This study investigates the bonding properties of modified wood by considering three different aspects: water related characteristics, mechanical performance and optical (fluorescence microscopy and X-ray micro-computed tomography) observation of adhesive penetration into modified wood structure. In recent years, the new wood modifications have become more commercially available in the market for both exterior and interior applications due to improved properties that modification can bring to the wood e.g. the improved biological durability, dimensional stability, hardness and weathering resistance of the wood as well as the environmentally friendly nature of the wood modification processes (Militz and Hill 2005).
Besides these advantages, modification can affect some technological aspects of the wood such as its bonding performance. For example, it can alter the strength of adhesion as a result of changes in chemical, physical and structural characteristics of the wood. For example, the less polar and less porous modified wood surfaces can result in reduced adhesion due to formation of less free OH groups for bonding leading to poorer adhesive wetting of the wood surface and weaker chemical bonds between the two adherents (Hunt et al. 2007). As modified wood becomes a more demanded material for different applications, there is a need to study its bonding performance where the challenge is to bond different modified materials as their physical and chemical characteristics are substantially changed by modification.
In this thesis, measurements of capillary water uptake, contact angle and surface energy were used to determine the water related properties and hydrophobic behavior of furfurylated (FA40 and FA70, which represent 65 and 75 % WPGs) and N-methylol melamine (NMM) (10, 20 and 30%) modified Scots pine and thermally treated Scots pine and beech (modified through an industrial scale vacuum press dewatering method at 195 and 210 °C). The capillary water uptake results indicated a considerable reduction of water uptake for all modifications in all directions both after short (24 h) and long contact times (168, 336 h). Contact angle measurement data revealed an increased hydrophobicity of modified wood. However, some exceptions were observed, mainly for thermally treated wood. Modifications provided radial and tangential surfaces with a non-polar character. Penetration of adhesives into the wood structure plays an important role in the production of glued wood-based panels and products by affecting the bond quality (Frihart 2005, Kamke and Lee 2007). The gross penetration of emulsion polymer isocyanate (EPI), polyurethane (PU) and polyvinyl acetate (PVAc) adhesives into modified wood, both with and without pressure, were determined by using fluorescence microscopy based on measurements of effective (EP) and maximum penetration (MP). Without application of pressure, the EP of EPI adhesive reduced after NMM modification and furfurylation (FA70) and also PU adhesive after NMM modification while the EP of PVAc adhesive increased into furfurylated and NMM modified (10 and 20%) wood. For thermally treated Scots pine, increasing the treatment temperature improved EP of all adhesives. Among used adhesives, PU penetrated much deeper into thermally treated wood for both treatment temperatures. Comparison of penetration of adhesive with and without pressure revealed that with the exception of EP of PU and EPI adhesives into NMM-modified wood and PVAc into thermally treated beech at 195°C, application of pressure led to rather different results as compared to the EP data when no pressure was applied. Visual observation and analysis of fluorescence microscopy photomicrographs provided more detailed information on modality of penetration. Due to the large and deep penetration of PU adhesive into thermally treated Scots pine observed in both studies (with and without pressure), the 3D pattern of penetration of this adhesive was obtained by X-ray micro- computed tomography indicating the pathways which were used by this adhesive for penetration. In another study, the bonding shear strength of the same modified wood materials glued with the same adhesives was also investigated. For all adhesives used, the shear strength significantly reduced after furfurylation and NMM modification of Scots pine samples, mainly due to the brittle nature of the wood after modification rather to the failure of the bondline. Bonding strength of both Scots pine and beech was also negatively affected by thermal modification and the bondline was found to be the weakest link in thermally modified wood. The EP of adhesives and the bondline thickness did not relate to the shear strength of all modified wood materials. It was indicated that the lower shear strength of modified wood could be attributed to other factors, such as the decreased chemical bonding or mechanical interlocking of adhesives, and the reduced strength of brittle modified wood substrate.
The effect of two important bonding variables, wood moisture content and open assembly time on penetration of PU adhesive into thermally modified wood (195 and 210 °C) was also studied. The equilibrium moisture content (EMC) level of 8.6% was found to be the optimum for an effective penetration of PU adhesive in thermally modified Scots pine treated at 195°C. In most of the cases, penetration of PU adhesive did not change significantly by increasing the open assembly time, which suggested using a shorter open assembly time of 15 min than 30 min for bonding of thermally modified Scots pine with PU adhesive, in order to save time and reducing the production costs. For samples treated at both treatment temperatures and after shorter open assembly time, the highest MP values observed at moderate EMC levels of 8.6 and 8.2% and the lowest at the higher EMC levels of 13.2 and 12.5%.
In another study, the effect of phenol formaldehyde (PF) treatment on bonding performance of beech glued with PVAc and phenol resorcinol formaldehyde (PRF) adhesives was also investigated. The results of both dry and wet conditions indicated higher shear strength for samples bonded with PRF than PVAc. With the exception of 25% PF treated wood bonded with PVAc, the PF modified wood can be glued with both adhesives satisfactorily under dry condition, while under wet condition only the 25% PF modified samples bonded with PRF provided acceptable bonding. For both adhesive systems, PF modification caused a reduction of adhesive penetration into wood structure, especially in the case of higher load treatment.
The development of bonding strength of modified birch veneers glued with hot curing phenol formaldehyde (PF) adhesive was investigated in different pressing (20 s , 160s) and open assembly times (20s , 10 min). Generally, the bonding strength improved by extending the pressing time. In 20 s pressing, increasing assembly time did not change the bonding strength in most of the cases while at 160 s pressing, prolongation of assembly time developed a better bonding for controls, NMM modified and thermally treated veneers at 180°C. The combination of 10 min assembly time and 160 s pressing time provided the highest bonding strength for controls, NMM modified and thermally treated veneers at 180°C while furfurylated samples achieved the highest values in 20 s assembly and 160 s pressing times. In general, modification affected negatively the bonding performance of the veneers, especially for furfurylated and NMM modified samples. In General, the overall results obtained in this thesis showed that modified wood has lower bonding ability and performance than unmodified wood as result of the decreased water related properties, less penetration of adhesive into wood structure and decreased bonding strength after modification. However, the increased dimensional stability and low water uptake of modified wood might lead to better performance in long term.