Лако-красочные материалы — производство

Organic isocyanates are very reactive with compounds that possess ‘‘active’’ hydrogens, as in the case of carboxylic acids, primary and secondary amides, primary and secondary amines, alcohols, phenols, and water. The relative reactivity of these compounds with isocyanate depends upon the nucleophilicity and steric structure of the attacking com­pound. Primary and secondary amines are typically most reactive, followed by primary alcohols, water, secondary alcohols, and phenols in that general order. Carboxylic acids and amides are the least reactive with isocyanate. Of course, deprotonation may increase nucleophilicity. For example, a phenoxide anion will be much more reactive than a phenol, and probably more reactive than a primary alcohol. Regarding the isocyanate, one should realize that the various NCO groups in pMDI also have variable reactivity. Isocyanate groups without ortho substitution (nothing ortho to the NCO group) are more reactive than the ortho substituted NCO groups. Consequently, 4,40-MDI is more reactive than 2,2′-MDI. Furthermore, and on average, pMDI is less reactive than the pure 4,40-MDI because the oligomeric polyisocyanates have ortho substituted NCO groups as shown above; this assumes reaction in the liquid state. Reactions most pertinent to wood bonding are with alcohols, phenols, and water, as shown below in general outline.

During urea formation, note that each mole of consumed water leads to an equivalent of CO2 gas. There has been significant controversy surrounding the precise nature of pMDI — wood cure chemistry, which continues even today. The importance of urethane formation with wood has long been recognized; such a linkage will enhance weather durability. However, the abundance and mobility of water vapor during panel hotpressing could preferentially promote urea formation. The significance of urethane formation has been debated in the literature since 1975, and possibly earlier [12-17]. Some have argued that urethane formation is unlikely [17], while others have predicted some combination of urea and urethane formation [12-16]. Wittman is credited with making one of the first analy­tical contributions to this issue [18]. By measuring CO2 production during panel hotpres­sing, Wittman showed that about 50% of the applied pMDI was consumed by polyurea formation. Thus it has long been established that the essence of pMDI cure is the reaction with water to form a polyurea network. In other words, pMDI wood binders are two-part systems where wood water is the second integral component. However, we are still uncer­tain of the significance of urethane formation, although this may be a moot point. Polyureas are unsurpassed in their capacity for hydrogen bonding, and they have excellent thermal and hydrolytic stabilities. Consequently, polyurea networks will adhere strongly to wood through secondary forces. Under current industrial practice, it may be safe to say that the incidence of urethane formation is little more than an academic issue. However, the future will no doubt provide ever advanced materials from wood, so the detection and promotion of urethane formation will become more than an academic exercise. A brief review of the search for urethane formation follows immediately.

Some investigators have analyzed wood-pMDI cure using differential scanning calorimetry [19,20]. This method demonstrates that wood has a significant impact on the heat of pMDI cure. However, calorimetry cannot identify cure reactions; so chemical identifications are highly suspect when based upon this method. Others have used infrared (IR) spectroscopy. For example, Weaver and Owen used IR to reveal the reactions of glucose, cellulose, lignin, and wood with phenyl isocyanate and pMDI [21]. They found that urethanes formed more readily with lignin than with cellulose, but that the water reaction was predominant in all cases. Urethanes were only detected when huge excesses of isocyanate were used [21]. Rosthauser et al. conducted a thorough model study and an in situ IR analysis; a remote sensing fiber optic probe was imbedded in laboratory scale particleboard bonded with pMDI [22]. Based upon carbonyl stretching, the authors claim no indication of urethane formation [22]. However, IR identifications within a wood particle mat are complicated by two effects: (1) signal overlap, and (2) the lack of depend­able reference spectra that accurately represent conditions within the mat. The study by Rosthauser and colleagues also showed that uretidione formation (isocyanate dimeriza­tion) might be more important than previously suspected [22]. Others have used isotopic labeling coupled with nitrogen-15 solid-state nuclear magnetic resonance (NMR) to directly probe the cure chemistry of intact bondlines [7,23-25]. This method reveals that biuret and polyuret formation is very common, that bondline chemistry is very sensitive to cure time and temperature, and that even the wood species may affect cure chemistry. These findings suggest that bondline chemistry (and thus performance) will vary according to the local conditions through the thickness of the wood composite. Unfortunately, nitrogen-15 solid-state NMR cannot clearly detect urethane formation because the urea and urethane nitrogen signals are almost perfectly overlapping. Indirect evidence for urethane formation has been found through relaxation measurements [7]. From the same study, it was suggested that the putative urethane linkages were subject to thermal cleavage, consistent with the thermal instability known of aromatic urethanes. All of the previously described work has a common shortcoming—there is no unambiguous method to detect urethanes. This failing may be resolved with another solid-state NMR strategy that builds upon the nitrogen-15 method. It was recently shown that urethanes were detectable when the pMDI wood binder was prepared with a double isotopic label, e. g., nitrogen-15 and carbon-13 in the isocyanate group [26]. Using small flake samples, this study found: (1) that urethane formation was abundant with cure conditions of 3.45 MPa (500 psi), 22-165°C, and 3min cure time, (2) that urethane linkages in the wood bondline were subject to thermal cleavage above 165°C, and (3) that the industrial significance of these findings remained to be seen [26]. In other words, the double label method may provide the final answer, but it has yet to be conducted under pilot scale conditions that better reflect industrial practice. Nevertheless, the abundance of urethane seen in this study was striking [26]. The probable cure chemistry of the pMDI-wood bondline is depicted below, but again much is still unknown.

If urethanes are common, we should also expect allophanate formation. However, reliable confirmation of this reaction may never occur due to interference from structurally similar linkages such as biurets. Besides, the extreme thermal instability of allophanates may preclude their contribution to bonding. Biurets and urethanes are substantially more thermally stable, in that order. Furthermore, alkyl urethanes (which arise from alcohols) are more stable than aryl urethanes (arising from phenols). Of course, both alkyl and aryl urethanes are possible with wood. As mentioned, ureas are very stable; their formation is essentially irreversible. While the thermal cleavage of urethane and biuret linkages has been demonstrated, the corresponding effect on bondline performance has not been deter­mined. In any event, it is clear that all of these chemistries promote hydrogen bonding with wood. Regardless of how important urethane formation might be, it is certain that pMDI wood binders promote strong secondary interactions with wood. As with most adhesives, secondary interactions are completely adequate for dry strength properties. In other words, the dry strength of unweathered pMDI bonded composites reveals nothing about urethane formation. Only durability will be impacted by hydrolytically and thermally stable covalent bonding.

The debate over urethane formation in the wood/pMDI bondline may soon yield to studies of the peculiar morphology of this interphase. Recall that pMDI is a low viscosity, low molecular weight, low surface tension organic liquid. Consequently, pMDI wood binders readily wet and deeply penetrate into wood, as demonstrated by Shi and Gardner [27]. In fact the deep penetration of pMDI into wood contradicts traditional views on wood adhesion. The truly polymeric wood binders such as phenol-formaldehyde (PF) and urea-formaldehyde (UF) are formulated for only moderate levels of wood penetration; overpenetration is undesirable with these resins. By traditional standards, pMDI wood binders overpenetrate and yet they perform as well or better than other wood binding thermosets. One then wonders what becomes of the resin that does not span the gap between bonded wood particles? Does it polymerize into a bulk phase within wood cell lumens, providing no benefit? Or does the deep penetration provide some performance gain? How deep is the penetration?

Certainly, pMDI flows into the micrometer size voids of wood via capillary action. Furthermore, it is apparent that penetration occurs down to the angstrom scale. In other words, pMDI actually penetrates into the amorphous components of the wood cell wall, mixing on the molecular level. The wood cell wall plasticization by pMDI was first demonstrated by Marcinko et al. with aspen (Populus tremula) wood using solid-state NMR [28]. The corresponding wood swelling caused by pMDI can be measured using a thermomechanical analyzer, an instrument capable of measuring minute dimensional changes. For example, Figure 1 shows a swelling profile of a small block of aspen wood in pMDI. The maximum swelling is slight, about 0.48%; nearly half of this swelling occurs within 5-10 min.

The swelling by pMDI of the amorphous wood polymers has interesting implications for the interphase morphology. With pMDI monomers and oligomers dispersed among amorphous wood chains, the subsequent cure might provide a type of interpenetrating polymer network (IPN) where a synthetic polyurea/biuret network interpenetrates amor­phous wood polymers [29]. Urethane formation may then be less important because the disruption of an IPN morphology would require covalent bond cleavage. Proof for the IPN morphology is not yet available, but supporting evidence appears in the literature. For example, Marcinko and coworkers have shown that cured pMDI restricts wood polymer motions according to dynamic mechanical analysis and solid-state NMR [30,31]. Furthermore, the literature suggests that pMDI wood binders impart dimensional

stability and thickness swell resistance to wood panels [13,32,33]. Such moisture resistance would be consistent with the hypothetical IPN interphase morphology, but not proof.

It is also possible that the cured adhesive may exhibit phase heterogeneity that could impact mechanical properties, toughness in particular. The predominant chemical linkages within the pMDI/wood bondline appear to be urea and biuret/polyuret. Nitrogen-15 solid-state NMR studies suggest that these chemical groups may exist separately within nanometer scale phases [7,25]. One study demonstrated the onset of phase heterogeneity only after extended cure times [7]; while another found phase heterogeneity at shorter cure times, more representative of industrial practice [25]. This issue requires additional study because the NMR pulse sequence (which was different in the above-mentioned works) has a strong influence on the detection of phase separation in the wood/pMDI bondline [34]. Finally, recent work raises the possibility that crystallinity may develop during pMDI cure [35,36]. Such crystallinity is known to occur in MDI based elastomers. Since pMDI is about 50% MDI monomer, crystallinity formation in the pMDI/wood bondline is a reasonable hypothesis. Additional research will shed light on this interesting possibility.