Abstract

Adhesive preparation

Manufacture of the adhesive was based on the explanation of Santoso et al. (2014) that air-dried merbau wood sawdust was mixed with water at a 1:4 ratio (wt/wt) and heated at 80°C for 3 hours. The mixture was filtered to separate the liquid extract from the sawdust, and the extraction process was repeated twice, with a final ovendried extractive yield of 6.5 percent (wt/wt) based on the original ovendried weight of the sawdust. The extract was mixed with resorcinol and formaldehyde and called MRF. MRF consisted of two components: resin (merbau wood liquid extract, resorcinol technic, and 40% NaOH) and formaldehyde (37%) as a cross-linker or curing agent. Merbau liquid extract (M) was mixed with resorcinol (R) at weight ratios of 100:5, respectively, and NaOH was added to adjust the pH to 10 to 11. The mixture was added to formaldehyde (F) at 10 percent M (wt/wt) and mixed until homogenous at room temperature, which was followed by a 1-hour conditioning period. The solid content of MRF was 29.5 percent with pH 11.

Reactivity of polyphenolic compounds in the liquid extract with formaldehyde (Stiasny precipitation number) was determined as follows. Twenty grams of dehydrated extract was weighed into a beaker containing 40 mL of water. The solution was adjusted to pH 7 by adding sodium hydroxide pellets. One hundred milliliters of 37 percent formaldehyde and 15 mL of concentrated HCl were added, and the mixture was heated at 90°C for 85 minutes. The resulting solution was vacuum filtered, and the solid residue was washed with hot water, dehydrated, and dried at a temperature of 105°C in an oven until reaching a constant weight. The Stiasny number was determined as follows:

The total phenolic content present in the extract was quantitatively determined by colorimetric assay using the Prussian blue reagent (Akoto and Brefoh 2014). The MRF resin was produced according to Santoso et al. (2014) by copolymerizing the liquid extract with formaldehyde and resorcinol in alkaline conditions to yield 1,3-benzenediol, 4,5-dimethyl-resorcinol.

Glulam manufacture and testing

The sample specimens were two types of five-layer boards, namely, cross-laminated timber (CLT) and glulam. The CLT was prepared with cross-band laminas that were perpendicular to the fiber direction of the face, core, and back laminas (Fig. 1), and for the glulam all laminas were arranged parallel to the fiber direction. Each board was manufactured from a single wood species—pangsor (Ficus callosa), mindi (Melia azedarach), pine (Pinus merkusii), or mangium (Acacia mangium)—and was 70 mm thick, 150 mm wide, and 980 mm long. Each board type was replicated five times for each wood species.

View Figure

• Download as Powerpoint
• Download Figure

The adhesive gluemix was spread at a rate of 150 g/cm², and the laminas were cold pressed for approximately 3 hours at room temperature, followed by conditioning for 7 days. The boards were tested for moisture content, density, shear strength, delamination, modulus of rupture (MOR), modulus of elasticity (MOE), and formaldehyde emission based on the Japanese Agricultural Standard 234 (JAS 2003).

MALDI-TOF analysis

Samples of merbau wood liquid extract were dissolved in water-acetone for analysis (Merck, supplied by VWR, France; 4 mg/mL, 50/50 volume). The solutions were mixed with 2,5-dihydroxy benzoic acid (10 mg/mL in acetone; Laser Bio Labs, France) as the matrix solution, which facilitated the deposition of a sample in the instrument. Calibration was done with red phosphorous (Laser Bio Labs). To enhance ion formation, a concentrated solution of sodium chloride (10 mg/mL in distilled water; Carlo Erba Reactifs [SDS], France) was added to the matrix. The resultant solution contained three parts of the matrix solution, three parts of the sample solution, and one part of the NaCl solution, and 0.5 to 1 liter of it was placed on the MALDI target. The dry droplet sample preparation method was used (Hoong et al. 2010). After evaporation of the solvent, the MALDI target was placed in the spectrometer. The MALDI-TOF spectra were recorded on a Axima Performance instrument (Shimadzu Scientific Instruments, Manchester, UK). The irradiation source was a pulsed nitrogen laser with 3-ns intervals at a wavelength of 337 nm. The measurements were made under the following conditions: polarity, positive; flight path, linear; mass, high (20-kV accelerating voltage), and 100 to 150 pulses.

To obtain the molecular weight of the chemical species of each peak, 23 Da (the molecular weight of Na⁺ from the NaCl in the matrix and attached to the oligomers) must be subtracted from each peak value in the resulting positive mode spectrum.

Measurement of formaldehyde emission

Formaldehyde emissions from the laminated lumber boards were determined using the acetyl acetone method (Nash reagent) at a wavelength of 412 nm (Roffael 1993) according to the flask method of the European Standard EN 717‐3 (European Committee for Standardization [CEN] 1996), and thus the following procedure. A test sample of 25 by 25 by 10 mm was hung by string inside a 500-mL bottle containing 50 mL of distilled water about 1 cm above the water level. Another bottle containing 50 mL of distilled water was used as a control. The bottles were placed in a 40°C oven for 24 hours, and the sample was removed from the bottle. The bottles were put in the room for 30 minutes to increase formaldehyde emission captured, and the water was poured into a volumetric flask. The test setup is illustrated in Figure 2.

View Figure

• Download as Powerpoint
• Download Figure

We next evaluated the amount of formaldehyde emitted by the samples in the water. From a formaldehyde standard solution of 3 mg/liter, we pipetted 0, 5, 10, 20, 50, and 100 mL into six 100-mL bottles, respectively, and then diluted each by adding water up to the 100-mL line. The bottles contained concentrations of 0, 0.15, 0.3, 0.6, 1.5, and 3 mg/liter formaldehyde, respectively. We pipetted 25 mL of sample solution into an Erlenmeyer flask (100 mL) with a stopper, and put 25 mL of acetyl acetone–ammonium acetate into the Erlenmeyer flask and mixed it until homogeneous. The solution was heated for 10 minutes in a water bath at 65°C and then cooled to room temperature. The same procedure was used for the standard solutions, and both solutions were evaluated with a spectrophotometer at the 412-nm wavelength.

Data analysis

To evaluate the effect of product types (two levels) and wood species (four levels) on the boards properties, a completely randomized factorial two by four design was used to analyze the data. The first factor was the product type (namely, CLT and glulam), and the second factor was wood species (namely, pangsor, mindi, pine, and mangium).

MALDI-TOF analysis of merbau wood liquid extract

Figure 3 shows a representative MALDI-TOF spectrum of merbau wood extract with a molecular weight ranging from 100 to 600 Da. The spectrum shows that merbau wood extracts contain flavonoid oligomers up to six repeating units, based on the largest oligomer components detected in the sample, which had a peak at 445 Da. The MALDI-TOF mass spectra showed clear repetitive patterns of peaks that allowed identification of specific oligomer series present in the merbau wood extract sample. The major peak-to-peak mass increments were 133, 227, 297, 315 to 319, 429, and 445 Da, which corresponded to structures (I) resorcinol, (II) stilbene type, (III) fisetinidin, (IV) catechin, (V) fisetinidin-resorcinol, and (VI) catechin-resorcinol repeating units, respectively (Table 1).

View Figure

• Download as Powerpoint
• Download Figure

Table 1 Extraction of merbau wood liquid through maceration at a warm temperature.

View Fullscreen

Table 1 shows the possible combinations of different repeating units and the degree of polymerization corresponding to each. These values were obtained based on the m/z value of flavonoid structures I to VI. The dominant series of peaks at 133, 297, 319, and 445 Da for flavonoid extracts were monomers, dimers, trimers, and tetramers, respectively. The main repeating unit in each oligomer group was identified as those composed of catechin (IV and VI). Chemical compounds of 155 Da or 2-methyl-5-hydroxybenzofuran + Na⁺, 227 Da or stilbene type, 297 Da or fisetinidin + Na⁺, 315 to 319 Da or catechin or robinetinidin multi protonated + Na⁺, 429 Da or fisetinidin-resorcinol adduct + Na⁺, and 445 Da or catechin-resorcinol adduct + Na⁺ are demonstrated in Figures 4 to 9, respectively.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

It is interesting to note that cold-setting adhesives based on flavonoid-resorcinol direct adducts have been prepared synthetically (Pizzi and Roux 1978a) and that flavonoid-resorcinol-formaldehyde coreacted adhesives have already been reported (Pizzi and Roux 1978b) and commercialized. What is extremely new in the present work is that such flavonoid-resorcinol adducts exist naturally in the extract of merbau, rendering such an approach to cold-setting glulam adhesives particularly of interest both technically and economically.

The results support previous research (Santoso et al. 2014) that found merbau wood liquid extract consisted of the functional groups C–H and C=O with absorption bands below 2,850 to 2,750 per cm and 1,820 to 1,600 per cm, which indicated functional aldehyde (CHO) groups. Further analysis by pyrolysis–gas chromatography–mass spectrometry showed that merbau wood liquid extract was dominated by resorcinol compounds, especially 1,3-benzenediol.

Stiasny numbers close to 100 have been found for polyphenol-containing extracts (Food and Agriculture Organization of the United Nations 2000). The higher the Stiasny number, the higher the amount of reactive polyphenols available in the liquid extract for polymerization. The high Stiasny number (92.4%, wt/wt) obtained from the extract in the current study indicates its suitability for the resinification process.

The concentration of polyphenols (expressed as tannic acid) equivalent in the extract was found to be higher than that of phenol. The test for phenolic types yielded a pink color, which indicated the presence of flavan-3,4-ols and therefore the presence of condensed polyphenols, which are characterized as having polymerized flavonoid units. This was in agreement with Hedges (2000), who stated that condensed oligomers are ubiquitous phenolic compounds, representing the second most abundant natural phenolics after lignin and the fourth most common components in vascular plant tissues.

Properties of laminated lumber

Table 2 shows the properties of the glulam specimens of four wood species bonded using the merbau extract adhesive. The moisture content of the products ranged from 10.58 to 11.21 percent. The moisture content fulfilled the guidelines of Japanese Standard JAS 234‐2003 (2003), which requires the allowable moisture content in glulam to be less than 15 percent.

Table 2 Properties of boards.

^a All values are the average from five replicates. Delamination for all samples was 0 percent. Values followed by the same letter are not significantly different. MOR = modulus of rupture; MOE = modulus of elasticity.

View Fullscreen

The shear strength of glulam was higher than that of CLT, indicating that adhesion in glulam was better than in CLT because the adhesive in glulam penetrated the adherents with greater ease in the parallel lumber components. Pangsor wood joints presented the lowest shear strength because this species has the lowest density, and the other wood species presented higher strength values. Mindi and mangium glulam fulfilled the Japanese standard requiring a shear strength of more than 54 kg/cm² (dry test).

The joints shear strength obtained was better compared with similar products (five plies) described in previous research (Supartini 2012). The previous products had a low density (<0.45 g/cm³) and were manufactured from Jabon (Anthocephalus cadamba) and manii (Maesopsis eminii) wood using a commercial isocyanate adhesive. The resulting shear strength was 49 to 59 kg/cm² in the dry test but 0 kg/cm² in the wet test. Similarly, high-density products (>0.56 g/cm³) made from mangium wood had a shear strength of 66 kg/cm² in the dry test and 0 kg/cm² in the wet test.

The MOR of the boards ranged from 63.24 to 547.73 kg/cm² (Table 2). Mangium glulam had the highest MOR, and mindi CLT had the lowest. Pine CLT and glulam and mangium glulam had the highest MOR, which at more than 300 kg/cm² fulfilled the requirements of JAS 234‐2003 (2003). These results were lower than those of Supartini (2012) on five-layer CLT from manii, jabon, and mangium wood (394 to 604 kg/cm²).

The MOE of the boards was lower than those of Supartini (2012), who produced five-layer CLT of manii, jabon, and mangium wood that had MOE values of 49,000 to 74,000 kg/cm². The boards in this study did not fulfill the requirements of JAS 234‐2003 (<75,000 kg/cm²).

The formaldehyde emission of the boards ranged between 0.18 and 1.10 mg/liter, with the result that all the boards were classified into E 2 (more than 0.124 mg/m³ air; CEN 2003). Formaldehyde emission from pangsor glulam (1.10 mg/liter) was not significantly different from the glulam made from mindi (1.04 mg/liter) or mangium (1.10 mg/liter), according to the Japanese Standard the boards were classified into F** category (maximum, 2.1 mg/liter). However, the formaldehyde emission of pangsor glulam was significantly different from that of pine glulam (maximum, 0.7 mg/liter), which was classified in the JAS F*** category. The mindi CLT (0.18 mg/liter) was equivalent to CLT made from pine (0.26 mg/liter) or mangium (0.22 mg/liter) in terms of formaldehyde emission, and thus it satisfies the requirements of the type F**** (maximum, 0.4 mg/liter) of the JAS standard.