Introduction

With increasing use of lumber for exterior applications the concerns about insect attack, fungal degradation, or weathering lead to treatment of lumber with various wood preservatives. Thus, woods treated with waterborne preservatives are being used for various exterior applications such as structural and non-structural assemblies. An enhanced adhesion of treated woods to bond treated lumber and their greater durability obviously increase the use of treated woods for exterior applications.

In general, glued wood products for structural applications include laminated veneer lumber (LVL), laminated strand lumber (LSL), parallel strand lumber (PSL), and glulams. The manufacture of these products essentially requires the use of wood adhesives. Exterior grade wood adhesives such as phenol–formaldehyde (PF) resin, resorcinol–formaldehyde (RF) resin, or melamine–formaldehyde (MF) resins are mainly used due to their high moisture resistance in outdoor environment. In particular, the use of woods treated with preservatives for the manufacture of these products requires assessment of the performance of wood adhesives.

Most of the research on adhesive performance of treated woods concern copper–chrome–arsenic (CCA) treated woods (Calusen et al. 2001; Li et al. 2004; Mengeloglu and Gardner 2000; Munson and Kamdem 1998; Hong et al. 1997) and reported inferior strength of CCA-treated woods (Sellers and Miller 1997; Vick 1980, 1994, 1995, 1997; Vick and Kuster 1992; Vick et al. 1996). Many factors contribute to the development of insufficient adhesive strength of treated woods. The presence of contaminants such as waxy, oily, and inorganic materials hinders the development of cohesive adhesion bonds between wood substrate and adhesive (Pizzi 1994). The insoluble metallic components of wood preservatives, for example, impede the formation of interfacial adhesion between wood and adhesives leading to a poor adhesive strength in products (Vick 1997). The effects of CCA treatment on the curing behavior of PF resin in the presence of CCA-treated wood have been investigated using differential scanning calorimetry (DSC) (Vick and Christiansen 1993). Treated wood often contains many defects such as twist, checks, or splits on its surfaces after its drying and curing processes. Furthermore, the formation of effective adhesive bonds in treated woods is interfered by the preservative compounds present on the surface of treated woods (Vick 1994). Using a scanning electron microscopy (SEM) coupled with energy-dispersive X-ray analysis (EDXA), Vick and Kuster (1992) showed that the inside of cell lumen of CCA-treated southern pine was completely covered with a mixture of chrome, copper, and arsenic, which prevented bond formation between wood and adhesives. In addition, one attempt to improve adhesion of CCA-treated wood was to use hydroxymethylated resorcinol (HMR), which led to an increase in adhesive strength (Hong et al. 1997; Munson and Kamdem 1998; Vick 1997).

However, there were positive results regarding the adhesive performance of treated woods. It was reported, for example, that bivalent metallic ions present in CCA could accelerate the reaction of formaldehyde to phenol while trivalent metallic ions retarded the same reaction (Pizzi 1979). Trivalent chrome ions present in CCA could form stable chelates with RF or PF resins, which retarded the reaction of formaldehyde to phenolic nucleus. In addition, the ions of chrome, copper, and arsenic present in CCA could play a catalytic role in accelerating its curing reaction while no acceleration reaction occurred after the ingredients of CCA were chemically fixed in the wood cell wall (Vick and Christiansen 1993). When bonded with phenolic adhesives, CCA-treated woods could have strong internal bond (IB) strength providing good durability, which was proportional to the depth of adhesive penetration into the wood (Vick and Kuster 1992).

The use of alternative preservatives to replace CCA as wood treatment has increased in recent years. For example, the production of laminated lumber treated with alternative preservatives such as ammoniacal copper quat (ACQ), copper azole (CuAz; CY), and alkyl ammonium compound (AAC) was feasible to obtain sufficient bond strength (Junko et al. 1999; Junko and Takato 2002a, b, 2003). The authors showed that the surface properties of treated woods depend to a great extent on the preservatives used, and further processing like surface planing, incising, etc. heavily affected the adhesive wetting and strength.

Therefore, this study attempts to evaluate the adhesion performance of four wood adhesives such as urea–melamine–formaldehyde (UMF) resin, MF resin, PF resin, and RF resin for two softwoods treated with three preservatives such as CCA, CB-HDO, and CY.

Materials and methods

Materials

Logs of two softwood species, Korean pine (Pinus koraiensis Sieb) and Japanese larch (Larix leptolepis), were sawn to obtain lumber with quarter-sawn surfaces, following air-drying for 6 months. Small wood strips of 30 cm×200 cm and 5 cm thickness were planed for specimen preparation. About 1,000 specimens of 15 mm×25 mm×30 mm were prepared for preservative treatments and gluing with adhesives.

Preservative treatment and determination of its retention levels

In this study, four wood preservatives, including CCA, copper azole (Tanalith CY), and CB-HDO were used. Both CCA and copper azole (Tanalith CY) were obtained from the Koppers Arch (North Sydney, Australia) and CB-HDO (trade name CX-10) from Dr Wolman GmbH, Sinzheim, Germany. Specimens with a density of about 450–500 kg/m3 without visible defects were selected for pressure impregnation (full-cell method) of preservatives at a level of 3% (w/v). Treated specimens were cured at 25°C and 75% relative humidity for at least 3 weeks. Characteristics of these preservatives are shown in Table 1.

Table 1 Ingredient compositions of the preservatives used in this study
Full size table

In order to avoid the destruction of wood blocks, the gauge retention of each specimen treated with a preservative prior to adhesive bonding was determined by measuring the weight gain of treated wood blocks after drying at 60°C for 48 h following preservative treatment, which was a similar procedure to the one used by Lebow et al.(1999).

Adhesives and gluing of treated woods

Wood adhesives used were UMF resin, MF resin, PF resin, and RF resin. These commercial grade adhesives were supplied from a local resin company in the Republic of Korea. The properties of these adhesives are shown in Table 2. About 3 g of the adhesives were dried for 3 h at 105°C to obtain non-volatile solid contents. The viscosity was measured with a Brookfield viscosity meter using spindle #2 at 60 rpm.

Table 2 Characteristics of wood adhesives used in this study
Full size table

Prior to gluing, the specimens were further dried in a drying oven at 60°C for 2 days to obtain a moisture content (MC) of about 4–6%. The prepared specimen blocks were glued with adhesives at a loading of 100 g/m2 for both surfaces and then clamped in a steel jig under a pressure of 785 kPa at room temperature for 24 h.

Dry shear strength determination

In order to compare the adhesion strength of wood adhesives used, a compressive dry shear test was done according to the procedure specified in the American Standard (ASTM D 905 1994). For the compressive shear test 25 specimens were used for each combination of adhesive and preservative. These samples were tested in compression using an universal testing machine (Model H50K-S, Hounsfield, UK) with the crosshead speed of 2 mm/min. An average dry shear strength with 25 replications for each experimental unit was reported.

Results and discussions

The determined preservative retentions are summarized in Table 3. The retention levels are greater for CCA-treated pinewood. Lower retention of the CCA in larch wood could be attributed to inherent characteristics of larch wood. This result is compatible with another report (Kang et al. 1995), but the retention levels of CY and CB-HDO were not significantly different between the two wood species.

Table 3 Preservative retentions (kg/m3) of treated wood samples prepared
Full size table

In order to compare the performance of adhesives used, dry shear strengths of preservative-treated and adhesive-bonded samples are shown in Fig. 1. Average values of shear strength for the combinations of wood species, preservatives, and adhesives are also summarized in Table 4. Figure 1a shows dry shear strengths of both untreated and treated woods when they were bonded with UMF resin. When the untreated wood blocks were bonded with the UMF resin, dry shear strengths of both pine and larch samples were 1.8 and 1.63 MPa, respectively. For pinewood specimens, the dry shear strengths of the untreated, CY-treated, and CB-HDO-treated samples were not statistically different from each other. While that of the CCA-treated sample was lower than that of the untreated sample.

Fig. 1
figure 1

Dry shear strength of treated woods bonded with different wood adhesives

Full size image
Table 4 Dry shear strengths of untreated and treated woods depending on types of preservatives and adhesives used
Full size table

Statistical analysis was done to compare adhesive strengths of each group of wood species for each wood adhesive type. Dry shear strengths of larch wood specimens were not statistically different for CCA- and CB-HDO-treated woods when they were bonded with the UMF resin. When the UMF resin was used as samples of both species, the maximum dry shear strength was obtained for the CY-treated wood. This might be due to the amine groups in CY, which accelerate the curing rate of amino resins (i.e. UMF and MF resins) in their polymerization for adhesive bond formation. In fact, the addition of amines into UF resin accelerates a cross-linking reaction (Meyer 1979).

When treated with CCA, the specimens of the two species showed a decrease in their dry shear strengths. This result is compatible with other results reported (Vick 1980, 1994; Vick and Kuster 1992). As shown in Table 3, a greater strength of a larch specimen treated with CCA than the counterpart of pinewood could be attributed to lower CCA retention for this species (Kang et al. 1995). As two possible reasons for lower preservative retention large heartwood percentage and smaller radius of pit-pore in the cell wall were reported (Wang and DeGroot 1996).

Figure 1b shows dry shear strengths of treated wood blocks when they were bonded with MF resin. A greater strength was found for all treated samples than those of untreated samples for both wood species. All dry shear strengths of treated pinewood were greater than those of treated larch wood. The CY-treated samples showed relatively greater strength than those of other preservatives for both wood species. This result can also be explained in an enhanced curing rate of amino resins (i.e. UMF and MF resins) due to the large amount of amines in the CY (Meyer 1979).

Figure 1c shows dry shear strengths of treated woods when they were bonded with PF resin. For pinewood, the greatest shear strength was found when treated with CB-HDO while the CCA-treated samples showed the smallest shear strength. In general, the PF resin requires high temperature and low moisture content for its cure. The greatest shear strength of CB-HDO-treated samples could be attributed to the cure acceleration of carbonates present in the CB-HDO (Table 1). It was already shown that the addition of carbonates into PF resin accelerated the cure of PF resin under alkaline condition (Park and Riedl 2000). However, lower shear strength of CCA-treated pinewood could be attributed to copper, which interfered the curing by reacting with methylol groups (CH2OH). In addition, a strong acid level of CCA could retard the cure of PF resin when bonded with CCA-treated pinewood. When the larch wood samples were bonded with PF resin, the shear strength of the untreated samples was greater than those of all treated samples. The shear strength of CY-treated samples was comparable with that of untreated ones.

Figure 1d shows shear strengths of treated wood samples when they were bonded with RF resin. The shear strengths of samples bonded with RF resin were much greater than those of other adhesives. The pinewood samples treated with CB-HDO showed the greatest shear strength, followed by CY-treated and CCA-treated samples in descending order. As for PF resin, the presence of a relatively large amount of carbonate in the CB-HDO could be a possible cause for this result. In other words, the carbonate present in the CB-HDO played a role in accelerating the cure of RF resin used for treated pinewood samples.

The shear strengths of treated pinewood samples were lower than that of untreated ones. This might be due to the presence of copper in all preservatives used in this study. It was reported that the copper present in copper-based preservatives had an influence in retarding the cure of RF resin, resulting in lower adhesive strength (Sellers and Miller 1997). However, all dry strengths of treated pinewood samples were not statistically significant for the different preservatives.

All dry shear strengths of treated larch wood were lower than those of treated pinewood and also lower than that of the untreated larch wood. This result indicated that the copper present in all preservatives influenced the shear strength of treated larch wood. The CB-HDO used for the treatment of larch wood showed the best shear strength among the preservatives although its shear strength was relatively small. It was believed that the carbonate present in the CB-HDO played the same role for treated larch wood.

The greatest shear strength of treated woods bonded with RF resin among four adhesive types could be attributed to greater reactivity of the RF resin (i.e. cold setting adhesive). Differences in shear strengths could be ascribed to various factors such as types of preservatives, their chemical composition, pH levels of preservative and wood species, surface properties, extractive content of each wood species, etc. The presence of extractives dissoluble in ethanol and water, for example, accelerated the cure of UF resin used for preparing particleboards (Chen and Paulitsch 1974).

The shear strength of treated wood showed differences depending on wood species as well as adhesives types, although the samples were treated with the same preservatives. This result suggests that other factors rather than the ingredients of preservatives affect adhesion strength of treated wood. For example, lower pH of CCA could have an impact on the curing of all adhesives except MF resin. Also, greater pH levels of CB-HDO and CY could accelerate the curing of MF resin, PF resin, and RF resin except UMF resin that cured under acidic condition.

In summary, the results obtained from this study show that both UMF and MF resins produce relatively strong adhesive strength for CY-treated pine and larch woods. When larch wood was treated with CY, PF resin also produced good bond strength. In addition, the combination of CB-HDO and PF resin was good for developing bond strength of pinewood. RF resin gave the best result when CB-HDO was used for the treatment of both wood species. However, to use the treated woods for exterior applications, their water resistance should be evaluated by measuring wet strength. The above results also indicate that many factors as well as interactions of the factors impact the development of adhesive bond strength in treated woods.

Conclusions

In order to extend the use of treated woods for exterior applications, this study attempts to evaluate the performance of wood adhesives for two softwood species, pine and larch lumber, treated with traditional and alternative preservatives such as CCA, CY, or CB-HDO using UMF resin, MF resin, PF resin, and RF resin. The following results are obtained from this study:

  1. 1.

    Both UMF and MF resins produced relatively strong adhesive strength for CY-treated pine and larch woods.

  2. 2.

    The PF resin was also good for developing bond strength of pinewood treated with CB-HDO or CY. RF resin gave the best result when CB-HDO was used for the treatment of both wood species. These results indicated that alternative wood preservatives could be replaced with traditional CCA.

  3. 3.

    A proper combination of adhesive, preservative, and wood species should be selected for the best bond strength development, which was affected by the characteristics of these three parameters as well as their interactions.