SiOSiOSiOSiSiSiOSiOSiRRRRRRROOOOOOSiOROSiHO*nSiOSiOSiSiSiOSiOSiRRRRRRROOOOOOSiOOROR = c-C 6H11Zr*n 20 A59,60 B51 C2,52,61 -63 D3,25,64 Figure 2-7. Fully condensed DDSQ structures. made from polymer BA had to be made hydrophilic. This was accomplished by exposure to deep UV light, which allowed the formation of silanol (Si -OH) groups, making a hydrophilic surface. Thus, polymer BA provides a novel material in the field of flexible printed electronics. AA59 Figure 2-8. Polymer AA from DDSQ A . SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeY1MeY2H2CSiOSiOSi*nzyY2 =SiOSiOSiOSiOSiOPhPhPhPhOSiOSiOPhOOSiOSi*OSiPhPhxY1 = 21 Additionally, compound A was used to synthesize polymer AB (Figure 2-9) for develop ing new fluorinated functional materials. 60 Polymer BB did not exhibit a clear weight loss until T > 800 ¡C, much higher than that of the parent compounds : RF-(ACMO) n-RF, RF-(DMAA) n-RF and R F-(DOBAA) n-RF, which exhibited a clear weight loss at only 250 Ð 350 ¡C. It was also determined that light absorbance was enhanced in BB and the fluorescent ability was good when compared to the parent polymer. Therefore , polymer BB not only has surface -active propertie s from the fluorine, but also possesses light emitting characteristics, and can potentially be used in fluorinated functional materials. AB60 Figure 2-9. Polymer A B from DDSQ A . Compound B (Figure 2-7) was used to synthesize polymer B A ( Figure 2-10) in order to increase the Tg of polysiloxanes and expand their applications. Additionally , the cis and trans isomers were separated so that polymers of each isomer and an equivalent mixture of isomers could be synthesized. The Tm of each isomer and mixture was evaluated before polymerization and increased in the following order: mix (295 ¡C), cis (304 ¡C), and the n trans (357 ¡C). Tg values were determine d by DSC at a heating rate of 30 ¡C /min, and the softening temperature (T s) was determeined by TMA at a heating rate of 10 ¡C /min. The T g and T s were highest for the trans SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeRFCo-MRFxyRF =Co-M =R =H2CCHROONACMO (a) DMAA (b)DOBAA (c) NMe2NHC(CH 3)2C(=O)CH 3CFCF3OC3F7 22 polymer (35 and 82 ¡C, respectively). Th ese values were lowest for the cis polymer ( 30 and 39 ¡C, respectively). The mixed isomer had Tg and T s values that were in between these values (34 and 45 ¡C, respectively). The neat polymer has a T g at approximately -120 ¡C. Under nitrogen, 5 % weight lo ss at 10 ¡C /min was at T d = 500 ¡C for both isomers, whereas T d = 450 ¡C for the mixture of isomers. Under air, 5 % weight loss occurred at 400 ¡C (mix), 460 ¡C (cis), and 470 ¡C (trans ), with a heating rate of 10 ¡C /min. This study demonstr ated that the incorporation of DDSQ C significantly increased the T g of polysiloxane, and the individual isomeric state could provide additional control over the polymeric system. BA51 Figure 2-10. Polymer BA from DDSQ B . Compound C ( Figure 2-7) was used to synthesize compound CA and polymers CB and CC (Figure 2-11) in order to further enhance the thermal resistance, mechanical and dielectric properties of polyimide (PI) materials. 2 These polymers can be applied as interlayer dielectrics in integrated circuit fabrication. 65,66 Monofunctionalized SQs have been incorporated into PIs as side chains, for the purpose of decreasing dielectric constants; however, there was not a significant effect on the mechanical properties. 67-69 Material CA provides linear DDSQ co -polymerized with PIs and thermo mechanical improvements were realized for these PIs over the neat PI, PMDA/ODA (Table 2-2). Generally, T g was decreased and T d was increased for the 23 majority of these PIs. Only CBc exhibited a decrease in T d. There was no significant change in density. The initial modulus increased for all DDSQ/PIs; however, the tensile strength only increased for CBb. Percent elongation increased only for CBb and CBc. Material CBb also exhibited lower water absorption (< 1 %), when compared to the neat PI film. CA2 CB2 CC2 Figure 2-11. Material applications from fully condsensed C. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeOOOOOOSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeY1MeY2NOOO*Y1 =NOOONOONOO*XnY2 =SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeNNOOOOO**n 24 Table 2-2. Properties of materials CB and CC .2 Polymer Tg (¡C) + Td (¡C)* ! (g/cm 3)# " (10 kHz) ^ Tensile strength (MPa) Elongation (%) Initial modulus (GPa) CBa 264 541 1.41 2.43 42.1 2.9 2.32 CBb 261 538 1.42 2.69 74.1 6.0 2.15 CBc 255 521 1.43 2.39 58.0 5.9 2.15 CBd 262 537 1.42 2.59 52.3 5.0 1.82 CBe 267 551 1.44 2.79 65.8 5.4 1.51 CB 248 537 1.40 - - - - PMDA/ODA 362 530 1.44 3.46 - - - + 2nd heating at a heating rate of 30 ¡C/min. * 10 % weight loss at a heating rate of 10 ¡C/min in air. # density. ^ dielectric constant. Compound C (Figure 2-7) was also used to synthesize compound CD and polymer CE (Figure 2-12).70 Compound C D was synthesized as a model reaction and demonstrated that the reaction formed quantitatively with almost no side reactions or further hydrosilylation. Polymers CEa and CE b were soluble i n a large variety of common org anic solvents, whereas polymer CE c was not. CEa and CE b showed a T g just over 150 ¡C, whereas CE c showed no Tg. Under N2, at a heating rate of 10 ¡C/min, CE a showed an onset of degradation at 518 ¡C. The other polymers CD70 CE70 Figure 2-12. Structure CD and polymer CE from DDSQ C . SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMePhPhPhPhSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMe*MeArRR*nAr =R =BPEP (a) HBPEA (b)DEB (c) PhPhAr =R =BPEP (a) HBPEA (b)DEB (c) PhPh 25 had considerably lower degradation onset temperatures. Again, the incorporation of DDSQ into the polymer backbone demonstrated high thermal stability. Additionally, c ompound C was use d to synthesize polymer CF (Figure 2-13) for preparing Langmuir -Blodgett (LB) films , or amphiphilic hybrid polymer nano -sheets, with a well-defined molecular structure and uniform molecular weight. 61 Previous studies have incorporated monofunctionalized SQs in to LB films, but showed molecular weight distribution s since they used SQ co -polymers or mi xed SQs as a core .71-73 Studies involving compound C, demonstrated that the DDSQ core with di(ethylene glycol) coronae formed a stable and unifor m monolayer. Additionally , this structure demonstrated an amphiphile in which the core wa s hydrophobic and the coronae were hydrophilic. This is unique when compared to typical amphiphilic structures, with a polar/hydrophilic head and tail comprised of a hydrocarbon chain. Thus , the Òcore -coronaeÓ type molecule is a potenti al candidate for LB film forming materials. CF61 Figure 2-13. Polymer CF from DDSQ C . Compound C was also used to synthesize compound CG (Figure 2-14) for the purpose of improving thermomechanical properties of polyurethanes (PU). 62 Generally, the Tg increased with increasing weight % of DDSQ in the PU backbone. Overall, the Tg increased to 7 ¡C (48 weight % DDSQ) from -28 ¡C for neat PU. The te mperature at the maximum degradation rate and the yield of degradation temperatures were also improved with the addition of DDSQ. Moreover , DDSQ-PU displayed increased surface hydrophobicity when compared to neat PU. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeOOOOHOOH 26 CG62 Figure 2-14. Co-polymer CG from DDSQ C . Furthermore, c ompound C (Figure 2-7) was used to synthesize structure CH and polymer CI (Figure 2-15) to improve thermo mechanical properties of poly(hydroxyether of bisphenol A). 63 This polymerization proceeded with a 1:1 ratio of bisphenol A hydroxyl groups to epoxide groups (combine for both com pound CH and diglycidyl ether of bisphenol A , DGEBA) . CH63 CI63 Figure 2-15. Compound CH and Polymer CI from DDSQ C . SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeY1MeY2O*xY2 =OHNOPhH2CPhHNOOOHNOPhH2CPhHN*Y1 =OHNOH2CPhHNOOOHNOH2CPhHN*OO*1-xPhPhSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeOOOOSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeY1MeY2Y2 =OOH*1-xOOHOOHOOO*OHOxY1 = 27 It was determined that the Tg was dependent on the composition of the copolymers (Table 2-3). Although the onset of degradation temperature decreased with the addition of DDSQ, the maximum rate of degradation and the yield of degradation residues si gnificantly increased. Similar to other polymers, the surface hydrophobicity was increased for DDSQ -PH polymers when compared to neat PH. This is yet another example where DDSQ incorporated into the bac kbone of a polyme r improved thermomech anical properties of the neat polymer. Compound D ( Figure 2-7) was selected as the primary structure in this dissertation for several reasons: first, DDSQ was selected over other cage -like SQs since it can eas ily be difunctionalized in gram quantities, and be incorporated into the main chain of a polymer system. Table 2-3.63 Thermal properties of polymer CI with different wt % of CH. Compound CH (wt %) DGEBA:DE (wt) Tg (¡C)* Td (¡C)^ Residue (%) + Neat PH 0 10:0 91.5 424.3 2.2 PH9DDSQ1 6.4 9:1 94.5 424.4 7.9 PH7DDSQ3 20.3 7:3 94.1 418.4 13.7 PH5DDSQ5 35.9 5:5 83.6 385.2 25.3 PH3DDSQ7 53.5 3:7 57.4 385.1 36.9 PDDSQ 84.6 0:10 58.6 318.3 46.4 * DSC heating rate of 20 ¡C/min. ^ 5 % mass loss and a heating rate of 20 ¡C/min. + The yield of degradation residue at T = 800 ¡C. Additionally, DDSQ incorporates cis and trans isomers about the Si -O cage. Secondly, the particular reactive moiety (X) of these DDSQ not only has a significant impact on the characteristics of the molecule itself, b ut it also limits the applications. The aminophenyl moiety of compound D (Figure 2-7) can be thought of as a reactive version of the phenyl moieties attached to each silicon atom. Furthermore, the amine is particularly reactive an d can be used in a large variety of potential applications in areas such as: aromatic polyamides, 74 polyimides, 75 ionic -liquids, 21 photochemistry, 76 and catalysis 77. 28 In previous work , compound D was used to synthesize polymer D A ( Figure 2-16) in order to increase thermal stability, and mechanical properties while maintaining low dielectric constants of PIs. Good thermal stability was observed up to 500 ¡C (Table 2-4). At 700 ¡C in air, res idual weights range from 64 Ð 76 %. The polymers also displayed good mechanical properties DA3 Figure 2-16. Polymer D A from DDSQ D . and solubility in com mon organic solvents. These high thermo mechanical properties coupled with low dielectric constants make DA a suitable candidate for polymeric materials in advanced microelectronic applications. Table 2-4.3 Thermal and mec hanical properties of polymers D A. Polymer Tg (¡C) + Td (¡C)* ! (1 MHz) Tensile s trength (MPa) Elongation (%) Initial modulus (GPa) DAa None 555 2.56 82.1 10.9 2.6 DAb None 550 2.59 78.1 13.6 2.4 DAc 325 540 2.65 76.0 15.9 1.9 DAd None 553 2.43 72.3 8.0 1.8 + 2nd heating at a heating rate of 30 ¡C/min. * 10 % weight loss at a heating rate of 10 ¡C/min in r. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeYMeY*nNNOOOOY =Ar =OOOODA (a) 1,4,4-APB (b)1,3,4-APB (c) TFMB (d)OOCF3F3C 29 Additionally, c ompound D (Figure 2-7) was used to synthesize polymer D B (Figure 2-17) to form a thermosetting material based on phenylethynylphthalic anhydride (PEPA). The onset of the crosslinking reaction of neat phenylethynylphthalimide (PEPI) is above 300 ¡C. For DDSQ-PEPI, the onset tem perature was raised to 343 ¡C with a heating rate of 2 ¡C/min and t o 395 ¡C with a heating rate of 20 ¡C/min. Additionally the DDSQ cage impacted the activation energy of the reaction, the heat of fusion, and the overal l reaction profile. This study demons trated that these DDSQ -PIs could be materials used in fiber -reinforced composites. DB25 Figure 2-17. Polymer DB from DDSQ D . SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNNOOOOPhPh 30 !"!#$%&'()*+'*%,'#-./',01(,*2 #!"!"#$!%&'$()* $ As de scribed above, DDSQ with various che mical moieties, including aminophenyls and 2-methylpropyl -hydroxyl have been synthesized and their cis and trans isomers partially isolated. 25,51 These isomers have been identified using one -dimensional 29Si NMR spectroscopy. 2,3,25,51,52,64 Specificall y 29Si NMR spectra of DDSQ D (Figure 2-7), since it is the primary struc ture for this dissertation, were previously determined ( Figure 2-18).25 Mixed a) b) c) Figure 2-18.25 29Si NMR spectra of (a) cis and trans , (b) majority trans , and (c) majority cis DDSQ D. 31 isomer s from DDSQ D show 29Si resonances at ! -29.9, -78.4, -79.3, -79.5, -79.7 in a ratio of 2:4:1:2:1 ( Figure 2-18a). Majority trans isomers show characteristic 29Si resonances at ! -29.9, -78.4, -79.5 in a ratio of 2:4:4 (Figure 2-18b). Majority cis isomers show characteristic 29Si resonances at ! -29.9, -78.4, -79.3, -79.7 in a ratio of 2:4:2:2 (Figure 2-18c). It was initially expected that the 29Si NMR spectra of the mixed isomers ( Figure 2-18a) would display seven resonances representing all seve n silicon environments in DDSQ D . However, since environments between the cis and trans isomers are very similar, some of the resonances are isochronus, or overlapping. Symmetry argume nts were used to determine peak assignments of the individual silicon atoms. The 29Si resonance at ! -29.9 has been assigned to the D -group silicon atoms, silicon atoms bonded to 2 oxygen atoms (Figure 2-19).78 The 29Si resonance at ! -78.4 has been assigned to the T -group silicon atoms, silicon atoms bonde d to 3 oxygen atoms, nearest the D -group silicon atoms. 29Si resonances at ! -79.4 (cis), -79.6 (trans ), and -79.8 (cis) have been assigned to the internal T -group silicon atoms. Figure 2-19. DDSQ silicon atom labels. 32 !"!"!#$%#&'( #While 29Si NMR does benefit from large chemical shift dispersion typically leading to reduced spectral congestion, the technique is not id eal.79 29Si NMR delivers lower relative sensitivity when compared to that of 1H NMR ; requires a long er recycle delay and experimental time for quantitative measurements, more concentrated samples, and a broadband probe with appropriate hardware (i.e. br oadband amplifier, RF filters, and specific capacitor sticks for tuning) for accurate results . These factors make 29Si NMR less desirable when compared to the high er sensitivity, shorter relaxation and acquisition times, more dilute samples, and standard NMR equipment required for 1H NMR. The ability to identify the isomeric ratio of DDSQ molecules using 1H NMR spectroscopy would save time and material, and ultimately provide a more accurate quantification of the isomeric ratio than utilizing 29Si NMR spect roscopy. In order to unambiguously assign the proton resonances and use them for quantitation, two -dimensional (2D) NMR techniques are necessary . Specifically, proton correlations to the silicon nuclei of the silsesquioxane core not only facili tates 1H sp ectral assignment but also confirm s previous 29Si assignments for this class of DDSQ . Heteronucle ar multiple bond coherence ( HMBC) provides a 2D inverse correlation of hydrogen connectivity to specific carbon atom s, or other hetero atom s (X) (Figure 2-20). Additionally, using a gradient -selected version (gHMBC) reduces unwanted signal artifacts. 80 A gHMBC will provide couplings that are in the range of 2 -4 bonds. On a 2D heteronuclear plot, cross -peaks appear at the intersection of X -H peaks. These cross -peaks are displayed as a 33 contour plot, which is similar to a topographical map. 81 The correlations peaks are in actuality a cross -section (slice) of a 3 -dimensional image of an NMR spectrum. Figure 2-20. Representation of a hetereonuc lear spectrum. 29Si-1H gHMBC NMR has been used to correlate Si -H atoms. 82,83 For these DDSQ, gHMBC can be used to correlated atoms that are separated by 2- or 3-bonds, 2JSi-H or 3JSi-H (Figure 2-21). Proton atoms that have a larger separation than 3 bonds from a silicon atom can be spectroscopically determined from 1H-1H correlation spectroscopy (COSY) experiments Figure 2-21. 2J-coupled Si-H atoms (red , top ), and 3J-coupled Si -H atoms (blue , bottom ). 7.57.3-79.5-79.0-78.57.47.2Silicon spectrum (ppm) Proton spectrum (ppm) 34 (Figure 2-22). Unlike gHMBC, a gCOSY contour plot displays the 1D spectrum traced on the diagonal of the plot. 81 Peaks that do not appear on the diagonal are cross -peaks, or correlations peak s resulting from J -coupling. Figure 2-22. 1H-1H COSY spectrum. Correlation peaks that are not on the diagonal represent J -coupled peaks. ! 35 !"#$%&'()*$+),-*-./'0 $Fractional crystallization provides a platfo rm for larger quantities of material to be separated into fewer fractions , as compared to other methods such as chromatography. 84-87 Furthermore, fractional crystallization provides a much lower energy demand as op posed to an energy -intensive thermal separation method such as distillation. Hence , it is accepted as a n appropriate economic approach for an industrial scale .87 DDSQ B 51 and D 25 isomers have been separated in previous studies ( Figure 2-7) through methods of fractional crystallization that exploit differences in the solubility of the two isomers , and through chromatography columns . However, these studies do not provide a means to accurately quantify the purity of the isomer after isolation. They also lack models representing the solubility behavior and thermodynamic properties of t hese isomers. Solid -liquid equilibria ( SLE) measurements allow for thermodynamic properties of multiple component mixtures to be measured. Simultaneously , solubility behavior of the individual components in these multicomponent mixtures can be determined. A common method for experimentally measuring SLE is through sampling a satur ated solution and analyzing the sample with spectroscopy or chromatography. 88,89 Experimental d ata obtained through this method can then be modeled using common thermodynamic equations . !"#"$%&'()*+,-./*01%2+,(30.4%+5%678 %A criteria for phase equilibria is for the Gibbs energy to be equivalent in each phase at equilibrium (E quation 2 -1):90-92 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 36 where G is the Gibbs energy of component ( i) the solid (S) and liquid ( L) phase s. For a pure component, Gibbs energy is the same as the chemical potential of the liquid, µ (Equation 2 -2):92 !!!!!!!!!!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! To form a liquid, the Gibbs energy change is (Equation 2 -3):92 !!!!"# !!!!!"#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! The pure liquid is a hypothetical pure liquid at the equilibrium temperature because it is typically below the pure component melting point. Manipulating e quation 2 -2 to account for the hypothetical Gibbs energy, gives (Equation 2 -3):92 !!!!!!!!"#!!!!!"#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! With thermodynamic manipulations, e quation 2 -3 becomes (Equation 2 -4):92 !!!!"# !!!!!!"# !!!!!"#!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where !Hfus and !Sfus are the heat and entropy of fusion based on its melting transition ( Tm). The difference between the hypothetical Gibbs ener gy and the chemical potential is the change in the chemical potential for mixing a component. For a mixture , equation 2 -5 can be used : 92 !!!"#"$% !!!!!!!!!!!!!"!"!!!!!!!!!!!!!!!! where x is the mole fraction of component ( i) and " is the activity coefficient. Thus for a single component this becomes (E quation 2 -6):92 !!!!!!!!"#!!"#$ !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 37 Combining equations 2 -4 and 2 -6 results in the general equation for predicting the saturation mole fraction of a solid in a liquid , known as the Schrıder -van Laar equation (E quation 2 -7):92 !"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where x is the mole fraction of solute (i) that remains in solution (liquid phase) at a given temperature (T), based on its melting transition ( Tm) and heat of fusion ( !Hm), R is t he ideal gas constant , and " is the activity coefficient in the liquid phase . The activity coefficient ( ") quantifies deviations from an ideal solution due to molecular interactions. An ideal solution is a solution formed with no accompanying energy or vol ume change on mixing and no excess entropy ; the intermolecular attractive/repulsive forces (or intermolecular interactions) between the various pair types are all similar, and " = 1. When " > 1, the solubility is lower than an ideal solution (an unfavorab le interaction), and when " < 1, the solubility is greater (a favorable interaction). 92 The structu res of the solid material are generally large in comparison to the molecules of many common organic solvents . This will present a solution with non -uniformities and energetic interactions between molecules of the mixture and will give rise to non -idealitie s. There are many plausible activity coefficient models to be considered that demonstrate these non -idealities . A multicomponent model is necessary when considering two isomers and the solvent system . Additionally, when modeling the solubility based on the fully saturated mole fractions of the solvent and the solute, the activity coefficient s will be determined in the liquid state . The selected activity coefficient model will then be combined with equation 2 -7 to calculate the SLE . 38 Several models based on t he idea of local compositions that can be easily extended from a binary to multicomponent system and can be used for liquid activity coefficients are the universal quasi -chemical ( UNIQUAC ), the Wilson, and the nonrandom -two-liquid ( NRTL ) models .90 The UNIQUAC activity coefficient model accounts for different molecular sizes and intermolecular forces thr ough structural factors proportional to external surface area of the molecule (q) and the radius of the molec ule ( r). For many functional groups q and r are tabulated. However, they are not tabulated for silsesquioxane derivatives and would need to be estimated from experimental data, or by molecular group contribution methods. 90 The Wilson activity co efficient model also accounts for different molecular sizes between solvent and solute and intermolecular forces, but does not rely on values of q and r. Instead, it relies only on the idea of local composition is used to capture these effects . Local mole fractions (xij) are scaled with bulk mole fractions ( xi) and a Boltzman factor that is proportional to the probability of finding a molecule of t ype i in the vicinity of a molecule of type j (Pij). Considering a binary (1 -2) mixture provides a depiction of this idea ( Figure 2-23).90 Ratios of the local mole fractions are analog ous to ratios of the Pijs and are expressed as a product of the bulk mole fractions and the Boltzmann factors in terms of gijÕs. For a more detailed explanation of local composition and to understand how it is used in the derivation of the Wilson model, th e Figure 2-23.90 Interaction energy of the molec ules in the Wilson activity coefficient model, a central molecule of type 1 (left), and a central molecules of type 2 (right). g11g12g21g22Binary: = 1= 2g12 = g21 through symmetry gij = interaction energy 39 reader is encouraged to consult the literature. 90 A disadvantage of the Wilson model, however, is that it is applicable only to systems that are completely miscible. The NRTL activity coefficient model is based on the same theo ry as the Wilson model, but can be used for systems that are partially immiscible, and far from ideality. Additionally the NRTL model provides a parameter , !ij, to account for binary interaction characteristics of th e non-randomness of the mixture. This will provide a more flexible model for th e system in this dissertation . When !ij = 0, the solution is completely random. Generally !ij is approximately in the range of 0.2 to 0.4 and Prausnitz has provided guidelines for specifying the magnitude of !ij by considering comparisons to other classes of mixtures. 90 The NRTL activity coefficient model is ( Equation 2 -8):93 !"!!!!!!!!"!!"!!!!"!!!!!!"!!!!!!"!!!!!"!!"!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!! !!"!!"# !!!"!!"!!!!!!!!!"!!!"!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where ! is t he activity coefficient of solute ( i), x represents the mole fraction of each of the remaining constituents in this quarternary system, " represents the non -randomness parameter, and # represents a binary interaction parameter characteristic of the differe nce between the binary energy parameters for the i + j and i + i interactions which is developed from the idea of local compositions. In order to use the above models for determining SLE behavior, there are several steps that need to be taken. F irst , the solubility limit of the solute in a selected solvent solution will be determined experimentally. The solvent solution is generally compr ised of two solvents. The 40 solute will be co mpletely soluble in one solvent and insoluble in the other solvent, designate d the anti -solvent. As the ratio of the solvent to anti -solvent is varied, the solubility limit of the solid material changes, and data points are generated for each ratio. The thermophysical properties of each isomer, Tm and !Hm, will also be determined a t this step. Second ly, the behavior will be modeled using the experimental data and equations 2-7 and 2-8 in order to determine the theoretical solubility limit at any solvent ratio . Since there are multiple unknown variables, this is generally accomplish ed numerically through a regression. For simplification, a binary system has been modeled in a flow chart selected ( Figure 2-24). Generally, a n initial value of "1 is select ed, usually 1.0, and is used in equation 2 -7 along with the thermophysical data. This provides a value of x1 for that component. Other component mole fractions are determined through material balances; i.e. x2 = 1- x1 in a binary. The initial v alues of x1 and x2 are then used in e quation 2 -8 with guessed values for the interaction parameters, which will generate a new value of "1 and "2. The new values of "1 and "2 are used again in equation 2-7 and the process is repeated unti l the change in x1 is minimized. Interaction parameters are continuously adjusted in order to satisfy the minimization. Simultaneously the differences in the calculated and experimental values for x1 are minimized. After the minimization is complete, calcu lated values of xl and "l can be compared with experimental values. Experimental values of "l are determined by inputting experimental mole fractions and thermophysical data into e quation 2 -7. 41 Figure 2-24. Flow chart for determi ning calculated xl and !l. ! 42 !"#$%&'()&('*+$,-*'*)&.'/0*&/12 $Configurational modifications, such as varying the conformation of a small organic molecule, tend to have large effects on the thermodynamic, physical, and structural properties. 94 Para -di-substituted benzenes typically have a higher symmetry than meta -di-substituted benzenes and thus can more easily pack in a three -dimensional crystal lattice , forming more stable , densely packed crystals that exhibit a higher Tm (Table 2-5).94-97 The difference in the Tm for these Table 2-5. Tm of meta vs. para isomers. Compound Tm (K) meta para C6H4Et2 189 230 C6H4I2 310 403 C6H4(CN) 2 434 495 C6H4(NC) 2 377* 439 C6H4(Cl)NO 2 319 357 C6H4(Br)NO 2 329 400 C6H4(OH)(NO) 2 370 387 C6H4(OH)Cl 306 317 C6H4(OH)Br 306 339 C6H4(Br)COOH 428 527 C6H4(I)COOH 461 543 C6H4(Me)COOH 385 455 C6H4(Br)CONH 2 429 463 C6H4(I)CONH 2 458 488 * Inconsistent data also seen as 350 disubstituted benzenes is as large as 99 K (C6H4(Br)COOH) and as small as 9 K (C6H4(OH)Cl). Furthermore , many organic para - structu res ( p) that exhibit a higher symmetry number ( !), also exhibit a higher "Hm and "Sm than their met a-counterparts ( m) (Table 2-6). More over, 43 compounds that are more symmetric and higher melting generally exhibit lower solubility. 94 For additional examples, the reader is encouraged to consult the literature. 94,95 Table 2-6. Melting transitions based on meta - and para - substituted benzenes. Compound ! "Hm (kJ/mol) Tm (K) "Sm (J/mol*K) m-Dichlorobenzene 2 12.56 248.3 50.6 p-Dichlorobenzene 4 19.92 326.1 61.1 m-Xylene 2 11.55 225.2 51.3 p-Xylene 4 17.09 286.3 59.7 Similarly, trans isomers typically exhibit a higher order of symmetry over their cis counterparts; which also provides them with higher melting temperatures. 97-99 Trans isomers of disubstituted ethylenes have a higher T m than their cis counterparts. 100 Specifically, trans dichloroeth ylene has a T m at approximately 225 K, whereas cis dichloroeth ylene has a Tm below 200 K. Additionally, trans dimethyleth ylene has a T m above 160 K, and cis dimethyl ethylene has a Tm below 150 K. Other cis and trans isomer s demonstrate similar behavior (Table 2-7). Extensive studies on melting points of geometrical and constitutional isomers have been accomplished and the interested reader is encourage d to consult this literature. 97 Table 2-7. Tm of cis/trans isomers. Compound P Tm (K) cis-But-2-ene C2 134 trans -But-2-ene C2h 167 cis-Pent -2-ene C1 122 trans -Pent -2-ene C1 137 44 Additionally, increasing the molecular weight (MW) , while maintaining the same geom etry, can also affect the symmetry, molecular packing, and the Tm of a small organic molecule. 95,101 This has been demonstrated using alkane chains ( Table 2-8).97 Increased MW Table 2-8. Tm of increasing MW. Compound MW Tm (K) C3H8 44 83 C4H10 58 137 C5H12 72 146 C6H14 86 181 C7H16 100 188 C8H18 114 213 C9H20 128 219 C10H22 142 240 C11H24 156 249 C12H26 170 264 C13H28 184 270 C14H30 198 282 C15H32 212 289 provides increased number of site -site interactions per molecule , which increase the T m. However, increasing the MW of substituted benzene by introducing a larger substituent , or branching of an alkane chain may actually lower the T m. For example, adding a bulky moiety such as tert -butyl, cyclohexyl, or phenyl could interrupt the symmetry and distort the lattice for steric reasons, and would decrease T m. Changing a moiety from hydrogen to a propyl group of an azopyridine carboxylic acid demonstrated such a decrease in T m (Table 2-9). 45 Table 2-9. Tm for altered moiety. Compound n R Tm (¡C) 1a 5 H 229 1b 5 C3H7 172 2a 10 H 173 2b 10 C3H7 152 Moreover, the way molecules pack and their affinity for hydrogen bonding can also provide thermodynamic and structural modifications. 94,101 -103 Molecules that exhibit hydrogen bonding in the crystal lattice have the potential to form stable dimers and exhibit an increase in Tm over a more symmetric isomer. 104,105 For example, meta -anisylpinacolone exhibits a T m of 58 ¡C, whereas para -anisylpinacolone exhibits a T m of 39.5 ¡C (Figure 2-25).105 The higher T m of the meta -structure was determined to be a result of its hydrogen bonding potential. The crystal structure shows two methoxy hydrogen atoms that exhibit h ydrogen bonding with the carbonyl oxygen. Additionally, the methoxy oxygen atom shows hydrogen bonding with one methylen e hydrogen and one aromatic hydrogen. Other r eports demonstrate that a completely different crystal structure can be obtained from deute rating hydrogen bond ing protons, which could potentially change the Tm along with other physical and chemical characteristics. Figure 2-25. Meta- and para - anisylpinacolone. 46 !"#"$%&'()*+,%-'./0*12% /+3%4./5' %3*/,2/65% 17%6*8'3%*516'25 %%Melting points can be altered when a ÒguestÓ molecule is able to substitute or diffuse into the crystal lattice of a ÒhostÓ molecule. 106 Lattice defects and solid solutions result in lower melting temperature, but are not necessary for melting depression, which can occur in mixtures where the solid and liquid compositions differ. This type of Òmelting point depressionÓ is indicative of a eutectic melting system in which the solid phases are immiscible and the liquid phases are completely miscible . Binary mixtures that exhibit melting point depression can be classified as a general , congruent or incongruent -melting eutectic , and can be portrayed in phase diagrams. 107 General e utectic phase diagrams display two immiscible solid phases in equilibrium with a single liquid phase at the melting temperature (Figure 2-26).108 The eu tectic temperature (T E) represents the temperature at which the first drop of liquid is observed. On a Figure 2-26.107 Binary phase diagram of solids A and B. E SB + L L SA + L SA + S B TA TB A B Temperatur eMole fraction TE 47 differential scanning calorimetry (DSC) trace, this is sometimes observed as a change in the slope of the baseline (not to be confused with T g), or an endothermic transition prior to the melt. TE is constant regardless of composition. The composition at which the T m is equivalent to T E is the eutectic composition. On a congruent melting eutectic phase diagr am, the liquid and solid phases have identical compositions at the equilibrium melting temperature, Tm (Figure 2-27).109 When a sample of mole fraction C ( Figure 2-27) is heated, all of solid A (SA) and part of solid B (SB) will melt at the metastable eutectic temperature. 107 The melted portion will then recrystallize to form a co -crystal solid (S C) and then SC and the remainder of SB completely melts at the congruent melting point T c. A DSC trace will show a n endothermic transition representing melting at T m-E, which is immediately followed by an exothermic transition , representing co-crystal formation. A final endothermic transition then occurs representative of the Tc. On an i ncongruent meltin g eute ctic phase diagram, the liquid and solid phases have different compositions at the peritectic , or incongruent, melting point , Tp (Figure 2-27).109,110 Similar to a congruent system, when a sample of mole fraction C is heated, all of S A and part of SB will melt at the metastable eutectic temperature, and the melted portion will then recrystallize to form a co -crystal solid (S C).107 Different from the congruent system, SC will melt at the incongruent melting point, T p, followed by a recrystallization of SB that finally melts at T B. A DSC trace will s how an endothermic transition representing melting at T m-E, which is 48 immediately followed by an exothermic transition , representing co -crystal formation. This is followed by another endothermic transition representative of the co -crystal melting, and an exothermic transition associated with the recrystallization of SB at T p. Finally, a broad endothermic transition follows at T B. Figure 2-27. Phase diagram representing two systems of co -crystallization, congr uent melting system (left), and incongruent melting system (right). Both congruent and incongruent eutectic -type binary mixtures are capable of co -crystal formation. A co -crystal is a crystalline solid containing at least two unique solid components. 111-113 Co-crystal formation is particularly appealing in the pharmaceutical industry because it has been seen to improve physiochemical properties such as solubility, physical stability, mechanical properties, and bioav ailability. 107 A phase diagram in which both the solid and thee liquid phases are completely miscible is known as an isomorphous melting system , or a homogeneous solid -state solution .109 In order m-E SB + L L SA +SC SA + STA TB A B Temperatur eMole fraction Tm-E C SB + S C E2 E1 SC + L SA + L TE2 TE1 TC M-E SB + L L SA +SC SA + S B TA TB A B Temperature Mole fraction Tm-E C SB + S C TP E SA + L TE P T SC + L 49 for a system to be isomorphous, both components must so lidify in the same crystal system, with similar lattice dimensions, and related chemical constitution. 109 Solid solutions may be substitutional , where ÒguestÓ molecules replace ÒhostÓ mole cules in the crystal lattice, or interstitial, where ÒguestÓ molecules occupy empty spaces in the ÒhostÓ lattice. Ideal behavior of an isomorphous system is achieved when the lattice dimensions of both components match and the compounds are chemically similar. Additionally, the ÒguestÓ molecule cannot interrupt the attractive -repulsive forces within the crystal lattice of the ÒhostÓ component. Positive deviations from ideality would result from disruption and negative deviations from large attractive forces. A typical phase diagram for an isomorphous solution has liquidus and solidus curves that represent the onset of melting and the onset of cooling of the sample (Figure 2-28). Above the liquidus curve is a miscible liquid of A and B, a nd below the solidus curve is a solid solution of A and B. The area between the curves is the two -phase region. A smaller two -phase region is suggests a Figure 2-28. Phase diagram of an isomorphous, solid -solu tion melting system . Solidus Liquidus Liquid + Solid TB TA A B Temperatur eMole fraction Miscible Liquid A + B Solid Solution A + B 50 mixture that is nearer to ideal. 109 The DSC trace of a solid solution is represented by a single endothermic melting transition upon heating. Typically, as the composition moves further from a pure component, the endotherm gets bro ader. The broadness is indicative of the distance between the solidus and liquidus curves. SLE methods that were discussed previously can also be used to model experimental melting point depression and determine a phase diagram .114-116 For all three eutectic -type systems, the Schrı der -van Laar equation (Equation 2 -7) and the NRTL activity coefficient model (Equation 2 -8) can both be used to determine the solid -liquid melt equilibria . However, for an isomorphous syst em; activity coefficients have to be considered in both the solid and the liquid phase s if the system is not ideal . Thus a correction can be made to the Schrı der -van Laar equation that accounts for this (Equation 2 -9):109 !"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where x is the mole fraction of (i) in the liquid ( L) or solid ( S) phase, T ¡m and !H ¡m refer to the melting transition of the pure component , Tm refers to the depressed melting point at a specified mole fraction, R is the ideal gas constant , and " is the activity coefficient . This equation relates the composition in both phases with the non-ideality of each phase and the thermophysical properties of the pure components. 117 The thermodynamic, physical, and structural implications of configurational modifications discussed in this background section have been well established and can be easily predicted for small organic molecules . However, the same implications are mu ch more complex 51 for larger molecules, especially fo r compounds containing a large inorganic portion. Previous to this dissertation, there have not bee n any structured studies on the implications of configurational modifications to these DDSQ molecules and their phase diagrams . 52 !"#$%&'&()&*$+,,'-(.)-/01 $$!"#"$"%&'()*+,(-, %Thermally cured polyimides are known as high performance materials (HPM) due to their excellent thermal and mechanical properties, in addition to their thermo -oxidative stability .118 In light of their remarkable properties, polyimides have been applied to a wide variety of applications such as : matrices for high performance advanced composite materials, membranes for gas separation, thin films in electronic devices , structural a dhesives and sealants, and high temperature indicators for aircraft wire coatings. 119 However, polyimides exhibit high viscosity , which requires high pressure to be used in order to process and fabricate the structural composites and adhesive joints. Therefore , research on polyimides has been directed towards decreasing the viscosity without sacrificing HPM characteristics. 120-124 Phenylethynyl groups are commonly used as polyimide thermosetting materials ( Figure 2-29). Upon heating , the triple bond undergoes a polymerization through a cross -linking, curing reaction to form thermosets . In order to achieve a reduction in the viscosity phenylethynyl groups are placed on the terminal end of a molecule. Efforts in overcoming this complication include structural modifications to the bridging molecule , which involve adding no ncoplanar phenylene moieties, kinked comonomers, and bulky lateral groups. Figure 2-29. An example of a phenylethynyl group. OPhOO 53 Ar = A B C D E F G H Figure 2-30.119 Oligo imide structures used to reduce viscosity. NOOArNOONOOArNnPETI-5K OOAr = 85 %O3,4'-ODA + 15 %OOAPBNOOArNOONOOArNnCF3CPE-3FDA OOO4,4'-ODA O3,4'-ODA OOAPBO4-BDAF OCF3CF3F3CC F3TFMB p-PDA &m-PDA O1,4,4,6-FAPB OF3CCF3OO1,3,5 APBNOOArNOONOOArNnCF3CCF3PE6FDA NOOArNOONOOArNnOPI-A OO 54 For PETI -3K and PETI -5K (Figure 2-30), minimum viscosity occurs over 350 ¡C. Additionally, thes e oligomers do not fully melt until 319 and 349 ¡C, respectively. Since the curing reaction occurs at approximately 350 Ð 371 ¡C, this make the processing window very small. Due to this, t he initial viscosity may vary each time it is measured because the extent of melting and curing will vary. A reduction of viscosity was achieved through substitutions to the structures of PETI -5K and PETI -3K, where the 3 and 5 represent the number average MW (Mn) as 3000 and 5000 g/mol, respectively ( Figure 2-30, Table 2-10).119 The minimum viscos ity was decreased for all oligomers when compare d to their corresponding PETI -3K and -5K oligoimides. Thus , it was concluded that the replacement of the planar, biphenylene dianhydride (s-BPDA) with a bulkier group causes minimum viscosity to decrease . Fur thermore, the crystalline melting transition no longer exists, and all oligomers only exhibit a Tg. As the Tg Table 2-10.119 Complex viscosity. Complex viscosity (Pa S) at the following T ( ¡C) Compound 290 300 310 350 Mmv* (Pa s, ¡C) Tg (¡C) Tm (¡C) PETI-3K 1823 1170 302 14 9 at 353 195 319 PETI-5K 71057 134770 161050 3353 2237 at 354 225 349 PE-3F-PETI-3K 27 17 7 29 6 at 314 190 PE-3F-PETI-5K 860 350 154 65 51 at 339 218 PE-3F-5K-A 1192 230576 230 347 91 at 328 235 PE-3F-5K-B 22 18 16 21 14 at 330 226 PE-3F-5K-C 38 32 31 19 11 at 355 184 PE-3F-3K-D -- -- 7 -- -- -- PE-3F-5K-D 367 165 99 312 68 at 325 237 PE-3F-5K-E 161 53 21 260 14 at 328 213 PE-3F-5K-F 1135 602 218 149 87 at 329 221 PE-6F-PETI-3K 26 14 7 16 3.5 at 324 195 PE-6F-PETI-5K 255 129 77 61 48 at 337 212 *MMV: minimum viscosity 55 decrease for compounds that d o not exhibit a melting transition, the minimum viscosity value and temperature appear to decrease. It was also recognized that at higher MW, there are more considerable influences on viscosity from the structural changes. When compound PI -A was used with different Ar moieties, a definitive trend in M n was determined ( Table 2-11).125,126 As the Mn decreased, the minimum viscosity and T g both decreased . Table 2-11.125,126 Complex viscosity according to MW of PI -A. PI-A Complex viscosity (Pa S) at the following T ( ¡C) G:Ha Mn (g/mol) 250 275 300 325 350 Mmv (Pa s, ¡C) Tg (¡C) 1 10000b 27010 8349 1389 392.6 418.8 328.1 at 334 183 1 3234 3801 491.8 121.3 70.48 1162 69.88 at 322 187 1 2093 63.31 15.37 7.17 7.18 29.61 6.39 at 313 161 3 2200 0.61 0.36 0.30 0.33 1.78 0.28 at 3 08 126 1 2139 0.58 0.32 0.31 0.36 0.44 0.25 at 288 124 0.3 1762 54.1 0.60 0.50 0.57 0.69 0.50 at 30 0 124 G:B 1 10000b N/A 27590 12475 3030 1472 1451 at 351 208 1 3173 12430 1421 323 135 3449 131 at 327 197 1 2569 156 32.0 10.9 7.54 10.2 7.54 at 325 169 3 2200 1.0 0.70 0.71 0.64 0.70 0.60 at 301 129-132 1 1991 1.0 0.39 0.29 0.33 0.57 0.27 at 299 132 0.3 1272 19.8 0.84 0.69 0.82 1173 0.67 at 294 129-132 a Ratio of Ar moieties used b This is a calculated value, experimental was not determined Isothermal viscosity measurements at 310 ¡C also demonstrated a decrease after ho lding for a 30 and 60 minutes ( Table 2-12 and Table 2-13). It was suggested that the bulkier groups from 3FDA and 6FDA make the structure amorphous, providing weaker intermolecular interactions. Additiona lly, these units stabilize the phenylethynyl end caps, which will reduce the 56 rate of cross -linking at the minimum viscosity temperature. For PI -A, the variation in the viscosity , when isothermed at 280 ¡C for 2 hours , decreased as the Mn decreased (Table 2-14). These studies depict the shortcoming s of this class of polyimide thermosets; the minimum melt viscosity needs to be reduced in value and in temperature, and the solid -liquid phase transition needs to occur at a lower tempera ture in order to increase the processing window. Additionally, the MW for the oligomers is not constant, since n varies, and therefore the properties such as the melt viscosity are not constant ( Figure 2-30). Thus, in addition to decreasing the viscosity and processing window, current research efforts have been directed towards developing a monodispersed oligomer to eliminate variation in M W. Table 2-12.119 Viscosity stability of PETI -3K and PE -3F. Viscosity (Pa s) at 310 ¡C Rate of viscosity increase Compound initial 30 min 60 min 30 min 60 min PETI-3K 1501 9913 >27500 280 >433 PE-3F-5K-B 1781 8528 >14000 5.5 >286 PE-3F-5K-C 53 207 978 5.1 15.4 PE-3F-PETI-3K 60 298 2098 7.9 33.9 PE-3F-PETI-5K 1433 4620 16791 106 256 PE-3F-3K-D 821 3269 10266 81 255 PE-3F-5K-D 72 147 2610 2.5 42.3 Table 2-13.119 Viscosity stability of PE 6F. Compound Viscosity (Pa s) at 310 ¡C initial 60 min PETI-3K 899 22500 PE-3F-PETI-3K 9 15 PE-6F-PETI-3K 17 28 PETI-5K 20759 412205 PE-3F-PETI-5K 321 527 PE-6F-PETI-5K 199 348 57 Table 2-14.125,126 Viscosity stability for samples PI -A held at 280 ¡C for 2 hours. PI-A G:H Mn (g/mol) Variation (Pa s) 1 10000 896.4 Ð 5379 a 1 3234 264.8 Ð 1563 b 1 2093 11.56 Ð 81.79 3 2200 0.48 Ð 3.30 1 2139 0.38 Ð 0.95 0.3 1762 0.57 Ð 4.01 G:B 1 10000b 8268 Ð 59300 a 1 3173 479 Ð 3145 b 1 2569 25.0 Ð 178.6 3 2200 0.35 Ð 1.25 1 1991 0.56 Ð 2.09 0.3 1272 0.65 Ð 11.9 a 290 ¡C b 310 ¡C A variety of cage like SQs have been end -capped with phenylethynyl groups (Figure 2-31). All SQ -based oligoimides displayed monodispersed MW. It was revealed that changing the R -group from isobutyl to phenyl resulted in a retardation of the phenylethynyl reactions. Additionally, changing the spacer group from propyl to phenyl effected the initial curing with the monofunctionalized, and octaf unctionalize d compounds. The DDSQ cage , in chain structure provided an increase in the extent of cure reaction. These studies also revealed that SQs end -capped with phenylethynyl groups are comparable with their organic counte rparts and can be used for fib er-reinforced composites. Additionally, the solid to liquid phase transition of these materials was decreased. Studies based on how the SQ cage has a ffec ted the viscosity were not accomplished prior to this dissertation. 58 Figure 2-31.25 Phenylethynyl end -capped SQs. SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiRR = R1 or R2NOOPhR1 =NOOPhR2 =SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiR1R = iBu, PhSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiR = iBu R2SiOSiOSiOHSi SiOSiOSiORRRRRROOOOOOSiR = iBu, PhOR1SiMeR1SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOR1SiMeSiMeR1 59 !"#"!$%&'()*+,-./01. $Aromatic polyamides , or polyaramids, are also considered to be HPMs due to their excellent thermo mechanical properties, making them suitable for advanced engineering technologies .74 The most common polyaramids are Kevlar ¨ and Nomex ¨, or poly( p-phenylene terephthalamide) (PPD -T) and poly( m-phenylene isophthalamide) (MPD -I), respectively (Figure 2-32). There are three common polymerization methods for synthesizing Figure 2-32. Kevlar ¨ (PPD -T) and Nomex ¨ (MPD -I) polymer structures. these polymers: (1) interfacial polymerization with M/P -DA and PCL , (2) low temperature solution polymerization with a diamine (DA) and a phthaloyl chloride (PCL) 127-130, and (3 ) low temperature solution polymerization with N -silylated DA and PCL 131-133 (Scheme 2-4). The method producing the highest molecular weight (MW) polymer is generally preferred. 127 NHHN*OO*HN*HNOO*nnKevlar Nomex a) b) Scheme 2-4. Nomex ¨ synthesis us ing (a) DA with PCL, and (b) N -silylated DA with PCL. ClClOONH2H2N+Nomexdiaminephthaloyl chlorideClClOONHHN+NomexN-silylated diaminephthaloyl chlorideMe3SiSiMe 3 60 Low temperatures solution polymerization with N -silylated DA and PCL resulted with higher inherent viscosity (!inh ) than their DA equivalents (Table 2-15). 131-133 Inherent viscosity is a relative measure of the MW of the polymer. Additionally, these reactions proceeded faster. Amide solvents are used for this polymerization at Ð10 ¡C. N -silylated PPD -T was synthesized in a solvent mix ture comprised of NMP and HPT containing LiCl at Ð10 ¡C to Ð5 ¡C for 12 h. N -sylilated MPD -I was synthesized in NMP with LiCl at Ð10 ¡C to Ð5 ¡C for 5 h. Table 2-15. !inh of synthetic Nomex ¨. Polymer !inh N-silylation method Diamine PPD-T 7.41 3.23 MPD-I 2.45 1.03 Silicon has a strong affinity for oxygen, and the carbocation on the § -position to the silicon can be stabilized through the silicon "-# effect in a two-step mechanism ( Scheme 2-5).134-136 The carbonyl oxygen is attracted to the silicon atom on the amine, making it possible for a nucleophilic attack of the nitrogen at the carbonyl carb on. A tetrahedral intermediate is formed. The second step involves the elimination of the chloride ion from t he tetrahedral intermediate t hat is stabilized by the "-# effect. Scheme 2-5. Mechanism for reaction of N -silylated DA and PCL. As previously mentioned, !inh provides a relative, and simple measure of the MW of these polymers. Inherent viscosity is an approximation of intrinsic viscosity ([ !]). Intrinsic CArClONHArSiMe 3CArClOSiMe 3HNArCArClHNOArSiMe 3CArHNOArSiMe 3ClCArHNOAr+Me3SiCl 61 viscosity can be related to molecular weight through the Mark -Houwink (Equation 2 -10):127 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!"! where k and a are constants dependent on the polymer -solvent pair, but are independent of MW. However, the Mark -Houwink equation does not hold true for viscosity -MW relationships at a polyaramid MW > 40,000 . Inherent viscosity measurements for PPD -T are typically done with a viscometer in 96 -98 % sulfuric acid, and for MPD -I in amide solvents. Gel permeation chromatography (GPC) ca n also be used to obtain these values. Other, more comp lex methods for determining MW of a polymer include: l ight scattering and elemental analysis. These polyaramids are known to have higher specific strength and modulus than glass and steel due to their low density, and have been used to make products to protect personnel from fire, bullets, and cuts, reduce the weight of aircraft and automobiles, and hold drilling platforms in place. 127 However, they exhibit extremely high transit ion temperatures, which lie above their decomposition temperatures, and very poor solubility in common organic solvents, making them difficult to process and limiting their applications. PPD -T is even more insoluble than MPD -I. An all para - wholly aromati c structure creates a stiff, rod -like macromolecule with high cohesive energy and a high tendency for crystallization due to favorable intramolecular H -bonding and pi -stacking. The tendency for crystallization is so strong, that it is even maintained in th e liquid state. MPD-I is not as linear, which causes a reduction in its cohesive energy and crystallization tendency, making it a HPM that is easier to process than PPD-T due to its inherent flexibility from the ÒkinksÓ in its structure. However, for the s ame reason it slightly underperforms PPD -T ( Table 2-16). 62 In view of the remarkable characteristics displayed by polyaramids, research efforts are directed in two areas: (1) reducing the cohesive energy resulting from H -bonding an d pi-stacking Table 2-16.74 Properties of PPD -T and MPD -I. Property PPD-T MPD-I Density (g /cm 3) 1.44 1.38 Water uptake (%) 3.9 5.2 Thermal Properties Tg (¡C) --- 275 Tm (¡C) > 500 db 365 db Td (¡C, in N 2) 520-540 400-430 Tensile Properties 2.9-3.0 Strength (GPa) 2.9-3.0 70-112 0.59-0.86 Modulus (GPa) (GPa 70-112 7.9-12.1 Elongation (%) 2.4-3.6 20-45 Crystallinity (%) 100 68-95 Flammability (L.O.I.) 29 29 a 65 % RH b Decomposes (d) so these materials are easier to process, and soluble in common organic solven ts without sacrificing high performance properties, and (2) expanding their high -performance properties to additional applications in new and promising fields, such as: optically active, luminescent, ionic exchange, flame -resistant and fiber -forming materi als. Research on the stereo -isomers of aromatic polya mides has been given little recognition. However, Koning et. al. have looked at cis/trans isomers of a non -aromatic polyamides, formed from the reaction of 1,4 -cyclohexanedicarboxylic acid, and 1,4 -diam inocyclohexane acid ( Figure 2-33).137 It was observed that the trans isomer exhibited a more ÔstretchedÕ configur ation due to its symmetry, and wa s capable of increasing the Tm, whereas t he cis isomer did not significantly alter the Tm. Additionally, the intersheet distance of the polyamides increased with increasing trans content, and rem ained the same with increasing cis content. These results clearly 63 demonstrate that the incorporation of the more highly symmetric trans isomer into the polyamide back bone gives rise to a unit -cell expansion, resulting from the larger intersheet distance without any effect on the interchain distance. The reason for this occurrence has been at tributed to the reduction of rotational ability intrinsic to the trans polymer, hampering the formation of intersheet H -bonds. Furthermore , the ÒkinkyÓ cis structure is not incorporated into the crystalline region of the polymer. Conversely, the trans structure is prese nt in both the crystalline and amorphous regions. All these effects were seen with aliphatic polyamides and thus do not incorporate additional effects such as pi-stacking that could be present in a polyaramid . Figure 2-33.137 Stereoisomers of a polyamide. An important feature of polyaramids to be considered is that w ater or polar, protic solvents can diffuse into the amorphous regions and disrupt H -bonding (Figure 2-34). Interchain H-bonds can be re placed by interactions with water, or other polar molecules, thus weakening interchain interactions and allowing freedom of movement about the amide group. This can have a large effect on polymer properties, specifically decreasing T g.138 Figure 2-34.138 Diffusion of wa ter into the chain of a polyaramid. HONHONHHHONHHNOHtrans cisHN*HNOO*nHOHN*NOO*nHOH 64 There have not been any cage -like SQ modified polyaramids of Kevlar ¨ or Nomex ¨. However, there have been reports of cage -like monofunctionalized SQs ( SQ F, Figure 2-3) to modify PA6, which is the non -aromatic version of these polymers ( Table 2-1).34 These reports dem onstrate that the incorporation of monofunctionalized SQ F through blending with PA6 decreases the T g for injection molded samples . For melt -spun fiber samples, there is an increase in tensile modulus and strength of ca. 50 % for up to a composition of 2.5 % SQ. This value was then decreased at higher composition s of SQ. Additionally; it was revealed in DMA experimenta that above T g, the rubbery modulus for the blends was significantly higher than neat PA6. Although these studies have demonstrated improved thermomechanical properties for PA6, the SQ itself is not covalently bon ded to PA6, which could potentially have a more significant effect on the polymer properties . !"#"$%&'()*%+),-). % An ionic liquid (IL) is defined as a salt with a T m below 100 ¡C.21 They a re very attractive material s due to their wide variety of properties: wide liquid range, excellent electrochemical, thermal, and chemical stability, dispersant capabilities, high ionic conductivity, tunable dielectric constants, and biocompatibility. Additionally, they have low vapor pressure and are soluble in a wide variety of common solvents. Furthermore, they can be recycled in liquid -liquid processes, and potentially recovered. These combined properties make ILs part of the Ògreen chemistryÓ classification. There applications include: replacing volatile organic solvents in an indus trial setting for conventional as well as catalytic reactions and separation processes, absorbents in gas separations, CO 2 storage, SO 2 and hydrofluorocarbon absorption, 65 electrolytes for energy storage devices such as dye sensitized solar cells, electric d ouble -layer capacitors, and fuel cells. 139 Properties of ILs can be enhanced through the incorporation o f nanopa rticles such as: silica and metal supporting oxides. These nanoparticles will act as IL -supporting substrates and have found applications in catalysis and the development of ion lithium batteries. 140-142 Scheme 2-6. Synthesis of SQ ionic liquids. SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiR =NH2MeI, K2CO3Acetonitrile 70 - 75 C48 hours¡SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiNIR =50 ¡Covernight SiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiNYSiOSiOSiOSiSiOSiOSiORRRRRRROOOOOOSiNIM + Y -+H2OY- =F3CSNSF3COOOOCl-BF4-PF6-SOOO(CF 2)3F3CABCDE 66 More recently, cage -like SQs have been nanoparticles selected for IL -supporting substrates. 21,143 -145 Reacting a corner -cappe d aminopropyl SQ with methyl iodide has successfully produced an IL with I - as the anion and -NMe3+ as the cation ( Scheme 2-6).21 Ion exchange then led to additional SQ ILs. All of these ILs had relatively high T g values, low -polarity, and were amorphous materials when compared with tetraalkylammonium ILs. Addition ally, when compared to standard ILs, dielectric constants as well as room -temperature conductivity were found to be much lower. The extraction capability for the sodium salt [H2TPP4-] from aqueous solutions confirmed that these ILs have cationic surfactant behavior. There was no considerable aggregation in solution with the exception of the iodide and chloride ILs. These studies demonstrate that cage -like SQs have great potential in the field of ILs. No studies on ILs and DDSQs were carried out prior to thi s dissertation. 67 !!!!! "#$#"#%&#' ! 68 !"#"!"$%"& ' (1) Yang, B.; Li, M.; Wu, Y.; Wan, X. Polymers & Polymer Composites 2013, 21, 37. (2) Wu, S.; Hayakawa, T.; Kikuchi, R.; Grunzinger, S.; Kakimoto, M. Macromolecules 2007, 40, 5698. (3) Wu, S.; Hayakawa, T.; Kakimoto, M.; Oikawa, H. Macromolecules 2008, 41, 3481. (4) Wu, J.; Mather, P. T. Polymer Reviews 2009, 49, 25. (5) Fan, H.; Yang, R. Industrial & Engineering Chemistry Research 2013, 52, 2493. (6) Rakesh, S.; Dharan, C. P. S.; Sella durai, M.; Sudha, V.; Sundararajan, P. R.; Sarojadevi, M. High Performance Polymers 2013, 25, 87. (7) Vahabi, H.; Ferry, L.; Longuet, C.; Otazaghine, B.; Negrell -Guirao, C.; David, G.; Lopez -Cuesta, J. M. Materials Chemistry and Physics 2012, 136, 762. (8) Vahabi, H.; Eterradossi, O.; Ferry, L.; Longuet, C.; Sonnier, R.; Lopez -Cuesta, J. M. European Polymer Journal 2013, 49, 319. (9) Li, L.; Li, X.; Yang, R. Journal of Applied Polymer Science 2012, 124, 3807. (10) Fox, D. M.; Lee, J.; Zammarano, M.; Katsoul is, D.; Eldred, D. V.; Haverhals, L. M.; Trulove, P. C.; De Long, H. C.; Gilman, J. W. Carbohydrate Polymers 2012, 88, 847. (11) Wang, X.; Xuan, S.; Song, L.; Yang, H.; Lu, H.; Hu, Y. Journal of Macromolecular Science Part B -Physics 2012, 51, 255. (12) Chrissafis, K.; Bikiaris, D. Thermochimica Acta 2011, 523, 1. (13) Ni, Y.; Zheng, S. X. Chemistry of Materials 2004, 16, 5141. (14) Gnanasekaran, D.; Reddy, B. S. R. Polymer Composites 2012, 33, 1197. (15) Guenthner, A. J.; Lamison, K. R.; Lubin, L. M.; Hadda d, T. S.; Mabry, J. M. Industrial & Engineering Chemistry Research 2012, 51, 12282. (16) Rizvi, S. B.; Yildirimer, L.; Ghaderi, S.; Ramesh, B.; Seifalian, A. M.; Keshtgar, M. International journal of nanomedicine 2012, 7, 3915. (17) Vila Ramirez, N.; Sanch ez-Soto, M. Polymer Composites 2012, 33, 1707. (18) Blanco, I.; Abate, L.; Bottino, F. A.; Bottino, P. Polymer Degradation and Stability 2012, 97, 849. (19) Jin, L.; Ishida, H. Polymer Composites 2011, 32, 1164. 69 (20) Zheng, L.; Waddon, A. J.; Farris, R. J.; Coughlin, E. B. Macromolecules 2002, 35, 2375. (21) Cardiano, P.; Lazzara, G.; Manickam, S.; Mineo, P.; Milioto, S.; Lo Schiavo, S. European Journal of Inorganic Chemistry 2012, 5668. (22) Geng, Z.; Ba, J.; Zhang, S.; Luan, J.; Jiang, X.; Huo, P.; Wan g, G. Journal of Materials Chemistry 2012, 22, 23534. (23) Ke, F.; Zhang, C.; Guang, S.; Xu, H. Journal of Applied Polymer Science 2013, 127, 2628. (24) Phillips, S. H.; Haddad, T. S.; Tomczak, S. J. Current Opinion in Solid State & Materials Science 2004, 8, 21. (25) Seurer, B.; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (26) Kim, G. M.; Qin, H.; Fang, X.; Sun, F. C.; Mather, P. T. Journal of Polymer Science Part B-Polymer Physics 2003, 41, 3299. (27) Kannan, R. Y.; Salacinsk i, H. J.; Sales, K. M.; Butler, P. E.; Seifalian, A. M. Cell Biochemistry and Biophysics 2006, 45, 129. (28) Markovic, E.; Constantopolous, K.; Cordes, D. B.; Lickiss, P. D.; Ward, A. J.; Masters, A. F.; Williams, K. G.; Gido, S. P.; DeArmitt, C.; Iacono, S. T.; Peloquin, A. J.; Hartmann -Thompson, C.; Brandhorst Jr, H. W.; Ghanbari, H.; Marashi, S. M. Applications of Polyhedral Oligomeric Silsesquioxanes ; Springer: Netherlands, 2011. (29) Li, G.; Wang, L.; Ni, H.; Pittman, C. U. Journal of Inorganic and Org anometallic Polymers 2001, 11, 123. (30) Haddad, T. S.; Lichtenhan, J. D. Macromolecules 1996, 29, 7302. (31) Lichtenhan, J. D.; Otonari, Y. A.; Carr, M. J. Macromolecules 1995, 28, 8435. (32) Mather, P. T.; Jeon, H. G.; Romo -Uribe, A.; Haddad, T. S.; Lich tenhan, J. D. Macromolecules 1999, 32, 1194. (33) Tsuchida, A.; Bolln, C.; Sernetz, F. G.; Frey, H.; Mulhaupt, R. Macromolecules 1997, 30, 2818. (34) Milliman, H. W.; Ishida, H.; Schiraldi, D. A. Macromolecules 2012, 45, 4650. (35) Zeng, K.; Liu, Y.; Zheng , S. European Polymer Journal 2008, 44, 3946. (36) Xu, H. Y.; Kuo, S. W.; Lee, J. S.; Chang, F. C. Polymer 2002, 43, 5117. (37) Liu, S.; Yi, L. -Z.; Liang, Y. -Z. Journal of Separation Science 2008, 31, 2113. (38) Kuo, S. -W.; Chang, F. -C. Progress in Polymer Science 2011, 36, 1649. (39) Willard, J. J.; Wondra, R. E. Textile Research Journal 1970, 40, 203. 70 (40) McCusker, C.; Carroll, J. B.; Rotello, V. M. Chemical Communications 2005, 996. (41) Gitli, T.; Silverstein, M. S. Polymer 2011, 52, 107. (42) Amir, N.; Levina, A.; Silverstein, M. S. Journal of Polymer Science Part a -Polymer Chemistry 2007, 45, 4264. (43) Gurevitch, I.; Silverstein, M. S. Journal of Polymer Science Part a -Polymer Chemistr y 2010, 48, 1516. (44) Normatov, J.; Silverstein, M. S. Polymer 2007, 48, 6648. (45) Normatov, J.; Silverstein, M. S. Journal of Polymer Science Part a -Polymer Chemistry 2008, 46, 2357. (46) Normatov, J.; Silverstein, M. S. Chemistry of Materials 2008, 20, 1571. (47) Silverstein, M. S.; Shach -Caplan, M.; Khristosov, M.; Harel, T. Plasma Processes and Polymers 2007, 4, 789. (48) Wright, M. E.; Schorzman, D. A.; Feher, F. J.; Jin, R. Z. Chemistry of Materials 2003, 15, 264. (49) Brunsvold, A. L.; Mi nton, T. K.; Gouzman, I.; Grossman, E.; Gonzalez, R. High Performance Polymers 2004, 16, 303. (50) Minton, T. K.; Wright, M. E.; Tomczak, S. J.; Marquez, S. A.; Shen, L.; Brunsvold, A. L.; Cooper, R.; Zhang, J.; Vij, V.; Guenthner, A. J.; Petteys, B. J. Acs Applied Materials & Interfaces 2012, 4, 492. (51) Hoque, M. A.; Kakihana, Y.; Shinke, S.; Kawakami, Y. Macromolecules 2009, 42, 3309. (52) Seino, M.; Hayakawa, T.; Ishida, Y.; Kakimoto, M. -a.; Watanabe, K.; Oikawa, H. Macromolecules 2006, 39, 3473. (53) Gilman, J. W.; Schlitzer, D. S.; Lichtenhan, J. D. Journal of Applied Polymer Science 1996, 60, 591. (54) Lichtenhan, J. D. Comments on Inorganic Chemistry 1995, 17, 115. (55) Katoh, T.; Kikukawa, T.; Ito, K.; Yoshida, K.; Yamamoto, Y.; Japan . 2006, p 21pp. (56) Lichtenhan, J. D.; Vu, N. Q.; Carter, J. A.; Gilman, J. W.; Feher, F. J. Macromolecules 1993, 26, 2141. 71 (57) Espinas, J.; Pelletier, J. D. A.; Abou -Hamad, E.; Emsley, L.; Basset, J. -M. Organometallics 2012, 31, 7610. (58) Hay, M. T.; Seurer, B.; Holmes, D.; Lee, A. Macromolecules 2010, 43, 2108. (59) Aminuzzaman, M.; Watanabe, A.; Miyashita, T. Journal of Materials Chemistry 2008, 18, 5092. (60) Goto, Y.; Yoshida, K.; Sawada, H. Colloid and Polymer Science 2011, 289, 1493. (61) Kucuk, A. C.; Matsui, J.; Miyashita, T. Journal of Colloid and Interface Science 2011, 355, 106. (62) Wei, K.; Wang, L.; Zheng, S. Polymer Chemistry 2013, 4, 1491. (63) Wang, L.; Zhang, C.; Zheng, S. Journal of Materials Chemistry 2011, 21, 19344. (64) Vij, V.; Haddad, T. S.; Yandek, G. R.; Ramirez, S. M.; Mabry, J. M. Silicon 2012, 4, 267. (65) Watanabe, Y.; Sakai, Y.; Shibasaki, Y.; Ando, S.; Ueda, M.; Oishi, Y.; Mori, K. Macromolecules 2002, 35, 2277. (66) Maier, G. Progress in Polymer Scien ce 2001, 26, 3. (67) Leu, C. M.; Chang, Y. T.; Wei, K. H. Chemistry of Materials 2003, 15, 3721. (68) Leu, C. M.; Reddy, G. M.; Wei, K. H.; Shu, C. F. Chemistry of Materials 2003, 15, 2261. (69) Leu, C. M.; Chang, Y. T.; Wei, K. H. Macromolecules 2003, 36, 9122. (70) Seino, M.; Hayakawa, T.; Ishida, Y.; Kakimoto, M.; Watanabe, K.; Oikawa, H. Macromolecules 2006, 39, 3473. (71) Mitsuishi, M.; Zhao, F.; Kim, Y.; Watanabe, A.; Miyashita, T. Chemistry of Materials 2008, 20, 4310. (72) Lee, W.; Ni, S .; Deng, J.; Kim, B. -S.; Satija, S. K.; Mather, P. T.; Esker, A. R. Macromolecules 2007, 40, 682. (73) Gunawidjaja, R.; Huang, F.; Gumenna, M.; Klimenko, N.; Nunnery, G. A.; Shevchenko, V.; Tannenbaum, R.; Tsukruk, V. V. Langmuir 2009, 25, 1196. (74) Garcia, J. M.; Garcia, F. C.; Serna, F.; de la Pena, J. L. Progress in Polymer Science 2010, 35, 623. (75) Li, G.; Wang, Z. Macromolecules 2013, 46, 3058. 72 (76) Slegt, M.; Overkleeft, H. S.; Lodder, G. European Journal of Organic Chemistry 2007, 5364. (77) Wang, L.; Du, W.; Wu, Y.; Xu, R.; Yu, D. Journal of Applied Polymer Science 2012, 126, 150. (78) Stock, A. Berichte Der Deutschen Chemischen Gesellschaft 1916, 49, 108. (79) Baudrillard, V.; Davoust, D.; Ple, G. Magnetic Resonance in Chemistry 1994, 32, 40. (80) Holmes, D.; Johnson, K. Determining C -H Connectivity: gHMQC and gHMBC ; http://www2.chemistry.msu.edu/facilities/nmr/handouts/gHMQC.pdf , 2010. (81) Holmes, D.; Johnson, K. The Gradient -Selected COSY (gCOSY) Experiment; http://www2.chem istry.msu.edu/facilities/nmr/handouts/gCOSY.pdf 2011. (82) Raya, J.; Hirschinger, J.; Ovarlez, S.; Giulieri, F.; Chaze, A. -M.; Delamare, F. Physical Chemistry Chemical Physics 2010, 12, 14508. (83) Liepinsh, E.; Sekacis, I.; Lukevics, E. Magnetic Reson ance in Chemistry 1985, 23, 10. (84) Myasnikov, S. K.; Uteshinsky, A. D.; Kulov, N. N. Theoretical Foundations of Chemical Engineering 2009, 43, 227. (85) Tadie, M.; Bahadur, I.; Reddy, P.; Ngema, P. T.; Naidoo, P.; Deenadayalu, N.; Ramjugernath, D. Journal of Chemical Thermodynamics 2013, 57, 485. (86) Wang, T. -C.; Li, Y. -J.; Chen, Y. -P. Journal of Chemical and Engineering Data 2012, 57, 3519. (87) Tam Le, M.; Lorenz, H.; Seidel -Morgenstern, A. Chemical Engineering & Technology 2012, 35, 1003. (88) Ji, H. -Z.; Meng, X. -C.; Zhao, H. -K. Journal of Chemical and Engineering Data 2010, 55, 2590. (89) Lohmann, J.; Ropke, T.; Gmehling, J. Journal of Chemical and Engineering Data 1998, 43, 856. (90) Prausnitz, J. M.; Lichtenthaler, R. N.; Azevedo, E. G . d. Molecular thermodynamics of fluid -phase equilibria ; 3rd ed.; Prentice Hall PTR: Upper Saddle River, N.J., 1999. (91) Gmehling, J. G.; Anderson, T. F.; Prausnitz, J. M. Industrial & Engineering Chemistry Fundamentals 1978, 17, 269. 73 (92) Elliott, J. R.; Lira, C. T. Introductory chemical engineering thermodynamics ; 2nd ed.; Prentice Hall: Upper Saddle River, NJ, 2012. (93) Renon, H.; Prausnit.Jm Industrial & Engineering Chemistry Process Design and Development 1969, 8, 413. (94) Gavezzotti, A. Journ al of the Chemical Society -Perkin Transactions 2 1995, 1399. (95) Pinal, R. Organic & Biomolecular Chemistry 2004, 2, 2692. (96) Plass, K. E.; Engle, K. M.; Matzger, A. J. Journal of the American Chemical Society 2007, 129, 15211. (97) Slovokhotov, Y . L.; Neretin, I. S.; Howard, J. A. K. New Journal of Chemistry 2004, 28, 967. (98) Lloyd, M. A.; Patterson, G. E.; Simpson, G. H.; Duncan, L. L.; King, D. P.; Fu, Y.; Patrick, B. O.; Parkin, S.; Brock, C. P. Acta Crystallographica Section B -Structural Science 2007, 63, 433. (99) Sysoev, S. V.; Cheremisina, T. N.; Zelenina, L. N.; Tkachev, S. V.; Zherikova, K. V.; Morozova, N. B.; Kuratieva, N. V. Journal of Thermal Analysis and Calorimetry 2010, 101, 41. (100) Slovokhotov, Y. L.; Batsanov, A. S.; How ard, J. A. K. Structural Chemistry 2007, 18, 477. (101) Aoki, K.; Nakagawa, M.; Ichimura, K. Journal of the American Chemical Society 2000, 122, 10997. (102) Patra, A.; Richert, C. Journal of the American Chemical Society 2009, 131, 12671. (103) Aake roy, C. B.; Forbes, S.; Desper, J. Journal of the American Chemical Society 2009, 131, 17048. (104) Simperler, A.; Watt, S. W.; Bonnet, P. A.; Jones, W.; Motherwell, W. D. S. Crystengcomm 2006, 8, 589. (105) Kuduva, S. S.; Sarma, J.; Katz, A. K.; Carre ll, H. L.; Desiraju, G. R. Journal of Physical Organic Chemistry 2000, 13, 719. (106) Plato, C.; Glasgow, A. R. Analytical Chemistry 1969, 41, 330. (107) Yamashita, H.; Hirakura, Y.; Yuda, M.; Teramura, T.; Terada, K. Pharmaceutical Research 2013, 30, 70. (108) He, B.; Martin, V.; Setterwall, F. Fluid Phase Equilibria 2003, 212, 97. 74 (109) Lane, G. A. Solar heat storage : latent heat materials ; CRC Press: Boca Baton, Fla., 1983. (110) Walas, S. M. Phase equilibria in chemical engineering ; Butterwort h: Boston, 1985. (111) Jones, W.; Motherwell, S.; Trask, A. V. Mrs Bulletin 2006, 31, 875. (112) Stahly, G. P. Crystal Growth & Design 2007, 7, 1007. (113) Bhogala, B. R.; Nangia, A. New Journal of Chemistry 2008, 32, 800. (114) Yoshimoto, N.; Naka mura, T.; Suzuki, M.; Sato, K. Journal of Physical Chemistry 1991, 95, 3384. (115) Wei, D.; Jin, K. Journal of Chemical Thermodynamics 2009, 41, 145. (116) Wei, D.; Zhang, X.; Li, H. Journal of Chemical Thermodynamics 2013, 60, 94. (117) Coutinho, J. A. P. Fluid Phase Equilibria 1999, 158, 447. (118) Thompson, C. M.; Hergenrother, P. M. Macromolecules 2002, 35, 5835. (119) Simone, C. D.; Scola, D. A. Macromolecules 2003, 36, 6780. (120) Hergenrother, P. M.; Connell, J. W.; Smith, J. G. Polymer 2000, 41, 5073. (121) Smith, J. G.; Connell, J. W.; Hergenrother, P. M. Journal of Composite Materials 2000, 34, 614. (122) Connell, J. W.; Smith, J. G.; Hergenrother, P. M.; Rommel, M. L. High Performance Polymers 2000, 12, 323. (123) Smith, J. G.; C onnell, J. W. High Performance Polymers 2000, 12, 213. (124) Jensen, B. J.; Bryant, R. G.; Smith, J. G.; Hergenrother, P. M. Journal of Adhesion 1995, 54, 57. (125) Chen, J.; Zuo, H.; Fan, L.; Yang, S. High Performance Polymers 2009, 21, 187. (126) Zuo, H. J.; Chen, J. S.; Yang, H. X.; Hu, A. J.; Fan, L.; Yang, S. Y. Journal of Applied Polymer Science 2008, 107, 755. (127) J., G. In Encyclopedia of Polymer Science and Technology ; John Wiley & Sons: New York, 2005; Vol. 3, p 558. 75 (128) J., P. In Encyclopedia of Polymer Science and Engineering ; John Wiley & Sons: New York, 1988; Vol. 11, p 381. (129) Fedorov, A. A.; Savinov, V. M.; Sokolov, L. B. Polymer Science U.S.S.R. 1970, 12, 2475. (130) Chaudhuri, A. K.; Min, B. Y.; Pearce, E. M. Journal of Polymer Science: Polymer Chemistry Edition 1980, 18, 2949. (131) Imai, Y.; Oishi, Y. Progress in Polymer Science 1989, 14, 173. (132) Oishi, Y.; Kakimoto, M.; Imai, Y. Macromolecules 1987, 20, 703. (133) Oishi, Y.; Kakimoto, M. A.; Imai, Y. Macromol ecules 1988, 21, 547. (134) Cai, H.; Yu, J.; Qiu, Z. Polymer Engineering & Science 2012, 52, 233. (135) Arai, F.; Takeshita, H.; Dobashi, M.; Takenaka, K.; Miya, M.; Shiomi, T. Polymer 2012, 53, 851. (136) Scott, R. L. Journal of Chemical Physics 1949, 17, 268. (137) Vanhaecht, B.; Willem, R.; Biesemans, M.; Goderis, B.; Basiura, M.; Magusin, P.; Dolbnya, I.; Koning, C. Macromolecules 2004, 37, 421. (138) Murthy, N. S. Journal of Polymer Science Part B -Polymer Physics 2006, 44, 1763. (139) Arman d, M.; Endres, F.; MacFarlane, D. R.; Ohno, H.; Scrosati, B. Nature Materials 2009, 8, 621. (140) Lu, Y.; Moganty, S. S.; Schaefer, J. L.; Archer, L. A. Journal of Materials Chemistry 2012, 22, 4066. (141) Lu, W.; Goering, A.; Qu, L.; Dai, L. Physical Chemistry Chemical Physics 2012, 14, 12099. (142) Moganty, S. S.; Jayaprakash, N.; Nugent, J. L.; Shen, J.; Archer, L. A. Angewandte Chemie -International Edition 2010, 49, 9158. (143) Tanaka, K.; Ishiguro, F.; Chujo, Y. Polymer Journal 2011, 43, 708. (144) Tanaka, K.; Chujo, Y. Journal of Materials Chemistry 2012, 22, 1733. (145) Colovic, M.; Jerman, I.; Gaberscek, M.; Orel, B. Solar Energy Materials and Solar Cells 2011, 95, 3472. 76 !"#$%&'() ! "#$%&'"("!)*!+)$)+'," ! Keywords double decker silsesquioxane , NMR, 1H; 29Si NMR , cis/trans isomers 77 !" #$%&'()*)+,-+.,%,/(0) +!"1 2%&0,345&*,% +As previously mentioned in the background section , a recently developed class of nano -structured silsesquioxanes provides a unique opportunity to investigate and characterize the influence of cis and trans configurations on the physical and chemical properties of inorganic -organic hybrid materials (Figure 3-1). The cis and trans descriptors characterize the orientation of the X and R moieties with respect to the Si -O core of the silsesquioxane . This class of Figure 3-1. Trans (left) and cis (right) isomers of DDSQ(X)(R). silsesquioxanes are formally known as double decker silsesquioxanes (DDSQ) because they are comprised of two ÒdecksÓ of silsesquioxanes stacked on top of one another forming a cage -like structure. 1 Prior to the advent of DDSQ, the majority of cage -like silsesquioxanes did not incorporate cis and trans isomers. 2-5 Of the few cage -like silsesquioxanes that did incorporate geometric isomers, none have been synthesized in large quantities. 6,7 Cage -like silsesquioxanes have demonstrated superior properties over their organic counterparts in areas such as: thermal and mechanical properties, 8-11 solubility, 12-14 flame retardance, 15-23 oxidative resistance, 24-27 and dielectric properties .28-30 For this dissertation, four different DDSQ molecules were synthesized ( Figure 3-2). The R -moiety was varied from methyl to cyclohexyl while the X -SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOXSiRXSiRtrans DDSQSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOXSiRRSiXcis DDSQ 78 moiety was meta -amninophenyl. Additionally, the X -moiety was changed from meta -aminophenyl to para -aminophenyl while the R -moiety was methyl. !Figure 3-2. Cis and trans isomers of compounds 3a-d. The four structures studied in this work and the ir notation is described in this section. The naming scheme for these DDSQ will be abbreviated as DDSQ(X)(R), where X is the X -moiety, SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2Ncis DDSQ(m-AP)(Me) trans DDSQ(m-AP)(Cy) cis DDSQ(m-AP)(Cy) cis DDSQ(p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CNH2H2NSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOMeSiSiH2NNH2trans DDSQ(p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCySiH2NSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCySiNH2H2Ntrans DDSQ(m-AP)(Me) cis DDSQ(m/p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2NH2Ntrans DDSQ(m/p-AP)(Me) H3CH2NNH2NH2NH2NH2NH2 79 and R is the R -moiety ( Figure 3-1). The first set of cis and trans isomers, [( meta-amino phenyl) methylsilyl] -bridged -(phenyl) 8-double -decker silsesquioxane, will be designated as DDSQ(m -AP)(Me), and will be represented with shorthand notation -3a. The second set of isomers [( meta-amino phenyl) cyclohexylsilyl] -bridged -(phenyl) 8-double -decker silsesquioxane, will be designated DDSQ(m -AP)(Cy) and represented by shorthand notation -3b. The third set of isomers [(para -amino phenyl) methylsilyl] -bridged -(phen yl) 8-double -decker silsesquioxane, will be designated by DDSQ(p -AP)(Me) and represented by shorthand notation -3c, and [(meta/para -amino phenyl) methylsilyl] -bridged -(phenyl) 8-double -decker silsesquioxane, will be designated by DDSQ( m/p-AP)(Me) and represente d by shorthand notation -3d. Compound s 3a and 3b are both meta - with respect to the aminophenyl (X) moiety, but differ at the organic (R) moiety (Figure 3-2). Compound 3a has a methyl moiety, whereas 3b has a cyclohexyl moiety. Compound 3c is para - with respect to the aminophenyl group an d has a methyl moiety . Compound 3d incorporates once side t hat is meta - and one side that is para - with respect to the aminophenyl group and has a methyl moiety. !"#$%&'()*+,$-*.$/)-0)*+, $Tricycle[7.3.3(3,7)]octasiloxane -5,14,17-tetraol -1,3,5,7,9,11, 14,17-octaphenyl (Ph 8tetrasilanol -POSS) was obtained from Hybri d Plastics (Hattiesburg, MS). Tetrahydrofuran (THF), hexanes, diethyl ether, magnesium turnings, triethylamine, trichloromethyl silane, trichloro cyclohexyl silane, 3-[bis(trimethylsilyl)amino]phenyl -magnesium chloride , and 4 -[bis(trimethylsilyl)amino]phenyl (bromo)magnesium were obtained from Sigma -Aldrich. The solvents were distilled under nitrogen and degassed using Freeze -Pump -Thaw methods. 80 !"!#$%&#'()*+,-'*-(.# #Compounds 3a-c were measured at 25 ¼C on a Varian UNITY -Inova 600 spectrometer equipped with a 5 mm Pulsed -Field-Gradient (PFG) switchable broadband probe and operating at 599.80 MHz ( 1H) and 119.16 MHz ( 29Si). 1H NMR data were acquired using a recycle delay of 20 s and 32 scans to ensure accurate integration. The pulse angle was set to 45 ¡. The 1H-chemical shifts were referenced to that of residual protonated solvent in CDCl 3 (7.24 ppm). 29Si{1H} NMR data were acquired using a recycle delay of 12 s with inverse -gated decoupling. The pulse angle was set to 90 ¡. 29Si{1H} spectra were referenced agains t the lock solvent using vendor supplied lock referencing. 13C{1H} NMR data were acquired using a recycle delay of 1 s and 256 sc ans. The pulse angle was set to 45 ¡. !"/#0.1+2)'3'# #All compounds, 3a-d, were synthesized by c apping DDSQ tetrasilanol with dichlorosilanes ( Scheme 3-1).31,32 The synthetic procedure generated a mixture of cis and trans isomers ( Figure 3-2). Scheme 3-1. Synthesis of DDSQ(AP )(R); X = Cl/Br, R = Me/Cy. SiOSiOSiOHSiOSiHOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOHHOMgY RCl3SiTHFSiClRCl1. Et3N, THF2. H+, MeOH NSiMe 3Me3SiNSiMe 3Me3SiPh8tetrasilanol then Et2OSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiRSiRH2NDDSQ(X)(R) H2N 81 !"#"$%&'()'*+,%!- %%Under a nitrogen atmosphere in a drybox, a solution of 1 M 3-[bis(trimethylsilyl)amino] phenyl(chloro)magnesium (60 mL, 60 mmol ) was added dropwise via an addition funnel to a stirred solution of trichloromethyl silane (11.077 g, 112 mmol ) and THF (20 mL). The solution was stirred for 20 h at 25 ¡C and then purifi ed by fra ctional distillation i to obtain (N-trimethylsilyl) 2-aniline -3-(dichloromethylsilane) (18.66 g, 53.24 mmol , 88 % yield) as a colorless liquid . 29Si{1H} NMR, !: 18.40 (1 Si), 5.22 (2 Si). 13C{1H} NMR, !: 148.43, 134.74, 133.82, 133.62, 130.35, 128.8, 128.58, 128.16, 123.73, 5.71, 2.27. 1H NMR !: 7.48 (1 H, multiplet), 7.35 (1 H, multiplet), 7.30 (1 H, multiplet), 7.12 (1 H multiplet), 1.06 (3 H, CH 3, singlet), 0.14 (18 H, TMS, singlet) (Figure 3-3). Figure 3-3. 1H NMR spectrum of ( N-trimethylsilyl) 2-aniline -3-(dichloromethylsilane). A solution of (N -trimethylsilyl) 2-aniline -3-(dichloromethylsilane) (3.03 g, 8.64 mmol) and triethylamine (2.01 g, 19.90 mmol) in THF (10 mL) was added dropwise via an addition funnel into a stirred solution of Ph 8tetrasilanol -POSS (4.50 g, 4.21 mmol) at 25 ¡C THF (40 mL). After 30 minutes, the HNEt 3Cl precipitate was separated by filtration, and the solvent was removed from the filtrate under vacuum. To the resultant residue diethyl ether (2 mL) was added followed by acidified methanol, which gave a white suspension that was stirred at 25 ¡C for 20 h . 82 The heterogeneous mixture was filtered, and the retentate was dried under nitrogen to yield 3a, (5.379 g, 4.03 mmol, 96 % yield). 29Si{1H} NMR, !: -30.6 (4 Si), -78.4 (8 Si), -79.4 (2 Si), -79.6 (4 Si), -79.8 (2 Si). 13C{1H} NMR, !: 145.99, 137.24, 134.27, 134.15, 132.03, 131.05, 130.87, 130.50, 130.43, 130.34, 128.96, 127.91, 127.81, 127.72, 127.62, 123.70, 119.92, 116.81, -0.40. 1H NMR !: 7.54 -6.92 (40 H, overlapping multiplets), 6.92 (4 H, multiplet), 6.67 (4 H, multiplet ) 3.16 (4 H, NH 2, broad singlet), 0.51 (6H, CH 3, 2 overlapping singlet s)ii (Figure 3-4). Figure 3-4. NMR spectr a of compound 3a (a) 29Si and (b) 1H. !"#"$%&'()'*+,%!-% %Under a nitrogen atmosphere in a drybox, a solution of 1 M 3-[bis(trimethylsilyl) amino]phenyl(chloro)magnesium ( 16.99 mL, 16.99 mmol ) was added dropwise via an addition funnel to a stirr ed solution of cyclohexyl trichloro silane ( 3.69 g, 16.99 mmol ) and THF ( 5 mL). The solution was stirred for 20 h at 25 ¡C and then purified by Kugelrohr distillation i to obtain (N-trimethylsilyl) 2-aniline -3-(cyclohexyl dichloro silane) ( 12.64 g, 53.24 mmol , 74 % yield) as a 7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppm7.77.57.37.16.96.76.5b) 83 colorless liquid. 29Si{1H} NMR, !: 18.45 (1 Si), 5.12 (2 Si). 13C{1H} NMR, !: 148.17, 135.55, 133.40, 131.77, 130.32, 129.04, 128.68, 128.53, 123.64, 30.72, 27.29, 26.45, 26.02, 2.20. 1H NMR !: 7.41 (1 H, multiplet), 7.30 (1 H, multiplet), 7.23 (1 H, multiplet), 7.04 (1 H multiplet), 1.78, 1.71, 1.29 (11 H, C 6H11, multiplets), 0.0 8 (18 H, TMS, singlet) ( Figure 3-5). Figure 3-5. 1H NMR spectrum of ( N-trimethylsilyl) 2-aniline -3-(cyclohex yl dichlorosilane). A solution of (N -trimethylsilyl) 2-aniline -3-(dichloromethylsilane) ( 1.00 g, 2.39 mmol) and triethylamine (0.48 g, 4.73 mmol) in THF (5 mL) was added dropwise via an addition funnel into a stirred solution of Ph 8tetrasilanol -POSS (1.26 g, 1.18 mmol) at 25 ¡C THF (5 mL). After 30 minutes, the HNEt 3Cl precipitate was separated by filtration, and the solvent was removed from the filtrate under vacuum. To the resultant residue diethyl ether (2 mL) was added followed by acidified methanol, which gave a white suspension that was stirred at 25 ¡C for 20 h . The heterogeneous mixture was filtered, and the retentate was dried under nitrogen to yield 3b, (1.19 g, 0.81 mmol, 69 % yield). 29Si{1H} NMR, !: -34.1 (4 Si), -78.5 (8 Si), -79.3 (2 Si), -79.5 (4 Si), -79.7 (2 Si ). 13C{1H} NMR, !: 145.82, 135.62, 134.26, 134.23, 132.24, 131.10, 130.86, 130.71, 130.43, 130.32, 128.81, 127.89, 127.76, 127.63, 127.49, 124.43, 120.72, 116.75, 27.64, 26.81, 26.43, 26.22, 2.15, 1.49. 1H NMR !: 7.57 -6.97 (40 H, overlapping multiplets), 6.83 (4 H, 84 multiplet), 6.67 (4 H, multiplet ) 3.24 (4 H, NH 2, broad singlet), 1.79, 1.58, 1.19, 1.08, 1.03 (11 H, C 6H11, multiplets ) (Figure 3-6). Figure 3-6. NMR spectr a of compound 3b (a) 29Si and (b) 1H. Two impurities were found and removed through solvent washing. Tolu ene was added to dry 3b until most solid was dissolved. The remaining precipitate was mostly impurity and was separated by filtration. Toluene was removed from the filtrate under vacuum . Diethyl ether was then added to the resultant residue, which gave a w hite suspension that was separat ed by filtration and the rententate was dried under Nitrogen to yield pure 3b. It was later determine d the product could also be purified by column chromatography. !"#"!$%&'(&)*+$!,$ $Compound 3c was synthesized based on the same described method , except the Grignard was synthes ized and not purchased (Scheme 2) . Briefly, under a nitrogen atmosphere in a drybox, a solution of 4-bromo -N,N -bis(trimethylsilyl)aniline (9.48 g, 30.0 mmol ) was added 7.06.05.04.03.02.01.0ppm7.77.57.37.16.96.76.5b) 85 dropwise via an addition funnel into a stirred solution of magnesium turnings (0.91 g, 37.4 mmol ) and THF (25 mL). The solution was stirred for 20 h and then used crude in the following reaction. 29Si{1H} NMR, !: 12.20 (2 Si) . Crude 4 -[bis(tri methylsilyl)amino]phenyl(bromo)magnesium (30.0 mmol ) was added dropwise via an addition funnel into a stirred solution of trichloromethyl silane (5.431 g, 36.3 mmol ) and THF (10 mL). The solution was stirred for 20 h and then purified by fra ctional distill ation i to obtain (N-trimethylsilyl) 2-aniline -4-(dichloromethylsilane) (9.02 g, 25.7 mmol , 95 % yield) as a colorless liquid. 29Si{1H} NMR, !: 18.92 (1 Si), 5.18 (2 Si). 13C{1H} NMR, !: 152.31, 133.72 130.11, 127.78, 5.95, 2.32. 1H NMR !: 7.60 (2 H, multiplet), 7.01 (2 H, multiplet), 1.03 (11 H, CH3, singlet), 0.10 (18 H, TMS , singlet) ( Figure 3-7). Figure 3-7. 1H NMR spectrum of ( N-trimethylsilyl) 2-aniline -4-(cyclohecyl dichlorosilane). A solution of (N -trimethylsilyl) 2-aniline -4-(dichloromethylsilane) (2.95 g, 8.42 mmol ) and triethylamine (1.86 g, 18.4 mmol ) in THF (10 mL) was a dded dropwise via an addition funnel into a stirred solution of Ph 8tetrasilanol -POSS (4.46 g, 4.17 mmol ) in THF (40 mL). The reaction mixture became cloudy after the addition of a few drops of the dichloromethylsilane/ triethylamine solution, indicating fo rmation of insoluble HNEt 3Cl salt. 86 After 30 min, the HNEt 3Cl was separated by filtration, and the solvent was removed from the filtrate under vacuum. To the resultant residue diethyl ether (2 mL) was added followed by acidified methanol, which gave a white suspension that was stirred at 25 ¡C for 20 h . The heterogenous mixture was filtered, and the retentate was dried under nitrogen to yield 3c, (4.69 g, 3.51 mmol , 84 % yield). 29Si{1H} NMR, !: -29.7 (4 Si) , -78.2 (8 Si), -79.1 (2 Si), -79.3 (4 Si), -79.5 (2 Si). 13C{1H} NMR, !: 148.27, 135.13, 134.39, 134.35, 134.30, 134.22, 132.29, 131.40, 131.19, 130.98, 130.52, 130.47, 130.38, 130.24, 127.93, 127.88, 127.69, 127.52, 124.64, 114.56, -0.10. 1H NMR !: 7.55 -7.10 (44 H, overlapping multiplets), 6.53 (4 H, multiplet ), 3.69 (4H, NH2, broad singlet), 0.51 (6H, CH 3, 2 overlapping singlets) ii (Figure 3-8). Figure 3-8. NMR spectr a of compound 3c (a) 29Si and (b) 1H. 7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppm7.77.57.37.16.96.76.5b) 87 !"#"#$%&'(&)*+$!+$,' -./)01$&2$!3$3*+$!45$ $Compound 3d, was comprised of an AP moiety w ith a combination of meta- and par a-isomers (Figure 3-2). However , the synthetic procedure is anticipated to form 25 % compound 3a, 25 % compound 3c, and 50 % compound 3d. A solution of (N -trimethylsilyl) 2-aniline -3-(dichloromethylsilane) (0.679 g, 1.94 mmol) , (N-trimethylsilyl) 2-aniline -4-(dichloromethyl silane) (0.679 g, 1.94 mmol) , and triethylamine ( 0.792 g, 7.82 mmol) in THF (10 mL) was added dropwise via an addition funnel into a stirred solution of Ph 8tetrasilanol -POSS (2.05 g, 1.92 mmol) at 25 ¡C THF (40 mL). After 30 minutes, the HNEt 3Cl precipitate was separated by filtration, and the solvent was removed from the filtrate under vacuum. To the resultant residue diethyl ether (2 mL) was added followed by acidified methanol, which gave a white suspension that was stirred at 25 ¡C for 20 h . The heterogeneous mixture was filtered, and the retentate was dried under nitrogen to yield compound 3d, (2.348 g, 1.75 mmol, 92 % yield). 29Si{1H} NMR, !: -29.96 (1 Si), -29.99 (1 Si), -30.59 (1 Si), -30.61 (1 Si), -78.38 (4 Si), -78.49 (4 Si), -79.39 (2 Si), -79.59 (2 Si), -79.61 (2 Si), -79.79 (2 Si). 13C{1H} NMR, !: 1438.19, 145.94, 137.27, 135.08, 134.30, 134.28, 134.16, 132.25, 132.20, 132.11, 132.06, 131.30, 131.24, 131.17, 131.15, 130.94, 130.47, 130.41, 1130.32, 130.29, 128.96, 127.89, 127.85, 122.83, 127.80, 127.71, 127.70, 127.61, 127.52, 127.43, 124.68, 124.65, 123.76, 119.98, 116.83, 114.52, -0.37. 1H NMR !: 6.91 (1 H, mu ltiplets), 6.94 (2 H, multiplet), 6.68 (2 H, multiplet), 6.55 (4 H, multiplet), 3.45 (4 H, NH2, broad singlet), 0.51 (6 H, CH 3, 6 overlapping singlet s)iii (Figure 3-9). 88 Figure 3-9. NMR spectr a of compound 3d (a) 29Si and (b) 1H. !"#$%&'() $*+*(,-&-$./$-,+01'0&2 $34.2')54'- $ All experiments for compounds 3a-c were performed multiple times, and the results analyzed based upon yields ( Table 3-1). ÒLowest Ó and ÒhighestÓ refer to the lowest and highest yield reported of al l experiments for that reaction/compound . Reaction one refers to the dichlorosilane product, 2, and reaction two refers to the final product, 3 (Scheme 3-1). Only experiments that were performed more than twice have a reported ÒaverageÓ yield. For reaction one, all compounds had yields within a similar range. Overall, for reaction one, compounds 3a and c had higher yields than compound 3b. The scale of the reaction did not appear to have an affect on the yield. There were several more experiments associated with reaction two for all compounds. Compounds 3a and c had much higher average yields ( 88 ± 9 and 83 ± 15, respectively) over four experiments than compound 3b (43 ± 11) did over 10 experiments. Again, the scale of the reaction did not appear to have an affect on the yield. 7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppm7.77.57.37.16.96.76.5b) 89 Table 3-1. Yield analysis of the products of the reactions involved for compounds 3a-c. Reaction 1: Dichlorosilane Product mass (g) Yield % 3a lowest 16.68 77 highest 6.60 94 average -- -- 3b lowest 1.09 69* highest 5.29 74+ average -- -- 3c lowest 12.8 47 highest 9.00 81 average -- 81 ± 15 5 experiments Reaction 2: DDSQ(AP)(R) Product mass Yield % Overall yield % 3a lowest 4.76 84 65 highest 5.38 96 90 average -- 88 ± 9 4 experiments 3b lowest 0.057 29 20 highest 1.19 69 51 average -- 43 ± 11 10 experiments 3c lowest 4.69 84 51 highest 9.80 88 84 average -- 83 ± 15 4 experiments * Contained unreacted starting material. +Kugelrohr significantly improved purity . !"#$%&'()*+,'-$./01.23$&'$ 34'56/3,3 $ In this chapter, compounds 3a-d were successfully synthesized. A yield analysis was prepared for all experiments that were performed more than once, and high yields were acheived for all compounds except 3b. For additional information on synthesis including a deprotection analysis, plea se see appendix A . 90 !!! "#$%& ! 91 !"#$% & (i) The first fraction had slight impurities and was discarded. All remaining material was pure product. (ii ) One singlet represents the cis isomer protons (CH 3, 3H), and the other represents the trans isomer protons (CH 3, 3H). (iii ) There is one singlet representative of the methyl protons (CH 3) for each of the following geometries: cis meta, trans meta, cis para, trans para, and the reamining two singlets are representative of the cis and trans isomers of the structure that has 1 meta -aminophenyl and 1 para -aminophenyl on a single DDSQ. 92 !!! "#$#"#%&#' ! 93 !"#"!"$%"& ' (1) Takashi, K.; Takashi, K.; Masaya, I.; Kazuhiro, Y.; Yasuhiro, Y. Japan, 2006. (2) Li, G. Z.; Wang, L. C.; Ni, H. L.; Pittman, C. U. Journal of Inorganic and Organometallic Polymers 2001, 11, 123. (3) Phillips, S. H.; Haddad, T. S.; Tomczak, S. J. Current Opinion in Solid State & Materials Science 2004, 8, 21. (4) Joshi, M.; Butola, B. S. Journal of Macromolecular Science -Polymer Reviews 2004, C44, 389. (5) Kannan, R.; Salacinski, H.; Butler, P.; Seifalian, A. Accounts of Chemical Research 2005, 38, 879. (6) Lichtenhan, J. D.; Vu, N. Q.; Gilman, J. W.; Feher, F. J. United States, 1995; Vol. 5,412,053. (7) Lichtenhan, J. D.; Vu, N. Q.; Carter, J. A.; Gilman, J. W.; Feher, F. J. Macromolecules 1993, 26, 2141. (8) Yang, B.; Li, M.; Wu, Y.; Wan, X. Polymers & Polymer Composites 2013, 21, 37. (9) Wu, S.; Hayakawa, T.; Kikuchi, R.; Grunzinger, S.; Kakimoto, M. Macromolecules 2007, 40, 5698. (10) Wu, S.; Hayakawa, T.; Kakimoto, M.; Oikawa, H. Macromolecules 2008, 41, 3481. (11) Wu, J.; Mather, P. T. Polymer Reviews 2009, 49, 25. (12) Gnanasekaran, D.; Reddy, B. S. R. Polymer Composites 2012, 33, 1197. (13) Guenthner, A. J.; Lamison, K. R.; Lubin, L. M.; Haddad, T. S.; Mabry, J. M. Industrial & Engineering Chemistry Research 2012, 51, 12282. (14) Rizvi, S. B.; Yildirimer, L.; Ghaderi, S.; Ramesh, B.; Seifalian, A. M.; Keshtgar, M. International journal of na nomedicine 2012, 7, 3915. (15) Fan, H.; Yang, R. Industrial & Engineering Chemistry Research 2013, 52, 2493. (16) Rakesh, S.; Dharan, C. P. S.; Selladurai, M.; Sudha, V.; Sundararajan, P. R.; Sarojadevi, M. High Performance Polymers 2013, 25, 87. 94 (17) Vahabi, H.; Ferry, L.; Longuet, C.; Otazaghine, B.; Negrell -Guirao, C.; David, G.; Lopez -Cuesta, J. M. Materials Chemistry and Physics 2012, 136, 762. (18) Vahabi, H.; Eterradossi, O.; Ferry, L.; Longuet, C.; Sonnier, R.; Lopez -Cuesta, J. M. European Polym er Journal 2013, 49, 319. (19) Li, L.; Li, X.; Yang, R. Journal of Applied Polymer Science 2012, 124, 3807. (20) Fox, D. M.; Lee, J.; Zammarano, M.; Katsoulis, D.; Eldred, D. V.; Haverhals, L. M.; Trulove, P. C.; De Long, H. C.; Gilman, J. W. Carbohydrate Polymers 2012, 88, 847. (21) Wang, X.; Xuan, S.; Song, L.; Yang, H.; Lu, H.; Hu, Y. Journal of Macromolecular Science Part B -Physics 2012, 51, 255. (22) Chrissafis, K.; Bikiaris, D. Thermochimica Acta 2011, 523, 1. (23) Ni, Y.; Zheng, S. X. Chemistry of Materials 2004, 16, 5141. (24) Vila Ramirez, N.; Sanchez -Soto, M. Polymer Composites 2012, 33, 1707. (25) Blanco, I.; Abate, L.; Bottino, F. A.; Bottino, P. Polymer Degradation and Stability 2012, 97, 849. (26) Jin, L.; Ishida, H. Polymer Composites 2011, 32, 1164. (27) Zheng, L.; Waddon, A. J.; Farris, R. J.; Coughlin, E. B. Macromolecules 2002, 35, 2375. (28) Cardiano, P.; Lazzara, G.; Manickam, S.; Mineo, P.; Milioto, S.; Lo Schiavo, S. European Journal of Inorganic Chemistry 2012, 5668. (29) Geng, Z.; Ba, J.; Zhang, S.; Luan, J.; Jiang, X.; Huo, P.; Wang, G. Journal of Materials Chemistry 2012, 22, 23534. (30) Ke, F.; Zhang, C.; Guang, S.; Xu, H. Journal of Applied Polymer Science 2013, 127, 2628. (31) Seurer, B. ; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (32) Vij, V.; Haddad, T. S.; Yandek, G. R.; Ramirez, S. M.; Mabry, J. M. Silicon 2012, 4, 267. 95 !"#$%&'( !!!" #$%&'#(#)*'#+&"*&$",-*&'#(#)*'#+&"+(" !"# "*&$" $%&'#"!"#$"./"-0#&1 "2345607"839:)"&9; " Keywords NMR; 1H; 29Si; 2D NMR; 1HÐ29Si HMBC; double decker silsesquioxanes; cis/trans isomers * This chapter has already been published with the following citation: Schoen, B. W.; Holmes, D.; Lee, A. Magnetic Resonance in Chemistry 2013, 51, 490. 96 !" #$%&'()(*+'(,&-+&$-./+&'()(*+'(,&-,)- !"#-+&$-$%&'# -0+1*-/2(&3- 451678(-359:;-<9= -!"-4-#&'>,$/*'(,& - In this chapter , the 1H NMR spectra of cis and trans isomers of compounds 3a-c (Figure 4-1) are assigned and their ratios quantified . In order to unambiguously assign the proton Figure 4-1. Cis and trans isomers of (a) 3a, (b) 3b, (c) 3c; AP = aminophenyl . SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2Ncis DDSQ(m-AP)(Me) trans DDSQ(m-AP)(Cy) cis DDSQ(m-AP)(Cy) cis DDSQ(p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CNH2H2NSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOMeSiSiH2NNH2trans DDSQ(p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCySiH2NSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCySiNH2H2Ntrans DDSQ(m-AP)(Me) cis DDSQ(m/p-AP)(Me) SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2NH2Ntrans DDSQ(m/p-AP)(Me) H3CH2NNH2NH2NH2NH2NH2 97 resonances and use them for quantitation, two -dimensional ( 2D) NMR techniques was necessary . Specifically, proton correlations to the silicon nuclei of the silsesquioxane core not only facili tates 1H spectral assignment but also confirm s previous 29Si assignments for these DDSQ . !"#$%&'()*+,(+&-($( ./*/()'*-0/)-+1 $2').+3, $Compounds 3a-c were measured at 2 5 ¼C on a Varian UNITY -Inova 600 spectrometer equipped with a 5 mm Pulsed -Field-Gradient (PFG) switchable broadband probe and operating at 599.80 MHz ( 1H) and 119.16 MHz ( 29Si). One -dimensional (1D) 1H NMR data were acquired using a recycle delay of 20 s and 32 scans to ensure accurate integration. The pulse angle was set to 4 5¡. The 1H-chemical shifts were referenced to that of residual protonated solvent in CDCl 3 (7.24 ppm). 1D 29Si{1H} NMR data were acquired using a recycle delay of 12 s with inverse -gated decoupling. The pulse angle was set to 90 ¡. 29Si{1H} spectra were referenced against the lock solvent using vendor supplie d lock referencing. All two -dimensional (2D) NMR spectra were obtained using gradient pulses on a 5 mm PFG switchable broadband probe without sample spinning. Phase -sensitive spectra were acquired using the hyper -complex States method. Threefold linear p rediction was applied to the F1 dimension as implemented by standard Varian software. 1H-29Si-gHMBC (gradient enhanced Heteronuclear Multi -Bond Correlation) spectra were acquired with spectral widths of 6600 Hz and 14370 Hz for F2 and F1, respectively. 1 The pre-acquisition delay was set to 1.0 s, and 400 increments with 8 transients per increment containing 1069 data points were acquired. The three -bond J-filter was set to 7 Hz (Varian parameter; Jnxh = 7). Unshifted sine -bell windows, which were matched to the acquisition time, were used for processing both dimensions. Zero filling to 4096 data points was applied to F1 98 prior to 2D Fourier transformation. 1H-1H-gCOSY (gradient enhanced COrrelation SpectroscopY) spectra were obtained using a spectral width of 8000 Hz in both dimensions. The pre-acquisition delay was set to 1.0 s and 128 increments with 4 transients of 11 52 data points were acquired. Both F2 and F1 were multiplied by unshifted sine bell weighting functions that were matched to the acquisiti on or evolution time. Prior to 2D Fourier transformation, F2 and F1 were zero filled to 2048 and 1024 data points, respectively. !"#$%&'()*+*,-)*.($-(&$/0-()*+*,-)*.($-(-123*3 The isomeric mixture of each compound was asses sed by 29Si and 1H NMR, and diagnostic chemical shifts for each isomer were observed in the spectra. 2 The isomeric mixture of 3a shows the expected 29Si resonances at ! -30.6, -78.4, -79.4, -79.6, -79.8 in a ratio of 2:4:1:2:1 ( Figure 4-2a). Trans 3a shows characteristic 29Si resonances at ! -30.6, -78.4, -79.6 in the ratio of 2:4:4 ( Figure 4-2b) and cis 3a has 29Si resonances at ! -30.6, -78.4, -79.4, and -79.8 in a ratio of 2:4:2:2 ( Figure 4-2c).2 The 29Si resonance at ! -30.6 has been assigned to the D -group silicon atoms (Si -3), silicon atoms bonded to 2 oxygen atoms ( Figure 4-1).3 The 29Si resonance at ! -78.4 has been assigned to the T -group silicon atoms, silicon atoms bonded to 3 oxygen atoms, nearest to the D -group silicon atoms (Si -2). 29Si resonances at ! -79.4, -79.6, and -79.8 have been assigned to the internal T -group silicon atoms; Si -1cR, Si -1t, Si -1cL, respectively. Each Si -1t has the same chemical environment with proximity to one methyl and one aminophenyl group giving rise to a single silicon resonance . In contrast, Si -1cR atoms are only proximal to the methyl group and Si -1cL to the aminophenyl group thereby leading to two 99 resonances. The 29Si resonances for 3b and 3c are nearly identical; ! -34.1, -78.5, -79.32, -79.5, -79.7 (Figure 4-2d), and ! -29.7, -78.2, -79.1, -79.3, -79.5 in the ratio of 2:4:1:2:1 ( Figure 4-2e), resp ectively. Figure 4-2. 29Si NMR spectra of (a) cis/trans 3a, (b) majority trans 3a, (c) majority cis 3a, (d) cis/trans 3b, and (e) cis/trans 3c. The 1H NMR spectra for these DDSQ derivatives reveal that the proton resonances in the phenyl region (6. 5 Ð 8 ppm) are the best candidates for the quantitation of isomeric ratios (Figure 4-3). Proton signals belonging to the R -substituents (Me or Cy) proved to be ambiguous when attempting to decipher between cis and trans isomers. First, the proton signals belonging to the cyclohexyl groups are broad, significantly overlapped , and have complicated coupling patterns 100 for cis and trans isomers . The proton signals belonging to the methyl groups of each isomer are broad singlets and overlap, making it difficult to integrate the individual peaks in this region. Secondly, t he 29Si signals associated with both organic R -groups (Me and Cy) , as evidenced by Figure 4-3. 1H NMR spectrum of (a) cis/trans 3a, (b) majority trans 3a, (c) majority cis 3a, (d) cis/trans 3b, (e) cis/trans 3c, (f) majority trans 3c, (g) majority cis 3c. 7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppma)7.67.57.40.0516.00b)7.67.57.47.9116.00c)7.77.57.37.16.96.76.54.2416.00d)8.9815.7716.007.77.57.37.16.96.77.8510.0416.00g)ppmppmppmppm7.77.57.37.16.96.77.77.57.37.16.96.76.54.0116.00f)7.57.06.56.05.55.04.54.03.53.02.52.01.5ppm7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppm7.77.57.37.16.96.76.57.9211.84 16.00e) 101 1H Ð 29Si gHMBC cross -peaks, are isochronous for cis and trans isomers. This makes it impossible to use the previously established 29Si shift assignments to determine the cis and trans proton signa ls in these regions, even if the signals were deconvoluted. !"#"$%&'()'*+,%#- The proton resonance at !H 7.54 was assigned to H -2a and H -2a! (m, 16 H) based on the gHMBC three -bond correlation with the 29Si single resonance at !Si -78.4 (Si -2), which is isochronous for both cis and trans isomers (Figure 4-4a). Whereas the trans isomer exhibits a single resonance for Si -1t ( !Si -79.6), the cis isomer, due to its asymmetry, has two pairs of Si-1 atoms, labeled Si -1cL (left) and Si -1cR (right), ( !Si -79.4 and -79.8) in a ratio of 2:2. The proton multiplet at !H 7.48 was assigned to H -1acR and H -1acR« (cis, m, 4 H) as evidenced by the gHMBC correl ation with the two Si-1cR atoms ( !Si -79.8). The proton multiplet at !H 7.16 was also assigned to H -1acL and H -1acL« (cis, m, 4 H) based on the gHMBC cor relation with the two Si -1cL atoms ( !Si -79.4). For the trans isomers, the proton multiplet at !H 7.32 was assigned to H -1at and H -1at« (trans , m, 8 H), which is confirmed by the gHMBC correlation with the Si -1t atoms ( !Si -79.6). Both the cis and trans isomers show isochronous resonances at !H 6.92 and 7.03 and were assigned to H -3a and H -3e ( m, 4 H and m, 4 H) based on the gHMBC correlation with the four Si -3 atoms ( !Si -30.5). Analysis of the 1H-1H gCOSY spectrum reveal ed that proton signals at !H 6.67 and 7.08 show correlations with !H 6.92 and 7.03 thereby establishing their assignment as H-3b an d H-3d. Proton multiplet s assigned to H-1acL and H -1acL« (!H 102 7.16) and H -1at and H -1at« (!H 7.32) exhibit too much overlap for accurate quantification. Thus, the proton resonances assigned to H-1acR and H -1acR« (!H 7.48) and H -2a and H -2a! (!H 7.54) are the best candidates for quantification. An equivalent ratio of the cis to trans isomers would, therefore, be represented by a ratio of 1:4 for these resonances. The remaining protons were identified by 1H-1H gCOSY cro ss-peaks and coupling patterns ( Table 4-1). Figure 4-4. 1H- 29Si gHMBC connectivity of (a) cis/trans 3a and (b) cis/trans 3c. 103 Table 4-1. 29Si and 1H resonances of cis/trans 3a. Atom 1H (mult.) 29Si (mult.) 1 (Si) --- -79.4(2 Si) *, -79.6(4 Si) €, -79.8(2 Si) * 1-(C6H5) --- --- -a,a ! (CH) 7.16(m, 4H) *, 7.32(m, 8H) €,7.48(m, 4H) * --- -b,b! (CH) 7.31(m, 4H) *, 7.15(m, 8H) €, 7.23(m, 4H) * --- -c (CH) 7.39(m, 4H) ¤, 7.38(m, 4H) ¤, 7.25(m, 4H) ¤, 7.10(m, 4H) ¤ --- 2 (Si) --- -78.4 (8 Si) # 2-(C6H5) --- --- -a,a ! (CH) 7.54(m, 16H) # --- -b,b! (CH) 7.24(m, 16H) # --- -c (CH) 7.39(m, 4H) ¤, 7.38(m, 4H) ¤, 7.25(m, 4H) ¤, 7.10(m, 4H) ¤ --- 3 (Si) --- -30.6 (4 Si) # 3-(CH3) 0.51 (s, 12H) ¤ --- 3-(C6H6N) --- --- -a (CH) 6.92 (m, 4H) --- -b (NH 2) 3.16 (s, 8H) # --- -c (CH) 6.67 (m, 4H) # --- -d (CH) 7.08 (m, 4H) # --- -e (CH) 7.03 (m, 4H) # --- € = trans, * = cis, ¤ = cannot distinguish between cis/trans, # = isochronous cis/trans !"#"$%&'()'*+,%#- The 1H and 29Si NMR spectra of 3b shows an analogus coupling pattern to that of 3a with only a slight variation of chemic al shifts. Additionally, the 2D 1H Ð 29Si gHMBC is analogous to that of 3a (Figure 4-3d). !"#"#%&'()'*+,%#. Similar to compounds 3a and 3b, the proton resonance at "H 7.55 was assigned to H -2a and H -2a! (m, 16 H) based on the correlation with the 29Si single resonance at "Si -78.2 (Si-2), 104 isochronous for both cis and trans isomers (Figure 4-4b). The proton multiplet at !H 7.50 was assigned to H -1acR and H -1acR« (cis, m, 4 H) as evidenced by the gHMBC correlation with the Si-1cR atoms (!Si -79.5). The proton multiplet at !H 7.19 was also assigned to H -1acL and H -1acL« (cis, m, 4 H) based on the gHMBC correlation with the other t wo Si -1cL atoms ( !Si -79.1). For the trans isomers, the proton multiplet at !H 7.33 was assigned to H -1at and H -1at« (tran s, m, 8 H) based on the gHMBC correlation with the Si-1t atoms ( !Si -79.3). The proton multiplet at !H 6.53 (8H) was assigned to H -3a and H -3a« as evidenced by the gHMBC correl ation with the four Si -3 atoms, and is isochronous for cis and trans isomers. The assignment of H -3a and H -3a« i s unique to the para -structure ( 3c). All 8 H -3 protons that demonstrate a three -bond gHMBC correlation with the Si -3 atoms of 3c (H-3a and H-3a«) , are represented by a single mul tiplet. The same 8 H -3 protons of the meta -species ( 3a and 3b), are represented by two multiple ts ( H-3a, 4H and H -3e, 4H ), due to the asymmetry inherent to the meta -structure. Analysis of the 1H-1H gCOSY spectrum reve als that proton signal at !H 6.53 shows correlations with !H 7.47 (H-3b and H-3b«; isochronous for cis and trans , m , 8H). However, also contrary to the 1H NMR spectra of the meta-species, the proton multiplet assigned to H-3b and H-3b« (!H 7.47) overlaps the proton multiplet assigned to H -1acR and H -1acR« (!H 7.50), which was used for quantification of the isomeric ratio of the previous two compounds. Similar to compounds 3a and 3b, the proton multiplets assigned to H-1acL and H -1acL« (!H 7.19) and H-1act and H -1act« ( !H 7.33) exhibit too much overlap for accurate quantification , and to a much larger extent than the overlap 105 of proton multiplets assigned to H-3b and H-3b« (!H 7.47) and H-1acR and H -1acR« (!H 7.50) (Figure 4-3). Thus, the proton resonances assigned to H-2a and H -2a! (!H 7.55) and H-1acR and H-1acR« (!H 7.50), combined with H-3b and H-3b« isomers is represented by the proton signals at !H 7.50 and 7 .47 integrating to 12 H and the proton signal at !H 7.55 integrating to 16 H, a ratio of 3:4. As the ratio of these signals increase, the ratio of isomers in the sample becomes majority cis 3c (Figure 4-3e-g). A ratio of 1:2 would signify an isolated trans 3c. The remaining protons were identified by 1H-1H gCOSY cr oss -peaks and coupling patterns ( Table 4-2). Table 4-2.29Si and 1H resonances of cis/trans 3c. Atom 1H (mult.) 29Si (mult.) 1 (Si) --- -79.1(2 Si) *, -79.3(4 Si) €, -79.5(2 Si) * 1-(C6H5) --- --- -a,a " (CH) 7.19(m, 4H) *, 7.33(m, 8H) €, 7.50(m, 4H) * --- -b,b" (CH) 7.32(m, 4H) *, 7.18(m, 8H) €, 7.20(m, 4H) * --- -c (CH) 7.39(m, 4H) ¤, 7.38(m, 4H) ¤, 7.25(m, 4H) ¤, 7.11(m, 4H) ¤ --- 2 (Si) --- -78.2 (8 Si) # 2-(C6H5) --- --- -a,a " (CH) 7.55(m, 16H) # --- -b,b" (CH) 7.23(m, 16H) # --- -c (CH) 7.39(m, 4H) ¤, 7.38(m, 4H) ¤, 7.25(m, 4H) ¤, 7.10(m, 4H) ¤ --- 3 (Si) --- -29.7 (4 Si) # 3-(CH3) 0.51 (s, 12H) ¤ --- 3-(C6H6N) --- --- -a,a " (CH) 6.53 (m, 8H) # --- -b,b" (CH) 7.47 (m, 8H) # --- -c (NH 2) 3.69 (s, 8H) # --- € = trans , * = cis, ¤ = cannot distinguish between cis/trans , # = isochronous cis/trans 106 With the established 1H NMR chemical shift assignments , integration was possible and samples containing cis and trans isomers could be quantified (Table 4-3). All quantitative 1H were run with a sufficient recycle delay to allow for complete relaxation as determined by T 1 analysis. 29Si NMR experiments were run with an empirically determined recycle delay of 12 seconds even though the T 1Õs ranged from 50 to 55 seconds. i Determining the percentage of cis isomer present in samples of m aterial 3c was complicated by signal overlap around !H 7.50-7.47.ii For this system it was necessary to subtract the integrated value of the protons at !H 6.60 (H-3b and H-3b«of cis-3c) from that of !H 7.50-7.47 in order to determine the percentage of cis isomer. Table 4-3. Integrated values of 1H NMR spectra from various mixtures of 3. compound !1 !2 ratio % cis 3a 0.05 16 0.0031 < 1 4.08 16 0.26 51 7.91 16 0.49 99 3b 4.24 16 0.27 53 3c 12.35 16 0.77 50 10.04 16 0.63 27 15.77 16 0.99 85 Integrals are taken at the following ! and can be seen in Figure 3: 3a: !1 = 7.48 ppm, !2 = 7.54 ppm; 3b: !1 = 7.48 ppm, !1 = 7.57; ppm 3c: !1 = 7.50 + 7.47 ppm, !2 = 7.55 ppm For comparative purposes, the percentage of the cis isomer in various samples of 3a was also determined using 29Si NMR spectra and compared with their 1H NMR counterparts (Table 4-4). Analysis times for 29Si NMR spectra varied from 10 minutes to 12 hours, while all 1H NMR spectra were acquired in 13 minutes. Increasing the time of acquisition for the proton experiment 107 did not lead to an appreciable change in calculated ratios. Contrariwise, as the length of a nalysis time for 29Si NMR spectra was extended, the percentage of cis isomer determin ed approached the percentages determined using the 1H NMR spectra. No im provement was seen when extending 29Si experiments past 4 hours . The limiting factor was the poorer signal -to-noise (S/N) in the 29Si NMR spectra. This not only made it more difficult to determine the limits of integration on 29Si NMR spectra as compared to that of the 1H NMR spectra but also increased the uncertainty in the former (Figure 4-2). Eve n after 12 hours, the S/N of the 29Si NMR spectrum was not comparable to its 1H NMR spectrum (Table 4-4), which was run for 13 minutes . For example, at 12 hours, the S/N of the smallest signal integrated was 19:1 and the S/N Table 4-4. Integrated values of 1H NMR spectra vs. 29Si NMR spectra from 3a mixtures . Exp # nucleus time ratio % cis S/N range (low -high) 1 29Si 10 min. 0.43 43 6-15 1H 13 min. 0.26 51 57-183 2 29Si 30 min. 0.43 43 11-28 1H 13 min. 0.25 49 52-201 3 29Si 2 hr. 0.48 48 12-23+ 1H 13 min. 0.25 51 33-90 4 29Si 4 hr. 0.97 97 2-34* 1H 13 min. 0.49 99 91-149 5 29Si 12 hr. 0.37 37 19-64 1H 13 min. 0.2 39 49-239 29Si: !1 = -79.4 ppm, !2 = -79.6 ppm, & !3 = - 79.8 ppm + effect of concentration *lower peak is low quantity trans S/N = range of signal to noise of the largest integrated signal was 64:1 (exp 5 -29Si). The lowest and highest S/N for the 1H NMR spectrum of the same material was 49:1 and 239:1, respectively (exp 5 -1H). For mixtures 108 containing predominately one isomer, it becomes increasingly more difficult to accurately determine isomeric purity . As an example, a fter 4 hours , the 29Si NMR sp ectrum of a sample that is majority cis isomer had a S/N for the trans isomer signal of 2:1 while the S/N of the cis isomer signal was 34:1 (exp 4 -29Si). The signal representing the trans isomer could potentially be mistaken for , or hidden under , noise had the analysis time been shortened, the purity of the cis isomer increased , or the concentration of the sample decreased . All 29Si samples had to be heavily concentrated, approximately 50 mg in 0.6 mL of solution, whereas the 1H NMR spectra could be obtained with a concentration of < 5 mg in 0.6 mL of solution. Overall, more accurate data was acquired with 1H NMR using less material and shorter analysis time . !"!#$%&'() *+&,#-./0-12 #%.'4-%2'%35 #In this study, 1H Ð 29Si gHMBC and 1H Ð 1H gCOSY were used to assign the chemical shifts of the protons for the cis and trans isomers of 3a-c as well as to verify the chemical shifts of 29Si- NMR. Once the proton chemical shifts of these compounds were identified , a routine access methodology to determine the cis and trans isomeric ratio was developed . The method described herein offer s a more accurate quantification of the isomeric ratio in a shorter analysis time using standard NMR equipment with less material required per sample when compared to utilizing 29Si NMR spectroscopy . With the ability to identify the isomeric ratios, attempts to isolate the individual isomers can be made, and the influence of stereo -configurations on the physical and chemical properties of these hybrid nanostructuered chemicals can be studied and utilized . 109 !!! "#$%& ! 110 !"#$% & (i) A comparison of 29Si NMR data collected with a 300 second delay, which would be appropriate for quantitative work, and with a 12 second delay yielded integration ratios within experimental error. E xperiments run with the 12 second delay gave improved Signal -to-Noise per unit t ime and were, thus, preferred. For additional information on the T1 analysis, please see appendix B . (ii ) A europium shift reagent also provides a solution to this problem. Fo r more inf ormation, please see appendix B . &&&&& 111 !!!!!!!!!!!"#$#"#%&#' ! 112 !"#"!"$%"& ' (1) Hurd, R.E.; Journal of Magnetic Resonance 1990, 87, 422. (2) Seurer, B.; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (3) Stock, A. Berichte Der Deutschen Chemischen Gesellschaft 1916, 49, 108. 113 !"#$%&'( !!!" #$%&#'(%)*&')+,"+-".+/01)/)'2"1($&3)+%"+- "!"# "&,4"$%&'#").+5(%. "),",&,+ 6.'%0#'0%(4 "4+01/("4(#7(%".)/.(.80)+9&,(. " Keywords double decker silsesquioxane, fractional crystallization, solubility , solid -liquid equilibria, NRTL model, isomer separ ation, cis/trans isomers * This chapter has been submitted for publication under the following citation: Schoen, B. W.; Lira, C. T.; Lee, A. Fluid Phase Equilibria 2013, submitted July 8, 2013 . 114 !" #$%&'&()*'+,&)+-.#-/#)%*#0-123+1+)4#3*%&5+-'#-/# !"##&.6#$%&'# #+0-7*'0#+.#.&.- 80)'2()2'*6# 6-231*#6*(9*'#0+10*:2+-;&.*0 #!"<#=.)'-62()+-. ##DDSQ with various reactive chemical moieties have been prepared and their cis and trans isomers partially isolated. 1,2 With the ability described in the previous chapter to identify cis and trans isomers using 1H NMR , a ratio of cis and trans isomers of 3a-c can now be accurately quantified .3 This allows for verification of purity and the development of a model representing the quantitative measurements and parameters need ed for fractional crystallization of the se isomer s. Furthermore, it provides an opportunity to understand how these configurations influence the structure -property relationship of these DDSQs . Fractional crystallization provides a platform for larger quantities of material to be separated into fewer frac tions as compared to other methods such as chromatography. 4-6 Furthermore, fractional crystallization provides a much lower energy demand as opposed to an energy -intensive thermal separation method such as distilla tion. Hence , it is accepted as a n appropriate economic approach for an industrial scale .7 A fractional crystallization method involving the v ariation in solvent polarity was used to separate the cis and trans isomers of compounds 3a-c and is described in this chapter . For the precipitates, molar r atios of the individual is omers were obtained using 1H NMR data .3 Experimental solubility results were modeled using the NRTL activity coefficient method . Activity coefficients, thermodynamic properties , and structural characteristics all contributed to the solubility model for the solution of mixed isomers. In this chapter , solubility, separation, and chemical properties were studied for isomers of compounds 3a-c. 115 !"#$%&'()*&+,$&-.$%('/0., $!"#"$"%&'()*+, -%.+/%0*.1*+,- %Triethylamines were obtained from JT Baker¨ Chemicals (Pleasant Prairie, WI) when used for column chromatography. Dichloromethane was obtained from Macrom TM Chemicals Products (Nashville, TN). SiliaFlash ¨ P60 Silica gel was purchased from S ilicycle Ultra Pure Silica Gels (Quebec City, Quebec, Canada). Silica gel 60 F 254, (0.2 mm thick, Sigma Aldrich) pre-coated plastic plates were used for thin layer chromatography (TLC). !"#"#"%2-'3*4%- *5.4.,6'+ %A fractional crystallization method was used to separate the cis and trans isomers. Each structure , 3a-c, was studied individually. Solvent polarity was varied by changing the ratio of solvent (THF) to anti -solvent (hexanes) using the procedure outlined in Figure 5-1. Purit y of the individual isomers was confirmed using 1H-NMR, and DSC data. A typical example of the separation procedure for compounds 3a-c follows: s olid DDSQ for one of the variants 3a-c (varying between 0.1 Ð 1.1 g, 0.077 Ð 0.81 mmole) was placed in a round bottom flask. The ratio of isomers in the initial sample varied. Minimal THF was added to the flask, at 25 ¼C, until the solid completely dis solved. Hexanes were added drop wise until a white suspension persisted. The precipitate (designated ppt1) was isol ated by filtration and dried under a stream of nitrogen. The first precipitate, ppt1, was usually predominantly the trans isomer, since it is less soluble. Depending on the initial cis:trans ratio and the solvent ratio, some cis isomer can be present . Within this filtrate, the trans was at the solubility limit, however the cis was not always at the solubility limit. The filtrate solvent mix of THF and hexanes was removed under vacuum, leaving a solid cake (designated ppt2). For collection of this precipitat e, hexanes were added to 116 create a pourable suspension. The solids (ppt2) were collected by filtration and dried under a stream of nitrogen. In general, the average recovery of total material in all experiments was greater than 90 % for all three compounds. By a material balance, the quantities of DDSQ and solvent in the supernatant of the first precipitation were the compositions used in solubility modeling. Cases where the cis isomer was below the solubility limit are noted below. Figure 5-1. Flow chart describing the fractional crystallization/isomer separation procedure of trans and cis isomers of compounds 3a, 3b, and 3c. Start Step 2: Add THF until solid is completely dissolved !"#$%&'% !"#$%#&&'(#$$()*+#,*-(%,./&( !"#$%&'(%)#'#$*$%&*''%*(+ Step 4: Record the volume of each solvent added Step 5: Remove ppt1 by filtration and collect the solution, ppt2 is still dissolved in the solution Step 6: Remove the solvent by vacuum Step 1: Place solid A-C in a flask End ppt1 solution containing ppt2 Step 7: Add hexanes to create a suspension Step 8: Isolate ppt2 by filtration 117 !"#"$%&'()*+,)-(+.'/0%12(/3/0+,/)4 % Thermal properties for pure isomers of 3a-c were important i n modeling the saturation curves . In order to obtain accurate values, the highest purity feasible of each isomer was desired. The trans isomer solid can be isolated easily by selectively precipitating trans while maintaining a solution undersaturated in cis. However , the cis isomer is more difficult to isolate at high purity by precipitation . Hence , chromatography columns were utilized to further purify the cis isomer s from selected ppt2 solids (mixture with high cis fraction) . Flash silica gel chromatography was performed using siliaflash ¨ P60 silica gel with a particle size of 0.043 -0.063 mm (230 -400 mesh) at room temperature. The column had a length of 203 mm and a diameter of 31.8 mm. The column was packed acc ording to methods described by Still et. a l.8 To prevent loss of the stationary phase through the bottom, the column was plugged with cotton. Silica gel was packed approximately th ree quarters of the total length of the column. A thin layer of sand (approximately 6.35 mm) was positioned above and below the silica gel in order to provide an even base for the stationary phase, and prevent concentration and s treaking of the bands as th ey came off the column and were collected. Solid DDSQ (0.2 Ð 1.0 g) was dissolved in minimal dichloromethane and loaded on the top of the column bed. The mobile phase was 1% triethylamine in dichloromethane. Silica gel 60 F 254, (0.2 mm thick) pre -coated pl astic plates were used for TLC . A small spot of each column fraction was applied to a TLC plate, approximately 1.5 cm from the bottom edge. A small amount of dichloromethane was placed in a glass beaker to a depth of less than 1 cm. The TLC plates were pla ce in the beaker so that the spots do not touch the dichloromethane , and the beaker was closed with a lid . Dichloromethane moved up each plate by capillary action, carrying the sample. The TLC plates were removed 118 before the dichloromethane reached the top of the plates , and spots were visualized by viewing them under short -wave range ultraviolet light. !"#"$%&'(%)*+,-./),/*0% %Compounds 3a-c were measured at 25 ¼C on a Varian UNITY -Inova 600 spectrometer equipped with a 5 mm Pulsed -Field-Gradient (PFG) switc hable broadband probe and operating at 599.80 MHz ( 1H). 1H NMR data were acquired using a recycle delay of at least 20 s and 32 scans to ensure accurate integration. The 1H-chemical shifts were referenced to that of residual protonated solvent in CDCl 3 (7.24 ppm). !"#"!%1233+.+4-256%),544247%, 56/.28+-.0 % Melting behavior was studied using a TA Instrument Q2000 equipped with a mechanical cooling system under a nitrogen atmosphere. Samples were placed in a Q -zero TM aluminum pan and sealed with a lid to evalua te the cis and trans isomers. Samples were first equilibrated at 40¡C for 1 min and subsequently heated to 350 ¡C with a constant heating rate of 10 ¡C/min. !"#"9%'/:+6247 % Solubility was modeled by the Schrıder -van Laar e quation (Equation 5 -1):9,10 !"!!!!!!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where x is the mole fraction of isomer ( i) that remain s in saturated solution at a given temperature ( T = room temperature ), based on its melting transition ( Tm), and heat of fusion (!Hm), and R is the ideal gas constant. The activity coefficient ( ") quantifies deviations from an 119 ideal solution. An ideal solution is a solu tion formed with no accompanying energy or volume change on mixing and no excess entropy ; the intermolecular attractive /repulsive forces (or intermolecular interactions) between the various pair types are all similar , and ! = 1. For a given temperature, the solution of E quation 5 -1 (RHS) is a constant and depends only on the Tm and "Hm of the pure substance. The left hand side (LHS) of E quation 5 -1 is associated with the mole fraction of the substance soluble in the given solvent system and its activity c oefficient. For ! = 1.00, or ideal solubility, the solubility of the substance in the solvent depends only on the experimental temperature . When ! < 1.00, the substance becomes mo re soluble in that solvent than ideal solubility (a favor able interaction) and when ! > 1.00, the substance is less soluble in that solvent than the ideal solubility (an unfav orable interaction) .11 In order to develop a comprehensive model for the separation procedure, the solubility/fractional crystallization data of the quaternary system, cis + trans + THF + h exanes (c + t + T + H), were included in the parameter fitting. It is recognized that the structure s of compounds 3a-c are so large in comp arison to the solvent molecules. Therefore, it is expected that this solution is non -ideal. The activity coefficient model used to fit the experim ental data was a simplified version of the non -random two liquid model (NRTL, Equation 5-2):12 !"!!!!!!!!"!!"!!!!"!!!!!!"!!!!"!!!!"!!!!!"!!"!!!!!"!!!!!!!!!!!!!!! !!"!!"# !!!"!!"!!!!!!!!!"!!!"!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where ! is the activity coefficient of solute ( i), x represents th e mole fraction of each of the remaining constituents in this quaternary system, "ij is an intermediate parameter representing the 120 binary interaction parameter s aij and bij, and !ij represents the non -randomness parameter. For the THF + hexane binary pair , a full set of parameters are available from Aspen Plus ¨ ver 7.3.13 For the limited data collected in this work at a single temperature, the number of parameters will be reduced by equating the ij interactions for some pairs and by setting some parameters to zero. For each of the three DDSQ structures, since cis and trans isomer s posses s the same functional groups, the binary energy interaction parameter s for cis to trans are assumed to be the same as cis to cis, and trans to trans , and thus "ct = "tc = 0. For the same reason, the cis to hexane s interaction is assumed to be the same as the trans to hexane s interaction, and the cis to THF interaction is assumed to be the same as trans to THF. These assumptions resulted in the simplifications, acis, H = atrans ,H and acis ,T = atrans ,T. These parameters will thus be abbreviated aDDSQ, H and aDDSQ, T without distinction of the cis and trans isomers. It was further assumed that the non randomness param eter, !, for DDSQ to THF interactions and DDSQ to hexanes are equal to zero. These simplifications reduce the binary NRTL equation to the two suffix (one parameter) Margules equation (E quation 5 -3):11 !"!!!!!!!!"!!!"!!!!"!!!!!!!!!"!!!!"!!!!!!!!!"!!!!!!!!!! This equation is shown as a binary in order to help the reader understand how the binary contribution to the multi -component model simplifies with these assumptions. !"#$%&'()*' $+,-$./'0(''/1, $$!"#"$%&'()*)+,-. %-/%0,1%).2%+*).1% #)30% Sections 3. 3.1.1 Ð 3.3.1.3 describe the recoveries and purities of the samples. Within each section, the discussion starts with conditions providing the highest purity of the cis and trans 121 Table 5-1.i Isomers were obtained from fractional c rystallizat ion/solubility experiments; ppt1 (step 5, Figure 3 -1), and ppt 2 (step 8 , Figure 3 -1). Their purity was determined by 1H NMR spectroscopy , experimental data was determined from m aterial recovered in step 8 (ppt2) of Figure 3 -1 for compounds (a) 3a, (b) 3b, and (c) 3c. Compound Exp. SMa ppt1 (% trans ) ppt2 (% cis) xH:xTc xcis xtrans 3a 1 1:1 84 90 0.137 1.35E-02 1.50E-03 2 1:1 90 95 0.616 3.47E-03 1.96E-04 3 1:1 58 86 1.361 5.32E-04 8.49E-05 4 1:1 59 92 1.155 7.75E-04 6.45E-05 5 1:1 96 99 0.899 4.29E-03 4.33E-05 6b 1:1 85 77 0.286 6.03E-03b 1.84E-03 7b 1:1 93 83 0.362 3.97E-03b 8.07E-04 8b 4:6 95 75 0.342 3.06E-03b 1.04E-03 9b 2:8 97 31 0.123 1.45E-03b 3.27E-03 10b 2:8 96 60 0.347 1.54E-03b 1.02E-03 11b 1:9 99 16 0.240 2.52E-04b 1.37E-03 12b 2:8 91 93 0.521 3.35E-03b 2.35E-04 13 8:2 35 86 0.205 2.60E-02 4.29E-03 14 9:1 20 91 0.539 6.98E-03 6.90E-04 15 9:1 12 89 0.411 1.61E-02 1.92E-03 16 9:1 15 85 0.240 2.55E-02 4.50E-03 17 9:1 7 85 0.308 1.46E-02 2.59E-03 18 8:2 33 84 0.220 2.64E-02 4.92E-03 3b 1 7:3 36 84 0.88 8.22E-04 1.61E-04 2 8:2 31 86 0.994 8.68E-04 1.37E-04 3 8:2 15 91 0.616 1.79E-03 1.74E-04 4 6:4 29 73 0.37 2.51E-03 9.46E-04 5 8:2 20 85 0.264 5.45E-03 9.90E-04 3c 1 6:4 71 99 0.880 5.91E-03 8.34E-05 2 6:4 68 95 0.513 7.85E-03 4.46E-04 3 6:4 78 96 0.616 5.74E-03 2.68E-04 4b 6:4 82 87 1.174 7.45E-04b 1.09E-04 5b 6:4 88 86 1.115 8.21E-04b 1.34E-04 6b 3:7 92 71 0.689 2.59E-04b 1.03E-04 7b 3:7 92 63 0.603 2.75E-04b 1.59E-04 8b 3:7 93 39 0.123 4.86E-04b 7.47E-04 9b 9:1 8 87 1.400 1.52E-03b 2.27E-04 10 9:1 3 88 1.155 4.52E-03 5.99E-04 11 9:1 7 92 1.369 1.42E-03 1.30E-04 12 9:1 32 97 0.822 4.63E-03 1.61E-04 a starting material in a ratio of cis:trans , b undersaturated cis isomer, c hexanes to tetrahydrofuran molar ratio 122 fractions. Table 5-1 provides a summary of the composition of each precipitate and solubility of each isomer that will be discussed in sections 3.3.1.1 -3.3.1.3. The trans isomer was often predominant in ppt1 while the cis isomer was predominant in ppt2 samples with exceptions discusse d below. Solubilities were determined based on composition data from the solid obtained in Step 8 of Figure 5-1. The mole fractions were determined by using t he material balance on the total mass of each solvent and the mass of DD SQ with 1H NMR analysis. Samples that were fully saturated with both isomers were evaluated for the quantity of each isomer that remains in solutio n (ppt2) as the solvent is varied from polar (THF -rich) to nonpolar (hexane -rich) ( Figure s 4-6). As the solut ion became hexane -rich, less of each isomer remained in solution. The cis isomer predominated throughout the range of solvents used. !"#"$"$"%&'()'*+,% #-%% From the 18 experiments summarized in Table 5-1, the highest purity of trans (ppt1) was obtained in experiment 11, when 0.447 g 3a (3.34e-4 moles) were placed in a n anti -solvent to solvent molar ratio of 5:10 (9:13 v/v) to yield 0.369 g (2.76e -4 moles, > 99 % trans ). The highest purity of cis was obtained in ppt2 of experiment 5 , when 0.710 g 3a (5.31e-4 moles) were placed in a 9:10 mixture of anti -solvent to solvent ratio by mole (7:4, v/v) to yi eld 0.327 g (2.45e -4 moles, > 99 % cis). Generally, the average recovery of total material in all experiment s was 92 ± 9 %. 3a -Experi ments that comprised a starting material (SM) of approximately 1:1 cis/trans ratio and a supernatant fully saturated with both isomers displayed a high trans purity in 3a -ppt1 and a high cis purity in 3a-ppt2, according to 1H NMR results (experiments 1-5). Those comprised of a supernatant fully saturated with trans isomer s and an undersaturated cis isomer 123 displayed high trans purity in 3a-ppt1 and a lower cis purity in 3a-ppt2 ( experiments 6-8). Additional separations were then carried out using the product s from the first 8 experiments as the SM (9 -21). Experiments using the first ppt1 (with a higher trans ratio) as a SM mostly displayed a supernatant fully saturated for the conditions tested with trans isomer s, cis isomer s were undersaturated, and displaye d a high trans purity in ppt1 and a much lower cis purity in ppt2 ( experiments 9-12). An attempt was then made to obtain the highest purity cis isomer with experiments comprised of a SM with a higher cis ratio ( experiments 13-18). These experiments Figu re 5-2. Experimental, circle (!) = cis, square (") = trans , and modeled, solid line (!) = cis, dashed line (---) = trans , solubility limits in a h exanes to THF (X H : X T) solvent solution for isomers of compou nd 3a. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1E-5 1E-4 1E-3 0.01 0.1 xH : x Tcis 3atrans 3axisomer 124 contained a supernatant fully saturated with both isomers, and displayed a low trans purity in 3a-ppt1 and a high cis purity in 3a-ppt2. The ratio of isomers in the SM did not have an effect on the solubility of the isomers. For all experiments tha t were comprised of a supernatant fully saturated with cis isomer, the average percentage of cis isomer in 3a-ppt2 was 90 ± 4.6 % (Figu re 5-2). !"#"$"%"&'()*(+,-& #.&& Experiment al purities are summarized in Table 5-1. The highest purity of the trans (3b-ppt1) was obtained in experiment 1, when 0. 328 g (2.22e-4 moles) were placed in a 6:7 mixture of anti -solvent and solvent by mole ( 3:2 v/v) to yield 0.257 g (1.75e -4 moles, 36 % trans ). The highest purity of the cis (3b-ppt2) was obtained in experiment 3 , when 0.270 g 3b (1.82 e-4 moles) were placed i n an anti -solvent to solvent molar ratio of 1:1 (1:1, v/v) to yield 0.058 g (3.91e-5 moles , 91 % cis). Generally, t he average recovery of total material in all expe riments was 94 ± 3 %. The SM of all 3b experiments was majority cis isomer due to a more complicated synthetic procedure , whic h provided not only a lower reaction yield, but also additional by-products and required further purification . One of the by -products had a similar solubility to the trans isomer , and in order to purify 3b, some trans isomer was lost with by -product. However, since fraction ppt 2 (majority cis isomer) is the principal solid , and the experimental data from compound 3a demonstrate d that the ratio of isomers in the SM did not have any effect on the solubility of the isomers, having all SM for experiments 3b comprised of majority cis isomer helped ensure that the cis isomer was not undersaturated. It was determined that t hese experim ents contained a supernatant fully saturated with both isomers, and displayed a low trans purity in ppt1 and a high cis purity in ppt2. The average percentage of cis isomer in ppt2 was 84 ± 7.0 % (Figure 5-3). 125 Figure 5-3. Experimental, circle (!) = cis, square (") = trans , and modeled, solid line (!) = cis, dashed line (---) = trans , solubility limits in a h exanes to THF (X H : X T) solvent solution for isomers of compound 3b. !"#"$"#"%&'()'*+,% #-%% Experimental are summarized in Table 5-1. The highest purity of the trans (ppt1) was obtained in Experiment 8 , when 0.322 g 3c (3.34e-4 moles) were placed in a 7.5:5.5 mixture of anti -solvent and solvent by mole ( 9:4 v/v) to yield 0.0273 g (2.04e -5 moles, > 99 % trans ). The highest purity cis fraction (ppt 2) was obtained in Experiment 1, when 0.331 g 3c (2.48e-4 moles) were placed in a 7.5:8.5 mixture of anti -solvent and solvent by mole (1:0.7 v/v) to yield 0.131 g (9.79e-5 moles, > 99 % cis). Generally, the average recovery of total ma terial in all experiments was 96 ± 5 %. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1E-5 1E-4 1E-3 0.01 0.1 cis 3btrans 3bxisomer xH : x T 126 As with 3a, experiments using a starting material (SM) of approximately 6:4 cis/trans ratio and a supernatant fully saturated with both isome rs displayed a moderately high trans purity in ppt1 and a high cis purity in ppt2, according to 1H NMR results ( experiments 1-3). Additional separations were then carried out using the products from the first 5 experiments as the SM (experiments 6-12). Expe riments that comprised a SM with a higher trans content mostly displayed a supernatant fully saturated for the conditions tested with trans isomer s, cis isomer s were undersaturated, and displayed a high trans purity in ppt1 and a much lower cis purity in Figure 5-4. Experimental, circle (!) = cis, square (") = trans , and modeled, solid line (!) = cis, dashed line (---) = trans , solubility limits in a h exanes to THF (X H : X T) solvent solution for isomers of compound 3c. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1E-5 1E-4 1E-3 0.01 0.1 xH : x Ttrans 3ccis 3cxisomer 127 ppt2 ( experiments 6-8). An attempt was then made to obtain the highest purity cis isomer with experiments comprised of a SM with a higher cis ratio ( experiments 9-12). These experiments contained a supernatant fully saturated with both isomer s, and displayed a low trans purity in ppt1 and a high cis purity in ppt2. Similarly to the meta -species 3a, it was demonstrated that the ratio of isomers in the SM did not have an effect on the solubility of the isomers for the para - species. For all expe riments that were comprised of a supernatant fully saturated with cis isomer, the average percentage of cis isomer in ppt2 was 93 ± 4.6 %. !"#"$%&'()*+,)-(+.'/%01234,2 % Chromatography columns were used to obtain the highest purity of each isomer. ii There were up to 60 fractions (5 Ð 10 mL each) from one chromatography column. TLC plates and 1H NMR spectra were used to identify the pure fractions. The trans isomer eluted before the cis isomer for all materials. There was sufficient separation in the order of elution to achieve complete separation of the two isomers. !"#"#%56&%01234,2 %Pure component properties (Tm and !Hm) were determined from DSC heating traces for compounds 3a, 3b, and 3c. Results are summarized in Table 5-2 and were used to fit the Table 5-2. Melting temperature (Tm) and the h eat of fusion (!Hm) for compounds 3a-c as determined by differential scanning Calorimetry, from T = 40 ¡C Ð350 ¡C with a heating rate of 10 ¡C/min. Compound 3a-cis 3a-trans 3b-cis 3b-trans 3c-cis 3c-trans Tm (¡C) 289.5 312.0 265.9 270.0 272.6 309.3 !Hm (kJ/mol) 42 56 39 46 38 58 128 sharp endothermic peak was observed, which is representative of a melting transition. The onset of melting (Tm) is calculated by extrapolating the slope from the peak width at half height to the baseline and the total heat of f usion ( !Hm) is the total area of the melting endotherm . Figure 5-5. Example of a melting endotherm ( trans 3a) showing the melting temperature and heat of fusion as determined by differential scanning Calorimetry, from T = 40 ¡C Ð350 ¡C with a heating rate of 10 ¡C/min. !"#"$%&'()*+(%,-%.,/'*%01++123 %Binary parameters for THF + hexanes were taken from the VLE regression sets in Aspen Plus ¨ ver . 7.3 (aH,T = aT,H = 0, bH,T = -15.0959 K, bT,H = 233.6258 K and !T,H = 0.3).13 Only the binary aij for cis/trans + hexanes and cis/trans + THF (aDDSQ,H and aDDSQ,T ) were 280290300310320-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 pk width at 1/2 height Extrapolated onset T Area under the curve (Heat of fusion) Temperature ( ¡C)Heat Flow (W/g) 129 adjusted while xi values were fitted along the saturation curves. Since all data were coll ected at room temperature only , the temperature dependence is ignored; thus bDDSQ, H = bDDSQ, T = 0. For the interaction between DDSQ and each solvent !DDSQ,H = aDDSQ,H and !DDSQ,T = aDDSQ,T . Therefore, a negative value of aij will provide a negative value of !ij and " < 1.00. A positive value of aij will provide a positive value of !ij and " >1.00. As the magnitude of aij increases, the solution becomes less ideal. For large negative values of aij, the material is more soluble than in an ideal solution; for large positive values the solution is an anti -solvent, and the material is less soluble than in an ideal solution. The model results are compared to experiment al results in Figu re 5-2 through Figure 5-4 and compared between compounds 3a-c in Figure 5-6. The scatter seen in Figu re 5-2 through Figure 5-4 can be attributed to the experimental uncertainties due to the small sample sizes used. The relative uncertainty that provides the most error (according to differential error analysis) is from the solid composition obtained by 1H NMR ( Table 5-1). The experimental measurements in Table 5-1 lie within the computed confidence intervals of +/ - 5%. For example, the point in Figu re 5-2 with a solid xH : xT ratio of 0.899 has a 500% relative error (% RE), determined by the error in the composition of the trans isomer ( Table 5-3). The RE is very high for this data point because 99% of the 326 mg sample is the cis isomer. Therefore, t he mass of the trans isomer is only 3.3 mg, or 1 % of the sample, which is below is threshold of detection for 1H NMR. The +/ - 5% confidence interval, as discussed previously is applied to the entire sample mass, yielding a confidence interval of approxima tely +/ - 16 mg, which is 5 times larger than mass of the trans isomer. Clearly, as the 130 Figure 5-6. Modeled sol ubility limits in a h exanes to THF (X H : X T) solvent solution for the cis and trans isomers of compounds 3a, 3b, and 3c. The three upper curves represent cis and the lower three curves represent trans . mass of a component decreases, the relative uncertainty for that component increases significantly which explains the larger scatter in the experime ntal trans data points. Another feasible cause for error is experimental masses and volumes of all four components , which provide a % RE of < 10 for the trans isomer and < 1 for the cis isomer . Additional causes for error in the experimental values and the model may be attributed to: room temperature and filtration methods, which are more ambiguous to estimate. When the model was adjusted to 0.0 0.4 0.8 1.2 1E-4 1E-3 0.01 3a 3b 3cxisomer xH : x T 131 account for the % RE in x i, activity coefficients and binary interaction parameters were minimally affected. Table 5-3. % Relative error (% RE) determined by a +/ - 5% confidence interval in 1H NMR measurements. Compound xH:xTa % REcis % REtrans Compound xH:xTa % REcis % REtrans 3a 0.123 16 7 3b 0.264 5 83 0.137 6 50 0.370 7 18 0.205 6 35 0.616 5 56 0.220 6 32 0.880 5 59 0.240 6 33 0.994 6 37 0.240 32 6 3c 0.123 8 13 0.286 7 21 0.513 5 93 0.308 6 33 0.603 8 14 0.342 7 20 0.616 5 112 0.347 8 13 0.689 7 18 0.362 6 30 0.822 5 148 0.411 6 47 0.880 5 359 0.521 5 76 1.115 6 36 0.539 5 56 1.155 6 43 0.616 5 93 1.174 6 39 0.899 5 500 1.369 5 60 1.155 5 65 1.400 6 39 1.361 6 36 a hexanes to tetrahydrofuran molar ratio Activity coefficients as a function of the mole fraction of solvent, THF or hexanes, for all three DDSQ compounds investigated are presented in Figure 5-7. Adjustable parameters representing the interaction between DDSQ compounds to solvent ( aDDSQ, T and aDDSQ, H) are summarized in Table 5-4. For compounds 3a-c the solubility in 100 % THF (left side of Figure 5-6), is higher than the ideal value ( Table 5-5). From the model the ln( !solute -solvent ) < 0 and the ln( !solute -anti -solvent ) > 0, which confirms that the THF is a good solvent and hexan es are an 132 Figure 5-7. Activity coefficients of the solu te vs. the mole fraction of (a) anti -solvent (hexanes) and (b) so lvent (THF), when each compound is considered a binary. When the natural logarithm of the activity coefficient is larger than zero, the interaction is between DDSQ and hexanes. When the natural logarithm of the activity coefficient is below zero, the inter action is between DDSQ and THF. 0.0 0.2 0.4 0.6 0.8 1.0 01234 3a 3b3cln( !solute )xsolvent (a) 0.0 0.2 0.4 0.6 0.8 1.0 -6-4-20 3a 3b 3cln( !solute )xsolvent (b) 133 Table 5-4. Binary interaction coefficients for 3a, 3b, and 3c. Compound aDDSQ,H aDDSQ,T 3a 1.7904 -3.4085 3b 2.1676 -1.5328 3c 0.1474 -1.7338 anti -solvent for these three DDSQ compounds . The !solute is highly non -ideal for all conditions tested, since xsolven t is very close to 1.00 and the saturated solutions are close to infinite dilution for DDSQ . The relation between the activity coefficients and the binary interaction p arameter s are most obvious in Equation 5 -3 for a binary, recalling that "ij = aij, and !"!!!!!!!". Table 5-5. Result of Schrıder -van Laar equation (R HS) at room temperature and the corresponding s olubility limits based on ideal solution assumptions ( ! = 1.00) for compounds 3a, 3b, and 3c. Compound RHS xideal a 3a-cis -7.97 3.47E-04 3a-trans -11.08 1.54E-05 3b-cis -7.03 8.83E-04 3b-trans -8.37 2.32E-04 3c-cis -6.96 9.54E-04 3c-trans -11.42 1.10E-05 a xideal : ( != 1.00) The magnitude of the binary interaction parameter of DDSQ and THF ( aDDSQ, T) combined with the RHS of the Schrıder -van Laar equation (E quation 5 -1) can be used to explain the overall solubility of compounds 3a-c. For all three compounds investigated , the solubility of the cis species is always higher than the trans species in a solvent with a fixed ratio of THF to hexanes. Experimental s aturation curves for both the cis and trans species in a THF/hexanes mixture, as shown in Figu re 5-2 through Figure 5-4, appear to exhibit the same slope . This 134 observation validated the assumption that the interaction parameters between each isomer and solvent is the same , aDDSQ,T = aDDSQ,H . Howe ver, t he magnitude of the separation between the cis and trans saturation curves for each compound (Figure 5-6) varies and is dependent on the physical melting characteristics , which is given by the ratio of the ideal solubility for the two isomers ( Table 5-5). Isomers of c ompound 3c exhibited the largest magnitude of separation in their saturation curves ( Figure 5-6). Cis isomers of compound 3c were 33 times more soluble than th eir trans isomers, a partitioning threshold of 33:1. The separation of the saturation curves of the isomers of compound 3a was slightly closer; cis isomers were 22 times more soluble than trans isomers. Isomers of compound 3b exhibited the smallest magnitu de of separation in their saturation curves; cis isomers were only 3.5 times more soluble than trans isomers. Isomers of 3c exhibit the largest difference in Tm and !Hm, and the isomers of 3b exhibit the smal lest difference, which is consistent with this solubility data. 14,15 The cis isomer s, which are the more soluble species, of compounds 3b and 3c approach ed the same solubil ity in THF (left side of Figure 6) . This is because they have very similar values for the ir binary interaction parameter between DDSQ and THF , a3b,T = -1.5328 and a3c,T = -1.7338, and similar Tm and !Hm values (RHS ! -7). Cis 3a possesses a higher solubi lity in THF than the other two compounds because its value of a3a,T is most negative (- 3.4085) and the T m and !Hm values (RHS ! -8) are larger than the other compounds . The slopes of the saturation curves in Figure 6 are very different for cis 3b and 3c, which results from their binary interaction coefficients with the anti -solvent (hexanes), a3b,H = 2.1676 and a3c,H = 0.1474. The magnitude of aDDSQ,H provides information on how quickly the solubility will decrease as hexanes are 135 added to THF . Since, a3c, H is so close to zero, the interaction of cis 3c and hexanes is nearly ideal. Therefore, the solubility of cis 3c with the addition of hexanes will not decrease as rapidly as the other compounds and the lower solubility is largely due to dilution of the st ronger THF solvent . The slope of the saturation curve of cis 3a is similar to that of cis 3b because they have similar values for their interaction with hexanes, a3a,H = 1.7904. The t rans isomers, which are the less soluble species, of compounds 3a and 3c have similar values for T m and !Hm (RHS ! -11), =but since a3a,T (-3.4085) has a larger magnitude than a3c,T (-1.7338), trans 3a is more soluble in THF than trans 3c. Additionally, trans 3a and trans 3b have different Tm and !Hm values (RHS ! -11 and -8, respectively) and different values for aDDSQ,T (-3.4085 and -1.5328, respectively), but coincidentally have similar solubilities in THF. This further confirms that the Schrı der -van Laar e quation and binary interaction coefficients are both necessary when determining solubility. !"#$%&'()* +,'-$./01.23$&'$,3&0/.$3/41.15,&'3 $Cis and trans isomers of a series of DDSQ(X)(R) were successfully separated using a fractional crystallization method . The Schrıder -van Laar equat ion (Equation 1), and the NRTL model (Equ ation 2) were used to fit the experimental data. The model was simplified to reduce the number of adjustable parameters. The NRTL model parameters are consistent with the solubility trends, suggesting they have physical meaning. In this paper, the solubili ty of the cis and trans isomers were represented using the same binary interaction coefficients for DDSQ + solvent and DDSQ + anti -solvent for each isomer. The two isomers of each compound do not interact differently with a given solvent even though they h ave different solubilities. The 136 difference in solubility of the cis and trans isomers is due to the pure component properties (RHS), not the DDSQ + solvent and DDSQ + anti -solvent interactions. Between the different compounds, the differences in solubility are due to both the Schrı der -van Laar e quation and binary interaction coefficients. Overall it was determined that c hanging the R -group from a methyl to a bulkier cyclohexyl moiety, 3a to 3b, decreases the solubility of the compound and also decreases th e partitioning threshold between the cis and trans isomers. The difference in Tm and !Hm were consistent with the change in the isomeric partitioning threshold of these compounds . Additionally the solubility of the meta -structures, compounds 3a and 3b, dec rease at similar rates with increasing content of hexanes, which was verified by the interaction parameters with hexanes. Changing the X-group from meta - to para -, 3a to 3c, leads to an increase in the partitioning threshold , and a decrease in the rate of solubility. Overall, the models show that assumptions were valid over the range of experiments , and the determined saturation curves can be used for the separation of the cis and trans isomers of compounds 3a-c. 137 !!!!! "#$%& ! 138 !"#$% & (i) A total mass balance of each isomer was performed to verify solubility data and can be seen in appendix C. (ii ) The retardation factors (R F) are reported in appendix C. 139 !!!!!!!!"#$#"#%&#' ! 140 !"#"!"$%"& ' (1) Hoque, M. A.; Kakihana, Y.; Shinke, S.; Ka wakami, Y. Macromolecules 2009, 42, 3309. (2) Seurer, B.; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (3) Schoen, B. W.; Holmes, D.; Lee, A. Magnetic Resonance in Chemistry 2013, 51, 490. (4) Myasnikov, S. K.; Uteshinsky, A . D.; Kulov, N. N. Theoretical Foundations of Chemical Engineering 2009, 43, 227. (5) Tadie, M.; Bahadur, I.; Reddy, P.; Ngema, P. T.; Naidoo, P.; Deenadayalu, N.; Ramjugernath, D. Journal of Chemical Thermodynamics 2013, 57, 485. (6) Wang, T. -C.; Li, Y. -J.; Chen, Y. -P. Journal of Chemical and Engineering Data 2012, 57, 3519. (7) Tam Le, M.; Lorenz, H.; Seidel -Morgenstern, A. Chemical Engineering & Technology 2012, 35, 1003. (8) Still, W. C.; Kahn, M.; Mitra, A. Journal of Organic Chemistry 1978, 43, 2923. (9) Prausnitz, J. M.; Lichtenthaler, R. N.; Azevedo, E. G. d. Molecular thermodynamics of fluid -phase equilibria ; 3rd ed.; Prentice Hall PTR: Upper Saddle River, N.J., 1999. (10) Gmehling, J. G.; Anderson, T. F.; Prausnitz, J. M. Industrial & Enginee ring Chemistry Fundamentals 1978, 17, 269. (11) Elliott, J. R.; Lira, C. T. Introductory chemical engineering thermodynamics ; 2nd ed.; Prentice Hall: Upper Saddle River, NJ, 2012. (12) Renon, H.; Prausnit.Jm Industrial & Engineering Chemistry Process Des ign and Development 1969, 8, 413. (13) In Aspen Physical Property System. Physical Property Methods and Model Burlington: 2010. (16) Pinal, R. Organic & Biomolecular Chemistry 2004, 2, 2692. (17) Gavezzotti, A. Journal of the Chemical Society -Perkin Tra nsactions 2 1995, 1399. 141 !"#$%&'() ! "#$%&!'&#$()*+!,*+ !!"# !$-.!$%&'#!)%*/&+%!*,!0#+&&!$/)-*"#&-12!.*3'2&!.&45&+! %)2%&%63)*7$-&%! ! Keywords double decker silsesquioxane, solid -liquid equilibria, NRTL model, eutectic , isomorphous cis/trans isomers ! 142 !" #$%&'()'$%*+,-(.,-( !"#(%/0($%&'# (+&,1'-&(,.(2$-''(%1+/,3$'/45(0,6)5'(0'78'-( &+5&'&96+,:%/'& ( !";( 99.3 % ( Table D -I, appendix D ). All crystal structures obtained are shown in appendix D . !"#"$"/%01231)45% #*%%One distinct crystal structure for trans 3c was determined ( Figure D -2, appendix D ). It was observed as a needle -like structure from a layered mixture of methylene chloride and hexanes , and crystallize d in the monoclinic space group P2(1)/n with a packing density (!) of 1.39 Mg/m 3. For cis 3c, one distinct crystal structure was also determined (Fi gure D -2, appendix D). It was observed as a flat sheet from a layered mixture of benzene and hexanes , and crystalliz ed in the triclinic space group P-1 as a twinne d structure with ! = 1.366 Mg/m 3. 152 Figure 6-3. Second heating of cis 3a, which demonstrates a glass transition (T g), and a recrystallization (T r); from T = 45 ¡C Ð200 ¡C with a heating rate of 10 ¡C/min. !"#"$"$%&'()'*+,% #-%%Two distinct crystal structures for trans 3a (polymorph A and B) were determined ( Figure D -3, appendix D ). Polymorph A was observed as a solid block, whereas polymorph B was observed as a long needle -like structure from a layered mixture of benzene and hexanes and both crystallized in the triclinic space group P-1 with ! = 1.397 and 1.399 Mg/m 3, respectively. One distinct crystal structure for cis 3a was determined ( Figure D-3, appendix D ). It was also observed as a solid block from a layered mixture of benzene and hexanes and crystallized in the monoclinic space group C 2 / c with ! = 1.365 Mg/m 3. The a minophenyl groups of two adjacent cis 3a molecules were connected via hydrogen bonding within the crystal packing s tructure (Figure 6-4). There is no hydrogen bonding within the crystal packing structure of either trans polymorph . 50100150200-0.2 0.0 0.2 TrTgHeat Flow (W/g )cis 3aexo^ Temperature (oC) 153 Figure 6-4. Hydrogen bonding (dotted line) in the crystal lattice of cis 3a; red = O, blue = Si, black = C, light blue = N, and pink = H. Phenyl moieties and other HÕs not shown for simplicity. !"#"$"#%&'()'*+,% #-%Three distinct crystal structures for trans 3b (polymorph A, B, and C) were determined (Figure D -4, appendix D). Polymorph A , B and C were all observed as a solid block s from a layered mixture of toluene and hexanes (A) and benzene and hexanes (B and C) . Polymorph A crystallized in the mon oclinic space group P 2(1) / c, whil e b oth polymorphs B and C crystallize d in the triclinic space gr oup P-1. The ! for polymorphs A -C were 1.35, 1.349, and 1.33 Mg/m 3, respectively. Both triclinic polymorphs contain disorder in the unit cell, whereas polymorph A does not. Exhaustive attempts were made to crystallize cis 3b in a large variety of solvents , but were unsuccessful . Diffraction patterns did not result far passed 2 ". 154 !"#"$"%&'()*+(*),-&./001()/ &After attaining the crystal structures, point group symmetry was examined. All trans isomers exhibit C i symmetry and all cis isomers exhibit C 2 symmetry as determined by Platon ©.24 This gives all six isomers 2 irreducible representations. !"#"#&21-(345&617,839)&9:&634,)/&+3.;(),4.&03<(*)1. &Melting behavior was investigated for samples with systematically varying cis to trans ratios. For compounds 3c and 3a, as the content of tran s isomer in the sample increased from 100 % cis, the onset of melting generally decreased, and the melting endotherm became broader until the eutectic composit ion. After the eutectic composition, the onset of melting began to increase and the melting endotherm again became broader until the sample approached 100 % trans isomer. For compound 3b, th e onset of melting constantly in creased from xtrans = 0 to xtrans = 1 and the width of the melting endotherm remained narrow and constant. !"#"#"=&>,(,&,4,-/.3. &In order to describe the melting behavior of binary cis/trans mixtures, the DSC data was analyzed with solid -liquid equilibrium equations. For systems that are immiscible in the solid state and miscible in the liquid state, the melting behavior can be analyzed according to the Schrıder -van Laar equation (E quation 6 -1):14 !"!!!!!!!!!"!"!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where xi is the mole fraction s of isomer ( i) that is in the liquid phase, Tmi is the melting temperature of the pure isomer ( i) and !Hmi is the heat of fusion of the melting transition, T is 155 the altered melting temperature of the binary sample , and R is the universal gas constant. The activity coefficient ( !) quantifies deviations from the ideal behavior in the mixture for isomer (i).25 The activity coefficient model used to fit the experimental data was the non -random two liquid model (NRTL , Equation 6 -2):26 !"!!!!!!!!"!!!!!!!!"!!!!"!!!"!!"!!!!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!"!!"# !!!"!!"!!!!!!!!!"!!!"!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! ! where i and j are trans and cis isomers, respectively, ! is the activity coefficient , x is the mole fraction, "ij is an intermediate parameter representing the binary interaction parameters aij and bij, and #ij represents the non -randomness parameter. It was assumed that the non randomness parameter, #, for trans to cis interactions and cis to trans interactions are equivalent and equal to 0.3, which is within the suggested range (0.2 Ð 0.47).26 A system that is completely miscible in both solid and liquid states, a correction to the Schrıder -van Laar equation, must be made (E quation 6 -3): !"!!"!!"!!"!!"!!!!!"!"!!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!where xiS and !iS, and xiL and !iL are the mole fraction s and activity coefficient of isomer ( i) that are in the solid state and liquid state, respectively. 156 !"#"$%&'(')*+,-(+.,%./%(0'%1.2+3%4%2+56+3%70-1'%3+-8)-*1%./%9+1:()-,1%;+,-)<%*+=(6)'1 % Sections 4.3.5.1 -4.3.5.4 show the results generated from the solid -liquid equilibrium data analysis for the melting behavior of binary cis/trans mixtures. Sections 4.3.5.1 -4.3.5.2 show phase diagrams for the melting behavior of compounds 3c and 3a. Both we re of the incongruent -eutectic type. Section 4.3.5.3 shows the binary interaction parameters and activity coefficients that resulted from these analyses. Finally, Section 4.3.5.4 shows the isomorphous phase diagram of compound 3b. !"#"$">%?0-1'%3+-8)-*%./% ;+,-)<%9+1:()-,1%*+=(6)'1%/.)% 9.*7.6,3%#9 %The phase diagram of compound 3c is of an incongruent type that has a eutectic point at approximately T E = 256.5 ¡C and xtrans = 0.33 and a meta -stable eutectic point at approximately Tm-E = 247.0 ¡C and xtrans = 0.40 (Figure 6-5). When viewing the DSC trace of the eutectic composition, one sharp melting endotherm appears, which is similar to that of the pure isomers (Figure 6-6). Above the eutectic composition, where xtrans = 0.50, three melting endotherms are observed. When this mixture is heated, all of the cis isomer and part of the trans isomer in the mixture melt at the meta -stable eutectic temperature (T m-E), and cocrystal behavior is formed. Subsequently, the cocrystal will melt and the trans isomer will recrystallize at the incongruent melting temperature, or peritectic point (T P), until a liquid of composition P is reached. After which, the trans isomer will melt gradually at a temperature above T P. As seen in Figure 6-6, this is represented by a clear endothermic peak associated with the Tm-E, which is followed by an exothermic peak associated with cocrystal formation. Then, another endothermic peak is observed associated with cocrystal melting, which is followed by another exothermic peak that is 157 Figure 6-5. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3c, plotted as melting temperature (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental Tm measured with DSC apparatus , star ( ) = experimental Tp, starburst ( !) = experimental T c, triangle ( !) = experimental T E, circle ( ") = experimental T m-E, solid line ( Ñ) = calculated T m from NRTL model (equat ion 6 -2), and dash dot ( úøúø) = ideal Tm, from Schrıder -van Laar equation (e quation 6 -1) where ! = 1.00. 0.0 0.2 0.4 0.6 0.8 1.0 240250260270280290300310TCTPTm-E TELtrans cis ideal xtrans St + S cm-E St + Sco-crystal ETemperature (oC)Sc + Sco-crystal Sco-crystal + LSt + LSc + LP 158 Figure 6-6. Melting behavior of binary cis/trans mixtures for compound 3c; T = 225 ¡C Ð325 ¡C with a heating rate of 10 ¡C/min. All thermal traces have the same scale bar as 100 % trans unless otherwise noted. * Tf,m is the final melting of 50 % trans 3c. 23024025026027028029030031032033 % trans 3c (Eutectic point) 0.2 W/g 0 % trans 3c (100 % cis )exo^ 100 % trans 3c1 W/g 0.2 W/g Temperature (oC)90 % trans 3c Tf,m TCTPTm-E TE80 % trans 3c Tm-E *10 % trans 3c 15 % trans 3c 20 % trans 3c 50 % trans 3c 159 associated with recrystallization of the trans isomer at T P. Finally, a third, more broad endothermic deflection is observed, which is associated with the complete melting of the trans isomer. Below the eutectic point, where xtrans = 0.20, the melting endotherm appears to be very broad, and is representative of two overlapping melting endotherm s, T m-E and T m-0.20 . !"#"$"%&'()*+&,-)./)0& 12&3-4)/5&6-*78/)4*&0-98:/+*&21/& 610;1:4,&#) &The phase diagram of compound 3a is also of an incongruent type that has a eutectic point at approximately T E = 260 ¡C and xtrans = 0.25 and a meta -stable eutectic poin t at approximately Tm-E = 252 ¡C and xtrans = 0.33 (Figure 6-7). Similarly to compound A, when viewing the DSC trace of the eutectic composition, one sharp melting endotherm appears ( Figure 6-8). Above the eutectic composition, where xtrans = 0.35, multiple melting endotherms are observed. As seen in Figure 6-8, two broad endother mic transitions are associated with this composition. The first transition is clearly representative of the T E and the second represents the T m. The endotherm representing the T m-E is not distinctly observed because it is overlapped by the T E endotherm. However, the existence of a T m-E is evidenced by a descent in the baseline that occurs earlier than the T E endotherm. The peritectic temperature was found at approximately T P = 268 ¡C. This was evidenced by the DSC traces of samples that were xtrans = 0.60 and 0.80, since there was a clear endothermic transition before the T m. For other samples that were above the eutectic composition, the T P endotherm was overlapped by the T m-E and T m endotherms. Similar to 160 Figure 6-7. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3a, plotted as melting temperature (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental Tm measured with DSC apparatus , ex ( x) = experimental T p, triangle (!) = experimental T E, circle ( ") = experimental T m-E, solid line ( Ñ) = calculated T m from NRTL model (equation 6 -2), and dash dot ( úøúø) = ideal T m, from Schrıder -van Laar equation (equation 6 -1) where ! = 1.00. 0.0 0.2 0.4 0.6 0.8 1.0 250260270280290300310320trans cis TcTPTELTm-E ideal xtrans Temperature (!C)St + Scm-E ESc + Sco-crystal Sco-crystal + LSt + Sco-crystal PSt + LSc + L 161 Figure 6-8. Melting behavior of binary cis/trans mixtures for compound 3a; T = 225 ¡C Ð325 ¡C with a heating rate of 10 ¡C/min. All thermal traces have the same scale bar as 8 0 % trans unless otherwise noted. 2402602803003200.4 W/g TE0 % trans 3a10 % trans 3a20 % trans 3a25 % trans 3a35 % trans 3a60 % trans 3a 70 % trans 3a 80 % trans 3aTP100 % trans 3aTm-E 2 W/g exo^ Temperature (!C) 162 compound 3c, the melting endotherm becomes broader below the eutectic point until it approaches 0 % trans (100 % cis). !"#"$"#%&'()*('+,-%./0'12%/0+,1')+/30%4'1'5,+,16%37 %,*+,)+/) 8+24,%53-,(%7/++/09 % Binary interaction parameters between trans and cis isomers, atc and btc, and between cis and trans isomers, act and bct, were adjusted while fitting xi values along the solid -liquid curves on the phase diagrams and are summarized in Table 6-4. For compound 3c, the experimental Table 6-4. Binary interaction parameters of compounds 3c and 3a. Compound atc act btc bct 3c 19.68 -53.85 -11272.78 30229.48 3a -21.60 -42.80 10291.98 23535.62 values on the left hand side of the phase diagram ( Figure 6-5) are near ideal values, and the right hand side is slightly below ideal values. For compound 3a, the experimental values on both sides of the diagram ( Figure 6-7) are largely below ideality. The determined parameters provided theoretical models for binary melting behavior that fit experimental data and experimental activity coefficients well ( Figure 6-9). !"#"$"$%:;'6,%-/'91'5%37%./0'12%)/6<+1'06%5/=+*1,6%731% )3543*0-%#. % Unlike the other compounds, the phase diagram of compound 3b is of an isomorphous (homogeneous solid solution) type. There is complete miscibility of both the liquid and solid states. Applying the pure component properties of cis/trans 3b to the modified Schrıder -van Laar equation (e quation 6 -3), and assuming both the liquid and solid phases are ideal, provides calculated solidus and liquidus values ( Figure 6-10). Experimental liquidus values are very close 163 Figure 6-9. Activity coefficients ( !) of isomer ( i) vs. the mole fraction of ( xtrans ) for compounds (a) 3c and (b) 3a; left hand side i = cis isomers, and right hand side, i = tran s isomer; square ( !) = experimental xi, and solid line ( Ñ) = calculated xi. 0.0 0.2 0.4 0.6 0.8 1.0 -0.16 -0.12 -0.08 -0.04 0.00 0.04 0.08 0.12 Experimental Fitted trans Fitted cistrans cisxtrans ln(!i)(a) 0.0 0.2 0.4 0.6 0.8 1.0 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 Experimental Fitted trans Fitted cistrans cisxtrans ln(!i)(b) 164 to ideal calculations ( Table 6-5). Additionally, the two phase region is very small; solidus and liquidus values are within 0.2 ¡C. The DSC trace of a solid solution is represented by a single, narrow endothermic melting transition upon heating for all binary mixtures ( Figure 6-11). Figure 6-10. Solid -liquid equilibrium phase diagram for binary mixtures of cis/trans isomers for compound 3b, plotted as melting temperatur e (T m) as a function of trans mole fraction ( xtrans ); square ( !) = experimental Tm measured with DSC apparatus , solid line ( Ñ) = liquidus line, dashed line ( ---) = solidis line. Solidus and liquidus lines are determined from an ideal solution to the Schrıder -van Laar equation for an is omorphous solution (E quation 6 -3) with an ideal approximation, where ! = 1.00. 0.0 0.2 0.4 0.6 0.8 1.0 265266267268269270271cis LSxtrans trans Temperature (!C) 165 Figure 6-11. Melting behavior of binary cis/trans mixtures for compound 3b; T = 250 ¡C Ð300 ¡C with a heating rate of 10 ¡C/min . All thermal traces have the same scale bar as 100 % trans unless otherwise noted. 2502602702802903003 W/g 100 % trans 3b90 % trans 3b70 % trans 3b60 % trans 3b50 % trans 3b40 % trans 3b 35 % trans 3b 0 % trans 3b exo ^ Temperature ( !C)20 % trans 3b 166 Table 6-5. Experimental melting temperatures (T experimental ) determined f rom DSC apparatus vs. calculated melting temperatures (T calculated ) determined from an ideal solution to the Schrıder -van Laar equation for an is omorphous solution (e quation 6 -3) with an ideal approximation, where ! = 1.00of compound 3b. xtrans Texperimental (¡C) Tcalculated (¡C) 1.00 270.0 270.0 0.92 269.8 269.7 0.68 268.5 268.8 0.60 268.0 268.5 0.52 267.5 268.1 0.42 267.1 267.7 0.35 267.2 267.4 0.20 266.3 266.8 0.00 265.9 265.9 !"#$#%!&'()*)+),-.)'/ !012-3)'4 !Solidification behavior for samples of trans 3a-c with increasing quantities of cis 3a-c was also investigated . The onset of solidification of compound 3c generally decreased from xtrans = 1 to xtrans = 0 (Figure 6-12). For compound 3a, the onset of solidification decreased with increasing content of cis isomer until the eutectic composition where this transition was no longer visible. For samples of 3a that did not exhibit solidification upon cooling, a glass transition and recrystallization was visible during the second heating and both increased with increasing content of cis isomer. The onset of solidification for compound 3b was more complicated. It ge nerally increased as it moved towards a less pure composition, and was largest at xtrans = 0.50. 167 Figure 6-12. Solidification behavior of binary cis/trans mixtures for compound 3c; T = 270 ¡C Ð120 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as the trace immediately above it unless otherwise noted. 1201401601802002202402601 W/g 90 % trans 3c100 % trans 3c0.2 W/g 2 W/g 80 % trans 3c20 % trans 3c50 % trans 3c33 % trans 3cexo ^ 0 % trans 3cTemperature ( oC) 168 Figure 6-13. Solidification behavior of binary cis/trans mixtures f or compound 3a; T = 300 ¡C Ð170 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as the trace immediately above it unless otherwise noted. 1801952102252402552702853004 W/g 70 % trans 3a0.2 W/g 2 W/g 85 % trans 3a25 % trans 3a60 % trans 3aexo ^ 100 % trans 3a90 % trans 3aTemperature ( !C) 169 Figure 6-14. Solidification behavior of binary cis/trans mixtures for compound 3b; T = 230 ¡C Ð180 ¡C with a cooling rate of 5 ¡C/min. All cooling traces have the same scale bar as 100 % trans . 18019020021022090 % trans 3bTemperature ( oC)15 W/g 100 % trans 3b60 % trans 3b70 % trans 3b 50 % trans 3b40 % trans 3b0 % trans 3b20 % trans 3b35 % trans 3bexo ^ 170 !"#$%&'()''&*+$ The discussion is split into several sections that generally mirror the results sections. Section 4.4.1 discusses trends observed in the melting behavior of the pure isomers, and is followed by the solidification behavior of the pure isomers in section 4.4 .2. Structural information such as packing density and hydrogen bonding within the crystal lattice were also discussed in these two sections to provide rationale of experimental observations. Additional structural information and observed polymorphs were d iscussed further in section 4.4.3, and moleulcar symmetry in section 4.4.4. Section 4.4.5 discusses the phase diagrams that resulted from the data analysis for the melting behavior of the binary cis/trans isomer mixtures. Finally, section 4.4.6 discusses t he solidification behavior of the binary cis/trans isomer mixtures. !"#"$%&'()*+,%-'./0*12 % Melting behavior of the pure isomers is divided into several smaller discussion sections so that trends can be more easily observed and discussed (4.4.1.1 -4.4.1.3). Trans vs. cis isomers are generally compared in section 4.4.1.1. Then, the melting behavior of the analogous para - and meta-aminophenyl DDSQ (compounds 3c and 3a, R = Me) is compared in section 4.4.1.2. Finally, the melting behavior of the meta -structures where R = Me (compound 3a) and where R = Cy (compound 3b) is also compared. !"#"$"$%32/+4%04"%5*4%*416'24 %The trans isomer s of all three compounds exhibit ed a hi gher T m than their cis counterparts (Table 6-3). A general rule of thumb is that structures that exhibit a higher T m are less soluble , at a given temperature, and hav e higher lattice energies .1 Lattice energy is a quantitative measure of the internal cohesion in the crystal structure, and depends on the 171 intermolecular forces. 3 It can be determined experimentally by measuring the total !Hm, or the thermal energy necessary for the se molecules to overcome intermolecular forces and fully melt. It was previously known that all three trans isomers were less soluble 16 and, as seen in Table 6-3, the trans isomer s had a larger !Hm as a result of their higher packing densities. Trans 3c had a packing density of 1.39 Mg/m 3, and for cis 3c " = 1.366 Mg/m 3. Compound 3a had very similar packing densities for trans and cis, " = 1.399 and 1.365 Mg/m 3, respectively . For compound 3b, trans had a " of 1.35 Mg/m 3 and it was anticipated that this is higher than cis 3b. A higher density increases intermolecular forces between molecules, thus illustrating why all three trans isomers exhibit ed a higher !Hm and T m than their cis counterparts .1,8 !"#"$"%&'()(&*+"& ,-.( &Trans 3c and 3a exhibit ed a similar Tm (3c = 310.6 and 3a = 312.0 ¡C), suggesting that there are minimal differe nces in lattice energy, which was demonstrated by similar !Hm (3c = 55 and 3a = 56 kJ/mol) and " (3c = 1.39 and 3a = 1.399 Mg/m 3). Conversely, the Tm of cis 3c (275.4 ¡C) and B (291.8 ¡C) was approximately a 16 ¡C difference. However, both exhibit ed a similar !Hm (3c = 38 and 3a = 42 kJ/mol) and " (3c = 1.366 and 3a = 1.365 Mg/m 3), suggesting other factors contribute d to their difference in T m. It was apparent from the crystal structures that cis 3a exhibited intermolecular hydrogen bonding (Figure 6-4), wh ile cis 3c does not. Hydrogen bonding increases intermolecular interactions and requires more thermal energy to melt.27-29 To verify this, the entropy of fusion, !Sm = !Hm/Tm, which normalizes !Hm with 172 respect to Tm and is a measure of the disorder of the system, was determined. 25 For cis 3a, !Sm = 74.3 J/mol !K which is larger than cis 3c (!Sm = 69.2 J/mol !K). More disorder throughout the melt transition confirms that it takes more energy to melt the ordered crystal lattice. !"#"$"%&'()*+,&-."&/+/,0*(1+, &Changing the moiety from methyl to cyclohexyl had the largest effect on the melting behavior . Trans and cis 3b exhibit ed approximately a 42 and 26 ¡C decrease in T m, as compared to compound 3a, respectively . More bulky substituents can in fact decrease the T m in a set of closely related substances by disrupting the lattice packing. 30 Two approaches were used to verify this fact: (1) a direct " determination using single crystal XRD data, and (2) !Sm. With single crystal XRD studies, it was determined that the " for trans 3b was 1.35 Mg /m3, which wa s 0.04 Mg/m 3 less than trans 3b. Additionally, the !Sm of trans and cis 3b (trans = 84.7 J/mol !K and cis = 72.3 J/mol !K) were smaller than the !Sm of trans and cis 3a (trans = 95.7 J/mol !K and cis = 74.3 J/mol !K). Thus, the isomers of compound 3a originally had a more ordered and dense crystal lattice, which explained the determined trend in the melting data. !"#"2&30,45464/7)408&9(*7-40: &&Solidification behavior of the pure isomers is also divided into several smal ler discussion sections so that trends can be more easily observed and discussed (4.4.2.1 and 4.4.2.2). Trans isomers are compared in section 4.4.2.1, and cis isomers are compared in section 4.4.2.2. Except for compound cis-B, all pure isomers investigated exhibited a complete recrystallization upon 173 cooling from the liquid state and the values for the heat of crystallization are nearly equal to the heat of melting during heating. !"#"$"%&'()*+&,+-./(+ &During cooloing, trans 3c and 3a exhibit ed a sharp exothermic crystallization transition peak where the endset point was nearly equivalent to the supercool point. This suggested that the crystallization process was rapid and dominated by the nucleation process. To examine the ef fect of para vs meta on the energy barrier associated in forming stable nuclei for crystallization, the degree of undercooling ( !L = T c Ð Tm), was evaluated. 59,60 It was found that the value of !L for trans 3c was m uch larger than trans 3a, (45.0 ¡C vs. 27.6 ¡C). Since the melting temperature for both was nearly identical, this result can only be used to explain the multiple polymorphs, or multiple stable configurations, observed for trans 3a. The multiple configurations allowed for crystalline trans 3a to require less of a driving force to for m stable nucle i. Once a particular crystalline configuration is formed the others will quickly follow due to the increase in the change of surface free energy. Although similar trans 3a, the trans 3b isomers also exhibited multiple stable configurations (polymorphs A -C), the value of !L for tran s 3b was about 68.0 ¡C which was 40 ¡C larger than trans 3a. This result suggested that the bulky cyclohexyl moiety of trans 3b effectively inhibited the molecular rotation needed to form equilibrium crystalline solids. 33 !"#"$"$&0,+&,+-./(+ &Cis 3c also exhibited a crystallization transition upon cooling. However, its exothermic crystallization peak was much broader , and the degree of undercooling was quite large (~ 145 ¡C) as compare d to trans 3c. This is indicative of a crystallization process dominated by a 174 diffusional crystal growth process that is kinetically controlled rather than thermodynamically controlled. Therefore, cis 3c crystals take longer to grow than the other isomers. Furthermore, !Hc indicates that cis 3c does not fully crystallize upon cooling, !Hc / !Hm = 84 %, while !Hc for all other isomers is greater than 96 %. This behavior suggests that the configuration nece ssary for cis 3c to crystallize upon cooling is more difficult to attain than the other isomers, and the cooling rate may need to be slower. Cis 3a did not exhibit crystallization upon cooling at a cooling rate of 5 oC/min from its liquid state . However, cis 3a did exhibit a glass transition ( Tg = 56.0 ¡C) followed by a cold crystallization (T r = 90.8 ¡C) during a second heating. Therefore, slower cooling rates were investigated to determine if this was a kinetic effect. Lowering the cooling rate to 1 ¡C/m in and holding it isothermally at anticipated crystallization temperatures still did not allow it to crystallize upon cooling . This was attributed to the hydrogen bonding in the crystal lattice. During cooling, it becomes difficult for cis 3a to arrange in the orientation necessary for intermolecular hydrogen bonding. 62 Thus, energy was needed for the recrystallization process to occur, which was provided through heating. The crystallinity associated with T r is 63 % (!Hr / !Hm = 0.63), which wa s much lower crystallinity than the other isomers, and indicative of semi -crystalline behavior . Cis 3b exhibited a crystallization transition upon cooling most similar to the trans isomers, i t also had an endset point nearly equivalent to the supercool point, and !L = 65.1 ¡C. Surprisingly, this was approximately 3 ¡C smaller than trans 3b. Since both isomers have a bulky cyclohexyl moiety, it was anticipated this was a result of the number of stable configurations, with cis 3b having more than trans 3b, and thus able to form stable nuclei faster. 175 !"#"$%&'()*+,%)*'-.*-'/) %Polymorphs were only found for trans 3a and 3b, and the different polymorphs did not affect packing density. It wa s anticipated that trans 3b has more polymorphs than trans 3a due to the additional stable confirmations from the bulky cyclohexyl moiety inhibiting rotation . It wa s also anticipated that trans 3a has more polymorphs than trans 3c, but for different reasons. There are two lo cation s meta - to the D -Si atom bonded to the aminophenyl moiety . Since the meta -moiety can freely rotate, it can be crystallized in either of these two positions. Conversely, there is only one location para - to the D -Si atom. Furthermore, when attempting t o grow single crystals, trans 3c proved much more difficult. Exhaustive attempts in numerous solvent solutions provided only one structure that was not co -crystallized with solvent. On the other hand, two polymorphs of trans 3a were simultaneously crystall ized in the first solvent system attempted. Cis 3c did not exhibit polymorphs, and was the only crystal with a twinned structure. The key to attaining cis 3c was purity. Samples that were of high purity (> 95 %) failed to form single crystals in a large variety of solvents. Successful crystals were grown only after cis 3b was placed in a liquid chromatography column and further purified (> 99 %). It is unlikely that there are any polymorphs for cis 3a due t o the hydrogen bonding potential of the aminophenyl groups in the determined crystal structure . The distance between these hydrogen atoms and the intermolecular nitrogen atoms was 2.582 (†) of cis 3a. Of all the other structures, polymorph A of trans 3a was the only one that had a similar conformation to cis 3a, where the intermolecular aminophenyl groups are pointed towards one another. However, the distance between theses hydrogen atoms and the intermolecular nitrogen atom was 2.69 (†), which was too large for hydrogen bonding. Failed crystallization attempts showed that cis 3b was polycrystalline, and t he 176 bulky cyclohexyl group suggested that it was possible that there are multiple stable conformations , thus it was anticipated that cis 3b has a lot of polymorphs. !"#"#$%&'($ ')*(+,*&-$ ./''(0-/1$2344(-(50$ ')*(+,*&-$ .0-,+0,-(. $For a small molecule, para - and trans are typically more symmetric than meta - and cis and therefore, generally they will exhibit will generally, they will exhibit a more dense packin g arrangement. 1,3 As was previously mentioned, all trans isomers exhibit ed Ci symmetry and all cis isomers exhibit ed C2 symmetry, giving all isomers 2 irreducible representations within the molecule itself . Therefore , one isomer wa s n ot more symmetric than the other . Thus, the fact that the trans isomers exhibit ed a higher packing density is unique, and can only be attributed to desired arrangements of the molecules within the unit cells. Furthermore, since all isomers have the same mo lecular symmetry, it was proposed that the cis and trans isomers could be mixed without completely destroying the crystalline order. !"#"6$7(*0358$9(:&;3)-$ 4)-$<5+)58-,(50 =0/>($>:&.($23&8-&'.$?+)'>),52.$@+$&52$@&A $ For the phase diagrams of compounds 3c (Figure 6-5) and 3a (Figure 6-7) the melting point wa s d epressed from the pur e material and the melting endotherms beca me broader u ntil the eutectic composition was reached . The left hand side of both phase diagrams is indicative of the melting point depression of the gradual addition of trans isomer to pure cis isomer. Whereas the right hand side of both diagrams shows the melting point depression upon the gradual addition of cis isomer to pure trans isomer. The melting point depression for compound 3a was more rapid and further from ideal than compound 3c on both si des of the phas e diagram. Since hydrogen bonding is necessary in the crystal lattice in order to form stable crystallites for cis 3a, it is speculated that the melting point will depress more rapidly for this compound. 177 !"#"!$%&'()*+$,&-./)01$201$) 304015-063$30')7$ 30'6()0* 8(95&$5-.3&$7).+1.4$:;04506*7$<,= $ Melting points can be altered when a ÒguestÓ molecule is able to substitute into the crystal lattice of a ÒhostÓ molecule. If the ÒguestÓ molecule is very similar in size and dimensions to the ÒhostÓ molecular, th e mixture exhibits solid solubilities at all compositions because the lattice is not distorted . The melting point decreases with increasing content of the lower melting molecule, or increases with increasing content of the higher melting material. This typ e of phenomena is consistent with an isomorphous solid solution melting system, such as the phase diagram of compound 3b (Figure 6-10), where the melting point decreas ed linearly from xtrans = 1 (Tm = 270.0 ¡C) to xtrans = 0 (Tm = 265.9 ¡C). Melting point behavior was modeled using the Schrıder -van Laar equation for all compounds. From the solid -liquid equilibria results, it wa s apparent that the experimental liquidus values wer e very close to ideal calculations for compound 3b ( Table 6-5). Additionally, th e solidus and liquidus values wer e very similar, with in 0.2 ¡C, suggesting that the two -phase region hardly exists ( Figure 6-10). This was verified by DSC traces for binary mixtures of compound 3b (Figure 6-11). Typ ically, as the composition moved further from a pu re component, the endotherm became broader. The broadness is indicative of the distance between the solidus and liquidus curves, the more broad, the larger the distance. Thus, the fact that experimental melting endotherms of all compositions of compound 3b were narrow, suggesting that the ideal ca lculation of the solidus curve wa s accurate. In order for a system to be isomorphous, both components must soli dify in the same crystal system with similar lattice dimensions and related chemical constitution. 13 Ideal behavior of an isomorphous system is achieved when the lattice dimensions of both components match and the compounds are chemically similar. Additionally, the ÒguestÓ molecule cannot interrupt 178 the attractive -repulsive forces within the crystal lattice of the ÒhostÓ component. Positive dev iations from ideality would result from disruption and negative deviations from excess attractive forces. Since, it was determined that compound 3b was an ideal, i somorphous solution, it supported the hypothesis that the crystal structures of cis/trans 3b are similar. Additionally it was anticipated that trans 3b acts as nucleation sites for the cis isomer. !"#"$%&'()*)+),-.)'/% '+%0)/-12%,)34.1-/3%5)6.7183 % Compounds 3a-c exhibit ed a large degree of undercooling , or supersaturation when they were cooled. S upersaturation occurs when there is more solute than is permitted by stable thermodynamic equilibrium. 13 For compound 3c, as the composition moved further from a pure isomer, the transition bec ame broade r and the undercooling increased until it reached the eut ectic composition where there was no solidification observed ( Figure 6-12). Additionally, whe n the eutectic composition wa s heated a second time, there was no melting endotherm observed. The solidification of compound 3c was generally linear, unlike the me lting transition, which exhibited an incongruent eu tectic. The undercooling followed the same linear trend ( Figure 6-15). Compound 3a followed a similar tre nd; however, after it approached the eutectic composition there was no solidification observed for the remaining compositions. Compound 3b was t he only structure that continued to exhibit a supercool point equivalent to the onset point of solidification throughout the whole range of compositions, suggesting that even as a mixture , the material is highly crystalline. The undercooling generally increased as it moved towards a less pure composition ( Figure 6-15). However, at xtrans = 0.50, the so lidification and undercooling had the smallest value. 179 Figure 6-15. Undercooling (Tm-Tc) of binary cis/trans mixtures vs. xtrans for compounds (a) 3c (b) 3a, and (c) 3b. 0.0 0.2 0.4 0.6 0.8 1.0 406080100120140Temperature ( oC)xtrans (a) 0.0 0.2 0.4 0.6 0.8 1.0 255075100Temperature ( oC)xtrans (b) 0.0 0.2 0.4 0.6 0.8 1.0 606570758085Temperature ( !C)xtrans (c) 180 !"#$%&'()* +,'-$./01.23 $&'$3&),+ 4),5*,+$67/.01)$/5*,),8.,1 $ Solid -liquid equilibrium phase diagrams based on the thermal behavior for binary mixtures of cis and trans isomers for a series of DDSQ(X)(R) were successfully determined. The Schrıder -van Laar equat ion (E quation 6 -1), and the NRTL model (E quation 6 -2) were used to fit the experimental data. Compounds 3c and 3a both exhibit ed incongruent melting, while compound 3b was an ideal isomorphous solution. Changing the X -moiety from para - to meta -aminophenyl shifted the solid -liquid equilibrium further from ideal values and increased activity coefficients. Changing the R -moiety f rom methyl to cyclohexyl created an isomorphous system, which was anticipated to result from cyclohexyl being more similar to the phenyl moieties. Additi onally, it wa s anticipated tha t both isomers have similar, multiple stable configurations and thus can form solid solutions easily when one acts as a nucleation site for the other. Single crystal XRD was used to explain the thermal behavior. All isomers exhibit ed the same 2 irreproducible representations according to point group symmetry. Thus, trans was not more symmetric than cis, but still melted at a higher temperat ure for all compounds, which could only be attributed to the desired arrangements of th e molecules within the unit cell that increased the packing density; increasing the !Hm and T m. Similarly, para was not more symmetric than meta, and the fact that cis 3a (meta) had a higher T m than cis 3c (para) was a result of its intermolecular hydrogen bonding potential within the unit cell. Although cyclohexyl (compound 3b) increase d the surface are a of the DDSQ, it actually melted at a lower temperature than the analogous methyl DDSQ (compound 3a). This was attributed to the bulkiness of the cyclohexy l moiety disrupting the crystal packing. Exhaustive crystallization attempts were made to crystallize cis 3b, but without any success. However, cis 3b demonstrated a large affinity towards crystallization upon cooling in the DSC, which wa s attributed to trans 3b acting as 181 nucleation sites for the cis isomer. This was entirely possible due to the fact that compound 3b was an ideal isomorphous solid solution with very similar solidus and liquidus curves, which suggested that the crystal structures of its isomers are nearly identical. Overall, the determined phase diagrams c ould be used to select an isomer, or ratio of isomers necessary for reaction conditions, and/or desired characteristics for a novel engineering design . 182 !!! "#$#"#%&#' ! 183 !"#"!"$%"& ' (1) Pinal, R. Organic & Biomolecular Chemistry 2004, 2, 2692. (2) Plass, K. E.; Engle, K. M.; Matzger, A. J. Journal of the American Chemical Society 2007, 129, 15211. (3) Gavezzotti, A. Journal of the Chemical Society -Perkin Transactions 2 1995, 1399. (4) Slovokhotov, Y. L.; Neretin, I. S.; Howard, J. A. K. New Journal of Chemistry 2004, 28, 967. (5) Lloyd, M. A.; Patterson, G. E.; Simpson, G. H.; Duncan, L. L.; King, D. P.; Fu, Y.; Patrick, B. O. ; Parkin, S.; Brock, C. P. Acta Crystallographica Section B -Structural Science 2007, 63, 433. (6) Sysoev, S. V.; Cheremisina, T. N.; Zelenina, L. N.; Tkachev, S. V.; Zherikova, K. V.; Morozova, N. B.; Kuratieva, N. V. Journal of Thermal Analysis and Calor imetry 2010, 101, 41. (7) Aoki, K.; Nakagawa, M.; Ichimura, K. Journal of the American Chemical Society 2000, 122, 10997. (8) Bansal, R. K. A Textbook of Organic Chemistry ; 5th ed.; New Age International: New Delhi, India, 2007. (9) Gilbert, A. S. Thermochimica Acta 1999, 339, 131. (10) Aakeroy, C. B.; Forbes, S.; Desper, J. Journal of the American Chemical Society 2009, 131, 17048. (11) Plato, C.; Glasgow, A. R. Analytical Chemistry 1969, 41, 330. (12) He, B.; Martin, V.; Setterwall, F. Fluid P hase Equilibria 2003, 212, 97. (13) Lane, G. A. Solar heat storage : latent heat materials ; CRC Press: Boca Baton, Fla., 1983. (14) Walas, S. M. Phase equilibria in chemical engineering ; Butterworth: Boston, 1985. (15) Yamashita, H.; Hirakura, Y.; Yuda, M.; Teramura, T.; Terada, K. Pharmaceutical Research 2013, 30, 70. (16) Schoen, B. W.; Lira, C. T.; Lee, A. Fluid Phase Equilibria 2013, submitted July 8, 2013 . 184 (17) Schoen, B. W.; Holmes, D.; Lee, A. Magnetic Resonan ce in Chemistry 2013, 51, 490. (18) Still, W. C.; Kahn, M.; Mitra, A. Journal of Organic Chemistry 1978, 43, 2923. (19) Software for the CCD Detector Systems for Determining Data Collection Parameters ; 1.61 ed.; Systems, B. A. X. -r., Ed. Madison, Wisconsin, 2009, Software for the CCD Detector Systems for Determining Data Collection Parameters . (20) 2010.11-3 ed.; Systems, B. A. X. -r., Ed. Madison, Wisc onsin, 2010, Software for the CCD Detector System . (21) Software for the Integra tion of CCD Detector System ; 7.68A ed. Madison, Wisconsin, 2010, Software for the Integration of CCD Detector System . (22) Blessing, R. Acta Crystallographica Section A 1995, 51, 33. (23) Sheldrick, G. Acta Crystallographica Section A 2008, 64, 112. (24) TA Instruments Universal Analysis 2000, TA instruments Inc. version 4.5A Build 4.5.0.5, Waters LLC, 2000. (25) Spek, A. L. Acta Crystallographica Section D -Biological Crystallography 2009, 65, 148. (26) Elliott, J. R.; Lira, C. T. Introductory chemical engineering thermodynamics ; 2nd ed.; Prentice Hall: Upper Saddle River, NJ, 2012. (27) Prausnitz, J. M.; Lichtenthaler, R. N.; Azevedo, E. G. d. Molecular thermodynamics of fluid -phase equilibria ; 3rd ed.; Prentice Hall PTR: Upper Saddle River, N .J., 1999. (28) Kaiser, T. E.; Stepanenko, V.; Wuerthner, F. Journal of the American Chemical Society 2009, 131, 6719. (29) Kuduva, S. S.; Sarma, J.; Katz, A. K.; Carrell, H. L.; Desiraju, G. R. Journal of Physical Organic Chemistry 2000, 13, 719. (30) Manin, A. N.; Voronin, A. P.; Perlovich, G. L. Thermochimica Acta 2013, 551, 57. (31) Slovokhotov, Y. L.; Batsanov, A. S.; Howard, J. A. K. Structural Chemistry 2007, 18, 477. (32) Baird, J. A.; Van Eerdenbrugh, B.; Taylor, L. S. Journal of Pharmaceutic al Sciences 2010, 99, 3787. (33) Wunderlich, B.; Springer,: Berlin, 2005, p xvi. 185 (34) Karis, T. E.; Kim, S. J.; Gendler, P. L.; Cheng, Y. Y. Journal of Non -Crystalline Solids 1995, 191, 293. (35) Zhou, J.; Kye, Y.; Harbison, G. Journal of the American Chemical Society 2004, 126, 8392. (36) Ahn, H. W.; Clarson, S. J. Journal of Inorganic and Organometallic Polymers 2001, 11, 203. 186 !"#$%&' !!!! "##$%&"'%()* ! Keywords polyimide, thermoset, polyamide, Nomex, ionic liquid 187 !" #$$%&'()&*+, -!".-/*%0&1&23-)4351*,3) -!".".-6+)5*27')&*+ - Polyimides exhibit high viscosity, which requires high pressure to be used in order to process and fabricate the structural composites and adhesive joints. Additionally, since the solid to liquid phase transition is so close to the curing reaction, the processing window of polyimide thermosetting materials is quite small. Furthermore, the characteristics of current polyimide thermosetting material are polydispersed since the MW of the oligomers themselves are polydispersed. Therefore, research on polyimide s has been directed towards decreasing the viscosity and increasing the processing window without sacrificing HPM characteristics. It w as anticipated that using DDSQ as a backbone for these polyimide thermosetting materials would allow these objectives to be reached, while providing monodispersed characteristics. Since an amorphous material provides a lower viscosity than a crystalline one, only compounds 3a, 3c, and 3d were used for this application. !"."8-9*%:3+),-(+2-53(;3+), - Tetrahydrofuran (THF) and toluene were obtained from Sigma -Aldrich and distilled under nitrogen. Phenylethynyl phthalic anhydride (PEPA) was obtained from Chriskev Company. !"."<-=7'%3(5-1(;+3)&'-5 3,*+(+'3 -Compounds were measured at 25 ¼C on a Varian UNITY -Inova 6 00 spectrometer equipped with a 5 mm Pulsed -Field-Gradient (PFG) switchable broadband probe and operating at 599.80 MHz ( 1H). 1H NMR data were acquired using a recycle delay of at least 20 s and 32 188 scans to e nsure accurate integration. The 1H-chemic al shifts were referenced to that of residual protonated solvent in CDCl 3 (7.24 ppm). 29Si NMR data were acquired using a recycle delay of 12 s with inverse -gated decoupling. The pulse angle was set to 90 ¡. 29Si spectra were referenced against the lock solvent using vendor supplied lock referencing. 13C NMR data were acquired using a recycle delay of 1 s and 256 scans. !"#"$%&'()*+,-, %Scheme 7-1 shows the synthesis of cis and trans [(meta-phenylethynylphthalimide ) methylsilyl] -bridged -(phenyl )8-double -decker silsesquioxane, DDSQ( m-PEPI)(Me) -5a, [(par a- phenylethynylphthalimide )methylsilyl] -bridged -(phenyl )8-double -decker silsesquioxane, Scheme 7-1. Synthesis of compound 5. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNNOOPhOOPhSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiSiMeMeH2NNH2DDSQ(AP)(Me) +OOOPhPEPA THF/Toluene 105 C20 hrs. ¡DDSQ(PEPI)(Me) 189 DDSQ(p-PEPI)(Me) -5c, and [(meta/par a- phenylethynylphthalimide )methylsilyl] -bridged -(phenyl )8-double -decker silsesquioxane, DDSQ(p-PEPI)(Me) -5d, through the reaction of DDSQ(AP)(Me) , compound 3, with phenylethynyl phthalic anhydride (Figure 7-1).1,2 Figure 7-1. Compounds (a) 5a- all meta, (b) 5c- all para, (c) 5di- one side meta and one side para. All compounds also have cis and trans isomers about the SQ cage with respect to the D-Silicon, as seen in the previ ous chapters for compounds 3a, 3b, and 3d. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNPhOOa)b)SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNPhOOcompound 5aSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNPEPIPhOOc)compound 5ccompound 5dPEPIPEPI 190 !"#"$"#%&'()'*+,%-. %Under a nitrogen atmosphere, a well -stirred solution of 3a (5.35 g, 4.00 mmol) and PEPA ( 2.18 g, 8.80 mmol) in anhydrous THF ( 70 mL) and toluene ( 70 mL) was stirred at room temperature (25 ¡C) for 1 h. The solution was heated to 60 ¡C for 2 h and refluxed at 115 ¡C for 20 h. Solvent was removed under vacuum and subsequently washed and precipitated with m ethanol . The product was filtere d and dried in a vacuum oven at 125 ¡C for 24 h ( 6.52 g, 3.63 mmole, 91% yield). NMR resonances (ppm) follow. 29Si NMR, !: -31.5, -77.9, -79.1, -79.2, -79.3. 13C NMR , !: 166.6, 166.5, 137.7 137.6, 137.2, 134.1, 133.3, 132.2, 132.0, 131.6, 131.4, 131.3, 131 .0, 130.7, 130.5, 130.4, 130.0, 129.4, 128.8, 128.7, 128.2, 128.0, 127.9, 127.7, 127.6, 126.55, 123.8, 122.14, 94.2, 87.9, -0.3. 1H NMR, !: 8.10 -7.11 (64 H, overlapping multiplets), 0.66 and 0.65 (6 H, overlapping singlets) ii (Figure 7-2). Figure 7-2. NMR spectra of compound 5a (a) 29Si and (b) 1H, the resonances at !H 3.5 and 1.9 ppm are residual methanol. 8765432108.007.757.507.257.00b)(ppm) 191 !"#"$"%&'()*(+,-&./ &Under a nitrogen atmosphere, a well -stirred solution of 3c (8.00 g, 5.99 mmol ) and PEPA (3.04 g, 12.23 mmol) in anhydrous THF (65 mL) and toluene (78 mL) was stirred at room temp erature (25 ¡C) for 1 h. The solution was heated to 60 ¡C for 2 h and refluxed at 115 ¡C for 20 h. Solvent was removed under vacuum and subsequently washed and precipitated with m ethanol . The product was filtere d and dried in a vacuum oven at 125 ¡C for 24 h (10.12 g, 5.63 mmole, 94% yield). NMR resonances (ppm) follow. 29Si NMR, !: -31.3, -78.1, -79.0, -79.2, -79.4. 13C NMR , !: 166.6, 166.5, 137.3 136.4, 134 .4, 134.2, 134.1, 133.3, 132.2, 132.0, 131.8, 130.8, 130.6, 130.5, 130.2, 129.4, 128.7, 128.0, 127.9, 127.8, 127.7, 126.7, 125.7, 123.9, 122.2, 94.4, 87.9, -0.3. 1H NMR, !: 8.14 -7.16 (64 H, overlapping multiplets), 0.63 and 0.62 (6 H, overlapping singlets) ii (Figure 7-3). Figure 7-3. NMR spectra of compound 5c (a) 29Si and (b) 1H. 8.007.757.507.257.00876543210(ppm) b) 192 !"#"$"%&'()*(+,-&.- /&Under a nitrogen atmosphere, a well -stirred solution of Ph 8bisaniline -POSS ( 0.997 g, 0.744 mmol) and PEPA (0.744 g, 2.97 mmol) in anhydrous THF (8 mL) and toluene (8 mL) was stirred at room temp erature (25 ¡C) for 1 hour. The solution was heated to 60 ¡C for 2 h and refluxed at 115 ¡C for 20 h. Solvent was removed under vacuum and subsequently washed and precipitated with m ethanol . The product was filtere d and dried in a vacuum oven at 125 ¡C for 24 h (1.12 g, 0.623 mmole, 84% yield) . Due to complications with the syn thetic procedure (Appendix V), additional excess PEPA was used for the synthesis of compound 5d. Unreacted PEPA was removed using silica gel chromatography columns ( dichloromethane : hexane = 7 : 3), NMR resonances (ppm) follow. 29Si NMR, !: -31.1, -31.3, -77.7 (2 overlapping singlets), - 78.87, -78.93, -78.98, -79.02, -79.04, -79.07, -79.14, -79.17, -79.19. 13C NMR, !: 166.4, 138.7, 137.2, 137.1, 134.3, 134.1, 133.2, 132.0, 131.7, 131.4, 131.0, 130.5, 129.7, 129.3, 128.6, 127.9, Figure 7-4. NMR spectra of compound 5d (a) 29Si and (b) 1H. 8.007.757.507.257.00876543210(ppm) b) 193 127.8, 127.7, 127.6, 127.5, 126.6, 126.5, 125.6, 123.8, 122.1, 94.3, 94.2, 87.9, 87.8, -0.3. 1H NMR, !: 8.08 -7.03 (64 H, overlapping multiplets), 0.56, 0.55, and 0.54 (6 H, 6 overlapping singlets )iii (Figure 7-4). !"#"$%&'()*+,%- ('+./0) %Melting behavior was studied using a Mettler Tolido DSC -1 equipped with a mechanical cooling system under a nitrogen atmosphere. Samples were placed in a Q -zero TM aluminum pan and sealed with a punctured lid to ev aluate solid to liquid transitions and the curing polymerization reaction. Samples were first equilibrated at 25 ¡C for 1 min and subsequently heated to 500 ¡C with a constant heating rate of 10 ¡C/min. !"#"1%2/3403/56%*(+37)(*(853 % Viscosity measurements were conducted on a TA Instruments AR 2000, equipped with an environmental control system. Steady -state shear viscosity (10 s-1) as a function of temperature was measured using an 8 mm diameter parallel plate fixture with a 0.4 mm gap . For handling purposes, sample powders (approximately 70 mg) were prepared through press -molding at room temperature. The fixture was equilibrated to the initial temperature ( 170 ¡C) for 30 min, and the sample was then placed betw een the gap. T he temperature was then increased at a constant heating rate until the final temperature ( 250 ¡C ) was reached. Steady -state viscosity as a function of shear rate was also measured at 170 ¡C, 200 ¡C, and 250 ¡C , from a shear rate of 1 s-1 through 100 s -1. 194 !"#"!$%&'()*'$+,-$-.'/(''.0, $!"#"!"#$%&'()*+$*,*+-./. $Differential scanning Calorimetry (DSC) was used to evaluate the glass (Tg) and melting transition s (Tm and !Hm) of the uncured oligomers and the subsequent curing reaction (TRXN and !HRXN ). Results are summarized in Table 7-1, and DSC curves are shown in Figure 7-5. Table 7-1. DSC data for polyimide thermosets . Compound Tg (¡C) Tm (¡C) !Hm (kJ/mole) TRXN (¡C) !HRXN (kJ/mole) 5a -- 127.1 9 326.3 256 5c -- 256.7 13 337.4 206 trans 5c -- 302.5 59 343.6 226 cis 5c 123.4 186.8 19 327.6 221 5d 106.4 -- -- 332.2 287 Compounds 5a and 5c show ed several transitions before the curing reaction, which were anticipated to be from the incomplete mixing of the isomers ( Figure 7-5a,b). In order to determine if this were true, trans and cis oligomers of 5c also evaluated. The DSC curve of trans 5c showed a sharper melting transition (T m = 302.5 ¡C) and a smoother baseline ( Figure 7-5c) than the mixed isomer of 5c (Figure 7-5d). The second T m shown in the mixture of isomers of 5c (Figure 7-5b) was at approximately 256.7 ¡C, which is likely representative of the trans 5c melting transition. It is depressed from its pure value as a result of the incorporation of the cis isomers. Similarly, the T g and the first T m (Figure 7-5b) are rep resentative of the cis isomers in 5c (Figure 7-5d), which wer e also depressed as a result of the incorporation of the trans isomer . It is not surprising that the cis isomer exhibit ed a larger depression in the T g and the first T m 195 Figure 7-5. DSC curves of (a) 5a, (b) 5c, (c) trans 5c, (d) cis 5c, and (d) 5d, from 30 to 500 ¡ C with a constant heating rate of 10 ¡C / min. 100200300400500-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 (a) Temperature ( !C)Heat flow (W/g) exo ^ Tm = 127.1 100200300400500-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 (b) Tm cis Tg cis Temperature ( !C)Heat flow (W/g) exo ^ Tm trans 100200300400500-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 (c) Temperature ( !C)Heat flow (W/g) exo ^ Tm = 302.5 100200300400500-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 (d) Tg = 123.4 Temperature ( !C)Heat flow (W/g) exo ^ Tm = 186.8 100200300400500-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 Tg = 106.4 Heat flow (W/g) exo ^ Temperature ( !C)(e) 196 since a similar trend was observed in the melting data for binary cis/trans mixtures of compound 3 (Chapter 6 ). This data demonstrates that compounds 5a and 5c are semi -crystalline. Compound 5d is the only oligomer that is completely amorphous, and its Tg is lower than other phenyl ethynyl end -capped oligomers. 3 Using a melting point apparatus, it was determined that at this Tg, compound 5d complet ely under goes a solid -liquid transition. The cure reaction occurred at approximately 330 ¡C for all compounds. The large st difference between the solid to liquid tran sition and the curing reaction wa s for compound 5d. Compound 5d also had the largest heat of reaction ( !HRXN = 287 kJ/mole) . It was anticipated that this wa s a result of the degree of crosslinking. The curing reaction occurred considerably lower (326 Ð 344 ¡C) than the onset of degradation for all compounds (475 ¡C). !"#"!"$%&'()*('+, % Eliminating c rystallinity and the processing window of the oligomer are the principal motifs for this application, and c ompound 5d had the most desirable combination of the two . It was fully amorphous and exhibited the largest processing window , completely melting below 170 ¡C which has been attributed to the inherent asymmetry of the molecule . Thus, only the viscosity of compound 5d was studied . Steady state shear viscosity was measured from 170 ¡C through 250 ¡C for compound 5d. Results were evaluated using TA Rheology advantage data analysis. 4 Two different batches were compared to test for consistency of the synthesis and characteristics (Figure 7-6). Results indicate d that there is consistency in the two different 197 batches, which is an indication that the oligomer properties are mono -dispersed. Additionally, the visc osity c ontinuously decreased with increasing temperature throughout the temperatures Figure 7-6. Complex viscosity of two batches of compound 5d; filled square ( !) = batch 1, and unfilled square ( ) = batch 2. 18020022024005101520253035Viscosity (Pa!s)Temperature (oC) batch 1 batch 2 198 range chosen. Compared to literature values, 3,5,6 the viscosity of compound 5d was much lower throughout the range of temperatures observed ( Table 7-2). Table 7-2. Complex viscosity for compound 5d. Complex viscosity (Pa !s) at the following T ( ¡C) Compound 170 200 230 250 Mmv (Pa s, ¡C) Batch 1 33.37 3.16 0.94 0.47 0.42 at 249 Batch 2 34.04 3.03 0.80 0.49 0.49 at 250 PI-A (G:A) a,6 -- -- -- 0.58 0.25 at 288 PI-A (G:B) a,5 -- -- -- 1.0 0.27 at 299 PI-A (3G:A) a,6 -- -- -- 0.61 0.28 308 PETI-3K3 -- -- -- -- 9 at 353 a Background information (p. 48) !"#"$%&'()*+,-(.%/012/34%'(% 5'*6-1-,0%780/1'4074 %A series of phenylethynyl -encapped oligomides were prepared from compound s 3a, 3c, 3d, and PEPA. The incorporation of the silsesquioxane cage decreased the glass transition temperature, while maintaining high thermomechanical features of the oligoimide, which provided a larger processing window. According to DSC investigations, all oligomers exhibited an irreversible cross -linking reaction above 320 ¡C to form a polyimide thermoset. DSC studies were also used to study the existence of specific thermal features , such as a crystalline melting and glass transition of the resulting oilgoimides. Results sho wed that b oth compounds 5a and 5c with mixtures of cis/trans exhibited a glass transition and broad endothermic crystalline melting transitions, whereas isolated cis and trans isomers exhibited sharper endothermic melting transitions. Thus, suggesting that mixtures of isomers were more amorphous than isolated isomers. Compound 5d did not exhibit any form of an endothermic crystalline melting transition, which suggested that it was wholly amorphous . Additionally, compound 5d exhibited the lowest 199 Tg and solid to liquid transition , it was completely liquidus below 170 ¡C. Thus, it appears that the assymetric mixture of stereo - (cis and trans ) and regio - (para - and meta -) isomers eliminates crystallinity and increases the processing window. Steady state shear viscosity of compound 5d was also investigated. It was determined that the viscosity continuously decreases with increasing temperature throughout the temperatures range chosen and is lower than literature values. Additionally, two batches of compound 5d were examined, with similar finding s, suggesting that these oligomers and their properties are monodispersed. Therefore compound 5d is a viable alternative for this class of polyimide thermosetting materials. 200 !"#$%&'()*)+,-$./0*+&1')2.,32 $!"#"4$56.*&-73., &6$ In light of the remarkable characteristics displayed by polyaramids, research efforts are directed in two areas: (1) reducing the cohesive energy resulting from H -bonding and pi -stacking so these materials are easier to process, and soluble in common o rganic solvents without sacrificing high performance properties, and (2) expanding their high -performance properties to additional applications in new and promising fields, such as: optically active, luminescent, ionicexchange, flame -resistant and fiber -forming materials. Prior to this dissertation, studies have demonstrated that SQ based polyamides improve thermomechanical properties for PA6. However, no studies have used an SQ cage -like structure that is covalently bonded to a polyamide backbone, which co uld potentially have a more significant effect on the polymer properties. This section investigates the synthesis of DDSQ -based Nomex and the affect that the cage has on the melting and degradation characteristics of these polymers. Various mass fractions of compound 3a and MPDA are reacted with ICL to determine if the amount of DDSQ in the backbone affects the MW of the polymer. It was proposed that the bulky Si -O core would reduce the cohesive energy without sacrificing the thermomechanical properties of neat Nomex. !"#"#$8&'906.2$)6-$*0):06.2 $ Trimehtylchlorosilane, triethyl amines, toluene, m-phenylene diamine, and isophthloyl chloride were purchased from Sigma Aldrich. Solvents were distilled under nitrogen. Lithium chloride was purchased from Jade Scientific (Canto n, Michigan) and used without further purification. Dry n -methylpyrrollidine was purchased from Alfa Aesar (Ward Hill, Massachussettes) and used without further purification. 201 !"#"$%&'()*+,%-+./*01(%,*23/+/(* %Compounds were measured at 25 ¼C on a Varian UNI TY-Inova 6 00 spectrometer equipped with a 5 mm Pulsed -Field-Gradient (PFG) switchable broadband probe and operating at 599.80 MHz ( 1H). 1H NMR data were acquired using a recycle delay of at least 20 s and 32 scans to e nsure accurate integration. The 1H-che mical shifts were referenced to that of residual protonated solvent in CDCl 3 (7.24 ppm). 7.2.4 Synthesis DDSQ-based Nomex was synthesized from compound 3a, N-silylated m -phenylene diamine (MPDA), and isophthloyl chloride (ICL) ( Scheme 7-2).7 MPDA was silylated prior to polymerization. Scheme 7-2. DDSQ-based Nomex scheme. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCH3SiH3CNN*ONOHHxHNHOO*yClClOONNHTMS HTMS SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCH3SiH3C++Ysilylated MPDAICL3a(X + Y)XNH2H2N60 ¡C 20 hrs. 6 + hrs. - 10 to - 20 ¡CMPDATMSCl TEA, Toluene LiCl NMPH2NNH2 202 !"#"$"%&'()*)+,-.&/012 &&&Under a nitrogen atmosphere, TMSCl was added dropwise to a stirring solution of solid MPDA (30.0 g, 277 mmole), triethylamines (72.3 g, 665 mmole), and T oluene (300 mL). The solution turned yellow almost immediately. It was then heated to 60 ¡C. After 20 h, the solution became a viscous liquid. Solvent was removed under vacuum and the product was purified by fractional distillation to yield a colorless, vi scuous liquid (60.93 g, 243 mmole). 1H NMR !: 6.94 (1 H, mu ltiplet), 6.12 -5.96 (3 H, overlapping multiplets), 6.51 (2 H, NH2, broad singlet ), 0.25, 0.24, 0.14, 0.05 (18 H, singlets) (Figure 7-7). Figure 7-7. 1H NMR spectrum of silylated MPDA. !"#"$"#&11'34)(5-&06)*+7+8(.- & All polyaramids were prepared according to the following general procedure. The value of X and Y was varied ( Table 7-3). Under a nitrogen atmosphere, a solution of ICL (X + Y mmole) and NMP was added dropwise to a stirred solution of 3a (X mmole), silylated MPDA ( Y mmole), LiCl (5 wt %), and N -methyl pyrrolido ne (NMP) (ca. 0.5 mmole (X+Y)/mL), 2 at -10 ¡C to -20 ¡C. After 5 h, the solution became viscous and was precipitated into cold Methanol to form a white solid. The product was isolated by filtration and dried under vacuum. A total of 17 203 polymers were synthesized according to this procedure. Nomex (0 mole % DDSQ) was synthesized as a model polymer for these studies (P -1 - P-5). Table 7-3. Polymer experiments. Exp # Mass % DDSQ Mole % DDSQ a Product mass (g) Yield % P-1 0 0 0.198 84 P-2 0 0 1.32b -- P-3 0 0 0.538 90 P-4 0 0 0.579 98 P-5 0 0 0.718 100 P-6 1.1 0.2 -- -- P-7 1.1 0.2 0.710 76 P-8 5.3 1 -- -- P-9 5.3 1 0.501 78 P-10 10.2 2 0.550 83 P-11 14.5 3 -- -- P-12 14.5 3 0.535 78 P-13 22.3 5 -- -- P-14 22.3 5 -- -- P-15 22.3 5 0.580 P-16 78.3 50 0.051 43 P-17 100 100 0.051 63 a Mole % of total amines used: X/(X+Y). b Not fully dry. -- not calculated. Three additional polymers ( Table 7-4) were synthesized in order to achieve the highest MW feasible. All were comprised of 1.1 mass % DDSQ (0.2 mole %). Under a nitrogen atmosphere, a solution of ICL (0.609 g, 2.99 mmole) and NMP was added dropwise to a stirred solution of silylated MPDA (0.631 g, 2.99 mmole), LiCl (0.318 g , 5.14 mmole ), and NMP (5 mL), at -10 ¡C to -20 ¡C. After (1 Ð 4.5 h), a solution of 3a (0.008 g, 0.006 mmole) and NMP (0.013 mL) was added to the reaction and allowed to stir. Subsequently, a solution of ICL Table 7-4. Additional polymer reactions. Exp # Mole % DDSQ Mass % DDSQ time added a Product (g) Yield P-18 1.1 0.2 1 0.733^ 102 b P-19 1.1 0.2 3 0.765^ 107 b P-20 1.1 0.2 4.5 0.705 98 a Time DDSQ added b May not be fully dry 204 (0.0001 g, 0.005 mmole) and NMP was added dropwise to the reaction. After 5 additional h, the solution became viscous and was precipitated into cold m ethanol to form a white solid. The product was isolated by filtration and dried under vacuum. According to 1H NMR spectra, the resonance at approximately ! 3.5 (NH) in the silylated MPDA product ( Figure 7-7) shifted upfield to approximately ! 10.5 (NH) after polymerization (Figure 7-8). The figure demonstrat es that there are two distinct resonances in this region for the co-polymers ( Figure 7-8b). One is associated with the amine in Nomex ( Figure 7-8a) and the other is associated with the amine on the DDSQ ( Figure 7-8c). Figure 7-8. 1H NMR spectra of (a) 0 mole % DDSQ (Nomex), (b) 50 mole % DDSQ, and (c) 100 mole % DDSQ. 205 !"#"$%&'()*+(,-%.)/012 % Molecular weight of the polymers was determined using two different methods. First, gel permeation chromatography (GPC) was used for the polymers that were soluble in DMF. Second, the polymers that were not soluble in DMF were measured relative to the same polymer/solvent pair using viscometer measurements. !"#"$"%&'()&*(+,(-./01&23+0,-.04+-*35 & For pol ymers that were soluble in DMF, GPC could be used in order to determine the number average MW (M n) was attainable using GPC analysis. The M n can be defined by the following formula (E quation 7 -1): !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where N i Mi is the total weight of polymers and N i is the total number of polymers. Typical M n for Nomex polymers is in the range of 10 -30 x 103 g/m ol. This would provide approximately 42 Ð 126 units of y per polymer chain ( Scheme 7-2). The PDI for these polymers is typically near 2.0 for polymers with a lower M n and 3 for polymers with an M n > 35 x 103 g/mol. Table 7-5. Mn for select polymers that are soluble in DMF. Exp # Mass % DDSQ Mn (103 g/mol) units PDI P-2 0 93900 394 3.8 P-3 0 109388 459 1.66 P-6 1.1 58250 231 2.5 P-8 5.3 68036 224 2.17 P-16 78.3 11547 14 2.0 P-17 100 11000 8 2.1 206 It appears that P -2, 3, 6, and 8 are all high MW polyaramids according to their M n, PDI and number of units per polymer chain. Conversely, P -16 and P -17 appear to be oligomers. Furthermore, it was determined that P -16 and 17 are soluble in THF, whereas all other polymers were not. This confirms that P -16 and 17 are low molecular weight polymers, or oligomers. It is anticipated that the remaining polymers that were not soluble in DMF have a higher MW, since increasing MW of Nomex polymers has been seen to exhibit decreased solubility. 8 !"#"$"#%&'()*('+,%-./(01.-.2+( % Viscosity was used as a relative measure of the MW of the polymers, since according to the Mark -Houwink equation it is proportional to MW (E quation 7 -2): !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where [ !] is intrinsic viscosity, M is molecular weight, and K and " are constants that are unique to the specific polymer and solvent pair. Since K and " are unknown for these polymer and solvent pairs, [ !] alone was used as a relati ve measure of the molecular weight of the polymers. To measure the intrinsic viscosity experimentally, first the specific viscosity reduced by the concentration ( !sp / c) of the solution is extrapolated to zero. A simple linear extrapolation of the Huggin s equation is a common method for determining the intrinsic viscosity (E quation 7 -3): !!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where c is concentration and k ! is a constant that depends on the specific polymer, solvent, and temperature of the solution, but is independent of MW. Alternatively, extrapolating inherent viscosity ( !inh ) to zero can also provide a measurement of the intrinsic viscosity (E quation 7 -4): 207 !!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! where k !! is a constant dependent on the polymer, solvent, and temperature of the given system. The inherent viscosity ( !inh ) and specific viscosity ( !sp) can be determined by comparing the time (t) that it takes t he polymer solution to flow through the capillary to the flow time of the pure solvent (t 0) (Equation 7 -5). !!!!!"!!!!!!!!!!"!!!!"!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Since viscosity can only be measured relative to the same polymer, p olymers comprised of the same mass % DDSQ were compared ( Figure 7-9). At 1.1 mass %, it was important to see if the time that the DDSQ was added had an affect on the M W of the polymer. It does not appear that the time of addition largely affects the MW of the polymer with the exception of P -20. For P-20, DDSQ was added to the reaction solution 4.5 h after the polymerization began. When all the components were added simu ltaneously, the polymeri zation takes approximately 5 h Figure 7-9. Extrapolation to intrinsic viscosity for 1.1 mass % DDSQ. 0.0 0.4 0.8 1.2 1.6 2.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 !inh (P-7) !sp (P-7) !inh (P-19) !sp (P-19) !inh (P-18) !sp (P-18) !inh (P-20) !sp (P-20) 1.1 mass % DDSQ Viscosity Concentration 208 Therefore, when DDSQ was added for P -20, the polymerization may have been complete. Both intrinsic viscosity and specific viscosity were linearly extrapolated to zero, from three different concentrations in NMP, in order to determine an accurate value of the intrinsic viscosity, and both constants, k ! and k !!. The y -intercept re presents intrinsic viscosity, and the constants k ! and k!! were determined from equations 7 -2 and 7 -3. Intrinsic viscosity, and values for the constants of several of the polymers were evaluated ( Table 7-6). Once k ! and k !! are determined, the intrinsic viscosity of the same polymer, solvent pair can be determined from one concentration. Table 7-6. Intrinsic viscosity measu rements of selected polymers in NMP. Polymer Mass % DDSQ [!] (dl/g) k! k!! P-4 0 0.69 0.34 0.58 P-5 0 0.59 0.25 0.42 P-7 1.1 0.92 0.18 0.34 P-18 1.1 1.0 0.046 0.24 P-19 1.1 0.92 0.23 0.39 P-20 1.1 0.64 0.21 0.38 P-9 5.3 0.83 0.42 0.43 P-10 10.2 1.2 0.17 0.31 P-12 14.5 0.65 0.14 0.34 P-15 22.3 0.64 0.28 0.44 !"#"$%&'()*+*,-./%*/*012-2 % Degradation characteristics of the polymers were determined using thermal gravimetric analysis under a nitrogen atmosphere. Samples were equilibrated at 50 ¡C and then heated to 500 ¡C with a heating rate of 10 ¡C/min. All polymers appear to have an onset of degradation (T d) over 400 ¡C (Table) which is typical of Nomex polymers (400 -430 ¡C for neat Nomex). 8 It appears that as the mass % of DDSQ is increased, then T d also appears to increase. Adding the DDSQ to neat Nomex at a later t ime during the reaction did not appear to have any affect on T d. 209 Features of degradation were also evaluated by observing the % weight loss vs. the temperature and the derivative of the % weight lost and the temperature ( Figure 7-10). Even Table 7-7. Onset of degradation (T d) of Nomex polymers. Exp # Mole % DDSQ Td (¡C) P-1 0.0 395 P-2 0.0 408 P-4 0.0 402 P-3 0.0 411 P-5 0.0 408 P-7 1.1 400 P-6 1.1 406 P-8 5.3 406 P-9 5.3 404 P-10 10.2 405 P-12 14.5 403 P-15 22.0 401 P-16 50.0 423 P-17 100.0 444 P-18 1.1 (1 h) 405 P-19 1.1 (3 h) 407 P-20 1.1 (4.5 h) 406 though P -2 and P -3 may only be oligomers according to their M n, they actually exhibit a much higher T d than the other polymers. There appear to be two characteristics to the degradation curves, which are emphasized when viewing the derivative of the % weight loss vs. the temperature ( Figure 7-10b). For P -18 (0 mass % DDSQ), the % weight loss exhibits a rapid onset at approximately 400 ¡C and then a maximum at approximately 425 ¡C. The % weight loss then becomes more gra dual until approximately 475 ¡C where the % weight loss becomes more rapid again, but not as rapid as the initial weight loss. As the mass % DDSQ is increased in the polymer, the initial weight loss is not as rapid, and the first maximum is much lower. At 78.3 mass % DDSQ, the second maximum is higher than the first. At 100 mass % DDSQ, there first 210 and second maximum appear to have converged. These features demonstrate that the addition of DDSQ increased the degradation properties of neat Nomex. Figure 7-10. TGA analysis of select polymers, (a) degradation (b) derivative of the degradation. 50100150200250300350400450500707580859095100a) P-5 (0 mass % DDSQ) P-14 (22.3 mass % DDSQ) P-3 (78.3 mass % DDSQ) P-2 (100 mass % DDSQ) % Weight loss Temperature (oC)3003253503754004254504755000.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 P-5 (0 mass % DDSQ) P-14 (22.3 mass % DDSQ) P-3 (78.3 mass % DDSQ) P-2 (100 mass % DDSQ) Temperature (oC)Derivative % weight loss b) 211 !"#"!$%&'(($)*'+(,),-+$'+.$/0&),+1$)0/20*')3*0( $ Differential scanning Calorimetry (DSC) was used to determine he gla ss transition (T g) and the melting transition (T m and !Hm) of the polymers. According to the literature, neat Nomex generally has a T g of approximately 275 ¡C, and a T m that is very close to its degradation temperature. From these studies, it appears that as the mass % of DDSQ is increased, then the onset of degradation is decreased. Additionally, the melting transition is decreased and the heat of fusion associated with the melting transition is also decreased. Table 7-8. Tg and T m of selected Nomex polymers. Exp # Mass % DDSQ Tg onset Tm onset !Hm (J/g) P-3 0 261 411 70 P-4 0 266 408 47 P-5 0 263 404 30 P-7 1.1 268 407 66 P-9 5.3 263 405 16 P-10 10.15 252 404 15 P-12 14.5 256 406 31 P-15 22.26 243 401 11 P-18 1.1 243 408 35 P-19 1.1 265 410 73 P-20 1.1 262 407 19 !"#"4$5,&/( $ Films were made for some of the polymers. Films were made using three different methods: (1) slow evaporation, (2) vacuum oven, and (3) UV lamp. For method 1, P -4 (0 mass % DDSQ) was dissolved in a minimal amount of NMP, and added to a glass, evaporating dish. Solvent was allowed to evaporate, and a white solid appeared ( Figure 7-11a). The films were not transparent. For method 2, polymer was dissolved in minimal NMP, and the solution dripped onto glass slides. The slides were placed in a vacuum o ven at 80 ¡C for two hours. Some of the 212 films were transparent ( Figure 7-11b). However, the thickness of the films was not well controlled, and therefore the transparency of the films varied. For method 3, from left to right, P -12 (14.5 mass % DDSQ), two slides of P -15 (22.3 mass % DDSQ), and P -4 and P -5 (0 mass % DDSQ) were dis solved in minimal NMP and then placed on aluminum foil. The thickness of the films were controlled to 100 µm and left under a UV lamp for 20 h. These films were all transparent. a) b) c) Figure 7-11. Films made from select polymers (a) slow evaporation, (b) vacuum oven, and (c) UV lamp. 213 !"#"$%&'()*+,-(.%/012/34%'(%5567 89240,%:'10; % DDSQ-based Nomex was synthesized and inherent viscosity was determined. From the inherent viscosity studies, it appears that the polymers have high MW, but further analyses such as determining the MW using a high temperature GPC with NMP need to be carried out. Polymers were also characterized according to their characteristics. These features demonstrate that the addition of DDSQ increased the degradation properties of neat Nomex. Additionally, the melting transition decreased and the heat of fusion associated with the melting transition is also decreased. Films could be made with these polymers. Once accurate MW are determ ined, mechanical testing can be carried out to determine if these properties are improved from the DDSQ macromer. 214 !"#$%&'()$*(+,(- $!"#".$%'/0&-,)/(&' $ Properties of ionic liquids (ILs) can be enhanced through the incorpor ation of nanoparticles such as silica a nd metal supporting oxides. Cage -like , monofunctionalized silsesquioxanes have been used as IL -supporting substrates with the intention of applying them as electrolytes in fuel cells .9 These silsesquioxane -based ILs have demonstrated improved thermal properties such as decreasing the melting temperature of ionic salts such that they become ILs .10 Furthermore, they improved the thermal stability of salts that had intrinsically low decomposition temperatures. To date, no difunctionalized ionic liquids have been synthesized, which would provide a new series of IL -based materials that have the added feature of geometrical isomers. !"#"1$2&345'/6$7'-$05785'/6 $ Tetrahydrofuran (THF) and iodomethane were purchased from Sigma -Aldrich . Iodomethane was filtered through silica gel prior to use. SiliaFlash ¨ P60 Silica gel was purchased from Silicycle Ultra Pure Silica Gels (Quebec City, Quebec, Canada). Potassium carbonate was purchased from Spectrum (Canton, Michigan) and used as received. Solvents used for liquid chromatography -mass spectroscopy were aqueous ammonium acetate and acetonitrile, obtained as HPLC -grade from Sigma Aldrich and used as received. !"#"#$29'/:56(6$&;$(&'()$3(+,(-$<=7> $ DDSQ Ionic liquid ( 6a) was synthesized according to Scheme 7-3. Under a nitrogen atmosphere, iodomethane (0.568 g, 4 mmole) was added dropwise to a stirring solution of 215 compound 3a (0.334 g, 0.25 mmole), potassium carbonat e (0.276 g, 2 mmole), and THF (11 mL) at room temperature (25 ¡C). After two weeks, a white precipitate occurred, which was removed by filtration and the supernatant was dried under vacuum. The supernatant solid was then extracted with acetonitrile and wat er to isolate the product. Acetonitrile was removed under vacuum and the product was dried under nitrogen. 1H NMR, !: 8.04 -7.01 (40 H, overlapping multiplets), 3.30 (18 H, N -(CH3)3, multiplet), 0.55 (6 H, Si -CH3, multiplet) ( Figure 7-12a). When not enough MeI is used, then the ionic nature does not fully form ( Figure 7-13). 1H NMR, !: 8.15-6.73 (40 H, overlapping multiplets), 3.35 (18 H, N -(CH3)3, multiplet), 2.63 (12 H, N -(CH3)3, multiplet), 0.57 (6 H, CH 3, multiplet), 0.47 (6 H, CH 3, multiplet) ( Figure 7-12b). Scheme 7-3. Synthesis of IL, 6a. MeI, K2CO3THF2 weeks SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiCH3SiH3CNIISiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOASiCH3SiH3CH2NNH2N3a6a 216 Figure 7-12. 1H NMR of (a) 6a and (b) 7a; resonances at 3.72 and 1.83 ppm are residual THF. Figure 7-13. Product without enough MeI. 217 !"#"$%&'((%()*+,-.(+.)/ %In order to confirm that the structures were accurate, the samples were analyzed using mass spectroscopy . Mass spectrometry analyses were conducted using a Waters Xevo G2 -S QToF (Waters, USA) mass spectrometer equipped with Acquity UPLC chromatography system. All mass spectrometric analyses and data processing were accomplished using MassLynx v. 4.1 software. 11 Flow injection analysis without chromatography using electrospray in positive ionization mode was performed to acquire high -resolution mass spectra of compound 6a. Dominant ions are shown at m/z 1336.17, which is consistent with the expected protonated molecular ions, ([M+H +]). Protonated molecular ions are expected since the sample was analyzed under positive ioni zation conditions. These m/z ions are singly charged , and so the m/z value is consistent with the mole cular mass, as the value of z (number of charges) equals 1. Other ions are also seen at lower intensities (m/z 668.59). The fact that the 13C ions (669.09, 669.59, ...) for this ion are 0.5 Da higher on the m/z scale than the 12C ions is an indication that z = 2, and hence the sample ions are doubly charged . So, this compound is generating both singly ([M+H ]+) and doubly ([M+2H]2+) charged ions upon electrospray ionization in positive mode. !"#"0%1.2+345627%-*8'-9(%.2%::;< =5*-6>*5%6.26+%36?465 (% Difunctionalized (doubly charged) ILs based on DDSQ -were successfully synthesized and characterized. Single ionized DDSQ resulted when not enough MeI was added to the reaction. 1H NMR spectra and mass spectroscopy results indicate the desired product was produced and there is minimal byproduct without purification. These DDSQ will benefit from 218 conductivity studies for the potential use as electrolytes in fuel cells. The added feature of the geometrical isomers could potentially provide these DDSQ with disti nct characteristics as IL materials that can be used in specific applications. ! !!! 219 !!!!!"#$%& ! 220 !"#$% & (i) Like compound 3d, compound 5d is actually a mixture of the following compounds: 25 % 5a, 25 % 5c, 50 % 5d as a result of the statistics of the synthetic procedure for compound 3d that is highlighted in chapter 3. (ii ) One singlet represents the cis isomer protons (CH 3, 3H), and the other represents the trans isomer protons (CH 3, 3H). (iii ) There is one s inglet representative of the methyl protons (CH 3) for each of the following geometries: cis meta, trans meta, cis para, trans para, and the reamining two singlets are representative of the cis and trans isomers of the structure that has 1 meta -aminophenyl and 1 para -aminophenyl on a single DDSQ. & 221 !!!!!!!!"#$#"#%&#' ! 222 !"#"!"$%"& ' (1) Seurer, B.; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (2) Vij, V.; Haddad, T. S.; Yandek, G. R.; Ramirez, S. M.; Mabry, J. M. Silicon 2012, 4, 267. (3) Simone, C. D.; Scola, D. A. Macromolecules 2003, 36, 6780. (4) TA Rheology advantage data analysis version 5.8.0; Waters LLC: New Castle, Delaware, 2002. (5) Chen, J.; Zuo, H.; Fan, L.; Yang, S. High Performance Polymers 2009, 21, 187. (6) Zuo, H. J.; Chen, J. S.; Yang, H. X.; Hu, A. J.; Fan, L.; Yang, S. Y. Journal of Applied Polymer Science 2008, 107, 755. (7) Oishi, Y.; Kakimoto, M. A.; Imai, Y. Macromolecules 1988, 21, 547. (8) Garcia, J. M.; Garcia, F. C.; Serna, F.; de la Pena, J. L. Progress in Polymer Science 2010, 35, 623. (9) Cardiano, P.; Lazzara, G.; Manickam, S.; Mineo, P.; Milioto, S.; Lo Schiavo, S. European Journal of Inorganic Chemistry 2012, 5668. (10) Jeon, J. -H.; Tanaka, K.; Chujo, Y. Rsc Advances 2013, 3, 2422. (11) MassLynx version 4.1; Waters LLC: New Castle, Delaware, 2012. 223 !!!!!"##$%&'($) ! 224 !""#$%&' (A)(((*+$,-#*&*(./(0.$.0#1* ( This appendix addresses the following key features th at were not covered in chapter 3 : (1) deprotection analysis and (2) additional features and further remarks on synthesis. 225 A. !"#$%&'(')*+),*#*-&.' )A/0)1&2.*$&3$(*#)4#45 "'(')) Acidified methanol with diethyl ether was used for the deprotection of the silylated amines ( Scheme A-1). A general rule of thumb when considering the compatibility of an acid or base with the silsesquioxane cage is: anythi ng that can etch glass can also break down the cage. Scheme A-1. Deprotection of aminophenyl groups. Thus, a wea k acid (acetic acid) was chosen for the deprotection. Since too much acetic acid may also bre ak down the cage, different solvent qualities and acid concent rations were evaluated to determine the best combination for deprotection (Table A-1). The aminophenyl moiety was completely deprotected for all conditions tested. The concentration s of acetic acid ( HoAc ) and DDSQ vs. the % yield were evaluated for optimal conditions of deprotection (Figure A-1a,b ). Table A-1. Deprotection analysis. XDDSQ XAA XDEE * XMeOH conc . AA conc . DDSQ % yeild 2.08E-03 4.91E-04 9.62E-03 9.88E-01 7.20E-04 5.43E-02 94 2.10E-03 5.01E-04 9.62E-03 9.88E-01 7.34E-04 5.49E-02 94 1.86E-03 8.24E-04 1.91E-02 9.78E-01 1.19E-03 4.80E-02 85 2.11E-03 4.91E-04 9.62E-03 9.88E-01 7.20E-04 5.51E-02 84 1.66E-03 7.19E-04 1.40E-02 9.84E-01 1.05E-03 6.54E-02 81 1.79E-03 6.47E-04 1.28E-02 9.85E-01 9.43E-04 7.07E-02 80 1.86E-03 6.59E-04 1.91E-02 9.78E-01 9.52E-04 4.79E-02 75 Avg. 1.92E-03 ± 1.75E-04 6.19E-04 ± 1.30E-04 1.34E-02 ± 4.23E-03 9.84E-01 ± 4.24E-03 9.01E-04 ± 1.84E-04 5.66E-02 ± 8.53E-03 85 ± 7 * DEE: diethyl ether SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiRSiRNTMS TMS NTMS TMS Acetic AcidDiethyl ether MeOHDDSQ(AP)(Me) 226 Figure A-1. Deprotection analysis, while investigating the concentrations of (a) DDSQ and (b) acetic acid , and (c) the ratio of X DEE : X MeOH vs. % yield. Generally, as the concentration of AA and DDSQ is increased, the % yield is decreased. This was expected since too much AA could potentially break down the silsesquioxane cage. Additionally, the ratio of the mole fraction of diethyl ether and methanol vs. % yield were also investigated ( Figure A-1c). In general, as this ratio increases the % yield decreases. Solid DDSQ is slightly soluble in diethyl ether and not soluble in methanol. Thus, it would be expected that as 75808590950.05 0.06 0.07 Concentration DDSQ (g/mL) % Yield a) 75808590950.00075 0.00100 0.00125 Concentratin Acetic Acid (g/mL) % Yield b)75808590950.010 0.015 0.020 XDEE:XMeOH % Yield c) 227 the solubility of the DDSQ is increased (a larger ratio) the silsesquioxane cage has more opportunity to interact with the AA. However, it is difficult to formulate an explicit conclusion since there are multiple changin g variables, and only one or two are considered in each graphical analysis ( Figure A-1). Furthermore, the scatter in the plots may also be caused by experimental error such as: measurement of masses and solvent volume, the exact v alue of the room temperature, filtration methods, and flask transfer. Determining an exact trend in the data would require additional experimentation where one variable would be adjusted while all others were held constant. However, after this analysis was completed, over 20 deprotection experiments were successful when the average conditions were applied and resulted within the range of the average % yield. Therefore , the average conditions are sufficient for deprotection of the aminophenyl group. A!"#$%%& '&()*+#,-*'./-0#*)%#,./'1-/#/-2*/30#()# 04)'1-0&0 ## The synthetic procedures for compounds 3a and c are well established 1,2 and provide high yields. However, the synthetic procedure for compound 3b provides low yields and additional experimentation was perf ormed in order to determine how to improve the yield. Additionally, the synthetic procedure for compound 3d has only been attempted two times with varying yield results. For compound 3b, the boiling point of the trichlorocyclohex ylsilane (90 ¡C / 10 mmHg) was much higher than methyltrichlorosilane (66 ¡C / 760 mmHg) , which made excess trichlorocyclohexylsilane more difficult to remove . However, this matter was easily resolved by using stoichiometric equivalents of trichlorocyclohexylsilane and 3-[bis(trimet hylsilyl)amino] phenyl -magnesium chloride . Additio nally, the boiling point of the dichlorocyclohexylsilane product (> 135 ¡C) was much higher than the methyldichlorosilane product ( ! 75-95 ¡C) .i Such a high boiling point required the flask to be heated at temperatures 228 above 200 ¡C, which was difficult with standard oil baths. Additionally, any unreacted trichlorocyclohexylsilane was difficult to separate. Therefore, Kugelrohr distill ation was used, which improved the yield and the purity. The synthetic procedure for the reaction between dichlorocyclohexylsilane and Ph8tetrasilanol had the lowest yield. Several parameters were varied in an attempt to improve the yield (Table A-2). Changing the solvent from THF to toluene actually decreased the yield to approximately 5 %. Using additional triethylamines also decreased the yield to approximately 20 %. Increasing the reaction time, increased the less soluble im purity, or the impurity that was Table A-2. Parameters varied in an attempt to improve the yield of the dichlorocyclohexylsilane to Ph 8tetrasilanol coupling reaction (Compound 3b). Parameter varied Notes THF to toluene Decreased yield to 5 % Doubled triethylamines Decreased yield to 20 % Increasing reaction time Increased less soluble impurity Refluxing at T = 105 ¡C Deprotected amines and increased less soluble impurity removed by washing with toluene ( Figure A-2).ii The more soluble impurity was removed by washing with diethyl ether. Refluxing the reaction at 105 ¡C increased the mass of the p roduct, and deprotected the amines, which provided one less step during the work -up procedure. Figure A-2. Compound 3b impurities. 229 However, since the less soluble impurity was also increased, it was difficult to tell whether the increase in the mass was an increase in yield or an increase i n impurities. It was also difficult to tell if the increase in the less soluble impurity was a result of the increased reaction time, or the heat of the reflux. Thus with the increased impurity, refluxing was not a viable method for synthesis. It was als o found that the yield of compound 3b might not actually be as low as reported, and the product is actually stuck in solution. During experimentation, it was recognized that the product was very difficult to precipitate from solution. For several of the pr oducts, solvent washing was attempted. Toluene and other higher boiling solvents were added to the solution, and then removed by vaccum. Afterwards the product was precipitated and isolated by filtration. The filtrate solvent was then removed by vaccum unt il a viscous slurry remained. This slurry was precipitated in anti -solvent and more product was recovered and isolated by filtration. For some experiments, repeating this procedure with the new filtrate solvent again yielded additional product. Furthermore , leaving the filtrate solvent for several days to weeks and then attempting to precipitate also yielded additional product. Thus, a better method to precipitate the product must be devised and it also must be determined if the product is low yielding or s tuck in solution. The synthetic scheme for c ompound 3d was attempted three times . The first time it was very low yielding ( ! 50 %) and had similar impurities to compound 3b. The second time the addition of th e dichlorosilane was slower, the yield was improved to 92 % and there were fewer impurities. iii The third time, the yield was again approximately 50 % initially, but more product was recovered with solvent washing. Additionally, the remaining filtrate solvent after solvent washing shows product in the 1H NMR spectrum. However, this solution has been saved because 230 attempts to recover this product have so far been unsuccessfu l. Again, a better method to precipitate the products, specifically compounds 3b and d, needs to be devised. 231 ! !!!"#$%& ! 232 !"#$% &&(i) This is not measured at a certain pressure, but is based on the Schenck line and vaccuum pump conditions. The reported values are according to the set -up used in this laboratory, but are reported for future work. (ii) Refer to chapter 3 for information on the solvent washing. (iii) Is is of interest to try this method with compound 3b. 233 !!!!!!!!"#$#"#%&#' ! 234 !"#"!"$%"& ' (1) Seurer, B.; Vij, V.; Haddad, T.; Mabry, J. M.; Lee, A. Macromolecules 2010, 43, 9337. (2) Vij, V.; Haddad, T. S.; Yandek, G. R.; Ramirez, S. M.; Mabry, J. M. Silicon 2012, 4, 267. 235 !""#$%&' (B)( *"#+,-.*+."/ ( This appendix addresses the following key features th at were not covered in chapter 4 : (1) chemical shift reagents, (2) T1 analysis, (3) isolating protected amines, (4) analyzing NMR spectra of compound 3d, and (5) quantifying 3d using Liquid -chromatography/mass spectroscopy (LC/MS). 236 B. Spectroscopy B!"#$%&'()*+#, %(-.#/&*0&1., # For integration purposes, it was of interest to resolve the overlapping proton multiplets (!H 7.47: H-3b and H-3b« and !H 7.50: H -1acR and H -1acR«) of compound 3c using tris(6,6,7,7,8,8,8 -heptafluoro -2,2-dimethyl -3,5-octanedionate)europium (III) hydrate, or Eu(FOD) 3, as a shift reagent (Figure B-1). Eu(FOD) 3 complexes w ith basic electrons, such as the aminophenyl group , and shifts associated resonances to a lower field with a small amount of line broadening (Figure B-2). Accordingly , protons nearest the aminophenyl group (H -3 protons) Figure B-1. Europium complex, where B: is the aminophenyl group. were the most affected. As the content of Eu(FOD) 3 was increased, the proton multiplet that was originally at !H 7.47 gradually shifted to downfield, while the proton multiplet originally at !H 7.50 remained virtually unaffected. When the ratio of 3c to Eu(FOD) 3 was 4:1 ( Figure B-2c), the proton multiplet that was originally at !H 7.47 no longer exhibited overlap with other resonances. Multiplets that were originally at !H 7.55 and 7.50 could then be used for quantification, similar to the quantification methods for compounds 3a and b. However, the proton multiplet that was originally at !H 6.53 (H-3a and H -3a«) was also shifted downfield. Initially, this resonance OCF2CF2CF3OEu+3BBEuFODFODFOD+2 BEu(FOD) 33 237 Figure B-2. 1H NMR spectra of increasing content of europium; reported as Compound 3c : Eu(FOD) 3; (a) 1 : 0 , (b) 10 : 1 , (c) 4 : 1, (d) 2 : 1 , (b) 1 : 1. 238 exhibited no signal overlap. Yet, with increasing content of Eu(FOD) 3, it exhibited signal overlap that , again , complicated the capability of determining the percentage of cis isomer present in samples of 3c. When the ratio of 3c to Eu(FOD) 3 was 4:1 ( Figure B-2c), it exhibited signal overlap between !H 7.35 and 7.45, and when the ratio of 3c to Eu(FOD) 3 was 2 :1 ( Figure B-2d), it exhibited signal overlap between 7.59 and 7.71 . Bet ween these ratios , this resonance will exhibit overlap with multiplets used for quantification. However, when the ratio of 3c to Eu(FOD) 3 was 1:1 ( Figure B-2e), it no longer exhibited overlap and shifted downfield to !H 8.53. Ther efore, the best ratios of 3c to Eu(FOD) 3 for quantification purposes are 4:1 and 1:1. If the content of Eu(FOD) 3 is increased past a 1:1 ratio, quantification may still be possible, but line broadening must be considered. Line broadening increases with inc reasing content of Eu(FOD) 3. B!"#$%#&'&()*+*## A T 1 analysis for both the 29Si and 1H atoms was performed on compound 3a. It is anticipated that all other compounds have similar T 1 values. Actual T 1 values for 29Si NMR spectra can be seen in Table B-1. The indexes of interest are 3 -5. Actual T 1 values for 1H NMR Table B-1. 29Si T 1 analysis; peaks can be seen in FigureII -3a. Peak Frequency (ppm) T1 Error 1 -30.6 17.14 0.1597 2 -78.4 59.81 0.932 3 -79.4 50.14 2.469 4 -79.6 54.93 1.273 5 -79.8 54.24 1.931 239 Table B-2. 1H T1 analysis; peaks can be seen in FigureII -3b. Peak Frequency (ppm) T1 Error 1 8.0286 2.401 0.01836 2 8.01453 2.444 0.009801 3 7.95594 2.328 0.05108 4 7.94188 2.358 0.04495 5 7.85516 3.515 0.05858 6 7.83992 3.314 0.06761 7 7.82469 3.26 0.03582 8 7.81766 3.172 0.06714 9 7.80242 3.202 0.115 10 7.78133 3.343 0.1491 11 7.76609 3.117 0.03866 12 7.76375 3.02 0.05968 13 7.75672 2.491 0.03449 14 7.74266 2.533 0.04013 15 7.72859 3.083 0.1194 16 7.71336 3.696 0.1343 17 7.70984 3.914 0.1323 18 7.69695 7.818 0.1534 19 7.68875 7.784 0.6443 20 7.68641 6.681 0.8989 21 7.68055 1.766 0.08184 22 7.66648 2.074 0.01171 23 7.65125 2.125 0.01925 24 7.63836 2.027 0.03045 25 7.62312 2.012 0.09292 26 7.57273 2.062 0.02297 27 7.57039 2.063 0.02349 28 7.5575 2.042 0.01133 29 7.54344 2.024 0.02645 30 7.49187 1.924 0.03527 31 7.48484 2.134 0.08385 32 7.47781 1.925 0.01734 33 7.46492 1.966 0.01806 34 7.44968 1.997 0.02583 35 7.40984 2.204 0.02388 36 7.40046 2.228 0.04198 37 7.39461 2.24 0.02957 38 7.27859 2.583 0.03189 39 6.99616 2.508 0.05217 40 6.98913 2.551 0.04157 41 6.98093 2.545 0.04397 240 spectra can be seen in Table B-2. The relaxation delay for the 1H NMR spectra used for quantification was 32 seconds. Therefore, all the T1 values were met. 29Si and 1H NMR spectra are shown again, below ( Figure B-3). Figure B-3. Spect ra of compound 3a (a) 29Si NMR spectrum and (b) 1H NMR spectrum. B!"#$%&'()*+,#-.&)/0)/1#(2*+/%# # The N -silylated amines are less reactive than the deprotected aminophenyl group. Therefore, it was of interest to isolate the silylated amines prior to deprotection and determine if the 1H NMR spectrum could be evaluated to the same degree as the deprotected amines. However, this was not possible since the line broadening of the N -silylated amine s in the 1H NMR spectrum was too wide (Figure B-4a). Additionally, the resonances representing cis and trans isomers were no longer distinct in the 29Si NMR spectrum (Figure B-4b). 7.57.06.56.05.55.04.54.03.53.02.52.01.51.0ppm7.77.57.37.16.96.76.5b) 241 Figure B-4. N-silylated amines (a) 1H NMR spectrum and (b) 29Si NMR spectrum. B!"#$%&'()*%+#,-.#/012348# 257058%9# :9# Compound 3d, was comprised of an AP moiety w ith a combination of meta- and par a-isomers. However , the synthe tic procedure is anticipated to form 25 % compound 3a (all meta) , 25 % compound 3c (all para) , and 50 % compound 3d (one side meta and one side para) , based on statistics. The 1H NMR spectrum of compound 3d appears to be too convolut ed to quantify the AP m oieties, and additionally to quantify the ratio of cis and trans isomers in a sample using the methods from chapter 4 (Figure B-5). The location of the resonances unique to the meta - and para - AP moieties is observed in the 1H NMR spectrum. H owever, it cannot be determi ned whether the meta -AP moiety is attached to a DDSQ with another meta -AP moiety, or a para -AP moiety since there is theoretically 50 % meta - and 50 % para -AP moieties. Furthermore, the 242 location of the meta - or para -resonance does not appear to be affected by the other AP moiety within the same structure. To clarify, a structure that is comprised of both meta- and para -AP appears to have the same 1H NMR chemical shifts for each meta - and para - AP since there are no ad ditional resonances in the 1H NMR spectrum. Figure B-5. 1H NMR spectrum of compound 3d. 243 The 29Si NMR spectrum of compound 3d was also examined in more detail (Figure B-6). From t his spectrum, it is apparent that compound 3d was produced . There are four resonances within the range unique to D -group silicon atoms , !: -29.96, -29.99, -30.58, and -30.61 (Figure B-6a). According to the 29Si NMR spectra in chap ter 4 , there is only one resonance representative for the D -group silicon atoms for compound 3a (!: -30.57) and one resonance for compound 3c (!: -29.73). For the 29Si NMR spectrum of compound 3d, there are two resonances in each of these areas. This indic ates that there are two distinct D -group silicon atoms for each para - and meta - AP moiety in this range . Therefore, the D -group silicon atoms that have meta -moieties on both sides of the DDSQ ( 3a) are represented by one resonance (2 Si) , and the D -group si licon atoms that have para -moieties on both sides of the DDSQ ( 3c) are represented by one resonance (2 Si) . The D -group silic on atoms that have one para - and one meta -moiety on each side of the DDSQ (3d) are represented by two distinct resonances; o ne for the D -group silicon atom attached to the para -moiety and one for the D -group silicon atom attached to the meta-moiety. Each of these resonances would be equivalent to 1 Si, which is half as many as the 29Si NMR resonances for compounds 3a and 3c in this re gion. However, since there is twice as much of compound 3d statistically, all four resonances are anticipated to integrate to the same relati ve value. From the 29Si NMR spectrum, it is apparent that all four regions are near equivalent, thus it can be said that compound 3d was produced. The region of the 29Si NMR spectrum that repre sents the T -group silicon atoms appears to be more convoluted (Figure B-6b). There are two resonances representative of the external T-group silicon atoms (!: -78.38 and -78.49). For compounds 3a and 3c, there is only one 244 Figure B-6. 29Si NMR spectrum of compound 3d; (a) D -group silicon atoms, (b) T -group silicon atoms, and (c) whole spectrum. 245 resonance in t his region for each compound ( !: -78.38 and -78.24, respectively). Therefore, there are two feasible scenarios for the reson ances in this region of compound 3d. The first being that one resonances is indicative of all external T-group silicon that are near a meta -moiety and the other resonances is indicative of all external T-group silicon that are near a para -moiety. It is equally possible that one of these resonances is indicative of the external T -group silicon of both compounds 3a and 3c, while the othe r resonance is indicative of the external T -group silicon of compound 3d. The same argument can be made for the internal T -group silicon atoms representing the trans -isomers, !: -79.59 and -79.61 (Figure B-6b). From the 29Si NMR resonances representative of the internal T -group silicon atoms, it is apparent that the silicon atoms that are further from the para - or meta - moieties exhibit less deconvolut ion for these resonances (they are closer together) . Moreover, there is little to none deconvolution of the resonances associated with the cis-isomers ( !: -79.39 and -79.79). The quantity and correlation of these resonances with the 1H NMR spectrum has no t been determined. However, it is apparent from the convolution of the 1H NMR, that this task would be very difficult. Additionally, no new resonances were easily observed in the 1H NMR spectrum, this makes correlating what may be Si-atoms associated with 3d particularly challenging . B!"#$%&'()*+)',#-.#%/)',#0)1%). 234567&(6,5&84+97&//#/8:3(56/368+#;0<9=>? #Since, the 1H and 29Si NMR spectra of compound 3d appear to be too convoluted to determine the exact ratio of regio - and geometrical isomers in the sample , LCMS was used. Solvents used for liquid chromatography -mass spectroscopy were aqueous ammonium acetate and acetonitrile, obtained as HPLC -grade from Sigma Aldrich and used as received . Mass spectrometry analyses were conducted using a Waters Xevo G2 -S QT oF (Waters, USA) mass 246 spectrometer equipped with Acquity UPLC chromatography system. All mass spectrometric analyses and data processing were accomplished using MassLynx v. 4.1 software. 1 LC/MS analysis was used to characterize isomers. Analytes were sepa rated using an Ascentis Express C18 column (5 cm !2.1 mm; 2.7 µm particles) with a reversed phase binary gradient. Solvent A was 10 mM aqueous ammonium acetate and solvent B was acetonitrile. Total solvent flow was maintained at 0. 25 mL/min, and gradient el ution was performed using the following solvent compositions: initial, 1 % B, held for 1 min; linear gradient to 99 % B at 15 min and hold at this composition until 17 min; sudden dec rease to 1 % B at 17.01 min and hold at this composition until 20 min. In jection volume and column temperature were 5 µL and 40 ¡C . Four pea ks are seen in the m/z spectrum for compound 3d; e xtracted in chromatograms for ions with m/z 1336.17 shows four peaks at 15.96, 16 .12, 16.22 and 16.39 . There are two peaks for each the all meta- (3a) and all para - (3c) structures ( Figure B-7). Each of the two peaks in the spectra for compounds 3a and 3c are anticip ated to be representative of individual cis and trans isomers. Therefore, six peaks would be anticipat ed for the spectra of compound 3d, but only four are visualized. The peak at 15.96 is also apparent in the spectrum of compound 3c. Additionally, peak s at 16.22 and 16.39 are apparent in the compound 3a spectrum, and also overlap the peak on the right hand side of the compound 3c spectru m. There is one definitive peak (16.12) that is not accounted for in the othe r spectra, which may be indicative of compound 3d, but it is difficult to tell if this is representative of trans - and/ or cis- isomers . Overall, th ese studies address that it is possible to quantify and separate the different regio -isomers that are produced in the synthesis of compound 3d, but additional LCMS studies to further deconvolute the spectrum need to be carried out. 247 Figure B-7. LCMS spectra of compound 3d (top), 3c (middle), and 3a (bottom). 15.0 15.5 16.0 16.5 17.0 17.5 18.0 0500000100000015000002000000250000015.0 15.5 16.0 16.5 17.0 17.5 18.0 0500000100000015000002000000250000015.0 15.5 16.0 16.5 17.0 17.5 18.0 05000001000000150000020000002500000Relative intensity compound 3dmixture of para and meta Retention time Relative intensity Retention time compound 3cpara only Relative intensity Retention time compound 3ameta only 248 !!! "#$#"#%&#' ! 249 !"#"!"$%"& ' (1) MassLynx v. 4.1 software, Waters Corporation, 2012, USA 250 APPENDIX C. SOLUBILITY BEHAVIOR This appendix addresses the following key features th at were not covered in chapter 5 : (1) total mass balance of each isomer for the solubility studies, (2) RF values associated with chromatography, and (3) the computational program that was designed for solubility modeling . 251 C. Solubility behavior C.1!"#$%&!'%((!)%&%*+, ! Table C-1 summarizes the results of performing a total mass balance on the cis and trans isomers for the solubility studies. An example of the mass balanc e can be seen in equation C-1: !"#$% !!!!!"!!!!!!!!"!!""!!!!!!"!!""!"!"# !!!!!!!!!!!!!!!!!!!!!!!!!!!!! where yield (% cis) is the recovery of the cis isomer, P1 and P2 are the total masses (mg) of the products associated with ppt1 and ppt2, respectively, and ppt1 is % trans isomers in P1 and ppt2 is the % cis isomer s present in P2. Approximately 96 ± 8 % of the cis isomer is recovered for 3a, 101 ± 10 % for 3b and 86 ± 6 %for 3c. For the trans isomer, approximately 96 ± 1 5 %is recovered for 3a, 79 ± 24 % for 3b and 90 ± 29 % for 3c. Material balances that provide > 100 % yield, are mostly a result of the large % RE associated with that sample , and is majorly a result of the error associated with 1H NMR ( ± 5 % confidence interval) . Specifically, several of the trans isomer samples have > 100 % yield; since the quantity of trans isomer is so small that it is near or below the detection limit of 1H NMR spectroscopy. Additionally, if the material did not fully dry, the mass c ould potentially be higher ; specific examples follow. Compound 3a, experiment number 17 has a yield = 148 % trans . For this value, ppt2 = 85 % cis. When ppt2 = 90 % cis, which is within the 1H NMR confidence interval (± 5 %), then the yield = 110 % trans . Although, 110 % is still high and relatively unrealistic, it shows that this experiment is with in the error associated with 1H NMR. Additionally, several other factors can be used to reinforce that this MB, although > 100 % is within ex perimental error. Fr om chapter 252 5, it was determined that RE trans = 33 % for experiment 17. Most of this error is a direct result of the mass of the SM (20 mg trans ), which is lower than all the other SM for trans . Furthermore, Table C-1. Total mass balance of the cis and trans isomers for compounds 3a, 3b, and 3c. Compound Exp. SMa (mg cis) SMa (mg trans ) ppt1 (% trans ) ppt2 (% cis) P1c (mg) P2d (mg) Yield (% cis) Yield (% trans ) Overall Yield % 3a 1 551 477 84 90 450 512 97 90 94 2 390 338 90 95 279 392 102 80 92 3 237 255 58 86 427 58 97 100 99 4 238 257 59 92 404 72 97 95 96 5 342 368 96 99 369 327 99 97 98 6b 525 520 85 77 423 470 81 90 85 7b 533 575 93 83 496 549 92 97 94 8b 279 425 95 75 350 328 94 97 96 9b 66 337 97 31 163 220 110 92 95 10b 178 738 96 60 545 457 115 108 109 11b 29 417 99 16 298 136 84 98 97 12b 583 505 91 93 429 586 100 86 93 13 505 89 35 86 8 559 96 91 96 14 376 77 20 91 267 157 95 88 94 15 249 39 12 89 5 257 94 74 91 16 879 167 15 85 448 569 98 91 97 17 265 20 7 85 104 151 85 148 89 18 429 76 33 84 20 455 92 105 94 3b 1 219 103 36 84 47 257 113 56 95 2 188 48 31 86 113 117 95 107 97 3 217 51 15 91 58 194 104 51 94 4 142 80 29 73 86 128 109 74 96 5 146 26 20 85 104 46 84 108 87 3c 1 199 133 71 99 161 131 89 87 88 2 200 133 68 95 164 125 86 88 87 3 433 289 78 96 323 322 88 92 89 4b 214 142 82 87 130 193 90 93 91 5b 211 141 88 86 129 193 86 100 91 6b 86 211 92 71 159 86 85 81 82 7b 93 227 92 63 202 105 89 99 96 8b 144 337 93 39 177 286 86 101 96 9 224 25 8 87 138 38 71 64 71 10 322 36 3 88 263 73 99 47 94 11 290 32 7 92 244 27 87 60 84 12 493 55 32 97 268 217 80 168 88 a starting material ; cis:trans ratio , b undersaturated cis, c P1 = product 1 (ppt1) , d P2 = product 2 (ppt2) 253 ppt1 has the lowest % trans of all the other experimental points. The same argument can be used for compound 3b, experiment 5. For compound 3c, experiments 7 -12 a very similar argument can be made. The SM is approximately 90 % cis, which provided the values seen in Table C-1. If the SM were 95 % cis, then the following would be the reported yield % (Table C-2). Thus, overall, the material balances of the cis an trans isomers for compounds 3a-c confirm that the solubility experiments were accurate. Table C-2. The affect of the isomeric ratio in the SM on the MB of the cis and trans isomers for compound 3c. Compound Exp. SMa (mg cis) SMa (mg trans ) ppt1 (% trans ) ppt2 (% cis) P1b (mg) P2c (mg) Yield (% cis) Yield (% trans ) Overall Yield % 3c 9 237 12 8 87 138 38 68 128 71 10 340 18 3 88 263 73 94 93 94 11 306 16 7 92 244 27 82 120 84 12 521 27 32 97 268 217 75 337 88 a starting material in a ratio of cis:trans , b P1 = product 1 (ppt1), c P2 = product 2 (ppt2) C.1!"#$%&'%$()*!+%,$)&!-" +.!The retardation factor (R F) provides information on the ability for the compounds to move through a chromatography column. Experiments to determine the RF for cis and trans isomers of compound s 3a-c are accomplished with TLC plates. Mixtures of cis and trans isomers for each compound were dissolved in dich loromethane and a spot of each solution was place on a TLC plate simil ar to the procedure in chapter 5 . The R F values were determined by the ratio of the distance that th e samp le spot traveled up the TLC plat e relative to the distance the solvent traveled (Equation II -2) and are summarized in Table C-3: !!!!"#$%&'( !!!!!!"#$%& !!"#$%&%' !"#$%&'( !!"#$%&' !!"#$%&%' !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 254 Table C-3. Retardation factor determined from equation C-2, using TLC plates with spots of compounds 3a-c dissolved in dichlorom ethane. Compound RF tran s RF cis RF tran s - RF cis 3a 0.563 0.438 0.125 3b-1 0.867 0.733 0.133 3b-2 0.533 0.667 0.133 3c 0.556 0.389 0.167 A smaller R F value indicates that the material will move slower through the column. However, this can be dependent on the concentration of the sample. Thus, the actual R F value of each isomer is not as important as the difference in the R F values for the cis and trans isomers (RF cis - RF trans ). A larger difference indicates that the isomers exhibit a larger degree of separation. As expected, the difference in the R F valu es for compound 3c is the largest and this compound exhibited the largest partitioning threshold of its isomers in terms of solubility, as seen in chapter 5 . All values of R F that are reported in Table C-3 are for a near equivalent mixture of cis and trans isomers. To see if the ratio of isomers affects the difference in the isomer RF value s, another sample of 3b was used with a slightly higher cis isomer fraction. RF cis - RF trans = 0.133 for both, thus the isomer ratio does not have an affect on the difference in isomer RF value s. C.3!"#$%&'$'()!*#+,$'-. !Matlab1 was used for t he modeling portion of chapter 5 . The functions that were programmed are below. Anything after a Ò%Ó is a comment. function [xc,gc,parms]=ZcD 255 % xc are the calculated mole fractions, gc are the calculated activity coefficients, parms are the calculated binary parameters, and ZcD is the program name xtrans=[x1 x2 x3 x4;]; % values of the experimental data, can be as many rows as needed xcis=[x1 x2 x3 x4;]; ntrans = size(xtrans,1); % size of the vectors, ntrans is the number of rows ncis = size(xcis,1); ntotal = ntrans + ncis; dHv=[27*1472.2352 20*1472.2352 ]; % J/mol Tmv=[265+273.15 250+273.15]; % K temp=298.15; % K R=8.314; % J/mol*K % optimize parameters parms = lsqnonlin(@calcobj,[3.5 -1],[-2 -4],[5.5 5.5]); % calls calcobj function (see below) to optimize parameters % set rang e of n1on2, extend c alcs slightly beyond expt range llim = min(xtrans(:,1)./xtrans(:,2)) - 0.2; 256 ulim = max(xtrans(:,1)./xtrans(:,2)) + 0.5; % set up vector with 11 pts for n2on1 and n4on2 and initialize vectors npts = 11; n1on2v = linspace(llim,ulim,npts); xc = zeros(npts,3); gc = zeros(npts,3); xc = zeros(npts,4); gc = zeros(npts,4); for j = 1:npts xc(j,2) = 1/(1 + n1on2v(j)); % material balances to find new x, g xc(j,1) = 1 - xc(j,2); [xc(j,:) gc(j,:)] = findx4(xc(j,:),j,dHv(1),dHv(2),Tmv(1),Tmv(2),parms); end % for j % now make a comparison plot n1on2ex = xtrans(:,1)./xtrans(:,2); semilogy(n1on2ex,xtrans(:,3),'o'); hold on n1on2ex = xcis(:,1)./xcis(:,2); semilogy(n1on2ex,xcis(:,4),'or'); 257 % plot model semilogy(n1on2v,xc(:,3),'b'); semilogy(n1on2v,xc(:,4),' -r'); xlabel('ratio of solvents, n1/n2') ylabel('mole fraction of POSS at solubility limit') hold off hold off %--------------------------- function obj = calcobj(guessvalues) disp(' ---------------------') guessvalues; obj = zeros(ntotal,1); % parameters to adjust are in guessvalues(1) and guessvalues(2) for k = 1:ntotal k; dH = dHv(1); dH2= dHv(2); Tm = Tmv(1); Tm2= Tmv(2); 258 if k<= ntrans x(k,:) = xtrans(k,:); [xc(k,:) gc(k,:)]= findx3(x(k,:),k,dH, dH2, Tm, Tm2, guessvalues); obj(k) = abs(log(xc(k,3)/x(k,3))); else x(k,:) = xcis(k -ntrans,:); % saturated with cis and trans [xc(k,:) gc(k,:)]= findx4(x(k,:),k, dH, dH2, Tm, Tm2, guessvalues); obj(k) = abs(log(xc(k,3)/x(k,3)))+abs(log(xc(k,4)/x(k,4))); end %if end %for end %function calcobj %-------------------------- function [xcrow gcrow ]= findx3(xrow, k, dH, dH2, Tm, Tm2, guessvalues) % find x based on saturation to fit only x(3) xold = 2; % force a step gcrow(3) = 1; % assume ideal solution gcrow(4) = 1; n1on2 = xrow(1)/xrow(2); n4on2 = xrow(4)/xrow(2); k; xcrow(3) = (exp(( -dH/(R*temp))*(1 -(temp/Tm))))/gcrow(3); 259 xcrow(2) = (1 -xcrow(3))/(1+n1on2+n4on2 ); xcrow(4) = xcrow(2)*n4on2; xcrow(1) = xcrow(2)*n1on2; gcrow = AAANRTL3(xcrow, temp, guessvalues(1),guessvalues(2)); c=0; res=abs((xold - xcrow(3))/(xcrow(3))); while (res > 0.0001) gcrow = AA ANRTL3(xcrow, temp, guessvalues(1),guessvalues(2)); xcrow(3) = (exp(( -dH/(R*temp))*(1 -(temp/Tm))))/gcrow(3); xcrow(2) = (1 -xcrow(3))/(1+n1on2+n4on2); xcrow(4) = xcrow(2)*n4on2; xcrow(1) = xcrow(2)*n1on2; c=c+1; res=abs((xold - xcrow(3))/(xcrow(3))); xold = xcrow(3); end %while end %function findx3 %-------------------------- function [xcrow gcrow ]= findx4(xrow, k, dH, dH2, Tm, Tm2, guessvalues) %function to fit x(3) and x(4) xold = [2 2]; gcrow(3) = 1; 260 gcrow(4) = 1; n1on2 = xrow(1)/xrow(2); k; xcrow(3) = (exp(( -dH/(R*temp))*(1 -(temp/Tm))))/gcrow(3); xcrow(4) = (exp(( -dH2/(R*temp))*(1 -(temp/Tm2))))/gcrow(4); xcrow(2) = (1 - xcrow(3) -xcrow(4))/(1 + n1on2); xcrow(1) = 1 - xcrow(2) - xcrow(3) -xcrow(4); gcrow = AAANRTL3(xcrow, temp, guessvalues(1),guessvalues(2)); c=0; res=abs((xold(1) - xcrow(3))/ (xcrow(3)))+abs((xold(2) - xcrow(4))/(xcrow(4))); while (res > 0.0001) gcrow = AAANRTL3(xcrow, temp, guessvalues(1),guessvalues(2)); xcrow(3) = (exp(( -dH/(R*temp))*(1 -(temp/Tm))))/gcrow(3); xcrow(4) = (exp(( -dH2/(R*temp))*(1 -(temp/Tm2))))/gcrow(4); xcrow(2) = (1 - xcrow(3) -xcrow(4))/(1 + n1on2); xcrow(1) = 1 - xcrow(2) -xcrow(3) -xcrow(4); c=c+1; res=abs((xold(1) - xcrow(3))/(xcrow(3)))+abs((xold(2) - xcrow(4))/(x crow(4))); xold = [xcrow(3) xcrow(4)]; 261 end %while end %function findx4 %-------------------------- end %USElsq function gamma=AAANRTL3(moleFracs, temp, A13, A23) %Row vector [Hexanes, THF, POSS] x=moleFracs; %this function defines the number of components nComp=length(x); %this will create a matrix with all the rows equal to the vector of %compositions Y = kron(x,ones(nComp,1)); %binary parameters for NRTL 262 aij=[0 0 A13 A13; 0 0 A23 A23; A13 A23 0 0 A13 A23 0 0; ]; bij=[0 -15.0959 0 0; 233.6258 0 0 0; 0 0 0 0 0 0 0 0;]; cij=[0 0.3 0. 0; 0.3 0 0. 0; 0. 0. 0 0 0 0 0 0;]; %See Aspen reference for NRTL model %tau is truncated after second term %alpha is truncated after first term tau=aij+bij/temp; alpha=cij; G=exp( -alpha.*tau); term1=x*(tau.*G); 263 term2=x*G; inverseterm2 = (1./term2); squareinverseterm2=inverseterm2.^2; part1=(term1./term2); %part2=x*(tau.*G)'./term2 part2=inverseterm2*((Y.*tau.*G)'); %part3= -((x.*term1)*G')./(term2.^2); part3=(term1.*squareinverseterm2)*((Y.*G)'); %logGamma=(term1./term2) + (x*(tau.*G)')./term2 - ((x.*term1)*G')./(term2.^2); logGamma = part1 + part2 - part3; gamma=exp(logGamma); gamma(isnan(gamma))=1; end 264 !!!!!!!!!!!"#$#"#%&#' ! 265 !"#"!"$%"& ' (1) MATLAB version R2011a ; The Mathworks Inc.: Natick, Massachusetts, 2011. 266 !""#$%&' (D)((*+,-.!/(-.+0*.0+#-(!$%("1!-#(%&!2+!3- ( This appendix addresses the following key features th at were not covered in chapter 6 : (1) the tables and figures associated with the crystal structures and (2) the computation program used for modeling. ( 267 D. !"#$%&'($%")*%)"+$(&,-(./&$+(-0&1"&2$ (D34(!"#$%&'($%")*%)"+$ (Crystallographic data for all compounds are reported in Table D-1 through Table D-3. All structures are shown in Figures 6 -8. Phenyl moieties and hydrogen atoms are not shown for simplicity. Table D-1. Crystallographic data of compounds 3c. Compound trans -3c cis-3c Empirical formula C62 H58 N2 O14 Si10 C62 H58 N2 O14 Si10 FW (g/mol) 1336 1336 Crystal system Monoclinic Triclinic Space group P 2(1) / n P-1 a (†) 10.0493(9) 13.7676(14) b (†) 43.567(4) 17.5029(18) c (†) 14.5921(13) 27.499(3) a (¡) 90 87.1100(10) b (¡) 91.6600(10) 79.161(2) g (¡) 90 87.3850(10) V († 3) 6386 6495.6 Z 4 4 !calcd (Mg/m 3) 1.39 1.366 µ (mm -1) 0.272 0.267 F (0 0 0) 2784 2784 Crystal size (mm 3) 0.72 x 0.13 x 0.10 0.39 x 0.15 x 0.15 " range (¡) 1.47 - 25.42¡ 1.55 - 25.46¡ Reflections collected 52876 23931 Independent reflections 11749 23931 Data / restr. / param. 11749 / 0 / 795 23931 / 0 / 1585 R(int) 0.0725 0 Final R 1 [I > 2 # (I)] 0.0495 0.0654 Final wR (F 2) [I > 2 # (I)] 0.106 0.1454 Final R 1 (all data) 0.0931 0.1264 Final wR (F 2) (all data) 0.1278 0.1655 Completeness to: 25.42¡ = 99.7% 25.46¡ = 99.3% G-O-F on F 2 1.021 1.063 Lg. diff. peak/hole (e / † 3) 0.9741 and 0.8272 0.9610 and 0.9029 268 Table D-2. Crystallographic data of compounds 3a. Compound trans -3a Polymorph A trans -3a Polymorph B cis-3a Empirical formula C62 H58 N2 O14 Si10 C62 H58 N2 O14 Si10 C62 H58 N2 O14 Si10 FW (g/mol) 1336 1336 1336 Crystal system Triclinic Triclinic Monoclinic Space group P -1 P -1 C 2 / c a (†) 10.4549(6) 9.911(2) 20.8538(9) b (†) 12.7243(7) 13.532(3) 17.7674(7) c (†) 13.2241(8) 14.086(3) 17.6323(8) a (¡) 100.5230(10) 65.100(2) 90 b (¡) 90.8660(10) 72.134(2) 95.9742(3) g (¡) 112.7660(10) 70.678(2) 90 V († 3) 1587.72(16) 1585.8(6) 6500.3(5) Z 1 1 4 !calcd (Mg/m 3) 1.397 1.399 1.365 µ (mm -1) 0.274 0.274 0.267 F (0 0 0) 696 696 2784 Crystal size (mm 3) 0.44 x 0.28 x 0.25 0.24 x 0.04 x 0.04 0.42 x 0.41 x 0.28 " range (¡) 1.77 Ð 25.28 1.63 Ð 25.32¡ 1.51 Ð 25.42¡ Reflections collected 17266 18911 28342 Independent reflections 5763 5756 5974 Data / restr. / param. 5763 / 0 / 398 5756 / 0 / 398 5974 / 0 / 399 R(int) 0.0223 0.1235 0.035 Final R 1 [I > 2 # (I)] 0.0316 0.0729 0.0503 Final wR (F 2) [I > 2 # (I)] 0.0839 0.1787 0.1311 Final R 1 (all data) 0.036 0.1475 0.0653 Final wR (F 2) (all data) 0.0874 0.2145 0.1452 Completeness to: 25.28 ¡ = 99.8 % 25.32 ¡ = 99.7 % 25.42 ¡ = 99.4 % G-O-F on F 2 1.049 1.034 1.047 Lg. diff. peak/hole (e / † 3) 0.397 and -0.317 0.656 and -0.513 0.499 and -0.437 269 Table D-3. Crystallographic data of compounds 3b. Compound trans -3b Polymorph A trans -3b Polymorph B trans -3b Polymorph C Empirical formula C72 H74 N2 O14 Si10 C72 H74 N2 O14 Si10 C72 H74 N2 O14 Si10 FW (g/mol) 1472 1472 1472 Crystal system Monoclinic Triclinic Triclinic Space group P 2(1) / c P-1 P-1 a (†) 13.8808(12) 11.4196(7) 10.9653(9) b (†) 20.5920(18) 12.2071(8) 12.9878(11) c (†) 14.3386(13) 14.5294(9) 14.4400(12) a (¡) 90 70.9960(10) 67.8970(10) b (¡) 117.9000(10) 76.2550(10) 81.1680(10) g (¡) 90 74.0800(10) 74.6940(10) V († 3) 3622.07 1816.91 1834.37 Z 2 1 1 !calcd (Mg/m 3) 1.35 1.346 1.33 µ (mm -1) 0.247 0.246 0.244 F (0 0 0) 1544 772 772 Crystal size (mm 3) 0.20 x 0.17 x 0.15 0.29 x 0.20 x 0.14 0.34 x 0.33 x 0.26 " range (¡) 1.66 - 25.33¡ 1.81 - 25.39¡ 1.74 - 25.41¡ Reflections collected 29155 25277 30213 Independent reflections 6617 6689 6767 Data / restr. / param. 6617 / 0 / 442 6689 / 1018 / 612 6767 / 990 / 597 R(int) 0.0465 0.0405 0.0269 Final R 1 [I > 2 # (I)] 0.051 0.084 0.0529 Final wR (F 2) [I > 2 # (I)] 0.1313 0.2578 0.1643 Final R 1 (all data) 0.0729 0.1180 0.0612 Final wR (F 2) (all data) 0.1443 0.2943 0.1772 Completeness to: 25.33¡ = 99.8 % 25.39¡ = 99.9 % 25.41¡ = 99.8 % G-O-F on F 2 1.03 1.038 0.87 Lg. diff. peak/hole (e / † 3) 0.582 and -0.406 0.873 and -0.538 1.210 and -0.568 270 Figure D-1. Crystal structures for compound 3c. 271 Figure D-2. Crystal structures for compound 3a. 272 Figure D-3. Crystal structures for compound 3b. 273 D!"#$%&'(#)*&+,&-#-.)(/*0+ # In order to fit the experimental data, a Matlab program, which was very similar to the one in the previous appendix was written and is shown below. function [xcurve,gcurve,xcurve2, gcurve2, p arms,temps,temps2]=tc1c5L % This file has cis also. %% global ntotal dH dH2 Tm Tm2 x1 x2 temp R xc gc %% These are the values of the mole fraction and the temperature. db = load('fix1c.mat'); x1 = db.fixx(:,1); temp = db.fixx(:,2) + 273.15; %K x2 = 1-x1; ntotal = size(x1,1); dHv=[55000 38000]; %J/mol Tmv=[310.6+273.15 275.7+273.15]; % K R=8.314; %J/mol*K dH = dHv(1); 274 dH2 = dHv(2); Tm = Tmv(1); Tm2 = Tmv(2); options = optimset('Display','iter'); % The parms are the initial guesses fo r the NRTLDSC file (A12 and A21). parms = lsqnonlin(@calcobjtc1c5,[ -0.03 3.35 100 100], [], [],options); xcOUT = xc; gcOUT = gc; % Corresponding temperatures for trans. llim = 150; %C ulim = 310.6; %C % % % New temperature values temps = linspace(llim+273.15,ulim+273.15,100); npts = length(temps); % for j = 1:npts [xcurve(j,:),gcurve(j,:)] = findx123(temps(j),dH, Tm, parms); end % for j 275 % Corresponding temperatures for cis. llim2 = 150; %C ulim2 = 275.7; %C % New temperature values temps2 = linspace(llim2+273.15,ulim2+273.15,100); npts2 = length(temps2); for j = 1:npts2 [xcurve2(j,:),gcurve2(j,:)] = findx223(temps2(j),dH2, Tm2, parms); end % for j figure(1) %plot all experimental values ph(1) = plot(x1(:,1),temp,'bo'); hold on; %plot values from findx1 function, using trans dH and Tm ph(2) = plot(xcurve(:,1),temps,'r -'); hold on; %plot values from findx2 function, using cis dH and Tm ph(3) = plot((1 -xcurve2(:,1)),temps2,'r -'); hold off; set(gcf,'Color',[1 1 1],'Position',[1000 600 1000 700],'InvertHardcopy','off'); 276 set(gca, ... 'Color' , [0.8 0.8 0.8], ... 'FontName' , 'Times New Roman' , ... 'FontSize' , 20 , ... 'FontWeight' , 'bold' , ... 'TickDir ' , 'in' , ... 'TickLength' , [.02 .02] , ... 'XMinorTick' , 'on' , ... 'YMinorTick' , 'on' , ... 'XGrid' , 'on' , ... 'YGrid' , 'on' , ... 'LineWidth' , 1 ); ylabel('Temp','Interpreter','la tex'); xlabel('XcNEW','Interpreter','latex'); legend(ph,{'xcNEW','x1'}); end function obj = calcobjtc1c5(guessvalues) global dH dH2 Tm Tm2 ntotal x1 x2 xc gc temp %disp(' ---------------------'); 277 obj = zeros(ntotal,1); xc = zeros(ntotal,2); gc = zeros(ntotal,2); for k = 1:ntotal if k < 7 [xc(k,:), gc(k,:) ]= findx123(temp(k),dH, Tm, guessvalues); obj(k) = xc(k,1) - x1(k); else [xc(k,:), gc(k,:) ]= findx223(temp( k),dH2, Tm2, guessvalues); obj(k) = xc(k,2) - x2(k); end end end function [xcrow, gcrow ]= findx123(temp, dH, Tm, guessvalues) global R activity = (exp(( -dH/(R*temp))*(1 -(temp/Tm)))); xcrow = fzero(@iterx,[1 0]); 278 %xcrow(2) = 1 -xcrow(1); function obj = iterx(x1c) gcrow = NRTLDSCa([x1c 1 -x1c], temp, guessvalues(1),guessvalues(2),guessvalues(3),guessvalues(4)); obj = x1c*gcrow(1) - activity; end end %function findx1 function [xcrow, gcrow ]= findx223(temp, dH2, Tm2, guessvalues) global R activity2 = (exp(( -dH2/(R*temp))*(1 -(temp/Tm2)))); xcrow = fzero(@iterx,[0 1]); function obj = iterx(x2c) gcrow = NRTLDSCa([1 -x2c x2c], temp, guessvalues(1),guessvalu es(2),guessvalues(3),guessvalues(4)); obj = x2c*gcrow(2) - activity2; end end %function findx2 279 function gamma=NRTLDSCa(moleFracs, temp, A12, A21,B12,B21) %Row vector [Hexanes, THF, POSS] x=moleFracs; %this function defines the number o f components nComp=length(x); %this will create a matrix with all the rows equal to the vector of %compositions Y = kron(x,ones(nComp,1)); %binary parameters for NRTL aij=[0 A12 ; A21 0 ; ]; bij=[0 B12 ; B21 0 ;]; cij=[0 0.3 ; 280 0.3 0 ;]; %See Aspen reference for NRTL model %tau is truncated after second term %alpha is truncated after first term % I took bij out because of the varying temp. I will need to fix this lat er tau=aij+bij/temp; alpha=cij; G=exp( -alpha.*tau); term1=x*(tau.*G); term2=x*G; inverseterm2 = (1./term2); squareinverseterm2=inverseterm2.^2; part1=(term1./term2); %part2=x*(tau.*G)'./term2 part2=inverseterm2*((Y.*tau.*G)'); %part3= -((x.*term1)*G')./(term2.^2); part3=(term1.*squareinverseterm2)*((Y.*G)'); 281 %logGamma=(term1./term2) + (x*(tau.*G)')./term2 - ((x.*term1)*G')./(term2.^2); logGamma = part1 + part2 - part3; gamma=exp(logGamma); gamma(isnan(gamma))=1; end 282 !!!!!!!!"#$#"#%&#' ! 283 !"#"!"$%"& ' (1) MATLAB version R2011a; The Mathworks Inc.: Natick, Massachusetts, 2011. 284 !""#$%&' (E)(((!""*&+!,&-$. ( This appendix addresses the following key features regarding the following th at were not covered in chapter 7 : (1) polyimide thermosets a nd (2) aromatic polyamides. 285 E. !""#$%&'$()* +E,-+.(#/$0$12+'3240(*2' +E!"!"#$%&'()*+* #,-#./#Cis and trans [(meta- phenylethynylphthalimide )cyclohexyl silyl] -bridged -(phenyl )8-double -decker silsesquioxane , DDSQ(m-PEPI)(Cy)-5b, was also synthesized, but due to its highly crystalline nature, optimization of the product was not pursued. Under a nitrogen atmosphere, a well -stirred solution of Ph 8bisaniline -POSS ( 0.164 g, 0.093 mmol ) and PEPA (0.051 g, 0.19 mmol) in anhydrous THF (2 mL) and anhydrous toluene (2 mL) was stirred at room temp erature (25 ¡C) for one hour. After which, it was heated to 60 ¡C 2 h and then to 115 ¡C and refluxed for 20 h. Solvent was removed under vacuum until a white precipitate appeared. Methanol was added until no more precipitation occurred, and the product was filtered, and dried in a vacuum oven at 125 ¡C for 24 h . 1H NMR, !: 8.1 -7.1 (64 HÕs, overlapping multiplets), 1.28-0.87 (10 HÕs) . E!"!0#1)&)234#*%&'()*+* #It was determined that using a large excess of PEPA (> 100 % excess) in the reaction provided an optimized yield. When not enough excess is used, it is most not iceable in the 29Si NMR spectrum. Figure E-1 depicts a trans 5c with additional resonances in the region between - 78 and - 80 ppm. It is anticipated that the most common impurity would be product that is not fully immidized ( Figure E-2). During the reaction, THF and Toluene form an azeotrope with the water that is produced thr ough the condensation reaction, allowing the i midization to occur and pushing the reaction towards completion. However, as depicted in Figure E-1, refluxing was not 286 Figure E-1. 29Si NMR spectrum of trans 5c that is not a pure product. enough to drive the product to completion. To ensure the imidization does o ccur, molecular sieves were added to the reaction flask, the yield improved, and 29Si spectra did not contain additional resonances. When the reaction is driven to completion, the yield is generally > 80 %. Figure E-2. Anticipated impurity in the synthesis of 5. E!"!#$%&'()*+,(+&-$+.$/0 $ The 29Si NMR spectrum of compound 5d ( Figure E-3) was distinctly different than the spectrum of compound 3d. The D -group silicon region that previously showed four resonances in the spectrum of compound 3d only shows two resonances in the s pectrum of compound 5d. Additionally, there are now four resonances in the region representing the external T -group silicon, wh ile there were previously only two . The argument that was made for the four 29Si SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNHPhOOPEPIHO 287 NMR resonances in the region of the D -group sili con in appendix B can also be made for the external T -group silicon here. The most noticeable difference is in the region of the internal T -group silicon, which is representative of the cis and trans isomers. For compounds 5a and 5c, these resonances appea r in the order: cis trans cis . For compound 5a, the distance from cis to cis (peak to peak) is approximately 0.2 ppm, and for compound 5c it is 0.4 ppm. For the 29Si NMR spectrum of compound 5d, there are nine total resonances in this region ( !: -78.80, -78.87, -78.93, -78.98, -79.02, -79.07, -79.14, -79.17, -79.19). The resonance at -79.02 may actually be two overlapping resonances (which would make 10 resonances) , similar to the two overlapping resonances representative of the trans isomer in the 29Si N MR spectrum of compound 3d. Three regions are circled on the spectrum in Figure E-3. Each of these regions appears to have the cis trans cis pattern, where the summati on of the integration for both cis resonances is equivalent to the integration for one trans resonance . If this were accurate, then each region would be representative for 5a, 5c, or 5d. However, according to the cis to cis distance that was mentioned previously, this is unlikely. Figure E-3 also shows the typical spacing of compounds 5a and 5c, on the spectrum for compound 5d. Clearly, these can be the resonances depicting the all meta - and all para - DDSQ, and the remaining resonances (!: -78.87, -78.98, -79.07, -79.17) could be representative of compound 5d. The resonances in regions 1 and 3 may represent cis isomers, which suggests tha t the shifts at !: -78.87 and -79.17 are associated with cis 5d. There is twice as much 5d as 5a and 5c due to the statistics of the synthetic procedure, and according to the 29Si NMR, these resonances appear to fit that ratio. The resonances in region 2 ( !: -78.98 and -79.07) represent trans 5d, and it is anticipated that each resonance is representative of a specific regio - environment; silicon atoms near the meta - and para -moieties. 288 Figure E-3. 29Si NMR s pectrum of compound 5d. 289 E!"#$%&'()(*+,-#./-)*%0&(1.+2 #Before the polyaramid synthesis, a model reaction with N-(trimethylsilyl)aniline with benzoyl chloride was investigated ( Scheme E-1).1 The silylated am ine had to be synthesized first Scheme E-1. Model reaction for Nomex. E!"!#$%&'(')*+,$)-&'&-+ $Under a nitrogen atmosphere, chlorotrimethylsilane (14.30 g, 0.132 mole) was added dropwise to a stirring solution of aniline (10.22 g, 0.1097 mole) and triethylamin es (33.30 g, 0.329 mmole) at 60 ¡C. After 20 hours, solvent was removed by vacuum and the product was isolated by fractional distillation (T BP = 40 Ð 49 ¡C, P = 435 Ð 444 mTorr) (14.0 g, 0.0847 mole, 77 % yield). 1H NMR !: 7.28 (2 H, t, J = 7.9 Hz ), 6.85 (1 H, t, J = 7.4 Hz), 6.79 (2 H, d, J = 8.5 Hz), 3.54 (1H, NH, broad singlet) , 0.41 (9 H, singlet ) (Figure E-4). Figure E-4. 1H NMR spectrum of silylated aniline. ClONHSiMe 3+silylated aniline Benzoyl chloride NOHHexanes NH2TMSCl 290 E!"!"#$%&'(#)%*'+ #Under a nitrogen atmosphere a solution of benzoyl chloride (1.7 g, 0.01 2 mmole), was added dropwise to a stirring solution of silyl ated amines (2 g, 0.012 mmole) and hexanes (25 mL) at 25 ¡C. The product formed a white precipitate and was washed with hexanes and isolated by filtration ( 1.98 g, 0.011 mmole, 90 % yield ). The IR spectrum showed characteristic absorptions at 3340 (NH) and 1655 (C=O) cm -1. 291 !"#"!"$%"& 292 !"#"!"$%"& ''(1) Oishi, Y.; Kakimoto, M. A.; Imai, Y. Macromolecules 1988, 21, 547. 293 !""#$%&' (F)((*+,+-#(./-0 ( This appendix addresses ideas for future work in this area. ( 294 F. !"#"$%&' ($) &F*+&,%-.$.#%&/%#.0-.$.&/12#"$%3 & One area of research that can easily be expanded upon would be evaluating the synthesis for compound 3d such that the all meta -structures could be separat ed from the all para -structures and the one side meta - one side para -structure as shown again in Figure F-1. This c ould potentially be accomplished using further 2D NMR spectroscopy studies, or the LCMS analysis provided in appendix II. Figure F-1. Compound (a) 3a-all meta -structure, (b) 3c-all para -structure, and (c) 3d-one side meta- and one side para -structure. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2NH2Ncis 3acis 3cSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CNH2H2Ntrans 3ca)b)trans 3acis 3dtrans 3dc)SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOMeSiSiH2NNH2SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CSiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiH3CH2NH2NH3CH2NNH2NH2H2N 295 F!"#$%&'()*+#,#-.'-+/0&(1) # Substit uting the X -moiety for a propyl amine as opposed to an aminophenyl ( Figure F-2) moie ty may provide evidence of why compound 3b, with a cyclohexyl moiety, is an isomorphous solution. It has been suggested that since the cyclohexyl moiety is similar in size and shape t o the aminophenyl moiety , the cis and trans isomers are not as different, thus providing similar crystal str uctures. Since a linear, propyl amine, is more similar to the methyl moiety, this can be used to evaluate this theory. Additionally, determining the crystal structure for cis 3c is also relevant to this future work section. Figure F-2. DDSQ(propylamine )(Me). F!2#344(*('10/#0--/(50*('16 # Developing additional applications, or expanding upon the ones developed in this dissertation, that incorporate the distinctive features of the cis and trans isomers is central to the future of this research . Since a detailed analysis of the chemical and physical features of these cis and trans isomers has been established , the next step would be to apply this knowledge to the development of novel materials whose properties can be adjusted according to the quantity of cis and trans isomers used. SiOSiOSiOSiOSiOSiOSiOSiOPhPhPhPhPhPhPhPhOOOOOOSiMeSiMeNH2H2N