UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

NMR investigations of cyanate resins Niu, Junning 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_phd_niu_junning.pdf [ 8.45MB ]
Metadata
JSON: 831-1.0061801.json
JSON-LD: 831-1.0061801-ld.json
RDF/XML (Pretty): 831-1.0061801-rdf.xml
RDF/JSON: 831-1.0061801-rdf.json
Turtle: 831-1.0061801-turtle.txt
N-Triples: 831-1.0061801-rdf-ntriples.txt
Original Record: 831-1.0061801-source.json
Full Text
831-1.0061801-fulltext.txt
Citation
831-1.0061801.ris

Full Text

NMR INVESTIGATIONS OF CYANATE RESINS by JUNNING NIU B.Sc., Peking University, 1982 M.Sc., Nanjing University, 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (DEPARTMENT OF CHEMISTRY)  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA July 1993 © Junning Niu, 1993  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  L /Le nfri_s-trj,-  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  "M  ABSTRACT Heterogeneous polymer products are very complex systems. In many cases, the properties of these polymer systems are determined by the nature and the degree of reaction of a small number of functional groups in the system. However, it is very difficult to study them by most analytical techniques due to the influence of the polymer bulk and the irregularity and insolubility of the final product. In this thesis, both high resolution solution and solid-state NMR spectroscopies combined with specific isotopic enrichments are used to study such kinds of heterogeneous polymers with a particular focus on cyanate resin polymer systems, which are newly-developed thermoset resins used for electronic circuit boards and many structural composites. The mechanism of the curing reactions of cyanate resins based on bisphenol-A dicyanate has been investigated both in solution and in the solid state by NMR spectroscopy. To facilitate the study, 13C and 15N isotopic labelled cyanate resins were used. The main curing reaction was found to be the formation of triazine rings and no NMR evidence for the formation of dimeric or other intermediate species prior to triazine ring formation was found. Side products are found in the solution curing due to the reaction of the cyanate group with trace water present in the solvent. In the bulk curing, the reaction is remarkably efficient, and no detectable side reaction occurs. This can be rationalized in terms of the very strong intermolecular interactions between cyanate groups on different molecules, which is observed in the crystal structure of the bisphenol-A dicyanate monomer obtained from a single-crystal X-ray diffraction experiment. The possible cross reactions indicated by solid-state NMR spectroscopy between cyanate and epoxy resins have been investigated by using both natural abundance and 13C and 15N labelled monofunctional model compounds. These 11  soluble products were isolated and purified by adsorption chromatography and gel permeation chromatography, and were fully characterized by high resolution 1H, 13C, 15N NMR spectroscopies and by mass spectrometry. The major crossreaction product is a mixture of enantiomers which contain an oxazolidinone ring formed by one cyanate molecule and two epoxy molecules. However, triazine formation from the cyanate is much faster than the two competing reactions (the cross reaction between cyanate and epoxy and the selfpolymerization of epoxy) under the conditions investigated. In addition to the cross reactions of epoxy and cyanate, the reactions of epoxy with carbamate which is the major side product for the solution curing of cyanate resin have also been investigated, and several products related to the cross reaction have been isolated and identified. It is suggested that the reaction of epoxy and carbamate is one of the pathways in the overall cross reaction between epoxy and cyanate resins. A desired cross-linking monomer for mixed cyanate and epoxy resin systems, the monoglycidyl ether of bisphenol-A-monocyanate, has been synthesized and characterized. The cyanate group in the cross-linking monomer is more reactive than the epoxy group and can be cured independently under heat or by base. A more practical approach for the application of the crosslinking monomer is discussed and tested. A very tough and strong resin material was obtained using this approach. A bifunctional cross-linking monomer, 2-allylphenyl cyanate, for the cyanate resin (thermoset) and olefinic polymers (thermoplastic) has also been synthesized and characterized. As a cross-linking agent, it not only reacts with itself, but also reacts with other cyanates to form hetero-triazine structures. It can also be copolymerized with the olefinic monomer, methyl methacrylate, to form a cross-linked polymer. 111  TABLE OF CONTENTS Page ABSTRACT^ TABLE OF CONTENTS^  iv  LIST OF TABLES ^  x  LIST OF FIGURES ^  xi  ABBREVIATIONS ^  xx  ACKNOWLEDGEMENTS ^  x)di  1. INTRODUCTION ^  1  1.1. Polymer Types ^  1  1.2. Cyanate Resins ^  3  1.3. Epoxy Resins ^  9  1.4. Coupling Agents and Glass Reinforced Resins ^  15  1.5. Polymer Resin Characterizations ^  19  1.6. High Resolution Solid State NMR Spectroscopy of Polymers ^ 22 1.6.1. Principles of Pulsed FT—NMR ^  23  1.6.2. Dipolar Coupling in Solids and High Power Proton Decoupling ^ 27 1.6.3. Chemical Shift Anisotropy and Magic Angle Spinning ^ 32 1.6.4. Relaxation Times ^  37  1.6.5. Cross Polarization ^  40  1.7. Purpose of the Thesis Research  ^46  2. INVESTIGATIONS OF THE CURING REACTIONS OF THE CYANATE RESIN SYSTEM^  50 iv  2.1. Synthesis and Characterization of Specifically Labelled Cyanate Monomers ^  51  2.2. Investigation of the Curing Reaction in Solution ^ 56 2.2.1. 13C NMR Investigation of the Curing Reaction of BPADCN in Solution ^  56  2.2.2. '5N NMR Investigation of the Curing Reaction of BPADCN in Solution ^  59  2.3. Reactions of Monocyanate Model Compounds ^  61  2.3.1. 13C NMR Investigation of the Curing Reaction of PTBPCN 25a in Solution ^  61  2.3.2. '5N NMR Investigation of the Curing Reaction of PTBPCN 25b in Solution^  62  2.4. Synthesis and Characterization of the Carbamates 28 and 29^ 64 2.5. Isolation and Characterization of the Triazine 31 Obtained from p-tert-Butylphenyl Cyanate ^  67  2.6. The Mechanism of the Curing Reaction for Cyanate Resin in Solution ^  72  2.7. Investigation of the Curing Process in the Solid State^ 73 2.7.1. 13C Solid State NMR Investigations ^  74  (1) BPADCN Monomer^  74  (2) Cured Resin from Solution Polymerization ^ 75 (3) Cured Resin from Solid State Curing ^  78  2.7.2. 15N Solid State NMR Investigations ^  78  (1) BPADCN Monomer^  78  (2) Cured Resin from Solution Polymerization ^ 78 (3) Cured Resin from Solid State Curing ^  79  2.7.3. Quantitative Investigation of Solid State BPADCN Curing ^ 79 V  2.8. The Relation of the Crystal and Molecular Structure of BPADCN to Its Curing Efficiency ^  86  2.9. Conclusions ^  91  3. INVESTIGATIONS OF THE POSSIBLE CROSS REACTIONS BETWEEN CYANATE AND EPDXY RESINS ^  93  3.1. Solid State NMR Investigation of the Neat Curing Reaction of the Mixed Dicyanate / Epoxy Resins ^  94  3.2. Neat Curing Reactions of Monofunctional Cyanate and Epoxy Compounds ^  96  3.2.1. Neat Curing Reactions of Unlabeled PTBPCN 25 and PTBPGE 37^  96  3.2.2. Neat Curing Reactions of Using 15N labelled PTBPCN ^ 102 3.3. Further Characterization of the Major Cross-reaction Product ^ 104 3.3.1. 1D and 2D NMR Spectra ^  104  3.3.2. Model Compounds for Structure 39^  108  3.3.3. Variable Temperature NMR Experiments ^  111  3.3.4. NOE Experiments ^  114  3.3.5. A Possible Mechanism for the Main Cross-linking Reaction ^ 119 3.4. Investigation of the Second Unidentified Product ^ 120 3.5. Reaction of Carbamate and Epoxy ^  123  3.5.1. Reaction of 15N Enriched p-tert-Butylphenyl Carbamate 28 with PTBPGE 37^  123  3.5.2. Reaction of Phenyl Carbamate with Phenyl Glycidyl Ether ^ 126 3.6. A Possible Mechanism for the Reaction Between Epoxy and Carbamate ^  133  vi  3.7. Investigation of Imidocarbonate as a Possible Cross-reaction Product^  136  3.7.1. Reaction of Cyanate with p-tert-Butylphenol (PTBP) ^ 136 3.7.2. Reaction of Cyanate with Isopropanol ^  137  3.8. The Mechanism of the Curing Reaction for Dicyanate / Diepoxy Mixed Resins ^ 3.9. Conclusions ^  139 141  4. SYNTHESIS AND CHARACTERIZATION OF CROSS-LINKING AGENTS FOR MIXED CYANATE / EPDXY RESIN SYSTEMS AND FOR MIXED CYANATE / OLEFIN RESIN SYSTEMS ^ 143 4.1. A Cross-Linking Agent for Mixed Cyanate / Epoxy Resin Systems ^  144  4.1.1. Strategy for the Synthesis of the Cross-Linking Agent, the Monoglycidyl Ether of Bisphenol-A-monocyanate 61^ 145 4.1.2. Synthesis and Purification of the Monoglycidyl Ether of Bisphenol-A 62 ^ 4.1.3. Synthesis of Cross Linking Monomer 61^  147 154  4.1.4. Curing Reaction of Cross-Linking Monomer 61 with Heat ^ 155 4.1.5. Curing Reaction of Cross-Linking Monomer 61 with Base ^ 158 4.1.6. A More Practical Approach to the Application of Cross-Linking Monomer 61^ 4.1.7. Conclusions ^  163 166  4.2. A Cross-Linking Agent for Mixed Cyanate / Olefin Resin Systems ^  167  4.2.1. Synthesis and Characterization of the Cross-Linking Monomer 76 ^  169 vii  4.2.2. Self Curing Reaction of the Cross-Linking Monomer 76^ 170 4.2.3. Curing Reaction of the Cross-Linking Monomer 76 with a Cyanate Resin ^  173  4.2.4 Copolymerization of the Cross-Linking Monomer 76 with an Olefinic Monomer ^  177  4.2.5. Conclusions ^  178  5. EXPERIMENTAL ^  180  5.1. High Resolution NMR Experiments ^  180  5.1.1. Solution NMR Experiments ^  180  5.1.2. Solid State NMR Experiments ^  180  5.2. Mass Spectrometry Experiments ^  181  5.3. X-ray Diffraction Experiments ^  181  5.4. Syntheses ^  182  5.4.1. Labelled Cyanogen Bromide ^  182  5.4.2. Labelled BPADCN (2a and 2b) ^  183  5.4.3. Labelled PTBPCN (25a and 25b) ^  184  5.4.4. Triazine 31 formed from PTBPCN ^  185  5.4.5. 2,6-Dimethy1-4-phenoxycarbonylmorpholine 46^ 185 5.4.6. 5-Phenoxymethy1-2-oxazolidinone 53 ^  186  5.4.7. Bisphenol-A Monoglycidyl Ether 62^  187  5.4.8. Bisphenol-A-Monocyanate Monoglycidyl Ether 61. ^ 188 5.4.9. Triazine 68 Formed from the Monoglycidyl Ether of Bisphenol-A-Monocyanate 61^  189  5.4.10. A More Practical Way to Synthesize the Cross-Linking Monomer 61^ 5.4.11. 2-Allylphenyl Cyanate 76^  190 191 viii  5.4.12. Triazine 77 formed from 2-Allylphenyl Cyanate 76^ 192 5.5. Chromatographic Separation of the Reaction Products ^ 192  6. PROPOSALS FOR FUTURE WORK^  194  7. REFERENCES ^  197  8. APPENDICES ^  205  A. Crystal Structure Data for Compound 31^  205  B. Crystal Structure Data for Compound 2 ^  215  C. Crystal Structure Data for Compound 53^  221  ix  LIST OF TABLES Table^  Page  2.1.^Characteristic 13 C and 15 N Chemical Shift Values of the Functional Groups Derived from the Cyanate Group in Solution Curing Reactions ^ 74 2.2.^Intermolecular Distances between Cyanate Groups in the BPADCN Crystal ^  91  x  LIST OF FIGURES Figures^  Page  1.1.^Schematic representation of different polymers. (A). Linear polymer; (B). Branched polymer; (C). Highly branched polymer; and (D). Cross-linked polymer^ 3 1.2.^Vector diagrams describing the pulsed NMR experiment in the rotating frame of reference ^  26  1.3.^Schematic representation of dipolar interactions ^ 29 1.4.^The effect of dipolar decoupling and magic angle spinning on' 3C solid-state NMR spectra of poly(butylene terephthalate) ^  31  1.5.^Schematic representation of chemical shift anisotropy powder patterns ^  34  1.6.^Schematic representation of magic angle spinning^ 36 1.7.^Larmor frequencies for 1H and 13C in a 9.4 T magnetic field. There is no frequency overlap, and thus no overlap in the energies ^ 39 1.8.^Vector diagram for a 1H — 13C cross-polarization experiment ^ 43 1.9.^Schematic representation of the 1H-13C cross-polarization pulse sequence for solid state NMR experiment ^  43  1.10. A more detailed representation of part C in Figure 1.8 ^ 44 1.11. A demonstration of the advantages of combining dipolar decoupling (DD), magic angle sinning (1VLkS) and cross polarization (CP) techniques for obtaining 13C solid state NMR spectra of poly(methyl methacrylate) ^ 45 1.12. Schematic representation of some important active sites in a heterogeneous polymer resin system ^  46 xi  ^  1.13. The effect of selective isotopic enrichment ^  49  2.1.^Solution NMR spectra (1H at 300 MHz) of bisphenol-A dicyanate (BPADCN) monomer in CDC13. (A). 13C NMR spectrum of natural abundance monomer 2; (B). 13C NMR spectrum of the 13C enriched monomer 2a; (C). 15N NMR spectrum of the 15N enriched monomer 2b  54  2.2.^13C solution NMR spectra (1H at 200 MHz) of (A). natural abundance p-tert-butylphenyl cyanate (PTBPCN, 25) and (B). 13C enriched PTBPCN 25a in methyl ethyl ketone (MEK) and acetone-d6 ^  55  2.3.^Solution NMR spectra (1H at 300 MHz) of 15N enriched PTBPCN 25b in acetone-d6. (A). 15N spectrum with 1H decoupling; (B). 15N spectrum without 1H decoupling; (C). 13C spectrum with 1H decoupling ^  57  2.4.^13C solution NMR spectra (1H at 200 MHz) of 13C enriched BPADCN 2a in MEK and acetone-d6 cured with 200 ppm zinc octanoate as catalyst. (A). Before heating; (B). After heating for 1 hour at 60 °C; (C). After 16 hours at 60 °C; (D). After 5 days at 60 °C ^  58  2.5.^15N solution NMR spectra (1H at 300 MHz) of 15N enriched BPADCN 2b in MEK and acetone-d6 cured with 200 ppm zinc octanoate as catalyst. (A). Before heating and without 1H decoupling; (B). After heating at 90 °C for 1 day and with 1H decoupling; (C). Same sample as in (B) without 1H decoupling ^ 60 2.6.^13C solution NMR spectra (1H at 200 MHz) of the 13C enriched PTBPCN 25a in MEK and acetone-d6 with 200 ppm zinc octanoate added as a catalyst. (A). Before xii  ^  the addition of water; (B). After addition of excess water and standing at room temperature for 24 hours; (C). After heating at 100 °C for 1 hour^  63  2.7.^15N solution NMR spectra (1H at 300 MHz) of 15N enriched PTBPCN 25b in MEK and acetone-d6 without 1H decoupling. (A). After heating at 100 °C for 5 hours without catalyst; (B). After heating at 100 °C for 1 hour with 200 ppm zinc octanoate as catalyst ^  65  2.8.^(A). 13C NMR spectrum (1H at 200 MHz) of p-tert-butylphenyl carbamate 28 in acetone-d6; (B). 15N NMR spectrum (1H at 300 MHz) without 1H decoupling of 15 % 15N enriched 28 in acetone-d6 ^  66  2.9.^13C solution NMR spectra (1H at 200 MHz) of the carbamate 29 in DMSO-d6. (A). Before heating; (B). After heating at  120 °C for 28 hours ^  68  2.10. (A). 13C NMR spectrum 0-H at 300 MHz) in acetone-d6 of natural abundance triazine 31 formed from PTBPCN 25; (B). 15N NMR spectrum 0-H at 300 MHz) in acetone-d6 of 15N enriched triazine formed from PTBPCN 25b^ 70 2.11. Perspective view of the triazine molecule 31 formed from PTBPCN 25. 50% probability thermal ellipsoids are shown for the non-hydrogen atoms ^  71  2.12. (A). Solid state 13C CP/MAS NMR spectrum (1H at 100 MHz) of the natural abundance BPADCN 2; (B). NQS spectrum (1H at 100 MHz) of 2; (C). NQS spectrum 0-H at 100 MHz) of the 13C enriched BPADCN 2a; and (D) 13C CP/MAS/TOSS spectrum (1H at 400 MHz) of 2a^  76  2.13. 13C solid state CP/MAS/TOSS NMR spectra (1H at 400 MHz) of (A). The solid sample obtained by evaporation of the solvent after curing the 13C enriched BPADCN 2a in MEK and acetoned6 with 200 ppm zinc octanoate as a catalyst; (B). The same sample as in (A), NQS experiment; and (C). The solid sample obtained from bulk curing the 13C enriched BPADCN 2a at 250 °C for 15 minutes ^  77  2.14. 15N solid state CP/MAS/TOSS NMR spectra (1H at 400 MHz) of (A). 15N enriched BPADCN 2b; (B — D). The solid sample obtained by evaporation of the solvent after curing the 15N enriched BPADCN 2b in MEK and acetone-d6 with 200 ppm zinc octanoate as a catalyst, with contact times: (B). 1 ms; (C). 5 ms; (D). 1 ms with NQS pulse sequence ^  80  2.15. Solid state 15N CP/MAS NMR spectra (1H at 400 MHz) of the resin obtained after bulk curing the 15N enriched BPADCN 2b at 250 °C for 15 minutes, with contact time 1 ms. (A).  With TOSS sequence; (B). With TOSS/NQS sequence ^ 81 2.16. Series of 13C CP/MAS NMR spectra (1-H at 400 MHz) with variation of the contact time (CT) without sideband suppression. The sample resin was obtained by bulk curing a mixture of 50% 13C enriched and 50% 15N enriched BPADCN monomers for 15 minutes at 250 °C ^  83  2.17. Solid state CP/MAS/TOSS NMR spectra 0-H at 400 MHz) of the same resin sample as Figure 2.16 with CT = 5 ms. (A). 13C spectrum; (B). 15N spectrum^  84  aciv  ^  2.18. Series of 15N CP/MAS NMR spectra (1H at 400 MHz) with contact time variation but without sideband suppression. The resin sample was the same as Figure 2.16 ^ 85 2.19. Molecular structure of BPADCN monomer 2 from the single crystal X-ray diffraction study showing the numbering of the atoms (Tables in the Appendix B) ^ 88 2.20. Perspective view of part of the unit cell contents from the crystal structure of BPADCN monomer 2 showing the intercyanate interactions. (see Table 2.2)^  89  2.21. A plane through the three-dimensional network formed by the intermolecular intercyanate interaction. The BPADCN molecules are cyanate-connected to form parallel "strings" throughout the structure ^  90  3.1.^15N solid-state NMR spectra (1H at 400 MHz) of the resin obtained by curing EPON-825 and BPADCN (50%13C enriched 2a and 50% 15N enriched 2b) at 180 °C for 2.5 hours. (A). CP/MAS/TOSS spectrum; (B). CP/MAS/TOSS spectrum combined with the NQS technique ^  97  3.2. Series of 15N solid-state CP/MAS/TOSS NMR spectra (1-H at 400 MHz) of the same sample as Figure 3.1 with variation of the contact time as indicated. The triazine resonance is indicated by T and the imidocarbonate by I ^ 98 3.3.^NMR spectra in acetone-d6 of the crude reaction mixture obtained by heating PTBPCN (12% 15N enriched) and PTBPGE at 180 °C for 7 hours. (A). 13C NMR spectrum (1H at 200 MHz); (B). 15N spectrum 0-H at 300 MHz) with NOE and no 1H decoupling^  100 XV  ^ ^  3.4.^NMR spectra of the major cross-reaction product after both silica gel column and Sephadex LH-20 colurnn separations of the same reaction mixture as Figure 3.3. (A). 15N NMR spectrum (1H at 300 MHz) with NOE and no 111 decoupling; (B). 1H (500 MHz) NMR spectrum; and (C). 13C NMR spectrum (1H at 500 MHz) ^  105  3.5.^(A). 1H (400 MHz) 2D COSY NMR spectrum and (B). 1H/13C chemical shift correlated 2D NMR spectrum 1H at 500 MHz) of the same sample as Figure 3.4 ^ 106 3.6.^13C NMR spectrum (111 at 200 MHz) of 2,6dimethylmorpholine 44 in acetone-d6 ^  110  3.7.^Aliphatic regions of the 13C NMR spectra (11i at 300 MHz) of 2,6-dimethy1-4-phenoxycarbonylmorpholine 46 in acetone-d6 at the temperatures indicated ^  112  3.8.^Aliphatic regions of the 13C NMR spectra (1H at 300 MHz) of the cross-reaction product (Figure 3.4), at the temperatures indicated ^  113  3.9.^1H (400 MHz) NMR spectra in aromatic regions of the major cross-reaction products derived (A). From PTBPCN 25 and PTBPGE 37; and (B). From PTBPCN 25 and OMPGE 38^ 116 3.10. 1H (400 MHz) NOE difference NMR spectra of the major cross-reaction product derived from the reaction of PTBPCN 25 and PTBPGE 37, which is the same as Figure 3.4^ 117  3.11. 1H (400 MHz) NOE difference NMR spectra of the major cross-reaction product derived from the reaction of PTBPCN 25 and OMPGE 38 ^  118  xvi  3.12. NMR spectra of the second unidentified product after both silica gel column and Sephadex LH-20 column separation of the crude reaction mixture obtained by heating PTBPCN (-12% 15N enriched) and PTBPGE at 180 °C for 7 hours. (A). 15N spectrum (1H at 300 MHz) without 1H decoupling; (B). 13C spectrum (1H at 200 MHz); and (C). 1H (200 MHz) spectrum ^  122  3.13. 15N NMR spectrum (1H at 300 MHz) in acetone-d6 with NOE and no 1H decoupling of the reaction mixture obtained by heating p-t-butylphenyl carbamate (-12%15N enriched) and PTBPGE at 180 °C for 3.5 hours ^  124  3.14. NMR spectra in acetone-d6 of the second product after silica gel column separation of the same reaction mixture as in Figure 3.13. (A). 15N spectrum (1H at 300 MHz) with NOE and no 1H decoupling; (B). 13C spectrum (1H at 200 MHz); and (C). 1H (200 MHz) spectrum^  125  3.15. NMR spectra of product 53 in acetone-d6. (A). 15N spectrum (1H at 300 MHz); (B). 13C spectrum 0-H at 200 MHz); and (C). 1H (200 MHz) spectrum ^  127  3.16. Molecular structure of product 53 from the single crystal X-ray diffraction experiment with the numbering of the atoms indicated. Complete structural data are given in Appendix C ^ 128 3.17. NMR spectra of product 55 in acetone-d6. (A). 13C spectrum (1H at 200 MHz); (B). 1H (500 MHz) spectrum ^ 131 3.18. (A). 1H-13C 2D heteronuclear correlation NMR spectrum 0-H at 500 MHz) and (B). 1H (500 MHz) 2D COSY NMR spectrum of product 55 in acetone-d6 ^  132 xvii  ^  3.19. NMR spectra of product 59 in acetone-d6. (A). 13C spectrum (1H at 200 MHz); (B). 15N spectrum (1H at 300 MHz) with no 1H decoupling^  138  4.1.^13C NMR spectrum 0-H at 200 MHz) in acetone-d6 of the crude product prepared by reaction at 56 °C for 90 min. and using a reactant mixture with a 1:1:1 molar ratio of bisphenol-A, epichlorohydrin and potassium carbonate. (A). Full spectrum; (B). The expanded aromatic region 151 4.2.^NMR spectrum (1H at 200 MHz) of the monoglycidyl ether of bisphenol-A 62 in acetone-d6. (A). 13C spectrum; (B). 1H spectrum ^  153  4.3.^NMR spectrum (1H at 200 MHz) of the crude monoglycidyl ether of bisphenol-A-monocyanate 61 in acetone-d6. (A). 13C spectrum; (B). 1H spectrum ^  156  4.4.^(A). 13C NMR spectrum (1H at 300 MHz) in acetone-d6 and MEK of the crude product 68 from curing the monoglycidyl ether of bisphenol-A-monocyanate 61 by heating; (B). 1H NMR spectrum (200 MHz) in acetone-d6 of the crude product 68^  157  4.5.^13C NMR spectrum (1H at 200 MHz) in acetone-d6 of the crude product from curing the monoglycidyl ether of bisphenol-A-monocyanate 61 at room temperature with diethylamine base (monomer 61 is in excess) ^ 160 4.6.^13C NMR spectrum (1H at 200 MHz) in acetone-d6 of the crude product from curing the monoglycidyl ether of bisphenol-A-monocyanate 61 at room temperature with excess cliethylamine base ^  161 xviii  ^ ^  4.7.^NMR spectrum (1H at 200 MHz) in acetone-d6 of the cyanate product mixture obtained from the intermediate mixture without separation. (A). 13C spectrum; (B). 1-H spectrum ^ 165 4.8.^NMR spectra (1H at 200 MHz) of 2-allylphenyl cyanate 76 in acetone-d6. (A). 13C spectrum; (B). 1H spectrum ^ 171 4.9.^NMR spectra (1H at 200 MHz) of 1,3,5-tri(2-allylphenoxy)2,4,6-triazine 77 in acetone-d6. (A). 13C spectrum; (B). 1H spectrum ^  172  4.10. NMR spectrum in acetone-d6 of the product mixture obtained from curing a mixture of cross-linking monomer 76 and 12% 15N enriched PTBPCN 25. (A). 13C spectrum (1H at 200 MHz); (B). 15N spectrum (1H at 300 MHz) ^ 175  xix  ABBREVIATIONS The following abbreviations have been used throughout this thesis. a.m.u.^= atomic mass unit Ar^= aromatic substitute b.p.^=^boiling point BPADCN = bisphenol-A dicyanate BT^= bismaleimide and triazine CI^= chemical ionization COSY^= proton homonuclear correlation spectrum CP^= cross polarization CT^= contact time DGEBA^= diglycidyl ether of bisphenol-A DMMP^= dimethyl morpholine El^= electron impact EPON-825 = low molecular weight epoxy resin FID^= free induction decay FM^= formula mass FT^= Fourier transformation GPC^= gel permeation chromatography HETCOR = heteronuclear correlation spectrum HP^= high power decoupling hr.^=^hour(s) H.T.^= high temperature IR^= infrared L.T.^= low temperature M+^= parent ion XX  MAS^= magic angle spinning MEK^= 2-butanone (or methyl ethyl ketone) min.^= minute(s) MMA^= methyl methacrylate m.p.^= melting point MS^= mass spectrometry ms^= millisecond(s) NMR^= nuclear magnetic resonance NOE^= nuclear overhauser effect NQS^= non quaternary suppression; pulse sequence for selecting X nuclei with no attached proton OMPGE^= ortho-methylphenyl glycidyl ether ppm^= part per million PTBPCN = 4-tert-butylphenyl cyanate PTBPGE = 4-tert-butylphenyl glycidyl ether Tg^= glass transition temperature TMS^= tetramethyl silane TOSS^= total suppression of spinning sidebands; pulse sequence for spinning sideband suppression O^= chemical shift  ACKNOWLEDGEMENTS Firstly, I would like to sincerely thank my supervisor Prof. C. A. Fyfe for his guidance, encouragement, advice and support throughout the course of thesis studies. I am also very grateful to my supervisory committee members, Prof. R. E. Pincock and Prof. F. G. Herring for their advice and proof-reading of this thesis. Many thanks are extended to the following people: All of my past and present labmates and friends for their interaction socially and academically. In particular, Dr. N. E. Burlinson and Dr. H. Grondey for their valuable discussion, helpful advice and proof-reading of the thesis. Dr. G. Fu and Mr. K. Wong-Moon for their proof-reading of the thesis. Mr. K. Mok for his valuable collaboration. Dr. S. Rettig for the determination of crystal structures from X-ray diffraction experiments. Dr. S. 0. Chan, Ms. L. Darge, and Ms. M. Austria in the NMR laboratory, Mr. T. Markus and Mr. Kam Sukul in the electronic shop for their frequent and kind help. Dr. D. Wang, Dr. M. Poliks, and Dr. C. Reidsema at IBM for their helpful collaboration. Finally, I am greatly obliged to my parents and sisters for their constant love, encouragement and support in all aspects. I also wish to extend special thanks to my wife, Jane and son, Simon for their love, patience and support over last few years.  CHAPTER 1. INTRODUCTION The work described in this thesis is the investigation of cyanate resins and related mixed heterogeneous polymer systems primarily by high-resolution solution and solid-state NMR spectroscopies using specific isotopic enrichment at the reaction sites. In the Introduction, the general features of polymer resins will be outlined and a description of the solid-state NMR techniques will be given to facilitate the presentation of the experimental investigations in subsequent chapters of the thesis.  1.1. Polymer Types Polymers or macromolecules are usually made up of different sequences of repeating chemical structural units, which may be arranged regularly or irregularly to form linear or three dimensional networks. They have very large molecular weights, which could be 10,000 or greater. In polymer chemistry, different classifications have been used based on the polymer compositions, the polymerization mechanisms, and the polymer structures. Polymers were originally classified by Carothers[11 in 1929 into condensation and addition polymers on the basis of the compositional differences between the polymer and the monomer from which the polymer was synthesized. Condensation polymers are those polymers which are formed from polyfunctional monomers by various condensation reactions which involve the elimination of some small molecules such as water. Addition polymers are classified as those which are formed from monomers without the loss of any small molecule. Unlike condensation polymers, the repeating unit of an addition polymer has the same composition as the monomer. 1  Flory[21 pointed out the significant difference between the two polymerization mechanisms, called step and chain growth polymerizations. Based on this, polymers were divided into step-growth polymers and chaingrowth polymers. Step-growth polymerizations proceed by the stepwise reaction of the functional groups on different reactants. Any two molecular species can react with each other throughout the course of the polymerization. The size of the polymer molecules increases at a relatively slow rate in such polymerizations. In chain polymerizations, the situation is quite different and full-sized polymer molecules are produced almost immediately at the beginning of the reaction. Chain-growth polymerizations require an initiator to produce a reactive center, which may be a free radical, cation or anion. A monomer can react only with the reactive center, not directly with other monomers.  B  C  Figure 1.1. Schematic representation of different structures of polymers. (A). Linear polymer; (B). Branched polymer; (C). Highly branched polymer; and (D). Cross-linked polymer.  2  Polymers can also be classified as linear, branched, and cross-linked polymers depending on their structures[3] (Figure 1.1). The linear or lightly branched polymers are usually called thermoplastics, and they are reversibly fusible and can be re-shaped by the application of heat or pressure. In the case of cross-linked polymers, the molecular chains are joined together by covalent bonds and so the chains cannot slide past each other upon the application of heat or pressure. Highly cross-linked polymers formed by the action of heat can not be re-shaped by heating and are termed thermosets or thermoset resins. The term resin is generally used to indicate a precursor of a cross-linked polymeric material; sometimes, however, it is applied to any material whose molecules are polymers. [4]  L2. Cyanate Resins Cyanate resins, or cyanuric esters, all contain the cyanate group —0—CoN attached to a benzene ring, and the polymerization reaction proceeds by the thermally induced reaction of this group. The development of cyanate resins started in the mid 1960's, with cyanate compounds first being successfully synthesized from aromatic phenols[5,6]. A practical synthetic route for manufacturing cyanate esters was invented and developed by Bayer AG in the later 1960's[7,8]. Aliphatic cyanates readily isomerize to the corresponding isocyanates and then subsequently can trimerize to form isocyanurates.[9] By application of heat or organo-metallic compounds as catalysts, aromatic cyanates directly trimerize to form a stable cyanurate, s-triazine ring structure. [10,11 ] Although in both cases, based on calculations, the isocyanurate form is thermodynamically more stable, aryl cyanates and aryl cyanurates do not seem to isomerize.[12]  3  Cyanates can be prepared from the reaction of cyanogen halides and phenols in the presence of a hydrogen halide acceptor,[5,6] or from thiatriazoles by thermo1ysis[13,14]. For the manufacture of cyanate resins, the process using cyanogen halides and phenols is the most important. The general method for preparation of the monomers of cyanate resins can be represented by Equation 1.1.  OCN^(CH3CH2)3NHX^[1.1]  OH XCN  Commercially available cyanate resins are bisphenol derivatives containing cyanate functional groups (—OCEIN). A typical monomer of the cyanate resin has the general structure 1,  X - Bisphenol Linkage R - Ring Substituent The cyanate resin most commonly used is based on bisphenol-A dicyanate (BPADCN, 2) N=-C-0  0-CEN  2 derived from bisphenol A (3), which is also a raw material for epoxy resins. Bisphenol A (3) is so-called since it is formed from two phenols and one acetone molecule as shown in Equation 1.2.  4  2  HO  9H3 + 9=0 CH3  H+  HO  OH + H2O^[1.2]  The curing reaction of cyanate resins is assumed to be three cyanate functional groups on different monomers being cyclotrimerized to a triazine ring upon heating[- 0'-1-I [Scheme 1.1] to form the three-dimensional network of a thermoset resin. Therefore chemically, this family of thermosetting monomers and their prepolymers are esters of bisphenols and cyanic acid. These three dimensional networks of oxygen linked triazine rings and bisphenol units should be correctly termed polycyanurates. Since no leaving groups or volatile byproducts are formed during the curing process, the cyanate resin produced through a cyclotrimerization curing reaction is classified as an addition polymer. Because of the step polymerization mechanism of the curing reaction, it can also be classified as a step-growth polymer. The cyclotrimerization of the cyanate resin is found to be facilitated by soluble transition metal compounds as catalysts. The role of the soluble transition metal compounds is thought to be primarily by coordination, gathering cyanate groups in close proximity to facilitate the ring formation . [11 ,15] Curing reactions of cyanate resins have been investigated by infrared spectroscopy.[16,17] Following the disappearance of the strong cyanate —0CraN doublets at 2240 and 2270 cm-1 during the curing process, new absorbances appear simultaneously at 1565 cm-1 and at 1365 cm-1 (cyanurate). However, although NMR spectroscopy is a very diagnostic and reliable technique, there is  5  NEC-0  0-C-=N  2  Scheme 1.1  -11-0  ^N  N N  Y 0  N  6  no information from NMR studies to date regarding the species involved during the curing reaction, or the nature and efficiency of the curing process itself. Fang has recently described the use of 13C solution NMR spectroscopy to determine the molecular weights of bisphenol-A dicyanate (BPADCN, 2) oligomers.[18] By using modern solid-state and solution NMR techniques, the direct detection and characterization of all the species formed during the curing process will be reported in this thesis. Laminating resins based on bisphenol-A dicyanate prepolymer were not commercially available until 1975[19]. The Mitsubishi Gas Chemical company later introduced a BT resin system, which is blends of bismaleimide and triazine (cyanate ester) resins, in the mid 1980's[20]. While BT resins solved the laminate moisture problem, performance was compromised to some extent by their brittleness, and higher dielectric constants. However, pure and unblended materials of cyanate resins were not available until 1985 when Interez Inc. marketed a series of resin systems for both circuit boards and advanced composite applications[21]. Hi-Tek Polymers Inc. introduced an improved process for producing >99% pure monomers from a variety of bisphenol precursors, leading to the introduction of three aryl dicyanates and their prepolymers in 1985-1986. The moisture absorption problem of cyanate resins was overcome by eliminating carbamate impurities in the monomers and avoiding use of aliphatic amines as curing catalysts. The dielectric constant of the cured resin decreased to 2.66 (and ultimately 2.5 with experimental products), and moisture absorption at saturation was lowered to 1%. These developments were commercialized by Hi-Tek Polymers Inc. in the 1985-4989 period122,23,24]. A new cyanate resin with low temperature cure capability has been recently developed125 ,2 6]•  7  The principal end uses for cyanate resins are as matrix resins for printed circuit board laminates[22, 23, 27] and structural composites[16, 24, 28, 29]. Since the 1980's, there have been major changes[30] in the requirements and performances of high speed logic circuits, such as in RF/microwave telecommunication switches, radar and some military devices. These requirements are higher circuit density, faster signal transfer speed, and greater reliability. The recently developed cyanate resins help to meet these requirements. In the last ten years, aerospace composites which have a high damage tolerance have been developed by utilizing a mixture of both thermoset and thermoplastic resins.[24,31] Cyanate resins develop approximately twice the fracture toughness of multifunctional epoxies and are able to operate at 150 °C (300 °F). The unusually low capacitance properties of cyanate resins, due to their very low dielectric constants which are in the 2.5 — 3.1 range, are also utilized in military aircraft which have reduced radar signatures.[17] In the electronic market, cyanate resins show some very attractive features[' 7, 22, 23, 27], such as very low dielectric constant, high dimensional stability at molten solder temperatures, and excellent adhesion to conducting metals at temperatures up to 250 °C. Also, the processing characteristics in ketone solutions and the drillability of conventional diepoxide laminates have been retained. Since cyanate resins exhibit good solubility in organic solvents, they can be mixed with a wide variety of thermosets such as epoxies, bismaleimides or acrylates to form compatible blends.[32, 33] Mixing with thermoplastics usually results in interpenetrating network type structures.[16, 29, 34] After being fully cured, cyanate resins exhibit high glass transition temperatures (Tg) of at least 250 °C. Usually they are blended with epoxies up 8  to 60 – 70 % by weight in order to modify the Tg, increase toughness, reduce the cost and achieve non-flammability requirements. Blending with epoxies allows BT and cyanate resins to reach their full cure at 177 °C. The laminates made from such blends exhibit good moisture and delamination resistance. However, the nature and the mechanism of the curing reactions which occur in the mixed cyanate and epoxy resins have not been clearly established to date,[30] even though the formation of oxazoline (5) and isooxazoline (6) ring structures has been proposed[32, 35, 36]. 0— ..--0^,11 0^ 0 6 5^ oxazoline ring^isoxazoline ring  L3. Epoxy resins The first patent for the synthesis of the materials designated as epoxy resins was made by Pierre Castan of Switzerland about fifty years ago[37]. Since that time there has been intense activity in the synthesis of new epoxy compounds with different structures[38,39,40]. In addition, investigations of relationships between structure and macroscopic properties have increased our understanding of these polymer systems[41 ,42,43]• In a broad sense, the term epoxy refers to a chemical group consisting of an oxygen atom bonded with two carbon atoms which are already linked in some other way. The simplest epoxy is a three-membered ring to which the term aepoxy or 1,2–epoxy is applied. Ethylene oxide (7) is an example of this type. The terms 1,3– and 1,4–epoxy are applied to trimethylene oxide (8) and tetrahydrofuran (9).  9  ,0 CH2 —CH2 7  8  9^10  In general, only those epoxy compounds which contain the threemembered epoxy rings (ie. ethylene oxide derivatives) are referred to as epoxy resins. There is no universal agreement on the nomenclature of the threemembered epoxy ring[391. In fact, there is division even on the term epoxy itself; European authors generally prefer the term epoxide, while in America the term epoxy is more common. The epoxies may be designated as oxides, as in the case of ethylene oxide (epoxyethane) (7) or cyclohexene oxide (1,2-epoxycydohexane) (10). Several of the more common monoepoxies have trivial names, such as  epichlorohydrin (11), glycidic acid (12), and glycidol (13). Glycidyl (14) is used to refer to the terminal epoxy group, the name being modified by ether, ester, amine, etc., according to the nature of the group attached to the third carbon. An epoxy resin is defined as any molecule containing more than one epoxy group (whether situated internally, terminally, or on cyclic structures) which is capable of being converted to an useful thermoset material.[39] The term is applied to the resins both in the thermoplastic (uncured) or thermoset (cured ) state. 0^0 / CH2 -CHCH2CI^CH2 -CHCOOH^CH2 -CHCH2OH  11^12^13  /\  CH2 -CHC H 2 -  14  A number of properties have led to the rapid development of epoxy resins and their use in a wide range of industries. Epoxy resins belong to the thermosetting class of polymers. They offer great versatility, low shrinkage, excellent chemical resistance, outstanding adhesion, high mechanical strength,  10  and very good electronic insulation. Epoxy resins can be cured quickly and easily at practically any temperature from 5 to 150 °C, depending on the choice of curing agents.[39] Because of their versatility, the epoxy resins are used in thousands of industrial applications as adhesives, caulking compounds, casting compounds, sealants, coating materials, and as laminated resins for both constructional materials and electronic circuit boards[39,40,44]. These uses have stimulated an extensive research in the synthesis, structure and properties of epoxy resins. [45,46,47,48] The epoxy resin is normally synthesized by the base catalyzed reaction of epichlorohydrin (11) with an appropriate molecule, which has several phenolic hydroxyl groups.[4,39,40] The reaction proceeds in two steps as shown in Scheme 1.2.  OH + CICH2CH—CH2 \ / 11 0  OCH2CHCH2CI  NaOH  OH 15^  NaOH  0CH2cH—cH2 \/ 16^0  Scheme 1.2  First, the epoxy groups react with the phenolic hydroxyls to form a chlorohydrin intermediate (15), and then chlorine and hydrogen are stripped off to regenerate the epoxy groups in the epoxy resin (16). The first commercial epoxy resin was made by the reaction of epichlorohydrin (11) and bisphenol A (3), which gives the diglycidyl ether of bisphenol-A (DGEBA) (17) and higher molecular weight species (18). The 11  molecular weight of the resulting DGEBA resin will depend on the ratio of epichlorohydrin to bisphenol A employed. The greater the excess of epichlorohydrin is used, the lower the molecular weight of the resulting resin will be. In order to obtain high yields of the monomeric product, excess epichlorohydrin is employed, usually two or three times the stoichiometric amount. [4,391  OH^CICH2CH—CH2 \ /  HO  11 0  3  NaOH  ^  [1.3]  OC H2C 11 -/ C H2  17 •\  C H2-CHCH20  OH^ OCH2CHCH 0^  0 OCH2CH CH2  18  The commercial DGEBA epoxy resins are mixtures of high and low molecular weight oligomers, the distribution of molecular weights varying with the conditions of synthesis. The general structure of an epoxy resin is represented as 18. The low molecular weight resins, having an n value about 1 or below, are generally liquid. Above n=1, the resins are brittle solids. In fact, the pure diglycidyl ether of bisphenol-A (17), n=0, is a solid material melting at about 45 °C. It is the impurities present in the commercial epoxy resins which transform them to "supercooled" liquids.[491 In the synthesis of commercial  12  epoxy resins, epichlorohydrin is the principal epoxidizing reagent. Other epihalohydrins may be used, but are not economically attractive. Epoxy resins can also be derived from different polyolefinic compounds by addition of oxygen to the unsaturation. The direct oxidation with molecular oxygen would be an ideal synthesis process, but this is practical only in the synthesis of ethylene oxide[50]. For other materials the oxygen is transferred from a source compound, either a peracid, hypochlorous acid, or hydrogen peroxide.[511 The principal industrial processes usually use the peracid method [Equation 1.4 and 1.511. Peracetic acid is more favoured for economic reasons. H202 + RCOOH ^■-- H20 + RCOOOH^[1 .4] ----"C=C + RCOOOH  0 C--C + RCOOH^[1 .5] ^"-,--^-.  Higher cross-linking densities, giving increased solvent resistance and elevated glass transition temperature, Tg, can be obtained by using compound 19 and especially 20.[521 A molecule of dual functionality 21, which can crosslink two types of polymer systems, may be used to give a resin system with improved thermal and electrical properties.[53]  CH\2 -;CHCH 0  OC H2Ctl-CH 2  o  '  H2C CH,2 -,CHCH20  OCH2Ctl-,CH2  o  19  H2C CH,2- C H C H2 0 0  OCH2C,tH CH2  o  "  13  CH,2 C H C H2 0 CH2 - C H C H2 0  HC CHO ^  ^  OCH2CH-CH2 \0 /  \CD)^OCH2C,1-10-/CH2  20  CH2CH =CH2  OCH2C 1-1 -/C H2 0  C171,2-;CHCH20  21  CH2=CHCH2  The most valuable property of the epoxy resins is their ability to transform readily from the liquid (or thermoplastic) state to tough, hard thermoset solids. The conversion is accomplished by the addition of a chemically active compound known as a curing agent (hardener, activator, or catalyst). This may be either a Lewis acid or base compound. Basically the cured structure may be a homopolymer or a heteropolymer or a mixture of both types. The epoxy group may react in one of two different ways during curing: anionically or cationically.[4,391 In the anionic mechanism, the epoxy ring may be opened by an anionic curing agent x- to produce an epoxy anion: z"--‘ - \^.---  X^ C—C...,_  ^0.-  [1.6]  0 Epoxy ring The epoxy anion is an active chemical species, capable of further reaction producing additional covalent bonds and forming the resin framework. In the cationic mechanism, the epoxy group may be opened by an active hydrogen to produce a hydroxyl group in a number of ways:  14  ^  HX i"  X -.........1^„......C—C^+ HX^[1.7]  — C'--. (1 )^C ----^/ L\0 +IX  OH X -.. ' HX^[1.8] C—C^+ --,. OH  `..^..--^ - -+./ C — C^ —C /C 1 10/' rHX^OH m-  (2)  ...  `.^./ C — C^--..^7-' \ /^C — C X^0^i^i __ ^ X OH H  (3)  [1.9]  The hydroxyl group can also react further with other epoxy groups to form a ring opened polymer chain.  X --^1^-OH  +  n-1  .^.---...-^\ /^-.. . 0  I^I X C—C-0—H I^I^_n [  [1 . 1 0 ]  1.4. Coupling Agents and Glass Reinforced Resins Coupling agents are defined as materials that improve the adhesive bond of a polymer to glass, mineral or metal surface, and improve the chemical resistance (especially to water) of the bond across the interface. [54] Organofunctional silanes, which are hybrid organic-inorganic compounds, can be used as coupling agents, or adhesion promoters, between organic polymers and inorganic mineral, or glass substrates. For example, organosilanes are widely used as coupling agents in glass fiber-reinforced plastics{55,561. In the polymer industry, reinforced polymer systems are very important material types. Some of the advantages of the reinforcement are improved 15  mechanical properties, improved electrical insulation properties, increased dimensional stability, and improved processing characteristics[57]. The reinforcements are typically glass fibers, which are often treated with organosilane coupling agents to promote adhesion between the glass fibers and the polymer matrix[54]. The macroscopic properties of composites depend on the properties of the glass fiber, the interface of coupling agent and fiber, the interface of coupling agent and polymer, and the polymer matrix. All of these contributions must be optimized in order to increase the potential of using such a polymer system[581. It is therefore necessary to investigate the nature of these structural components at the molecular level in order to understand the performance properties of these reinforced polymer systems. Various theories for the adhesion promotion through a silane coupling agent have been proposed. The chemical bonding theory remains the most important[591. These organosilanes used as coupling agents have the general structure R—Si—X3, where R is a functional alkyl group and X is a hydrolyzable group, such as an alkoxy group or halogen. During application, it is assumed that (i) the X groups react with the silanol groups on the glass surface to form an ether linkage [Equation 1.11]; (ii) the hydrolyzed silanes also self-condense to form polysiloxanes; and (iii) the functional groups on the R group can react with appropriate chemical groups in the resin, and thus chemically couple the resin to the glass [Equation 1.12]. The X groups serve only to provide hydrolytic sites. Their chemical nature does not affect the structures of the final products of the coupling action. The R group is chosen for its chemical reactivity with the resin component. In general, the effectiveness of the silane as a coupling agent is reported to parallel the reactivity of the organofunctional group with the resin.[54,60]  16  I I^ 0 Si —R + HX  I^ OH + X—Si—R  CH:=-CH2  I 0 Si—CH=CH2  +  1  22  23 [1.12]  ---ECH—CH2IFICH— C H2 -ECH — CH2Vri-Si 1  0  24  Assuming that this theory describes the processes correctly, the coupling agent acts as a bridge to covalently bond the glass to the resin. This could be expected to lead to the strongest interfacial bond. Indirect evidence for coreaction of the organofunctional group with thermosetting resins has been obtained. For example, in terms of strength properties, vinyltrichlorosilanefinished glass gives unsaturated polyester laminates with dry and wet strengths about 60% greater than those of laminates of ethyltrichlorosilane-finished glass[54]. An unsaturated polyester laminating resin should be able to copolymerize with the olefinic groups in the coupling agent, vinyltrichlorosilane, but not with ethyltrichlorosilane. However, chemical bonding at the interface is difficult to detect directly, because of the thinness of the interface. Covalent bonding between the coupling agent and glass surface was very difficult to demonstrate. Extraction studies reveal that the silane anchored on a glass surface is stable against long time extraction. This suggests that there is covalent chemical bonding between the 17  silane and the glass fiber.[611 With the development of Fourier transform infrared spectroscopy (FT-IR), it was possible to detect the chemical reaction of the silane with the glass fiber.[621 It was observed that silanol condensation was enhanced by the glass surface and cross condensation between the coupling agent and the glass occurred during drying. [63] Because of the complexity of composites and their insolubility, FT-IR has been one of the few spectroscopic techniques capable of characterizing them. Consequently, a great deal of work has been done in this field[641. However, recent developments in NMR spectroscopy have made it possible to obtain highresolution NMR spectra from solids, making it another potential spectroscopic technique in this area.[651 The combined use of high power decoupling (HP), magic angle spinning (MAS) and cross polarization (CP) techniques makes it possible to obtain high resolution 13C NMR spectra for insoluble polymers[66,67,68,69] as well as organosilane moieties bound to silica surfaces[70,71,72,73]. Combined with specific isotopic enrichment, NMR spectroscopy is a much more diagnostic and selective technique than IR spectroscopy. In surface studies, 13C NMR has the particular advantage of avoiding interference effects from the glass reinforcement matrix, which is a problem that may arise in other surface characterization methods. The usefulness of solid state 13C CP/MAS NMR spectroscopy has been shown for the characterization of the glass/coupling agent interface of glass fibre reinforced polymer systems. 13C NMR studies of a range of organosilane coupling agents adsorbed on silica surfaces have been carried t[74]. These studies show that it is possible to probe the glass/coupling agent interface of a composite and obtain the information on chemical bonding between the coupling agent and the glass, as well as on the structure of the coupling agent on the glass surface. However, the investigation of the interface between the coupling 18  agent and polymer is much more difficult because of the strong background interferences from the polymer matrix. The proposed covalent bonding between the polymer and the coupling agent anchored on the glass reinforcement has never been directly demonstrated.  1.5. Polymer Resin Characterizations Polymer structure can be described at two different levels. The most fundamental one concerns the chemical microstructure, which is defined by the chemical composition, chemical bonding and internal sequences of different arrangements in the macromolecule, i.e. its structure, configuration and conformation. The second structural level can be termed as phase structure, which describes the intermolecular packing of the polymer molecules as crystals, semi-crystals, amorphous solids, rigid or rubbery solids or liquids. The portion that crystallizes is a solid in the classical sense, although there may be some type of molecular motion in the crystal and the ordering may not be perfect over a long range. The disordered portions that do not crystallize can be glassy and rigid if the temperature of observation is below the glass transition temperature, Tg, or rubbery if above Tg. The primary purpose for characterizing a polymer system is to determine the polymerization mechanism and the structure of the polymer, and finally, to relate these to the performance properties of the polymer in its end use. If the structure of a polymer and the polymerization mechanism are completely characterized and the properties of the structural components are known, the polymerization process may possibly be optimized and controlled to get the optimum desired properties of the polymer system. However, synthetic polymers are very complicated systems. Even for a single linear polymer chain with only two different structural elements, A and B, and total n number of elements in 19  the polymer chain, the number of possible different chain structures is 21, which could easily be 10,000 or greater. To make it even more complicated, most synthetic polymers have a large number of different structural elements in the polymer chains, or three dimensional networks. For many real polymers, many possible structural variables can also exist. The very important one is the chemical and stereochemical arrangement, which includes monomer composition, arrangement sequence, and tacticity. Branches, crosslinks, end groups, and chain defects can also be very important aspects of polymer structure, even if they are present in low concentration. Molecular weight and its distribution are other important factors. Morphological, conformational effects and chemical defects, which include impurities, monomer isomerizations, and side reactions, can also be considered as additional variables. Because a number of these structural variables can exist simultaneously in a polymer, the number of possible structures for the polymer molecule can be very large. Therefore, it is usually not possible to completely define the spatial coordinates of every atom in polymer molecules. The problem is further complicated by the nature of the distribution of the structural variables along the polymer chain or on the polymer networks which greatly influences the polymer properties. Distributions of these structural elements can be random, blockwise or alternating, and are determined by the nature of the polymerization process. The distribution can influence the characterization of the polymer in two ways. Firstly, the chain or network structure is highly variable because the polymerization process is a statistical process. Therefore, the polymer sample is always a complex multicomponent mixture. Secondly, the detailed local structures can not be obtained because the measurement techniques which are available today can only provide weighted average structural data.[75] 20  Ideally, spectroscopic techniques for the study of polymers should yield narrow linewidth, high-resolution spectra which provide diagnostic and selective structural information. Since polymer systems are always complex mixtures of different structural components and molecules, a suitable spectroscopic method must be capable of selectively monitoring more than one structural component at one time. It must have sufficient sensitivity to detect and monitor very low concentrations of structural components in the polymer, since small structural changes may produce very large changes in the physical and mechanical properties. The spectroscopic technique should provide very specific information, since we will need to determine not only the structure of the single repeating units, but also how they are connected together and to what extent the units are ordered. A non-invasive and nondestructive technique is preferable, because it allows the study of the same polymer sample by other methods. The technique should also be capable of studying the polymer in its end-use form, such as a fiber, film, composite, coating, or adhesive. Most spectroscopic techniques, such as UV / visible spectroscopy or mass spectrometry, do not meet these requirements. However, some techniques have evolved for polymer analysis that do, such as high-resolution solid state NMR spectroscopy, Fourier transform infrared (FTIR) and Raman spectroscopies, and X-ray diffraction analysis. Each of them has both advantages and certain limitations. By using them in combination, detailed structural information of polymers for analysis, quality control, and research can be obtained. Even though it is not possible to study polymers in their final engineering form, usually a solid, the contributions that high-resolution solution NMR spectroscopy of polymer studies has made to our understanding of the structure of polymers are great. For the investigation of polymerization reactions, the corresponding monofunctional compound is frequently used as a model of the 21  multifunctional monomer to simulate the real polymerization process. In this case, the polymer chains or frameworks are not formed and all the products are soluble.  1.6. High Resolution Solid State NMR Spectroscopy of Polymers In the last 20 years, tremendous progress has been made in the area of NMR spectroscopy. The development of Fourier transform techniques and high magnetic fields have changed the NMR method from essentially proton NMR spectroscopy to multinuclear spectroscopy. More recently, rapid progress has been seen particularly in three areas[76]. One is the NMR imaging technique. Another is the use of multipulse sequences and multi-dimensional NMR spectroscopy to obtain additional information which was not available previously. The last is the capability of obtaining high-resolution NMR spectra from solids. The combination of the techniques of high power decoupling, magic angle spinning, and cross polarization have opened new areas of chemistry and physics to high resolution NMR[651• Most chemists are familiar with high resolution solution NMR for characterizing the structure of a molecule in solution. Thus, the first expectation for many polymer chemists would be that high resolution solid state NMR could be the counterpart of the solution technique, which could provide not only the same structural information as solution NMR, but also with some advantages. These supposed advantages would be (i) increasing sensitivity, because the sample is not diluted; (ii) the ability to study insoluble materials; and (iii) the ability to observe species that are unstable or short lived in solution because of their chemical reactivity or thermal instability. On the other hand, those interested in the physical properties of polymers are more concerned with the intrinsic properties of a polymer in the solid state, because polymers in the end use are often in the solid 22  form. High resolution solid-state NMR spectroscopy gives the possibility of providing a more direct link between a polymer structure and its physical properties in its end use.  1.6.1. Principles of Pulsed FT-NMR When the magnetic moment m, of the nucleus interacts with a magnetic field 110, the torque exerted on the spinning nucleus causes the nuclear magnetic moment to precess about 110. The Larmor frequency COo is the frequency of this precession, and is proportional to the gyromagnetic ratio y of the nucleus and the magnetic field strength 1/0 as shown in equation  coo = y H0.  ^  (1.6.1)  When a nucleus with a non-zero spin number I is placed in a magnetic field, 21+1 quantized energy levels will be generated. For example, two energy levels are generated for 1H or 13C nuclei, which have 1=1/2. The separation of these energy levels, AE, depends upon the magnetic field strength Ho, and can be expressed as:  AE = Y H 2ic  (1.6.2)  More spins tend to align their magnetic moments along the field direction than against it. The population of spins in each energy level (N+ and NJ is described by the Boltzmann distribution:  N+ e —dE 1 kT NJN+  ^  (1.6.3)  23  where  N_ is the population of the upper energy level, and N+ is the population of  the lower energy level. Transitions between these energy levels form the basis of NMR spectroscopy. Transitions between the nuclear spin energy levels can be produced by bringing each chemically different nucleus in the sample into its resonance condition. In practice, the resonance condition can be obtained by one of three methods: Either the field is swept at a fixed frequency, a range of frequencies is swept at a fixed field, or a band of frequencies around the Larmor frequency is simultaneously excited at a fixed 1/0 field. Fourier transform (FT) NMR uses the third method, where a short radio frequency (rf) pulse at the Larmor frequency, (Os, is applied to the system. The pulse is generally powerful enough to cover the entire frequency range of a given type of nucleus in the sample. The Boltzmann distribution of spin populations is thus disturbed from equilibrium, and the system tends to reestablish its equilibrium state through transitions between the energy levels. The process of returning to equilibrium state is a time dependent process involving both spin-lattice and spin-spin interactions, which are described by the spin-lattice relaxation time (T1) and the spin-spin relaxation time (T2) respectively. Following the rf pulse, an rf receiver is turned on to receive the signal from the nuclei. This signal, recorded as a function of time  f(t), is called the free induction decay (FID). The time domain function f(t)  and the frequency domain function F(co) are Fourier transforms of each other. Thus, the NMR spectrum, which is the frequency domain function  F(co), can be  obtained by a Fourier transformation of the FID. It is often convenient to describe FT NMR from the viewpoint of a rotating frame of reference, i.e. a reference frame coincident with z axis and rotating around z with angular velocity coo, where x', y', z' are used to represent the three axes of the frame. The behavior of spins can be described by simple vector 24  diagrams in a coordinate system rotating at the Larmor frequency con. The equilibrium distribution of spins leads to a net magnetization which can be represented by the M0 vector in the field direction along z (Figure 1.2A). Figure 1.2B shows the effect of an on-resonance rf field H1 which is applied perpendicular to the magnetic field direction. The H1 field creates a torque on the net magnetization M0, causing M0 to precess in the y'z' plane at a rate co = y H1. If the duration of the H1 pulse, t, is set to that cot = 7c/2, then the magnetization M0 is tipped by 900 into the x'y' plane (Figure 1.2C), where it is detected. Figures 1.2D and 1.2E show how the magnetization relaxes back to its equilibrium state by T1 and T2 relaxation processes. The decaying component of the magnetization in the x'y' plane is recorded as the FID signal. The return to equilibrium is shown in Figure 1.2F. The experiment can be repeated once equilibrium is reestablished. Because T1 T2, the relaxation time T1 controls the length of time required to obtain a FT NMR spectrum. The spin-lattice relaxation times (T1) for carbons in polymer solids are often very long (several minutes), while the corresponding T1 values for some polymers in solution are in the order of tenths of seconds.[76] The advantage of the pulsed Fourier transform NMR technique is primarily its efficiency. The data can be collected all at once, rather than from a slow sweep of the field. In addition, the technique is useful for improving the signal to noise ratio by data averaging. Many free induction decays from a weak signal can be repeatedly co-added in the computer. The noise is random and is therefore accumulated more slowly, while the signal is coherent and continually added. The signal-to-noise ratio increases as the square root of the number of accumulated spectra. The pulsed Fourier transform method also makes multipulse NMR experiments for various purposes possible.  25  x'  ( D)  ,  (E)  ,  Ho  4^  (F)  \ , H1=0^7Hi=0  Figure 1.2. Vector diagrams describing the pulsed NMR experiment in the rotating frame of reference. (x', y', z' axes are used to indicate the use of the rotating coordinate system, which rotates around z' at the Larmor frequency.) (A). The net magnetization  M0 is aligned the magnetic field direction 1/0;  (B, C). An rf field H1 is applied perpendicular to  Ho and tips  the net magnetization by 90'; (D, E). The magnetization begins to relax in the x'y' plane by spin-spin (T2) processes and in the z' direction by spin-lattice (T1) processes; (F). The equilibrium magnetization is reestablished along  Ho. 26  1.6.2. Dipolar Coupling in Solids and High Power Proton Decoupling In the NMR spectra of small molecules in solution, the dipole-dipole interaction usually can not be directly observed since rapid molecular tumbling averages this interaction to zero. In solids, however, the molecules are not free to tumble isotropically and the static dipole-dipole interaction generally causes peak broadening so large that all spectral details are lost. The dipole-dipole interaction arises from direct spin-spin coupling. For example, the magnetic dipole of one spin (1H, for example) influences the magnetic dipole of another spin (13C, for example), which results in a splitting of 13C spectral line. For a single crystal in which the coupled C—H pairs are faraway from each other, two peaks will be observed, whose position and separation depend on the orientation of the two interacting nuclei in the magnetic field. For powders, all orientations are present, and the observed pattern is a broad envelope of overlapping peaks The dipolar interaction between two spins can be written in the general form:  where and  b  XD = /1*-6/2^  (1.6.4)  xi, is the magnetic dipolar Hamiltonian, li and  12 are two spin vectors,  is the dipolar coupling tensor.[77] The dipolar Hamiltonian for two spin  species I and S can be written as:[781  XD = XII + XSS + XIS  ^  (1.6.5)  For abundant nuclei I, (eg. 1H), the major portion of this dipolar interaction is that between like spins, which is XII. For dilute nuclei S (eg. 13C in natural abundance), the major dipolar interaction is usually with the abundant and nearby protons, ie. Ytis. For a normal polymer sample, the 13C atoms in natural abundance are sufficiently diluted and far enough away from each other  27  to make the 13C-13C dipolar interaction ;s very weak and hence negligible in most cases. Mathematically, the heteronuclear dipole-dipole interaction, Xis, has the form[781 hN1 Ns _3  EEr  Xis = – YiYs ( — )2 27r^k  ^- 1)/iz/kz  (1.6.6)  Where r is the distance between the nuclei and 0 is the angle between the vector connecting the two interacting nuclei and the external magnetic field Ho. The dipolar interaction strongly depends on the internuclear distance r. The dipolar coupling for a set of magnetically isolated 13C-1H spin pairs at a certain angle, 0, relative to H0 results in a splitting in the 13C spectrum (corresponding to the two allowable proton spin states) given by:  AED Yc YH 271  (3COS20 - 1)/r3  (1.6.7)  This is shown schematically in Figure 1.3B. The variables in Equation (1.6.7) are defined in Figure 1.3A. However, for a normal powder sample, the various C–H vectors have all possible angles with respect to the external magnetic field. For a single type of magnetically isolated C–H vector in such material, the complete range of dipolar couplings is expected, which produces the Pake doublet pattern[79] (Figure 1.3C). Actually, the individual C–H vectors are not magnetically isolated and many additional dipolar interactions can also occur. This produces an inhomogeneously broadened line with a Gaussian shape (Figure 1.3C). Therefore, the static dipole-dipole interaction in solids causes a large line broadening, which must be removed in order to obtain high resolution NMR spectra of solids.  28  A  Figure 1.3. Schematic representation of dipolar interactions. (A). Dipolar interaction between a spin pair of 13C and 1H nuclei. t are the z components of the magnetic moments. (B). Dipolar splitting of isolated C—H pairs at one angle relative to the magnetic field. (C). Pake pattern expected for isolated C—H spin pairs distributed at all angles in a powder sample. (D). Approximate Gaussian line shape observed for non-isolated C—H spin pairs, where all dipolar interactions are operative. (Reproduced from reference [78])  29  Equation 1.6.6 for is shows the spatial and spin parts separately. If either of these parts can be averaged to zero by some coherent process, the effect of Xls can be removed from the spectrum. One way to remove the dipolar coupling involves modulation of the spatial part in Equation 1.6.6 by mechanically spinning the sample at 6 = 54.74°, the magic angle, which reduces the (1-3cos2e) term to zero[80] (see on). The other option is to average /igkz the spin part of Ytis, to zero. This can be achieved by high power dipolar decoupling. Dipolar decoupling is accomplished by applying an additional rf field to the solid sample at the proton Larmor frequency, as is done to remove the scalar J-coupling in solution. For solids, however, the amplitude of the decoupling field must be larger compared to the static dipole-dipole interaction. Instead of the 1 gauss field used in solution NMR spectroscopy, an approximately 12 gauss (corresponding to approximately 50 kHz) field is required to remove the effect of the C—H dipolar interaction from solid state 13C NMR spectra. Thus, high resolution solid state NMR spectroscopy requires amplifiers capable of providing sufficient power for dipolar decoupling, and rf probes capable of withstanding these high power levels. The effect of dipolar decoupling on solid state NMR spectra is illustrated in Figure 1.4.[81] All three 13C NMR spectra are of bulk poly(butylene terephthalate). The top spectrum is obtained under the conditions used for a solution NMR experiment, whereas in the middle spectrum the dipolar interactions with the protons are decoupled. The spectrum in Figure 1.4B still does not exhibit the high resolution associated with solution state 13C NMR spectra. Although the static dipole-dipole interactions with the protons have been removed in this spectrum, the lines are still broad, primarily because of chemical shift anisotropy. The bottom spectrum (Figure 1.4C) combines magic angle spinning (MAS) with high power proton decoupling. (MM averages all the 30  i epsoonouniesi Ole 1^ 300^  \wwwi werf Ii004,00 1^1^1  100^  ppm FROM TMS  -100  Figure 1.4. The effects of dipolar decoupling and magic angle spinning on 13C solid state NMR spectra of poly(butylene terephthalate). (A). Spectrum obtained using low power decoupling. (B). Spectrum obtained using high power (dipolar) decoupling. The primary source of line broadening in this spectrum arises from the chemical shift anisotropy. (C). Spectrum obtained using dipolar (high power) decoupling in addition to magic angle spinning (MAS). (Reproduced from reference [81]). 31  overlapping chemical shift anisotropies to their isotropic values, as will be discussed in the following section.)  1.6.3. Chemical Shift Anisotropy and Magic Angle Spinning The chemical shift arises because the electrons in a particular atom interact with the magnetic field,  Ho. The external magnetic field induces electric  currents in the molecule, and these currents produce a local magnetic field HL at the nucleus. Thus, the nucleus no longer experiences only the external field 110, but  Ho+HL, which alters the resonance frequency of the nucleus. The chemical  shift tensor  a describes the orientation and the magnitude of this three  dimensional local field. In general, the chemical shift interaction has the formrn  h  5€0- = y — IAA) 2TE  (1.6.8)  where 6 is the chemical shift tensor and Yea is the chemical shift Hamiltonian. Equation 1.6.8 can also be simplified as[78] h ,ea = Y — azz 1z Ho 211  (1.6.9)  where azz is the projection of the chemical shift tensor onto the z axis, which can be expressed in terms of the diagonal elements of a, aii, and the respective direction cosines, Xii, with respect to  Ho:  2^2^2 ^ (1.6.10) azz = a11X11+ a22A22+ a33X33  The principal values of the chemical shift tensor describe the magnitude of the tensor in three mutually perpendicular directions in the molecule, and the  32  direction cosines specify the orientation of the tensor with respect to the external field. In solution, molecules tumble freely and isotropically, averaging the chemical shift tensor to its isotropic value ciso. Thus, the isotropic chemical shift is the only part of the chemical shift tensor observed in solution state NMR. However, motion in solids is limited, and a broad anisotropic chemical shift pattern is usually observed for a solid sample. The chemical shift observed in solution state NMR, or the isotropic chemical shift, is one-third of the sum of the diagonal elements of the tensor: 1,  cis° = 5 kan + a22 + a33)  (1.6.11)  Equations (1.6.9) and (1.6.10) indicate that the chemical shift of a particular nucleus in the solid state depends upon the orientation of the molecule with respect to the magnetic field. For a single crystal in absence of other interactions, a single and relatively narrow signal is observed whose position depends on the orientation of the crystal with respect to the magnetic field 110. This can be seen schematically in Figure 1.5A. For a powder sample, rather than a single crystal, all crystallite orientations are present and the resultant NMR spectrum shows a characteristic chemical shift anisotropy powder pattern. Figures 1.5B and C illustrate two theoretical chemical shift anisotropy patterns. For carbon monoxide, the axial symmetry of the molecule is reflected in a pattern as shown in Figure 1.5B.[82] The position of  all  or al  represents the observed resonance frequency when the principal axis system is parallel or perpendicular to the field direction. Figure 1.5C shows a generalized chemical shift anisotropy powder pattern for the more common nonsymmetrical case. The isotropic chemical shifts ais0=1(1 ,--,/3,)( ,a11 +sr:322+033)i are indicated by the dotted lines in Figure 1.5.  33  Ho  CEO  I ^1 ,^ 1 I 1  A^1  B  I  0 HI C  I 1  1  1 I 1 I 1 I  1  C  Figure 1.5. Schematic representation of chemical shift anisotropy powder patterns. (A). Chemical shifts observed for two orientations of a carbon monoxide single crystal relative to the static magnetic field. (B). Axially symmetrical chemical shift anisotropy powder pattern would be observed for axially symmetrical molecules such as polycrystalline carbon monoxide. (C). General anisotropic chemical shift anisotropy powder pattern for nonsymmetrical molecules. The dotted lines represent the isotropic chemical shifts, oiso. (Reproduced from reference [78]) 34  Molecular motion causes a partial narrowing of the chemical shift tensor powder pattern. The exact shape of the motionally narrowed chemical shift pattern contains information concerning the axis and angular range of the motion. In general, carboxyl, carbonyl, and aromatic carbons have the largest anisotropies I ail — a33I, or powder pattern widths (approximately 180 — 250 ppm). The anisotropies for methyl, methylene, and methine carbons are usually less and are in the order of 30 —60 ppm. The chemical shift anisotropy contributes a line broadening in solid state NMR spectra that often obscures the structural information available from the isotropic chemical shift. In addition, powder data give only the principal components not their orientations. For example, less structural information can be deduced from the spectrum in Figure 1.4B than from the spectrum in Figure 1.4C. For this reason, high resolution solid state NMR spectra are usually obtained using magic angle spinning to reduce the chemical shift anisotropy and obtain the isotropic value. Magic angle spinning (MAS) is carried out by mechanically spinning a sample about an axis making the "magic angle" (0 = 54.7°) with respect to the magnetic field direction (Figure 1.6). MAS is not a new technique. In 1959 Andrew et al.[831 used it to narrow the 23Na line in NaC1, and in the same year Lowe independently used MAS to narrow dipolar broadened 19F lines in CaF2 and Teflon[84]. These early attempts to remove dipolar interactions with MAS were not totally successful as it was impossible to spin the sample rapidly enough to remove the large homonuclear dipole-dipole interactions. However, Schaefer and Stejskal (1976) showed that if one removed the static dipolar interaction by high power proton decoupling, the remaining chemical shift anisotropy could be averaged to its isotropic value using mAs.[66c]  35  A  54°44' \ 90°^180° 9  Figure 1.6. Schematic representation of magic angle spinning. (A). The geometric arrangement for mechanical sample spinning. (B). Variation of the term (3cos20-1) as a function of 0. The curve crosses the axis at 0=54°44', the "magic angle". (Reproduced from reference [65]) Under rapid mechanical sample rotation about an angle 0 with respect to the magnetic field direction (Figure 1.6), the direction cosines in Equation (1.6.10) become time dependent. Therefore, the chemical shift Hamiltonian can „avg  be divided into a time-independent part, it cy , and a time-dependent part, 0t): [78] vg  Xcy^a + Za(t)  (1.6.12)  Taking the time average under fast sample rotation, only the time-independent , Avg  part, if^, will be left. 1^0^1 evg = y 11— /z1/0 [ — sin443 (011 +022+c733) + (3cos20 —1) 271^2 x(functions of direction cosines)]^(1.6.13)  36  The angle 0 in Equation (1.6.13) is the angle that the rotation axis makes with the static magnetic field direction (Figure 1.6). For 0 = 54.7° (the magic angle), the (3cos20 — 1) term becomes zero and sin20 is 2/3. The first term in Equation (1.6.13) becomes one-third of the trace of the tensor, ie. the isotropic chemical shift. Using MAS, the chemical shift powder pattern is thus reduced to its isotropic average value. As mentioned in the previous section, MAS can also reduce the dipole-dipole interaction. The effect of MAS on the 13C NMR spectrum of the solid, poly(butylene terephthalate), is shown in Figure 1.4C. In this spectrum the broad overlapping carbonyl and aromatic resonances in Figure 1.4B have been reduced to their isotropic averages, and a truly high resolution spectrum is obtained.  1.6.4. Relaxation Times As in conventional FT-NMR spectroscopy of liquids, spin relaxation times also play an important role in obtaining high resolution NMR spectra of solids. The most fundamental relaxation times, which are important for both solution and solid state NMR spectroscopy, are the spin-lattice relaxation time (T1) and spin-spin relaxation time (T2). T1 characterizes the regrowth of the magnetization back to its equilibrium value along the direction of the static field after it has been perturbed by a radio frequency pulse. It dictates the repeat time which may be used for signal collection. The spin system must couple to its surroundings, or lattice, in order for T1 relaxation to occur. In this process the spin system gives its excess energy to its surroundings, or lattice. This nonradiative process of spin relaxation occurs via a modulation of local magnetic fields by molecular motion at the proper frequency. In polymer systems, the local fields usually arise from the proton magnetic moments. Many important motional processes in solids, particularly in polymers, have characteristic 37  frequencies in the range of tens of kHz, which are not high enough to produce efficient T1 relaxation under normal circumstances. The spin-lattice relaxation is most efficient when the correlation frequencies of these motions are near the Larmor frequency, typically in several to several hundreds of MHz range. Since Larmor frequencies which are effective for T1 relaxation are determined by the strength of the static magnetic field, one way to match the Larmor frequencies to the frequency of molecular motion in tens of kHz range is to perform the NMR experiment in a very low magnetic field. However, this approach is not practical because NMR sensitivity and resolution decrease as the magnetic field is lowered. In general, each type of nucleus has its own T1 (protons are an exception, see later). T1 values are different for different nuclei because of differences in Larmor frequencies, motions and available relaxation mechanisms. Because different nuclei can be coupled by various interactions, their relaxation times may not be independent. As described previously, the isotopically abundant and nonisolated nuclei (eg. 1H) are strongly coupled in solids. The relaxation times of different chemical types of these nuclei are often averaged by spin diffusion, or mutual spin flips among strongly coupled nuclei. Therefore, the protons in a solid polymer usually all have the same T1 value. In addition, the proton T1 is generally shorter than most of the carbon T1 values.[851 Nonequilibrium magnetization in any part of the proton spin system is transferred to surrounding protons in times on the order of 100 !Is by spin diffusion.[781 As expected, solids, with less molecular motion in the megahertz frequency range, often have long 13C T1 times. Even though the 13C in a methyl group which generally rotates rapidly often has relatively short T1 time, the T1 times for other carbons can be extremely long. The 13C in the rotating methyl group rarely exchanges its short T1 time with the other carbons in the sample 38  via spin diffusion, because the natural abundance of 13C is very low and the distance between the 13C nuclei is very long on average, and thus, the coupling between the 13C nuclei is very weak. In addition, the methyl carbon cannot transfer its short T1 to its nearby protons, and then from the protons back to other carbons, because the carbon and proton Larmor frequencies are very far apart and thus there is no overlap in the carbon and proton energies (Figure 1.7). On the other hand, the protons on the methyl group also have short T1 values, and they can transfer this short T1 to all the other protons via spin diffusion, as the protons are essentially 100% naturally abundant and close enough to strongly couple to each other.  1H 13c _ A 0^200^400 MHz  Figure 1.7. Larmor frequencies for 1H and 13C in a 9.4 T magnetic field. There is no frequency overlap, and thus no overlap in the energies.  Spin diffusion is not an actual molecular diffusion but the transportation of spin energy within the spin system by mutual, energy conserving spin flips. Two adjacent protons in a solid strongly coupled by the dipolar interaction essentially have the same resonance frequency even if they are chemically different because of their large line widths caused by strong dipolar coupling. If the two have antiparallel magnetic moments to each other, it is an energetically favorable process for both moments to change orientation simultaneously or "flip", again yielding an antiparallel pair but with magnetic moments reversed. Such flips can occur rapidly among neighboring, coupled protons in a solid and 39  serve to distribute excess energy or magnetization among all the coupled protons. This process, like any diffusion process, is driven by a concentration gradient, which is the spatial gradient of the magnetization in this case. It is spin diffusion that ensures that the 1H linewidth is homogeneous. Because of spin diffusion, the whole proton spin system is coupled to the lattice via the most efficiently relaxing portions of the system. Energy transfer to the lattice is very efficient near paramagnetic impurities, lattice defects, and molecular segments in rapid motion, such as methyl side chains or possibly end groups. When x and y components of g magnetization are created by a rf pulse, the decay of these components is characterized by T2. There is no energy exchange associated with this decay and it is a pure entropy process. This decay occurs because all the spins do not precess at the exactly same rate and tend to get out of phase. The linewidth of a NMR spectrum, v112, is related to T2 by: 1 V1/2 -  TCT2  (1.6.14)  for a Lorentzian line. Molecular motion at the Larmor frequency affects T2; low frequency motion, on the order of 100 — 1000 Hz, can also affect T2.[78] Inhomogeneity in the static field also appears to affect T2, because the spins are in different fields and hence get out of phase more rapidly. However, for high resolution solid state NMR this last factor is usually negligible in a well shimmed magnet. For polymers in both solution and solid state, Ti. is usually much greater than T2, T1 » T2.  1.6.5. Cross Polarization (CP) In addition to removing the strong dipolar interaction and averaging chemical shift anisotropy, one other difficulty must be overcome before high-  40  resolution solid state 13C NMR spectra can be obtained in an efficient way. As described previously, T1 characterizes the return of the spin system to its equilibrium state after being perturbed by a rf pulse, and controls the rate at which the experiment can be repeated. For rare nuclei, such as 13C, which require more signal accumulating to obtain their spectra, the repetition rate becomes very important. Since the relaxation times of rare nuclei 13C are normally not short, the signal accumulation time can be very long. However, this problem can be circumvented by transferring magnetization from the abundant 1H nuclear spins to the rare 13C nuclei under observation. Through this process, the repetition rate for 13C signal collecting is now determined by the shorter T1 of the 1H nuclei. The process of magnetization transfer from abundant spins to rare spins is termed cross polarization (CP) and was first introduced by Pines et al.(861 Although 13C and 1H have Larmor frequencies different by factor of four (Figure 1.7), Hartmann and Hahn in 1962[87] demonstrated that energy may be transferred between them in the rotating frame. Thus, energy transfer between nuclei with very different Larmor frequencies such as 1H and 13C can be achieved if the following "Hartmann-Hahn condition" is matched:  7C H1C = YH H1H (1.6.15) Equation 1.6.15 results in a "match" of the rotating frame energies for 1H and 13C. Since yif is four times of 'ye, it is necessary for the match that the strength of the applied carbon field (Hic) is four times the strength of the applied proton field (HIH). Cross polarization gives increased S/N (the signal to noise ratio) in a given time period.  41  The vector diagrams for this experiment are shown in Figure 1.8, and the pulse sequence is shown in Figure 1.9. The vector diagrams in Figure 1.8A show the proton and carbon spin systems equilibrated in the magnetic field. A 90° rf Pulse (Hm) is applied to the proton along the x' axis and brings the proton magnetization along the y' axis (Figure 1.8B). Then, a long proton pulse(HiH) with its phase shifted by 900 locks the proton magnetization along the y' axis (Figure 1.8C). The strong Ili ji spin-locking pulse force the proton spins to precess around the y' axis of their rotating frame with a frequency co -H = Yll H1H • Magnetization decay processes occur during this time[66e1. Meanwhile, the carbons are exposed to a long carbon pulse I/1c along the direction of the spinlock field (y' axis), which causes the carbon magnetization to precess along the I/1c field with the frequency coc = Tc Hic. If coH and (pc are equal, an energy exchange between both nuclei becomes possible, which causes the 13C magnetization to grow along the Hic field (Figures 1.8C and 1.9). At this point, the cross polarization process can be described in more detail in Figure 1.10. When the Hartmann-Hahn condition (Th matched by adjusting the power levels of the  Hai  Iiiii  = 7c Hid is  and H1c fields, the z-  components of both the 1H and the 13C magnetizations have the same time dependence (Figure 1.10) as the two precession frequencies are the same, con = O. Because the z-component time dependence is common to both spin systems, mutual spin flips can occur between the 1H spins and the 13C spins. This process can be visualized as a "flow" of magnetization from the abundant proton spins to the rare 1-3C spins. An alternative way to visualize polarization transfer is from a spin temperature or thermodynamic point of view. This concept has been described in detail by Pines et al.[881  42  1H^z'^z'  (1)spin lock ^z' protons along y axis  90,  (2)protons x precess around y' axis  c°H =  (1) apply rf z' power /-iic along y axis ^> x' (2)13C magnetization^_ grows up^(pc — Yc Hic along y' axis  13C^z'^  (3) 13C spins precess A^B^around y' axis  Figure 1.8.  Vector diagram for a 1H  —  13C cross-polarization experiment.  The carbon reference frame and the proton frame are rotating at different frequencies wc, coH. (Reproduced from reference [81]) 1H 90° (SPIN LOCK)y,  lac  DECOUPLE  ALLOW PROTONS TO RE-EQUILIBRATE  CONTACT I^TIME OBSERVE FID  WidT TIME  ^a. TIME  Figure 1.9. Schematic representation of the 1H--13C cross-polarization pulse sequence for solid state NMR experiments. 43  Figure 1.10. A more detailed representation of part C in Figure 1.8. The 13C spins and the 1H spins are precessing about the direction of the spin-lock field with frequency coc=ycHic, and coH=ThHill respectively. When the Hartmann-Hahn condition is matched (coc = o3H), the two spin systems have zcomponents with the same frequency dependence. Thus mutual spin flips can occur between 1H and 13C. (Reproduced from reference [81])  The advantages of cross polarization are twofold. Firstly, as mentioned previously, it circumvents the problem of the long carbon Ti. values normally found in solids. The 13C nuclei obtain their magnetization from the protons, and thus it is the proton Ti. which controls the cross polarization experiment repetition rate. Secondly, the 13C shows an enhancement in its signal intensity, which can be as large as the ratio of yH / yc, (a factor of nearly four). Thus, cross polarization saves experimental time by reducing the waiting period as well as by improving the signal to noise ratio. Figure 1.11 demonstrates the advantage 44  B  C  D  I;H-3-1 CH2C=0 OCH3 CH3  Figure 1.11. A demonstration of the advantages of combining DD, MAS, and CP techniques for obtaining 13C solid-state NMR spectra of poly(methyl methacrylate). (A). Stationary sample; no cross polarization; low power decoupling. (B). Stationary sample; cross polarization; high power decoupling. The spectrum shows the effects of chemical shift anisotropy. (C). Magic angle spinning; no cross polarization; high power decoupling. The spectrum shows "high-resolution" isotropic shifts but the signal to noise ratio is low as cross polarization is not used. (D). Complete experiment: cross polarization; magic angle spinning; and high power decoupling. The spectrum shows similar resolution as (C) but the signal to noise ratio is greatly enhanced by the use of the cross polarization technique. (Reproduced from reference [891) 45  of combining dipolar decoupling (DD), magic angle spinning (MAS) and cross polarization (CP) techniques for a typical glassy polymer, poly(methyl methacrylate) (PMMA). Comparison of C and D clearly illustrates the value of cross polarization. [89]  1.7. Purpose of this Thesis Research As briefly described above, polymer resin systems are normally complex heterogeneous materials. The final resins are usually multiphase, amorphous and insoluble solids.  Resin Reaction Sites I Cross Reaction Sites  Zz Glass /Surface  Coupling Agents  End Groups  I  L  Cross Linking Agents  Figure 1.12. Schematic representation of some important active sites in a heterogeneous polymer resin system  Figure 1.12 shows a typical heterogeneous polymer resin system. The higher the degree of cross-linking in the polymer system, the more the 46  individual molecules are immobilized, and thus, the greater the strength and stability of the polymer matrix. In many cases, the mechanical, physical and chemical properties of a polymer are determined by the nature and the degree of the reaction for a very small number of functional groups in the polymer system or in added molecules. The interference from the bulk polymer matrix makes it very difficult for most of analytical techniques to study the reaction of these small number of functional groups (eg. the different reaction sites, the end groups, the coupling agent, the cross linking agent or the curing agent in the polymer system, as shown in the Figure 1.12). The NMR signal of an isotopically enriched element will be many times higher than the signal of the corresponding low natural abundance element. For 13C NMR, the signal enhancement from isotopic enrichment can be up to 99 fold (the natural abundance of 13C is 1.1%). For 15N NMR, even higher increases in signal intensity may be obtained, since the natural abundance of 15N is only 0.37%. By using specific isotopic enrichment in a functional group, e.g. in a coupling agent or a cross-linking agent, we can greatly reduce the contributions from the nuclei in the sample bulk. By subtracting the spectrum of a corresponding sample with no isotopic enrichment, the spectrum of only the enriched group may be obtained since the NMR spectrum has greatly enhanced intensity for the enriched nucleus in the functional group. It is thus possible to unambiguously trace the enriched functional group during a polymerization process and characterize its product species. The effect of isotopic enrichment can be demonstrated in Figure 1.13, which shows 13C CP/MAS spectra for a cured phenolic resin.[901 The top spectrum was obtained from a sample with natural abundance. The importance of using isotopic enrichment can be seen by comparison of both spectra. In  47  spectrum B, using only about 5% 13 C enriched formaldehyde makes the reaction products from the formaldehyde much more clearly distinguishable. By the use of solid state NMR spectroscopy combined with specific isotopic enrichment, it should be possible to efficiently study the properties and the reaction mechanisms of the small number of very important chemical groups which determine the physical, mechanical, and chemical properties of a polymer. Several projects are included in this thesis. Each of them corresponds to the investigation of one of the different reaction sites or different agents in heterogeneous polymer systems as shown in Figure 1.12 with particular emphasis on the case of cyanate resin related system. They will be discussed in detail in the following chapters.  48  CH2  200  160 0  PPM from TMS  Figure 1.13. The effect of selective isotopic enrichment. (A). Solid-state 13C NMR spectra obtained at 22.6 MHz from a cured phenolic resin (phenol/formaldehyde/sodium hydroxide 1/2/0.01, cured at 110 °C for 24 hours), magic-angle spinning at 3.6 kHz; (B). conditions as in (A) except that the sample was prepared by using formaldehyde 13C enriched to —5%. The small peaks marked s denote spinning sidebands. (Reproduced from reference [90])  49  CHAPTER 2. INVESTIGATIONS OF THE CURING REACTIONS OF THE CYANATE RESIN SYSTEM Conventional electronic circuit boards are made from glass-fibre reinforced epoxy resins. However, they often fall short of the thermal and electrical performance demands of many modern high speed devices. With current trends toward increased circuit densities, shorter propagation delays, elevated operating temperatures, and higher reliability, new advanced materials are being developed to satisfy these demands. Among these materials, cyanate resins are considered to be very promising systems. As described in the previous chapter, the cyanate resins are commonly derived from the bisphenol type of monomer. The cured resins exhibit good thermal, mechanical and insulating characteristics and have been considered for many electronic packaging and structural materials applications. [17,27,30] Bisphenol A dicyanate (BPADCN, 2) is one of the cyanate resin monomers most often used. The curing reaction is postulated to proceed by reaction of three cyanate (—OCN) groups on different molecules to form a triazine ring as shown in Scheme 1.1. Although the reactions of some model systems have been studied and IR investigations have been carried out to identify some of the functional groups during curing, [16,22] there is little direct evidence to date regarding the species involved, or the nature and efficiency of the curing process itself. The purpose of this project was to carry out an investigation of the curing reactions of these resins by high resolution 13C and 15N NMR spectroscopy both in solution and in the solid state. To facilitate the study and to clearly monitor the reactions of the functional groups during the curing process, not only the natural abundance dicyanate monomer 2, but also the 13C and 15N enriched 50  dicyanate monomers, 2a and 2b, were used. This greatly enhances the contribution of the cyanate group and its reaction products to the NMR spectra. In addition, to provide reference spectra for the identification of reaction products, the reactions of the analogous monocyanates, p-tert-butylphenyl cyanate (PTBPCN) in natural abundance (25) and with specific isotopic enrichment (25a) and (25b), were also studied. Since the monocyanates can not form a cross-linked network during triazine formation, the better solution NMR spectra can be obtained.  OCN  NCO  2  N13C0  2a  013CN^15NC0 ^  OCN  25  ^  0C15N  2b  013CN  25a  ^  0C15N  25b  2.1. Syntheses and Characterizations of Specifically Labeled Cyanate Monomers The 13C and 15N enriched cyanate monomers were synthesized by a twostep procedure. First, the labeled cyanogen bromide (Br13CN or BrC15N) was made by using either K13CN or KC15N as the labeled starting material [Equation 2.1], following a similar procedure to that of Hartman and Dreger[91].  51  K13CN^+^Br2 ^..- Br13CN + KBr ^[2.1] KC15N^+^Br2^BrC15N +^KBr The labeled cyanogen bromides were then reacted with the appropriate phenol to produce the corresponding labeled cyanate compounds [Equation 2.2].  OH + BrC*N + Et3 N  ^MM.  OC*N + Et3NHBr  [2.2]  BrC*N = Br13CN or BrC15N The 13C and 15N enriched dicyanate monomers 2a and 2b were prepared from bisphenol-A by reaction with labeled cyanogen bromide as indicated in Equation 2.3. However, before using labeled materials, the preparations were carried out with unlabeled reagents, and a small scale reaction suitable for producing appropriate quantities for NMR studies was optimized. The synthesis procedures are modifications of literature methods[602] and are given in detail in the Experimental chapter at the end of the thesis.  013CN  N13 CO  Br13Cyl HO  2a  OH^  [2.3]  3 0C15N  2b  52  Figures 2.1A and B show the 13C solution NMR spectra (reference: TMS) with assignments for natural abundance and 13C enriched BPADCN monomers, 2 and 2a, respectively. The large increase in the intensity of the peak at — 109 ppm due to the —OCN carbon indicates that it is possible to follow its reaction in the curing process by comparison of spectra from the labeled and the unlabeled materials. Figure 2.1C shows the 15N solution NMR spectrum of the 15N enriched BPADCN monomer 2b. The resonance for the nitrogen in the —OCN group is clearly observed at — 53 ppm (reference: neat formamide). The natural abundance signal is not detectable at all under these conditions. The 15N spectra during curing will thus show signals only from the cyanate group and its subsequent reaction products. The labeled monocyanates, 25a and 25b, were synthesized by the reaction of p-tert-butylphenol 26 with labeled cyanogen bromide according to Equation 2.4. This reaction is a modified literature procedure{931 and is described in detail in the Experimental chapter. The corresponding unlabeled compound 25 was also synthesized by the same method in larger quantities.  01 3C N  Brl 3CN OH  25a  [2.4]  BrC15N  26  0C1 5 N  25b Figure 2.2A shows the 13C NMR spectrum of the natural abundance monocyanate PTBPCN 25 together with the assignment. The corresponding spectrum of the 13C labeled material 25a, given in Figure 2.2B, indicates the  53  c2  c6 c3 c2  c7 OCN  c6  cl c4  I  c5  c7 I^I  B  c7 —0C*N  CDC13  1 150^100  i  ^T  50  ^  PPM  C  ..r...••••••wirsomeaLawa..............eorolmarroauqm.sl  r^  80^  I  40^  1  -10^-40  PPM1  Figure 2.1. Solution NMR spectra 0-H at 300 MHz) of bisphenol A dicyanate (BPADCN) monomer in CDC13. (A). 13C NMR spectrum of natural abundance monomer 2; (B). 13C NMR spectrum of the 13C enriched monomer 2a; (C). 15N NMR spectrum of the 15N enriched monomer 2b. 54  c6^A  c6 c3 c2 m2 ml m3 m4 C1-13COCH2CH3 c3 c2  a'2 a'l CD3C0CD3  c5 m2^m4 m3  ^cl c4 ml^ a'l li I  c7  a'2  c7^  c6 c3 c2  B  c7 OC*N  J  200^160^120^80^ao^0 PPM  Figure 2.2. 13C solution NMR spectra (1H at 200 MHz) of (A). natural abundance p-tert-butylphenyl cyanate (PTBPCN, 25) and (B). 13C enriched PTBPCN 25a in methyl ethyl ketone (MEK) and acetone-d6.  55  successful introduction of 13C into the cyanate group which gives the resonance line at 8 = 109 ppm. Figure 2.3A shows the 15N spectrum of the 15N labeled cyanate 25b. There is a single sharp resonance at - 53 ppm consistent with the BPADCN spectra. The spectrum obtained without decoupling shows that there are no coupled protons (Figure 2.3B). The 13C spectrum (Figure 2.3C) shows a splitting of the -OCN resonance, C7, into a doublet due to the coupling to the 15N nucleus which is consistent with the 15N spectrum and confirms the introduction of the 15N nucleus into the cyanate group. The coupling constant in the cyanate group is JC-N = 11.6 Hz.  2.2. Investigation of the Curing Reaction in Solution Because of its inherent higher resolution, solution NMR was first used to characterize the species involved in the curing reaction process, including reactants, main products, side products and any possible intermediates.  2.2.1. 13C NMR Investigation of the Curing Reaction of BPADCN in Solution The progress of the curing reaction in solution was investigated for the 13C enriched BPADCN monomer by high-resolution 13C NMR at 50 MHz (proton frequency of 200 MHz) with proton decoupling. The solvent used was 2-butanone (also called methyl ethyl ketone, MEK) with a small amount of acetone-d6 added to provide a deuterium lock signal. Zinc octanoate (200 ppm) was added as a catalyst, and in different experiments the solution was heated at 60 °C, 90 °C and 100 °C for various times in sealed glass tubes. Representative spectra are given in Figure 2.4. Figure 2.4A shows the 13C spectrum of the BPADCN monomer before reaction. The three large signals at high-field are due to the MEK solvent. The intense signal at 8 = 109 ppm is due to the enriched -013CN group. This resonance and those derived from it are used to monitor the progress 56  1  ^,^-,-^,^/^,^-,^.^.-^i^r^  200  ^  ^T^I^I  -'^T ^ ^ PPM 50 150^100  Figure 2.3. Solution NMR spectra (1H at 300 MHz) of 15N enriched PTBPCN 25b in acetone-d6. (A). 15N spectrum with 1H decoupling; (B). 15N spectrum without 1H decoupling; (C). 13C spectrum with 1H decoupling.  57  Figure 2.4. 13C solution NMR spectra (1H at 200 MHz) of 13C enriched BPADCN 2a in MEK and acetone-d6 cured with 200 ppm zinc octanoate as catalyst. (A). Before heating; (B). After heating for 1 hour at 60 °C; (C). After 16 hours at 60 °C; (D). After 5 days at 60 °C.  58  of the curing reaction. The other signals in the spectrum are relatively small and agree with those of the unlabeled monomer previously obtained (Figure 2.1B). After just one hour heating at 60 °C, two new resonances at 156 and 174 ppm (labeled a7 and t7, respectively) are observed (Figure 2.4B) which grow in intensity with time (Figure 2.4C and D). These are assigned to the expected triazine (8 = 174 ppm, t7) and one other species (8 = 156 ppm, a7). The nature of this second species will be discussed in more detail later. Subsequent spectra recorded after longer heating periods show the same two major resonances but their intensities diminish relative to those of the solvent signals, although the cyanate resonance of the monomer disappears at, the same time. This is because high-resolution NMR spectroscopy only detects those species present in solution. With the curing process taking place while the solution spectra of Figure 2.4 were obtained, considerable quantities of solid material had precipitated from solution. The precipitates are considered to be cured cyanate resin which contains mainly the triazine moiety (8 = 174 ppm) and some carbamate moiety (8 = 156 ppm) which will be discussed later.  2.2.2. 15N NMR Investigation of the Curing Reaction of BPADCN in Solution A series of 15N spectra 0-H, 300 MHz) were obtained under identical conditions to those of the 13C spectra discussed above and are presented in Figure 2.5. These spectra are particularly informative as the only signals observed are from the -0C15N group and its reaction products, with no interference at all from solvent or unlabeled monomer resonances. Figure 2.5B shows that a single sharp signal at - 53 ppm as expected and two additional resonances are observed at -40 ppm and 87 ppm during curing, in general agreement with the results from the 13C NMR spectra. The signal at 87 ppm can  59  B  -0--C-N*H2  II  o  1  1  80  i  1  O  1  . -40 PPM  Figure 2.5. 15N solution NMR spectra (1H at 300 MHz) of 15N enriched BPADCN 2b in MEK and acetone-d6 with 200 ppm zinc octanoate as catalyst. (A). Before heating and without 1H decoupling; (B). After heating at 90 °C for 1 day and with 1H decoupling; (C). Same sample as in (B) without 1H decoupling.  60  be assigned to the expected triazine ring. Figure 2.5C was obtained without proton decoupling during acquisition. The signal at —40 ppm shows a triplet structure indicating that it is coupled to two protons while the others are unaffected as expected. These general characteristics are maintained during further curing (spectra not shown), although again a considerable amount of solid material has precipitated and only soluble species are detected in the solution NMR spectra. The second product species is postulated to be the carbamate compound 27 formed by the addition of water to the cyanate group as shown in Equation 2•5.[6b]  9  OC*N + H20  0_C^H2  [2.5]  27  This structure fits all of the characteristics of the 13C and 15N NMR spectra, particularly the coupling of the 15N nucleus to two protons. Further confirmation of the structure of this species is given below. One very important feature of the spectra is that they rule out the formation of substantial amounts of any long lived intermediate "dimer" species on the route to triazine ring formation. These species would all have shown two resonances in both their 13C and 15N spectra. Thus the curing reaction appears to be remarkably clean!  2.3. Reactions of Monocyanate Model Compounds 2.3.1. 13C NMR Investigation of the Curing Reaction of PTBPCN 25a in  Solution As a complement to the solution NMR studies of BPADCN discussed above, the solution reactions of the analogous monocyanate were investigated,  61  with particular emphasis on the identification of the reaction products and the effect of added water. As indicated previously, in the case of the monocyanate, no cross-linking polymerization can occur and there will be no precipitation from the reaction mixture. Figure 2.6A shows the 13C spectrum of the 13C enriched monomer PTBPCN 25a in MEK solvent, together with its assignment. The large resonance at 109 ppm is due to the cyanate group (c7). An excess of water and 200 ppm zinc octanoate were added to the system and it was allowed to stand for 24 hours. The —OCN resonance is reduced and a new signal appears at 156 ppm (Figure 2.6B), which is consistent with the behavior of the dicyanate and assigned to p-tert-butylphenyl carbamate 28 in the present instance [Equation 2.6].  [2.6]  On heating the sample at 100 °C for 1 hour (Figure 2.6C), the intensity of this signal increases and small peaks appear at ö = 174 ppm (triazine, t7) and 171 ppm (unassigned). After prolonged heating at 100 °C, the triazine resonance is the major component in the spectrum. Further information on the nature of the other reaction products comes from the 15N spectra (see on).  2.3.2. 15N NMR Investigation of the Curing Reaction of PTBPCN 25b in  Solution An investigation of 15N spectra of PTBPCN in MEK and acetone-d6 gave results in agreement with the 13C data of the previous section. The experiments were carried out with the 15N enriched PTBPCN 25b whose spectrum was 62  Figure 2.6. 13C solution NMR spectra (1H at 200 MHz) of the 13C enriched PTBPCN 25a in MEK and acetone-d6 with 200 ppm zinc octanoate added as a catalyst. (A). Before the addition of water; (B). After addition of excess water and standing at room temperature for 24 hours; (C). After heating at 100 °C for 1 hour. 63  c7 c6 c3 c2  A m2 ml m3 m4 CH3COCH2CH3  c7 OC*N  m4 m2  c6  a'2 a'l CD3C0CD3  m3  c3 c2  m1  I,'  cl c4  a'2  c5  c7  B  a7 —o-C-NH2  II o  a7  I  I Ai  i  c7  C  a7  t7 II  H,  tl  200^160^120^80^40^0 PPM'  presented previously in Figure 2.3. Figure 2.7A shows that there is no reaction in MEK and acetone without catalyst on heating at 100 °C for up to five hours, even in the presence of added water. The corresponding 13C spectra which are not shown give the same result. It would appear that trace amounts of catalyst are needed at least at this temperature to induce the reaction. Just as expected, reactions were observed after adding 200 ppm zinc octanoate as catalyst and heating the sample at 100 °C for only one hour (Figure 2.7B). These reactions cause the appearance of two main resonances in the 15N spectrum, one at 87 ppm due to the formation of triazine 31 and the second one at —41 ppm which is split into a triplet due to coupling to two protons. This is consistent with formation of 28 by the addition of water to the cyanate group as described in the previous section. Two small singlet resonances are also observed due to very small amounts of other side products.  2.4. Synthesis and Characterization of the Carbamates 28 and 29 In order to confirm that the species with 8 = — 156 ppm in the 13C NMR spectrum and 8 = — —40 ppm in the 15N NMR spectrum is indeed a carbamate, both carbamates, 28 and 29, were synthesized by reaction of the appropriate cyanates (BPADCN and PTBPCN) with water using zinc octanoate or an acid[9] as a catalyst and were purified by recrystallization in acetone. Details of the procedures are given in the Experimental chapter later. The 13C and 15N NMR spectra of p-tert-butylphenyl carbamate 28 show resonances at 8 = 156 ppm for the carbonyl carbon (Figure 2.8A) and 8 = —41 ppm for the nitrogen which is coupled to two hydrogens giving a triplet resonance (Figure 2.8B). These spectra as well as the spectra of the carbamate 29 (not shown) are consistent with the previous results, and indicate that the carbamate is indeed formed in the curing process of the cyanate resin. 64  I  80  0  -40 1TM'  Figure 2.7. 15N solution NMR spectra (1H at 300 MHz) of 15N enriched PTBPCN 25b in MEK and acetone-d6 without 1H decoupling. (A). After heating at 100 °C for 5 hours without catalyst; (B). After heating at 100 °C for 1 hour with 200 ppm zinc octanoate as catalyst.  65  C6  C6 C3 C2  A  CS C4^CI  C7 N H2 0  C3  CI C7  CS C4  140  Err r-r7-r-  C2  ^  100^60  If  -20^-30^-40^-50^Ppm Figure 2.8. (A). 13C NMR spectrum (1H at 200 MHz) of p-tertbutylphenyl carbamate 28 in acetone-d6; (B). 15N NMR spectrum (1H at 300 MHz) without 1H decoupling of 15 % 15N enriched 28 in acetone-d6. 66  To monitor the possible involvement of the carbamate in further curing processes, the isolated carbamate 29 was heated in acetone and MEK at 120 °C in a sealed tube. After heating for 24 hours, a white precipitate appeared, which was not soluble in many solvents, such as ketone, chloroform, ethyl acetate, toluene, ether etc., and only dissolved sparingly in dimethyl sulfoxide (DMSO). This white precipitate was separated and identified as isocyanuric acid C31{3N303, 30. It shows signals at .5 = 151 ppm in its 13C NMR spectrum and at ö = 22 ppm in its 15N NMR spectrum. Its MS spectrum shows a parent ion at M+ = 129 a.m.u. as expected. Figure 2.9A shows the 13C solution NMR spectrum with assignments for the sample of carbamate 29 dissolved in DMSO-d6. Figure 2.9B gives the 13C NMR spectrum for the same sample after heating for 28 hours at 120 °C. The spectrum shows that the carbamate decomposed to yield two compounds, bisphenol A (3) and isocyanuric acid (30), as shown in Equation 2.7. It should be mentioned that the decomposition is favoured at high temperature.  3 H2N0,0 0  00NH 2  A  o  29  0  3 HO 3  H H 'N^NOH + 2 ONO H  [2.7]  30  2.5. Isolation and Characterization of the Triazine (31) obtained from p-tert-Butylphenyl Cyanate Because of the critical importance of triazine ring formation to the cyanate curing process, it was decided to isolate the anticipated triazine product  67  Figure 2.9. 13C solution NMR spectra (1H at 200 MHz) of the carbamate 29 in DMSO-d6. (A). Before heating; (B). After heating at  120 °C for 28 hours. 68  from the reaction of p-tert-butylphenyl cyanate and thoroughly characterize it to prove that it did indeed have the expected structure 31, 1,3,5-p-tertbutylphenoxy-2,4,6-triazine.  0 N 0  Yfly NN 0  31  As indicated in the Experimental chapter, the crystals obtained were suitable for a single crystal X-ray structure determination. Both NMR (see Figure 2.10) and single crystal X-ray diffraction experiments (see Figure 2.11) were carried out on this material. All of the spectral characteristics were in agreement both with the postulated structure 31 and with the 13C and 15N spectra recorded during the curing reaction in solution. The crystallographic data, the atomic coordinates, bond lengths and angles are given in the Appendix A. A perspective view of the molecular structure is shown in Figure 2.11. The triazine ring is clearly visible in the centre of the molecule and appears undistorted. The symmetrical 1,3,5substitution of the triazine ring appears to minimize intramolecular steric interactions between the substituent groups. The NMR spectra (Figure 2.10) of triazine 31 show chemical shifts 5 = 174 ppm for the carbons on the triazine rings and 5 = 87 ppm for the nitrogens on the triazine rings as expected. Therefore, the equivalence of the crystalline triazine product and the major reaction product from the curing reactions in solution, which has characteristic  69  t2 t3  A  t6  a' 1 t2^a'2 a' 1 CD3 COCD3  t3  t7  T  90  a'2  ti t4  80  710  60  ,^t PPM  Figure 2.10. (A). 13C NMR spectrum (1H at 300 MHz) in acetone-d6 of natural abundance triazine 31 formed from PTBPCN 25; (B). 15N NMR spectrum (I-H at 300 MHz) in acetone-d6 of 15N enriched triazine formed from PTBPCN 25b.  70  C24  C28^ CZ 7A  C27  Figure 2.11. Perspective view of the triazine molecule 31 formed from PTBPCN 25. 50% probability thermal ellipsoids are shown for the non-hydrogen atoms.  71  13C and 15N chemical shifts of 8 = 174 ppm and 8 = 87 ppm respectively, is clearly established. The postulation that the curing reactions of both monocyanate and dicyanate monomers in solution lead to analogous triazine products is verified.  2.6. The Mechanism of the Curing Reaction for Cyanate Resin in Solution The results of the investigations in previous sections therefore suggest that the main curing reaction for the cyanate resin in solution is to form triazine rings 33. There is no evidence for the formation of any long lived dimeric intermediates during the course of triazine formation. However, a side reaction does occur due to trace water present in the solvent used (MEK in the present instance). The reaction of the cyanate and water in the presence of catalyst and under heating up to 100 °C for a relatively short time gives a carbamate side product 27. With prolonged heating at high temperature, the carbamate 27 tends to decompose to the phenol 34 and the isocyanic acid (35) which is not stable and is immediately converted to its trimeric form, isocyanuric acid (30). According to the literature[9], the phenol can react with unreacted cyanate to form an imidocarbonate 36. However, this occurs only at low temperature. At high temperature, the reverse reaction is preferable. The synthesis and characterization of the imidocarbonate will be discussed in the next chapter. Therefore, the general reaction scheme for the curing of cyanate resin in solution can be given as in Scheme 2.1. Side reactions which involve the reaction of the cyanate group with water impurities in the MEK solvent could complicate the curing process in solution and make the final resin susceptible to attack by water. They also weaken the strength of the final resin because they decrease the number of cross-linking 72  sites by the loss of cyanate groups. Thus, bulk curing would be preferable. In the following sections this is discussed in detail.  OCN  32 H .T .  vs.A,^0-0-0-NH2 0 27  34  0 H^H [0=C=N-H]  35  O'N'O  30  Scheme 2.1  2.7.^Investigations of the Curing Process in the Solid State In order to probe the curing reactions in the solid state, 13C and 15N solidstate NMR investigations[94] were carried out on the 13C and 15N enriched BPADCN materials previously described. The chemical shift information obtained from the solution NMR experiments described above were used as reference data for structural assignments. The characteristic chemical shifts of the different functionalities determined in these studies are summarized in Table 2.1.  73  Table 2.1. Characteristic 13C and 15N Chemical Shift Values of the Functional Groups Derived from the Cyanate Group in Solution Curing Reactions. Functionalities  13c  15N  ppm from TMS  ppm from neat formamide  0  ^w  OCN  -(  0  \^N-4 02-0,pN.  •  109  53  174  87  156  —40  151  22  159  43  0 izz  0-C_NH2  6  0  H^,4^H N^*N0^N  H  .0  +0-0, *^*  w (o)-- o -  0=NH  2.7.1. 13C Solid State NMR Investigations (1) BPADCN Monomer Figure 2.12A shows the 13C CP MAS[66] spectrum of natural abundance BPADCN monomer 2 at 25 MHz 0-H frequency 100 MHz) together with the assignment in terms of the molecular structure. The NQS[951 (non-Quaternary suppression) experiment on the same sample is shown in Figure 2.12B. The NQS technique only detects the carbons with no attached protons. Exceptions  74  are methyl carbon resonances which are only partly eliminated due to their reduced dipolar interaction caused by the methyl group rotational motion.[95] The carbon in the —OCN group gives rise to three resonances due to residual dipolar coupling to the directly bonded 14N (I = 1) quadrupolar nucleus (these are reduced but not eliminated by MAS).[961 Figure 2.12C is the corresponding dipolar-dephased spectrum of the 13C enriched BPADCN monomer 2a and shows only these three signals as anticipated. Figure 2.12D shows the CP/MAS spectrum, with sidebands removed by the TOSS (total suppression of spinning sidebands) pulse sequence,[971 of the 13C enriched monomer obtained at 100 MHz (400 MHz for 1H). The spectrum is simplified as the three resonances have become almost degenerate because the dipolar coupling is independent of the magnetic field. The residual dipolar coupling is thus relatively reduced at the higher magnetic field while the quadrupolar coupling is reduced. Further 13C spectra of this system were therefore obtained mainly at high field (1H at 400 MHz). (2) Cured Resin from Solution Polymerization Figures 2.13A and B show the 13C CP MAS TOSS spectra of the solid obtained by evaporation of the solvent after the solution polymerization described previously. There is an intense signal at approximately 174 ppm assigned to the triazine ring carbons. The second broad signal at approximately 155 ppm is a composite signal corresponding roughly to the solution signal assigned to the carbamate compound 27 and its decomposed derivative isocyanuric acid, 30. It should be noted that these spectra are not quantitative and that the signals due to species such as carbamate 27 and isocyanuric acid 30 would be greatly enhanced by the cross-polarization process compared to the triazine signal as there are two protons only two bonds away from the 13C enriched carbons in their molecular structures. 75  Figure 2.12. (A). Solid state 13C CP/MAS NMR spectrum (3-H at 100 MHz) of the natural abundance BPADCN 2; (B). NQS spectrum 0-H at 100 MHz) of 2; (C). NQS spectrum 0-H at 100 MHz) of the 13C enriched BPADCN 2a; and (D) 13C CP/MAS/TOSS spectrum 0-H at 400 MHz) of 2a. 76  c6 c3 c2  A  cl & c4  cl & c4  c6 c3 c2 NC.  • *c5 c4.  c  c7 OCN  c 1 & c4  c7 —0C.N  L•••••••■•■••••  150^100  ^  50^ppm  Figure 2.13. 13C solid state CP/MAS/TOSS NMR spectra (1-H at 400 MHz) of (A). The solid sample obtained by evaporation of the solvent after curing the 13C enriched BPADCN 2a in MEK and acetone-d6 with 200 ppm zinc octanoate as a catalyst; (B). The same sample as in (A), NQS experiment; and (C). The solid sample obtained from bulk curing the 13C enriched BPADCN 2a at 250 °C for 15 minutes. 77  a7 t7  A  0 H ,J,^H 'N * II '  t7  zz  ......„..... 0 t7 *)--N * N a-0 ),--,-N  C  ,-t0  ^ILA_ 150^100^50^Ppm  (3) Cured Resin from Solid State Curing Figure 2.13C shows the 13C solid-state spectrum of the product from the bulk curing (no catalyst) of the 13C labelled BPADCN monomer 2a sample for 15 minutes at 250 °C. The spectrum is remarkably clean, showing a single major resonance at 174 ppm with no indication of substantial amounts of unreacted monomer or carbamate side product 27. This would appear to be a very viable polymerization process, and remarkably efficient. It further confirms that the carbamate side reaction in solution is caused by water impurities in the ketonic solvents used.  2.7.2. 15N Solid State NMR Investigations (1)  BPADCN Monomer Figure 2.14A shows the high-resolution solid state 15N NMR spectrum of  the labeled BPADCN monomer 2b at 40.6 MHz (1H at 400 MHz; reference: neat formamide). There are two sharp resonances, indicating that either the site symmetry is lower than the symmetry of the isolated gas phase molecule or that there are two nonequivalent molecules in one unit cell. This is also reflected in the two methyl resonances in the solid state 13C spectrum of BPADCN monomer, Figure 2.12A. The small peaks denoted ss are spinning sidebands due to incomplete sideband suppression. (2)  Cured Resin from Solution Polymerization Figures 2.14B — D show 15N spectra as a function of the contact time used  in 111/15N cross-polarization experiments for the solid sample, which was obtained from the solution polymerization by evaporation of the solvent. The resonance at lower field can be assigned to the triazine ring nitrogens, but there is substantial intensity at higher field due to other species, carbamate 27 and isocyanuric acid, 30. However, as before, these will be greatly enhanced by the 78  cross-polarization process, because there are directly attached protons coupled to the 15N nucleus in the two cases. This dipolar coupling interaction depends on 1/r3 ( where r is the distance between 1H and 15N nuclei), and this will have an extreme effect on the cross-polarization process. The relative intensities in spectrum 2.14B bear little relation to the actual concentrations of the different species. At longer contact times, this effect is less important and Figure 2.14C better reflects the concentrations of the different species. Figure 2.14D which only shows the signals from 15N nuclei without attached protons confirms that the higher field components all have directly bonded protons. Only one peak is observed here, corresponding to the triazine ring nitrogens. (3) Cured Resin from Solid State Curing Figure 2.15A shows the high-resolution solid-state 15N spectrum of the product obtained from a bulk curing of the 15N labelled BPADCN monomer 2b under identical conditions to those used for the 13C labeled BPADCN monomer 2a previously discussed (bulk curing at 250 °C for 16 minutes). The spectrum is much cleaner than those in Figure 2.14 obtained from the solution curing. Also, as before, the "non-triazine" signals are greatly enhanced. Figure 2.15B, which shows only non-proton bearing nitrogens, indicates as before that the other species have directly attached protons.  2.7.3. Quantitative Investigation of Solid State BPADCN Curing According to the previous results, the curing of BPADCN appears to be a very efficient process, especially when carried out on the neat material where side reactions with water are minimized. An attempt was made to quantify the curing reaction and also to relate the 13C and 15N spectra directly to each other by investigating a mixed BPADCN resin made up of 50%13C enriched monomer  79  Figure 2.14. 15N solid state CP/MAS/TOSS NMR spectra (1H at 400 MHz) of (A). 15N enriched BPADCN 2b; (B — D). The solid sample obtained by evaporation of the solvent after curing the 15N enriched BPADCN 2b in MEK and acetone-d6 with 200 ppm zinc octanoate as a catalyst, with contact times: (B). 1 ms; (C). 5 ms; (D). 1 ms with NQS pulse sequence.  80  B -  0- C-N*1-12 II 0  C  D  100  ^  0^-100^ppm  A^I  1 00  ^ ^ ^ 0 -100 ppm  Figure 2.15. Solid state 15N CP/MAS NMR spectra (1-H at 400 MHz) of the resin obtained after bulk curing the 15N enriched BPADCN 2b at 250 °C for 15 minutes, with contact time 1 ms. (A). with TOSS sequence; (B). with TOSS/NQS sequence.  2a and 50% 15N enriched monomer 2b. This mixed resin was cured for 15 minutes at 250 °C. Both 13C and 15N CP/MAS spectra were obtained in a high magnetic field (1H, 400 MHz). The disadvantage of a high magnetic field is that the large chemical shift anisotropies, which increase with the strength of the magnetic field, give rise to large numbers of spinning sidebands. In the more qualitative studies presented in the previous sections, these were eliminated using the 81  TOSS pulse sequence. However, for quantitative data it is important that the intensities of the spinning sidebands are taken into account. From a knowledge of where other resonances could occur, a spinning rate was chosen to avoid overlap of spinning sidebands and isotropic resonances. As noted previously, because the side products contain protons in close proximity to the labelled nucleus under observation, the spectra are very sensitive to the contact time used for cross polarization. A complete variation of contact times was therefore carried out in the experiments on both nuclei. The series of spectra shown in Figure 2.16 were obtained by the variation of contact time in the 13C CP/MAS NMR experiments for the labeled and cured BPADCN resin mixture. Only at very short contact times is there any indication of resonances due to species other than triazine. The contributions are small and it was very difficult to try to estimate them due to contributions from natural abundance resonances. Quantitation will be more reliable from 15N data. The maximum intensity of the 13C spectra occurs at about 5 ms of contact time, after which all of the signals decay due to the proton T1 p relaxation process. A spectrum obtained at this contact time using the TOSS sequence (Figure 2.17A) indicates that there are negligible contributions to the spectrum from the species other than triazine. Figures 2.17B, and 2.18 show an analogous series of 15N CP/MAS NMR spectra obtained from the same labelled sample. In the case of 15N spectra, the side products have protons directly attached to the 15N nucleus, greatly enhancing their contribution to the spectra. However, there is no natural abundance contribution to the spectrum. From an analysis of these data, the total contribution of species other than triazine to the spectrum is concluded to be less than five percent. Thus, the very high efficiency of the solid state curing reaction is verified. 82  CT(ms) 20.0  15.0  10.0  5.0  2.0  1.0  0.5  0.2  300  .^1^.  .^1^.  200  100  .^.^  Ppm  J  0.1  Figure 2.16. Series of 13C CP/MAS NMR spectra (1H at 400 MHz) with variation of the contact time (CT) without sideband suppression. The sample resin was obtained by bulk curing a mixture of 50% 13C enriched and 50% 15N enriched BPADCN monomers for 15 minutes at 250 °C. 83  A  i  .^.^t^.^.  150  .^.^1^.  100  50  .^1  Ppm  Figure 2.17. Solid state CP/MAS/TOSS NMR spectra (1H at 400 MHz) of the same resin sample as Figure 2.16 with CT = 5 ms. (A). 13C spectrum; (B). 15N spectrum.  84  15.0  10.0  5.0  0.2  1^.^1^•^1^.  300^200^100  0.1  0^-100 -200 Ppm  Figure 2.18. Series of 15N CP/MAS NMR spectra (1H at 400 MHz) with contact time variation but without sideband suppression. The resin sample was the same as Figure 2.16.  85  The solid-state NMR spectra thus indicate that the cured resin solid is composed mainly of triazine ring linkages, whether the curing reaction is carried out "neat" or in a solution. A catalyst is not necessary for the curing reaction if the temperature is high enough, and side-products come mainly from reaction with water present in the solvent. The reaction is very efficient, especially when the neat monomer is cured, more than 95% of the cyanate functionalities being converted to triazine rings in these experiments.  2.8. The Relation of the Crystal and Molecular Structure of BPADCN to Its Curing Efficiency From the results above, the cyanate to triazine conversion is confirmed as the basic curing reaction. It is demonstrated that triazine ring formation proceeds without the accumulation of substantial amounts of intermediate dimeric species, and that the small amounts of side products come from the reaction of the cyanate group with water present in the solvent. In the case of thermal curing of the neat resin, the reaction is found to be remarkably efficient, over 95% of the cyanate groups present being converted to triazine rings. In many ways, this high efficiency of the curing reaction of neat BPADCN is surprising because the Bisphenol-A moiety is quite rigid and tends to form a very rigid cross-linked framework, which one would intuitively feel it would be difficult to cure the resin near completion. To try to better understand the factors involved in the curing process, crystals of BPADCN suitable for single crystal X-ray diffraction studies were obtained by recrystallization from cyclohexane solution, and the crystal and molecular structures were determined. The crystallographic data are given in Table V in the Appendix B. The final positional parameters from the refinement are presented in Tables VI, VII, and VIII in the Appendix B with the numbering 86  of the atoms in the molecular structure shown in Figure 2.19. In the crystal, the two pseudo planes of symmetry through Cl, which exist in solution due to rotation about the C1—C4, Cl—C10, C13-02 and C7-01 bonds, are removed. All of the atoms in the molecule are now unique and in principle, all should give separate signals in the high-resolution solid state NMR experiments, although in practice not all differences are large enough to be resolved. This lack of molecular symmetry in the crystal explains the two resonances for the two methyl groups in the solid-state 13C NMR spectra (Figure 2.12) and the two nitrogen resonances for the two cyanate groups (—OCN) in the 15N spectra (Figure 2.14A) as described above. More interesting in the context of the curing reactions of this monomer is the crystal structure depicted in Figures 2.20 and 2.21. Figure 2.20 shows the arrangement of the molecules in the unit cell and indicates clearly that interactions between cyanate groups on different molecules are the dominant factors controlling the formation of the lattice structure. The interactions all involve four cyanate groups on different molecules. Two of these molecules form a four-membered ring by strong C17---N2 interactions and the remaining two cyanate groups each interact via an N1---02 interaction with one of the two oxygens in this four-membered ring. These intermolecular contacts are given in Table 2.2. Thus, in the solid-state, the functional groups needed for the curing reaction represented in Scheme 1.1 are all in very close proximity and strongly interacting. Figure 2.21 illustrates how these inter-cyanate group interactions form a complete three-dimensional structure, the lattice being made up of parallel "strings" of cyanate-bonded molecules.  87  Figure 2.19. Molecular structure of BPADCN monomer 2 from the single crystal X-ray diffraction study showing the numbering of the atoms (Tables in the Appendix B).  88  C11 N1  Cl2r it' iy cio C13^)  C15 C14  02  C17 C14^02 Cl5r1 C13 N1  C10.cf3C12  C11^C16 01  C8 C9  Figure 2.20. Perspective view of part of the unit cell contents from the crystal structure of BPADCN monomer 2 showing the intercyanate interactions. (see Table 2.2).  89  Figure 2.21. A plane through the three-dimensional network formed by the intermolecular intercyanate interaction. The BPADCN molecules are cyanate-connected to form parallel "strings" throughout the structure.  90  Table 2.2. Intermolecular Distances between Cyanate Grou s in the BPADCN Crystal Atom  Atom*  Distance (A)  0(2)  N(1)'  3.318(4)  N(2)  C(17)"  3.484(4)  N(2)  N(2)"  3.546(6)  * The symbols 'and "refer to different molecules  Although melting the sample will destroy the perfection of this ordering pattern, it is expected that strong intercyanate group interactions will still occur and substantial local ordering may persist in the melt. These results form a guide for modelling the curing process, and the high efficiency of the neat curing reaction becomes more understandable. The reduced curing efficiency in solution can be explained as due to the presence of solvents with strongly polar groups where competitive interactions with the cyanate groups can occur. When other dicyanates or derivatives are considered as potential resin systems in the future, it would be worthwhile to carry out a single crystal X-ray structural investigation to check whether the changes in molecular structure have destroyed the very strong intercyanate interactions which are present in the case of BPADCN.  2.9. Conclusions  The mechanism of the curing reactions of cyanate polymer resins based on bisphenol A dicyanate (BPADCN) has been investigated both in solution and in the solid state by NMR spectroscopic techniques. To increase the signal to noise ratio (S/N) and to unambiguously characterize the reactions of the cyanate  91  functional groups, 13C and 15N enriched cyanate resins and monocyanate model compounds were used, the latter yielding soluble and isolable analogs. In solution, the main reaction is formation of triazine rings as identified by NMR and MS techniques and characterized by single crystal X-ray diffraction on an isolated crystalline material from the monocyanate model compound. Side products are formed by the reaction of the cyanate functionalities with trace water present in the ketonic solvent, but there is no NMR evidence for the formation of dimeric or other intermediate species prior to triazine ring formation. The resins from the solution curing and also those formed directly by curing of the neat resin were characterized by high resolution solid-state NMR. In the former case triazine ring formation and the presence of side products were confirmed by both 13C and 15N solid-state NMR. In the case of curing the neat resin, the reaction is very clean and very efficient. It is shown to be almost quantitative. The efficiency of this process is rationalized in terms of the very strong intermolecular intercyanate bonding interactions which are observed in the crystal structure of the BPADCN monomer obtained from a single crystal Xray diffraction experiment.  92  CHAPTER 3. INVESTIGATIONS OF THE POSSIBLE CROSS REACTIONS BETWEEN CYANATE AND EPDXY RESINS A detailed investigation of the mechanism of the curing reaction of the cyanate resin both in solution and in the bulk was reported in the previous chapter. Cyanate resins have many excellent properties as mentioned previously. They can be mixed with many kinds of thermosetting resins to form compatible formulation blends. In commercial applications, they are usually modified with epoxy resins giving complete curing at temperatures as low as 177 °C.E3°1 However, the nature and even the existence of any cross-linking reactions between the cyanate and epoxy functional groups are not yet clear. Fivemembered oxazole 5 and isooxazole 6 rings have been proposed as cross-reaction products[32,35,361, but there is no general evidence or incisive spectroscopic information (such as from NMR, mass spectrometry, or X-ray diffraction techniques) to confirm or refute these proposed structures. In the present chapter, an investigation of possible cross-curing reactions between cyanate and epoxy resins is reported. In order to obtain soluble and isolable cross-reaction products, the monofunctional cyanate, p-tert-butylphenyl cyanate (PTBPCN, 25), and the monofunctional epoxides, p-tert-butylphenyl glycidyl ether (PTBPGE, 37) and ortho-methylphenyl glycidyl ether (OMPGE, 38), were used as model compounds for the two resins to prevent formation of a  cross-linked network. The reaction products were separated by absorption chromatography (silica gel column) and gel permeation chromatography (Sephadex LH-20 column), and fully characterized by NMR and mass spectrometry. The results demonstrate that the major cross-reaction product 93  contains a five-member oxazolidinone ring and is composed of one cyanate and two epoxy monomers. It is not an oxazole ring structure as proposed in the literature132,35,361. The reaction between the epoxy and the carbamate which is the side product produced during the cyanate resin curing process discussed in Chapter 2 was also investigated. Several related cross-reaction products have also been identified. Finally, the mechanism of the curing reaction for the cyanate/epoxy mixed resins will be discussed.  0CN  25  OCH 2 C H — CH2 \o /  37  OCH2CH\ ---/CH2 CH3  0 38  3.1. Solid State NMR Investigation of the Neat Curing Reaction of the Mixed Dicyanate / Epoxy Resins A neat mixture of the two resins [the mixed 13C and 15N enriched BPADCN resin (50% 2a and 50% 2b) and the natural abundance epoxy resin (EPON-825, 18, n=0.2), 1:1 molar ratio] was cured at 180 °C for 2.5 hours in air. Solid-state 13C and 15N spectra were obtained both on the crude products and also on the materials after exhaustive extraction (3 days) with MEK solvent. In addition, a sample cured under nitrogen protection which prevents the resin from reacting with moisture or oxygen in the air was also prepared. It gave the same results as the sample cured in air.  0^ ,^\^ CH2-CHCH 0^  OH 1 OCH2CHCH 0  o  / \  OCH2CH CH2  18  94  13C NMR Spectra The most intense resonance is due to triazine ring carbons (-173 ppm). The next largest peak is found at -155 ppm which is consistent with the 0u presence of carbamate 0 G-N— functional groups. Unreacted epoxy groups are present and their intensities are reduced somewhat after MEK extraction since unreacted EPON-825 dissolves in the MEK (spectra are not shown). 15N NMR Spectra These spectra are particularly important in this investigation because only the cyanate functional group and its derivatives give signals. As can be seen in Figure 3.1A, the spectrum shows several signals. The major signal at 87 ppm can be assigned to triazine. The cyanate signal at 52 ppm is very small, which means that most of the cyanate groups have reacted. Three other small signals appear at the higher field side of the cyanate. Two of them (at -42 ppm —o ,C=NH  and -23 ppm) can be assigned to be imidocarbonate ^o^and isocyanurate groups. The imidocarbonate species can be considered as an adduct of the cyanate and a hydroxyl group on the epoxy resin. Hydroxyl groups exist in the cured or partially cured epoxy resin as shown in Equations 1.7 - 1.9 and structure 18. Even EPON-825, which is a reasonably pure monomeric form of the epoxy resin, does contain a small number of hydroxyl groups (see structure 18). In addition, as will be shown later, some phenol moieties can be produced in the curing process itself. Besides these signals, there is also a very intense resonance at - -36 ppm.  0 From previous work, this could be assigned to carbamate -0-C-NH2 moieties.  0  However, the carbamate species, — 0-C-NH2, were not observed at all in the neat curing of BPADCN resin reported in the previous chapter due to the  95  ^  absence of water, although in the current case a small number of them might be expected from any entrained water in the epoxy resin. However, this signal grows relatively slowly as a function of contact time (Figure 3.2) whereas the 0  ^I  ^I  nitrogen in the —  0- C - NH2  functional group should cross-polarize very  efficiently due to having two directly bonded protons. Most importantly, in the non-proton-attached nitrogen selection experiment (Figure 3.1B), substantial spectral intensity remains. Thus it would appear that this resonance is due to a different nitrogen species in which the nitrogen has a similar local chemical 0 environment as in the^NE12 functional group, but has no directly 0, attached protons, i.e.^. It could thus be due to products from a cross reaction between the dicyanate and the epoxy resins. Based on these observations it was assumed that some cross reactions occurred, giving rise to the —36 ppm signal. Thus, attempts were made to investigate it further using monocyanate and monoepoxy analogues to obtain soluble and isolable products which could be more easily characterized.  3.2. Neat Curing Reactions of Monofunctional Cyanate and Epoxy Compounds An investigation of the curing reactions of a mixture of the two monofunctional compounds was carried out using unlabeled materials first to ascertain whether any cross reaction occurred between them.  3.2.1. Neat Curing Reactions of Unlabeled PTBPCN 25 and PTBPGE 37 In the 13C spectrum (not shown) of the unreacted mixture of the two reactants PTBPCN 25 and PTBPGE 37, the resonance at 109 ppm due to the cyanate group and the two resonances at 44 and 55 ppm due to the two carbons 96  ,0  >.-%-. N * N* -0- C-N*/ N^11^\ A^o^ii^II^o i /^00  yz... )._ •  ■ 0/ II  i  z  i•  4  L'1",....."•—••","."A"."."""r's.."0-•  1  11,11■1/11111TISIVIT  100  r •^•  0  1^1^,  -100^ppm  Figure 3.1. 15N solid-state NMR spectra (1H at 400 MHz) of the resin obtained by curing EPON-825 and BPADCN (50% 13C enriched 2a and 50% 15N enriched 2b) at 180 °C for 2.5 hours. (A). CP/MAS/TOSS spectrum; (B). CP/MAS/TOSS spectrum combined with the NQS technique.  97  CA  CT (ms) 5.0 4.0 3.0 2.0 1.5 1.0  I 0.5 0.2 0.1 0.05  200 100 -100 ppm Figure 3.2. Series of 15N solid-state CP/MAS/TOSS NMR spectra (1H at 400 MHz) of the same sample as Figure 3.1 with variation of the contact time as indicated. The triazine resonance is indicated by T and the imidocarbonate by I.  98  (Figure 3.2 continued)  CT (ms) 30.0  25.0 20.0 17.5 15.0 12.5 10.0 9.0 8.0 7.0  ioo  ^  6.0 I^. .^I  .^.^  .^.  -1 00 ppm 98-1  in the epoxy ring are particularly important, as reactions of the two functional groups should be reflected by changes in these resonances. After heating the mixture of the two monofunctional compounds, 25 and 37 (in an approximately 1:1 molar ratio), for 12 hours at 100 °C, there is no  change in the epoxy resonances while some of the cyanates (resonance at 109 ppm) have been converted to the triazine (resonance at 174 ppm) as would be anticipated from the results of Chapter 2. Heating the sample further for two more days at a higher temperature (125 °C) induces further reaction and the formation of a solid precipitate. The solid material was separated from the crude reaction mixture and identified as the triazine compound 31 by NMR. The unchanged epoxy carbon resonances indicates that the epoxy compound remains unreacted. Therefore, it is concluded that under these conditions, the epoxy groups do not react either with each other or with the cyanate groups and the only reaction which takes place is triazine formation by the cyanate groups. In order to observe the cross reaction between two resins, the temperature has to be raised even higher. The two monofunctional compounds, 25 and 37, were reacted neat in an approximate 1:1 molar ratio at 180 °C for 7 hours. The crude reaction mixture was dissolved in acetone-d6 and the 13 C NMR spectrum obtained is shown in Figure 3.3A. The most important features of the spectrum are that the cyanate group in PTBPCN 25 has completely reacted and has been converted mainly to triazine and that the most of PTBPGE 37 remains unreacted as shown by the three characteristic high field resonances in the aliphatic region. Thus, at least under the conditions of this reaction, the amount of cross reaction between the two compounds must be relatively limited. However, more careful inspection of the spectrum indicates that some of the epoxy groups have indeed reacted as indicated by the series of small resonances in the 45 — 80 ppm region which are 99  Figure 3.3. NMR spectra in acetone-d6 of the crude reaction mixture obtained by heating PTBPCN (12% 15N enriched) and PTBPGE at 180 °C for 7 hours. (A). 13C NMR spectrum 0-H at 200 MHz); (B). 15N spectrum (1H at 300 MHz) with NOE and no 1H decoupling.  100  A  0  / \  - OCH2- CH- CH2  =z  r"......14•■••■•■1,  11  •  ()  / 00  I^ I  no -OCN  160^' 120 ppm . 80 ^40  B I ^z^ \ ^..------- II ^ ^/\  /  0 —0—C—V ^II^\  00^0 I^I  , ^I ^ , 80  i^1^I^i 40  I  I  0  1  1^I^  I^I^I^I^I^  -40^IDIDni  I  characteristic of products from an epoxy ring opening reaction. However, it is not clear from the spectrum whether they are due to the cross reaction between cyanate and epoxy or due to self polymerization of epoxy groups. Separation and purification of the reaction products from the cyanate/ epoxy reaction was achieved using a combination of adsorption chromatography and gel permeation chromatography. A general protocol was developed for the separation of the reaction products, consisting of: 1) Extraction with pentane in which the triazine product is only slightly soluble and is thus largely removed from the reaction mixture. 2) Adsorption chromatographic separation on a silica gel (230-400 mesh, BDH No9385-48) column (2.5 cm x 17 cm) using eluants with gradually increasing polarity from 100% pentane to 100% diethyl ether. Finally the column was stripped clean using ethyl acetate. 3) When necessary, gel permeation chromatography (GPC) was carried out on a Lipophilic Sephadex LH-20 (Sigma) column (2.0 cm x 55 cm) with acetone as eluant. The reaction mixture was separated using the methods described above with most of the triazine product being removed by pentane extraction before the chromatographic separation. The approximate weight percentages of the components after the silica gel column separation of the reaction mixture are: triazine  -27%  PTBPGE and minor unknowns -38% imidocarbonate and phenol^ 0  -10%  ll  -0-C-NH2 and minor unknowns  major unidentified cross-reaction product fraction second unidentified reaction product fraction  101  others^  -5%  The structure of the second unidentified product will be discussed in detail later in a subsequent section. The major cross-reaction product fraction was further purified by GPC. El and CI mass spectra both show a parent ion peak at 587 a.m.u., which fits with a formula of C37H49N05. Such a molecule could be derived from one cyanate and two epoxy monomers. In the HPLC experiment on a Waters 945 spectrometer using 50% diethyl ether and 50% pentane as an eluant, the major cross-reaction product gives two incompletely separated fractions. This suggests that it is probably a mixture of isomers with the same molecular mass and very similar structures.  3.2.2. Neat Curing Reactions Using 15N Labeled PTBPCN In order to better characterize the species in the product mixture, -12% 15N labelled PTBPCN (25 and 25b) was reacted with PTBPGE 37 at 180 °C for 7 hours as before. The use of 15N enrichment means that the much more diagnostic 15N spectra can be obtained with no natural background, while limiting the enrichment to a level of -12% means that a large enough quantity for chromatographic separation could be processed. Figure 3.3B shows the 15N spectra of the crude reaction mixture obtained with NOE and without proton decoupling. There are clearly three major species in the mixture although there may be small amounts of others present which might be discriminated against by the experimental conditions or by their lack of directly bonded protons. The three major species, as indicated in the spectra, are: ^ Triazine 87 ppm ^ Imidocarbonate 43 ppm 102  Unknown species^—36 ppm There is no indication of carbamate resonance signals which would be expected at —41 ppm in 15N NMR spectrum (Figure 3.3B). This is because the carbamate functional groups react with PTBPGE 37 as will be discussed in more detail later. However, they were found after the chromatographic separation (spectra are not shown) and could be formed by hydrolysis of —OCN or imidocarbonate groupings on the silica gel column during the separation. It should also be mentioned that the spectral contribution from the imidocarbonate component is greatly enhanced by NOE in this spectrum. The signals of the unknown species at —36 ppm consist of two 15N resonances in the range of carbamates. The nitrogens have no directly attached protons, but have similar local environments to that of a carbamate (Figure 3.3B). They appear in the same chemical shift range where the non-protonated resonance was observed in the solid-state 15N spectra (Figure 3.1) from the mixed BPADCN / EPDXY resin curing. It is considered that these resonances are due to the same species in both preparations. Chromatographic separation of the mixture was first carried out on a silica gel column. After further purification of the cross-reaction product by GPC, a pure sample was obtained. The El mass spectrum of the purified major crossreaction product is in complete agreement with the previous mass spectral data. Most importantly, the parent ion peak mass of 587 a.m.u. confirms the compound is formed by the combination of one cyanate and two epoxy molecules. The composition of the major cross-reaction product was also confirmed by the synthesis, isolation and purification of the corresponding reaction product from PTBPCN 25 (FM = 175) and ortho-methylphenyl glycidyl ether (OMPGE, 38) (FM = 164). The El mass spectrum of the major cross-reaction product from this reaction shows the parent ion mass is now 503, again in exact agreement 103  with the proposed composition being one cyanate and two epoxy monomer units. Both of these compounds were subsequently investigated to unambiguously determine their structures.  3.3. Further Characterization of the Major Cross-Reaction Product 3.3.1. 1D and 2D NMR spectra Figures 3.4A, B and C show, respectively, the 15N, 1-H and 13C spectra of the major cross-reaction product from the reaction of PTBPCN 25 and PTBPGE 37 dissolved in acetone-d6. The 15N spectrum (Figure 3.4A) indicates that two similar nitrogen environments are present, both of which are non-protonated. In addition, since the MS indicates that there is only one cyanate moiety present, the molecule must exist in two, very similar isomeric forms. The 1H spectrum (Figure 3.4B) gives the expected ratios of methyl, aliphatic and aromatic protons for the proposed composition, but the numbers and relative intensities of the multiplets in the aliphatic region (see Figure 3.5A) again suggest the presence of isomers. This is confirmed by the 13C NMR spectrum (Figure 3.4C) where eleven signals (probably twelve signals with two degenerate) are observed in the aliphatic region. Since the MS shows that only two epoxy groups are present, this again infers the presence of two isomers. The 1H-coupled 13C spectrum (not shown) shows that the aliphatic carbons in the sample are present as four 0  N-CH2-6H-CH 2- 0  groupings as indicated.  From this information, the 1H 2D COSY NMR spectrum (Figure 3.5A) and the 1H/13C heteronuclear shift correlation 2D NMR experiment (Figure 3.5B) make it possible to assign the 13C and 1H resonances in the aliphatic region of the spectra as shown. These two experiments better define the system,  104  Figure 3.4. NMR spectra of the major cross-reaction product after both silica gel column and Sephadex LH-20 column separations of the same reaction mixture as Figure 3.3. (A). 15N NMR spectrum (1H at 300 MHz) with NOE and no 1H decoupling; (B). 1H (500 MHz) NMR spectrum; and (C). 13C NMR spectrum (1H at 500 MHz).  105  / —0—C— N II^\ 0 -CH3  aromatic protons aliphatic protons  i^i^I 7^5  H20  1  aromatic carbons (-----A---Th  'CH3 epoxy ring has opened  C  ill.■••■■/"..•■•■•■••,t  i  i 140  ^"amem.W.N.■•••■■■^  1  I 1^ .r^ 60^PPM 100^  A  4.8  i^1 4.4^4.0  3.6 Pm  Figure 3.5. (A). 1H (400 MHz) 2D COSY NMR spectrum and (B). 1H/13C chemical shift correlated 2D NMR spectrum (1H at 500 MHz) of the same sample as Figure 3.4.  106  106-1  but as will be seen, they are not sufficient to make a definite assignment of the structure. From the 1D and 2D NMR experiments, two possible structures (39 and  40) which are in agreement with the general features of the NMR and MS data are proposed as shown below.  CH / x 0-C-N^0 H^\^,r_ 0 CH — 39  0-CH2  Although the proposed structures 39 and 40 are quite different and it would appear at first sight that they should be easily distinguishable from the 1H and 13C NMR experiments, closer inspection reveals that both of them are made up from combinations of the exact same groupings of local environments. For example, the nitrogen local environment is 41, that of the carbonyl carbon is  42 and the moiety from the epoxy ring opening is 43 in both cases. Thus, both structures are in good general agreement with the observed resonances and chemical shift values of the 13C, 15N, and 1H NMR spectra and also show exactly the same local connectivities deduced from the 2D NMR experiments.  107  /^\ -0 - C-N^N-CH2 -CH -CH2-0 II^\^/ 0  01  '  42^  43  Furthermore, there are opportunities in both structures for the formation of closely related stereoisomers which would explain the multiplicity of signals observed in the spectra. As indicated, both structures have two chiral centers. Compound 39 is expected to exist in two isomeric forms (cis and trans), and thus 0 N-CH2 -H C -CH2 -0 yield two groupings (six 13C signals) and two 15N signals.  However, the restricted rotation about the partially double bond of CO—N in 39 could double the number of aliphatic carbon resonances without affecting the 15N spectrum.  0  In the case of 40, the two N-CH2 - CH - CH2 -0  groupings are not  equivalent in the structure, giving rise to six aliphatic carbon signals. In addition, the presence of two chiral centres as before will give rise to RR(SS) and RS(SR) isomers, doubling the number of carbon resonances to twelve, as observed. As will be described, various other experiments have been carried out in an attempt to make an assignment of structure. Although no one experiment is definitive, the balance of evidences favors the second structure 40.  3.3.2.  Model Compounds for Structure 39  In order to probe the influence of ring configuration and inversion on the chemical shift differences in 39, the model compound 2,6-dimethylmorpholine (DMMP, 44) shown below was studied.  108  CH3 / *( HN^0 *(  44  CH3  The 1H spectrum of DMMP 44 gives the correct relative intensities for the protons in different environments (spectrum not shown). The two stereoisomers (cis and trans) are present in different amounts and can be clearly identified. The 13C spectrum of DMMP 44 has much better separation of the different signals which can be assigned as indicated (Figure 3.6). However, it is clear that only six aliphatic carbon signals can be produced by the basic ring system of the reference compound 44 and the doubling of the number of environments would have to be due to a restricted rotation about the CO—N bond as previously indicated. In order to test this latter possibility, a suitable model compound was synthesized. The compound chosen was the derivative 46 formed by the reaction of compound 44 with phenyl chloroformate 45 as in equation [3.1].  x x  CH3  CH3  HCI / /K (0)^ 0 C CI + HN0 ^.^0-C-N^0 \ IF^ 16 \ ^* K 45^44  CH3  46  [3 .1 ]  CH3  The linkages of 46 are identical to those of compound 39. The reference compound has a bulky phenyl substituent and contains the essential CO—N unit where a restricted rotation about the CO—N bond would occur even though the ring substituents are different in two cases.  109  Figure 3.6. 13 C NMR spectrum ( 1 H at 200 MHz) of 2,6dimethylmorpholine 44 in acetone-d6.  The effect of the restricted rotation about the CO—N bond can be seen in the 13 C spectrum at room temperature (20 °C) in Figure 3.7. The carbon resonances in the aliphatic region almost double in number and become ten signals (twelve signals with two degenerates). In particular, the CH2 carbon  110  signals in the two isomers are furthest apart because the CH2 carbons are closer to the CO—N bond than all of the others.  3.3.3. Variable Temperature NMR Experiments To confirm that restricted rotation exists in the model compound 46, variable temperature 13C NMR experiments were carried out. The spectra at temperatures from 20 — 60 °C are shown in Figure 3.7. The line coalescence at elevated temperature is very obvious and occurs very easily (just slightly above room temperature). Similar changes are observed in the proton NMR spectra (not shown). In order to probe possible effects from restricted rotation around the CO N -  bond in the major cross reaction product, a variable temperature 13C NMR study was also carried out on this material (Figure 3.8). However, in contrast to the results from model compound 46, there is no clear evidence for line coalescence at higher temperatures. There is some broadening of the lines above room temperature which can be ascribed to temperature gradients within the sample. It is possible that the energy barrier for rotation could be too large to be averaged over the accessible temperature range but this is considered unlikely due to the similarity of the N substituent groups. Based on the results on the  model compound 46, if structure 39 were the correct structure of the crossreaction product, averaging should occur just above room temperature in the corresponding 1H NMR spectra due to the smaller frequency separations of the resonances. However, here as well no indication of averaging was observed, suggesting that restricted bond rotation is not present in the major crossreaction product. Furthermore, unlike the major cross-reaction product, the two isomers of compound 46 (trans and cis) could be easily separated using a silica gel column, 111  Figure 3.7. Aliphatic regions of the 13C NMR spectra (1H at 300 MHz) of 2,6-dimethy1-4-phenoxycarbonylmorpholine 46 in acetoned6 at the temperatures indicated.  112  0—C—N 0 \^( 0 Cis^and^Trans'  CH3  CH CH'  CH3'  '4Avr#4.*""%wefiL400.4  40  30  20 70^65 PPrn 50  15  C H' 0 0  N-CH25  r.)^rv  CH  N-C H2  o^  _ jt,_____,,Lili  T ( °C) 60  ^50  40  30  20 —T—I—Fir I  tilt  75^70  1.--^f^I^I —1  50  I I T-1  45 ppm  Figure 3.8. Aliphatic regions of the 13C NMR spectra (1H at 300 MHz) of the cross-reaction product (Figure 3.4), at the temperatures indicated. 113  because the physical properties of the two model isomers are more different than those of the isomers in the major cross-reaction product mixture. This again suggests that the major cross-reaction product does not have a similar structure to compound 46. Therefore, structure 39 can be ruled out.  3.3.4. NOE Experiments Although, as indicated previously, the local environments in structures 39 and 40 are identical, there are some differences when longer range effects are considered. In particular, the structural unit 47, as shown below, is unique to structure 40. In the proton NMR spectrum, it is possible to identify the resonances due to the aliphatic methine (CH) proton and the resonances of the ortho protons on the aromatic ring which are clearly separated from the aliphatic signals. In structural unit 47, an NOE effect should be observable between the resonances of the CH proton and those of the ortho protons on the aromatic ring.  CH2-0  47  Particularly importantly, the resonances due to the ortho protons on the aromatic ring which is derived from the cyanate monomer and not from PTBPGE can also be identified by the comparison of the spectrum of the major cross-reaction product made from PTBPCN 25 and PTBPGE 37 (Figure 3.9A) with that of the analog made from PTBPCN 25 and OMPGE 38 (Figure 3.9B). 114  Figure 3.10 presents the results of 1H NOE difference experiments on the major cross reaction product made from PTBPCN 25 and PTBPGE 37. The spectra 0 1  - CH2-CH-CH2-0 structural units within confirm that there are two separated N^  one molecule. The assignments of the 1H resonances deduced from the 1H 2D COSY experiment (Figure 3.5A) are also confirmed by these experiments. It is 0 1  very clear that in both N-CH2-CH-CH2- Ounits, NOEs exist between the CH (or CH') proton and the N—CH2 (or N-CH2') protons, between the CH (or CH') proton and the 0—CH2 (or 0—CH2') protons, and between the 0—CH2 (or 0— CH2') protons and the ortho protons on the aromatic ring. However, there is only 0 1  N-CH2-CH-CH2-0 unit which shows NOEs between the CH proton and the one^  ortho protons on an aromatic ring (see the spectrum in which the CH proton at 4.95 ppm was irradiated in Figure 3.10). Most importantly, it can be identified that this aromatic ring is derived from the cyanate monomer as shown in Figure 3.9, assuming that they have structures 40 and 48 as indicated. The NOE difference experiments for compound 48 (Figure 3.11) show the same results as described above. An NOE is seen clearly between the CH proton and the ortho protons on the cyanate-derived aromatic ring, providing the strongest evidence in favor of structures 40 and 48 for the cross-reaction products. The other NOE effects are 0 shown within the N-CH2-6H-CH2-0- unit or between the 0—CH2 (or 0—CH2') and ortho protons on aromatic ring, which do not distinguish between 39 and 40. They do, however, demonstrate the reliability of the experiments and are consistent with the connectivities deduced from the 2D NMR spectra (Figure 3.5).  115  ortho H on cyanate derived ring  ortho H on cyanate derived ring  I^I  7.5^IDPm^6.5  Figure 3.9. 1H (400 MHz) NMR spectra in aromatic regions of the major cross-reaction products derived (A). From PTBPCN 25 and PTBPGE 37; and (B). From PTBPCN 25 and OMPGE 38.  116  Figure 3.10. 1 H (400 MHz) NOE difference NMR spectra of the major cross-reaction product derived from the reaction of PTBPCN  25 and PTBPGE 37, which is the same as Figure 3.4.  117  °Rho H on epoxy derived ring ( 6.82 ppm )  O-CH2'  Irradiated at (ppm) 4.06 (0 M2') -  ortho H on epoxy derived ring ( 6.86 ppm )  O-CH2' CH'  4.08 (0-CH2')  4  (Mho H on epoxy i derived ring ( 6.89 ppm )  CH  O -CH2  4.27 (0-CH2)  N-CH2' CH' O-CH2' I  4.85 (CH') ortho H on cyanate / derived ring O-CH2 CH^N-CH2  4.95 (CH) ortho H on cyanate derived ring  z oz6  6 6..F  n^I ,F  101, h  kilo  7.0  -^ 6.0^ 5.0^4.0^ppm  Figure 3.11.1H (400 MHz) NOE difference NMR spectra of the major cross-reaction product derived from the reaction of PTBPCN 25 and OMPGE 38.  118  ortho H on epoxy derived ring  0-CH2'  t  Irradiated at (Ppm) 4.10 (0-CH2')  ortho H on epoxy derived ring  0-CH 2 '  .....A4vi r  Heow•-.-..0"—^N 4.13 (0-CH2') Mho H on epoxy derived ring  /  0 CH2 -  4.24 (0-CH2)  CH I  4.86 (CH') ortho H on cyanate / derived ring CH  4.97 (CH) ohho H on cyanate derived ring  i  ==  00  7.0  118-1  0- CH CH3  3.3.5. A Possible Mechanism for the Main Cross-linking Reaction Based on the above information, a possible mechanism for the formation of the major cross-reaction product 40 (and analog 48) can be proposed. The addition of two epoxy functional groups to a cyanate group gives the unstable intermediate species 49 which then rearranges by migration of the phenoxy group from cyanate to the CH carbon to give compound 48 [Equation 3.21. This is in agreement with the placement of the cyanate phenyl ring in the position indicated. Strong NOEs between the CH proton and the ortho protons on the aromatic ring which is derived from the cyanate are observed in both compounds 40 and 48 as discussed above.  119  OCH2CH-CH \ ,^2 0 CH3^38  OCN + 25  0-CH2 CH3  CH3  [3.2] 49  0-CH2 CH 3  3.4. Investigation of the Second Unidentified Product The second unidentified product fraction was also further purified by GPC (Sephadex LH-20 column). The 15N spectrum of the main component is shown in Figure 3.12A, and indicates that there is a single nitrogen present with one attached proton. The chemical shift at - -40 ppm is in the range previously 0 II found to be characteristic of a carbamate -°-C- NH2 environment. The 13C  120  spectrum (Figure 3.12B) indicates three carbon signals in the opened epoxy ring region. In addition to the major peaks, there are a number of very small additional signals in the spectrum indicating that the sample is not completely pure. The 1H NMR spectrum (Figure 3.12C) also shows some minor peaks, but those due to the major component are clearly identifiable as indicated. The chemical shifts and intensities of the signals in the 13C and 1H spectra reveal 0 the presence of a p-t-butylphenyl group, a N -CH2-CH-CH2 -0 moiety from the opening of the epoxy ring and also a carbonyl group. The 1H 2D COSY NMR spectrum (not shown) shows the expected connectivities within the 0  N-CH2-CH-CH2 -0  moiety and makes an assignment of this portion possible. In  addition, the nitrogen atom is known from the 15N spectrum (Figure 3.12A) to have a single attached hydrogen and a carbamate environment, eg. 0I^I -0-C-NH-  .  Furthermore, the NMR spectra show that the ratios of the  different components: phenyl ring, opened glycidyl unit, and nitrogen are 1:1:1 (or 2:2:2 etc.). The mass spectrum gives information regarding the molecular weight, but there is some ambiguity because of the presence of the minor impurities. However, the major peak in the spectrum is at M+.249 a.m.u. which would correspond to a 1:1:1 ratio of the components if it were the true parent ion. From all of these information, a possible structure for the second unidentified product can be proposed as 50. To confirm this and to make complete assignments in Figure 3.12 further experiments were carried out (see on).  0-CH2-  50  121  Figure 3.12. NMR spectra of the second unidentified product after both silica gel column and Sephadex LH-20 column separation of the crude reaction mixture obtained by heating PTBPCN (-12% 15N enriched) and PTBPGE at 180 °C for 7 hours. (A). 15N spectrum (1H at 300 MHz) without 1H decoupling; (B). 13C spectrum (1H at 200 MHz); and (C). 1H (200 MHz) spectrum. 122  9 8  4 „CH2 3 0-CH2-CH \ 2  o-  I  "  "  I^I  8.0  fl 0  • , " • -• i •^i^i^...,„„, 6.0 ppm 4.0^2.0  3.5. Reaction of Carbamate and Epoxy Carbamate is formed from the reaction of cyanate with water, and is one of the side products of the curing reaction of cyanate in solution. To determine whether a cross reaction product could be formed between the epoxy monomer and the carbamate, both 15N labelled and natural abundance model carbamate compounds were used in neat reactions with epoxy compounds.  3.5.1. Reaction of 15N Enriched p-tert-Butylphenyl Carbamate 28 with  PTBPGE 37 A mixture of 12% 15N enriched p-t-butylphenyl carbamate 28 (made from hydrolysis of 12% 15N labelled PTBPCN 25) and PTBPGE 37 (1:2 mole ratio) was heated at 180 °C for 3.5 hours. The 15N NMR spectrum of the reaction mixture (Figure 3.13) shows that all the carbamate has reacted and there are only two major nitrogen environments, one of them with a single attached proton (species A). The reaction mixture was separated using a silica gel column and a similar procedure described above. Two major reaction products, each of them with one single nitrogen environment, were isolated. One of the products has identical 15N, 13C, and 1H NMR spectra to those of the second unidentified reaction product obtained from the reaction of PTBPCN 25 and PTBPGE 37. It also shows same parent ion peak (M+.249 a.m.u.) in the mass spectrum. Therefore, they are considered to be the same compound with the carbamate derived sample being of high purity. The second compound shows a single nitrogen resonance at 23 ppm with no proton coupling (Figure 3.14A). The 13C NMR spectrum (Figure 3.14B) also indicates three resonances in the region corresponding to a ring-opened epoxy  123  •^N-CH2-CH-CH2 moiety^-(:). The 1H NMR spectrum (Figure 3.14C) gives the ratios of methyl protons, aliphatic protons and aromatic protons as 9:6:4. The structure of this product can be seen by comparison with the data from the product formed from the reaction of phenyl carbamate with phenyl glycidyl ether described in the next section.  A  8:0  1^I^I  40^0^-40^Ppm  Figure 3.13. 15N NMR spectrum (1H at 300 MHz) in acetone-d6 with NOE and no 1H decoupling of the reaction mixture obtained by heating p-t-butylphenyl carbamate (-12% 15N enriched) and PTBPGE at 180 °C for 3.5 hours.  124  Figure 3.14. NMR spectra in acetone-d6 of the second product after silica gel column separation of the same reaction mixture as in Figure 3.13. (A). 15N spectrum (1H at 300 MHz) with NOE and no 1H decoupling; (B). 13C spectrum (1H at 200 MHz); and (C). 1H (200 MHz) spectrum. 125  A  T-T-T--1-  1--- r-  80^40  7^-I-  0  -40 Ppm  epoxy ring has opened  160  1 tz:1^ lm^ro pp^s  410 -C H3  aromatic protons  aliphatic protons  3.5.2. Reaction of Phenyl Carbamate with Phenyl Glycidyl Ether As described later in the Experimental chapter, the product 53 was produced by the reaction of phenyl carbamate 51 with phenyl glycidyl ether 52 at 180 °C for 3.5 hours. It corresponds to the second unknown product of the reaction of PTBPCN 25 and PTBPGE 37 according to the similarities between their 1H and 13C NMR spectra. Since it can be isolated and well purified, the structure of this product can be well determined. In this case, the interpretation of the mass spectrum is unambiguous because the sample is pure, and the parent ion mass at M+.193 a.m.u. clearly identifies the ratio of phenyl ring to ring-opened epoxy moiety to nitrogen as being 1:1:1, indicating the loss of one phenol during the reaction. The 15N, 13C and 1H spectra shown in Figures 3.15 and the 1H 2D COSY NMR spectrum (not shown) are in complete agreement with the proposed structure 53. Furthermore, a good quality crystal of this sample was obtained from recrystallization in acetone and was investigated by single crystal X-ray diffraction yielding the molecular structure shown in Figure 3.16. It indeed contains an oxazolidinone ring confirming the conclusions from the NMR and mass spectra. More importantly, with the exception of the nuclei in the phenyl group which are affected by the substitution of H for t-butyl, the 13C and 1H spectra are identical with those of the second unknown product obtained from reaction of PTBPCN 25 and PTBPGE 37 . Therefore, the second unknown product can be assigned the corresponding structure 50.  0-C H2  NH  0-CH2  50^0^ 53  126  Figure 3.15. NMR spectra of product 53 in acetone-d6. (A). 15N spectrum (1H at 300 MHz); (B). 13C spectrum (1H at 200 MHz); and (C). 1H (200 MHz) spectrum. 127  4  98  r^,CF12 3  0-CH2-CH  A  3  , ^ , ^, -32^-36 9  -40  -44^ppm  8  7.0^6.0  5.0  4.0 ppm  H1  Figure 3.16. Molecular structure of product 53 from the single crystal Xray diffraction experiment with the numbering of the atoms indicated. Complete structural data are given in Appendix C.  128  Besides product 53, the reaction of phenyl carbamate and phenyl glycidyl ether also produced two other products and some polymerized materials (mostly polyether polymer). The reaction products were isolated and purified by absorption chromatography (silica gel column) and gel permeation chromatography (Sephadex LH-20 column). The approximate weight percentages of the components in this reaction mixture were: product 54, phenol and unreacted phenyl glycidyl ether —35% polymerized materials —25% product 55 fraction  —15%  product 53 fraction  —25%  On standing for several days, a product identified as 54 crystallized and was isolated from the first fraction. After further purification by recrystallization from acetone, MS analysis yielded a molecular weight of 244, and elemental analysis showed composition of C 72.8% H 6.6% 0 20.5% corresponding to the formula C15111603. From the 13C, 1H NMR spectra (not shown), the product can be identified to be 2-hydroxyl-1,3-diphenoxy propane 54 (shown in Equation 3.3 below). Another product fraction identified as 55 (shown in Equation 3.3 below) can be further purified using GPC (Sephadex LH-20 column) with acetone as the eluant. From the similarities between their 13C and 1H NMR spectra, it corresponds to the compound with a single 15N resonance at 23 ppm (Figure 3.14A) in the mixture from the reaction of —12 % 15N enriched p-tertbutylphenyl carbamate with PTBPGE. In the CI mass spectrum, it shows a parent ion peak at 579 mass units [ (M+1)+=580 a.m.u.]. Figure 3.17 shows the 13C and 1H NMR spectra of this product. The 13C NMR spectrum (Figure 3.17A) shows three resonances in the aliphatic region which indicate that there are one or more identical ring-opened glycidyl units 129  0 1 N—CH2—CH—CH2-0. From the 1H NMR spectrum (Figure 3.17B), the ratio of aromatic protons to aliphatic protons is approximately 5:6. This means that the 0  1 N-CH2 -CH -CH 2 -0 could be either ratio of phenyl groups to the structural units  5:6 or more likely 1:1 with a contribution of one additional hydrogen from a 0 1  hydroxyl group on the N-CH2 -CH- CH2 - o unit. By comparison with the proton ratio found for the corresponding product made from the reaction of p-tbutylphenyl carbamate and PTBPGE (Figure 3.14C), the 1:1 ratio of phenyl and 0  1 N -CH2 -CH - CH2 —0can be established. The 1H-13C 2D the structure unit  heteronuclear shift correlation and the 1H 2D COSY NMR spectra (Figures 0 1 N CH2 — CH C H2 0 3.18A and B) confirm the existence of the structural unit —  —  —  and a hydroxyl group which is connected to the -CH- carbon. The 1H resonance of the hydroxyl group is split into a doublet by coupling to the —CH— proton. By combining all of the information from the mass spectrum (M+ = 579) and the 0  1 N-CH2-CH-CH2 -0 NMR spectra, the ratio of phenyl, the structural unit and  nitrogen should be 3:3:3 and the structure of this product is deduced to be 55. Therefore, the reaction of phenyl carbamate 51 and phenyl glycidyl ether 52 can be represented as in Equation [3.3].  130  Figure 3.17. NMR spectra of product 55 in acetone-d6. (A). 13C spectrum (1H at 200 MHz); (B). 1H (500 MHz) spectrum.  131  67 0 ii^2 3 4^5 CI-W1-1 CHp OH  7  0  6  A  8 43 1  5  .^i^I^'^1  160^120 ppm^80  B  ^2  ^  1 ^I  aliphatic protons  aromatic protons  „ ^ ^ 7.0 6.5 ppm  ^ PP.  I^T^V^llll^I  4.5^4.0  40  A  Ca1Z413P 50  60  41:;=)  70  41=21> SP^  1  4.5  I  i 4.0  1  Ppm  Figure 3.18. (A). 1H-13C 2D heteronuclear correlation NMR spectrum (1H at 500 MHz) and (B). 1H (500 MHz) 2D COSY NMR spectrum of product 55 in acetone-d6.  132  N-CH2  B OH  CH  O-CH2  /\  4.0  4.5 I 4.5  1 4.0  IDPm  132-1  0-0—NH2 i 0 51  0- CH 2- C C H2  52 [3.3]  0-CH2  0-CH2-CH-CH2OH  53  54 0 0-CH2-CH-CH2-NN-CH2-CH-CH2OH^ 0^N^0  OH  CH2-CH-CH2-  55^OH  Phenol is necessary for the formation of 54 [Equation 3.4]. This is produced during the formation of compounds 53 and 55. (see Scheme 3.1).  0—CH2-0070H2 0 52  HO  [3.4] 0-CH2-CH-CH2OH  54  3.6. A Possible Mechanism for the Reaction Between Epoxy and Carbamate As will be described in the Experimental chapter, during the reaction of phenyl carbamate and phenyl glycidyl ether, some white solid formed at the  133  beginning and then disappeared at longer reaction times. The reaction was stopped after the white solid formed and this material was isolated by filtration and then washed with pentane, diethyl ether, and acetone. The 13C NMR spectrum which shows only a single resonance at 151 ppm and the mass spectrum which shows a parent ion peak at M+.129 a.m.u., corresponding to a formula C3H3N303 indicate that it is isocyanuric acid 30. This is the decomposition product of the carbamate as described in Chapter 2. Isocyanuric acid is obtained from three phenyl carbamates by elimination of three phenol moieties. In addition, the formation of product 53 also involves elimination of a phenol moiety. Thus, a possible mechanism for the overall reaction can be given, as shown in Scheme 3.1. At high temperatures, the carbamate 51 may first decompose to phenol and isocyanic acid (0=C=NH) which is a very unstable intermediate under these conditions. Isocyanic acid will immediately react either with phenyl glycidyl ether 52 to form the product 53 or with other isocyanic acid molecules to form a trimer, isocyanuric acid 30. Isocyanuric acid 30 is the intermediate for product 55 and only exists at the beginning of the reaction. The phenol produced in the  decomposition of the carbamate can react with phenyl glycidyl ether 52 to yield product 54. All of these reactions are shown in Scheme 3.1 below.  134  ^  0-C-NH2  0  51  (0)^ OH^+^[0=C=NH  0-cH2-CH-,cH2  0-CH2-CH-,CH2  52 ^52  ^0  (0)^ 0 CH2-CH-CH OH  (0)-0-CH2^  NH  53  54  o  0 HN^NH  30  0 N^0  o  o-cH2-cti-,cH2 52^0  0 0-CH2-CH-CH2-N^—CH2-CH-CH OH^ 0 N 0  55  OH  CH2-CH-CH2OH  Scheme 3.1  The intermediate 30, which is produced by decomposition of the carbamate, was also observed in the curing reaction of dicyanate resin in solution described in Chapter 2. This can be seen in the 15N solid-state NMR spectra of the resin cured in solution (Figure 2.14) and the 13C spectrum of 135  Figure 2.9. The peak at —22 ppm (Figures 2.14B and C) disappears in the nonprotonated nitrogen selection experiment (Figure 2.14D).  3.7. Investigation of Imidocarbonate as a Possible Cross-reaction Product Since some hydroxyl groups are always present in the commercial epoxy resins, it was thought that the reaction between cyanate and these hydroxyl groups might act as another way to cross-link the two resin systems. To prove this, p-tert-butylphenol (PTBP, 56) was first reacted with the cyanate PTBPCN 25 to provide a suitable reference compound (59), and then isopropanol (57) was used as a model compound for the secondary hydroxyl groups in the polyether units of the epoxy resin (58). CH3-CH-CH3^mAO-CH2-CH-CH2-ONAA,  1  OH^  OH  57^  58  3.7.1. Reaction of Cyanate with p-tert-Butylphenol (PTBP) 13C NMR spectra (not shown) show that heating the reaction mixture of the PTBPCN 25 and PTBP 56 in MEK/acetone-d6 solvent for up to six days at 100 °C gives complete conversion of the cyanate (8 = 109 ppm) to triazine (8 = 174 ppm), but there is no reaction at all for cyanate and the corresponding phenol under these conditions (spectra not shown). After a trace of sodium hydroxide has been added as a catalyst to the mixture, and the solution has stood for 10 hours at room temperature, the 13C NMR spectrum (not shown) shows that there is efficient conversion of the cyanate group to another species characterized by a 13C resonance at 159 ppm as well as the production of a  136  small amount of triazine. This major species is thought to have an imidocarbonate structure 59. A very clean conversion to the same product species is also observed at room temperature in acetone solvent with triethylamine as a catalyst. These conditions were used to prepare and isolate this compound as described in the Experimental chapter. The 13C and 15N NMR spectra of this purified compound are shown in Figure 3.19A and B together with the complete assignments. The 15N spectrum shows a resonance at 43 ppm due to a nitrogen with a single attached hydrogen. Combining this information with the data from the mass spectrum, which yields parent ion M+ = 325 a.m.u., confirms that it is the imidocarbonate 59. Therefore, basic conditions are necessary for the formation of the imidocarbonate, which can be represented as in Equation 3.5.  OCN  base  a-ON C=NH  0)—OH  25^56  [3.5]  59  3.7.2. Reaction of Cyanate with Isopropanol After heating PTBPCN 25 in an excess of isopropanol 57 at 100 °C for one hour with triethylamine as a catalyst, the 13C NMR spectrum shows that there is complete conversion of the cyanate group to a new species characterized by a resonance at 159 ppm together with a small amount of triazine (spectrum not shown). From the similarity of this shift to that of the reaction product with PTBP 56, it is thought that the product of the reaction of cyanate 25 with isopropanol 57 has structure 60 and that reactions of this general type could form a viable attractive route to the cross reaction between cyanate and epoxy resins. This is indeed observed in curing the mixed dicyanate (BPADCN) and 137  Figure 3.19. NMR spectra of product 59 in acetone-d6. (A). 13C spectrum (1H at 200 MHz); (B). 15N spectrum 0-H at 300 MHz) with no 1H decoupling.  138  6 5 4 13  12  32  1^8 10 9  07 . C=N H 0  59  B  IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 60^ 50^ 40^ 30  ppm  13  1  ^  1^f^1 140  1  1^I^I  100  I  1^i^1 60  f  Ppm  138-1  epoxy (EPON-825) resins, which gives a small peak at 43 ppm in the 15N solidstate NMR spectrum (Figure 3.1A) and disappears in the non-protonated nitrogen selection experiment (Figure 3.1B). However, this contribution to the cross reaction is quite limited compared to the major cross reaction which forms the oxazoliclinone ring basic structure 40. CH3 -CH -CH3  I 0 \  C=NH  60  (o)- o1  3.8. The Mechanism of the Curing Reaction for Dicyanate / Diepoxy Mixed Resins As discussed above, the reactions of monofunctional model compounds yield several quite interesting products which can be isolated and characterized. All of these products are directly related to the curing reaction of the dicyanate and diepoxy resin mixture. Each of them corresponds to a specific cross reaction between the two resins, and can be easily identified by comparison of the high resolution solution NMR spectra for the low molecular weight products and the solid-state NMR spectra for the cross-linked resins. Among these products, the most important one is the major cross reaction adduct 40 derived from monocyanate 25 and monoepoxy 37. The possible mechanism for this cross reaction is as proposed in Equation 3.2. Because it comes from the direct reaction between one cyanate and two epoxy molecules, it corresponds to a cross-reaction product between the two resins and makes a major contribution to the large peak at —36 ppm in Figure 3.1. From both the product percentage yields for the model compound reactions and the 15N solid  139  state NMR spectrum of the bulk cured resin mixture, it can be concluded that this cross reaction is the major one between the two resins, even though it is limited to —12%. Cross-linking between the two resins can also occur through the reaction of the cyanate with the hydroxyl groups in the epoxy resin, which gives the product with the imidocarbonate type of structure 60. However, its contribution to the overall cross reaction is much less than that of the major cross reaction mentioned above. The reactions of monocarbamate 51 and monoepoxy 52 give products 53, 54, and 55. Product 53 shows a resonance at —40 ppm in its 15N NMR spectrum  and corresponds to the product in the cured mixed resins which makes partial contribution to the peak at —36 ppm in Figure 3.1. However, this contribution is very limited in comparison to that of product with the structure similar to 40 from the major cross reaction. Its presence can be detected in the variable contact time experiment (Figure 3.2) and the non-protonated nitrogen selection experiment (Figure 3.1B). In the non-protonated nitrogen selection experiment, the intensity of the resonance at —36 ppm remains almost constant. Although product 53 is not formed from the direct reaction between epoxy and cyanate, it is related to the cross-reaction between the two resins. During its formation [Scheme 3.1], the phenol moiety is cleaved from the cyanate, and can continue to react with epoxy to form compound 54 [Equation 3.4]. Since this phenol moiety comes from the cyanate resin, it can be considered that this reaction [Equation 3.4] is a cross-reaction between the two resins and that product 54 is also a cross-reaction product. Product 55 has a resonance at 23 ppm in the 15N NMR spectrum and corresponds to the species in the cured mixed resins which yields a small peak at 23 ppm in the 15N solid-state NMR spectrum (Figure 3.1). As in the case of 140  product 53, the formation of product 55 also involves elimination of a phenol moiety from the cyanate (Scheme 3.1), and then reaction of this phenol with epoxy to form the cross-reaction product 54. Intermediate 30 is a trimer of moieties originally in the cyanate, with loss of the phenol part from the cyanate resin. It only links epoxy monomers, and thus acts as a curing agent for the epoxy resin. Therefore, compound 55 is not a cross-reaction product between the epoxy and cyanate monomers but the product of a curing reaction for the epoxy resin with intermediate 30 as the curing agent.  3.9. Conclusions The possible cross reactions which solid-state NMR indicates occur between cyanate-functionalized and epoxy-functionalized resins have been investigated using both natural abundance and labelled monofuctional model compounds. These soluble products were isolated and purified by adsorption chromatography and gel permeation chromatography, and then fully characterized by high resolution 1H, 13C, 15N NMR spectroscopy and by mass spectrometry. The major cross-reaction product between cyanate and epoxy resin monomers contains one cyanate monomer and two epoxy monomers, and very clearly indicates that cross reaction between the cyanate and the epoxy can occur during the mixed resin curing process. The balance of all information to date is considered to favor an oxazolidinone structure 40 for the major crossreaction product. The reaction between the cyanate and the hydroxy groups in the epoxy resin is also a cross-reaction pathway even though it is very limited. Besides the direct reaction of epoxy and cyanate, cross reactions for the epoxy and cyanate mixed resins can also occur through the reaction of epoxy and the carbamate derived from the cyanate group. 141  However, epoxy consumption lags cyanate consumption in the overall reaction as triazine formation from the cyanate is much faster than the two competing reactions, the cross reaction between cyanate and epoxy, and the selfpolymerization of epoxy, under the conditions investigated. Thus, cross reaction between cyanate and epoxy is limited; approximately 12% cross reaction between cyanate and epoxy was found in the overall reaction under the curing conditions used in the present study.  142  CHAPTER 4. SYNTHESIS AND CHARACTERIZATION OF CROSS-LINKING AGENTS FOR MIXED CYANATE / EPDXY RESIN SYSTEMS AND FOR MIXED CYANATE / OLEFIN RESIN SYSTEMS The end use of a polymer system is often decided from engineering requirements. The structure at the molecular level is not the only important factor when an application requires a bulk property to be within a specific range. Thus, many synthetic polymers are designed such that the physical properties will be optimum for a particular end use. In the past this meant developing completely new polymers, but in the last fifteen years, there has been much interest in the properties and applications of multicomponent polymer systems.[98] It is often found that mixing two polymer systems together can produce a synergistic effect combining both desirable advantages of the two systems. In this way, the desired mechanical properties and processing capabilities can often be obtained. Thus, a specialized polymer system can be obtained by combining two or more different but known polymers. Multicomponent polymer materials are defined as mixtures of two or more structurally different polymeric species, such as polymer blends, blocks, grafts, cross-linked polymers, or interpenetrating polymer networks (IPN's). They can be classified into two groups in terms of the type of bonding between different components.[991 Polymer blends and IPN's do not have chemical bonds between the different components, while block copolymers, graft copolymers and crosslinked polymers contain intermolecular covalent bonds which hold the different species together.  143  The cross-linking bonds in cross-linked polymers are formed by cross reactions which can proceed directly through the side groups on different components or through a cross-linking agent which can react with the two different components and hold them together. In the latter case, the desired degree of cross-linking can be exactly controlled by controlling the amount of cross-linking agent added.  4.1. A Cross-Linking Agent for Mixed Cyanate / Epoxy Resin Systems As described previously, the advantages of using epoxy resins for printed circuit boards are that the high mechanical performance and low shrinkage properties of these resins provide toughness and mechanical strength for the circuit boards. In addition, epoxy resins can be cured at any temperature between 5 °C to 150 °C depending on the choice of curing agents. They can also be modified in many ways for various specific applications or properties.[39] However, compared to the cyanate resins, epoxy resins have the disadvantage of relatively high dielectric constants, which make the signal transfer speed slower. [17,30] Furthermore, epoxy resins show high moisture absorption and low heat insulation relative to the cyanate resins. As a result, the performance of printed circuit boards made from epoxy resins may not be reliable when they are operated at elevated temperature or in humid environments. In general, cyanate resins have the advantages of low dielectric constants, low moisture absorption, and high thermal stability.[17,301 However, a serious disadvantage of cyanate resins is the lack of mechanical strength. A circuit board made from cyanate resin alone can be easily broken in use. Thus it can be seen that both cyanate and epoxy resins show some desirable properties for printed circuit boards but neither of them has all of the ideal characteristics. Therefore, cyanate resins are usually cured together with 144  epoxy resins to try to get the desirable properties from both. However, as seen in the previous chapter, although cross reaction between the cyanate resin and the epoxy resin can occur during the curing process, it is of limited efficiency, and is complex and difficult to control. Thus, obtaining an alternative to the direct cross reaction for cross-linking the two resins will be very important for better control of the properties of the final resin. This might be done by the addition of a cross-linking agent which is a "mixed monomer", ie which contains both cyanate and epoxy functional groups. This would give a predictable degree of cross-linking under controlled conditions which could be varied by changing the proportion of the cross-linking agent. In addition, since the curing reaction of a cyanate resin can be initiated thermally while a base is needed for curing an epoxy resin, a two stage curing process could be carried out. This is the main idea behind the research in the present chapter.  4.1.1. Strategy for the Synthesis of the Cross-Linking Agent, the Monoglycidyl Ether of Bisphenol-A-monocyanate 61  NCO  OCH2CH —CH2 \/  o  61  The mixed functional target molecule, the monoglycidyl ether of bisphenol-A-monocyanate 61 or 2-(4-cyanatopheny1)-244-(2,3-epoxypropoxy)phenylipropane, was chosen as the cross-linking agent to be synthesized. The two possible routes considered for the synthesis are shown as A and B in Scheme 4.1. They are different only in the order in which the functionalities are added to  145  the starting material, bisphenol A (3). Each route includes two reactions and one intermediate species. HO  OCH2CH-,CH2  o  62  Route A  OH  HO  NCO  OCH2Ctl-ICH2  o  3  61 {4}  Route B NCO  OH  63 Scheme 4.1  By examination of the conditions for the four different reactions, it was considered that route A was more likely to be successful than route B for several reasons. Although the addition reactions of the cyanate and the epoxy functional groups both involve base, heating is required for the addition of the epoxy group, whereas the formation of the cyanate proceeds at a much lower temperature (0 °C). From the curing mechanism for a cyanate resin discussed in Chapter two, the cyanate functional group formed in the reaction of step {3} in route B might be trimerized or destroyed by the high temperature used in the reaction of step {4} in route B. Furthermore, due to the presence of bases in both reaction steps {3} and {4}, the intermediate product of route B, bisphenol-A monocyanate or 2(4-cyanatopheny1)-2-(4-hydroxyphenyl)propane 63, may react with itself to form poly(bisphenol-A imidocarbonate) 64 as shown below. In contrast, in route A of 146  the synthesis, the epoxide intermediate, the monoglycidyl ether of bisphenol-A or 244-(2,3-epoxypropoxy)pheny1]-2-(4-hydroxyphenyl)propane 62, will remain unreacted under the low temperature (0 °C) conditions for the formation of the cyanate in step {2} in Scheme 4.1.  OCN  HO  64  Consequently, route A in Scheme 4.1 was chosen for the preparation of the target cross-linking monomer, the monoglycidyl ether of bisphenol-Amonocyanate 61. The procedure for the synthesis was first to synthesize, purify, and characterize the intermediate, the monoglycidyl ether of bisphenol-A 62 and then to prepare the target monomer 61 from this intermediate. After the successful synthesis of the intermediate 62 and the target cross-linking monomer 61, the independent curing reaction of the cyanate functionality in the cross-linking monomer 61 was studied by application of heat. The curing of epoxy functionality was also studied using a base as a curing agent. It is shown that the curing reaction of the epoxy group in the cross-linking monomer 61 is not independent of the curing of the cyanate group. Characterizations were carried out by NMR and MS experiments as described in the Experimental chapter. In the following sections, the investigations of these reactions and the characterization of the product species will be discussed in detail.  4.1.2. Synthesis and Purification of the Monoglycidyl Ether of Bisphenol-A, 62 In general, the formation of epoxy resins is described by the following scheme (Scheme 4.2):[100]  147  CI -CH2R1-/C H2  OH  HO  3  HO  OCH2C1-,1---,CH2  62 CI-CH2CH-/CH2 0  OCH2Ctl-/CH2  17 HO  CH2 -CHCH 0 / 0  OH  OCH2CHCH 0  OH  66 OH CI-CH2Cti-/CH2 0 OC H2 H C H2 0 OH  OCH2CH- C H2  \ /  o  67 HO  012-/CHCH  OH^01-CH201-1-CH2 0 OCH2CHCH 0 -^OH  18  OCH2CH CH2 -  \ /  o  Scheme 4.2  148  Structure 18 represents the general formula of epoxy resins based on bisphenol A and epichlorohydrin. Like compound 17, such compounds are also termed diglycidyl ethers since all of them contain two glycidyl ether groups, 0 /\ 0-CH2CH—CH2 , per molecule. In order for the final products to be commercially useful as epoxy resins, it is necessary that the polymers are terminated by epoxy groups, through which they may be subsequently crosslinked. This is achieved by carrying out the polymerization with an excess of epichlorohydrin. In a typical process for the preparation of a epoxy resin, a mixture of bisphenol A and epichlorohydrin (about 1:4 molar ratio) is heated to about 60 °C with stirring. Solid sodium hydroxide (2 moles per mole of bisphenol A) is added slowly. The unreacted epichlorohydrin is then removed by distillation under reduced pressure.[- 00] As seen in the scheme, the monoglycidyl ether of bisphenol-A 62 exists as an intermediate at the beginning of the course of an epoxy resin synthesis even though it has never been isolated as a synthetic target product. Thus, it was thought that it should be possible to synthesize the intermediate product 62 by reacting bisphenol A and epichlorohydrin with a suitable base, but that a 1:1:1 of molar ratio of bisphenol A, epichlorohydrin, and base should be used. The major product expected from this reaction should be the desired intermediate 62. However, since the two hydroxyl groups in bisphenol-A are identical, the selective epoxidation of one of them is very difficult. The product from over reaction, the diglycidyl ether of bisphenol-A, and some unreacted bisphenol-A are also to be expected as side products. Therefore, purification of the desired intermediate 62 must be conducted after the preparation. This was accomplished by extraction followed by adsorption chromatography in the following experiments.  149  A moderate reaction temperature of 56 °C and relatively short reaction time of 90 minutes were chosen for the initial reaction conditions.[1011 The resulting crude product was identified by NMR spectroscopy to be a mixture of 61, 17, 3 and some unknown side products as shown in Figure 4.1. As a  consequence, purification of this crude mixture was carried out to obtain pure intermediate product 62. The sequence chosen for the purification was first extraction of the crude product using different organic solvent with increasing polarities, followed by separation of each component using a silica gel column. The details of this will be presented later. However, it was found that the intermediate 62 could not be isolated as a single pure component by this purification procedure. The best fraction obtained from the silica gel column was a mixture of intermediate 62 and bisphenol A (3). Thus, in order to obtain pure 62 from the silica gel column, there should be no bisphenol A present in the  crude mixture. This means that all of bisphenol A must react completely at the synthesis step. Although other methods (such as using a combination of different chromatographic sequences) might also yield pure 62, varying the reaction conditions to eliminate bisphenol A at the synthesis step was thought to be the most straightforward approach. On the other hand, variations of the reaction conditions in the preparation step might not only allow to complete the consumption of bisphenol A, but also could produce a higher yield of the desired product 62. For investigations of the efficiency and selectivity of the epwddation process, a number of reactions were carried out under different conditions with variation of the reaction temperatures and times and the choice of a base used. The optimum reaction conditions for the preparation of intermediate 62 were found to be using potassium carbonate as the base and keeping the reaction temperature at 80 °C and the reaction time at 150 minutes (see Equation 4.1), 150  Figure 4.1. 13C NMR spectrum (1-H at 200 MHz) in acetone-d6 of the crude product prepared by reaction at 56 °C for 90 min. and using a reactant mixture with a 1:1:1 molar ratio of bisphenol A, epichlorohydrin and potassium carbonate. (A). Full spectrum; (B). The expanded aromatic region. 151  12g 13g 14g  CH2CH—CH2 \ /  o  A  1p  ip 12p 13p 14p  HO  OCH2CH-CH2 \ /  o  3b 4b 7b  lb  HO  9g 9p 4p  OH  4b  3p 3b lOg 10p  7g 7p 7b 12g 12p  14g 14p 13g 6g 13p 6p 6b  T^7^T  40 ppm  80  120  160  2P 2b  I1g  llp --2(r/fr--  8p  144.5  8g  5b^5p  747  144.0  ppm  156.5^156.0  157.5^157.0  143.5  143.0  142.5  ppm 15 1-1  HO  0 0  OH 4. CI-CH2C\H-,CH2  K2 CO3  0^80°C  HO  C• 0  OCH2CH-CH2 '0/^[4 .1]  150 min.  which will be given in detail in the Experimental chapter. Under these conditions, the major component is the desired intermediate product 62. Minor side products of this reaction are compounds 17 and 66 and some unidentified compounds. The most important feature is that there is no bisphenol A in the crude product mixture formed under these reaction conditions and pure intermediate 62 should be obtained using the same purification procedure as mentioned above, which are: Purification of the desired intermediate 62 was accomplished first by extraction using solvents of increasing polarities, then followed by chromatographic separation of the proper fraction from the extraction using a silica gel column. The purpose of using the extraction as a preliminary purification step is to remove any polymeric species formed during the reaction. Due to their adhesive properties, these species could damage the silica gel bed and affect the resolution of the column. Therefore, prior removal of these species gives better column separations. As a result, pure monoglycidyl ether of bisphenol-A 62 was obtained through this purification sequence as a crystalline material (m.p. = 86 — 88 °C). The overall yield is about 20 %. Both MS and NMR experiments are consistent with the structure. The mass spectrum shows M+ = 284 a.m.u.. Its 13C and 1H NMR spectra are shown in Figure 4.2. Since the monoglycidyl ether of bisphenol-A 62 is composed of half moieties of bisphenol-A (3) and bisphenol-A diglycidyl ether (17), compound 62 gives very similar NMR spectra to those of a  152  Figure 4.2. NMR spectrum (1H at 200 MHz) of the monoglycidyl ether of bisphenol-A, 62, in acetone-d6. (A). 13C spectrum; (B). 1H spectrum.  153  1p  3p 4p 7P 9p 10p 11p 12p 13p 14p 2^  OCH2CH-CH2  HO^  A  9p 4p  3p^ 10p  \ /  0  62 7p  14p 12p  13p 6p  2p lip  451)5  ^•■.0kMOMMT■IMWO I^I^I  160^120^80^40  ^  Ppm  7p  12p  9.0  7.0  5.0  14p  3.0  ppm 153-1  mixture of compounds 3 and 17. The assignments in Figure 4.2 are made from the known spectra of bisphenol A (3) and the diglycidyl ether of bisphenol-A (17). [102]  4.1.3. Synthesis of Cross-Linking Monomer 61 After successful synthesis of the intermediate compound 62, the target cross-linking monomer 61 was synthesized by using 1:1:1 molar ratio of purified 62, cyanogen bromide, and triethylamine as a base as shown in Equation 4.2.  The detailed procedure for the preparation is given in the Experimental chapter.  HO-(0)--^-(0)-0CH2CH-CH2 \ /  + BrCN 4. (CH3CH2)3N  0  62  < 0 °C  [4.2]  OcH2o\H-/CH2 + (CH3CH2)3NHBr  NCO  61  o  The crude product 61 is a viscous light yellow liquid. An attempt to purify this crude product was made using distillation under reduced pressure, but was not successful due to the high boiling point of compound 61 and its trimerization at elevated temperatures. However, the NMR spectra of the product 61 in Figure 4.3 show that there are only very minor impurities present (resonance b is due to the residual benzene solvent). Most importantly, the 13C spectrum (Figure 4.3A) shows that the cyanate functional group which gives a resonance at 109 ppm is indeed formed in the reaction. In addition, the 1H spectrum (Figure 4.3B) shows that only a very small phenolic proton signal at 8.2 ppm is left,  154  which means most of the hydroxyl groups are converted to cyanate groups. Meanwhile, the three characteristic peaks at high field, which come from the glycidyl ether group, remain unchanged. This proves that the formation of the cyanate at step {2} in Scheme 4.1 does not affect the glycidyl ether functional group formed in the previous step {1}. The mass spectrum shows the parent ion peak of product 61 at 309 a.m.u., which is consistent with the NMR experiments and confirms the formation of the cyanate, Therefore, the desired cross-linking monomer 61 has been successfully synthesized and characterized.  4.1.4. Curing of the Cross-Linking Monomer 61 with Heat Since the cross-linking monomer 61 has two different functional groups, cyanate and epoxy, and each of them can be cured by different mechanisms, it was thought that monomer 61 could react independently with the cyanate or the epoxy functional groups on other molecules. The first attempt was to cure the cyanate functional group of the monomer alone. The monomer 61 was simply heated at 180 °C for 15 minutes. A brown-yellow solid was produced which was identified as compound 68 from MS and NMR experiments (Figure 4.4). The mass spectrum gives the parent ion M+.927 for the solid product, which is three times the mass of the cross-linking monomer 61 and corresponds to the trimer structure 68. The 13C NMR spectrum (Figure 4.4A) shows that all of the cyanate groups have reacted (the peak at 109 ppm disappears) and triazine rings have formed (the resonance at 174 ppm appear). Particularly importantly, the 13C NMR spectrum shows the epoxy groups remain unreacted during this curing process as reflected in the three characteristic peaks at high field for the glycidyl ether moiety. The 1H spectrum (Figure 4.4B) is similar to that of the monomer 61 (Figure 4.3B), which also means the glycidyl ether  155  Figure 4.3. NMR spectrum ( 1 H at 200 MHz) of the crude monoglycidyl ether of bisphenol-A-monocyanate 61 in acetone-d6. (A). 13 C spectrum; (B). 1 H spectrum.  156  lc NCO 2c  A  3c 4c^7c^gc 10c 1 lc 12c 13c 14c  OCH2CH—CH2 \01  61 b 4c^3c  9c  10c 7c  13c 14c  12c  6c 11c  2c 5c  8c lc I 6...irmpod  !^  160  I^I  11111]^-1 120 80  t  1 40  1^1 ppm  B  c  Jil P 4c  .^910c  I^' -• -I ' -` -.- -' T '  9.0^7.0  5.0  3.0  PPIn 156-1  Figure 4.4. (A). 13C NMR spectrum (111 at 300 MHz) in acetoned6 and MEK of the crude product 68 from curing the monoglycidyl ether of bisphenol-A-monocyanate 61 by heating; (B). 1H NMR spectrum (200 MHz) in acetoned6 of the crude product 68. 157  7t  at 4t  9t 10t 11t  o  12t 13t 14t  00H20F1-0H2 \/ 0  se  A  9t 4t  10t  13t  3t  it  14t 7t  1 t  6t  lit^8t 2t 5t  ^ 1^i^1 ^ 1 ^ ^ 1 40 160 120 80 ^  "  9.0  I  '  7.0  ,  5.0  44.40,400.0•0_00,,  i ^ ,^ 1 Ppm  ,  3.0  1.0 Ppm  functional groups are unchanged. Therefore, it is confirmed that the cyanate functional group in the cross-linking monomer 61 can be cured independently without affecting the epoxy functional groups. OCH2 CH - C H2 \  o  /  68 OCH2CF\I / CH2  4.1.5. Curing of the Cross-Linking Monomer 61 with Base The independent reaction of the cyanate functional group in the crosslinking monomer 61 by heat has been verified. The next step was to try to find a way to independently cure the other functional group, the epoxy, in the monomer 61. Because cyanate groups are more reactive than epoxy groups under heating, the consumption of the cyanate resin is faster than that of the epoxy resin in a mixed resin system during a heat-curing process as shown in Chapter 3. For this reason, heat should be avoided if it is desired that only the epoxy functional group in monomer 61 be cured. From the curing mechanism for epoxy resin mentioned before, [39,1 00] a amine with an active hydrogen [diethylamine, (CH3CH2)2NH] was used as a curing agent to try to react only with the epoxy group in monomer 61 at room  158  temperature. A mixture of monomer 61 (0.26 mmole) and diethylamine (0.15 mmole) was stood at room temperature for one day. Its 13C NMR spectrum is shown in Figure 4.5. The spectrum shows that the reaction of the cyanate group is still faster than that of the epoxy group under these basic conditions since the cyanate resonance at 109 ppm disappears completely while the three characteristic glycidyl resonances at higher field are still present. Meanwhile, resonances derived from the cyanate group at 167 ppm and 174 ppm for an unknown are produced during the reaction, which indicates that most of the cyanate groups are transformed to this unknown. According to the literature,[9] aryl cyanates react with secondary amine to form N,N-disubstituted isoureas 69. If aryl cyanate is excess, the isoureas 69 can continue to react with aryl cyanates and form triazine 70 (see Equation 4.3). Therefore, the unknown product obtained from the reaction of monomer 61 and diethylamine can be characterized as the compound with the triazine structure 72. The MS spectrum shows a parent ion at 716 a.m.u., confirming the identification. The resonances at 167 ppm and 174 ppm in the 13C NMR spectrum (Figure 4.5) are due to the two different carbons on the triazine ring of 72. The small resonance at 158 ppm is assigned to the carbon in isourea 71, which is derived from the cyanate carbon in monomer 61. Furthermore, this can be verified by a reaction of monomer 61 with excess of diethylamine, which leads to formation of the isourea 71 only without further reaction to form the triazine 72. A mixture of monomer 61 (0.28 mmole) and diethylamine (0.43 mmole) was stood at room temperature for one day. Its 13C NMR spectrum (Figure 4.6) shows there is no triazine 72 formed. All of the cyanate carbons are converted to the isourea carbons of 71, showing a resonance at 158 ppm. The MS spectrum is consistent with the result, showing a parent ion at 382 a.m.u.. The overall reaction can be represented as in Equation 4.4. 159  Figure 4.5. 13 C NMR spectrum ( 1 H at 200 MHz) in acetone-d6 of the crude product from curing the monoglycidyl ether of bisphenol-A-monocyanate 61 at room temperature with diethylamine base (monomer 61 is in excess).  160  116 117 CH3CH2N,,CH2CH3  m_r, 14^ 101 ts..............a,v 12 vi 12 w t  111 115^114113^,,,N,.... CH2 \& —CHCH20 112  OCH2C Il'-'-/C H 2  o  0  111  115  15  114  14  18  113  1 116  I  I  160  i  T  120  1  1  80  I  5  40  117  i^• Ppm  110 19^17^14^i3 114^113 112^111 0  C4--ICHCH  .8  0  115^116  ,CH2CH3 0—C—N\ NH CH,CHQ II —  i4 i9  i3  115  110  116 17  114  112  113  16  11 1 11^15^ i2  18  ‘60,411.••■•~100..~4.ftivera v~oiemmillin 1^ 1^I^I^I^1^-,^1^1^.^1 BO^160^140^120^100^80^60^40^20 PPM '^  ,  Figure 4.6. 13C NMR spectrum (1H at 200 MHz) in acetone-d6 of the crude product from curing the monoglycidyl ether of bisphenol-A monocyanate 61 at room temperature with excess diethylamine base.  161  Ar-O-CEN + HN  /R  R^R' 'N'  r R 2 ArOCN Ar-O-C-N ii^\,, NH rl - ArOH  \ R'  [4.3] Ar0  69  OAr  70  0—C—N + HN  ,CH2CH3 ‘CH2CH3  61 0 /\ CH2 —CHCH 0  CH 2 CH 3 / 0-C-N II^\ " NH CH2ru  [4.4]  71  excess monomer 61 CH 3CH2,N,CH2CH3  0 /\ CH2 —CHCH -  72  o  /\  0-CH2 CH —CH2  From these experiments, it can be concluded that the epoxy functional group in the cross-linking monomer 61 can not be cured independently without affecting the cyanate functional group. However, it should be emphasized that even though the epoxy group could not be cured independently, monomer 61 still can be used as a cross-linking agent. It should be possible to use it in a mixed cyanate and epoxy resin system either by first curing monomer 61 with the  162  cyanate resin under mild heating and then curing it with the epoxy resin by addition of base, or by curing all of them together under the action of both heat and base curing agent.  4.1.6. A More Practical Approach to the Application of the Cross-Linking Monomer, 61 As seen previously, the most difficult part in the synthesis of the crosslinking monomer 61 is to obtain the pure intermediate compound 62, and considerable effort was expended for this purpose. Even though it was successfully obtained in pure form, its yield is perhaps too low (only approximately 20 %) to be used in practical applications. In addition, the chromatographic separation is not an efficient process for economic reasons. Fortunately, these procedures are necessary only for characterization purposes; it is not necessary to get the purified intermediate compound 62 in a real synthetic procedure for production of final mixed resins. From the results obtained, the major side products in the synthetic step {1} presented in Scheme 4.1 are the overreacted the cliglycidyl ether of bisphenol-A (17 or 18) and unreacted bisphenol A (3). The first one (17 or 18) is actually an epoxy resin monomer or precursor, one of the  components in the mixed cyanate/epoxy resin system. The second one (3) would be transformed in the next synthetic step {2} to bisphenol-A dicyanate 2, which is also one of the components in the mixed resin system. Therefore, without separation after step {1}, the final product after step {2} will be a mixture of the cross-linking monomer 61, the cyanate resin and the epoxy resin, which is just the desired mixed resin system. Thus, the final major side products are compatible with the real application system and do not affect the application of the cross-linking monomer 61 in any way. 163  To check these conclusions, a real mixture of the cross-linking monomer 61, the cyanate resin and the epoxy resin was synthesized and tested. A mixture of bisphenol-A, epichlorohydrin, and potassium carbonate with 1:1:1 molar ratio was reacted at 80 °C for 90 minutes. The NMR spectra show the crude intermediate product to be a mixture of three major components: bisphenol-A 3, the monoglycidyl ether of bisphenol-A 62, and the diglycidyl ether of bisphenolA 17 plus some high molecular weight species. The 13C spectrum also gives the approximate percentages of each component in this intermediate mixture: about 28% for bisphenol A, 38% for the monoglycidyl ether of bisphenol-A and 34% for the diglycidyl ether of bisphenol-A and some high molecular weight species. From these data, the amounts of cyanogen bromide and triethylamine base for adding the cyanate functional groups into the intermediate mixture were calculated. The details of the preparation procedure for both synthetic steps are given in the Experimental chapter. NMR spectra of the final product mixture containing the cross-linking monomer 61 were obtained (Figure 4.7). Particularly importantly, the 13C spectrum (Figure 4.7A) not only shows that the cyanate functional groups are formed, but also that there are two main kinds of cyanate group in the final cross-linking product mixture, as reflected in the two very similar peaks at the cyanate region (-109 ppm). In comparison with the cyanate spectra obtained before, the one at higher field can be assigned to bisphenol-A dicyanate 2 and the one at lower field can be assigned to the monoglycidyl ether of bisphenol-Amonocyanate 61. On careful examination, one or more additional small cyanate peaks are also found nearby at lower field, which are perhaps due to some higher molecular weight species. The 1H spectrum (Figure 4.7B) shows that the peak at — 8.2 ppm, which belongs to the hydroxyl groups on a phenol ring, almost completely disappears. This means that all of the hydroxyl groups on the 164  Figure 4.7. NMR spectrum (1H at 200 MHz) in acetone-d6 of the cyanate product mixture obtained from the intermediate mixture without separation. (A). 13C spectrum; (B). 1H spectrum.  165  A NCO  OCN  2 b  a  NCO  OCH2CH-CH, \ i^..  b  o  61  IIIIIIIIIIIIIIIIII  110^109 I  ....1._..I.A.  .a...............LIL......1J^  ^1  160^120^80^40 ppm  r  8.0  6.0  4.0  2.0 Ppm 165-1  different phenols have been converted to cyanate groups, and thus this preparative procedure is very efficient for practical applications of the mixed resin system. This final product mixture containing the cross-linking monomer 61 was cured using diethylamine as a curing agent at room temperature for one day. The results are similar to that of a relative pure monomer 61 discussed above. The final product of the curing reaction also depends on the molar ratio of the monomer 61 to diethylamine. 13C NMR spectra (not shown) also show that the cyanate functional groups react faster than the epoxy groups as before since all of the cyanate peaks at — 109 ppm have disappeared while the majority of the three high field peaks for the epoxy groups are still left. A very important goal for use of this mixed product system is to get a tough and strong cured mixed resin material. The product mixture containing the cross-linking monomer 61 was cured at 180 °C for 3 hours. A very tough and strong resin material was obtained, which is much tougher and stronger than the cured resin obtained by curing the cyanate resin and the epoxy resin together as described in Chapter 3. Further investigations of the preparation on a large and more practical scale under industrial conditions are warranted.  4.1.7. Conclusions The desired cross-linking monomer for a mixed cyanate resin and epoxy resin system, the monoglycidyl ether of bisphenol-A-monocyanate 61, has been synthesized and characterized. The intermediate compound, the monoglycidyl ether of bisphenol-A 62, was also synthesized and purified by extraction and chromatographic separation using a silica gel column. The cyanate functional group in the cross-linking monomer 61 can be cured independently by heat to form the triazine structure 68, but the epoxy functional group in the cross166  linking monomer 61 can not be cured independently of the cyanate group because the latter is more reactive than the epoxy group under both heat and basic conditions. By using a secondary amine, diethylamine, as a curing agent, the cyanate groups in the cross-linking monomer 61 react with diethylamine to form the types of structure 71 or 72, depending on the molar ratio of monomer 61 to diethylamine. A more practical approach for the application of the crosslinking monomer 61 has been discussed and tested. Most interestingly, under heat curing, a very tough and strong resin material was produced from this cross-linking mixed resin mixture.  4.2. A Cross-Linking Agent for Mixed Cyanate / Olefin Resin Systems An interpenetrating polymer network (IPN) is a particular kind of multicomponent polymer system. The original definition of INP requires both chemical species to be self cross-linked.[99] However, if one component is a linear thermoplastic polymer while the other is cross-linked thermoset polymer then the final system will be only partially, but selectively cross-linked, and is referred to as a Semi Interpenetrating Polymer Network (SIPN)[16]. The SIPN system is used to combine the advantages of both the thermoplastics and the thermosets. Thermoplastics are the dominant polymers in low temperature engineering applications because they are tough and easy to process. However, most thermoplastics have relative low Tg values, and will lose their hardness and mechanical strength at high temperatures. In addition, most thermoplastics must be heated to 150 °C or more above their highest end use temperature before their viscosity is low enough to allow for processing, while most organic molecules decompose at a significant rate at 350 °C or above. Therefore, very few thermoplastics have end use temperatures of 200 °C or above. 167  Unlike thermoplastics, thermosets are cross-linked polymers and usually have high Tg values. They can be processed at temperatures not far above their highest use temperature. Thus, the end use temperature of a thermoset can be close to its decomposition temperature. However, high temperature thermosets are generally brittle and thus lack the toughness of the thermoplastics. The concept behind SIPN is again to obtain the most attractive features of both materials, the toughness and the high end use temperature[l 6]• In a SIPN system, the linear polymer is not formally cross-linked, but it will be highly entangled. The final material should be less rigid than a full IPN when used above the glass transition temperature of the linear polymer, but it should be tougher since the linear molecules may be free to move. As a result of SIPN technology, materials, such as cyanate resins, conventionally regarded as being too brittle for certain applications, can now be used. Dicyanate Semi Interpenetrating Polymer Networks (SIPNs) are very useful as matrix materials.[16,29,34] They can be made by dissolving a thermoplastic in cross-linking dicyanates and then curing the resulting mixture. The SIPNs produced in this way could be very strong, with tensile strengths of 10,000 to 12,000 psi, and flexible, with elongations to break of 10 to 17 percent. Dicyanate SIPNs also have good thermal stability. The softening temperatures of the SIPNs are significantly higher than those of the corresponding thermoplastics. [34] However, there is no cross reaction between the cyanate resin and the thermoplastics, such as polyolefins, in the dicyanate SIPNs. It was thought that the addition of a certain amount of cross-linking agent, which could link the two polymers together, should make the SIPNs even stronger and tougher. Thus, the purpose of this portion of the research was to design and synthesize a cross-  168  linking agent for mixed cyanate and olefin resin systems as a complement to the SIPN technique.  4.2.1. Synthesis and Characterization of the Cross-Linking Monomer 76 For the cyanate resin and olefin resin the cross-linking agent should contain both cyanate —OCN and olefinic ^ CC ^ groups. This can be achieved by reaction of cyanogen bromide BrCN with an appropriate phenol with an attached olefinic group. 2-allylphenol 75 was chosen for use in this project as it is easily produced by a reaction of phenol and allyl chloride (or 3-chloro-1 -propene, 73) with a strong base, followed by Claisen rearrangement as shown in Equation 4.5,[103] and it is inexpensive and commercially available. OH CH2=CHCH2CI +  73  NaOH  74 200 °C  ^  [4.5]  OH CH 2C H ----CH 2  75  The target monomer 2-allylphenyl cyanate 76 was obtained by reaction of 2-allylphenol 75 and cyanogen bromide in the presence of a base, triethylamine, at low temperature (molar ratio 1:1:1) as shown in Equation 4.6. The details of the preparative procedure will be given in the Experimental chapter.  169  OH  CH2CH=CH2  < 0 °C  + BrCN + (CH3CH2)3N  75 OCN CH2CH=CH2  (CH3CH2)3NHBr^[4.6]  76  The final product is a colorless liquid which can be purified by distillation under reduced pressure (1.4 mmHg) at 55 °C. The mass spectrum shows the parent peak at M+ = 159 a.m.u.. Its NMR spectra are given in Figure 4.8 together with the complete assignments. The —OCN group gives a 13C resonance at 109 ppm which is consistent with the other cyanates previously studied.  4.2.2. Self Curing Reaction of the Cross-Linking Monomer 76 On heating 2-allylphenyl cyanate at 180 °C for 2 hours, a solid compound is produced. This solid product was purified by recrystallization from acetone, and had m.p. = 110 — 111 °C. Its mass spectrum gives M+ = 477 a.m.u.. The NMR spectra are shown in Figure 4.9. The 13C NMR spectrum shows a resonance at 174 ppm which is a characteristic of triazine ring carbons. Therefore, both MS and NMR experiments indicate the solid product obtained from curing 2-allylphenyl cyanate 76 is 1,3,5-tri(2-allylphenoxy)-2,4,6-triazine 77. This is consistent with the curing reactions of the other cyanates discussed previously and can be represented as in Equation 4.7.  170  Figure 4.8. NMR spectra (1H at 200 MHz) of 2-allylphenyl cyanate 76 in acetone-d6. (A). 13C spectrum; (B). 1H spectrum.  171  3  54 9 6 7  A 10 OCN 7 8 9 6 (.....\ 2 CH2CH=CH2 x .._.) '^3 76 4  8  /  1  2 10  1  I  810  1  410 ppm'  B  8  9.0  7.0  ,  5.0  •^I^'^•^•^•^I^I  3.0 Ppm  Figure 4.9. NMR spectra ( 1 H at 200 MHz) of 1,3,5-tri(2allylphenoxy)-2,4,6-triazine 77 in acetone-d6. (A). 13 C spectrum; (B). 1 H spectrum.  172  A 9 87  CH2=CHCH  3 54  6  7  9 8  2  1 10  160^10^80^Ppm  9  7  r-  8  •■•••  TrI•1171,111■VTT^  T/T7711-TIVITY^vlilir7111,1•  8.0^6.0^4.0^Ppm  CH2CH=CH2 OCN  CH2CH =CH2  3  0 N 0  A  YOY N^N Y  76  C H 2 C II = C H 2  [4.7]  0  CH2=CHCH2  77  4.2.3. Curing Reaction of the Cross-Linking Monomer 76 with a Cyanate Resin As a cross-linking agent, 2-allylphenyl cyanate 76 should not only react with itself. In order to link the cyanate resin, it must react with the cyanate group of the cyanate resin on which the cross-linking monomer 76 is desired to be anchored. This is a concern because the structure of the cross-linking monomer 76 is quite different from the cyanates studied previously which have no ortho substituents. To test this, a mixture of 2-allylphenyl cyanate 76 and p-tert-butylphenyl cyanate (PTBPCN, 25) was cured at 180 °C for 3 hours. As previously, the PTBPCN 25 was used as a model compound of the real cyanate resin to ensure that the final products were soluble and could be characterized by high resolution solution NMR spectroscopy. The 13C NMR spectrum (Figure 4.10A) of the final product shows the triazine ring resonance (— 174 ppm) as a group of several peaks, indicating that it is a mixture of several products with different triazine rings. Both El and CI MS experiments also confirm the NMR result and detect different triazine products with parent peaks at M+ = 477, M+ = 493, M+ = 509, and M+ = 525  173  a.m.u. respectively. From the MS experiments the approximate percentages of each of the components are as follows: M+ = 477  13%  M+ = 493  36%  M= 509  40%  M+ = 525  11%  In order to obtain the 15N NMR spectrum for the final products, 12% of 15N enriched p-tert-butylphenyl cyanate (PTBPCN) was reacted with 2allylphenyl cyanate 76. The 15N NMR spectrum (Figure 4.10B) also shows four peaks in the range of triazine ring nitrogen at — 87 ppm. Since only nitrogens derived from the enriched PTBPCN can give signals in the 15N NMR spectrum, the 15N spectrum should be cleaner and more easily interpretable than the 13C spectrum. The 13C spectrum gives all of the signals for all different carbons on different triazine rings. Some of them are degenerate and not distinguishable as seen in Figure 4.10A. A total of six carbon signals  for the four different triazines  should be obtained, but only four resolved signals are observed. Therefore, quantitative data can not be obtained from the 13C spectrum. As expected from Equation 4.8, there are only four different kinds of enriched nitrogens on three different triazines (31, 78, 79,) which should give 15N signals. This is exactly what is seen in the 15N spectrum in Figure 4.10B. All four signals are well resolved and distinguishable. Thus, quantitative data can be obtained from the 15N spectrum. From the 15N resonance in the 15N spectrum of triazine 31 obtained in Chapter 2, the peak at lowest field in Figure 4.10B corresponds to this compound. The peak at highest field in Figure 4.10B is thought to correspond to the enriched nitrogen on the triazine ring in 78. The two close peaks of equal intensity in the middle represent the two different enriched nitrogen positions on triazine 79. From 15N NMR spectrum, the molar 174  Figure 4.10. NMR spectrum in acetone-d6 of the product mixture obtained from curing a mixture of cross-linking monomer 76 and 12% 15 N enriched PTBPCN 25. (A).  13 C spectrum ( 1 H at 200 MHz); (B). 15 N spectrum 0-H at 300 MHz).  175  CH2CH=CHe 0 NO  YOY NyN`d^78 0 arCH2CH=CH2  a  89.0^88.0^87.0^86.0^ppm A  175.0^174.5  I 160^120^80^40 ' Ppm 175-1  ratios of triazines 78: 79 : 31 are estimated to be approximately 1 : 3 : 3. This is consistent with the previous results obtained from the MS experiments. OCN CH2CH=CH2  +  76  OCN  25  [4.8]  N* 0 0 N* 0  YOY *N^N* Y 0  YO'r N^N* -r0  31  CH2CH=C H2  79  CH2CH=CH2  CH2CH=CH2  0 N 0  0 N 0  Y1(5r N^N* -.'1"-. 0  YOY NN  ,r  CH2CH=C H2  0 CH2CH=CH2  C H2=CHCH2  77  78  There are three different units on each triazine ring. Each unit is from one cyanate group. If 2-allylphenyl cyanate 76 is designated to be A and PTBPCN 25 is designated to be B, the combinations for formation of a triazine ring should have eight different ways:  176  AAB^ABB AAA^ABA^BAB^BBB BAA BBA The combinations of AAA and BBB form triazines 77 and 31 respectively. All combinations in the second column form triazine 78 and all combinations in the third column form triazine 79. If each of the combinations have equal probability of formation, the ratios of the four different triazines 77 : 78 : 79 : 33 will be 1 : 3 : 3 : 1. This fits perfectly the results from both the NMR and the MS experiments. Thus, it has been shown that the copolymerization between the crosslinking monomer 76 and the cyanate 25 is an ideally random case. There is no preference at all between them for triazine ring formation. This is an ideal property for a cross-linking agent.  4.2.4. Copolymerization of the Cross-Linking Monomer 76 with an Olefinic Monomer  A preliminary investigation of the copolymerization of the cross-linking monomer 76 with an olefinic monomer was carried out. Firstly, the possible polymerization for the cross-linking monomer 76 alone under free radical initiation was tested to see whether it can be self polymerized through its ally' double bond without affecting the cyanate group. 1 g of cross-linking monomer 76 was mixed with 1 % benzoyl peroxide (BPO) and heated to 95 °C for one day.  The 13C INTMR spectrum (not shown) shows that all of the monomer 76 has trimerized to form 77 without reaction of the allyl double bond. Subsequently, a much lower temperature (45 °C) was used to polymerize the monomer 76 with 1 % azobisisobutyronitrile (AIBN) as initiator. This polymerization process was continued for ten days. The 13C NMR spectrum (not shown) still shows that all 177  of the monomer 76 has been converted to 77, but there is no any indication of reaction of the allyl double bond. These results suggest that the cross-linking monomer 76 can not be self polymerized through the allyl double bond without affecting the cyanate functional group (similar to the mixed cyanate/ epoxy system). The polymerization of triazine 77 alone was also tested at 120 °C for 5 days using 1 % BP0 as initiator. There is still no allyl double bond opening. However, it can be copolymerized with other olefinic monomer. This was verified by copolymerization of 0.2 g triazine 77 with 0.8 g methyl methacrylate (MMA) at 95 °C for one day using 1 % BP0 as initiator. The final polymer product was extracted by boiling benzene for 1 day using a soxhlet extractor. More than 60 % of the polymer product was left after the extraction. When poly(methyl methacrylate) (PMMA) is formed under the exactly same condition as above, nothing was left after extracting for only 16 hours. The difference between these indicates the final polymer product in the first case is cross-linked by the crosslinking agent 77 and therefore is not soluble in benzene. Further investigations of this reaction should be carried out in future work.  4.2.5. Conclusions As a complementary approach to the SIPN multicomponent polymer system, a bifunctional cross-linking agent for the cyanate resin (thermoset) and polyolefine (thermoplastic) mixed system, 2-allylphenyl cyanate 76, has been synthesized and characterized. Like the other cyanates as previously described, 2-allylphenyl cyanate 76 easily forms the cross-linking triazine compound 77 upon heating. 77 is a crystalline solid with m.p. = 110 — 111 °C. As a crosslinking agent, 2-allylphenyl cyanate 76 not only reacts with itself, but also reacts with another cyanate to form heterogeneous triazine rings, such as 178  triazines 78 and 79. Even though it can not polymerize with its own monomer through the allyl double bond, it can copolymerize with an other olefinic monomer, such as methyl methacrylate, to form a cross-linked and insoluble polymer.  179  CHAPTER 5. EXPERIMENTAL 5.1. High Resolution NMR Experiments In this section, the general conditions employed for the NMR experiments are described. The detailed conditions of specific experiments and the procedures used to prepare the NMR samples have been given in earlier sections.  5.1.1. Solution NMR Experiments Conventional 1H, 13C, and 15N solution NMR spectra were obtained using Bruker ACE 200 and Varian XL-300 spectrometers. The variable temperature experiments were carried out on the Varian XL-300 spectrometer. The NOE difference NMR spectra and 1H COSY 2D NMR spectra were obtained on Bruker WH-400 and ANIX-500 spectrometers. 1H-13C heteronuclear chemical shift correlation (HETCOR) 2D NMR spectra were obtained using a Bruker AMX-500 spectrometer with an inverse detection pulse sequence. All solution NMR spectra were obtained using 5 mm tubes. Deuterated solvents used were from Cambridge Isotope Laboratories. 13C and 1H chemical shifts are given with respect to TMS and the 15N chemical shifts are given with respect to neat formamide.  5.1.2. Solid State NMR Experiments 13C and 15N CP MAS solid-state NMR spectra were obtained using Bruker CXP-100 and MSL-400 spectrometers with commercial double resonance probes and with the magic angle set using the 79Br resonance of KBrE1041. All solid state NMR spectra were obtained using 7 mm od sample rotors. The 13C and 15N chemical shifts of the solid state NMR spectra are 180  given with respect to TMS and neat formamide, using adamantane and 15NH4C1 as the intermediate external references, respectively.  5.2. Mass Spectrometry Experiments Electron impact mass spectra (EIMS) and desorption chemical ionization mass spectra (DCIMS) were obtained using Kratos MS 50 (70 eV) and DelsiNermag R10-10B mass spectrometers. The ionizing gas in the latter experiments was NH3.  5.3. X-ray Diffraction Experiments All measurements were made on a Rigaku AFC6S cliffractometer with graphite monochromated CuKoc radiation.  5.3.1. Triazine 3 1  31  Crystallographic data appear in Table I in Appendix A. Final atomic coordinates, bond lengths and bond angles are given in Tables II—IV in Appendix A, respectively.  181  5.3.2. Bisphenol-A Dicyanate 2  OCN  NCO  2 Crystallographic data appear in Table V in Appendix B. Final atomic coordinates, bond lengths and bond angles are given in Tables VI—VIII in Appendix B, respectively.  5.3.3. 5-Phenoxymethy1-2-oxazolidinone 53  0-CH2  NH  \ 53^0  Crystallographic data appear in Table IX in Appendix C. Final atomic coordinates, bond lengths and bond angles are given in Tables X—XII in Appendix C, respectively.  5.4. Syntheses All chemical reagents used in the syntheses were supplied by Aldrich Chemical Co.and BDH unless otherwise indicated.  5.4.1. Labelled Cyanogen Bromide BrC*N (Br13CN or BrC15N)[91] 0.50 ml Br2 (0.01 mol) and 0.5 ml H20 were added to a 25 ml r.b. flask fitted with a magnetic stirring bar, sitting in a salted ice-water bath in a fume hood. To the stirred mixture, a solution of 0.65 g (0.01 mol) labelled KC*N (K13CN with 99% 13C or KC15N with 99% 15N) dissolved in 2 ml H20 was added dropwise from a pipet over a 20 minute period [Note: The rate of adding  182  KCN and the speed of stirring should be controlled properly to avoid KC*N being in excess in any local portion of the solution. Otherwise, the reaction mixture will turn to black, which is probably caused by the formation of (CN)x]. The correct color change should be from dark-red to fresh-red to orange to yellow to colorless. The exact amount of KC*N solution, which is needed to titrate the system just to a colorless or light yellow end point, was added. The purification of the formed cyanogen bromide can be performed by distillation immediately after reaching the end point.. The distillation was conducted at room temperature under vacuum with a dry ice acetone trap as product collector. The final BrC*N product is a colorless crystalline solid with m.p. = 48 — 50 °C (reported m.p. = 49 —51 °C for BrCN[91]), yield = 70 — 75 %.  5.4.2. Labelled Bisphenol-A Dicyanate (BPADCN 2a or 2b)[6a,92]  N1 3 CO  01 3 C N^1 5 NCO  2a  0C1 5N  2b  1.10 g (0.0105 ml) of BrC*N (Brl 3CN or BrC1 5N) and 1.14 g (0.005 mol) of Bisphenol A previously dissolved together in 10 ml of acetone were added to a 50 ml r.b. flask fitted with an equalizing pressure dropping funnel and a magnetic stirring bar, sitting in a salted ice-water bath in a fume hood. The mixture was stirred rapidly with cooling in the salted ice-water bath while 1.01 g (0.01 mol) of (CH3CH2)3N was added dropwise over a 20 minute period through the dropping funnel. A white solid, (CH3CH2)3NHBr, appeared after adding (CH3CH2)3N. Stirring was continued for 30 minutes while the mixture warmed to room temperature. The product was isolated by slowly pouring the mixture into 50 ml of ice-cooled water with vigorous stirring. In this step, the  183  (CH3CH2)3NHBr solid dissolved and the crude cyanate product precipitated. The precipitate was then isolated by filtration and washed with water until a neutral eluate was obtained. After vacuum drying, the crude cyanate product was recrystallized from cyclohexane, m.p. = 80 — 81 °C (80 °C in lit .[61), yield = 80— 85%. The 13C and 15N NMR spectra are shown in Figure 2.1.  5.4.3. Labelled p-tert-Butylphenyl Cyanate (PTBPCN 25a or 25b)[93}  0C15N  013 ON  25a  25b  To a 50 ml r.b. flask equipped with a magnetic stirrer and a pressure equalized dropping funnel, sitting in a salted ice-water bath in a fume hood, a solution of 1.10 g (0.0105 mol) of BrC*N (Br13CN or BrC15N) and 1.5 g (0.01 mol) of p-tert-butylphenol in 10 ml acetone was added. The mixture was stirred rapidly and 1.01 g ( 1.39 ml, 0.01 mol) of (CH3CH2)3N was added dropwise over a 20 minute period through a dropping funnel. (CH3CH2)3NHBr appeared as a white solid during the addition of (CH3CH2)3N. After an additional 15 minutes of stirring, the mixture was warmed up to room temperature. The white precipitate of (CH3CH2)3NHBr was removed by filtration and the solvent then removed by evaporation using a rotary evaporator under reduced pressure at room temperature. The final mixture was distilled under a vacuum of 0.5 mmHg at 75 °C. The final product is a colorless liquid, yield 76 — 87%. The 13C and 15N NMR spectra are shown in Figures 2.2 and 2.3.  184  5.4.4. Triazine 31 Formed from p-tert-Butylphenyl Cyanate  31  The triazine was prepared by heating a sample of p-tert-butylphenyl cyanate in a sealed tube for two hours at 150 °C. When the sample turned solid, it was recrystallized from acetone, yielding needle shaped colorless crystals, m.p. = 193.5 °C (Cal. for C33H39N303: C, 75.4; H, 7.5; N, 8.0 %. Found: C, 75.4; H, 7.5; N, 7.9 %). The 1H and 13C NMR spectra in deuterated chloroform and acetone (Figure 2.10) were in complete agreement with the postulated structure, as is the MS spectrum showing the parent ion M+ = 525 a.m.u. A small quantity of 15N labelled material was also synthesized and recrystallized from acetone. The 15N spectrum (Figure 2.10) was again in agreement with the proposed structure and showed no coupling to protons as expected.  5.4.5. 2,6-Dimethy1-4-phenoxycarbonylmorpholine 46  ^/  ^O  ^<  CH3  0-C-N^0  `-^<cH3 46  185  A solution of phenyl chloroformate (0.02 mole, 3.13 g) in 10 ml of dried toluene was added dropwise with stirring, through a dropping funnel equipped with a drying tube, into a mixture of 2,6-dimethylmorpholine (0.02 mole, 2.30 g) and pyridine (0.02 mole, 1.58 g), which had been pre-cooled in an ice bath. The solid formed during the reaction was removed by filtration. The toluene phase was decolorized by decolorizing neutral carbon, (Norit, Fisher Scientific Company) and was dried over MgSO4. The toluene solvent was then removed by evaporation. The crude liquid product (3.84 g, yield 81 %) solidified after sitting at room temperature overnight. The pure product was obtained by recrystallization from acetone as colorless crystals (mp. 68 - 73 °C). It is a mixture of cis and trans isomers (Cal. for C131-117NO3: C, 66.4; H, 7.3; N, 6.0 %. Found: C, 66.5; H, 7.3; N, 6.0 %). Two isomers were separated by chromatography on a silica gel (230-400 mesh, BDH No9385-48) column (2.5 cm x 17 cm) using eluants with gradually increasing polarity from 100% pentane to 100% diethyl ether. The cis isomer (mp. 77.5 — 78.2 °C) and the trans isomer (mp. 87.5 — 88.0 °C) show different NMR spectra (Figure 3.7).  5.4.6.  5-Phenoxymethy1-2-oxazolidinone 53  ^0—CH2  0-4 53^o Phenyl carbamate (0.01 mole, 1.37 g) and phenyl glycidyl ether (0.02 mole, 3.00 g) were mixed in a r.b. flask equipped with a condenser. The mixture was heated with stirring at 180 °C for 3.5 hours. Some white solid appeared after all the phenyl carbamate had dissolved at the beginning of the reaction and disappeared at the end of the reaction. (This white solid can be separated  186  from the reaction mixture at this stage by filtration. It was found to be cyanuric acid. Further details are given in Chapter 4.) The final brown sticky mixture was dissolved in diethyl ether leaving some residual solids. The solid was filtered and washed with Et20, and then recrystallized from acetone. It is a colorless crystalline material, 12% yield, mp. 120 —121 °C (Cal. for Ci °Hi iNO3: C, 62.2; H, 5.7; N, 7.3 %. Found: C, 62.4; H, 5.9; N, 7.2 %).  5.4.7. Bisphenol-A Monoglycidyl Ether 62  HO  0 C H2 C ---/C H2 62  Bisphenol A (0.02 mole, 4.56 g), anhydrous potassium carbonate (0.02 mole, 2.76 g) and 10 ml of methyl ethyl ketone (MEK) were mixed in a r.b. flask equipped with a condenser and a magnetic stirrer. The mixture was heated and 3 ml of H20 was added to dissolve the solid producing a clear solution. Then, 1 chloro-2,3-epoxypropane ( 0.02 mole, 1.85 g) was then added dropwise to the mixture with stirring through a dropping funnel. The reaction mixture was heated to reflux for 150 minutes with continuous stirring. Stirring was continued for another 45 minutes while the mixture cooled down to room temperature. The white precipitate formed in the reaction was removed by filtration. The clear product mixture was extracted by a mixture of 30 ml of dichloromethane and 20 ml of H20. After extraction and separation, the aqueous phase was washed twice with 10 ml of dichloromethane. The combined organic phases were washed twice with 15 ml of 1 M sodium hydroxide solution followed by four 25 ml portions of water until neutrality was obtained. The organic phase was dried overnight over anhydrous magnesium sulfate. After  187  filtration and evaporation, a light yellow viscous liquid product (3.50 g, 55 % crude yield) was obtained. Purification of the crude product was carried out by extraction and chromatography. The crude product mixture was extracted with the following solvents (the weight and yield of each extraction are given in parentheses respectively): (i) 25 ml n-pentane (36.7 mg, 2.0 %); (ii) 20 ml n-pentane (10.6 mg, 0.6 %); (iii) 30 ml 50 % n-pentane/ 50 % diethyl ether (1.01g, 60.3 %); (iv) 20 ml 50 % n-pentane/ 50 % diethyl ether (370 mg, 20.3 %); (v) 25 ml diethyl ether (307.7 mg, 16.7 %). Further purification was done by adsorption chromatography on silica gel (230 — 400 mesh, MERCK 9385) column (2.5 x 17.8 cm) using 50 % n-pentane/ 50 % diethyl ether as an eluant. 96.3 mg of pure crystalline product were obtained from a loading of 216.6 mg of sample onto the column, which was from the extraction fraction by 30 ml 50 % n-pentane/ 50 % diethyl ether above. This pure crystalline product was obtained in 44.5 % yield of the loading sample and in 18.9 % overall yield, m.p. = 86 — 88 °C (Cal. for C181-12003: C, 76.0; H, 7.1; 0, 16.9 %. Found: C, 75.9; H, 7.0 %). The structure of this product was completely characterized as 62 by NMR (see Figure 4.2) and MS experiments (parent ion M+ = 284 a.m.u.).  5.4.8.  The Monoglycid_v1 Ether of Bisphenol-A-Monocyanate 61  NCO  OC H 2 C H —ICH 2  o 61 A 50 ml r.b. flask fitted with an equalizing pressure dropping funnel and magnetic stirring bar, sitting in a salted ice-water bath in a fume hood, was charged with a solution of 0.60 g (0.0057 mole) of BrCN and 1.42 g (0.0050 mole)  188  of bisphenol A monoglycidyl ether in 10 ml acetone. The mixture was stirred rapidly with cooling in the bath while 0.70 ml (0.0050 mole) of (CH3CH2)3N was added dropwise over a 20 minute period through the dropping funnel. (CH3CH2)3NHBr appeared as a white solid after adding (CH3CH2)3N. Stirring was continued for 15 minutes while the mixture warmed to room temperature. The solid was removed by filtration and then by extraction with mixture of 20 ml of benzene and 20 ml of H20. The organic phase was collected, washed with two portions of 20 ml of H20 and dried over anhydrous MgSO4 for 2 hours. After filtration and evaporation of the solvent under vacuum, a clear light yellow viscous liquid product (1.36g, 88 % yield) was obtained. It was characterized by NMR (Figure 4.3) and MS experiment (parent ion M+=309 a.m.u.) as the cyanate 61 with minor impurities (Cal. for C191119NO3: C, 73.7; H, 6.2; N, 4.5 %. Found: C, 73.2; H, 6.3; N, 4.2 %).  5.4.9. Triazine 68 Formed from the Monoglycidyl Ether of Bisphenol-A-  Monocyanate 61  68 OCH2C1-\1-/ C I-12  o  189  2.0 g of the monoglycidyl ether of bisphenol-A-monocyanate 61 were charged into a r.b. flask and heated at 180 °C in vacuum (1.5 mmHg) for 15 minutes. A bright brown color cured solid was obtained. It was characterized by NMR (Figure 4.4) and MS (parent ion M+.927 a.m.u.) as the triazine product 68 with minor impurities (Cal. for C57H57N309: C, 73.7; H, 6.2; N, 4.5 %. Found: C, 72.8; H, 6.2; N, 4.2 %).  5.4.10. A More Practical Way to Synthesize the Cross-Linking Monomer 61 A mixture of bisphenol-A (0.05 mole) and potassium carbonate (0.05 mole) in 20 ml MEK and 6 ml H20 was heated to boiling for 10 minutes and all of the solid was dissolved. Then, 0.05 mole of epichlorohydrin was added to the mixture through a dropping funnel at a rate of 0.5 ml per minutes. After that, the mixture was kept refluxing and stirring for 90 minutes (at about 80 °C), and then cooled down to the room temperature with continuous stirring. The crude product mixture was extracted twice with 20 ml H20 and the dried over Mg504 overnight. After the drying agent was removed by filtration, the solvents were removed through a rotavapor at 80 °C under reduced pressure for 10 minutes. The '3C NMR spectrum of the product mixture shows that the monoglycidyl ether of bisphenol-A is about 32%, the bisphenol-A is about 36%, and the diglycidyl ether species are approximately 32%. From these estimated data, approximate 10% of excess amounts of cyanogen bromide and triethylamine were used to convert phenol moieties to cyanate groups. 3.963 g of the product mixture obtained in the previous step and 2.242 g of cyanogen bromide were dissolved together in 20 ml acetone and sat in an icewater-bath with stirring. 1.800 g of triethylamine was slowly added in through a dropping funnel in a 30 minute period. The white precipitate produced in the reaction was removed by filtration and then by extraction with 30 ml H20. The 190  crude products were dissolved in 30 ml acetone and dried over MgSO4 overnight. After filtration, the solvents were removed by a rotavapor at 70 °C for less than 5 minutes under reduced pressure. The final product mixture was obtained as a viscous liquid. The NMR spectrum shows that it contains three major products: bisphenol-A dicyanate 2, the diglycidyl ether of bisphenol-A 17, and the monoglycidyl ether of bisphenol-A-monocyanate 61, which still can be used as the cross-linking agent for the cyanate and the epoxy mixed resins.  5.4.11. 2-Allylphenyl Cyanate 76 OCN CH2CH=CH2  76 To a 50 ml r.b. flask equipped with a magnetic stirrer and a pressure equalized dropping funnel, sitting in a salted ice-water bath in a fume hood, a solution of 3.15 g (0.0300 mole) of BrCN and 3.5 g (0.0261 mole) of 2-allylphenol in 15 ml acetone was added. The mixture was stirred rapidly with cooling in the bath while 2.64 g (3.64 ml, 0.0261 mole) of (CH3CH2)3N was added through a dropping funnel over a 20 minute period. (CH3CH2)3NHBr appeared as white solid while adding (CH3CH2)3N. After an additional 30 minutes of stirring, the mixture was warmed up to room temperature. The (CH3CH2)3NHBr was removed by filtration and the solvent was then removed using a rotary evaporator under reduced pressure at room temperature. The final mixture was purified by distillation under a vacuum of 1.5 mmHg at 70 °C. The final product is a colorless liquid, yield 85 %. The NMR spectra are shown in Figure 4.8. The MS experiment gives the parent ion at M+.159 a.m.u.. (Cal. for Ci 0119NO: C, 75.5; H, 5.7; N, 8.8 %. Found: C, 75.6; H, 5.8; N, 8.7 %). 191  5.4.12. Triazine 77 Formed from 2-Allylphenyl Cyanate CH2CH=CH2  0 CH2CH=CH2  CH2=CHCH2  77  The liquid sample of 2-allylphenyl cyanate 76 changed to a white crystalline solid after sitting at room temperature for two months. (This trimerization process can be shortened to 2 hours if the sample is heated to 150 °C.) It was recrystallized from acetone yielding long needle shaped crystals, m.p. = 110 —111 °C (Cal. for C301127N303: C, 75.5; H, 5.7; N, 8.8 %. Found: C, 75.6; H, 5.8; N, 8.7 %). The 1H and 13C NMR spectra in deuterated chloroform and acetone are in complete agreement with the postulated structure (Figure 4.9). MS spectrum shows the parent ion at M+ = 477 a.m.u..  5.5. Chromatographic Separation of the Reaction Products Thin-layer chromatographic (TLC) separations were done on commercial aluminum-backed silica gel plates (E. Merck, Type 5554). Preparative Thinlayer chromatography was done on 20 cm x 20 cm plates coated with 2 mm of silica gel (E. Merck, Silica Gel 60). Visualization was accomplished with ultraviolet light. Adsorption chromatographic separation was done using a silica gel (230-400 mesh, BDH No9385-48) column (2.5 cm x 17 cm) and using eluants with gradually increasing polarity from 100% pentane to 100% diethyl ether. Preparative column chromatography was done on a silica gel (230 — 400 mesh,  192  BDH No9385-48) column (7 cm x 17 cm). Gel permeation chromatography was carried out using a Lipophilic Sephadex LH-20 (Sigma) column (2.0 cm x 55 cm) and acetone as eluant.  193  CHAPTER 6. PROPOSALS FOR FUTURE WORK The cyanate resin and its related heterogeneous polymer systems are newly developed materials and have many potentially useful features. From the results in the thesis, it is thought some work should be continued in the future.  (1)  Application of the Cross-Linking Monomer 61 As described above, one of the major applications for cyanate resin is for  making electronic circuit boards mixed with epoxy resin. The monomer 61 can be used as a cross-linking agent for this mixed resin system. Particularly importantly as shown in Chapter 4, the cross-linking agent 61 can be obtained in a very straightforward manner which is compatible with the normal mixed resin composition. The effect on the properties of the end product by using the cross-linking monomer 61 should be further studied on a large scale and in a board fabrication. The physical, mechanical, thermal and electrical properties of the mixed resin obtained from the practical approach as described in Section 4.1.6 should be tested under real application conditions. This should be done in collaboration with some industrial agency.  (2)  Application of the Cross-Linking Monomer 76 As discussed in Section 4.2.5, the triazine 77 obtained from 76 can be  copolymerized with methyl methacrylate monomers to form a cross-linked and insoluble copolymer. Further investigations should be carried out using other olefinic monomers and also using the monomer 76 directly. In the latter case, even though the cyanate group is quite active, it may still be possible to let the 194  allyl double bond in 76 react with olefinic monomers first without affecting the cyanate group, because the presence large number of olefinic monomers might sufficiently dilute the cyanate monomers to make the trimerization of the cyanate groups slower than the double bond opening of the allyl groups. If this prediction is correct, the copolymer formed in this way will be a linear molecule with some free cyanate groups attached. It will still be soluble since no cross-link or very few cross-links will have formed. This kind of copolymer is ready to form an insoluble cross-linked polymer network through the free cyanate groups on the polymer chain just by application of heat. Its curing behavior should be similar to a thermoset resin, but the properties of the end polymer product should be different. Again, the physical, mechanical and thermal properties of the polymer obtained by using this cross-linking agent should be tested under real application conditions.  (3) A Possible Coupling Agent for the Glass-Reinforced Cyanate Resin  Composites During these thesis studies, some work was done on the characterization of a silane coupling agent, triethoxyvinylsilane (TEVS). They show that TEVS can be firmly anchored on a silica gel surface by reaction of the ethoxy group in the TEVS with the silanol group on the silica gel surface. The Si-0---Si bonds formed show strong stability against long time extraction and acidic hydrolysis. NMR spectra reveal that the TEVS on the silica gel can be copolymerized with an unsaturated monomer, such as styrene and methyl methacrylate. The chemical bonding between the anchored coupling agent TEVS and the polymer matrix is directly observed by solid state NMR spectroscopy. These illustrate  195  that silane coupling agents can be useful in the formation of glass-reinforced polymer composites. However, there is currently no coupling agent for cyanate groups. From the reaction of cyanate with diethylamine described in Section 4.1.5, 4aminobutyltriethoxysilane is thought to be a very promising coupling agent for the glass-reinforced cyanate resin composites, such as electronic circuit boards. It is predicted that it should bond the cyanate resin to the glass surface by the following reactions according to Equation 4.4:  1  n ^ OH + (CH3CH20)3Si—(CH2)4 NH2  [6.1]  [1 0] nSiI ^ (CH2)4 NH2 +  n  CH3CH2OH  (OCH2CH3)3.n  [f^ otsi n  (cH2)4 NH2 +  OCN —  (OCH2CH3)3_n  [f ^  [6.2]  0 N--( oi-n!i—(CH2)4 —NH —( N (OCH2CH3 )3.n  If all of these can be verified, it will be a very useful coupling agent, not only for the cyanate resin, but also for the mixed cyanate/epoxy resin system, because it is currently used as a coupling agent for the glass-reinforced epoxy resin. 196  REFERENCES [1]  Carothers, W. H.; J. Am. Chem. Soc., 51, 2584 (1929).  [2]  Flory, P. J.; "Principles of Polymer Chemistry", Chap. 2, Cornell University Press, Ithaca, N.Y., 1953.  [3]  Miller, M. L.; "The Structure of Polymers", Chapters 3 and 7, Van Nostrand Reinhold, New York, 1966.  [4]  Odian, G. G.; "Principles of Polymerization", Chapter 1, John Wiley and Sons, New York, 1981.  [5]  Storch, R.; and Gerber, H.; Angew. Chem., 72, 1000 (1960).  [6]  (a) Grigat, E.; and Piitter, R.; Chem. Ber., 97, 3012-17 (1964); (b) Grigat, E.; and Piitter, R.; Chem. Ber., 97, 3018-21 (1964); (c) Grigat, E.; and Piitter, R.; Chem. Ber., 97, 3027 (1964).  [7]  (a) Grigat, E.; Schminke, H. D.; and Putter, R.; German patent 1190184 (1965). (b) Grigat, E. et al, "Esters of Cyanic Acid", US 3,553,244, 1/5/71.  [8]  Grigat, E.; et al.; Kunstoffe, 58(12), 827-32 (1968).  [9]  Jensen, K.A.; and Holm, A.; in "Chemistry of Cyanates and Their Derivatives", edited by Patai, S.; J. Wiley, New York, Vol.1, p.569—  618(1977). [10]  Shimp, D. A.; Intl. SAMPE Symp. Exhib., 32, 1063-1072 (1987).  [11]  Bauer, M.; Bauer, J.; and Kuhn, G.; Acta Polym., 37, 715-719 (1986).  [12]  Sergecv, V. A.; Shitikov, V. K.; and Pankratov, V. A.; Russ. Chem. Rev., 48(1), 79-93 (1979).  [13]  Jensen, K. A.; and Holm, A.; Acta Chem. Scand., 18, 826 (1964).  [14]  Martin, D.; Angew. Chem. Int. Ed., 3, 311 (1964).  [15] Shimp, D. A.; PMSE Preprints, ACS Meeting, New York, April, 1986, p.107-113. 197  [16]  Feldman, J. A.; and Huang, S. J.; "Cross-linked Polymers", Chapter 18, 244 (1988).  [17]  Shimp, D. A.; Christenson, J. R.; and Ising, S. J.; Proc. 34th International SAMPE symposium, 34, 222 (1989).  [18]  Fang, T.; Macromolecules, 23, 4553-56 (1990).  [19]  Weirauch, K. K.; IPC Technical Paper, TP-066 (1975).  [20]  Gaku, M.; IPC Technical Paper, TP-283 (1979); BT Resin—Glass Fabrics Multilayer Materials, Mitsubishi Gas Chemical Company Inc. (1984).  [21]  Hudock, F. A.; and Ising, S. J.; IPC Technical Paper, TP-618 (1986).  [22]  Shimp, D. A.; "Thermal Performance of Cyanate Functional Thermosetting Resins", SAMPE Quarterly, 19(1), p.41-46 (1987).  [23]  Ising, S. J.; and Christenson, J.; "An Update of Cyanate Ester Technology", IPC Fall Meeting, Anaheim, CA, Oct.23-28, 1988.  [24]  Shimp, D. A.; et al, "New Liquid Dicyanate Monomer for Rapid Impregnation of Reinforcing Fibers", 34th Int. SAMPE Symposium and Exhibition, Reno, Nevada, May 8-11, 1989, p.1336-46.  [25]  Shimp, D. A.; and Ising, S. J.; 35th Int. SAMPE Symp., 1045-56 (1990).  [26]  Shimp, D. A.; and Vanderlip, J. T.; US Pat. 5,149,863 (1992).  [27]  Bogan, G. W.; Lyssy, M. E.; Monnerat, G. A.; and Woo, E. P.; SAMPE Journal, 24(6), Nov./Dec., 19-25 (1988).  [28]  Shimp, D. A.; Hidock, F. A.; and Bobo, W. D.; "Toughening Cyanate Functional Resins for Structural Composite Matrix Applicationds" , 18th Int. SAMPE Tech. Conf., Seattle, October, 1986, p851-862.  [29]  Hartness, J. T.; "A Dicyanate Semi-Interpenetrating Polymer Network (SIPN) Matrix Composite", ASTM Spec. Tech. Pull., No. 937, 453-462  (1987).  198  [30]  Wang, D. A. 'Advanced Materials for Printed Circuit Boards' IBM Corporation,Systems Technology Division, Endicott, NY13706, (1990).  [31]  (a) Almen, G. R.; et al, "Thermoset I Thermoplastic Alloys: An Approach to Tough High Performance Resin Matrices", 20th Int. SAMPE Tech. Conf,  September, 1988, p.46-59; (b) Chu, S. J.; et al, US patent 4,656,208. [32]  Gotro, J. T.; Appelt, B. K. and Papathomas, K. I.; Polymer Composites, 8(1), 39 (1987).  [33]  Kubens, R.; et al, US patent 3,562,214 (1971).  [34]  (a) Prevorsek, V. A.; and Chung, D. C.; U.S. Pat. 4,456,360 (1979); (b) Wertz, D. H.; and Prevorsek, D. C.; Polym. Eng. and Sci., 25, 804 (1985).  [35]  Martin, D.; Angew. Chem., 76, 303 (1964).  [36]  Shimp, D. A.; Hudock, F. A.; and Ising, S. J.; 33rd Int. SAMPE Symp., March 7, p.754 — 766 (1988).  [37]  Castan, Erf. P.; Swiss 211,116 (1938); Brit. 518,057 (1938); U.S. 2,324,483 (1940).  [38]  Batzer, H.; Kunststoffe-Plastics, 14 (3/4), 77 (1967).  [39]  Lee, H.; and Neville,K.; "Handbook of Epoxy Resins", McGraw Hill Book Company, London, 1967.  [40]  May, C. A.; and Tanaka, Y.; "Epoxy Resins", Marcel Decker Inc., New York, 1973.  [41]  Batzer, H; Lohse, F.; and Schmid, R.; Angew. Makromol, Chem., 29, 330, 349 (1973).  [42]  Fischer, M.; Lohse, F.; and Schmid, R.; Makromol. Chem., 180, 1251 (1981).  [43]  (a) Morgan, R. J.; et al.; in Bauer, R. S.; "Epoxy Resin Chemistry IT', ACS Symposium Series 221, Amer. Chem. Soc., Washington, p.211 (1983); (b) Adv. Polym. Sci., 72, 1 (1985). 199  [44]  (a) Gillham et al., Adv. Chem. Ser., 208, 235 (1984); (b) Partabsjy, A, M,; "Epoxy Resins", Washington, 1970.  [45]  Bauer, R. S.; "Epoxy Resin Chemistry", ACS Symposium Series 221, Amer. Chem. Soc., Washington, 1983.  [46]  Rozenberg, B. A.; Advances in Polym. Sci., 75, 113 (1985).  [47]  Dusek, K; Advances in Polym. Sci., 76, 1 (1986).  [48]  Oleinik, E.F., Advances in Polym. Sci., 76, 49 (1986).  [49]  DeHoff, "Effect of Various Curing Agents on the Corrosion Resistance of Epoxy Resins", Plastics Symposium, 16th Annual NACE Conference,  Dallas, 1963. [50]  Glasson, et al., "Process for the Manufacture of Epoxies", U.S. 2,650,927, 1953.  [51]  Batzer, H.; "Cycloaliphatic Epoxies", Chem. Ind. (London), Feb.1, 1964.  [52]  Mitsubishi Petrochem Comp., JP 55 139 373 (1980); JP 58 150 581 (1983).  [53]  CIBA-GEIGY AG; DOS 2 726 821; DOS 2 726 846 (1976).  [54]  Plueddemann, E. P.; "Silane Coupling Agents", Plenum Press, New York (1982).  [55]  Plueddemann, E. P.; in "Interfaces in Polymer Matrix Composites", Plueddemann, E. P.; ed., Academic Press, New York, p.173 (1974).  [56]  Plueddemann, E. P.; J. Adhes., 2, 184 (1970).  [57]  Katz, H. S.; and Milewski, J. V.; "Handbook of Fillers and Reinforcements for Plastics", Van Nostrand Reinhold Company, New York (1978).  [58]  Ishida, H.; and Koenig, J. L.; Polym. Eng. Sci., 18, 128 (1978).  [59]  Plueddemann, E. P.; Clark, J. A.; Nelson, L. E.; and Hoffmann, K. R.; Mod. Plast., 39, 136 (1962).  [60] Wu, S.; "Polymer interface and adhesion", Marcel Dekker Inc., New York, 1982. 200  [61]  Schrader, M. E.; Lerner, I.; and D'Oria, F. J.; Mod. Plast., 45(1), 195 (1967).  [62]  Ishida, H.; and Koenig, J. L.; J. Polym. Sci., 17, 615 (1979).  [63]  Ishida, H.; and Koenig, J. L.; J. Colloid and Intel: Sc., 64(3), 565 (1978).  [64]  Ishida, H.; "Adhesion Aspects of Polymeric Coatings," Mittall, K.; ed., Plenum Press, New York, p.45 (1982).  [65]  Fyfe, C. A.; "Solid State NMR for Chemists"; C.F.C. Press, Guelph, 1984.  [66]  Schaefer, J.; Chin, S. H.; and Weissman, S. I.; Macromolecules, 5, 798 (1972); (b) Schaefer, J.; Stejskal, E. O.; and Buchdahl, R.; Macromolecules, 8, 291 (1975); (c) Schaefer, J.; and Stejskal, E. O.; J. Am. Chem. Soc., 98, 1031 (1976).  [67]  (a) Schaefer, J.; Stejskal, E. O.; and Buchdahl, R.; Macromolecules, 10, 384 (1977); (b) Schaefer, J.; and Stejskal, E. O.; "High-Resolution 13C NMR of Solid Polymer" in "Topics in Carbon-13 NMR Spectroscopy", Ed.  Levy, G. C.; Vol. 3, Wiley Interscience, New York, 1979. [68]  Fyfe, C. A.; Lyerla, J. R.; Volksen, W. and Yannoni, C. S.; Macromolecules, 12, 757 (1979).  [69]  Lyerla, J. R.; "High-Resolution Carbon-13 Studies of Bulk Polymers" in "Contemporary Topics in Polymer Science", Ed. Shen, M.; Vol.3, Plenum  Publishing Corp., New York, 1979. [70]  Maciel, G. E.; Sindorf, D. W.; and Bartuska, V. J.; J. Chromotog., 205, 438 (1981).  [71]  Hays, G. R.; Claque, A. D.; Huis, H. R.; and Van Der Velden, G.; Appl. Surf Sci., 10, 247 (1982).  [72]  Leyden, D. E.; Kendall, D. S.; and Waddell, T. G.; Anal. Chim. Acta, 126, 207 (1981).  201  [73]  Chiang, C. H.; Lin, N.; and Koenig, J. L.; J. Colloid Interf. Sci., 86, 26 (1982).  [74]  Zaper, A. M.; and Koenig, J. L.; Polym. Composites; 6(3), 156 (1985).  [75]  Allcock, H. R.; and Lampe, F. W.; "Contemporary Polymer Chemistry", Prentice Hall, Englewood Cliffs, New Jersey, 1990.  [76]  Komoroski, R. A.; and Mandelkern, L.; "Overview of High-resolution NAIR of Solid Polymers" in "High Resolution NMR Spectroscopy of Synthetic Polymers in Bulk"; Ed. by Komoroski, R. A.; VCH Publishers, Inc.  Deerfield Beach, Florida, 1986. [77]  Mehring, M.; "High Resolution NMR in Solids"; Springer-Verlag, Berlin, 1983.  [78]  Komoroski, R. A.; "Principles and General Aspects of High-Resolution NMR of Bulk Polymers" in "High Resolution NMR Spectroscopy of Synthetic Polymers in Bulk"; Ed. by Komoroski, R. A.; VCH Publishers,  Inc. Deerfield Beach, Florida, 1986. [79]  Pake, G. E.; J. Chem. Phys.; 16, 327 (1948).  [80]  Andrew, E. R.; Phil. Trans. R. Soc. London; A299, 505 (1981).  [81]  Jelinski, L. W.; "Chain Structure and Conformation of Macromolecules"; Ed. by Bovey, F. A.; Academic Press, New York, Chapter 8, 1982.  [82]  Beeler, A. J.; Orendt, A. M.; Grant, D. M.; Cutts, P. W.; Michl, J.; Zilm, K. W.; Downing, J. W.; Facelli, J. C.; Schindler, M. S.; and Kutzelnigg, W.; J. Am. Chem. Soc., 106, 7672 (1984).  [83]  Andrew, E. R.; Bradburry , A.; and Eades, R. G.; Nature, 182, 1659 (1959).  [84]  Lowe. I. J.; Phys. Rev.; 2, 285 (1959).  [85]  Lyerla, J. R.; Meth. Exp. Phys., 16A, 241 (1980)  [86]  Pines, A.; Gibby, M. G.; and Waugh, J. S.; J. Chem. Phys., 56, 1 776 (1972). 202  [87]  Hartmann, S. R.; Hahn, E. L.; Phys Rev., 128, 2042 (1962).  [88]  Pines, A.; Gibby, M. G.; Waugh, J. S.; J. Chem. Phys., 59, 569 (1973).  [89]  Fyfe, C. A.; Dudley, R. L.; Stephenson, P. J.; Deslandes, Y.; Hamer, G. K.; Marchessault, R. H.; J. Macromol. Sci. Rev. Macromol. Chem. Phys.; C23, 187 (1983).  [90]  Fyfe, C. A.; Rudin, A.; and Tchir, W.; Macromolecules, 13, 1320 (1980).  [91]  Hartman, W. W.; Dreger, E. E.; Org. Synth., 2, 30 (1931).  [92]  Cozzens, R.F., Walter, P.; J. Appl. Polym. Sci., 34, 601 (1987).  [93]  Martin, D.; Bauer, M.; Org. Synth., 61, 35 (1983).  [94]  Fyfe, C. A.; Niu, J.; Rettig, S. J.; Burlinson, N. E.; Reidsema, C. M.; Wang, D. W.; and Poliks, M. D.; Macromolecules, 25, 6289-6301 (1992).  [95]  Opella, S. J.; Frey, M. H.; J. Amer.Chem. Soc., 101, 5854 ( 1979).  [96]  (a) Zumbulyadis, N.; Henrichs, P. M.; Young, R. H.; J. Chem. Phys., 75, 1063 ( 1981). (b) Bohm, J.; Fenzyke, D.; Pfeiffer, H.; J. Mag. Res., 55, 197 ( 1 983). (c) Naito, A.; Ganapathy, S.; McDowell, C. A.; J. Chem. Phys., 74, 5393 (1 981 ). (d) Naito, A.; Ganapathy, S.; McDowell, C. A.; J. Mag. Res., 48, 367 (1982).  [97]  Dixon, W.T.; J. Mag. Res., 49, 341 (1982).  [98]  Frisch, K. C. et al.; "Topologically Interpenetrating Polymer Networks" in "Recent Advances in Polymer Blends, Grafts and Blocks" ed. by L. H.  Sperling, Plenum, Plenum Press, N.Y., 1978. [99] An, J. H.; and Sperling, L. H.; "Development of Multiphase Morphology in Sequential Interpenetrating Polymer Networks" in "Cross-linked Polymers", Ed. by Dickie, R. A.; Labana, S. S.; and Bauer, R. S.; American  Chemical Society, Washington, DC, 1988. 203  [100] Saunders, K. J.; "Organic Polymer Chemistry", Chapman and Hall, London, New York, 1988. [101] Mok, K; B.Sc. Thesis, Department of chemistry at the University of British Columbia, 1992. [102] Jagannathan, N. R.; and Herring, F. G.; J. Polym. Sci., Part A, 26, 1 (1988). [103] Matin, H; et al.; "Polyepoxy-substituted Aromatic Compounds, and Polymers", U.S. 2,938,875, (1960). [104] Frye, J. S.; Maciel, G.E.; J. Mag. Res., 48, 125 (1982).  204  APPEDICES A. Crystal Structure Data for Compound 81 Table I. Crystal Data Empirical Formula^  C 33 H 39 N 3 0 3  Formula Weight^  525.69  Crystal Color, Habit^  colorless, prism  Crystal Dimensions (mm)^0.250 X 0.300 X 0.450 Crystal System^  monoclinic  No. Reflections Used for Unit Cell Determination (29 range)^25 ( 59.3 - 96.0°) Omega Scan Peak Width at Half-height^ Lattice Parameters:  0.37  a= b . c . 8 -  12.214 6.556 37.981 90.644  (2)A (2)A (2)A (7)°  V . 3041 (1)A3 Space Group  P21 /c (#14)  Z value  4  D  1.148 g/cm 3  calc  F000  1128  P(CuKa)  5.50 cm -1  205  Table II.^Final Atomic Coordinates atom  x  Y  z  0(1)  0.1993(1)  0.3243(2)  0.25210(3)  0(2)  0.4550(1)  -0.1540(2)  0.27712(3)  0(3)  0.3737(1)  -0.0339(2)  0.16511(3)  N(1)  0.3240(1)  0.0899(2)  0.26633(4)  N(2)  0.4168(1)  -0.0878(2)  0.22098(4)  N(3)  0.2800(1)  0.1557(2)  0.20631(3)  C(1)  0.2704(1)  0.1849(3)  0.24057(4)  C(2)  0.3953(1)  -0.0437(3)  0.25435(4)  C(3)  0.3547(1)  0.0175(3)  0.19880(4)  C(4)  0.1360(1)  0.4320(3)  0.22687(4)  C(5)  0.1815(1)  0.5856(3)  0.20775(5)  C(6)  0.1168(2)  0.6921(3)  0.18434(5)  C(7)  0.0062(1)  0.6497(3)  0.17949(4)  C(8)  -0.0370(1)  0.4956(3)  0.20016(5)  C(9)  0.0270(2)  0.3867(3)  0.22388(5)  C(10)  0.4337(1)  -0.1405(3)  0.31349(5)  C(11)  0.3899(2)  -0.3072(3)  0.32910(5)  C(12)  0.3773(2)  -0.3066(3)  0.36537(6)  C(13)  0.4075(1)  -0.1428(3)  0.38586(5)  C(14)  0.4507(2)  0.0240(3)  0.36860(5)  C(15)  0.4648(2)  0.0258(3)  0.33247(5)  C(16)  0.3168(1)  0.0469(3)  0.13610(4)  0(17)  0.3032(2)  -0.0876(3)  0.10899(5)  C(18)  0.2542(2)  -0.0245(3)  0.07794(5)  206  (Table II continued) atom C(19)  0.2166(2)  0.1728(3)  0.07331(4)  C(20)  0.2325(2)  0.3041(3)  0.10126(5)  C(21)  0.2834(2)  0.2453(3)  0.13246(5)  C(22)  -0.0613(2)  0.7663(3)  0.15224(5)  C(23)  -0.1810(2)  0.6995(4)  0.15149(6)  C(24)  -0.0580(3)  0.9928(4)  0.16063(9)  C(25)  -0.0152(2)  0.7255(5)  0.11593(6)  C(26)  0.3925(2)  -0.1487(4)  0.42583(5)  C(27)  0.4561(5)  -0.325(1)  0.4418(1)  C(27A)  0.377(3)  -0.366(5)  0.4372(5)  C(28)  0.2714(4)  -0.180(1)  0.4341(1)  C(28A)  0.306(1)  0.003(4)  0.4336(3)  C(29)  0.4302(7)  0.045(1)  0.4442(1)  C(29A)  0.499(2)  -0.080(5)  0.4413(4)  C(30)  0.1577(2)  0.2382(3)  0.03936(5)  C(31)  0.1681(3)  0.4650(4)  0.03286(7)  C(32)  0.0371(2)  0.1866(6)  0.04286(8)  C(33)  0.2034(3)  0.1278(5)  0.00744(6)  207  80Z LOE6'0  ZESti'0-  EZEP*0  (EZ)H  8LZt7'0  ELTP*0-  OLOE*0  (YZZ)H  1/.917'0  LTEE*0-  SZWO  (ZZ)H  ZSTT*0  ZZLL'O  60900  (IZ)H  Z860'0  066C0  L8500-  (0Z)H  OTTT*0  88L5'0  T810'0-  (6T)H  Tb8T'0  1910'T  58800-  (8T)H  OEPT*0  9L901  tT0T*0-  (L.T)H  ZO9U0  8060T  6L10'0  (91)H  65T*0  LES5'0  EG8T*0-  (ST)H  SEET'0  SLLL'O  OTZZ'O  -  (ti)H  9PLT*0  EtZL'O  SETZ'O-  (EI)H  STST*0  8EtE*0  ES6Z*0  (zT)H  06600  IStt'0  890Z*0  (TT)H  9850'0  tZZT*0-  88tZ*0  (01)H  9TTT*0  06ZZ'0-  T8ZE*0  (6)H  80ZE'0  617PT*0  9966'0  (8)14  EZ8E*0  SttT'0  OZLI7'0  (L)H  L9LE'0  69Zt'0-  8SPE*0  (9)14  05TE*0  85P0-  5L90  (5)H  E8EZ*0  88L0  E5000-  (t7)H  6L6T*0  0E900  05TT*0-  (E)H  9OLT*0  0Z08'0  8010  (Z)H  LOTZ*0  G6T9'0  E650  (T)H  z^A^X^1.1101e (panupuoo II  am')  (Table II continued) atom  x  Y  z  H(23A)  0.4368  -0.4498  0.4282  H(24)  0.5345  -0.3055  0.4378  H(24A)  0.3763  -0.3731  0.4629  11(25)  0.2292  -0.0627  0.4253  14(25A)  0.2980  0.0171  0.4591  H(26)  0.2451  -0.3044  0.4225  H(26A)  0.3260  0.1357  0.4235  H(27)  0.2624  -0.1920  0.4596  H(27A)  0.2361  -0.0419  0.4232  H(28)  0.5086  0.0644  0.4402  H(28A)  0.4956  -0.0869  0.4670  H(29)  0.3895  0.1615  0.4346  H(29A)  0.5577  -0.1678  0.4329  H(30)  0.4168  0.0333  0.4695  H(30A)  0.5129  0.0614  0.4340  H(31)  0.2455  0.5003  0.0301  H(32)  0.1274  0.5016  0.0114  H(33)  0.1383  0.5400  0.0529  H(34)  0.0065  0.2624  0.0627  H(35)  -0.0019  0.2240  0.0211  11(36)  0.0290  0.0399  0.0470  11(37)  0.1970  -0.0199  0.0109  13(38)  0.1620  0.1680  -0.0137  13(39)  0.2807  0.1641  0.0046  209  Table III.  Bond Lengths (A)  atom  atom  distance  atom  atom  distance  0(1)  C(1)  1.338(2)  C(13)  C(14)  1.383(3)  0(1)  C(4)  1.414(2)  C(13)  C(26)  1.532(3)  0(2)  C(2)  1.337(2)  C(14)  C(15)  1.385(3)  0(2)  C(10)  1.411(2)  C(16)  C(17)  1.365(2)  0(3)  C(3)  1.346(2)  C(16)  C(21)  1.370(3)  0(3)  C(16)  1.400(2)  C(17)  C(18)  1.379(3)  N(1)  C(1)  1.326(2)  C(18)  C(19)  1.383(3)  N(1)  C(2)  1.320(2)  C(19)  C(20)  1.379(3)  N(2)  C(2)  1.329(2)  C(19)  C(30)  1.531(3)  N(2)  C(3)  1.321(2)  C(20)  C(21)  1.387(3)  N(3)  C(1)  1.322(2)  C(22)  C(23)  1.525(3)  N(3)  C(3)  1.319(2)  C(22)  C(24)  1.519(3)  C(4)  C(5)  1.364(3)  C(22)  C(25)  1.519(3)  C(4)  C(9)  1.367(2)  C(26)  C(27)  1.517(6)  C(5)  C(6)  1.374(3)  C(26)  C(27A)  1.50(3)  C(6)  C(7)  1.389(2)  C(26)  C(28)  1.529(5)  C(7)  C(8)  1.387(2)  C(26)  C(28A)  1.49(2)  C(7)  C(22)  1.523(2)  C(26)  C(29)  1.516(6)  C(8)  C(9)  1.385(2)  C(26)  C(29A)  1.49(2)  C(10)  C(11)  1.357(3)  C(30)  C(31)  1.513(4)  C(10)  C(15)  1.359(3)  C(30)  C(32)  1.519(4)  C(11)  C(12)  1.388(3)  C(30)  C(33)  1.523(3)  C(12)  C(13)  1.374(3)  210  Table IV.^Bond Angles atom  atom  atom^angle^atom  atom  atom  angle  C(1)  0(1)  C(4)  118.1(1)  0(2)  C(10)  C(15)  121.1(2)  C(2)  0(2)  C(10)  119.6(1)  C(11)  C(10)  C(15)  121.6(2)  C(3)  0(3)  C(16)  124.3(1)  C(10)  C(11)  C(12)  118.7(2)  C(1)  N(1)  C(2)  112.3(1)  C(11)  C(12)  C(13)  122.2(2)  C(2)  N(2)  C(3)  112.1(1)  C(12)  C(13)  C(14)  116.8(2)  C(1)  N(3)  C(3)  112.4(1)  C(12)  C(13)  C(26)  120.5(2)  0(1)  C(1)  N(1)  113.3(1)  C(14)  C(13)  C(26)  122.7(2)  0(1)  C(1)  N(3)  119.1(1)  C(13)  C(14)  C(15)  121.9(2)  N(1)  C(1)  N(3)  127.6(2)  C(10)  C(15)  C(14)  118.8(2)  0(2)  C(2)  N(1)  119.6(2)  0(3)  C(16)  C(17)  113.9(2)  0(2)  C(2)  N(2)  112.8(2)  0(3)  C(16)  C(21)  125.7(2)  N(1)  C(2)  N(2)  127.7(2)  C(17)  C(16)  C(21)  1'20.3(2)  0(3)  C(3)  N(2)  111.8(1)  C(16)  C(17)  C(18)  120.0(2)  0(3)  C(3)  N(3)  120.3(1)  C(17)  C(18)  C(19)  122.0(2)  N(2)  C(3)  N(3)  127.9(2)  C(18)  C(19)  C(20)  116.2(2)  0(1)  C(4)  C(5)  120.4(2)  C(18)  C(19)  C(30)  121.4(2)  0(1)  C(4)  C(9)  118.2(2)  C(20)  C(19)  C(30)  122.3(2)  C(5)  C(4)  C(9)  121.2(2)  C(19)  C(20)  C(21)  122.9(2)  C(4)  C(5)  C(6)  119.0(2)  C(16)  C(21)  C(20)  118.7(2)  C(5)  C(6)  C(7)  122.4(2)  C(7)  C(22)  C(23)  112.3(2)  C(6)  C(7)  C(8)  116.5(2)  C(7)  C(22)  C(24)  109.6(2)  C(6)  C(7)  C(22)  120.5(2)  C(7)  C(22)  C(25)  108.9(2)  C(8)  C(7)  C(22)  122.9(2)  C(23)  C(22)  C(24)  107.9(2)  C(7)  C(8)  C(9)  121.9(2)  C(23)  C(22)  C(25)  107.3(2)  C(4)  C(9)  C(8)  119.0(2)  C(24)  C(22)  C(25)  110.7(2)  0(2)  C(10)  C(11)  117.1(2)  C(13)  C(26)  C(27)  110.4(3) 211  (Table IV continued) atom  atom  atom  C(13)  C(26)  C(27A)  109(1)  C(13)  C(26)  C(28)  109.4(2)  C(13)  C(26)  C(28A)  105.8(6)  C(13)  C(26)  C(29)  113.3(3)  C(13)  C(26)  C(29A)  105.6(6)  C(27)  C(26)  C(28)  108.0(4)  C(27)  C(26)  C(29)  107.7(4)  C(27A) C(26)  C(28A)  119(2)  C(27A) C(26)  C(29A)  107(2)  C(28)  C(29)  107.9(4)  C(28A) C(26)  C(29A)  110(1)  C(27A) C(27)  C(29A)  123(2)  C(19)  C(30)  C(31)  111.9(2)  C(19)  C(30)  C(32)  108.1(2)  C(19)  C(30)  C(33)  111.3(2)  C(31)  C(30)  C(32)  108.5(3)  C(31)  C(30)  C(33)  107.7(2)  C(32)  C(30)  C(33)  109.2(2)  C(26)  angle  212  (Table IV continued) atom  atom  atom  angle  C(4)  C(5)  H(1)  C(6)  C(5)  C(5)  atom  atom  atom  120.50  C(22)  C(23)  H(15)  109.47  H(1)  120.50  H(13)  C(23)  H(14)  109.47  C(6)  H(2)  118.81  H(13)  C(23)  H(15)  109.47  C(7)  C(6)  H(2)  118.81  H(14)  C(23)  H(15)  109.47  C(7)  c(8)  H(3)  119.05  C(22)  C(24)  H(16)  109.47  C(9)  C(8)  H(3)  119.05  C(22)  C(24)  H(17)  109.47  C(4)  C(9)  H(4)  120.53  C(22)  C(24)  11(18)  109.47  C(8)  C(9)  H(4)  120.52  13(16)  C(24)  H(17)  109.47  C(10)  C(11)  H(5)  120.66  11(16)  C(24)  H(18)  109.47  C(12)  C(11)  H(5)  120.65  H(17)  C(24)  H(16)  109.48  C(11)  C(12)  H(6)  118.92  C(22)  C(25)  H(19)  109.47  C(13)  C(12)  H(6)  118.91  C(22)  C(25)  H(20)  109.47  C(13)  C(14)  H(7)  119.04  C(22)  C(25)  H(21)  109.47  C(15)  C(14)  H(7)  119.04  H(19)  C(25)  H(20)  109.47  C(10)  C(15)  H(8)  120.61  H(19)  C(25)  H(21)  109.47  C(14)  C(15)  H(8)  120.61  H(20)  C(25)  H(21)  109.48  C(16)  C(17)  H(9)  120.01  C(26)  C(27)  H(22)  109.49  C(18)  C(17)  H(9)  120.02  C(26)  C(27)  H(23)  109.47  C(17)  C(18)  11(10)  119.02  C(26)  C(27)  H(24)  109.50  C(19)  C(18)  13(10)  119.03  H(22)  C(27)  H(23)  109.44  C(19)  C(20)  13(11)  118.57  11(22)  C(27)  13(24)  109.48  C(21)  C(20)  H(11)  118.57  13(23)  C(27)  11(24)  109.45  C(16)  C(21)  13(12)  120.65  C(26)  C(27A) H(22A)  109.51  C(20)  C(21)  H(12)  120.65  C(26)  C(27A) H(23A)  109.38  C(22)  C(23)  H(13)  109.47  C(26)  C(27A) H(24A)  109.51  C(22)  C(23)  H(14)  109.47  H(22A) C(27A) H(23A)  109.42  angle  213  (Table IV continued) atom  atom  atom  angle  atom  atom  atom  H(22A) C(27A) H(24A)  109.60  C(30)  C(31)  13(31)  109.47  H(23A) C(27A) H(24A)  109.42  C(30)  C(31)  14(32)  109.50  C(26)  C(28)  H(25)  109.45  C(30)  C(31)  14(33)  109.47  C(26)  C(28)  H(26)  109.45  14(31)  C(31)  14(32)  109.47  C(26)  C(28)  H(27)  109.44  14(31)  C(31)  14(33)  109.43  13(25)  C(28)  H(26)  109.51  14(32)  C(31)  14(33)  109.48  11(25)  C(28)  H(27)  109.49  C(30)  C(32)  H(34)  109.46  H(26)  C(28)  H(27)  109.49  C(30)  C(32)  13(35)  109.47  C(26)  C(28A) 13(25A)  109.57  C(30)  C(32)  14(36)  109.44  C(26)  C(28A) H(26A)  109.51  H(34)  C(32)  14(35)  109.52  C(26)  C(28A) H(27A)  109.51  14(34)  C(32)  11(36)  109.46  H(25A) C(28A) H(26A)  109.44  13(35)  C(32)  14(36)  109.48  H(25A) C(28A) H(27A)  109.44  C(30)  C(33)  H(37)  109.45  H(26A) C(28A) H(27A)  109.35  C(30)  C(33)  14(38)  109.49  angle  C(26)  C(29)  13(28)  109.43  C(30)  C(33)  11(39)  109.47  C(26)  C(29)  13(29)  109.47  14(37)  C(33)  H(38)  109.48  C(26)  C(29)  H(30)  109.45  H(37)  C(33)  14(39)  109.44  H(28)  C(29)  H(29)  109.49  14(38)  C(33)  13(39)  109.51  13(28)  C(29)  11(30)  109.47  H(29)  C(29)  H(30)  109.52  C(26)  C(29A) H(28A)  109.62  C(26)  C(29A) H(29A)  109.53  C(26)  C(29A) H(30A)  109.54  H(28A) C(29A) H(29A)  109.42  H(28A) C(29A) H(30A)  109.43  H(29A) C(29A) H(30A)  109.30 214  B. Crystal Structure Data for Compound 2 Table V. Crystal Data Empirical Formula^  C 17 H 14 N 2 0 2  Formula Weight^  278.31  Crystal Color, Habit^  colorless, prism  Crystal Dimensions^(mm)  0.250 X 0.400 X 0.400  Crystal System  monoclinic  No.^Reflections Used for Unit Cell Determination (28 range)  25^( 20.0 -  Omega Scan Peak Width at Half-height  0.35  Lattice^Parameters:  a b c 0  =  26.10)  10.072 (2)A 11.410 (2)A 13.351 (3)A 108.49 (2)e  V . 1455 (1)A3 Space Group  P2 1 /a (#14)  Z value  4  D  calc  1.270 g/cm 3  000  584  P(MoKa)  0.79 cm -1  215  Table VI. Final Atomic Coordinates  atom  x  Y  z  0(1)  0.6290(2)  0.7298(2)  0.5398(1)  0(2)  0.0664(2)  0.0994(2)  0.1896(2)  N(1)  0.6805(3)  0.9327(3)  0.5819(3)  N(2)  0.1602(3)  -0.0415(2)  0.0952(3)  C(1)  0.1644(2)  0.5929(2)  0.1848(2)  C(2)  0.0261(3)  0.6547(2)  0.1776(2)  C(3)  0.2080(3)  0.6329(2)  0.0898(2)  C(4)  0.2827(2)  0.6291(2)  0.2843(2)  C(S)  0.4001(3)  0.5586(2)  0.3247(2)  C(6)  0.5141(3)  0.5933(2)  0.4078(2)  C(7)  0.5076(3)  0.6996(2)  0.4518(2)  C(8)  0.3965(3)  0.7725(2)  0.4177(2)  C(9)  0.2838(3)  0.7364(2)  0.3327(2)  C(10)  0.1424(2)  0.4601(2)  0.1847(2)  C(11)  0.0766(3)  0.4101(2)  0.2512(2)  C(12)  0.0537(3)  0.2909(3)  0.2520(2)  C(13)  0.0970(3)  0.2214(2)  0.1858(2)  C(14)  0.1627(3)  0.2648(2)  0.1205(2)  C(15)  0.1850(2)  0.3849(2)  0.1201(2)  C(16)  0.6532(3)  0.8390(3)  0.5601(3)  C(17)  0.1193(3)  0.0261(3)  0.1389(3)  216  (Table VI continued) atom H(1)  0.0026  0.6400  0.2423  H(4)  0.3024  0.6041  0.0976  H(2)  0.0363  0.7393  0.1692  H(3)  -0.0487  0.6245  0.1167  H(5)  0.1421  0.6014  0.0246  H(6)  0.2072  0.7187  0.0865  H(7)  0.4018  0.4814  0.2930  H(8)  0.5969  0.5432  0.4341  H(9)  0.3956  0.8484  0.4518  H(10)  0.2028  0.7885  0.3063  H(11)  0.0457  0.4607  0.2988  H(12)  0.0071  0.2570  0.2995  H(13)  0.1941  0.2127  0.0741  H(14)  0.2325  0.4171  0.0725  217  Table VII. Bond Lengths (A)  atom  atom  distance  atom  atom  distance  0(1)  C(7)  1.443(3)  C(4)  C(9)  1.383(3)  0(1)  C(16)  1.281(3)  C(5)  C(6)  1.377(3)  0(2)  C(13)  1.430(3)  C(6)  C(7)  1.359(3)  0(2)  C(17)  1.292(4)  C(7)  C(8)  1.352(3)  N(1)  C(16)  1.120(4)  C(8)  C(9)  1.389(4)  N(2)  C(17)  1.122(4)  C(10)  C(11)  1.388(3)  C(1)  C(2)  1.537(3)  C(10)  C(15)  1.378(3)  C(1)  C(3)  1.538(3)  C(11)  C(12)  1.380(4)  C(1)  C(4)  1.533(3)  C(12)  C(13)  1.359(4)  C(1)  C(10)  1.531(3)  C(13)  C(14)  1.345(3)  C(4)  C(5)  1.390(3)  C(14)  C(15)  1.389(3)  218  Table VIII. Bond Angles  atom  atom  atom  angle  atom  atom  atom  angle  C(7)  0(1)  C(16)  117.4(2)  C(6)  C(7)  C(8)  123.5(2)  C(13)  0(2)  C(17)  118.8(2)  C(7)  C(8)  C(9)  117.9(2)  C(2)  C(1)  C(3)  107.6(2)  C(4)  C(9)  C(8)  121.9(2)  C(2)  C(1)  C(4)  111.7(2)  C(1)  C(10)  C(11)  120.3(2)  C(2)  C(1)  C(10)  109.0(2)  C(1)  C(10)  C(15)  122.8(2)  C(3)  C(1)  C(4)  106.8(2)  C(11)  C(10)  C(15)  116.8(2)  C(3)  C(1)  C(10)  111.9(2)  C(10)  C(11)  C(12)  121.7(2)  C(4)  C(1)  C(10)  109.9(2)  C(11)  C(12)  C(13)  118.7(2)  C(1)  C(4)  C(5)  120.8(2)  0(2)  C(13)  C(12)  115.3(2)  C(1)  C(4)  C(9)  122.2(2)  0(2)  C(13)  C(14)  122.5(2)  C(5)  C(4)  C(9)  116.8(2)  C(12)  C(13)  C(14)  122.2(2)  C(4)  C(5)  C(6)  122.4(2)  C(13)  C(14)  C(15)  118.7(2)  C(5)  C(6)  C(7)  117.6(2)  C(10)  C(15)  C(14)  121.9(2)  0(1)  C(7)  C(6)  114.9(2)  0(1)  C(16)  N(1)  176.4(4)  0(1)  C(7)  C(8)  121.6(2)  0(2)  C(17)  N(2)  176.5(3)  (Table VIII continued) atom  atom  atom  angle  atom  atom  atom  angle  C(1)  C(2)  H(1)  109.47  C(5)  C(6)  H(8)  121.21  C(1)  C(2)  H(2)  109.47  C(7)  C(6)  H(8)  121.21  C(1)  C(2)  H(3)  109.47  C(7)  C(8)  H(9)  121.06  H(1)  C(2)  H(2)  109.47  C(9)  C(8)  H(9)  121.06  H(1)  C(2)  H(3)  109.47  C(4)  C(9)  H(10)  119.08  H(2)  C(2)  H(3)  109.47  C(8)  C(9)  H(10)  119.07  C(1)  C(3)  H(4)  109.47  C(10)  C(11)  H(11)  119.15  C(1)  C(3)  H(5)  109.47  C(12)  C(11)  H(11)  119.15  C(1)  C(3)  H(6)  109.47  C(11)  C(12)  H(12)  120.63  H(4)  C(3)  H(5)  109.47  C(13)  C(12)  H(12)  120.63  H(4)  C(3)  H(6)  109.47  C(13)  C(14)  H(13)  120.67  H(5)  C(3)  H(6)  109.47  C(15)  C(14)  H(13)  120.67  C(4)  C(5)  H(7)  118.81  C(10)  C(15)  H(14)  119.05  C(6)  C(5)  H(7)  118.81  C(14)  C(15)  H(14)  119.04  220  ^  C. Crystal Structure Data for Compound 53  Table DC. Crystal Data Empirical Formula^  C 10 H 11 NO 3  Formula Weight^  193.20  Crystal Color, Habit^  colorless, irregular  Crystal Dimensions (mm) ^0.150 X 0.350 X 0.450 Crystal System^  triclinic  No. Reflections Used for Unit Cell Determination (28 range) ^25 (104.2 - 122.9°) Omega Scan Peak Width at Half-height^ Lattice Parameters:  0.20 a b = c = a 0^"' Y m  8.1373^(6)A 10.489^(1)A 6.0086^(6)A 103.552 (7)° 97.495 (8)° 72.003 (7)0  V =  473.25 (8)A3  Space Group  PI^(#2)  Z value  2  Dcalc  1.356 g/cm 3  F000  204  P(CuKa)  8.00 cm -1  221  Table X.^Final Atomic Coordinates atom  x  Y  z 0.0362(2)  0(1)  0.0999(1)  0.67281(9)  0(2)  0.1227(1)  0.8322(1)  -0.1370(2)  0(3)  0.2053(1)  0.4590(1)  0.2836(2)  N(3)  -0.0537(2)  0.8848(1)  0.1611(2)  C(2)  0.0593(2)  0.8029(1)  0.0097(2)  C(4)  -0.0895(2)  0.8153(1)  0.3200(3)  C(5)  0.0058(2)  0.6662(1)  0.2206(2)  C(6)  0.1307(2)  0.5976(1)  0.3935(2)  C(7)  0.3233(2)  0.3755(1)  0.4113(2)  C(8)  0.3691(2)  0.4153(2)  0.6412(3)  C(9)  0.4890(2)  0.3195(2)  0.7522(3)  C(1O)  0.5613(2)  0.1886(2)  0.6378(4)  C(11)  0.5153(2)  0.1507(2)  0.4087(4)  C(12)  0.3959(2)  0.2430(1)  0.2941(3)  H(1)  -0.087(2)  0.977(2)  0.163(3)  H(2)  -0.038(2)  0.845(2)  0.470(3)  H(3)  -0.214(2)  0.832(2)  0.330(3)  H(4)  -0.076(2)  0.612(1)  0.151(2)  H(5)  0.068(2)  0.602(1)  0.529(3)  H(6)  0.223(2)  0.646(2)  0.449(3)  H(7)  0.320(2)  0.509(2)  0.725(3)  H(8)  0.519(2)  0.347(2)  0.913(3)  H(9)  0.646(2)  0.120(2)  0.719(3)  H(10)  0.563(3)  0.060(2)  0.323(3)  H(11)  0.366(2)  0.217(2)  0.134(3) 222  Table XL^Bond Lengths (A)  atom  atom  distance^atom  atom  distance  0(1)  C(2)  1.344(1)'  C(5)  C(6)  1.501(2)  0(1)  C(5)  1.452(1)  C(7)  C(8)  1.383(2)  0(2)  C(2)  1.217(2)  C(7)  C(12)  1.386(2)  0(3)  C(6)  1.424(2)  C(8)  C(9)  1.395(2)  0(3)  C(7)  1.371(2)  C(9)  C(10)  1.367(3)  N(3)  C(2)  1.332(2)  C(10)  C(11)  1.377(3)  N(3)  C(4)  1.434(2)  C(11)  C(12)  1.383(2)  C(4)  C(5)  1.529(2)  N(3)  H(1)  0.92(2)  C(8)  H(7)  0.98(2)  C(4)  H(2)  0.97(2)  C(9)  H(8)  0.96(2)  C(4)  H(3)  0.98(2)  C(10)  H(9)  1.00(2)  C(5)  H(4)  1.00(1)  C(11)  H(10)  0.96(2)  C(6)  H(5)  1.00(1)  C(12)  H(11)  0.96(2)  C(6)  H(6)  1.01(1)  223  Table XII.^Bond Angles atom  atom  angle  0(3)  C(6)  C(5)  107.2(1)  117.5(1)  0(3)  C(7)  C(8)  124.5(1)  C(4)  113.2(1)  0(3)  C(7)  C(12)  115.1(1)  C(2)  0(2)  121.0(1)  C(8)  C(7)  C(12)  120.4(1)  0(1)  C(2)  N(3)  110.1(1)  C(7)  C(8)  C(9)  118.7(1)  0(2)  C(2)  N(3)  128.9(1)  C(8)  C(9)  C(10)  121.2(2)  N(3)  C(4)  C(5)  101.5(1)  C(9)  C(10)  C(11)  119.4(1)  0(1)  C(5)  C(4)  104.83(9)  C(10)  C(11)  C(12)  120.8(2)  0(1)  C(5)  C(6)  109.6(1)  C(7)  C(12)  C(11)  119.4(2)  C(4)  C(5)  C(6)  112.6(1)  C(5)  C(6)  H(6)  110.3(9)  C(2)  N(3)  H(1)  118(1)  H(5)  C(6)  H(6)  108(1)  C(4)  N(3)  H(1)  129(1)  C(7)  C(8)  H(7)  121(1)  N(3)  C(4)  H(2)  109(1)  C(9)  C(8)  H(7)  120(1)  N(3)  C(4)  H(3)  112(1)  C(8)  C(9)  H(8)  119(1)  C(5)  C(4)  H(2)  110(1)  C(10)  C(9)  H(8)  120(1)  C(5)  C(4)  H(3)  114(1)  C(9)  C(10)  H(9)  121(1)  H(2)  C(4)  H(3)  110(2)  C(11)  C(10)  H(9)  120(1)  0(1)  C(5)  H(4)  107.2(8)  C(10)  C(11)  H(10)  122(1)  C(4)  C(5)  H(4)  111.9(8)  C(12)  C(11)  H(10)  117(1)  C(6)  C(5)  H(4)  110.3(8)  C(7)  C(12)  H(11)  120(1)  0(3)  C(6)  H(5)  110.7(8)  C(11)  C(12)  H(11)  121(1)  0(3)  C(6)  H(6)  110.7(8)  N(3)  H(1)  0(2)  170(2)  C(5)  C(6)  H(5)  109.8(8)  atom  atom  atom^angle^atom  C(2)  0(1)  C(5)  109.74(9)  C(6)  0(3)  C(7)  C(2)  N(3)  0(1)  224  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061801/manifest

Comment

Related Items