Melamine Formaldehyde Adhesives

Transcription

1 32 Melamine Formaldehyde Adhesives A. Pizzi Ecole Nationale Supérieure des Technologies et Industries du Bois, Université de Nancy 1, Epinal, France I. INTRODUCTION Melamine formaldehyde (MF) and melamine urea formaldehyde (MUF) resins are among the most used adhesives for exterior and semiexterior wood panels and for the preparation and bonding of both low- and high-pressure paper laminates and overlays. Their much higher resistance to water attack is their main distinguishing characteristic from urea formaldehyde (UF) resins. MF adhesives are expensive. For this reason, MUF resins which have been cheapened by addition of a greater or lesser amount of urea are most often used. Notwithstanding their widespread use and economical importance, the literature on melamine resins is only a small fraction of that dedicated to UF resins. Often MFs and MUFs are described in the literature as a subset of UF amino resins. This is not really the case, as they have peculiar characteristics and properties all of their own which in certain respects are very different from those of UF adhesives. II. USES FOR MF RESINS Melamine formaldehyde resins are used as adhesives for exterior- and semiexterior-grade plywood and particleboard. In this application their handling is very similar to that of UF resins for the same use, with the added advantage of their excellent water and weather resistance. MF resins are also used for the impregnation of paper sheets in the production of self-adhesive overlays for the surface of wood-based panel products and of self-adhesive laminates. In this application the impregnation substrate, cellulose paper, is thoroughly impregnated by immersing it in the resin solution, squeezing it between rollers, and drying without curing it to proper flow by passing it through an airdraft tunnel oven at 70 to 120 Cat 10 m/s. The dry MF-impregnated sheets can then be bonded by one of two main processes: 1. The sheets of MF-impregnated paper, consisting of one surface layer or a few surface layers, are bonded together and with a substrate of paper sheets impregnated with phenolic resins to form laminates of variable thickness. In the impregnated papers is the dry but still active MF resin, which functions as the adhesive of the MF-impregnated sheet to both MF-impregnated sheets and at

2 the interface between MF-impregnated and phenol formaldehyde (PF)-impregnated layers. These laminates are high-pressure laminates. 2. The MF in an impregnated paper sheet is not completely cured but still has a certain amount of residual activity and is applied directly in a hot press, in a single sheet, on a wood-based panel, to which it bonds by completing the MF adhesive curing process. Press platens are made from stainless steel or chromium-plated brass and copper. The chromium layer preserves surface quality longer than does ordinary steel. The MF laminates exhibit a remarkable set of characteristics. Because of their unusual chemical inertness, nonporosity, and nonabsorbance, they resist most substances, such as mild alkalies and acids, alcohols, solvents such as benzene, mineral spirits, natural oils, and greases. No stains are produced on MF surfaces by these substances. In addition to almost unlimited coloring and decorating possibilities, this remarkable resistance has resulted in the extensive use of MF laminated wood-based panel products for tabletops, sales counters, laboratory benches, heavy-duty work areas in factories and homes, wall paneling, and so on. III. CHEMISTRY A. Condensation Reactions The condensation reaction of melamine (I) with formaldehyde (Fig. 1) is similar to but different from the reaction of formaldehyde with urea. As for urea, formaldehyde first attacks the amino groups of melamine, forming methylol compounds. However, Figure 1 Methylolation (hydroxymethylation) and subsequent condensation reactions to form melamine formaldehyde adhesive systems.

3 formaldehyde addition to melamine occurs more easily and completely than does addition to urea. The amino group in melamine accepts easily up to two molecules of formaldehyde. Thus complete methylolation of melamine is possible, which is not the case with urea [1]. Up to six molecules of formaldehyde can be attached to one molecule of melamine. The methylolation step leads to a series of methylol compounds with two to six methylol groups. Because melamine is less soluble than urea in water, the hydrophilic stage proceeds more rapidly in MF resin formation. Therefore, hydrophobic intermediates of the MF condensation appear early in the reaction. Another important difference is that MF condensation to give resins, and their curing, can occur not only under acid conditions, but also under neutral or even slightly alkaline conditions. The mechanism of the further reaction of methylol melamines to form hydrophobic intermediates is the same as for UF resins, with splitting off of water and formaldehyde. Methylene and ether bridges are formed and the molecular size of the resin increases rapidly. These intermediate condensation products constitute the large bulk of the commercial MF resins. The final curing process transforms the intermediates to the desired MF insoluble and infusible resins through the reaction of amino and methylol groups which are still available for reaction. A simplified schematic formula of cured MF resins has been given by Koehler [2] and Frey [3]. They emphasize the presence of many ether bridges besides unreacted methylol groups and methylene bridges. This is because in curing MF resins at temperatures up to 100 C, no substantial amounts of formaldehyde are liberated. Only small quantities are liberated during curing up to 150 C. However, UF resins curing under the same conditions liberate a great deal of formaldehyde. At the condensation stage attention must be paid to the formation of hydrolysis products of the melamine before preparation starts. The hydrolysis products of melamine are obtained when the amino groups of melamine are gradually replaced by hydroxyl groups. Complete hydrolysis produces cyanuric acid (Formula 1). Formula 1 Ammeline and ammelide can be regarded as partial amides of cyanuric acid. They are acid and have no use in resin production. They are very undesirable by-products of the manufacture of melamine because of their catalytic effect in the subsequent MF resin production, due to their acidic nature. If present, both must be removed from crude melamine by an alkali wash and/or crystallization of the crude melamine.

4 B. Mechanisms and Kinetics The mechanism of the initial stages of the reaction of melamine with formaldehyde leading to the formation of methylol melamines is very similar to that of urea. The reaction mechanism of the acid-catalyzed condensation reactions of methylol melamines to form polymers and resins has been elucidated by Sato and Naito [4]. Melamine and formaldehyde react similarly to urea and formaldehyde, although basic differences are evident in the reaction rates and mechanism. The primary products of reaction are methylolmelamines, and evidence indicates that such compounds are formed only at ambient or higher temperature except in acid ph ranges. The reaction is reversible throughout the ph range. Its forward rate is proportional to either [melamine][hcho] or [melamine][h þ CHOH] or [melamine þ ][HCHO], according to the ph used. Methylolmelamine forms dimers by condensation with melamine under neutral and acid conditions (70 C); this process is irreversible. The initial hydroxymethylation is very rapid. Its rate is determined by the condensation of conjugated acids of methylolmelamines with melamine. The reaction rate is proportional to [melamine] 2 [HCHO] [5]. When the [mineral acid]/[melamine] ratio is 0.0 to 1.0, the early stage hydroxymethylation of melamine is dependent on the concentration of the melamine molecule (base species) MH and its conjugated acid MH þ 2 in the following manner [6]: rate ¼ k H2 O½MHŠ½HCHOŠþk H ½MH þ 2 Š½HCHOŠþk MH2 þ ½MHþ 2 Š½MHŠ½HCHOŠþ þk MH ½MHŠ 2 ½HCHOŠ in the absence of added acid, when the ratio [mineral acid]/[melamine] is ¼ 0, the rate of the reaction can thus be represented as rate ¼ k H2 O½MHŠ½HCHOŠþk MH ½MHŠ 2 ½HCHOŠ The condensation reaction has been studied by investigating the kinetics of the initial stage of the condensation of di- and trimethylolmelamine (MF 2 and MF 3 ) in the ph range 1 to 9. Regardless of ph, the initial rate is equal to [4]: rate ¼ k½mf n Š 2 ðwith n ¼ 2or3Þ In the presence of mineral acid, the main reaction at the early stage of the condensation is the reaction between the methylolmelamine molecule and its conjugated acid (MF n H þ )[7]. This was found at an [acid]/[mf n ](n ¼ 2 or 3) ratio lower than 1.0 (ph 2.7). With an [acid]/[mf n ] ratio higher than 1.0 to 1.2 (ph<2), the main condensation takes place between the conjugated acids themselves. At equal ph values the condensation rate of trimethylolmelamine is considerably faster than that of dimethylolmelamine. This is the opposite of the rates of mono- and dimethylolurea. This means that while the nitrogen of the amido group in the case of urea is more reactive and therefore more nucleophilic than the nitrogen of the amidomethylol group, the opposite is true in the case of melamine. The reaction for MF 2 is primarily between the carbon of the methylol group next to the nitrogen in HM þ CH 2 OH, and the nitrogen of the amino group in MCH 2 OH. For MF 3, the condensation is mainly between the carbon of the methylol group next to the charged nitrogen in H þ MCH 2 OH, and the nitrogen of the aminomethylol group in MCH 2 OH [4]. The condensation rate therefore increases with the increasing electrophilicity of the carbon of the methylol group and the increasing nucleophilicity of the nitrogen of the amino group or aminomethylol group. Therefore in MF 3 the carbon in HM þ CH 2 OH is more electrophilic than the same carbon in MF 2. On the other hand, the nitrogen of the aminomethylol group in

5 HM þ CH 2 OH of MF 3 is less nucleophilic, and therefore less reactive, than the nitrogen of the amino group of MF 2. The effects of the carbon and nitrogen atoms are consequently opposite to each other in the MF n condensation. Since the effect of the carbon is greater than the effect of the nitrogen on the reaction rate, MF 3 condenses faster than MF 2.At lower ph values the effect of the nitrogen becomes negligible and MF 3 is even faster than MF 2 in condensing to polymers. The difference between the kinetic behavior of urea and melamine can be ascribed to the different effect of the nitrogen atom in the two compounds. With regard to the formation of methylol compounds as a result of hydroxymethylation, the functionality of melamine has been observed to be 6 against formaldehyde [1,8]. Similarly, melamine reacts easily with formaldehyde to form MF 3 ; it also forms MF 6 in concentrated formaldehyde [1,4,8]. For example, urea readily forms dimethylolurea, but forms trimethylolurea with marked difficulty [1,8] and never forms tetramethylolurea. These results suggest that the nitrogen of the amidomethylol group in methylolurea is considerably less nucleophilic than the nitrogen of the amido group in urea. However, the nitrogen of the aminomethylol group in methylolmelamine is not markedly less nucleophilic than the nitrogen of the amino group in melamine. Presumably, this is due to the difference in basicity between urea and melamine. The same is also true of their condensation reactions. C. Mixed Melamine Resins With regard to melamine urea formaldehyde, copolymers can be prepared which are generally used to cheapen the cost of MF resins, but which also show some worsening of properties. Copolymerization was proven by means of model compounds and polycondensates [9]. MUF resins obtained by copolymerization during the resin preparation stage are superior in performance to MUF resins prepared by mixing preformed UF and MF resins, especially because processing of such mixtures is quite difficult [10]. The relative mass proportions of melamine to urea used in these MUF resins is generally in the melamine:urea range 50:50 to 30:70 [11]. Melamine phenol formaldehyde resins, which in some respects show better properties than those of their corresponding MF and PF resins, have also been prepared [12 14]. Analysis of the molecular structure of those resins in both their uncured and cured states appeared to show that no co-condensates of phenol and melamine form and that two separate resins coexist. This is due to the difference in reactivity of the phenolic and melamine methylol groups as a function of ph. Also, in their cured state an interpenetrating network of the separate PF and MF resins, as a polymer blend, is formed, not a copolymer of the two [15 18]. Today MUF resins are produced in greater amounts than MF resins in the field of adhesives due to the relatively high cost of melamine: their formulation has progressed to such a level that often no difference in performance exists between a good MUF resin and a pure MF resin. MF resins are still more extensively used at this stage in the paper impregnation/laminates fields although both MUF copolymers as well as separate, double application of UF (paper core) and MF (paper surfaces) resins are making considerable inroads in this area. MUF resins instead totally dominate today in the wood adhesives field. Paper laminates and wood adhesives are the two main application areas of these resins. A type of resin also used today is the so-called PMUF (or MUPF according to which author is writing) adhesives. These are fundamentally MUF resins in which a minor proportion of phenol (between 3 and 10%; phenol:melamine:urea by weight of 10:30:60 for example) has been assumed to have coreacted with to further upgrade weather resistance of the bonded joint. Unfortunately the alleged superior performance of such resins is

6 Figure 2 Schematic representation of the dependence on the type of formulation used of the fate of phenol in a PMUF resin. (1) Phenol only present as unlinked free phenol/phenol derivatives but mainly as a pendant group neither participating in resin cross-linking nor contributing to resin performance and water resistance. (2) Intermediate case. (3) Case in which phenol is co-condensed and participating in the cross-linked network. often only wishful thinking as the phenol has frequently not been properly reacted with the other materials, and consequently the PMUF resin will have a worse performance than a comparable top of the range MUF resin. This was confirmed by the demonstration that it depends exclusively on the resin manufacturing parameters and materials reaction order used whether or not the phenol coreacts within many PMUF adhesives, showing that often the phenol remains as a useless pendant group in the hardened aminoplastic (MUF) network without contributing at all to its performance [19,20] (Fig. 2). The best reaction order necessary to obtain PMUF resins in which phenol makes a positive contribution to the performance of the hardened network has been reported [19]. PMUF resins are still used and some good resins of this type are indeed used in the unrealistic hope that they outperform equivalent MUF resins, when it has been shown clearly that they perform at best as a MUF adhesive presenting the same number of moles of melamine for the total moles of phenol plus melamine of the PMUF itself. The idea that the addition of small percentages of phenol to a MUF resin yields resins of better exterior durability is then an incorrect myth perpetuated in the wood panels industry. Newer formulations of MUF resins always outperform the corresponding PMUF. PMUFs are not bad resins, they are simply resins in which one of the materials, phenol, is often wasted for no purpose. IV. RESIN PREPARATION, GLUE MIXING, AND HARDENING Because of their characteristic rigidity and brittleness in their cured state, when MF resins are used for impregnated paper overlays, small amounts (typically 3 to 5%) of modifying compounds are often copolymerized with the MF resin during its preparation to give better flexibility to the finished product and better viscoelastic dissipation of stress in the joint. Most commonly used are acetoguanamine, e-caprolactam, and p-toluenesulfonamide (Formula 2). The effect of these is to decrease cross-linking density in the cured resin due to the lower number of amidic or aminic groups in their molecules. Thus in resin segments where

7 Formula 2 they are included, only linear segments are possible, decreasing the rigidity and brittleness of the resin. Acetoguanamine is most used for modification of resins for high-pressure paper laminates, while caprolactame, which in water is subject to the following equilibrium (Formula 3), Formula 3 is used primarily for low-pressure overlays for particleboard. Small amounts of noncopolymerized plasticizers such as diethylene glycol can also be used for the same purpose. Due to the peculiar structure of the wood product itself, MF adhesives for particleboard generally do not need the addition of these modifiers. Often, a small amount of dimethylformamide, a good solvent for melamine, is added at the beginning of the reaction to ensure that all the melamine is dissolved and is available for reaction. Sugar is often added to lessen cost of the resin. The aldehyde group of sugars have been proven to be able to condense with the amine groups of melamine and hence to copolymerize in the resin. Their quantity in MF resins must be limited to very low percentages, and if possible, sugars should not be used at all, as with aging they tend to cause yellowing, crazing, and cracking of cured MF paper laminates and to have a bad effect on adhesive long-term water resistance in both plywood and particleboard. MF adhesive resins for plywood and particleboard must be prepared to quite different characteristics than those for paper impregnation. The latter must have lower viscosity but still high resin solids content because they need to penetrate the paper substrate to a high resin load, to be dried without losing adhesive capability, and only later to be able to bond strongly to a substrate. Instead, MF adhesive resins for plywood and particleboard are generally more condensed, to obtain lower penetrability of the wood substrate (otherwise, some of the adhesive is lost by overpenetration into the substrate). The reverse applies for paper substrates, where the contrasting characteristics desired good paper penetration and fast curing can be obtained in several ways during resin preparation. These characteristics can be achieved by producing, for example, a resin with a lower degree of condensation and high methylol group content. Typically, a MF resin of a lower level of condensation with melamine/formaldehyde molar ratio of 1:1.8 to 1:2 will give the desired characteristics. Its high methylol content and somewhat lower degree of polymerization will give low viscosity at a high resin solids content, favoring rapid wetting

8 and impregnation of the paper substrate, while the high proportion of methylol groups will give it fast cross-linking and curing capabilities. A second, equally successful approach in to produce a MF resin of lower methylol group content and higher degree of condensation to which a small second addition of melamine (typically, 3 to 5% total melamine) is effected toward the end of resin preparation. The shift to lower viscosity and higher solids content given by a second addition of melamine, shifting to lower values the average of the resin molecular mass distribution, yields a resin of rapid impregnation characteristics. Conversely, the higher degree of polymerization of the major part of the resin gives fast cross-linking, and curing, due to the lower number of reaction steps needed to reach gel point. Typical total melamine/ formaldehyde molar ratios used in this system are 1:1.5 to 1:1.7. Figures 3 and 4 show typical temperature and ph diagrams for the industrial manufacture of MF and MUF resins for adhesives and other applications. The important control parameters to take care of during manufacture are the turbidity point (the point during resin preparation at which addition of a drop of MF reaction mixture to a test tube of cold water gives slight turbidity) and the water tolerance or hydrophobicity point, which marks the end of the reaction. The latter is a direct measure of the extent of condensation of the resin and indicates the percentage of water or mass of liquid on the reaction mixture that the MF resin can tolerate before precipitating out. It is typically set for resins of higher formaldehyde/melamine ratios and lower condensation levels at around 170 to 190%, but for resins of lower formaldehyde/melamine molar ratios and higher condensation levels it is set at around 120%. As can be seen from the diagrams in Fig. 3, once maximum reaction temperature is reached, ph is lowered to 9 to 9.5 to accelerate formation of the polymer. Once the turbidity point is reached, ph is again increased to 9.7 to 10.0, to slow down and more finely control the end point, determined by reaching of the wanted value of the water tolerance point. Industrial MF resins are generally manufactured to a 53 to 55% resin solids content with a final ph of 9.9 to 10.4 (but lower ph values are also used for low-condensation resins). To have acceptable rates of curing if higher ph values are used, higher quantities of hardener need to be used, which is clearly uneconomical. For typical MF resins for low pressure (particleboard), Figure 3 Typical temperature and ph diagrams for the industrial manufacture of MF resins.

9 Figure 4 Typical manufacturing diagram for 40:60 to 50:50 melamine/urea weight ratio MUF resins. self-adhesive overlay pressing times of between 30 and 60 s at 170 to 190 C press temperature are required according to the type of resin used. Pressing conditions for particleboard and plywood adhesives are identical to those used for UF resins. Glue mixing presents different requirements according to the final use of the MF resin. Hardeners are either acids or materials that will liberate acids on addition to the resin or on heating. In MF and MUF adhesives for bonding particleboard and plywood, the use of small percentages of ammonium salts, such as ammonium chloride or ammonium sulfate, is well established and is indeed identical to standard practice in UF resins. In MF adhesives for low- and high-pressure self-adhesive overlays and laminates the situation is quite different. Ammonium salts cannot be used for the latter application for three main reasons. First, evolution of ammonia gas during drying and subsequent hot curing of the MF impregnated paper would cause high porosity of the cured MF overlay. Second, the stability of ammonium salts, in particular of ammonium chloride, might cause MF liquid resin whitening and the MF-impregnated paper to cure and deactivate at ambient temperature after a short time in storage, causing the resin to have lost its adhesive capability by the time it is needed in hot curing. Third, the elimination of ammonia during drying and curing would leave the cured, finished paper laminate essentially very acid due to the residual acid of the hardener left in the system. This badly affects the resistance to water attack of the cured MF surface defeating the primarily advantage for which such surfaces have justly become so popular. Thus a stable, self-neutralizing, non-gas-releasing hardener is needed for such an application. Several have been prepared and one of the most commonly used is the readily formed complex between morpholine and p-toluenesulfonic acid. Morpholine and p-toluenesulfonic acid readily react exothermically to form a complex of essentially neutral ph that is stable up to well above 65 C (Formula 4). Formula 4

10 Table 1 Typical Paper Impregnation Glue Mix for Self-Adhesive Low-Pressure MF Overlays Ingredient Parts by mass MF resin, 53% solids content 99.1 Release agent 0.08 Wetting agent 0.16 Hardener (morpholine/p-toluenesulfonic acid complex) 0.64 Defoamer 0.02 During heat curing of the MF paper overlay in the press, the complex decomposes, the MF resin is hardened by the acid that is liberated, morpholine is not vaporized and lost to the system, and on cooling the complex is reformed, leaving the cured glue line essentially neutral. In MF glue mixing for overlays and laminates, small amounts of release agents to facilitate release from the hot press of the cured bonded overlay are added. Small amounts of defoamers and wetting agents to further facilitate wetting and penetration of the resin in the paper are always added. A typical glue mix is shown in Table 1. Two strong trends have appeared reasonably recently in the preparation of melamine-impregnated paper laminates. First, impregnating machines capable of giving papers in which much cheaper UF resin is substituting as much as 50% of the more expensive MF resin have now been in operation for several years. This equipment is based on a double impregnating bath application: the paper passes through a first bath where it absorbs the UF resin first, the excess on the surfaces being scraped off in-line, and then passes through a second bath where it absorbs the MF resin. The concept is to limit the UF resin to the inside of the paper with the MF resin coating the outside of the paper: the hardened surface after final curing will then have all the waterproof characteristics of a MF paper laminate but at a lower price. Good results are obtained and many machines using this type of process are today in industrial operation. A more recent trend has been to develop MUF copolymers to use with the less costly single impregnating bath machines. A few cases of this route to coping with the high cost of melamine are on record. V. MUF ADHESIVE RESINS OF UPGRADED PERFORMANCE Several effective techniques to consistently and markedly decrease the melamine content in MUF wood adhesives without any loss of performance have also been recently developed. Some of these formulation systems and techniques are already in the early stages of industrialization. Among these melamine/acid salts, such as melamine acetate (Formula 5), function both as efficient hidden hardeners of UF resins for plywood as well as upgrading the performance of simple UF resins for plywood by approximately 10% by mass melamine grafting to yield comparable strength durability of premanufactured MUF resins of 30 to 40% melamine mass content, hence of resins of much higher mass content of melamine. In short a MUF resin of melamine:urea weight ratio 10:90 will perform in certain applications such as exterior plywood as a premanufactured MUF resin of melamine:urea between 30:70 and 40:60 [21 24]. The system works both (i) by simple addition of the melamine salt in the UF glue mix eliminating the need to premanufacture a MUF resin. The effectiveness of melamine grafting in the glue mix and during hot pressing has

11 been found to depend on the relative solubility of the melamine salt which depends on both the acid strength of the acid as well as the number of acid functions in the salt. (ii) By use of salts in which the excess acid has been eliminated from the salt, hence melamine monoacetate with no loose acid residue. The salt can be added in the resin factory to a UF resin and the mix sold as a MUF resin as pot life is indefinite and the resin needs the addition of a classical hardener for aminoplastic resins such as sodium sulfate or sodium chloride for hardening. The solubility of the salts used increases with increase in temperature. The reasons why traditional, premanufactured MUF resins waste 2/3 or more of the melamine used in them, and why such a melamine salt addition system is so much more effective by not wasting melamine were presented in the same study [21]. Formula 5 How is it possible that addition of a melamine salt to a UF glue mix in a melamine:urea mass ratio of 10:90 yields plywood of comparable water resistance to a prereacted MUF resin of melamine:urea mass ratio in the range 30:70 to 40:60? As a consequence of what is presented above it is now possible to answer such a question. In the preparation of precopolymerized MUF resins, hence of today s normal, commercial MUF resins, during the high temperature preparation reaction the melamine also reacts with formaldehyde to form short MF chains which are then bound to the more abundant UF chains. Hardening of MUF resins has been proven to occur almost exclusively by cross-linking through CH 2 bridges connecting two melamines [20,25] as, due to its much lower reactivity, urea is not greatly involved. The use of melamine salts at ambient temperature in the glue mix instead ensures that only single melamine molecules are singly and separately grafted on the UF resin chain. U CH 2 M CH 2 M CH 2 M against U CH 2 M to yield rather different cross-linked networks than those of a standard MUF reactormade resin [21 26]. As to cross-link the system only a very small amount of melamine molecules for each UF chain is needed to achieve the same effect, to have several chains of MF as in standard MUF resins does not improve the bond strength because (i) only one of the melamines in the chain will react, the other not participating at all in final cross-linking, and (ii) the bonding strength will also not be improved by having even all the melamines of the MF chain react all in the same space zone of the network as shown in the first network formula above: on the contrary, the highly localized position on vicinal sites in the network of a high density of cross-links might well render the resin far too rigid and far too brittle

12 (which indeed is the case for most melamine-based resins). It is then clear that at least 2/3 of the melamine presently used in MUF resins is actually wasted and does not contribute much to the final results other than in a damaging manner, this being unavoidable as a consequence of the system of preparation used. The new system presented greatly improves on the present situation, not only on ease of handling (only a UF resin and a melamine salt as a hardener are needed rather than a more sophisticated MUF resin), but also on the amount of melamine needed (just approximately 1/3 of present consumption for equal exterior-grade bonding performance) with potentially considerable economic advantages as melamine is generally expensive. The results of a 2 year field weathering test in Europe have confirmed that a UF resin to which has been added 15% melamine acetate salt at the glue mix stage, to obtain a melamine:urea mass ratio of 10:90 solids on solids, imparts a better durability and better exterior performance to plywood glue lines than traditionally reactor-coreacted MUF resins of melamine:urea mass ratio of 33:66 and even of commercial, prereacted PMUF resin where the relative mass proportions of the materials in the resin are 10:30:60 [23]. Postcuring of aminoplastic-bonded wood joints has always been avoided due to the evident degradation induced by heat and humidity on the aminoplastic resin hardened network. This is a known fact and it is for this reason that boards bonded with UF, MUF, and MF resins are traditionally cooled as rapidly as possible after manufacture. However, tightening of formaldehyde emission regulations has caused considerable progress in aminoplastic formulations, especially much smaller molar ratios, and hence today s aminoplastic adhesives are indeed very different materials than those of years ago. A recent study [27,28] has shown that the postulate on the avoidance of postcuring of aminoplastic resin bonded joints is under many conditions no longer valid. Thus, (1) postcuring (for example by hotstacking in the simpler cases, by an oven or other heat treatment in more sophisticated cases) can be used in principle and under well-defined conditions to improve the performance of UF and MUF-bonded joints and panels without any further joint and hardened adhesive degradation, as the value of the modulus reached during postcuring is always consistently higher than the value at which the modulus stabilizes after complete curing during the pressing cycle. (2) Postcuring could also be used in principle and for the same reasons to further shorten the pressing time of MUF-bonded joints and panels when well-defined postcuring conditions are used or to decrease the proportion of adhesive used at parity of performance [27,28]. (3) There is clear indication that under certain conditions, even when adhesive degradation starts, the application of the posttreatment reestablishes the value of the joint s strength to a value higher than its maximum value obtained during curing. Some of the best posttreatment schedules have also been presented [27] (Fig. 5). The performance improvements in the internal bond (IB) strength of bonded wood panels are introduced by the series of reactions pertaining to internal methylene ether bridge rearrangements to a tighter methylene bridge network which have already been observed and extensively discussed in the analysis of aminoplastic and phenolic resins [27,29 31]. These are able to counterbalance well the degradative trend to which the aminoplastic resin should be subjected. Furthermore, in modern resins of lower F:(U þ M) molar ratio the amount of methylene ether bridges formed in curing is much lower. Thus, disruption by postcuring of the already formed resin network by internal resin rearrangements will be milder, if at all present, and will definitely not yield the marked degradation and even collapse of the structure of the network which characterizes older resins of much higher molar ratio when postcured under the same conditions [32,33]. In short, notwithstanding the internal rearrangement the network will stand and stand

13 Figure 5 Thermomechanical analysis of a joint glued with a modern MUF adhesive showing the advantage of hot-post-stacking for modern, low molar ratio aminoplastic adhesive bonded panels. Note the maximum modulus achieved during isothermal heating (180 C for 8 min) (lower curve) and maximum modulus achieved after cooling and reheating at 100 C for 8 min (upper curve): the difference in modulus is the potential gain due to hot-post-stacking. quite strongly: no, or hardly any decrease of IB strength will be noticeable. For modern, lower molar ratio aminoplastic adhesives, since the resin network does not noticeably degrade or collapse with postcuring, only the tightening of the network derived by further bridge formation by reaction within the network of the few formaldehyde molecules released by the now mild internal rearrangement will be noticeable: the IB strength value will then improve with postcuring in boards bonded with modern, lower formaldehyde aminoplastic adhesives. There are important differences in the behavior of MUF resins prepared in different ways, and hence at the level of their performance as binders of wood panels, due both to their differences at the level of the resin structure and to the type and distribution of the molecular species formed before hardening, as well as to the differences in the structure of the final hardened networks. An example of three types of MUF resins examined can illustrate this point. (i) A sequential MUF in which the UF was prepared first and then melamine coreacted afterwards once the UF polymer had been formed [8], a last small urea addition also being carried out for a final (M þ U):F molar ratio of 1:1.5 and M:U weight ratio of 47:53, (ii) a MUF resin in which the great majority of the urea and of the melamine were premixed and then reacted simultaneously to form the resin, followed by addition of small amounts of both last melamine and last urea, for a (M þ U):F molar ratio of 1:1.5 and M:U weight ratio of 47:53, and (iii) a UF resin of molar ration 1:1.5 to which has been added 15% by weight on resin solids of monoacetate of melamine in the glue mix for a final (M þ U):F molar ratio of 1:1.39 and M:U weight ratio of 14:86. The proportion and type of chemical species formed which can be calculated by the molar proportions of the reagent, the manner in which these are combined during the reaction under different conditions as well as the rate reaction constants of urea and melamine with formaldehyde lead to the conclusion, confirmed by 13 C nuclear magnetic resonance (NMR), that the distribution of species for resins (i), (ii), and (iii) are as follows (their relative proportions are indicated in Formulas 6, 7, and 8). Case (i) above presents the following predominant chemical species (Formula 6), where M attached to the UF polymer is in the form of both a single melamine as well as in the form of a melamine formaldehyde short oligomer.

14 Formula 6 Case (iii) above presents instead just UF oligomers and melamine salts (Formula 7), Formula 7 where M is always in the form of a single melamine molecule. Case (ii) above presents the following predominant chemical species (Formula 8), Formula 8 Thus an MF resin drowned in mostly unreacted urea and where M attached to the UF polymer is in the form of both a single melamine (M and M framed) as well as in the form of a MF short oligomer (M framed). The structure of the three resins when still in liquid form explains the appearance of their structure after hardening. Thus, hardened MUF resins of formulation type (ii) will present structures as presented in Formula 8 and thus will waste the benefit of a considerable proportion of the melamine used. Hardened MUF resins of type (i) will present structures intermediate between those shown in Formulas 7 and 8 (but tending more to the type of Formula 8) and thus while also wasting a considerable proportion of the

15 melamine used, this will be less than for formulations of type (ii): the strength and water resistance results of MUFs of type (iii) will then be noticeably better at parity of all other conditions than what is obtainable with resins of type (ii), as indeed has been shown to be the case. MUF resin formulations of type (iii), those of melamine acetate type, will give hardened structures according to Formula 7 without wasting much melamine and giving hence the best performance, with the limitation of proportion already mentioned and explained above. This can be seen by comparing the strength results obtained by constant heating rate thermomechanical analysis (TMA) [26]. A MUF formulation of type (iii) containing 20% melamine acetate performs almost as well as a good formulation of type (i) which contains two and a half times more melamine. They both perform much better than a formulation of type (ii) [26] with some notable exceptions [38]. Another recent approach which has shown considerable promise in markedly decreasing the percentage of adhesive solids on a board, and hence in markedly decreasing melamine content, has been found almost by chance. It is based on the addition of certain additives to the MUF resin. Additives have been found that are both able to decrease melamine content in MUF resins at parity of performance, as well as able to decrease the percentage of any MUF resin needed for bonding while still conserving the same adhesive and joint performance. This second class of additives works for UF adhesives too, but less well, while it gives acceptable results for PF resins, but it is at its best in the case of MUF resins. This second class of additives is the acetals [34 36], methylal and ethylal being the two most appropriate due to their cost to performance ratio, which do not release formaldehyde at phs higher than 1 [35]. Methylal has according to results reported by the Environmental Protection Agency (EPA) an LD 50 value of 10,000 against that of 100 in the case of formaldehyde, and is thus classed as nontoxic. The addition of these materials to the glue mix of formaldehyde-based resin improves considerably its mechanical resistance and the performance of the bonded joint. This is in general valid for MUFs, UFs, and PFs, but the effect is particularly evident for the MUF resins [35]. Decreases in MUF resin solids content of as much as 33% while conserving the same performance are reported in the case of wood particleboard. In Fig. 6 are shown the continuous heating rate TMA curves of modulus as a function of temperature for an MUF resin of 1:1.2 (M þ U):F molar ratio. Similar but much less extreme Figure 6 Thermomechanical analysis graph showing the increasing maximum values of the modulus of a MUF-bonded joint with increasing amounts of methylal (an acetal) as an effectiveness upgrading additive.

16 trends are obtained also for UF and PF resins. In the case of MUF resins the addition of 10% additive on resin solids yields laboratory particleboard in which one can decrease the percentage of resin solids on the board by between 20% and 25% without any loss of performance. Similarly, at equal resin solids the strength of a particleboard is 33% higher when 10% additive on resin solids is added to the glue mix. Addition of 20% methylal on the board yields, in the case of the same resin, the same strength with 30% less adhesive (and hence less melamine) [35]. What is the mechanism of action of methylal, ethylal, and some other acetals to achieve such a feat? Their excellent solvent action on melamine and higher molecular weight ligomers. The cases shown earlier in this chapter referring to melamine salts and the loss of effectiveness due to wastage of melamine are applicable in this case too. Melamine when added to a reacting mixture during resin manufacture is not really soluble. It reacts then in heterogeneous phase with the other components of the resin, some of it being in a transient state in equilibrium between being in solution and being out of solution, and thus its efficacity is partially, but noticeably reduced. The introduction of an excellent solvent, none better than these acetals was known before, brings the totality of the reaction into homogeneous phase with a consequent, noticeable improvement in both the effectiveness of reaction and the effectiveness of melamine utilization. A different class of additives from those above but also able to decrease melamine content in MUF resins at parity of performance also exist. They are based on the addition in the glue mix of 1 to 5% additive and allow preparation of MUF copolymers, premanufactured in a traditional manner, in which either the proportion of melamine is lower, for example a 20:80 by weight M:U resin to which the additive has been added performing as well as a M:U 50:50 resin, or alternatively to upgrade a top of the range M:U 50:50 MUF adhesive to an exterior performance comparable and even superior to that of PF resins [37 39]. Several different types of additives can achieve this but they are all based on the preparation and acid stabilization of imines, or better of iminomethylene bases [38,39], and of their addition to the MUF resin. Thus, the effect is still the same whether the imines/iminomethylene bases, acid-anion stabilized, are prepared by coreaction of ammonia and formaldehyde [38,39], or for instance as described for acid-anion stabilized decomposition of hexamethylenetetramine [38,39] (Fig. 7). The structure of the imines and the iminomethylene bases yielding this effect are very similar indeed to the structure of the acetal additives presented above, the NH bridge of the imines having the same function of the O bridge of the acetals. The imines/iminomethylene bases have the added dimension, however, that the nitrogen can function as a knot of tridimensional cross-linking itself, which the oxygen bridge obviously cannot do. The amount of nitrogen-based additive that can be used is limited by its higher sensitivity to water in the hardened network. This is not the case of the possibly less effective oxygen-based additives, which can be used in greater amount: one property balances the other. The oxygen bridge conversely presents perhaps a better longer-term thermal stability than the nitrogen-based bridges. These are only very relative, rather subjective advantages. What is instead important is that the similarity of structure indicates that in the main (but not completely) the mode of action of all these additives may appear to be the same, but often different effects are at work, namely first a considerable improvement of the viscoelastic dissipation of the energy of the glue line and bonded joint without a drop in cross-linking density. The differences between the different additives is then due to additional, although rather important effects such as the solvent effect of the acetals in the MUF resins, and the increase in reaction rate [25] and buffer effect [38] of the iminomethylene basis, as well as others. It is on the basis of

17 Figure 7 Mechanism of hexamethylenetetramine decomposition leading to the formation of anionstabilized reactive iminomethylene bases. The same bases can be formed by reaction of ammonium salts such as ammonium sulfate and formaldehyde and constitute a metastable intermediate between hexamine and final decomposition products, and vice versa (after refs 26, 37 39). this similarity of structure and effect that a scale of additives providing similar effects to different levels has been established (see Formula 9) [40]. It must be pointed out that a TMA strength improvement of 100% on the MUF resin without methylal (this is achieved by addition of 20% methylal on resin solids) corresponds in the actual wood particleboard to an increase of IB strength of 33%. This means that of all the compounds shown above only the acetals, such as methylal and ethylal, as well as the similarly structured imine/iminomethylene bases discussed above (for which the effect on strength is more marked) are capable of marked improvements in IB strength at the actual wood panel level. These developments are of use for MUF resins not just in the field of wood adhesives, or of other binders in general, but also to improve and upgrade the performance of

18 Formula 9 resins in other applications such as that of melamine-based impregnated paper laminates, where they have been shown to improve considerably the storage stability of paper impregnating resins [36]. As regards the more application-bound physical aspects of MUF resins these can be applied in different ways, this too sometimes having a bearing on other types of additives used. Thus to the normal case of a MUF plus its hardener one can add cases in which a formaldehyde depressant such as a low condensation MF, UF, or MUF precondensate or one of their mixes is added; sometimes this is in combination with an accelerator based on the same principle. Such an approach is more used in other resins, but it has been shown and reported as being feasible also for MUF resins [41]. VI. TEMPERATURE TIME-TRANSFORMATION AND CONTINUOUS- HEATING-TRANSFORMATION CURING DIAGRAMS OF MUF RESINS WHEN ALONE AND HARDENING IN A WOOD JOINT (OR OTHER INTERACTIVE SUBSTRATE) Temperature time-transformation (TTT) and continuous-heating-transformation (CHT) curing diagrams for polycondensation resins are starting to acquire more importance in the deductions of the behavior of different resins during hardening. They are a type of state diagram. TTT and CHT diagrams of resins by themselves or on noninteracting substrates show similar trends as exemplified by the case of epoxy resins on glass fibers [42,43] (Fig. 8a). Different trends than those for TTT and CHT diagrams of epoxy resins reported in the literature (Fig. 8a) occur, however, in the higher and lower temperature zones of the diagrams of waterborne formaldehyde-based resins hardening on wood. CHT and TTT diagrams have already been reported for PF, UF, MUF, phenol resorcinol formaldehyde (PRF), and tannin formaldehyde thermosetting resins [31,44 46] (Figs. 8b and c). The higher temperature zone of the CHT diagrams for MUF resins in a wood joint, reported in Fig. 8c, shows the same trends (and for the same reasons, namely the interactive nature of the substrate and movement of water from resin to substrate and vice versa) observed for UF and PF resins. However, the experimental TTT diagram in Fig. 8b shows quite a different trend from the CHT diagram for the same resins and for the TTT diagrams reported in the literature for epoxies on noninteracting substrates such as glass fiber. To start to understand the trend shown in Figs. 8b and c it is first necessary to observe what happens to the modulus of the wood substrate alone (without a resin being present) when examined under

19 Figure 8 (a) Schematic classical TTT and CHT curing diagrams of epoxy resins and any other polycondensation resin, such as MUFs, on a nonreactive, noninterfering substrate. (b) Schematic TTT curing diagram of MUF, PRF, UF, and PF wood adhesives on wood as an interacting substrate. (c) Schematic CHT curing diagram of PRF, MUF, UF, and PF wood adhesives on wood as an interacting substrate.

20 the same conditions of a wood joint during bonding. No significant degradation occurs up to a temperature of 180 C as shown by the relative stability of the value of the elastic modulus as a function of time. Some slight degradation starts to occur at 200 C, but after some initial degradation the elastic modulus again settles to a steady value as a function of time and at a value rather comparable to the steady value obtained at lower temperatures. Evident degradation starts to be noticeable in the C range and this becomes even more noticeable at higher temperatures. The effect of substrate degradation on the TTT diagram in Fig. 8b can then only start to influence the trends in gel and vitrification curves at temperatures higher than 200 C and it is for this reason that the region of the curves higher than 200 C is indicated by dashed lines in Fig. 8b. At a temperature 200 C the trends observed are due to the resin only. In this range of temperature the eventual turning to longer time and stable temperature of the vitrification curve, characteristic of the TTT diagrams of epoxy resins, becomes also evident for the TTT diagrams of the waterborne PRF and MUF resins on lignocellulosic substrates indicating that diffusion hindrance at a higher degree of conversion becomes for these resins too the determinant parameter defining reaction rate. What differs, however, from previous diagrams is that the trend of all the curves, namely the gelation curve, initial pseudogel (entanglement) curve, and start and end of vitrification curve is the same. In epoxy resin TTT diagrams the trend of the gelation curve is completely different from that reported here. The result shown in Fig. 8b is, however, rather logical because if diffusion problems alter the trend of the vitrification curve, then the same diffusional problem should also alter the gelation and pseudogel curves. This is indeed what the experimental results in Fig. 8b indicate. It may well be that in waterborne resins the effect is more noticeable than in epoxy resins. This is the reason why it is possible to observe it for PF, UF, PRF, and MUF resins. With the data available and with the limitation imposed by the start of wood substrate degradation of higher temperatures it is not really possible to say if the gelation curve and the vitrification curve run asymptotically towards the same value of temperature at time ¼1 although the indications are that this is quite likely to be the case. What is also evident in the trend of the two curves is the turn to the left, hence the inverse trend of their asymptotic tendency towards T g1. This turn cannot be ascribed to substrate degradation because for very reactive resins, such as PRFs, such a turn already occurs at a temperature lower than 150 C, hence much lower than the temperature at which substrate degradation becomes significant. This inverse trend can only be attributed to movements of water coming from the substrate towards the resin layer as the trend of the curves indicates an easing of the diffusional problem already proven to occur at such a high degree of conversion [30,39]. Two other aspects of the TTT diagrams in Figs. 8b and c must be discussed, these being the trend of the curves at temperatures higher than 200 C and the trend of the devitrification (or resin degradation) curve. The dashed line trend and experimental points of all the curves at temperatures higher than 200 C are clearly only an effect caused by the ever more severe degradation of the substrate: degradation of the substrate implies a greater mobility of the polymer network constituting the substrate, hence the continuation of the curves as shown in their segmented part. That this is the case is also supported by the virtual negative times yielded by the TMA equipment when the temperature becomes extreme, as well as by the trend of the resin s higher degradation curve which tends to intersect the vitrification curve at about C or higher, this being a clear indication that one is measuring the changes in the reference system, the substrate itself, and that these are at this stage so much more important than the small changes occurring in the resin as to be able to dominate the whole complex system which is the bonded joint.