into the wood is determined by the resistance to the hydrodynamic flow. The least resistance is in the longitudinal direction through the lumens in the long and slender tracheids of softwoods, or through the vessels of hardwoods. Since vessels are connected end to end with perforation plates and no pit membrane, this cell type dominates the penetration of adhesives in hardwoods. Using optical microscopy, Kamke and Lee observed the presence of resin in pit chambers of both hardwood and softwood species and in cell lumens as a result of entry of the resin through a pit [44].
The rheology of the adhesives plays a significant role in the adhesive penetration. Polymeric adhesives exhibit non-Newtonian behavior [45]. Also, the capillary pathways through cell lumens and pits are tortuous. As a result, penetration of the adhesives through wood is further complicated.
Waterborne adhesives, such as the phenolics and amino resins, are heterogeneous with unique distribution of MWs. They are prone to separation (1) when the water is absorbed by the cell wall and (2) when the high MW polymer molecules are trapped by the pit membrane into different depths depending on the MW of the polymer (comparable to gel permeation chromatography). Because of the above wood-related factors, the influence of fluid dynamics of the adhesives on its penetration into wood cannot be generalized. Gross penetration can happen with most types of resin at low viscosity, while cell wall penetration can occur only with resins having small MW components. In order to determine what should be the critical molecular size of the adhesives below which they can penetrate into the cell wall, Tarkow et al. [46] studied the critical MW of polyethylene glycol (PEG) needed to penetrate the cell wall of Sitka spruce. Their study showed that the critical MW of PEG was 3000 at room temperature. MW fractions less than 3000 are common in the case of phenolic, amino resins, and isocyanate resins and therefore these resins can be expected to penetrate into the cell wall. Further, prior to polymerization, adhesives penetrating the cell wall swell and plasticize the wood. This is an additional factor in favor of penetration of these resins into the cell wall. This has been reported for pMDI and low MW PF adhesives [32, 47, 48].
While discussing the importance of gross and cell wall penetration with respect to the adhesion mechanisms, Frihart proposed four scenarios [49]. In the first case of micrometer penetration, the adhesive occupies the free volume within the cell wall—thus inhibiting shrinking and swelling. In the second case, there can be mechanical anchoring between the cured adhesive and the substrate. The third case is the formation of interpenetrating polymer network that is made up of the cross-linked adhesive within the free volume of the cell wall. The last case claims the formation of chemical cross-links with the cell wall polymeric components.
1.7 Wood Factors Affecting Penetration
The actual penetration depth of solution or adhesive depends on the permeability of wood to liquids, the technological methods, and physicochemical properties of the specific adhesive. In the case of liquid flow through porous material under ideal conditions with no interactions occurring between the liquid and the porous material, the permeability is defined by Darcy’s law [50]:
where Q is the liquid volume flow [m3 s–1], K is the specific permeability of wood [m2], A is the area perpendicular to the liquid flow [m2], L is the sample length in the direction of flow [m], η is the dynamic viscosity of the liquid [Pa s], and ΔP is the pressure gradient [Pa]. As described by Darcy’s law, the pressure gradient ΔP is the cause for the liquid penetration into wood.
The permeability and surface energy are the two wood-related factors controlling adhesive penetration [44]. Permeability varies with species and direction (e.g., tangential, radial, and longitudinal). However, longitudinal permeability may be as much as 104 times greater than transverse permeability [51]. Wood species with low permeability, such as Douglas-fir heartwood, severely restricts resin penetration in the radial and tangential directions. High permeability of the wood surface may be problematic to adhesive bonding if this leads to starvation at the bondline. Thus, bonding endgrain is difficult [44]. There are earlywood and latewood differences, as well as heartwood and sapwood differences. Pit aspiration sometimes occurs in softwoods during drying [51], thus severely reducing permeability. White [52] noted greater penetration of phenol-resorcinol into earlywood than latewood cells of southern pine.
1.8 Influence of Resin Type and Formulation on Penetration
The penetration of adhesives in wood is influenced by the resin viscosity, its MW and MW distribution, resin solid content, and the surface tension. It has to be kept in mind that the curing process may lead to a change in the above characteristics (e.g., viscosity, MW, etc.). Hence, penetration may be influenced during the curing process. Additives that influence the curing behavior may affect the penetration [44].
Hse reported a correlation between penetration and contact angle for PF and southern pine wood [53]. The author employed 36 formulations to determine the contact angle, and its influence on cure time, heat of reaction, plywood shear strength, percent wood failure, bondline thickness, and cure shrinkage. Penetration was not measured, but assumed to be inversely proportional to bondline thickness (thickness of cured adhesive between the veneers). Penetration increased with increasing caustic content. There were no clear trends observed for penetration in relation to adhesive solids content or formaldehyde–phenol mole ratio.
In the case of powdered adhesives, such as powdered PF used in OSB manufacture, it has to melt before penetration. Johnson and Kamke [54] noted that powdered PF resin remained on the surface of wood strands during the blending process and was able to flow and penetrate only during steam injection hot-pressing.
Frazier et al. [55] noted that low MW of pMDI resin would promote penetration into wood cell walls. They further hypothesized that the MDI forms an interpenetrating network of polyurea and biuret linkages within the cell wall. Swelling of the cell wall by pMDI was also observed [48].
The use of resin with low MW components has the potential for deeper penetration than that with high MW. Stephen and Kutscha separated a commercial PF resin into two MW fractions (approximately ±1000 MW) [56]. They observed the penetration of the resin by placing a drop of resin on the surface of aspen. The authors reported that no penetration through the wood was observed for the high MW fraction of the resin, while a penetration of one to two cells deep occurred for the low MW PF fraction. The addition of NaOH improved the penetration of the high MW fraction, which the authors assumed was due to swelling of the cell wall by the NaOH. The improved penetration may also have been due to a lower viscosity as a result of NaOH addition [57]. Gollob et al. reported that PF bond performance was associated with penetration in Douglas-fir plywood [58]. They noted that higher MW formulations tended to dry out and had little penetration.
Zheng studied the penetration of the blends of MDI and PF into yellow-poplar and southern pine [59]. The penetration of the adhesive blends was characterized by a phase separation, with pMDI penetrating deeper. PF tended to bulk the lumens and remain at the interface of the bondline. In general, the blends resulted in less penetration than either of the neat resins. The author attributed the reduction in penetration to increased MW, and subsequent increased viscosity, due to the formation of urethane bonds between the PF and the PMDI.
MW distribution of resin systems will impact their ability for cell wall penetration. Laborie [32] reported evidence of cell wall penetration for two PF formulations, one that had a number average Mn of 270 and a weight average Mw of 330. The other PF had Mn and Mw values of 2840 and 14,200, respectively. The more highly condensed PF resin had a broad MW distribution, including a low Mw component that was similar to the low MW PF resin. Using dynamic mechanical analysis, the author concluded that both resin systems penetrated the cell wall.
1.9
