UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of molar ratio and pH during synthesis on the performance of phenol-melamine-formaldehyde… Sidhu, Avtar Singh 1998

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

Item Metadata

Download

Media
831-ubc_1998-0303.pdf [ 5.06MB ]
Metadata
JSON: 831-1.0088445.json
JSON-LD: 831-1.0088445-ld.json
RDF/XML (Pretty): 831-1.0088445-rdf.xml
RDF/JSON: 831-1.0088445-rdf.json
Turtle: 831-1.0088445-turtle.txt
N-Triples: 831-1.0088445-rdf-ntriples.txt
Original Record: 831-1.0088445-source.json
Full Text
831-1.0088445-fulltext.txt
Citation
831-1.0088445.ris

Full Text

THE EFFECT OF MOLAR RATIO AND pH DURING SYNTHESIS ON THE PERFORMANCE OF PHENOL-MELAMINE-FORMALDEHYDE ADHESIVES by AVTAR SINGH SIDHU B.Sc, The University of British Columbia, 1988 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF REQUIREMENT FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Wood Science We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1998 © Avtar Singh Sidhu, 1998 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. Department of The University of British Columbia Vancouver, Canada Date DE-6 (2788) ABSTRACT In North America, almost all exterior grade plywood, particleboard, or flakeboard is made with phenol formaldehyde (PF) resins. Formulations are available which can satisfy a wide range of working and performance property demands of the users. PF resins suffer from a few disadvantages which include the distinct dark brown color in the cured glueline and the relatively high temperature (120°C at glueline) required for curing. Modification of PF resins with several different chemical groups has been tried to reduce or remove the drawbacks" associated with PF resin. One particular modification technique involves the addition of melamine into the PF resin system. Although modification using melamine has been carried out in Europe and in Japan, where the melamine is cheaper than phenol, no such attempt has been made in North America, where the price of melamine is higher than that of phenol. Even though melamine is an expensive chemical compared to phenol, its advantages may lie in the lower cure temperatures and shorter press cycles that are required during hot pressing operations. In this study, an array of phenol-melamine-formaldehyde (PMF) resins were synthesized by varying the formaldehyde/phenol and formaldehyde/melamine ratios at pH 7.5 and 9.0. Melamine formaldehyde (MF) and PF resins were also synthesized for comparison. The structure of all these resins, as well as commercial MF and commercial PF resins, was characterized using Fourier Transform Infrared (FTIR) spectroscopy, Proton Nuclear Magnetic Resonance (1H-NMR) spectroscopy, Gel Permeation Chromatography (GPC) and Differential Scanning Calorimetry (DSC). The bond performance of these resins was evaluated by producing 3-ply plywood panels ii and testing these panels for shear strength and wood failure under dry, wet and boiled conditions. Better bond performance was observed for PMF resins that were synthesized at pH 7.5 than the ones synthesized at pH 9.0. Very reasonable wood failure results were obtained for these resins and also for MF resins (synthesized and.commercial) at lower press temperatures(120°C) and lower press times (3 min.) compared with the PF resin. The existence of co-condensation in PMF resins was confirmed with the IR and NMR analysis. The majority of the co-condensation in resins prepared at pH 7.5 was by way of methylene bridges. The bond performance was attributed to the presence of melamine in the system and not to the level of coplymerization that occured between phenol and melamine. These resins and also the MF resins were of much lower molecular weight (<1,500) in comparison with the PF (>20,000) resins. DSC data of PMF resins cooked at pH 7.5 showed that these resins exhibit two exotherms, the first exotherm (150°C) corresponding to the condensation reactions which take place during curing while the second exotherm (220°C) possibly corresponding to the elimination of formaldehyde from the dimethylene ether links to form methylene cross-links. iii TABLE OF CONTENTS Page ABSTRACT ii TABLE OF CONTENTS .' iv LIST OF FIGURES vii LIST OF TABLES • • xi ACKNOWLEDGEMENT xii DEDICATION ' xiii QUOTATION xiv 1. INTRODUCTION 1 2. LITERATURE REVIEW 7 2.1 Resin Chemistry 7 2.2 Melamine Formaldehyde Synthesis : 9 2.3 Synthesis of Phenol Melamine Formaldehyde Resin 14 2.4 Resin Curing 17 2.5 Resin Characterization 27 2.5.1 Instrumental Techniques 28 2.6 Bond Performance -30 3. METHODOLOGY 33 3.1 Resins ' : • 33 3.1.1 Laboratory Resins 33 3.1.1.1 Phenol Melamine Formaldehyde Resin Synthesis 35 3.1.1.2 Phenol Formaldehyde Resin Synthesis 37 3.1.1.3 Melamine Formaldehyde Resin Synthesis 38 iv 3.1.2 Commercial Resins ••••39 3.2 Resin Analysis • 39 3.2.1 Fourier Transform Infrared Absorption Spectroscopy 40 3.2.2 Differential Scanning Calorimetry 41 3.2.3 Gel Permeation Chromatography 42 3.2.4 Nuclear Magnetic Resonance Spectroscopy 43 3.3 Wood Bonding Study 44 3.3.1 Gluing Procedure 44 3.3.2 Sample Preparation and Testing 45 4. Results and Discussion 50 4.1 Resin Synthesis : 50 4.1.1 PMF and MF Resins 50 4.1.2 PF Resins 53 4.2 Resin Properties ' 54 4.2.1 Fourier Transform Infrared Spectroscopy 54 4.2.1.1 Synthesized and Commercial PF and MF resins 54 4.2.1.2 PMF Resins 58 4.2.2 Proton Nuclear Magnetic Resonance Spectroscopy 63 4.2.3 Gel Permeation Chromatography 78 4.2.4 Differential Scanning Calorimetry 90 4.3 Wood Bonding Results 104 4.3.1 Dry Specimens 107 4.3.2 Soaked Specimens 114 4.3.3 Boiled Specimens 117 4.3.4 Summary of Wood Bonding 120 5. SUMMARY 123 6. CONCLUSION 126 7. LITERATURE CITED 127 APPENDIX I 132 APPENDIX II 135 vi LIST OF FIGURES Figure Page 1. Resonance structures of melamine 7 2. Additional resonance structures of melamine 8 3. Resonance structures of melamine as a result of tautomerism 8 4. Reaction of melamine with formaldehyde 10 5. Reaction mechanism for methylol melamine formation 11 6. Proposed structure of phenol melamine formaldehyde resin 14 7. Reaction products from the reaction between methylol phenol and methylol melamine -15 8. Formation of methylene and ether linkages 18 9. Formation of methylene linkage during the reaction of hexamethylolmelamine 19 10. Structure of melamine formaldehyde resin as proposed by Koehlerand Frey (1943) , 21 11. Structure of 2,4,6-trimethylolmelamine at low pH 22 12. Mechanism for the polymer growth at pH 6.3 according to Sato and Naito (1973) i 23 13. Alternative mechanism for polymer growth,according to Sato and Naito (1973) • 24 14. 'H-NMR spectra of PFM, PF, MF, and PF+MF 26 15. Models for the polymeric structures of the PMF resin 27 16. Resin kettle for preparation of resins 34 17. Gel Permeation Chromatography System 43 18. Veneer Orientation in Panel 45 19. Dimensions of the tension shear specimens 46 vii 20. "Sure Grip" Wedge Grips for tension shear test 48 21. Divisioned plexiglass for wood failure determinations 49 22. Hydrophobe solids and water tolerance profile for the PMF and MF resin cooks , 52 23 IR spectra for (a) synthesized and (b) commercial PF resins 55 24. IR spectra for (a) synthesized and (b) commercial MF resins 57 25a,b. IR spectra for PMF resins prepared at pH 9.0 59 25c. IR spectra for PMF resins prepared at pH 9.0 60 26. IR spectra for PMF resins prepared at pH 7.5 61 27. NMR spectra for synthesized MF resin 65 28. NMR spectra for commercial MF resin 66 29. NMR spectra for synthesized PF resin 67 30. NMR spectra for commercial PF resin 68 31. NMR spectra for synthesized MF resin 69 i 32. NMR spectra for synthesized MF resin after addition of D 2 0 70 33. NMR spectra for Cook #1 PMF resin 71 34. NMR spectra for Cook #2 PMF resin 72 35. NMR spectra for Cook #3 PMF resin : 73 36. NMR spectra for Cook #5 PMF resin 74 37. NMR spectra for Cook #6 PMF resin 75 38. GPC calibration curve using poystyrene standards 79 39. (a) Molecular weight distribution and (b) GPC detector response for synthesized PF resin 80 40. GPC spectra of synthesized and commercial PF resins 82 41. GPC spectra of synthesized and commercial MF resins 83 42. (a) Symmetric and (b) Asymmetric trimethylolomelamines 84 43. GPC spectra of synthesized PMF resins prepared at pH 9.0 87 viii 44. GPC spectra of synthesized PMF resins prepared at pH 7.5 88 45. Typical DSC spectra for PF resins showing (a) a single peak and (b) two overlapping peaks 92 46. The peak temperatures and the amount of heat evolved for (a) MF and (b) PF resins during different advancement stages of synthesis 93 47a,b. The peak temperatures and the amount of heat evolved for (a) Cook #1 and (b) Cook #2 during different advancement stages of synthesis 94 47c. The peak temperatures and the amount of heat evolved for Cook #3 during different advancement stages of synthesis 95 48a,b. The peak temperatures and the amount of heat evolved for Cook #5 and Cook #6 during different advancement stages of synthesis 96 49. DSC spectra for (a) synthesized PF and (b) commercial PF resins 99 50, DSC spectra for (a) synthesized MF and (b) commercial MF resins 100 51a,b. DSC spectra for (a) Cook #1 and (b) Cook #2 resin 101 51c. DSC spectrum for Cook #3 resin 102 52a,b DSC spectra for (a) Cook #5 and (b) Cook #6 resins 103 53. Pressure and temperature profiles for the 3 minute and 5 minute press times for panels pressed at 120°C .....105 54. Pressure and temperature profiles for the 3 minute and 5 minute press times for panels pressed at 150°C : 106 55. Two-way interactions (resin x time and time x temperature) for shear strength of dry specimens 111 56. Two-way interactions (resin x temperature) for wood failure of dry specimens : 113 57. Two-way interactions (resin x time and resin x temperature) for shear strength of wet specimens 115 58. Two-way interactions (resin x time and resin x temperature) for wood failure of wet specimens 116 59. Two-way interactions (resin x time) and three-way interactions (resin x time x temperature) for shear strength of boiled specimens 118 60. Two-way interactions (resin x time) and three-way interactions (resin x time x temperature) for wood failure of boiled specimens 119 ix I A1. Graph of the main effects of all resins from the ANOVA tables, (a) shear strength (b) wood failure 138 A2. Non-transformed average shear strength values for panels pressed at 120°C for 3 and 5 minutes ; 139 A3. Non-transformed average shear strength values for panels pressed at 150°C for 3 and 5 minutes 140 A4. Non-transformed average wood failure values for panels pressed at 120°C for 3 and 5 minutes 141 A5. Non-transformed average wood failure values for panels pressed at 150°C for 3 and 5 minutes 142 LIST OF TABLES Table Page 1. The selected mole ratios and pH conditions for PMF resins 35 2. Reactants for the synthesis of PF resin 37 3. Reactants for the synthesis of melamine formaldehyde resin 38 4. Resin solids, viscosity and free formaldehyde for all synthesized resins 50 5. Proportion of functional groups (relative to one another) present in all the resins ' 79 6. Molecular Weights (Mn and Mw) and the polydispersity index for all resins 89 7. Analysis of variance of shear strengths for dry specimens .: 108 8. Analysis of variance of wood failure for dry specimens 108 9. Analysis of variance of shear strengths for wet specimens 109 10. Analysis of variance of wood failure for wet specimens 109 11. Analysis of variance of shear strengths for boiled specimens 110 12. Analysis of variance of wood failure for boiled specimens 110 A1. Non-transformed average shear strength values for panels pressed at 120°C for 3 and 5 minutes 133 A2. Non-transformed average wood failure values for panels pressed at 120°C for 3 and 5 minutes . 134 A3. Non-transformed average shear strength values for panels pressed at 150°C for 3 and 5 minutes 135 A3. Non-transformed average wood failure values for panels pressed at 150°C for 3 and 5 minutes 136 xi Acknowledgements I would like to thank my late supervisor, Dr. Paul Steiner for giving me the opportunity to pursue my graduate studies while I was an employee in the Wood Science Department at the University of British Columbia. I also like to express my gratitude to Dr. Simon Ellis for offering me direction and guidance as well as kind criticism throughout the latter part of my M.Sc. program. A special note of thanks goes to Dr. David Barrett for allowing me to continue my M.Sc. program after Dr. Paul Steiner's death in May of 1995. I would also like to thank Dr. Frank Lam for his support and encouragement. I want to also acknowledge Dr. Laszlo Paszner, James T. White, Dr. Gary Troughton, Axel Anderson and Dr. Sergey Shevchenko for their help and their very insightful suggestions. I am very grateful to Martin Feng from Bordon Company Limited, Sammy Edwards from Melamine Chemicals Inc.,Tom Holloway from Neste Resins Corp., and Dr. Bunichiro Tomita for their very kind help in supplying the commercial resins. Most of all, I am very grateful to my wife, Karen, for her support and understanding without whom my graduate studies would not have been possible. xii To my sons, Kevier and Karmvier nephew, Harvier and my niece, Sukhjote "Poor indeed is the student who does not become better than his teacher." (Cited by Pizzi, in Advanced Wood Adhesives Technology, 1994) 1. INTRODUCTION Phenol formaldehyde (PF) and urea formaldehyde (UF) are resins with a long standing history remaining very important polymers among synthetic resins today. In the wood composite industry, UF resins have been used successfully for interior applications whereas PF resins have been utilized for both interior and exterior applications. Both resins offer several advantages to the consumer/user. One big advantage offered by PF resins is durability whereby wood composites bonded with them can withstand severe outdoor conditions over extended periods of time. UF resins offer faster curing rates and competitive pricing. However, together with these advantages, there are some inherent disadvantages for both resins. Wood composites manufactured with UF resin tend to delaminate when exposed to high moisture or high humidity environments. PF resins cure at a relatively high temperature and require long press times. Furthermore, the PF glue-line exhibits an unattractive dark color. To overcome some of these disadvantages, there has been an attempt in the last two decades to incorporate melamine in the PF and UF systems (Maylor,1995). Melamine copolymer resins (so called, melamine-urea-formaldehyde (MUF), melamine-urea-phenol-formaldehyde (PMUF) and phenol-melamine-formaldehyde (PMF) binders have been developed which give improved moisture resistance, low formaldehyde emissions and lighter colored glue-lines. Melamine (2,4,6-triamino-1,3,5-triazine) was investigated by Liebig as early as 1834 (Hodgins, 1941). It was almost forgotten, however, because no use was found for it until the development of the melamine resins around 1935. In 1936, Ciba A. G. in Basel, Switzerland, applied for a patent for a process producing melamine from calcium cyanamide (Widmer, 1965). This process made melamine comparatively inexpensive l and readily available. Since then, the chemistry of melamine has been extensively studied. Initially, the discovery of a cheap melamine synthesis threatened the use of UF adhesives. Since melamine offers six active hydrogen atoms for reaction with formaldehyde, it yields better cross-linking and thus better water-resistant adhesives. Thus, several resin producers, CIBA, Heubel (BP 455, 000 by W. Hentisch and R. Koehler 1936) and others shifted production to melamine-formaldehyde (MF) resins (Widmer, 1965). Initially, because of limited production capacity and relatively higher cost, the melamine resins were only used to fortify existing urea resin products and left the bulk plywood and particleboard markets to UF resins only. In fact, MF resin has established itself as a wood adhesive in Europe and Japan, but even as of today, has not yet reached a comparatively great importance in North America (Maylor, 1995). The MF resin consumption in North America is driven by the surface coatings and laminate markets, which account for approximately 71% of MF resin demand while wood adhesives account for only 6% of MF resin demand. In Japan, the wood adhesives market accounts for 58% of the demand for MF resins (Gorbaty et al., 1994). Initially, improved heat and water resistance imparted to molded and laminated products were the principal features noted upon the addition of melamine to PF and UF adhesives. With the advent of World War II, military requirements gave considerable impetus to the development and production of melamine resin materials (Widmer, 1965). Generally speaking, the period from 1940 to 1950 was very productive in the growth of amino resins. MF resins have been used in many thermosetting resin applications such as molding resins; adhesives (mainly for plywood and furniture); laminating resins for counter, cabinet, and table tops; textile resins to impart crease resistance, stiffness, shrinkage control, water repellency and fire retardance; wet-2 strength resins for paper. MF resins have also been used in alkyd resin preparations to give baking enamels such as for automotive finishes. MF and UF resins fall in the category of aminoresin adhesives or aminoplastic adhesives. These are important members of the thermosetting class of synthetic resin products made by combining an aldehyde with a compound containing an amino (-NH2) group (Wohnsiedler, 1952). The global amino resin capacity reached 18 billion pounds in 1993 and UF resin accounts for over 80% of this amount; MF resin accounts for most of the rest (Gorbaty et al, 1994). Therefore, urea and melamine resins are the most prominent members of the amino resin class. The less prominent members of the amino resin class are based on thiourea, aniline, ethylene urea and the guanamines. The advantages of aminoresin adhesives are their "(1) initial water solubility which renders them eminently suitable for bulk and relatively inexpensive production, (2) hardness after curing, (3) nonflammability after curing, (4) good thermal properties, (5) absence of color in cured polymers and (6) easy adaptability to a variety of curing conditions" (Pizzi, 1983). Urea-and melamine-formaldehyde condensation products have very much in common in their chemical behavior, production of the intermediates during curing, capacity to be cured to high molecular-weight condensation polymers and properties of the final products (which can be used, to a wide extent, in the same type of applications) (Maylor, 1995). In the manufacture of melamine-urea-formaldehyde (MUF) resins, urea and melamine together are reacted with formaldehyde, which results in the formation of addition products, such as methylol compounds. Further reaction, and the concurrent elimination of water leads to the formation of low molecular weight condensates which are still soluble in water. Higher molecular weight products that are insoluble and infusible, are obtained by further 3 condensing the low molecular weight condensates and cross-linking. The main differences between these two amino resins are the better water and heat resistance, greater hardness, and the capacity to be cured more rapidly and under slightly basic conditions of melamine- compared to urea-formaldehyde resins. Some researchers believe that the disadvantage of these aminoplastic resins is their bond deterioration caused by water and moisture. This is a result of the hydrolysis of the aminoplastic or aminomethylenic bond (Pizzi, 1983). This seems to be true more for UF and MUF resins but not for the MF and PMF resins. The higher resistance of melamine-formaldehyde resins to moisture attack is due to the considerably lower solubility of melamine in water. It is important to note, that melamine dissolves in hot water only, whereas, urea dissolves in both hot and cold water. Therefore, urea-formaldehyde adhesives are used for interior applications only whereas MF and MUF resins can be employed successfully for rather severe outdoor conditions. Melamine-formaldehyde adhesives have replaced urea-formaldehyde resins in wood gluing to only a certain extent. They produce high quality plywood because their adhesive joints are boil-proof (Pizzi, 1983). Phenol-modified melamine-urea-formaldehyde (PMUR) resins have been used for production of exterior grade particleboard and oriented strand board (OSB) in Germany and France (Clad and Schmidt-Hellerau, 1977). In the Asia-Pacific region, PMF resins have been used for the production of moisture resistant medium density fiberboard (MDF) (Maylor, 1995). In Japan, PMF resins, in combination with urea-formaldehyde resins, have been used for the manufacturing of concrete form plywood (Tamura etal., 1981). 4 Phenol-melamine-formaldehyde resins obtained by incorporating melamine in the PF resin system cure at a lower temperature than the PF resin alone, thus offering an advantage over phenolic resin itself from the aspect of saving energy in the production of plywood (Tamura, 1981). The copolymerization which occurs between the methylols of phenol and melamine might play a major part in the resin curing at lower temperatures and shorter press times. Aqueous phenol-melamine resins have good adhesion to softwood and resinous tropical hardwoods such as apitong. Many researchers uphold the good weather resistance of these recently developed phenol triazine resins. The PMF resins used in the manufacture of concrete form plywood have been traditionally made by first synthesizing PF and MF resins separately and then mixing the two in appropriate proportions prior to use. According to Higuchi et al. (1994), it is not possible to achieve copolymerization between melamine and phenol unless the resin is synthesized with both melamine and phenol in a single cook. While some work has been carried out by Higuchi et al. (1994) to elucidate the chemical structure of PMF resins, very little is known about the curing nature (i.e., speed of cure, temperature of cure) of these resins. There is also very little information available on how the bond performance is altered with change in formaldehyde/phenol and formaldehyde/melamine molar ratios; at what temperature the PMF resin can be optimally cured and what time frame is required to give optimal bond strength. The current study was undertaken to investigate the curing characteristics and bond performance of the phenol-melamine- formaldehyde resin systems. The first step involved synthesizing an array of PMF resins at several different molar ratios and under different pH conditions. Control resins (MF and PF resins) were also synthesized for comparison purposes. All resins were subsequently characterized by proton-nuclear 5 magnetic resonance (1H-NMR) spectroscopy and fourier transform infrared (FTIR) spectrophotometry to identify the relative proportion of different functional groups present. The molecular weight distribution of each resin was determined by gel permeation chromatography (GPC) and the curing behavior of each adhesive was determined by differential scanning calorimetry (DSC). Finally, the bond performance of each resin was assessed by first bonding three veneers together (3-ply plywood) and then testing these panels for shear strength and wood failure after conditioning the samples at three different conditions: (1) The first set was conditioned to 50% relative humidity at 20 °C. (2) The second set was vacuum-pressure soaked. (3) Finally, the third set was boiled, dried and then boiled again to evaluate the long-term performance of the bonded specimens. 6 2. LITERATURE REVIEW 2.1. Resin Chemistry In the field of organic chemistry melamine (2,4,6-triamino-1,3,5-triazine) enters into many reactions which are of considerable interest to the resin chemist. Hofmann (1874) first found that melamine produces resinous products when he was. working with tetra-phenylmelamine (Hodgins et. al., 1941). Chemically, the most important property of melamine is the capacity for combining with formaldehyde to give resins. To provide a background for this discussion, it is necessary to examine the structure of melamine more closely. Hughes (1941) made an intensive study of the structure of melamine in the crystalline state. According to this work the triazine ring is a resonance structure of delocalized electrons. Its resonance system extends to the lone pair of electrons of nitrogen atoms outside the ring (Figure 1). Figure 1. Resonance structures of melamine. Other less symmetrical structures may also make contributions, one set being (Figure 2): + H2N' 7 +NH2 +NH2 1 Jl Figure 2. Additional resonance structures of melamine. In view of the possibility of the occurrence of tautomerism in melamine, three additional structures have received consideration (Wohnsiedler, 1952) (Figure 3). NH2 NH2 N NT I  i f ^ H Normal or amino form Diamino-imino form N H 2 N H N H H I s r S s i H H N ^ N ^ N J H H N ^ N ^ N J H H H Amino-diimino form Iso-form Figure 3. Resonance structures of melamine as a result of tautomerism. 8 Which structure best expresses the properties of melamine is a problem in itself. Studies using ultraviolet absorption spectroscopy have shown that melamine undergoes structural changes in acid versus neutral or alkaline solutions (Dixon et al., 1947). From the point of view of this study and the fact that the triazine ring is a very stable structure at high temperatures, the benzenoid structure is favored with probably the diamino-imino form existing as a second or exclusive structure in acid solutions. In neutral solutions, the melamine is believed to exist as the iso-form. 2.2. Melamine Formaldehyde Synthesis The basic primary reactions between melamine, urea, substituted melamine or substituted urea and formaldehyde are very similar (Pizzi, 1994). Formaldehyde addition to melamine occurs more easily and more completely than its addition to urea. Under a specified set of conditions the primary reaction (addition) results in the attachment of formaldehyde to the nitrogen of an amino group to form methylol compounds and these methylol compounds further undergo condensation reactions with the splitting off of water and/or formaldehyde to form higher molecular weight intermediates. Since each melamine molecule contains three amino groups, in each of which either or both of the hydrogen atoms may be substituted, it is clear that a good many methylol derivatives of melamine may appear. Methylol derivatives which may be produced range from monomethylol to the hexamethylol melamine (Figure 4). The nature of these products is determined by the conditions under which.the reactions between the formaldehyde and melamine take place. Conditions which are important are: the relative proportions of formaldehyde and melamii,e present, the pH of the aqueous formaldehyde, the temperature and finally the time of the reaction. 9 Figure 4. Reaction of melamine with formaldehyde. A typical reaction mechanism that takes place during the hydroxy-methylation reaction in alkaline media involves an attack of a nitrogen anion of melamine on the carbonyl carbon of formaldehyde (Figure 5). 10 Figure 5. Reaction mechanism for methylol melamine formation. Because melamine is less soluble in water than urea, the hydrophilic stage proceeds more rapidly in MF resin formation than in UF resin formation. Therefore, hydrophobic intermediates of the MF condensation appear early in the reaction. Another important difference between MF and UF is that the MF condensation and curing occurs not only under acid conditions but also under neutral or even slightly alkaline conditions (Pizzi, 1983). This characteristic is often an advantage. Curing capacity in a non-acid medium is a highly valued property in electrical applications and in applications in which corrosion is a problem. 11 Of the nine methylol products possible from the condensation of melamine and formaldehyde, the most stable and readily isolated is hexamethylolmelamine (Ho.dgins et al., 1941). This compound may be produced either by heating melamine with an excess of neutral formaldehyde to 90°C, or by allowing the melamine to react with neutral formaldehyde at room temperature over a period of 15 to 18 hours. Elemental analysis indicates that the product formed in both cases is the same and contains one molecule of water of crystallization per molecule of hexamethylolmelamine. Another methylol compound that has been prepared is a trimethylolmelamine (one methylol group on each amino group). This compound is obtained by reacting one mole of melamine with three moles of neutral or slightly alkaline formaldehyde at room temperature for 15 hours. The entire separation and purification of this material must be carried out with extreme rapidity and at very low temperatures to prevent further reaction. Elemental analysis shows that under these conditions the compound crystallizes from aqueous solutions with two molecules of water of crystallization. Because of this extreme reactivity,,the material has no definite melting point and is not stable on standing at room temperature. The progressive polymerization of the methylol melamine is very dependent upon the reaction pH. Small changes in acidity or alkalinity have greater effect on the reaction rate than in the case of the methylol ureas. Therefore, the control of the pH during the reaction of melamine with formaldehyde is of paramount importance. The technical grades of formaldehyde often contains 0.5 -1.0% of formic acid. Even this small amount of formic acid cannot be allowed during resin synthesis due to the fact that it will often acts as a catalyst and renders the reaction uncontrollable. With poor pH 12 control, a soft gel like white precipitate will form which will quickly transform into a hard block of resin and is often difficult to remove from the reaction vessel. The optimum pH range for methylolation and further condensation to medium and high molecular weight resins is 8.5 - 9.0. The temperature of the reaction should be well controlled in the range of 80 - 90°C, otherwise, it may lead to viscous, hydrophobic MF resins with separation of the upper water layer. At pH 7, water-soluble, hydrophilic polymers are first formed. Under continued heating the resin solutions have less tolerance for dilution with water. In commercial practice, the reaction is carried only to the water-soluble stage and then inhibited by increasing the pH to about 10 and cooling. The solubility and viscosity of the reaction mixture can be repeatedly monitored by two methods, the water tolerance test or the hydrophobe test. The water tolerance is checked by adding water to a 10 mL sample of reactants until a definite cloud develops throughout the solution. The amount of water added to bring the solution to a cloud point is multiplied by 10 to give the water tolerance value (Ashland Chemical Co., standard quality control method). The hydrophobic solids are calculated in much the same way as is water tolerance. A 20 g sample of the reactants is weighed and to this sample water is added until a cloud point persists in the solution after shaking. The hydrophobe solids (H.S.) are calculated by the following formula (Gaylord, 1968). A*B H.S.(%) = -—- A = weight of sample A ~l~ B = % theoretical solids C = milliliters of water required to reach cloud point. It is important to realize that as the molecular weight of the resin syrup increases, the water tolerance decreases and the hydrophobe solids increase. Therefore, by controlling the pH and viscosity, the melamine formaldehyde solution can be kept at the 13 desired condensation stage by arresting the reaction. Further condensation is prevented by cooling and by bringing the pH to the level which guarantees best stability as mentioned earlier. 2.3 Synthesis of Phenol-Melamine-Formaldehyde Resin. Preparation of synthetic resin from melamine and phenol can be carried out by either co-condensation, where melamine and phenol are mixed together and then reacted with formaldehyde, or by the process of combining two phenolic and melamine resins, each being prepared in separate cooks. When a mixture of melamine and phenol are condensed with formaldehyde, there are two kinds of "homopolymers" (phenolic and triazine type) existing in the three-dimensional structure of macromolecules of resulting products (Chen-Chun et. al., 1982) (Figure 6). Figure 6. Proposed structure of phenol melamine formaldehyde resin. 14 This reaction is usually carried out in the pH range of 6.0 - 9.5 within a short time. When this reaction is carried out under alkaline conditions, the formaldehyde reacts with phenol and melamine separately to form various hydroxymethyl substitutes with different degrees of substitution. The reactivity of these substituted compounds is very high. Hydroxymethyl groups react not only with hydrogen atoms on the triazine ring and on the phenol ring, but also react with each other (Figure 7). OH Figure 7. Reaction products from the reaction between methylol phenol and methylol melamine. These prepolymers then undergo further dehydration. By means of methylene bonds and ether bonds, a three dimensional heteroaromatic complex structure is finally formed, as shown in Figure 6. 15 In the preparation of this resin, the phenol imparts a light brown to red-brown color to the resin and to any articles subsequently molded from it, as well as imparting a brown color in gluelines when used as an adhesive in plywood manufacture. Upon exposure to light or air this brown color has a tendency to deepen. Several methods of obtaining colorless melamine modified phenol formaldehyde resins have been proposed. According to British Patent No. 1,057,400 (Ibigawa Electric Industry Company Limited, 1963), the polycondensations are carried out at a pH value of 6.0 to 7.5. U.S. Patent No. 3,321,551 (Knutson, 1967) proposes a multistage process where the^polycondensation of a phenol formaldehyde precondensate with melamine is carried out at a pH value of 6.9 to 7.8. Also, in this process, the use of strong alkalis such as sodium hydroxide or potassium hydroxide is not permitted since they are believed to be responsible for producing brown-colored solutions due to the oxidation of free phenol. Instead, neutralizing agents (carbonates selected from the group consisting of the carbonates of calcium, barium and magnesium) are employed. In U.S. Patent No. 4,229,557 (Feinauer et. al.,1979), the production of melamine-phenol-formaldehyde resins, which are substantially white and resistant to yellowing, is carried out by the polycondensation of melamine with phenol and formaldehyde in aqueous basic reaction medium at a pH value between 8 and 11. This process comprises adding to the reaction mixture one or more water-soluble ammonium or alkali phosphates or ammonium or alkali borates in an amount of at least 0.05% by weight, based on the reaction mixture or the dry resin, and subsequently isolating the resulting resin by dehydration at a temperature of at least 70°C or heating the dry resin to at least 70°C. The reason for using this particular pH range is to improve the storage life of the resin. 16 2.4. Resin Curing During the curing process, the prepolymers formed during the addition and condensation reactions are transformed to highly branched, three-dimensional structures. An increase in molecular size by linear reaction and cross-linking leads to formation of macromolecules and the hardened or cured products become insoluble and infusible. These hardening, wetting or curing reactions are most easily accomplished by the application of heat and either acid or alkali. The methylol melamines readily undergo condensations which transform them from small molecules, simple chemical entities to very large molecular polymers. This change is of great importance, as the desirable properties of the melamine resins, such as insolubility, infusibility and ease of film formation, are associated exclusively with the highly polymerized state (Powers, 1947). The most active participants in the condensation process appear to be -NH 2 groups because the more of these get substituted, the less methylol melamine condensation occurs. For example, hexamethylolmelamine is much more stable than monomethylolmelamine (Moncrieff, 1947). The actual reactions which take place during polymerization are complex since many different reactions occur simultaneously. The final resinous polymeric products are characterized by the possession of both methylene and methylene ether linkages, which are probably derived from reactions of the type shown in Figure 8. 17 NH 2 N H 2 N ^ N r ( N ^ N , H 2 N — l ^ ^ ^ J — N H C H J O H + HjNr+-l^NJl—NHChfeOH N ^ N N ^ N —NHCH2NH —NHCHfeOH H 2 N Methylene Linkage NH 2 N H 2 "l j fX - A A H 2 N — \ 4?—NHCH2|OH + H!OCH 2Nrf ^IST N H 2 N L 1 H 2 N — — N H C H 2 O C H 2 H N J l ^ ^ J — N H 2 Methylene Ether Linkage Figure 8. Formation of methylene and methylene ether linkages. It must be realized that for all methylolmelamines containing one or more unreplaced hydrogen atoms, the mechanism can be readily formulated as going through the methylene linkage. However, with hexamethylolmelamine, it is necessary to split off formaldehyde in order to make such a linkage possible, as shown in Figure 9. In the polycondensation of melamine formaldehyde, Gams (1941) attached particular importance to the ether linkage and assigned a minor role to the methylene bridge as a polymer-forming linkage (Wohnsiedler, 1952). This conclusion was drawn after an 18 introductory study of the possible resinifying linkages in a hexamethylolmelamine and several other condensates. During this introductory study, it was realized that there were two linkages which may exist between molecules of methylolmelamine-methylene linkage and the methylene ether linkage. Figure 9 shows that with hexamethylolmelamine, it is necessary to split off formaldehyde in order to make such a linkage possible. In any other methylolmelamine which contains one or more unreplaced hydrogen atoms, the mechanism can be readily formulated as going through the methylene linkage. + n(3CH 20) Figure 9. Resinification through (1) Methylene and (2) Methylene Ether Linkages. 19 With such a multiplicity of functional groups and with at least two main reactions which are possible, very large molecules will soon be built-up. Moreover, these large polymeric molecules possess the desired properties of water and chemical resistance, heat stability and film-forming ability. Based on the work by Koehler and Frey (1943) a somewhat simplified possibility for the constitution of a melamine resin in the final cured stage is shown in Figure 10 (Wohnsiedler, 1952). Besides some unreacted methylol groups and methylene groups, the presence of many ether bridges is emphasized. This is because in curing melamine formaldehyde resins at temperatures of up to 100°C, no substantial amounts of formaldehyde are liberated. Only small quantities are liberated during curing up to 150°C. In contrast, urea formaldehyde resins cured under the same conditions liberate a great deal of formaldehyde. 20 CH 2OH NH JL X ^ Y 1 — N h f ^ N ^ N H C H 2 O C H 2 N ^ N N JvJ._ N H — N H C H 2 O C H 2 N H N H C H 2 -N ^ N V^J - N H y N ^ N H C H 2 N A N A N H C H 2 o C H 2 N H N S N N H C H ^ N H C H 2 O C H 2 N h r N N H C H 2 O C H 2 N H ^ N Y N / H | N H C H 2 O C H 2 N H -N H C H 2 O C H 2 N H -Figure 10. Structure of melamine formaldehyde resin as proposed by Koehler and Frey (1943). The reaction mechanisms of the acid catalyzed condensation of methylol-melamine to form polymers and resins, have been elucidated by Sato and Naito (1973). As mentioned earlier, the melamine assumes different resonance ionic forms at different pH values. Methylolmelamines can readily be visualized as existing in similar charged or ionic forms. For example, at low pH values, a symmetrical trimethylol-melamine molecule should exist in the following ionic form (Figure 11): 21 HOCH 2H •NHCH2OH NH +CH 2OH Figure 11. Structure of 2,4,6-trimethylolmelamine at low pH. It is believed that at pH values of approximately 6.3 some of the ions begin to form (Wohnsiedler, 1952). One of the possible mechanism for polymer growth is through the interaction of the ionic and molecular forms of trimethylolmelamine (Figure 12). The reaction takes place between a methylol group attached to an amine of one methylolmelamine and an amine hydrogen of another ring. With the formation of the dimer, a molecule of water is released. This process takes place to a major extent with charged and uncharged dimers coexisting. Coincident with the release of water, the triazine ring reverts to the aromatic structure. 22 N N HOHjCHN-C^ ^C-NHCHjOH K , N - C ^ C-NHCrt,OH MH N^. ^ N • NHj NHCHjOH (HtScHjOHofMFj) ( M C H 2 O H o f MF 2) H O - v — H HOHjCHN-Cj^ ^ C - N H — ^ • • • • < ^ N — C ^ ^ C - N H C ^ O H .NH H H H N ^ HCR.OH H O H / I H N - C j ^ ^ C NHCHjNH C ^ ^ C — N H CK.OH • N ^ NHCHjOH H O K ^ H N ^ - ^ ^ C NHCRjNH C ^ ^ C — N H C ^ O H q < J . ^ N L ^ N + ^ C w tilHj NHCKjOH Figure 12. Mechanism for the polymer growth at pH 6.3 according to Sato and Naito (1973). There are alternative mechanisms that are possible (Figure 13). With the participation of ether and azomethine-type structural mechanisms at least two other ionic mechanisms are possible (Wohnsiedler, 1952). 23 HOHjCHN-C-^ ~ ^ C — N H C H 2 O H HOHjCHN-Cj ^ C - N H C H 2 O H C C + UH2 + /1H2 ( HMCHjOH of M F Z ) (HMCHjOH of M F 2 ) HOHfJHN-rjj ^ C — N H c—-•—N—<jj C ~ * N H C H 2 0 H N. ,NH H H CH7 N^ ,NH *~ H OH I' • N K , + N H ? HOH.CHN-C^ C^ NHCH2N C^-NHC^ OH II I I I I + HO N^ N^H CH2 N^^ NH ^ Figure 13. Alternative mechanism for polymer growth according to Sato and Naito (1973). Phenol-melamine-formaldehyde curing behavior has been studied very extensively by Roh et. al. (1987a,b, 1989,1990,1991). In their early studies they used a two-step extraction with a 4% aqueous sodium hydroxide solution and formic acid to separate the three-dimensionally cross-linked fraction from soluble fractions. Reasonable relationships between cure time and the amount of the insoluble fraction, the degree of swelling and the nitrogen content were obtained. It was shown that the infrared and ultraviolet absorption spectroscopy methods, which were previously described for determining deqree of cure of phenolic resins, were not successful in PMF systems. Roh et al. (1987a), investigated the effects of pH and temperature on the curing behavior of PMF resin. It was found that under acidic conditions (pH 5.0-6.0) the 24 condensation of melamine was dominant in the rapid-first-stage and the polymerization of phenol took place in the slower-second-stage of the curing reaction. Moreover, it was realized that when a long hot pressing time was applied to promote the slow stage in the curing of the PMF resin, the durability of plywood was greatly increased. Roh et al. (1989) determined, by varying the molar ratios of phenol to melamine (P/M), formaldehyde to melamine (F/M) and formaldehyde to phenol (F/P), that the co-condensation of phenol with cross-linked melamine occurs in the second stage of PMF synthesis. By the use of proton-nuclear magnetic resonance (1H-NMR) spectroscopy, Roh et al. (1990) demonstrated that the formation of the insoluble fraction in PMF resin is caused mainly by the co-condensation of phenol with melamine and also that the curing of the PMF resin at pH 8.5 progresses mainly by way of the formation of a dimethylene ether linkage between phenol and melamine. Higuchi (1990) suggested that knowledge of the polymeric structure of adhesives is very important in understanding and predicting the performance of adhesives. Investigation of the curing behavior of phenol-melamine-formaldehyde as well as melamine formaldehyde resins was carried out in order to deduce polymeric structures of these resins. Even though, 1H-NMR and 1 3C-NMR techniques can be used to distinguish the bonds between phenolic nuclei and melamine residues, there are no means available to determine the amount of methylene and dimethylene ether bonds which might be formed between melamine and other components in cured resins. When PMF resins were cured under acid conditions, (pH 5.0), it was found that the condensation of the melamine fraction was predominate over that of the phenol fraction. Curing at pH 8.5, the rate of conversion of melamine fraction was about the 25 same as that of the phenol part. It was also deduced that the co-condensation between phenol and melamine occured at this latter pH. The 'H-NMR spectrum of the product of the reaction of methylolated phenol with melamine at pH 8.5 compared with that of a mixture of a PF resin and a MF resin was obtained as shown in Figure 14. The spectrum for PMF resin shows a peak at 4.3 ppm. This peak was assigned to methylene linkages formed between phenolic and the triazine nuclei. 5 4 5 4 5 4 6 5 4 ppm Figure 14. 1 H-NMR spectra of PFM, PF, MF, and PF+MF Note: PFM: Product of the reaction between methylolated phenol and melamine. PF: Phenol formaldehyde resin. MF: Melamine formaldehyde resin. PF+MF: Mixture of PF and MF (1:1). 26 In the course of curing at pH 10.0, the conversion of the phenol fraction preceded that of the melamine part. However, in the late stages of the curing course, the rate of conversion of the melamine part exceeded that of the phenol fraction. From this study Higuchi et al. (1990) proposed the following models for the polymeric structures of the PMF resin (Figure 15). — M — M — M — M — I I — M — M — M — M — (PH 5) -M — P — M — P — I I M M I I " M P — P — M — (pH 8.5) -P M — P M - P P I I p — p — M — P I M (pH 10) Figure 15. Models for the proposed polymeric structures of the PMF resin. Note: P,M and — denote a phenolic nucleus, a melamine residue and a methylene bond (or a dimethylene ether bond), respectively. The methylol groups are not illustrated. 2.5. Resin Characterization In order to gain a deeper understanding of the performance of wood adhesives, it is necessary to gain a wider knowledge of their polymeric structures. In the cured state, thermosetting polymers are generally insoluble, infusible materials which are difficult to examine by many analytical techniques. Many of the desirable physical 27 properties of adhesives are the result of their highly cross-linked structure (Ellis, 1989). In this regard, it is often desirable to determine the structure and concentration of each component formed when all reactants react under a specified set of conditions. The nature and distribution of molecular species formed during the preparation and subsequent thermal cure of PMF resins depend on reaction conditions such as F/M/P ratio, the nature and amount of catalyst present, and the reaction time and temperature. Full characterization of the resin is difficult because the complex distribution of molecular products formed are essentially insoluble, intractable materials (King et al., 1974). Nonetheless, several spectroscopic methods have been utilized for the elucidation of the chemical structure and to study the cure characteristics of these resins. 2.5.1. Instrumental Techniques Koeda (1954) originally demonstrated the chromatographic separation of monomeric methylol melamines of various degrees of substitution by means of paper chromatography. Braun and Legradic (1974a) used a silylation technique whereby the different methylo! compounds were separated and characterized by gel permeation chromatography (GPC). Silylation involved the modification of the reaction product of melamine and formaldehyde by treatment with N-diethyl-trimethylsilyl-amine to yield materials that were soluble in organic solvents, such as dimethyl formamide (DMF) or tetrahydrofuran (THF). Anderson et al. (1970) described the characterization of monomeric etherified methylolated melamines using GPC and 1H-NMR spectroscopy. GPC analysis permits a rapid and accurate determination of the monomer and polymer 28 content of the products. 1H-NMR spectroscopy allows the qualitative identification of the etherifying alcohols and a quantitative estimation of the alkoxy groups present. Chiavarini et al. (1976) presented a method based upon 1H-NMR and chemical analysis by which they claimed it is possible to determine the content of methylol, methylene and oxymethylene formaldehyde, moreover non-substituted, monosubstituted and disubstituted amino groups. The MF condensates were examined in dimethyl sulfoxide d6/CaCI2 solvent phase without any chemical pretreatment. Tomita and Ono (1979) used Fourier Transform 1 3C-NMR Spectroscopy to characterize random chemical structures of melamine formaldehyde resins, including methylated melamine formaldehyde resins and urea melamine formaldehyde resins. Tomita and Matsuzaki (1985) used 1 3C-NMR spectroscopy to analyze the chemical reactions between phenol or urea with melamine and formaldehyde and also to determine the conditions to produce their co-condensation. It was found that in the case of co-condensation between urea and phenol, the synthetic method to react methylolphenols with an excessive amount of urea itself under acidic conditions was desirable. In the case between melamine and phenol the co-condensation was achieved by reacting methylolphenols with methylolmelamines under a pH below neutral. Braun and Ritzert (1984) investigated the condensation reaction between phenol, melamine and formaldehyde at different conditions (pH, molar ratio of phenol and melamine) using gel permeation chromatography, 1 3C-NMR spectroscopy and thermogravimetric techniques. It was found that the reactivity of phenol in the reaction with formaldehyde was insignificant at low pH values. At alkaline conditions this 29 reactivity increased very quickly. Melamine showed an inverse behavior. Braun and Ritzert (1984) also described a method to determine the curing of phenol-melamine-formaldehyde condensate. The curing time was determined by estimation of free formaldehyde and by thermogravimetric analysis. The extractive separation of phenol (as well as cresol) and melamine was characterized by IR spectroscopy. Tomita and Matsuzaki (1985) used 1 3C-NMR spectroscopy and GPC to analyze the chemical reactions of phenol-melamine-formaldehyde and phenol melamine urea systems and to determine the conditions necessary to produce their co-condensation. Roh et al. (1990) studied the curing behavior of phenol-melamine-formaldehyde resin using 1H-NMR spectroscopy. It was deduced that at pH 8.5 co-condensation between phenol and melamine took place and also that curing of the phenol-melamine-formaldehyde resin at pH 8.5 progresses mainly by way of the formation of a dimethylene ether linkage between phenol and melamine. 2.6. Bond Performance Bond strength depends in part upon the polymerization process as the adhesive is transformed from the liquid to solid state. When adhesive is first applied to wood it wets and interacts with the surface. The degree of adhesive penetration into the wood during the wetting and pressing process often influences strength development and stress distribution in the cured glueline (Pizzi, 1983). Too little adhesive penetration can result in wood-adhesive interface failure in service. Too much adhesive penetration gives starved gluelines during pressing and results in poor initial bond quality. Ideally, the penetration should be sufficient to reinforce the interface while providing strength to bond the wood elements. Generally speaking, a glueline should be just as strong and 30 durable as the wood it joins. It should mimic the cyclic dimensional changes that can occur in the composite during service. Unfortunately, the breakdown of an adhesive bond is not a simple linear response that occurs with time. In some cases a gluebond may appear to be of good quality for several years then suddenly deteriorate rapidly and fast. In others, this deterioration may occur slowly even going through several plateau stages. Bond degradation may be fully attributed to the adhesive (i.e. cohesive failure in the glueline) or depend on the adhesive-wood interface (i.e. adhesive failure) or be the result of wood breakdown (i.e. cohesive failure in the wood). Often a combination of all these factors is involved. National Bureau of Standards (1974) have established a standard method for evaluating adhesive bonds in plywood. Typically, cross laminates (veneers at 90°to one another) are used in plywood for dimensional stability (Gollob, 1982). Both wood failure and breaking load are parameters used for glueline evaluation in shear specimens (ASTM D906-73). In order to forecast the long-term performance capabilities of resin systems, accelerated aging methods have been developed. Both a vacuum-pressure-soak and boil cycle are established aging treatments for plywood type bonds (Millet, 1977). Specifications for acceptance of good quality plywood differ from country to country (Chow, 1972a). In Canada, the Canadian Standards Association (CSA) standard is used as a criteria of bond quality and it stipulates an average of 80% wood failure must be achieved in panels tested. No mention is made of strength requirements. In the United States, the American Plywood Association (APA) has a similar average wood failure requirement of 85% and also no mention of strength. In Europe, if we examine the German DIN standard, the panels tested must have an average shear strength of 114 psi with no panels allowed below 105 psi. This standard 31 makes no mention of wood failure. Thus North American standards are based on wood failure while many other countries base their standards on shear strength. Chow and Warren (1972) found that percentage wood failure is a more sensitive measure of the undercure of glue bonds than shear strength. The percentage wood failure is the parameter for indicating the level of adhesion of plywood, while the ultimate shear strength can be a reflection of the veneer quality after adequate adhesion has been developed to bond the wood. Although extensive work has been carried out in evaluating the bonding performance of melamine based adhesives such as MF and MUF, very little has been reported on in evaluating the performance of PMF resins. Hse (1992) evaluated MUF resin systems for bonding structural flakeboards. The bond evaluations were carried out by cutting internal bond (IB) specimens from panels. The IB specimens were tested in accordance with American Society for Testing and Materials (ASTM) Standards for evaluating the properties of wood-base fiber and particle-panel materials (D-1037-72). Hse found that MUF resin could be used to bond flakes at higher moisture contents than with conventional PF resins. Thus, the MUF systems afford energy savings in drying. Various other authors, Blomquist and Olson (1955, 1964), Detweiler (1953) and Selbo(1965), have shown that the addition of melamine in the UF systems results in an increase in durability of these adhesives. Roh et al. (1989) evaluated the durability of PMF resin bonded joints. In relation to the melamine resins, no increase in durability was observed for PMF resin by itself but striking increase in durability was observed when a high molecular-weiyht novolak powder was added to the PMF resins. Novolak is a PF resin which is incapable of curing without the addition of hardening agents. 32 3. METHODOLOGY 3.1. Resins 3.1.1. Laboratory Resins Phenol-melamine formaldehyde resins were prepared to represent a range of synthesis variables and molecular characteristics. For comparison purposes, phenol formaldehyde and melamine formaldehyde resins were also prepared. Selected properties of each resin were monitored during the resin synthesis: viscosity, water tolerance, temperature, pH and thermal properties. The resins were synthesized in a two-liter Pirex resin kettle with a water jacket as shown in Figure 16. The pH electrode and thermometer probe were inserted through one of the ground glass ports of the lid to monitor the pH and temperature of each cook. The thermometer probe was connected to an Omega digital thermometer (model HH22) and the pH electrode was connected to a Fisher pH meter (model 915).The reaction was heated or cooled by the flow of hot or cold water through the water jacket as required. Samples were taken during the progress of each cook to measure viscosity (PF resins only), water tolerance (MF and PMF resins), and thermal behavior (using a DSC). Viscosity, solids content and free formaldehyde levels were measured on all end products. 33 Water out Stirrer Motor Temperature Compensator pH Electrode Thermometer Water in Condenser Sample Port > I Water in , —• Water out Stirrer Figure 16. Resin kettle for preparation of resins (Adapted from Ellis (1989)). 34 3.1.1.1. Phenol Melamine Formaldehyde Resin Synthesis The synthesis procedure and resin synthesis parameters were selected after considerable discussion with industry personnel. Several patents also provided insight into the preparation of PMF resins (U.S. Patents, 4,229,557; 4,611,020; 3,321,551 and 2,328,592). Three molar ratios (F/P/M) and two different pH conditions were selected for this particular study as shown in Table 1. With the exception of Cook #4 all cooks were possible to carry out. Cook #4 was attempted four times but without any success. Table 1. The selected mole ratios and pH conditions for PMF resins. Cook Mole Ratio # (F/P/M) pH Comments 1 4.0/1.4/0.7 9.0 O.K. 2 4.0/0.7/1.4 9.0 O.K. 3 4.0/1.071,0 9.0 O.K. 4 4.0/1.4/0.7 7.5 DID NOT WORK 5 4.0/0.7/1.4 7.5 O.K. 6 4.0/1.0/1.0 7.5 O.K. In all the cooks, the formaldehyde solution was charged into the resin kettle first, followed by phenol and melamine respectively. Formaldehyde solution (37%) was charged into a resin kettle and the pH was adjusted appropriately to either pH 7.5 or 9.0 with 25% sodium hydroxide or 25% formic acid solution. After adding the appropriate amounts of phenol and melamine, the pH was again adjusted: (1) For Cooks #1 to #3 the pH was adjusted to 9.0 with 25% sodium hydroxide solution and (2) For Cooks #4 35 to #6 the pH was adjusted to 7.5 with 25% formic acid. The addition of formic acid was suggested by industry personnel. Preliminary experiments suggested that Cooks #1 to #3 could be carried out safely at temperatures of about 85°C to 90°C where as Cooks #4 to #6 could be carried out at 75°C to 80°C. If these temperatures were exceeded the reaction would often become too vigorous and go out of control leaving a large precipitate formation in the resin kettle. Therefore, after the adjustment of the pH conditions, the solutions were heated to their respective temperature range in 20 - 30 minutes. The temperature and pH of each cook was closely monitored, and if required, adjustments were made to keep conditions constant throughout the cook. Samples were taken from the cooks at specified intervals to measure the water tolerance of the reactants and to measure the hydrophobe levels. The reactions were continued at their respective temperatures until the hydrophobe solids reached approximately 23%. This level corresponds to a water tolerance of approximately 200%. It is generally believed that at this stage, resins provide an adequate shelf life. After reaching the target hydrophobe solids levels, each reaction was arrested by quickly lowering the temperature to room temperature. The viscosity of each cook was determined by using a Brookfield digital viscometer (model DV-III). Free formaldehyde levels were determined for each resin by the standard hydroxylamine hydrochloride method. The solids content of each cook was determined by heating one gram of sample to a constant weight at a temperature of 125°C. 36 3.1.1.2. Phenol Formaldehyde Resin Synthesis The synthesis of phenol formaldehyde resin was based on that used for the production of commercial type resins described in Glued Wood Products Laboratory manual (U.B.C.,1991). This synthesis was carried out in a resin kettle as described in the previous section. The resin cook consisted of the following reactants: Table 2. Reactants for the synthesis for PF resin. Phenol 1.0 mole Formaldehyde 2.0 moles (52.8% solution) Sodium Hydroxide 50% solution (added in three portions) Water to make up 45% solids All the phenol and formaldehyde were first charged into the resin kettle with the first portion of sodium hydroxide and water. The solution was heated to 60°C in approximately 30 minutes. An exotherm was observed causing the temperature to rise to almost 100°C. After the exotherm subsided, heat was applied to the reactants to raise the temperature to 95°C until viscosity N (measured using Gardner-Holt viscosity tubes) was reached at which point the solution was cooled to 65°C. At this point the second sodium hydroxide portion was added and the temperature of the solution was raised to 90°C. The temperature of the solution was kept at 90°C until viscosity R was reached at which point the solution was cooled to 50°C and a third sodium hydroxide portion was added. After stirring the solution for five minutes it was cooled to room 37 temperature. The viscosity, free formaldehyde and solids contents were determined as described for PMF resin synthesis. 3.1.1.3. Melamine Formaldehyde Resin Synthesis The procedure followed for the synthesis of melamine formaldehyde resin was based on the commercial cook used by a company called Melamine Chemicals Inc. (unpublished). Table 3. Reactants for the synthesis of melamine formaldehyde resin. Melamine 1.0 mole Formaldehyde 3.0 moles Water Used for making a 37% formaldehyde solution and adjusting the solids content. Sodium Hydroxide 25% solution for adjusting the pH of the solution during the cook. Formic Acid 25% solution for adjusting the pH of the solution during the cook. Water was added to the formaldehyde solution (52.8% solution) to make 37% solution. The pH of the formaldehyde solution was adjusted to 9.0 using 25% sodium hydroxide solutions after which it was charged into a resin kettle. The melamine was slowly added with vigorous stirring. The solution was heated to reflux in 30-40 minutes and the pH adjusted to 9.4. After refluxing the solution at 90°C, it was cooled to 85°C and the pH was again adjusted to 9.4. The solution was held at 85°C until a water tolerance of 200% was reached. The progress of the cook was monitored by measuring 38 the water tolerance and the hydrophobe solids as described earlier. After the water tolerance endpoint was reached, the solution was cooled to room temperature and the pH was adjusted to 9.6. It is believed that the solution is much more stable at this pH than at lower pH conditions and further condensation or polymerization of the resin is prevented. As before, the viscosity, free formaldehyde and solids content were determined for the end product. 3.1.2. Commercial Resins For comparison purposes, two commercial liquid resins were used in the analysis and bonding study. One was a phenol formaldehyde resin (Borden Packaging and Industrial Materials, PF658) and the other was a melamine formaldehyde resin (from Neste Resins Corp., CB5044). 3.2. Resin Analysis All synthesized and commercial resins were analyzed using Fourier Transform Infrared (FTIR) absorption spectroscopy, differential scanning calorimetry (DSC), gel permeation chromatography (GPC) and proton nuclear magnetic resonance (1H-NMR) spectroscopy. DSC analyses were performed on liquid and solid samples but FTIR, GPC and NMR analyses were performed on only freeze-dried samples of each resin. The freeze-drying procedure used here is the same as the one used by Ellis (1989). Liquid resin samples were slowly poured into a tray containing liquid nitrogen after which they were ground with the bottom of a glass beaker. Small solid chunks of the resin (still in liquid nitrogen) were put into an Edwards High Vacuum Freeze Dryer. After 39 freeze drying for two days, the samples were further ground into a fine powder using a mortar and pestle. Due to the insoluble nature of the resins in most solvents that are required for NMR and GPC analysis, these powder resins were acetylated prior to their analysis. The acetylation procedure also followed that of Ellis(1989). To 0.6 g of freeze dried sample, 20 mL of pyridine and 20 mL of freshly-distilled acetic anhydride were added. This solution was placed in an ice bath for approximately two hours at which point almost all the solids were dissolved. After letting the reaction sit at room temperature for 72 hours, it was poured into 75 mL of ice water in a separatory funnel and extracted with 60 mL of dichloromethane. The dichloromethane layer was washed with three 50 mL portions each of 2% hydrochloric acid, 7% sodium carbonate and distilled water. To this end product, 2 to 3 scoops of magnesium sulphate were added for further drying. After 24 hours of drying over magnesium sulphate the solution was filtered, rotary evaporated and dried further over phosphorus pentoxide in a vacuum dessicator for five days. The end product was a solid or gummy substance and this product was used for the GPC analysis. 3.2.1. Fourier Transform Infrared Absorption Spectroscopy The FTIR spectra were obtained using potassium bromide (KBr) discs prepared in a die using approximately 0.5 mg of unacetylated dried resin and 200 mg of potassium bromide in the usual manner. The spectra were run on a Perkin Elmer 1600 series Fourier transform infrared instrument. The instrument was used in single ratio transmission mode. The sample chamber was continually purged with nitrogen while the spectra were obtained. 40 3.2.2. Differential Scanning Calorimetry (DSC) The differential scanning calorimetric analysis was performed on a Thermal Analyst 2000 system (Dupont Instruments) equipped with a DSC pressurized cell. DSC is a thermal analysis technique that measures the temperatures and heat flows associated with material transitions. Such measurements provide quantitative and qualitative information about endothermic (heat absorbed) or exothermic (heat evolved) processes, as well as changes in heat capacity. DSC is primarily used to characterize polymers and other organic materials, but is also applicable to metals, ceramics, and other inorganics. Since thermoset resin (i.e., PF, UF, PMF, etc.) curing is accompanied by the evolution of heat (exothermic reaction), DSC can be used to evaluate their degree of cure or degree of cross-linking. For the resin analysis, a sample of approximately 5-10 mg was run as an aqueous resin solution or as resin solids. This sample was carefully weighed into a hermetic pan which was later crimped with a cover. One small hole was made on top of the sealed capsule to ensure proper equilibration of sample during the application of pressure. This assembly was placed on one of the platforms in the DSC cell. On the other platform, a reference (empty) pan with the lid was placed. The calculations for heat flow are based on the differential heat flow to the sample and reference pans. The analysis was performed at 4.83 MPa (700 psi) pressure with heating temperature from 25 to 250°C at a heating rate of 10°C/min. Runs were always carried out using an empty pan as a reference. Prior to making any runs the DSC cell was calibrated at the running conditions for temperatures and heat of fusion using standard materials which melt in the temperature range of interest. Indium metal was particularly valuable for 41 temperature calibration since its melting point (156.6°C) lies in the middle of the temperature range of interest. 3.2.3. Gel Permeation Chromatography The system used for the GPC analysis was that of Ellis (1989). Figure.17 shows the basic setup of the equipment. Prior to running the acetylated resin samples, the system was calibrated using polystyrene standards of different molecular weights. The GPC system consisted of an isocratic Spectra Physics 8810 pump, Rheodyne 7125 injector loop, Varian Micropak TSK exclusion column system (GH8P guard column, and four analytical columns; 1000H, 2500H, 3000H and 4000H), Kratos Spectroflow 757 UV/VIS detector set at wavelength of 254 nm and Spectra Physics 4290 integrator. Raw slice data were stored on a floppy diskette and later analyzed using MS Excel software package. Since the solvent system used in this system was THF, sample solutions of 0.5% w/v of the acetylated resins in THF were used. A flow rate of 1.0 mL/min of THF was used in all the analysis. Three replicate analysis of each sample were performed. 42 Injector Valve Detector Waste Analytical Columns Guard Column Integrator Computer Figure 17. Gel Permeation Chromatography System 3.2.4. Nuclear Magnetic Resonance Spectroscopy Deuteriated dimethylsulfoxide (d6-DMSO) was the solvent of choice in the NMR analysis for all resins except the synthesized and commercial PF resins. These PF resins were acetylated prior to analysis in order to eliminate any problems from insolubility of the resins in selected solvents. A very small amount of the resin (2-5 mg) was dissolved in 1 mL of solvent (deuteriated chloroform (d6-CDCI3) for acetylated resins and deuteriated dimethylsulfoxide (d6-DMSO) for unacetylated resins). The spectra were obtained on a Bruker WH-400 400 MHz NMR instrument in the Department of Chemistry, U.B.C, using solvent as the internal standard. Integration was carried out on each spectra to determine the area for each peak. All the peaks were referenced to that of the solvent peak (7.24 ppm for d6-CDCI3 and 2.51 ppm for d6-DMSO). 43 3.3. Wood Bonding Study To study the bond performance of each resin, wood-glue bonds were produced between Douglas fir [Pseudotsuga menziesii] veneers (30 cm x 30 cm x 3.3 mm) using each resin. The average moisture content of these veneers was 7% prior to making the panels. Each of the resins, with the exception of the PF resin, required the addition of 14% (based on the "actual" resin) wheat flour (Rogers #2) to control the flow properties. The resulting panels were hot pressed at two temperatures (120°C and 150°C) and two time periods (3 and 5 minutes). Five replicate samples were prepared for each temperature and time combination. Eight resins were used for bonding. Therefore, the number of panels prepared was, 8x2x2x5=160. 3.3.1. Gluing Procedure The resin (containing only 14% wheat flour as filler) was applied to one veneer on the lathe checks (or loose) side by means of a hand roller spreader. A resin loading of 200g/m2 was used. The rough or porous stock required a slightly heavier spread (1 to 2% more) as opposed to the smooth veneer. A second veneer was gently placed on top of first veneer with the lathe checks facing down and the grain oriented perpendicular to the grain of the first of veneer (Figure 18). 44 Figure 18. Veneer Orientation in Panel The same spread of resin was applied to this veneer. The third veneer was placed on the first two with lathe checks facing down and the grain oriented perpendicular to the second veneer. This assembly was pre-pressed for approximately 10 minutes under approximately 10 kg weight after which it was hot pressed in an electrically heated Wabash hot press (30 cm x 30 cm). The panels were pressed to a compression level of approximately 85-90% of the original veneer thickness (3-ply) using pressure control. 3.3.2. Sample Preparation and Testing From the resulting pressed panels, nine specimens were taken for determination of the shear strength and the wood failure values. The dimensions and orientation of the shear specimens are shown in Figure 19. The grooves were made so that the cut extended right through the first veneer and about half-way through the second veneer. 45 Figure 19. Dimensions of the tension specimens. Since the shear specimens were going to be exposed to one of three different conditions (treatments) prior to testing, three specimens from each panel were randomly assigned to each different condition. The three conditions under which the shear specimens were conditioned were as follows:-(a) Samples were conditioned under constant temperature and humidity conditions (26.6 ± 1°C, 50±2% RH) until they equilibrated, this took approximately two weeks. 46 (b) A vacuum pressure cycle, where the specimens were submerged in cold water and a vacuum drawn (from a water aspirator) for one hour, followed by a pressure of 0.41 MPa (60 psi) for one hour. The samples were tested in the wet condition. (c) A boil treatment, where the specimens were boiled in hot water for four hours, dried for 20 hours at 60°C and then boiled again for four hours. The samples were placed in cold water prior to testing. Following the conditioning step, all the specimens were tested in tension on a Sintech screw-type testing machine using "Sure Grip" Wedge Grips to hold the specimen during testing (Figure 20). 47 Figure 20. "Sure Grip" Wedge Grips for tension shear test. 48 The rate of separation of the grips was kept at 10 mm/min for all specimens. The percentage wood failure values were estimated visually. The soaked and boiled specimens were dried at 60°C for 24 hours before their wood failure values were determined. To assist in the determination of wood failure, a piece of plexiglass with one end (25.4mm x 25.4mm) divided into sixteen equal divisions was used (Figure 21). Since the MF and PMF (Cooks #1 to #3) resins were of very light color, great care had to be exercised when looking at the wood failure for specimens bonded with these. The differentiation between the glue and wood was possible for the tested specimens. Figure 21. Divisioned plexiglass for wood failure determinations. 49 4. Results and Discussion 4.1. Resin Synthesis 4.1.1. PMF and MF resins Table 4 shows the percent resin solids, viscosity, and free formaldehyde levels for the resins that were successfully synthesized. Table 4. Resin solids, viscosity and free formaldehyde for all synthesized resins. (A takeout copy of this table is available in the Appendix). C o o k # Type of R e s i n Mole Ratio (F/P/M) P H R e s i n S o l i d s (%) V i s c o s i t y (cps) Free Formaldehyde (%) 1 PMF 4.0/1.4/0.7 9.0 53.0 37 0.13 2 PMF 4.0/0.7/1.4 9.0 50.3 45 0.15 3 PMF 4.0/1.0/1.0 9.0 52.4 42 0.17 4 PMF 4.0/1.4/0.7 7.5 - - -5 PMF 4.0/0.7/1.4 7.5 51.0 30 0.27 6 PMF 4.0/1.0/1.0 • 7.5 52.1 33 0.33 7 MF 2.0/1.0 9.5 54.0 55 0.18 8 PF 2.0/1.0 12.5 39.0 2400 0.23 Resin #4 was repeated three times as mentioned earlier in the methodology section without success. The combination of pH conditions and molar ratio resulted in a reaction which was very vigorous and uncontrollable. It was realized early in this work that the reactions carried out at lower pH conditions (i.e., pH 7.5) must be done with great care. The temperature had to be controlled so that it did not exceed 80°C at any time during the reaction. If the temperature exceeded 80°C, the reactants heated up too 50 quickly and formed a white precipitate which was worthless for further production. According to Pizzi (1983), the optimum methylolation pH for melamine and formaldehyde is in the range 8.5-9.0. Within this range, methylolation and further condensation to medium - and even high - molecular weight resins can be safely carried out at about 80°C. However, if the resin synthesis is not well controlled, it may lead to a viscous, hydrophobic MF resin, with separation of an upper watery layer. This phenomenon was quite evident in all the synthesized resins after storing them for about a month at room temperature. Careful pH control during the reaction of melamine with formaldehyde done in combination with repeated tests for solubility and viscosity were the methods used to control the molecular weight of the resins. Estimation of the extent of condensation stage of a given resin solution was carried out by diluting a 5-10 mL sample at room temperature with water. The number of mL of water that were necessary to add to reach permanent turbidity was a reasonable estimate of the extent of condensation. Figure 22 shows the trends in the water tolerance and percent hydrophobe levels during a MF resin cook. The same kind of trends were observed for all PMF resin cooks. A water tolerance level of approximately 200% and a hydrophobe level of approximately 23% gave a resin that was stable for approximately one month. After this time, separation of the resin into a watery layer and a solid layer was evident. All the resins produced at pH 9.0 were reddish brown in color and those synthesized at pH of 7.5 were light yellow. The reason for this color difference at different pH conditions may be due to the different reactivities of PF and MF. The co-condensation between phenol and formaldehyde takes place more readily at alkaline conditions whereas under neutral and slightly acidic conditions, co-condensation between melamine and formaldehyde takes place more readily. The reaction between phenol 51 and formaldehyde imparts a brownish color to the resin and the reaction between melamine and formaldehyde imparts a light color. Figure 22. Hydrophobe solids and water tolerance profile for the MF resin cook. 52 4.1.2. PF Resin During the synthesis of the PF resin, the purpose of adding NaOH in three separate portions was that the phenol remained ionized and thus soluble. Also, it is important that the methylation (addition) step of the synthesis occurs first over the condensation step. The cooling of the solution before addition of NaOH prevented the reaction from going out of control and thus prevented possible curing of the resin during the cook. The temperature and the pH were closely monitored during the cook to monitor the level of condensation (average molecular size). Control of the average molecular size of the finished resin is essential for the correct flow in plywood and other bonding operations while in the hot press prior to curing. Too low a level of condensation (i.e., low molecular weight resins) may give too much flow; the resin "runs away" from the wood or rapidly sinks into it under pressure, leaving "starved" glue lines. If a resin of too high a condensation stage (i.e., high molecular weight resins) is present, its flow under normal pressure and temperature may be too low to produce good results. The pH, viscosity and percent resin solids for this resin are given in Table 4. All these results are typical for a resol type of PF resin. Resol is a PF resin which is prepared under alkaline conditions and requires no further addition of hardners to cure to a final highly cross-linked structure. It only requires heat during the pressing operation to become fully cured. 53 4.2. Resin Properties 4.2.1. Fourier Transform Infrared Spectroscopy Fourier Transform Infrared Spectroscopy (FTIR) is a powerful tool for identifying types of chemical bonds and functional groups in a molecule by producing an infrared spectrum that is like a molecular "fingerprint". FTIR can be used to identify chemicals in spills, paints, polymers, coatings, drugs and contaminants. Molecular bonds, in these chemicals, vibrate at various frequencies depending on the elements present and the type of bonds. For any given bond, there are several specific frequencies at which it can vibrate. According to quantum mechanics, these frequencies correspond to the ground state (lowest frequency) and several excited states (higher frequencies). One way to cause the frequency of a molecular vibration to increase is to excite the bond by having it absorb light energy. For any given transition between two states, the light energy absorbed (determined by the wavelength) must exactly equal the difference in the energy between the two states (usually ground state (E0) and the first excited state (E,)). Therefore, the wavelength of light absorbed is characteristic of the chemical bond. 4.2.1.1 Synthesized and Commercial PF and MF resins The structural characteristics of PF and MF resins have been examined thoroughly using IR spectroscopy (Richard and Thompson (1946), Haslam and Willis (1965)). Since IR spectroscopy lends itself as more of a qualitative tool for determining the chemical structure of a compound, it is almost always used in conjunction with other techniques such as 1H-NMR and 1 3C-NMR spectroscopy. Figure 23 shows IR spectra for synthesized and commercial PF resins. 54 (a) Synthesized PF Resin / \ \ 1020 cml r • 1 \ J 1170 cmT A / A / / I 1 \ \ fJ v /» -780cm' l / / 1 /\ \ / 8 8 0 c m 1 J f / W-^-1220 cml I 1600 cml / 1260 crnl 3420 cml / 4000 3500 3000 2500 2000 1500 1000 cm-1 500 - i i I I 1 1 1 T -4000 3500 3000 2500 2000 1500 1000 cm"1 500 Figure 23. IR spectra for (a) synthesized and (b)commmercial PF resins. 55 The spectra of the PF resin show several very distinct peaks. The peaks at 1000 cm"1 and 1260 cm'1 correspond to the methylol groups which were not involved in the further polymerization of the resin and the peaks at 780 cm"1, 880 cm"1, 1475 cm"1 and 1600 cm"1 correspond to some of the phenolic ring peaks. Different substitution patterns on the benzene ring give rise to these particular peaks. The broad band at 3420 cm"1 is due to an O-H stretch from possibly the methylol groups and also the phenolic -OH group which may be present. The IR spectra of MF resins (Figure 24) also show several very distinct peaks. The peaks at 810 cm"1 and between 1350 - 1560 cm"1 correspond to the aromatic vibrations of the triazine ring. According to Braun and Legradic (1974b) the peak at 1170 cm"1 corresponds to the ether groups. This peak is indicative of the methylene ether bridges that form between the triazine rings during polymerization. The peak at 1000 cm"1 corresponds to the C-OH bond in hydroxymethyl groups. The broad peak between 3000 - 3500 cm"1 is due to the deformation vibrations of -NH 2 , >NH and -OH groups. According to Holmberg (1985), IR spectra of MF resin show an intense band at 1560 cm"1 corresponding to the in-place stretching vibration of the triazine ring. This peak usually superimposes a bond at 1650 cm"1 which is due to C=N stretching vibrations of terminal imino groups formed from -NHCH 2OR groups of the melamine resin. The commercial MF spectrum is almost identical to the synthesized resin whereas the commercial PF spectrum is quite a bit different from the synthesized PF resin showing cleaner lines and more pronounced peak structures. The methylol related peaks in the commercial PF spectrum are a bit smaller relative to 56 4000 3500 3000 2500 2000 1500 1000 C f f l H 500 Figure 24. IR spectra for (a) synthesized and (b) commercial MF resins.. 57 the phenolic ring peaks in the synthesized PF resin. According to Ellis (1989), this could be indicative of a higher molecular weight commercial resin. 4.2.1.2. PMF resins Figure 25 shows spectra of PMF resins prepared at pH 9.0 and Figure 26 shows spectra of PMF resins prepared at pH 7.5. All these spectra show the presence of peaks that were present in the individual PF and MF resin spectra as well as additional peaks possibly pertaining to the co-condensed resin. All the peak assignments could be made as in the last section. From Figure 25, it is very evident that as the proportion of phenol increases in the cooks, there is a small but distinct increase in the intensity of the peaks at 760 cm"1 and 1260 cm'1 and the peak at 590 cm"1 decreases in intensity. The 1600 cm"1 peak in the PF resin trace disappeared altogether. The peak that may pertain to the dimethylene ether bridges (1170 cm"1) is higher in intensity for Cook #2 which contains a greater proportion of melamine than Cook #1 and #3. Since Cook #1 contains more phenol with respect to melamine it seems that in a PMF resin, the triazine rings might be connected by dimethylene ether bridges whereas in Cook #2 where the phenol proportion is higher, may have lower proportion of dimethylene ether bridges. In Figure 26, most of the peaks between wavelengths of 1370 cm'1 - 1600 cm"1 pertain to different substitution patterns on both the phenolic and triazine nuclei. The peak at 1170 cm"1 is very distinct and the peak that was seen at 1260 cm"1 has almost disappeared in Cook #5, where the melamine level is lower than that of phenol. The peak that was seen for the MF resins at 590 cm'1 is very distinct in Cook #5 and other peaks which were not seen for the earlier spectra are present in the fingerprint region 58 ioo.oo-4 XT 4000 3500 3000 2500 2000 1500 1000 cm"1 500 0.00-) , , , 1 , , ) 4000 3500 3000 2500 2000 1500 1000 cm-* 500 Figure 25a,b. IR spectra for PMF resins prepared at pH 9.0. 59 loo.oo4 > i i 1 , , 4000 3500 3000 E500 3000 1500 1000 cm"' 500 Figure 25c. IR spectra for PMF resins prepared at pH 9.0. 60 o.ooH 1 , , 1 1 1 T -4000 3500 3000 2500 2000 1500 1000 C U T 1 500 Figure 26a.b. IR spectra for PMF resins prepared at pH 7.5. 61 (lower wavelengths). The peak at 690 cm"1 is very prominent in Cooks #5 and #6 and according to Raczniak (1983), this peak may be due to free phenol. In addition, the peak at 1370 cm'1 is very distinct and much larger than it was in Cooks #1 to #3. This peak possibly corresponds to different substitution patterns on both the phenolic and triazine nuclei. Care must be taken in interpreting spectra for resins at different pH conditions since reactions carried out at different pH's lead to different substitution patterns. It is very important to understand that all these data from the infrared spectroscopy only give us a qualitative indication of the structure of each resin and the question of whether there is any sign of copolymerization between the PF and the MF resins is still unanswered. 62 4.2.2 Proton Nuclear Magnetic Resonance Spectroscopy Proton Nuclear Magnetic Resonance (1H-NMR) spectroscopy has been the principal analytical tool in examining the resins immediately after synthesis (Bovey, 1972) . NMR methods appear to be satisfactory for studying branching and linkage types. This technique is also useful for characterizing the degree of substitution per monomer unit in polymer systems. The assignments of chemical shifts to specific functional groups for MF and PF resins have been carried out by numerous researchers including Gollob(1982), Chiavarini (1976) and Roh (1990). For PMF resin, the chemical shifts have been assigned to specific functional groups by only Roh (1990). Gollob (1982) made the following main chemical shift assignments for the acetylated PF resins. Chemical Shift (ppm) Functional Group 7.5 - 6.5 ArH 5.5 - 5.2 ArCH 2OCH 2OAc 5.6 - 4.8 ArCH 2OAc 4.8 - 4.5 ArCH 2OCH 2OAc + (ArCH2OCH2Ar) Roh (1990) assigned the following chemical shifts to the functional groups present in a MF resin. Chemical Shift (ppm) Functional Group 7.6-7.3 -NH 6.6-6.1 -NH 2, -CH 2OH 5.1 -CH 2 -0-CH 2 -4.7 -CH 2 -63 NMR spectra obtained in this study for both the synthesized (PF and MF) and for the commercial resins (PF and MF) are in agreement with these assignments (Figures 27 -30). The spectral assignments and the integrals are shown for all these figures. In order to determine which chemical shifts might represent the bridges in the final polymer, D 20 was added to the sample that had been already prepared for obtaining a NMR analysis (i.e., resin dissolved in DMSO). By using D 20, all the chemical shifts that are a result of exchangeable protons disappear and the chemical shifts that represent non-exchangeable protons (i.e. methylene and dimethylene ether bridges) remain. Figure 31 shows the NMR spectrum for the synthesized MF resin that was dissolved in DMSO. Figure 32 shows the NMR spectrum after the addition of D 20. The chemical shifts pertaining to both methylene and dimethylene ether bridges were present between 4-5 ppm in Figure 32. The other major peak present at approximately 3.8 ppm is the H 20 peak from the DMSO. The NMR spectra obtained for the acetylated MF and PMF resins were very difficult to analyze and therefore, a decision was made to only analyze the unacetylated MF and PMF resin spectra. For the PF resin, spectra for the acetylated resin was obtained and analyzed. Table 5 shows the relative proportion of the major functional groups present in all the PF, MF and PMF resins with respect to methylene bridges (note: for MF resins, these methylene bridges only pertain to the ones between the triazine rings) as well as their spectral assignments. For both the synthesized and the commercial PF resins the final polymer is dominated by methylene bridges. In both MF resins, there seems to be approximately seven times more methylene linkages than dimethylene ether linkages. All the PMF spectra are represented in Figures 33 - 37. Figure 33 shows the spectral assignments which are also true for all the subsequent 64 65 66 67 68 69 71 7 L— -4.1 CO o CD T— CO CO i CO o i o o o o o CO T ~ o o o T ~ < z I m CO CM LO LO CM LO O T— 1 1 o •<* O O O O o Ar-i z CN X o I o o CM CO CD O CO • 1 1 O d o O o Ar-CHS Ar-CHS < l CM X o 1 o 1 m CM CO co CD LO CD O CO • • o o o o o O O o o d Ar-i z CM X o o o o o O O O O o o o o o o 1 i o z 1 1 z 1 CM X o 1 o 1 LO O CD h- CM CD CD co T— i • LO O o O O o O o !H0-z 1 Chemical Shift (ppm) Resin Type PMF Cook #1 PMF Cook #2 PMF Cook #3 PMF Cook #5 PMF Cook #6 Synthesized MF Commercial MF Synthesized PF Commercial PF 76 spectra. It should be noted here that the peak assignments may shift to a slightly lower or higher chemical due to different chemical environments which are encountered in a different polymer system. From Table 5 it is evident that PMF resins cooked under different conditions (i.e., pH and molar ratio) show significant differences in the makeup of the final polymer molecules. Since co-condensation/co-polymerization could indirectly be represented by the amount of linkages (methylene and dimethylene ether bridges at 4.35 ppm and 4.40 ppm, respectively) between the phenolic and the triazine nuclei, PMF Cooks #5 and #6 show very little co-condensation between melamine and phenol whereas Cooks #1, #2 and #3 show a significant amount of co-condensation taking place between melamine and phenol. In Cook#1 the greater proportion of the linkages, both methylene and dimethylene ether, are between the phenolic nuclei. This makes sense since the molar ratio of phenol was twice as much as that of melamine. In Cooks #2 and #3, where the amount of melamine is equal to or greater than phenol, there are more methylene bridges between the triazine rings in the final polymer. Cooks #5 and #6 show greater amount of activity between the triazine groups than between just the phenolic groups or between the triazine and the phenolic groups. This indicates that the amount of co-condensation in these resins is much less than in Cooks #1 to #3. Nonetheless, it is important to understand that there is a certain level of co-condesation possible in all PMF resins synthesized in this study. This co-condensation might play a very important part in the bond performance of these resins. 77 4.2.3 Gel Permeation Chromatography Polymers need to have a degree of polymerization high enough to produce adequate mechanical properties; as it increases to 500 or above, useful cohesive strength, impact strength, etc., are developed (Haupt et al., 1991). Beyond this point, improvements are marginal. It should be noted that the synthetic processes for polymers produce a range or distribution of molecular weights, and all polymer molecules in the sample are not of the same size. The required solubility, viscosity of melts, and flow properties should determine the range of molecular weight (MW) chosen. Not only the average molecular weight, but the molecular weight distribution (MWD) of the polymer can have significant influence on adhesive performance. The presence of low molecular weight components can alter melt and flow properties. Furthermore, where porous adherends like.wood are involved, the deliberate addition of polymers of low molecular weight to the higher molecular weight adhesive has been found to enhance adhesion. Gel permeation chromatography (GPC) is a relatively fast technique for the determination of average MW and MWD of polymeric materials. In the present study, GPC was used for the determinations of MWD for all the resins. Prior to running the resin samples, a calibration plot was obtained for the columns by injecting polystyrene standards of known molecular weight (Figure 38). A typical detector output and the molecular weight distribution of a synthesized PF resin is shown in Figure 39. 78 6.00 5.00 4.00 3.00 2.00 1.00 Retention Time (min) Figure 38. GPC calibration curve using poystyrene standards. 79 Figure 39. (a) Molecular weight distribution and (b) GPC detector response for synthesized PF resin. 80 The molecular weight distributions for all the resins are shown in Figures 40,41,43 and 44. A summary of number average molecular weight (Mn), weight average molecular weight (Mw), and the polydispersity index (Mw I Mn) are given in Table 7. These are some of the MW values that are used to characterize a particular polymer. The values and/or changes occurring in specific MW averages and distributions of polymers often lead to better understanding of processing and end-use characteristics such as flow, cure, and strength properties (Haupt, 1991). Figure 40 shows GPC spectra of synthesized and commercial PF resin in their acetylated form. These spectra are very similar to those obtained by Ellis(1989). The synthesized PF resin contains greater proportion of higher molecular weight molecules whereas the commercial resin is composed of relatively equal proportions of lower and higher MW molecules. From these spectra, it is evident that the higher molecular weight species are not completely separated whereas the lower molecular weight methylol phenols are somewhat separated. The possible oligomers giving rise to the peaks below molecular weight of 1000 can be estimated as was done by Ellis (1989). The peaks present at a molecular weight of approximately 200 are possibly due to free phenol (unreacted) and mono substituted phenol. Peaks between 400 and 500 could be attributed to disubstituted and trisubstituted phenolic dimers. The peaks present at approximately 700 could be the result of trisubstituted and tetrasubstituted phenolic trimer or monosubstituted phenolic tetramer. The peaks around 1000 can possibly be attributed to pentasubstituted phenolic pentamers and trisubstituted phenolic hexamers. Figure 41 shows the GPC spectra for the synthesized and the commercial MF resins. Both these resins are of much lower molecular weight compared to the 81 1,000 10,000 Mol. Wt. 100,000 Figure 40. GPC spectra of synthesized and commercial PF resins. 82 10 20 50 100 200 500 Mol. Wt. 1,000 2,000 5,000 10,000 10 20 (b) Commerc Resin al MF 50 100 200 500 1,000 2.000 5,000 10,000 Mol. Wt. ure 41. G P C spectra of synthesized and commercial MF resins. 83 synthesized and commercial PF resins. For both MF synthesized and commercial resins, there are three very distinct peaks present at molecular weight of 130, 500 and 1000. The peak at 130 could possibly be attributed to unreacted melamine molecules (molecular weight 126). The other peaks are much more difficult to assign due to the many reaction combinations which are possible. Firstly, from the NMR data it was clear that the triazine nuclei could be connected by both, methylene and dimethylene ether linkages. Secondly, condensation products could be the result of mono- to hexa-methylol addition products. Finally, all the hydrogens on the triazine nuclei could have been acetylated during the acetylation procedure. According to Wohnsiedler (1952), the condensation leading to three dimensional polymers is a result of trimethylolmelamine addition products. The formation of a 3-dimensional polymer formed as a result of tetra-to hexamethylolmelamine addition product would undoubtedly impose severe steric strain upon the structure. It was further suggested by Blank (1979) that the main addition products formed by reacting melamine with formaldehyde were symmetric trimethylolmelamine (Figure 42a) and only small amounts of an asymmetric trimethylol-melamine compounds (Figure 42b). H x / C H 2 O H N H H N H H (a) (b) Figure 42. (a) Symmetric and (b) Asymmetric trimethylolomelamines. 84 Even if the assumption is made of the final polymer molecules being the result of trimethylolmelamine molecules, it is still very difficult to assign any one particular molecule to the peaks observed in Figure 41. It is clear that greater proportion of the polymers present are of low molecular weights. Comparing the MF spectra to the PF spectra, it is easy to realize that the MF resin possesses lower degree of polymerization. For the rest of the discussion, the peak at molecular weight of 500 will be referred to as the low molecular weight peak and the peak at 1000 as the high molecular weight peak. The GPC spectra profile for the synthesized and commercial MF resins are similar but the commercial MF resin contains a greater amount of the higher molecular weight fraction. Figure 43 shows GPC spectra for Cooks #1 to #3. These spectra have similar profiles to the synthesized MF resin spectra in Figure 41 with two exceptions. Firstly, the peak at 130 is missing and secondly, these spectra contain a slightly greater proportion of higher molecular weight molecules. The slight shift to lower molecular weight from Cook #1 to #3 could be due to the different degree of substitution on the triazine or phenolic nuclei. Figure 44 shows the spectra for Cooks #5 and #6. These spectra not only show the distinct peaks that were observed in Figures 41 (with the exception of the peak at 130) and 43, but show much higher molecular weights for both cooks. From Table 6, Mn, Mwand polydispersity values are higher for both of these cooks as compared with Cooks #1 to #3. The higher molecular weight molecules are probably not due to the co-condensation products between phenol and melamine but possibly due to individual PF and MF polymers. From the NMR data, a greater possibility for co-condensation existed in Cooks #1 to #3. Even though the pH conditions for Cooks #5 and #6 are not the best for MF and PF preparation, there is a 85 possibility that the higher molecular weight molecules are a result of PF polymers. It is clear from literature (Gollob, 1982) and also from Table 6 that the molecular weight of polymers formed during the PF cook is much higher than for the MF cook. 86 50 100 200 500 1,000 2,000 Mol. Wt. 5,000 10,000 Figure 43. G P C spectra of synthesized PMF resins prepared at pH 9.0. 87 50 100 200 500 1,000 2,000 5,000 10,000 Mol. Wt. 10 20 50 100 200 500 1,000 2,000 5,000 10,000 Mol. Wt. Figure 44. G P C spectra of synthesized PMF resins prepared at pH 7.5. 88 Table 6. Molecular Weights (Mn and Mw) and the polydispersity index for all Resins Resin Type Mn Mw Mw/Mn PMF Cook#1 637 801 1.26 PMF Cook #2 632 807 1.28 PMF Cook #3 642 866 1.35 PMF Cook #5 835 1325 1.59 PMF Cook #6 707 1115 1.58 Synthesized MF 259 606 2.34 Commercial MF 724 870 1.20 Synthesized PF 1549 25819 16.67 Commercial PF 1194 18458 15.46 89 4.2.4. Differential Scanning Calorimetry The IR, NMR and GPC data give information about the chemical structure of the resins at the preparation stage. The DSC traces give information about the cross-linking reactions which occur when the resin structures formed initially are subjected to a specific heating rate program. A full analysis of the complex reactions occurring during cure is not possible, but a broad general interpretation of the DSC data can be proposed in the light of the information derived from IR, NMR and GPC. Thermal cross-linking of all resins is undoubtedly complex. At normal curing temperatures two main types of reactions predominate. A methylol group may condense with an active hydrogen atom on a neighboring phenolic or triazine nucleus, with the elimination of water and the formation of a methylene cross-link. Self-condensation of methylol groups may also occur, with the elimination of water and the formation of a methylene ether cross-link, or with elimination of both water and formaldehyde and formation of a methylene bridge. At higher temperatures, other types of reaction can occur (Martin, 1956). DSC has been applied extensively to investigate the curing characteristics of PF resins including the degree of cure, temperature, duration of the curing reaction and the magnitude and variability of the heat of reaction as a function of time and temperature (Era and Matilla (1976), Westwood (1971), King et al. (1974), Chow and Steiner (1979)). It is well known that the exothermic cure reaction and the endothermic vaporization of water in the system are competitive processes in the 100 - 200°C range. Thus the net energy change detected by DSC in this region does not yield significant information. By raising the pressure to 1.38 Mpa (200 psi), the water vaporization is 90 shifted out of the phenolic-cure region and a one or two-stage curing exotherm is observed. This curing exotherm is a curve (Figure 45) which may show one or two overlapping peaks, probably owing to the formation of methylene and/or methylene ether bridges. The area under this exotherm is assumed to be proportional to the heat of reaction. In the present study, all the samples were run under pressure, in a small sealed aluminum sample pan which had a small hole punched in the top to equalize pressure. The pressurization insured the elimination of all the effects due to sublimation and evaporation, and insured equilibrium conditions. The DSC analysis was carried out to see: (1) the difference in the thermal behavior of each resin and; (2) the thermal behavior of each resin during the different advancement stages of resin synthesis. The latter was carried out by taking samples from each resin, during synthesis, at arbitrary intervals and then running a DSC spectra for each of these samples. The results for this analysis are summarized graphically in Figures 46 - 48. Figures 49 - 52 show the DSC spectra for the final resin products. From Figures 46 - 48, samples taken for DSC analysis at later stages during individual cooks show a lower heat of reaction (exothermic heat) with time. This makes sense, since there are more free formaldehyde and more additional compounds (i.e., hydroxy-methyl groups) available for further condensation reactions at early stages of resin synthesis as opposed to later stages. This indicates that a highly polymerized resin will yield lower heat of reaction. According to Chow (1972b), less advanced resins (lower molecular weight) show a greater amount of heat evolved due to greater amount of unreacted formaldehyde and more unreacted hydroxymethyl groups in the system. The above deductions could only be made for one specific type of resin and are not necessarily true between different types of resins. For example, the heat of reaction for 91 0 . 5 1 4 6 . 8 5 ° C Temperature (°C) General V4.1C DuPont 2000 Figure 45. Typical DSC spectra for PF resins showing (a) a single peak and (b) two overlapping peaks. 92 Figure 46. The peak temperatures and the amount of heat evolved for (a) MF and (b) PF resins during different advancement stages of synthesis. 93 Figure 47a,b. The peak temperatures and the amount of heat evolved for Cook #1 and Cook #2 during different advancement stages of synthesis. 94 0 50 100 150 200 Reaction Time (min.) Figure 47c. The peak temperatures and the amount of heat evolved for Cook #3 during different advancement stages of synthesis. 95 100 - i 250 CO o X U l 100 Reaction Time (min.) 4- 100 250 20 J 50 100 150 Reaction Time (min.) 200 g 6. E 0) H 150 ,? 100 200 250 -•—Exothermic Heat peak #1 (J/g) - ® — Peak Temp. #1 (C) •Exothermic Heat peak #2 (J/g) • Peak Temp. #2 (t ) Figure 48~,b. The peak temperatures and the amount of heat evolved for Cook #5 and Cook #6 during different advancement stages of synthesis. 96 each of the samples for the MF resin (Figure 46a) is much lower than the PF resin samples (Figure 46b) even though the molecular weight of the MF resin was much lower than the PF resin as shown by the GPC data. In contrast, as mentioned earlier, Chow (1972) found that the less polymerized resin yielded higher heat of reaction and a highly polymerized resin yielded lower heat of reaction. In addition to the lower heat of reaction for samples taken at later stages of each cook, the peak temperatures are also lower. Therefore, it is possible that as the resin becomes more cross-linked, the temperature required to cure the resin is lowered. Figure 48 shows the results obtained for Cooks #5 and #6. For both of these cooks, at a certain stage of reaction, two DSC exotherms were evident. The presence of the second exotherm is a clear indication that two separate reactions, or closely related sets of reactions, occur in each of these resins. Figure 49 shows the DSC spectra for a synthesized and commercial PF resin. The peak temperatures for both resins are similar but the heat of reaction for the commercial resin is higher. This indicates that the synthesized resin is much more advanced than the commercial resin and thus should be of higher molecular weight. This observation is consistent with Table 5 which shows the molecular weight values for all the resins. Both resins show overlapping peaks with the commercial resins showing a much more distinct overlap. Figure 50 shows the DSC spectra for synthesized and commercial MF resins. The synthesized resin shows only one exotherm whereas the commercial resin shows two exotherms. The first peak temperature and the heat of reaction for both resins are very similar. The commercial resin may contain additives which might account for the second exotherm. These two exotherm phenomena were 97 also true for DSC spectra obtained for Cooks #5 and #6 in Figure 52. The first peak temperature for Cook #5 is lower than that of Cook #6 with their heat of reaction being very similar. The second exotherm is still observed above 200°C. A possible explanation could be derived from the NMR data in Table 5. Cooks #5 and #6 possess greater amount of dimethylene ether groups in comparison with Cooks #1 - #3 and as do both synthesized and commercial MF resins. It is possible that some complex reactions occurred which involved these dimethylene ether bridges and these reactions are probably responsible for the second exotherm. As to the type of complex reactions, it is very difficult, if not impossible, to guess which reactions occur. King (1974) postulated that any DSC peaks that occur at temperatures above 200°C for PF resins correspond to degradation reactions and include a contribution from the widely accepted reaction whereby dibenzylether linkages eliminate formaldehyde to form methylene linkages. Kurachenkov and Igonin (1971) observed exothermic peaks at 220°C for various parasubstituted polybenzyl ethers. They quoted infrared evidence which supported the identification of the peak as arising from the elimination of formaldehyde from ether linkages to form methylene crosslinks. Figure 51 shows the DSC spectra for Cooks #1 to #3. The peak temperature for all these cooks is quite similar with the amount of heat evolved being noticeably greater for Cook #1 than Cooks #2 and #3. Some researchers believe that during the hot pressing operations, when MF or MUF resin is used, a large amount of formaldehyde is released. The DSC data indicates that this can only be possible if very high temperatures are utilized during the pressing operations. 98 0.0 Temperature (*C) General V4.1C OuPont 2000 Figure 49. DSC spectra for (a) synthesized PF and (b) commercial PF resins. 99 (b) OH 2 -5H a x 123.9-;* 6 5 . 6 9 J / g 226. 7 I T -10H 303.26* 42 .92J /g -15- —r-50 100 150 Temperature (*C) 200 250 General V4.1C OuPont 2000 Figure 50. DSC spectra for (a) synthesized MF and (b) commercial MF resins. 100 0 . 0 -0.5-4 ^ - 1 . 0 X -2.0-j -2.5- — r -50 151.08°C 100 150 Temperature (°C) (a) 200 250 General V4.1C OuPont 2000 0.0 - 0 . 5 -- 1 . 0 --1.5H -2.0H -2.5-(b) 147.53°C 119.38"CT\ \ 109.6J/Q \ 50 100 150 Temperature ("C) 200 250 Figure 51a,b. DSC spectra for (a) cook #1 and (b) cook #2 resins. 101 0.5 Temperature CC) General V4.1C DuPont 2000 Figure 51c. DSC spectrum for cook #3 resin. 102 (a) 128.42-0 107.62°C ~ 4i.63J/g ~~^f-^ 217.36'C 206.31/%\\ —. . * 1- —i 1 19.47J/Q ^ 1 50 100 150 Temperature (°C) 200 250 General V4.1C DuPont 2000 (b) ^ V ^ ^ 141.6B°C 211.47°C 115.27°C ""—-—-__\. 46.34J/Q 198.52*C ~-\ 30.51J/g \ 50 100 150 Temperature (°C) 200 250 General V4.1C DuPont 2000 Figure 52a,b. DSC spectra for (a) cook #5 and (b) cook #6 resins. 103 4.3 Wood Bonding Results An attempt was first made to use the neat resins for bond analysis but it was found, through trial and error, that all the resins, except PF resin, over-penetrated the veneer and left very little resin in the glueline for creating any adhesive strength in the final panel. Thus, a decision was made to add 14% wheat flour (Rogers #2) to all the resins except the PF resin. In industry, fillers and extenders are added to plywood glues to help control resin mix viscosity, moisture content, reduce glueline failure due to dryout or blows and to conserve resins (Gollob, 1983). The reason why neat resins were preferred for bonding was that any additives in the resin mix may mask the intrinsic properties of the resins. Figures 53 and 54 show the pressure and temperature profiles for panels pressed at 1.38 MPa (200 psi) and temperatures of 120°C and 150°C, consecutively. For panels that were pressed at temperature of 120°C, the core temperature of the panel reached approximately 116°C for both the 3 and 5 minute press times. For panels that were pressed at temperature of 150°C, the core panel temperature reached about 125°C for the 3 minute press times and 140°C for the 5 minute press times. Bond performance of all the resins was evaluated using both the wood failure and the shear strength values. The average values for both the shear strength and wood failure results are shown Table A-1 and Table A-2 in Appendix I. Since some of the data sets were non-normal, all the data were transformed using the arcsin transformation. The experimental array was a 8 x 2 x 2 (8 resins, 2 press times and 2 press temperatures) factorial design. The analysis of variance (ANOVA) was performed 104 Figure 53. Pressure and temperature profiles for the (a) 3 minute and (b) 5 minute press times for panels pressed at 120°C. 105 0 250 200 3 150 (!) i _ 3 (0 $ 100 0. 50 0 / Pressure (psi) — Core Temperature f*C) 50 100 150 200 Press Time (min.) 250 (a) 140 + 120 100 ? CD 80 | cu a. 60 E CO 40 20 0 cu o o 300 (b) > f Pressure (psi) Core Temperature (°C) 160 140 120 p cu 100 ^ ro 80 o Q. 60 40 20 0 E cu H o o 0 50 100 150 200 250 Press Time (min.) 300 350 400 Figure 54. Pressure and temperature profiles for the (a) 3 minute and (b) 5 minute press times for panels pressed at 150°C. 106 on each of the three parameters; (1) resin (2) press time (3) press temperature. The analysis was carried out, on the transformed data, using a statistical software package called STATISTICA. Tables 7-12 show the analysis of variance for shear strength and wood failure values for dry, wet and boiled specimens, respectively. Many significant effects between resin, time and temperature were observed for both the wood failure and shear strength measurements for the three different treatments that the samples were subjected to prior to testing. The interactions were further examined by using a multiple range test called the Scheffe multiple range test. For further analysis the results will be considered separately between each different treatments to which the samples were subjected. 4.3.1 Dry specimens. Tables 7 and 8 show the ANOVA results for the shear strength and wood failure for dry specimens. It was observed that the two-way interactions (resin x time and time x temperature) and all the main effects (resin, time and temperature) were significant at the 95% confidence level. For the wood failure results, all the main effects were significant as well as the two-way interaction (resin x temperature) at 95% confidence levels. Since some of the interactions are significant, it is necessary to interpret the results by looking at the interactions rather than the. individual main effects. However, for both the shear strength and the wood failure values, the most significant main effect seems to be the temperature from comparing the F-values. This is quite reasonable since the curing of the resins depends strongly on the temperature. Faster curing of the resin is accomplished by a higher temperature. Figure 55 shows the two-way 107 Table 7. Analysis of variance of shear strengths for dry specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 14369.76 9.46795 0,000003 Time* 1 24438.19 17.80117 0.000188 Temperature* 1 39532.66 24.81130 0.000021 Resin x Time* 7 3651.36 2.65971 0.027481 Resin x Temperature 7 3652.41 2.29231 0.051803 Time x Temperature* 1 6150.40 6.61813 0.014938 Resin x Time x Temperature 7 786.95 0.84680 0.557635 Error 126 1517.73 * - significantly different at 95% confidence level. Table 8. Analysis of variance of wood failure for dry specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 9444.06 34.31290 0.000000 Time* 1 2522.62 8.00310 0.007996 Temperature* 1 34822.10 195.56430 0.000000 Resin x Time 7 241.02 0.76470 0.620658 Resin x Temperature* 7 2354.34 13.22220 0.000000 Time x Temperature 1 457.35 1.99540 0.167430 Resin x Time x Temperature 7 80.06 0.34930 0.924425 Error 126 275.23 * - significantly different at 95% confidence level. 108 Table 9. Analysis of variance of shear strengths for wet specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 26143.69 30.55019 0.000000 Time* 1 7706.45 4.87985 0.034451 Temperature* 1 7144.66 6.63911 0.147930 Resin x Time* 7 4013.19 2.54121 0.336880 Resin x Temperature* 7 3944.64 3.66552 0.005134 Time x Temperature 1 2178.43 1.58848 0.216657 Resin x Time x Temperature 7 2255.32 1.64455 0.158733 Error 126 855.76 * - significantly different at 95% confidence level. Table 10. Analysis of variance of wood failure for wet specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 21127.05 128.39310 0.000000 Time* 1 1162.14 7.17880 0.011554 Temperature* 1 10236.32 53.76330 0.000000 Resin x Time* 7 616.58 3.80880 0.040790 Resin x Temperature* 7 1101.44 5.78500 0.000217 Time x Temperature 1 41.20 0.27740 0.602068 Resin x Time x Temperature 7 315.73 2.12550 0.069190 Error 126 164.55 * - significantly different at 95% confidence level. 109 Table 11. Analysis of variance of shear strengths for boiled specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 34661.66 34.93578 0.000000 Time* 1 22695.93 33.87894 0.000002 Temperature* 1 11520.42 10.22304 0.003118 Resin x Time* 7 4499.08 6.71591 0.000063 Resin x Temperature 7 2427.10 2.15377 0.065881 Time x Temperature 1 325.44 0.48261 0.492256 Resin x Time x Temperature* 7 3046.86 4.51832 0.001351 Error 126 992.15 * - significantly different at 95% confidence level. Table 12. Analysis of variance of wood failure for boiled specimens. Source of Variation DF Mean Square F-Ratio P-level Resin* 7 11232.02 45.70282 0.000000 Time* 1 2522.06 11.22559 0.002081 Temperature* 1 8774.04 42.70487 0.000000 Resin x Time 7 241.33 1.07414 0.402287 Resin x Temperature* 7 1124.66 5.47394 0.000335 Time x Temperature 1 237.56 1.18588 0.284296 Resin x Time x Temperature* 7 1061.12 5.29703 0.000430 Error 126 245.76 * - significantly different at 95% confidence level. no Plot of Means 2-way interaction F(7,32)=2.66; p<0275 280 ,_ 220 ro > ro level 1 level 2 RESIN CommMF RESIN LabMF RESIN LabPF RESIN Cook#5 RESIN Cook#6 RESIN Cook#2 RESIN Cook#3 RESIN Cook#1 TIME Plot of Means 2-way interaction F(1,32)=6.62; p<.0149 ] , , - ' ' level 1 level 2 TIME leveM TIME level 2 TEMP Figure 55. Two-way interactions (resin x time and time x temperature) for shear strength of dry specimens. in interaction results (resin x time and time x temperature) for shear strengths. The resin and time interactions show that wood shear strength values for all the resins increased with increase in temperature with the exception of Cooks #2 and #6 which show similar values at both temperatures (120°C and 150°C) or a slight decrease with increase in temperature. The large increase in shear strength from 3 minute press time to 5 minute press time was observed for Cook #1, #3 and the PF resin from the slope of the lines between the two times. From the (time x temperature) interaction graph, it is clear that the higher shear strength values were observed for panels pressed at 150 °C and also higher values for 5 minute press time. From the slope of the lines between the two temperatures, a significant increase in shear strength values was observed for panels pressed at higher temperature (150°C) for the 3 minute press time. Figure 56 shows the two-way interaction (resin x temperature) for the wood failure results for the dry specimens. Cooks #5 and #6 show the best results in terms of wood failure and also there is no increase observed between panels pressed at 120 °C and 150 °C. Large increases in wood failure results are observed for Cooks #1, #2, #3 and the PF resin when panels are pressed at the higher temperature as opposed to the lower temperature. It is interesting to note that the large increase in wood failure at higher temperature is observed for resins that performed poorly at the lower temperature. This indicates that the resins might not have been fully cured at the lower temperatures. 112 Plot of Means 2-way interaction F(7,32)=13.22; p<.0000 level 1 level 2 RESIN CommMF RESIN LabMF RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 TEMP Figure 56. Two-way interactions (resin x temperature) for wood failure of dry specimens. 113 4.3.2. Soaked specimens. The ANOVA of shear strength and wood failure results for these specimens are shown in Tables 9 and 10, respectively. The two-way interactions (resin x time and resin x temperature) as well as the main effects (resin, time and temperature) were significant at 95% confidence level. Figures 57 and 58 show these interactions graphically for shear strength and wood failure results, respectively. From the resin x time interactions in Figure 57, Cooks #3, #5 and #6 show decreases in shear strengths for panels pressed at 150°C. All other resins show an increase in the shear strengths for panels pressed at this temperature. Again, there is a large increase in shear strength observed at higher temperatures with resins that performed poorly at lower temperatures. These resins include the Cook#1 and the PF resins. From the resin x temperature interaction graph in Figure 57, panels pressed with the commercial MF resin, and Cook #6 decreased in shear strength when pressed at higher temperature (150°C). The resins that showed significantly higher shear strength values include synthesized MF resin, Cook #3 and Cook #6. Figure 58 shows the two-way interactions (resin x time and resin x temperature) for the wood failure results for the soaked specimens. From the resin x time interaction, the wood failure results for PF resin and Cooks #1 and #3 are significantly different from the rest of the resins. All resins show an increase in the wood failure results with panels pressed at 150°C with the exception of Cook #2. The resins that show the best results include commercial MF, synthesized MF, Cooks #2, #5 and #6. From the resin x temperaiure interaction, wood failure results for PF resin, Cook#1 and #3,again, show statistically significant differences from the rest of the resins. The commercial MF, 114 280 260 240 220 200 180 160 140 120 100 80 Plot of Means 2-way interaction F(7,32)=2.54; p<0337 | k- I _ _ ! " " " " " " " " " " " * — y ^ r - r - - , - ^ - - - - - - - - - -— i ^ ____ Za. ... - " \ L level 1 level 2 RESIN CommMF RESIN LabMF RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 TIME Plot of Means 2-way interaction F(7,32)=3.67; p<.0051 280 260 240 220 200 180 160 140 120 100 ] ; :: z. •- — - — - 1 ' 1 — ~~ " tf-level 1 level 2 - o - RESIN CommMF -•a- RESIN LabMF -o- RESIN LabPF - - A - RESIN Cook#1 - * - RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 - A - RESIN Cook#6 TEMP igure 57. Two-way interactions (resin x time and resin x temperature) for shear strength of wet specimens. 115 Plot of Means 2-way interaction F(7,32)=3.81; p<.0041 j = _ — r r f - - - i i H~=.... r - ~ ~~ ! — — ! 1 i -A level 1 level 2 -o- RESIN CommMF a - RESIN LabMF -o- RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 TIME Plot of Means 2-way interaction F(7,32)=5.79; p<0002 i; < > 5. _ _ +— ~ .... ^<^t>. _._ - - - _ - - —- - - -• - - ~- -^-^j, - • _ — — — — ~- ~~ ~ ~ ' A — — — _ _ _ | level 1 level 2 o - RESIN CommMF a - RESIN LabMF •o- RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 TEMP Figure 58. Two-way interactions (resin x time and resin x temperature) for wood failure of wet specimens. 116 synthesized MF and Cook #5 performed much better under wet conditions followed by Cooks #2 and #6. 4.3.3 Boiled specimens. Tables 11 and 12 show the ANOVA's for the shear strength and wood failure results for boiled specimens. From Table 11, the two-way interaction (resin x time) and three-way interaction in the shear strengths are statistically significant. Figures 59 and 60 show these interactions graphically. The two-way interaction (resin x time) shows that the shear strength results for Cook #1 are significantly different from rest of the resins. Cook #2 and the PF resin show a very large increase in the shear strength for samples pressed at temperature of 150°C. At pressing temperatures of 120°C, Cooks #5, #6 and synthesized MF show higher shear strength values and at 150°C, synthesized MF, PF, Cook #2 and #5 show similarly high values. From the three-way interaction (resin x time x temperature) in Figure 59, the shear strengths for panels prepared with Cook #1 resin show significantly different results at both times and at both temperatures than rest of the resins. For the PF resin, the largest increase in shear strength was observed for panels pressed for 5 minutes at 120°C. An increase was also observed for such panels pressed at 150°C, but it was not statistically significant. For panels pressed at 120°C, Cooks #3, #5 and #6 show a decrease in shear strength for panels pressed for 5 minutes. For panels pressed at 150°C for 5 minutes, commercial MF, synthesized MF and Cooks #5 and #6 show a decrease in the shear strengths. Figure 60 shows the two-way interaction (resin x temperature) and three-way interaction (resin x time x temperature) for the wood failure of the panels. From two-way 117 Plot of Means 2-way interaction F(7,32)=6.72; p<0001 _ _.. . 0 -Ii • " " " 1 1 • - • _ . ^ . ^ •• - - • i 1 _ J± . - - • - "1 level 1 level 2 RESIN CommMF RESIN LabMF RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 TIME Plot of Means 3-way interaction F(7,32)=4.52; p<.0014 [ L T - - - . . - f i f ^ T ^ : i i 1 -6 ; ^ i '•• ^ \ \ ^ L c j TIME leveM level_2 TEMP level 1 TIME leveM level_2 TEMP level 2 - o - RESIN CommMF ~o- RESIN LabMF - o - RESIN LabPF RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 ~ * - RESIN Cook#6 ure 59. Two-way interactions (resin x time) and three-way interactions (resin time x temperature) for shear strength of boiled specimens. 118 Plot of Means 2-way interaction F(7,32)=5.47; p<.0003 level 1 level 2 - o - RESIN CommMF -•a- RESIN LabMF - o - RESIN LabPF RESIN Cook#1 - • - RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 - * - RESIN Cook#6 TEMP 100 80 60 ra > o 40 ra > 20 -20 Plot of Means 3-way interaction F(7,32)=5.30; p<0004 I. TIME level_1 level_2 TEMP level 1 TIME leveM level_2 TEMP level 2 -o- RESIN CommMF • • D - RESIN LabMF - o - RESIN LabPF • A - RESIN Cook#1 RESIN Cook#2 RESIN Cook#3 RESIN Cook#5 RESIN Cook#6 Figure 60. Two-way interactions (resin x temp) and three-way interactions (resin x time x temperature) for wood failure of boiled specimens. 119 interaction, it is clear that the wood failure values for panels made with Cook #1, #3 and Cook #6 resins show little differences for both temperatures. The wood failure values for panels made with commercial MF, synthesized MF, synthesized PF, Cook #2 and Cook #5 show statistically significant differences when different temperatures were used for pressing. The three-way interactions show that the wood failure of panels pressed with PF resin increased greatly at 120°C when the pressing time was increased from 3 to 5 minutes. At 150°C, wood failure for PF resin actually decreased when pressed for 5 minutes. Cooks #1, #2, and #3 performed very poorly in terms of wood failure at 120°C at both 3 and 5 minute press times. A similar performance was, again, observed for these resins at 150°C for both press times. For panels pressed at 150°C a significant increase in wood failure was observed for Cook #5 when pressed for 5 minutes. 4.3.4 Summary of Wood Bonding It will perhaps be easier to see the performance of all the resins if the main effects of resins from the ANOVA table are graphed (Figure A1). The non-transformed shear strength and wood failure results are shown in Table A1 and Table A2 and graphically depicted in Figures A2 - A5. From Figure A1 it is very clear that for both the shear strength and wood failure results, the type of condition that the specimen is tested under has a great effect on the relative bond performance. The samples that were tested dry showed the highest results for all resins, soaked samples showed lower performance and the boiled samples showed even poorer results. This is mainly due to the fact that, in general, the wood when exposed to more severe conditions, such as boiling, will weaken and thus result in the increase of wood failure values and a 120 decrease in the shear strength values. This was the case for all the resins synthesized in this study with the exception of lab PF resin which showed higher WF values as the severity of the conditions increased. From observing the non-transformed data in Tables A1 and A2, the resins that performed best under dry conditions include in terms of shear strength Cook #2, #5, #6 and both the MF resins. Best wood failure results under dry conditions were achieved by Cook #5, #6 and both MF resins. Under soaked conditions, all the resins exhibit similar shear strengths with the exception of Cook #1 and synthesized PF resin which had lower shear strength values. Wood failure results under soaked conditions are greater for Cook #5, #6 and both the MF resins. Under boiling conditions, Cook #5, #6 and both MF resins show better shear strengths whereas wood failure values are higher only for the synthesized MF resin. The synthesized PF resin shows higher wood failure for boiled samples. This is a possible indication that the resin was under-cured during the pressing cycle and during the boiling treatment the resin further cured to a higher level and thus yielded higher wood failure. When the wood failure results for the soaked samples are compared, the above explanation seems to be an unlikely scenario. If the under-cured gluelines were a problem, the specimens should perform poorly after vacuum-pressure treatment. This was not the case with specimens pressed at 150°C. The wood failure values are higher for these soaked specimens than the boiled specimens. Nonetheless, this result is true for the specimens pressed at 120°C for 5 minutes and specimens pressed at 150°C for 3 minutes. The lower wood failure values for the PF resin might be attributed to its higher molecular weight, calculated from the GPC analysis. It is well known that resins of very high molecular weight might be unable to penetrate the veneer and thus result in poor bonding. Typically, 121 commercial PF resins perform much better than the PF synthesized in this study. The average wood failure values for a commercial PF resins, which is fully cured, might be in the excess of 80% under all treatment conditions. If the bond performance is examined by just considering the wood failure values , it is very clear that Cook #5, #6 and both the synthesized and commercial MF resins perform much better than the PF resin at lower temperatures and shorter times. Cook #2 performs better than Cook #1 and #3 at both press temperatures and both press times. Cook #2 contains a greater proportion of melamine with respect to Cook #1 and #3. Cook #5 , which contains a greater proportion of melamine , exhibits better bond performance than Cook #6. Therefore, it is reasonable to assume that lower pH and greater proportion of melamine in the PMF resins results in better bond performance. It is also true that the melamine resins have a greater capacity to penetrate and swell the wood than the phenolics and this phenomenon might have something to do with their better bond perfromance. Surprisingly, MF resin which is historically known to be less durable than the PF resin, shows the highest durability under almost all treatment conditions. 122 5. SUMMARY. Successful syntheses of PMF resins at various pH conditions and molar ratios were carried out. In addition, MF and PF resins were also synthesized. Both, the commercial MF and the commercial PF resins were obtained from industry and were used as comparison with the synthesized resins. Various analytical techniques were utilized in order to define the chemical characteristics of each resin. A possible relationship between the chemical structure of each resin and bond performance was carried out by making 3-ply plywood panels and testing these panels for shear strength and wood failure. The PMF resins synthesized at pH 7.5 were of light color and these resins resulted in a bright colored glueline after bonding. This aspect of a light colored structural adhesive is advantageous since it could possilby result in a large increase in the market value of composites bonded with it. From the IR and NMR analysis, there was strong evidence that co-condensation between the phenolic and triazine nuclei did take place in the PMF resins. PMF resins cooked at pH 7.5 showed more activity between the triazine nuclei as opposed to the phenolic activity or the activity between the phenolic and triazine nuclei. GPC analysis showed that the PMF resins cooked at pH 7.5 had a higher molecular weight than those cooked at pH 9. This was possibly responsible for their better bond performance characteristics. The molecular weights for these PMF resins, as well as the MF resins, was much lower than that of the PF resins. The DSC analysis showed that PMF resins cooked at pH 7.5 exhibited two exotherms, one at approximately 150°C and the other at about 210°C. The first exotherm is attributed to condensation reactions whereby only water is released as one of the byproducts. The second exotherm is possibly due to elimination of formaldehyde from 123 dimethylene ether linkages to form methylene cross-links. This particular behavior was observed for the commercial MF resin, as well. One of the drawbacks of melamine resins is believed to be the excessive amount of formaldehyde released during pressing operations. From the DSC data, it is clear that formaldehyde release will only take place at temperatures above 210°C. The usual press temperatures for plywood manufacture do not exceed 150°C. The bond performance analysis revealed that the PMF resins cooked at pH 7.5 and both the MF resins performed much better than PMF resins cooked at pH 9.0 and the PF resin. The deficiency in the performance of PF resin was attributed to its very high molecular weight which possibly resulted in its poor penetration and wetting during the pressing cycle. Nonetheless, the pH 7.5 PMF resins showed excellent wood failure results under dry and soaked conditions for both press temperatures and both pressing times. Very good wood failure results for these resins, pressed at 150°C and for 5 minutes press time, were also obtained after the boiling treatment. This finding confirms the earlier hypothesis that melamine addition to PF resin systems results in their better performance at lower press temperatures and lower press times. This aspect of any resin will ultimately result in very significant energy savings. For the PMF resins which were cooked at pH 9.0, the one containing a higher molar ratio of melamine showed higher wood failure results. Therefore, there is a definite correlation between the pH and bond performance and also between the amount of melamine used relative to phenol in PMF synthesis. It is still generally believed that in PMF resins the presence of phenol will provide durability to the resin whereas melamine will impart a lower temperature of cure (Higuchi, 1990). Therefore, 124 under pH 7.5 conditions, there is an adequate amount of co-condensation between phenolic and triazine nuclei to give a resin with reasonable bond performance. At the present time many light colored adhesives are available for non-structural bonding but none for structural bonding. The appearance factor will make it advantageous in the future to have some structural composites made with light colored adhesives. Future work should consist of determining which pH is optimum for best bond performance, not necessarily for maximum co-condensation, of PMF resin. In light of the results obtained in the present study, the pH during synthesis of these resins should be very close to pH 7.5. An attempt to utilize weak acid catalysts should be made to further reduce curing temperatures and cure times. Acids are known to be very detrimental to the wood properties but currently various acid catalysts are used very successfully for UF resin curing. 125 6. CONCLUSIONS Several characteristics of PMF resins as they relate to the synthesis, chemical characterization and also the performance were identified. • The PMF resins prepared at lower pH conditions (pH 7.5) were lighter in color (light yellow) and when cured resulted in a very light glueline. • From the IR and NMR data, it was evident that the PMF resins prepared at pH 7.5 and 9.0 show co-condensation between the phenolic and triazine nuclei. Greater level of co-condensation existed in the resins prepared at pH 9.0; • GPC data showed that the resins prepared at pH 7.5 had higher molecular weights than the ones prepared at pH 9.0. • DSC data showed that the PMF resins prepared at pH 7.5 developed two exotherms, one at about 150°C and the other at 215°C whereas the ones prepared at pH 9.0 showed only one exotherm. • DSC data also showed that greater amount of heat was given off by Cooks #1 to #3 than Cooks #5 and #6. The amount of heat given off is determined by the number of reactive sites which are available during the curing reactions and do not necessarily indicate the level of bond performance. • Resins cooked at pH 7.5 gave better bond performance when bonded at lower press times and lower temperatures than the ones cooked at pH 9.0. They were also better than the control (commercial) PF resin. In particular, the resin that contained a greater proportion of melamine as starting reactant also showed significant durability. • Surprisingly, the control MF resins (commercial and synthesized) showed the best results in terms of bond performance and durability. 126 7. LITERATURE CITED American Society for Testing and Materials. 1973. "Standard Method of Testing for Strength Properties of Adhesive in Plywood Type Construction in Shear by Tension Loading". ASTM Designation: D906-64 (Reapproved 1970). Annual Book of ASTM Standards. Part 16. P. 255-258. American Society for Testing and Materials. 1973. "Standard Method of Evaluating the Properties of Wood-Base Fiber and Particle Panel Materials". ASTM Designation: D1037-72a. Annual Book of ASTM Standards. Part 16. P. 311-352. Anderson, D.C., D.A. Netzel and D.J. Tessari. 1970. "The Synthesis and Characterization of Monomeric Etherified Methylolated Melamines Using Gel Permeation Chromatography and Proton-Neutron Magnetic Resonance Spectroscopy". J . Appl. Polym. Sci.14: 3021-3032. Anonymous. 1974. U. S. Product Standard (PSI-74) construction and Industrial Plywood. National Bureau of Standards, Washington, D. C. August. 32 pp. Blank, W.J. 1979. "Reaction Mechanism of Melamine Resin." Journal of Coatings Tech. 51:61. Blomquist, R. F. and W. Z. Olson. 1955. "Durability of Fortified Urea-Resin Glues in Plywood Joints". For. Prod. J., 1: 50-56. . Blomquist, R. F. and W. Z. Olson. 1964. "Durability of Fortified Urea-Resin Glues Exposed to Exterior Weathering". Forest Products Journal, 13 (6):461. Bovey, F. A. 1972. "High Resolution NMR of Macromolecules." Academic Press. N.Y. 462pp. Bornstein, L. F. 1985. "Melamine Modified Phenolic Type Resin for Continuous Lamination." U. S. Patent No. 4,611,021. Braun, D., and V. Legradic. 1974a. "Characterization of Methylolmelamine Using Gel Permeation Chromatography." Angew. Makromol. Chem. 35: 101-114. Braun, D., V. Legradic. 1974b. "Investigation of the Melamine Formadehyde Reaction Products." Angew. Makromol. Chem. 36: 41-55. Braun, D., H-J. Ritzert. 1984. "Investigation of the Reaction of Phenol and Melamine with Formaldehyde under Basic Conditions." Angew. Makromol. Chem.125: 9-26. Chen-Chun Ku, Ting-YaO Sun and Jing-Si Gao. 1982. "Phenol-Triazine Resin." Polymer News, 8: 76-78. 127 Chiavarini, M., N. Fanti and R. Bigatto. 1976. "Composite Characterization of Melamine Formaldehyde Condensation by NMR Spectroscopy." Angew. Makrom. Chem. 56: 15-25. Chow, S. 1972a. "Thermal Analysis of Liquid Phenol Formaldehyde Resin Curing." Holzforsch., 26:229-232. Chow, S. 1972b. "Lathe-Check Influence on Plywood Shear Strength." Can. For. Serv., West. For. Prod. Lab., Inf. Rep. VP-X-122, Vancouver. 12 p. Chow, S. and P. R. Steiner. 1979. "Comparison of Cure of Phenol Formaldehyde Novolac and Resol Systems by Differential Scanning Calorimetry." J. of Appl. Polym. Sci., 23:1973-1985. Chow, S. and W. G. Warren. 1972. "Efficiency of plywood bond-quality testing methods." Can. For. Serv., West. For. Prod. Lab., Inf. Rep. VP-X-104, Vancouver, 13 P. Clad, W., C. Schmidt-Hellerau. 1977. "Particlboards for the building Industry: Performance and Suitability Test." Proc. 11 th WSU Particleboard Symp. 33-61. Detwiler, E. B. 1953. "Properties and Application of Amino Resin Adhesives and Binders." For. Prod. Jour., vol. (9): 46-49. Dixon, J.K., NT. Woodberry and G.W. Costa. 1947. J . Am. Chem. Soc. , 69:599. Ellis, S. 1989. "Some Factors Affecting the Flow and Penetration of Powdered Phenolic Resins into Wood." Ph. D. Thesis, The University of British Columbia, B.C. 214 pp. Era, V. A. and A. Matilla. 1976. "Thermal Analysis of Thermosetting Resins." J. of Thermal Analysis, 10:461-469. Feinauer et al. 1979. "Process for the Production of White Melamine-Phenol-Aldehyde Resins which are Resistant to Yellowing." U.S. Patent 4,229,559. Gams, A., G. Widmer, W. Fisch. 1941. Helv. Chim. Acta, 24, 320E (Cited by Wohnsiedler, 1952). Gaylord, G.G 1968. "Urea and Melamine Resins." Macromolecular Synthesis, John Wiley and Sons, Inc., New York, London, Sydney, Toronto. Vol. 3: P.45, 49. Gollob. L , 1982. "The Interaction of Formulation Parameters with Chemical Structure and Adhesive Performance of Phenol-Formaldehyde Resins." PhD Thesis. Oregon State University, 153 pp. 128 Haupt, R. A., F. R. Ahmed Kabir and T. Sellers, Jr. 1991. "Effect of Aqueous Gel Filtrration Chromatography Parameters on Molecular Weight Determination of Phenolic Compounds." Proceedings from the Adhesives and Bonded Wood Symposium. Seattle, Washington. P. 450-461. Higuchi, M., J-K Roh, S. Tajima, H. Irata, T. Honda and I. Sakata. 1990. "Polymeric Structures of Melamine-Based Composite Adhesives." Proceedings from the Adhesives and Bonded Wood Symposium. Seattle, Washington. 429-449. Hodgins, T.S., A.G. Hovey, S. Hewett, W.R. Barrett and C.J. Meeske. 1941. "Melamine-formaldehyde Film-Forming compositions." Ind. Eng. Chem. 33: 769. Hofmann, A.W. 1874. Ber., 7, 1746 (Cited by Hodgins et. al., 1941. Holmberg, K. 1985. "Curing of Melamine Resins". Journal of Coatings Technology, 56: 34. Hse, C. Y. and Z. He. 1992. "Melamine Modified Urea-Formaldehyde Resin For Bonding Flakeboard". Hughes, E. W. 1941. J . Am. Chem. Soc , 63:1737 (Cited by Wohnsiedler, 1952). Ibigawa Electric Industry Company Limited. 1963. "Method of Manufacturing Colorless or Colored Melamine Modified Phenolic Resins." Pat. No. 1,057,400. King, P.W., R.H. Mitchell and A.R. Westwood. 1974. "Structural Analysis of Phenolic Resole Resins." J . of Appl. Polym. Sci., 18:1117-1130. Knutson, K.E.S. 1967. "Novel Resin and Method for Producing Same." U.S. Patent No. 3,321,551. Koehler, R. and Frey, R. 1943. Kolloid Z., 103(2):138 (Cited by Wohnsiedler, 1952). Kurachenkov, V. I and L. A. Igonin. 1971. "Curing Mechanism for Phenol-Formaldehyde Resins." J . Polym. Sci. 9: 2283-2289. Martin, R. W. 1956. "The Chemistry of Phenolic Resins.". Wiley, New York. P. 124 Maylor, R. 1995. "New Melamine Modified Binders for Moisture Resistant MDF." Proceeding of the Wood Adhesives Symposium, Portland, Oregon 115-121. Millett, M. A., R. H. Gillespie and B. H. River. 1977. "Evaluating Wood Adhesives and Adhesive Bonds." Report Prepared by U.S. Department of Agriculture, Forest Service, Forest Products Laboratory, Madison, Wl for Department of Housing and Urban Development, Office of Policy and Research. Moncrieff, R. W. 1947. "Melamine Resins." Paint Manufacture, 17:5. 129 Millett, M. A., R. H. Gillespie and B. H. River. 1977. "Evaluating Wood Adhesives and Adhesive Bonds." Report Prepared by U.S. Department of Agriculture, Forest Service, Forest Products Laboratory, Madison, Wl for Department of Housing and Urban Development, Office of Policy and Research. Moncrieff, R. W. 1947. "Melamine Resins." Paint Manufacture, 17:5. Pizzi, A. 1983. "Wood Adhesive-Chemistry and Technology." Marcel Dekker Inc., New York, 59-105. Pizzi, A. 1994. "Advanced Wood Adhesives Technology." Marcel Dekker, Inc., New York, 67-87. Powers, P. O. 1947. "Amino Resins and Plastics." in Kirk-Othmer Encyclopedia of Chemical Technology, A. Standen and J . Scott, Eds., 1s l ed., Vol. 1, Interscience Publishers, Inc., New York, 1947. 741. Richards, R. E. and H. W. Thompson. 1947. "Vibrational Spectra of Phenolic Derivatives and Phenolic Resins." j . Chem. Soc. 78: 126127. Roczniak, K., T. Biernaka and M. Skarzynski. 1983. "Some Properties and Chemical Structure of Phenolic Resins and Their Derivatives." J . Appl. Polym. Sci. 28: 531-542. Roh, J . , M. Higuchi and I. Sakata. 1987a. "Curing Behavior and Bonding Properties of Thermosetting Resin Adhesives I." Mokuzai Gakkaishi 33 (3): 193-198. Roh, J . , M. Higuchi and I. Sakata. 1987b. "Curing Behavior and Bonding Properties of Thermosetting Resin Adhesives II." Mokuzai Gakkaishi 33 (12): 963-968. Roh, J . , M. Higuchi and I. Sakata. 1989. "Curing Behavior and Bonding Properties of Thermosetting Resin Adhesives III." Mokuzai Gakkaishi 35 (4): 320-327. Roh, J . , M. Higuchi and I. Sakata. 1990. "Curing Behavior and Bonding Properties of Thermosetting Resin Adhesives III." Mokuzai Gakkaishi 36 (1): 36-41. Roh, J . , M. Higuchi and I. Sakata. 1991. "Curing Behavior and Bonding Properties of Thermosetting Resin Adhesives III." Mokuzai Gakkaishi 36 (1: 42-48. Sato, K. and T. Naito. 1973. Oikym. J . 5(2): 144. (Cited by Widmer, 1965). Selbo, M. L. 1965. "Performance of Melamine Resin Adhesives in Various Exposures." For. Prod. J . 10: 475-483. Tamura, Y., T. Yamada, Y. Hasegawa and A. Fukazawa,. 1981. "Recent Progress of Phenol Melamine Formaldehyde Resin Adhesives." Mokuzai Kogyo, 36(42): 315-320. 130 Tomita, B. and H. Ono. 1979. "Melamine Formaldehyde Resin: Constitutional Characterization by Fourier Transform 1 3C-NMR Spectroscopy." J. of Polymer Sci. 17: 3205-3215. Tomita, B. and T. Matsuzaki. 1985. "Co-condensation between Resol and Amino Resins." Ind. Eng. Chem. Prod. Res. Dev. 24(1): 1-5. Westwood, A. R. 1971. "Analysis of the Curing Reactions of Thermosetting Polymers." Thermal Analysis, 3: 169-176. Widmer, G. 1965 "Amino Resins", Encyclopedia of Polymer and Science and Technology, Vol. II: 1-94. Widmer, G. and W. Fisch. 1943. "Mixed Aldehyde Condensation Products and Process of Making Same." U.S. Patent Number 2,328,592. Wohnsiedler, H. P. 1952. "Polymerization in Melamine Formaldehyde Molded Resins." Ind. Eng. Chem. 44(11): 2679. 131 APPENDIX I 132 Table A1. The non-transformed average shear strength values for all resins at 120°C for 3 and 5 minute press times. 3 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Shear Standard Average Shear Standard Average Shear Standard Resin Strength (MPa) Deviation Strength (MPa) Deviation Strength (MPa) Deviation Type Cook #1 0.79 0.48 0.22 0.27 0.00 0.00 Cook #2 1.49 0.37 1.34 0.24 0.77 0.13 Cook #3 0.77 0.50 0.51 0.35 0.30 0.29 Cook #5 1.46 0.24 1.47 0.27 1.16 0.10 Cook #6 1.72 0.25 1.76 0.30 1.20 0.27 Lab MF 1.47 0.21 1.38 0.25 1.14 0.15 Comm. MF 1.53 0.21 1.63 0.33 1.08 0.13 Lab PF 1.00 0.32 0.75 0.43 0.52 0.47 5 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Shear Standard Average Shear Standard Average Shear Standard Resin Strength (MPa) Deviation Strength (MPa) Deviation Strength (MPa) Deviation Type Cook #1 1.31 0.18 0.80 0.06 0.18 0.14 Cook #2 1.69 0.15 1.41 0.34 1.06 0.11 Cook #3 1.40 0.20 1.06 0.23 0.54 0.08 Cook #5 1.45 0.21 1.34 0.19 1.13 0:13 Cook #6 1.72 0.39 1.67 0.28 1.14 0.18 Lab MF 1.70 0.26 1.73 0.17 1.45 0.22 Comm. MF 1.59 0.26 1.59 0.15 1.21 0.21 Lab PF 1.43 0.31 1.35 0.41 1.29 0.42 133 Table A2. The non-transformed average wood failure values for all resins at 120°C for 3 and 5 minute press times. 3 minute press times Dry Spec imens Soaked Spec imens Bo i l ed Spec imens Average Wood Standard Average Wood Standard Average Wood Standard Resin Type Failure (%) Deviation Failure (%) Deviation Failure (%) Deviation Cook #1 24 29 0 0 0 0 Cook #2 39 26 75 29 3 3 Cook #3 11 11 3 4 0 1 Cook #5 93 6 86 9 38 33 Cook #6 84 16 66 14 25 15 Lab MF 50 28 87 17 45 26 Comm. MF 60 27 77 19 24 29 Lab PF 4 4 8 10 8 13 5 minute press times Dry Spec imens Soaked Spec imens B o i l e d Spec imens Average Wood Standard Average Wood Standard Average Wood Standard Res in Type Failure (%) Deviation Failure (%) Deviation Failure (%) Deviation Cook#1 37 20 3 8 0 0 Cook #2 51 30 50 39 4 5 Cook #3 32 24 8 8 2 3 Cook #5 93 6 88 13 38 30 Cook #6 88 11 80 17 36 18 Lab MF 74 28 89 11 67 27 Comm. MF 77 20 84 12 27 26 Lab PF 5 4 36 23 53 29 134 Table A3. The non-transformed average shear strength values for all resins at 150°C for 3 and 5 minute press times. 3 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Shear Standard Average Shear Standard Average Shear Standard Resin Type Strength (MPa) Deviation Strength (MPa) Deviation Strength (MPa) Deviation Cook#1 1.20 0.21 0.99 0.24 0.20 0.20 Cook #2 1.66 0.26 1.52 0.23 1.09 0.13 Cook #3 1.50 0.24 1.53 0.24 0.72 0.26 Cook #5 1.68 0.38 1.61 0.16 1.39 0.21 Cook #6 1.50 0.25 1.32 0.30 0.96 0.14 Lab MF 1.73 0.27 1.62 0.20 1.45 0.19 Comm. MF 1.49 0.18 1.34 0.18 1.16 0.15 Lab PF 1.54 0.24 1.26 0.30 1.05 0.27 5 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Shear Standard Average Shear Standard Average Shear Standard Resin Type Strength (MPa) Deviation Strength (MPa) Deviation Strength (MPa) Deviation Cook#1 1.47 0.44 1.09 0.48 0.60 0.42 Cook #2 1.88 0.33 1.88 0.24 1.44 0.13 Cook #3 1.69 0.31 1.49 0.27 1.13 0.18 Cook #5 1.64 0.32 1.45 0.23 1.35 0.18 Cook #6 1.59 0.28 1.62 0.27 1.15 0.13 Lab MF 1.80 0.35 1.54 0.43 1.32 0.33 Comm. MF 1.59 0.32 1.43 0.27 1.13 0.20 Lab PF 1.90 0.22 1.42 0.19 1.35 0.24 135 Table A4. The non-transformed average wood failure values for all resins at 150°C for 3 and 5 minute press times. 3 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Wood Standard Average Wood Standard Average Wood Standard Resin Type Failure (%) Deviation Failure (%) Deviation Failure (%) Deviation Cook #1 66 10 13 18 1 2 Cook #2 89 6 84 9 22 22 Cook #3 72 22 27 23 2 3 Cook #5 94 6 84 14 40 23 Cook #6 88 8 64 25 24 10 Lab MF 84 23 96 4 77 22 Comm. MF 88 19 90 9 43 30 Lab PF 53 35 62 21 78 17 5 minute press times Dry Specimens Soaked Specimens Boiled Specimens Average Wood Standard Average Wood Standard Average Wood Standard Resin Type Failure (%) Deviation Failure (%) Deviation Failure (%) Deviation Cook#1 29 34 12 31 1 3 Cook #2 95 5 80 17 27 20 Cook #3 88 11 51 31 9 7 Cook #5 97 4 88 11 78 17 Cook #6 88 11 90 13 21 7 Lab MF 91 12 89 3 87 15 Comm. MF 87 8 92 12 47 28 Lab PF 58 19 67 19 58 22 136 APPENDIX II 137 Figure A1. Graph of the main effect of the resin from the ANOVA tables, (a) shear strength (b) wood failure. 138 3 minute Press Time 0 Dry Specimens • Soaked Specimens Ei Boiled Specimens Cook # j 5 Minute Press Time ffl Dry Specimens Bi Soaked Specimens • Boiled Specimens Cook # Figure A2. Non-transformed average shear strengths for panels pressed at 120°C and at 3 and 5 minutes. 139 gure A3. Non-transformed average shear strengths for panels pressed at 150°C and at 3 and 5 minutes. 140 3 minute Press Time @ Dry Specimens B Soaked Specimens • Boiled Specimens Cook# 2 T S. 1-5 2 ? 1 Jj 0.5 00 0+i 1 2 3 5 6 8 5 minute Press Time M Dry Specimens H Soaked Specimens 1 Boiled Specimens Cook# Figure A4. Non-transformed average wood failure values for panels pressed at 120°C and at 3 and 5 minutes. 141 3 minute Press Time M Dry Specimens H Soaked Specimens • Boiled Specimens 1 2 3 5 6 7 8 C o o k * 100 _ 80 £ 60 40 20 9 88881 tu 2 3 5 6 Cook# JL 8 5 minute Press Time 1 Cry Specimens M Seated SjDedirens 1 Boiled Sperirrens Figure A5. Non-transformed average wood failure values for panels pressed at 150°C and at 3 and 5 minutes. 142 

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.831.1-0088445/manifest

Comment

Related Items