Synthesis of Bio - Oil-phenol-Formaldehyde Resins under Alkali Conditions: Physical, Chemical and Thermal Properties of Resins and Bonding Performance

• In the present study, bio-oil produced from vacuum pyrolysis of woody biomass has been investigated as a source of chemical feedstock. Bio-based resins were produced using the bio-oil with phenol substitutions ranging from 10 to 30 wt%. The conventional GC/MS analysis was carried out for the evaluation of the chemical composition of bio-oil. TGA, DSC and FT-IR analyses were used in order to characterize the bio-oil-phenol-form-aldehyde (BPF) resins. The bonding quality of wood samples bonded with the BPF resins was investigated under different pre-treatment conditions. The highest shear strength was observed for the control samples bonded with the laboratory PF resin. As the amount of bio-oil was increased up to 30 wt%, the shear strength of the samples decreased from 12.08 to 11.76 N/mm 2 . The bonding performance was not negatively affected by the combination of bio-oil under dry conditions. According to TS EN 12765 standard, the relevant performance requirements for bonded samples under dry conditions must be at least 10 N/mm 2 . Relating to the standard, all samples bonded with BPF resins obtained the requirements for durability class C1. Under wet conditions, the bonding performance was negatively affected by the addition of bio-oil. However, the BPF resins fulfilled the durability requirements for C1, C2, and C3 specified in EN 12765 (2002).


INTRODUCTION 1. UVOD
Currently, the global demand for chemicals is supplied largely from petroleum-based products. Because of the growing concern for the significant environmental problems and reduction of availability of fossil energy resources, researchers have been studying renewable feedstocks to replace petroleum. There are many similarities between biomass and crude oil in terms of their use as chemical feedstocks. Biomass offers viable greener, lower cost alternative feedstock to the chemical production (Gosselink et al., 2004;Bozell, 2002;Yanik et al., 2007;Connor and Piskorz, 1994).
Phenol-formaldehyde resins (PF) are cross-linkable polymeric resins made from phenol and formaldehyde in the presence of a catalyst. Depending on the pH of the catalyst, these monomers react to form one of two main resin types: resol and novolac reins. A basic catalyst is generally used in formulating the resol type of resins. They are used successfully for the production of wood composites, because of their high bonding strength and extreme resistance to moisture (Junming et al., 2010;Christjanson et al., 2010;Moubarik et al., 2009). They are, however, relatively expensive as regards the phenol price. Therefore, there have been many attempts to replace phenol with renewable resources, such as lignin, soybean, tannin or bio-oil, etc. in the production of phenol-formaldehyde (PF) resins In the present study, bio-oil produced from vacuum pyrolysis of woody biomass has been investigated as a renewable chemical feedstock. The chromatography− mass spectrometry (GC/MS) analysis was carried out for the evaluation of the chemical composition of biooil. The bio-oil was chemically synthesized with phenol and formaldehyde at 10 wt%, 20 wt%, and 30 wt% phenol replacement levels under alkaline conditions. BPF resins were characterized by several thermal and chemical analyses such as TGA, DSC, and FT-IR. In addition, the shear strengths of wood samples were investigated under both dry and wet conditions.

Biomass and chemicals 2.1. Biomasa i kemikalije
In this study, pine sawdust was used for the pyrolysis process as woody biomass. Prior to the experiments, the raw material was dried and ground. The particles with the size of 1-2 mm were selected for the experiments.
A commercial PF and laboratory PF resins were considered as reference resins in this work. The commercial PF adhesive was provided by POLİSAN chemical company in Izmit, Turkey. BPF resins were prepared by using phenol (liquid) and formaldehyde aqueous solution (37 %) supplied by GENTAŞ chemical company, İzmit, Turkey. Sodium hydroxide pellets were purchased from Sigma Aldrich.

Bio-oil production from biomass 2.2. Proizvodnja bioulja od biomase
The pyrolysis experiments were implemented in a vacuum pyrolysis reactor having 24 cm × 36 cm interior diameter and depth, respectively. The reactor was electrically heated externally. During the pyrolysis process, pyrolysis temperature was operated with a Proportional-Integral-Derivative controller. The total pressure was fixed at about 10 kPa. For the pyrolysis; 2000 g sample of biomass was placed into the reactor and the experiments were performed with a heating rate of approximately 15 °C/min to the final temperatures of 500 °C and held for 60 min. The bio-oil was collected by condensation in a condenser connected to the reactor exit. The bio-oil was filtered through the filter paper with a mesh size of 75 μm.

Chemical properties of bio-oil 2.3. Kemijska svojstva bioulja
The chromatography−mass spectrometry (GC/ MS) analysis was carried out for the determination of the chemical composition of bio-oil (Agilent 6890). Helium was used as the carrier gas. The oven temperature was set to start at 40 °C and gradually raised to the final temperature of 300 °C.

Chemical synthesis of resins 2.4. Kemijska sinteza smola
The reactions were performed in a glass reactor equipped with a stirrer, a reflux condenser, and a heating part. In a typical synthesis of a laboratory-made PF resin (lab. PF), the reactor was first filled with 1000 g of phenol and 1100 g formaldehyde (37 wt%). The temperature was increased to 60 °C and then 62.5 g of NaOH solution (50 wt%) was added to the process. The blend was mixed and heated up to 90 °C, and then held at that temperature for 90 min. Finally, the second part of NaOH (50 wt%.) solution was charged at a temperature of 60 °C. After the reaction, the process was cooled to room temperature. The same procedure was applied for bio-oil/PF resins synthesis. By varying the amount of bio-oil in the reaction mixture, a series of phenol substitutions were synthesized ranging from 10 to 30 wt%.

Physical characterization of resins 2.5. Određivanje fizičkih svojstava smola
The pH and dynamic viscosity values were determined with a digital pH meter and a rotational viscometer at 25 °C, respectively. Gel times were calculated at 100 °C. The solid contents of the resins were determined by weighing 2 g of the resin in a glass plate. The glass plate was placed in an oven at 120 °C for 120 min and it was desiccated and weighed to find the weight of dry resin remaining.

Termogravimetrijska analiza (TGA)
The thermal decomposition behavior of resins was analyzed by using a HITACHI STA 7300. For the thermogravimetric analysis (TGA), nearly 10 mg of the sample was heated at a 5 °C/min heating rate up to the final temperature of 700 °C. The carrier gas was nitrogen with a flow rate of 25 ml/min.

Differential scanning calorimetry analysis (DSC) 2.7. Diferencijalna pretražna kalorimetrija (DSC)
DSC analyses were carried out using a DSC7000X. About 5 mg of samples (resin) was applied to a stainless-steel crucible. Scans were performed at a temperature of 50 -550 °C at a 10 °C/min heating rate.

Fourier-transform infrared spectroscopy (FT-IR) 2.8. Fourierova transformacijska infracrvena spektroskopija (FT-IR)
FT-IR spectra were recorded on an Alpha FT-IR spectrometer by direct transmittance using KBr pellet technique. 10 scans were recorded for each sample with a resolution of 4 cm -1 over a range from 4000 to 400 cm -1 .

Sample preparation and property testing 2.9. Priprema uzoraka i ispitivanje svojstava
The beech wood (Fagus orientalis Lipsky) planks provided from a commercial company in Karabuk, Turkey were used as wood material. The lamellas (5 mm × 20 mm × 150 mm) were prepared from the planks for shear strength tests. Prior to tests, the lamellas were conditioned (20 °C and 65 % relative humidity) to reach the equilibrium moisture content of 12 %. The resin was implemented on one surface of the lamella using a hand brush at a rate of 180 g/m 2 as suggested by the manufacturer. The pressure, temperature, and duration were set as 0.2 N/mm 2 , 130 ºC, and 15 min, respectively. Bonded samples were conditioned for 7 days in the standard climate room and then cut into samples. The dimensions and shape of a typical sample are shown in Figure 1.  Classified samples were divided into three subgroups for different pre-treatments, prior to the evaluation according to TS EN 12765 standard. Table 1 summarizes the proposed criteria for evaluating the quality of bonds according to this standard. The first sample groups (pre-treatment 1) was evaluated in dry conditions after conditioning for 7 days (65±5 % relative humidity, 20±5 °C standard climate); the second group of samples (pre-treatment 2) was soaked in cold water (20 °C) for 24 h; the third group of samples (pre-treatment 3) was boiled for 3 h, then cooled in water (20 °C) for 2 h. The determination of shear strength was performed in a universal testing machine, according to BS EN 205. Table 2 presents the chemical characterization of the bio-oil. The bio-oil consists mainly of oxygenated organics (aldehydes, ketones, phenols, benzenes, alcohols, and polycyclic aromatic hydrocarbon). It was found that phenols were the main component in the bio-oil. The most abundant phenols in the bio-oil include phenol, 2-methyl-phenol, 4-methyl-phenol, 2-methoxy-phenol, 2,5-dimethyl-phenol, and 2-methoxy-4-methyl-phenol. The compounds found in the bio-oil correspond well to the literature (Ingemarsson

Karakterizacija smola
The results of the physical properties of the resins are displayed in Table 3. The pH values of all resins were similar due to the fact that pH values were continuously adjusted during the synthesis process. The viscosities of the BPF resins were higher than those of reference resins, and also, gradually increased with the increasing addition of bio-oil. This result may be due to the fact that high molecular weight compounds in the bio-oil cause high viscosity in bio-based resins. The solids content of 10-30 % BPF resins were 44.32 wt%, 46.61 wt% and 46.76 wt%, respectively. The solid content of the laboratory PF resin (40.67 wt%) was lower than that of BPF resins and the solids content increased as the bio-oil per- The TGA curves of the resins (i.e. the bio-oil modified/commercial/lab.) are presented in Figure 2. The thermal behavior of the commercial PF resin was different from that of the laboratory PF resins and the BPF resins after the temperature was higher than approx. 150 °C. This higher thermal degradation is due to the fact that commercial PF resin contains an amount of urea decomposing easily at high temperatures. It was observed that the lab. PF resin and all of the BPF resins had a similar trend of thermal behavior. The highest thermal degradation of the BPF resins was determined in the 30 % bio-oil PF resin. In the literature, thermal degradation of biobased PF resins obtained from the different renewable feedstocks were a rough match to our results Domínguez et al., 2013;Zhao et al., 2013). Figure 3 shows the DSC thermograms obtained at a heating rate of 10 °C/min for all resins from 50 to 550 °C. As shown in Figure 3, the DSC thermograms of all resins had a multiple peak structure. The main endothermic peak was detected between 125 and 136 °C for all resins. The formation of these peaks might be connected to the condensation of phenol or bio-oil with methylol groups in order to create a methylene bridge and the condensation of two methylol groups in order to create dibenzyl ether bridges and the cross-linking condensation of the methylene bridges (Christiansen and Gollob, 1985;Gabilondo et al., 2007;). All modified resins showed the second peak at a lower temperature than commercial PF. DSC thermograms obtained in this study were consistent with the earlier research of thermal behavior of bio-based adhesives (Khan et al., 2004;Cheng et al., 2012;Lin et al., 2014).
BPF resins were fractionally and structurally characterized by FT-IR spectroscopy; the 4000-400 cm -1 infrared spectral region is shown in Figure 4. The reference resins (commercial and lab. PF resins) were also analyzed as a comparison.
The O-H stretching vibrations between 3600 and 3200 cm -1 represent the existence of phenols. The peak around 1600 cm -1 was assigned to the aromatic ring vibrations, indicating the existence of phenols in resins. A new peak was observed at around 1650 cm −1 , attributed to C=O stretching, indicating carbonyl compounds in the BPF resin. All the resins have the aromatic rings around 1450 cm −1 , phenolic C-O stretching around 1200 cm −1 , C-O stretching around 1000 cm −1 . The peaks around 900 cm −1 , 800 cm −1 , and 750 cm −1 were related to the bending vibration of C-H in the aromatic rings. Cui and co-workers reported that the BPF resins have similar FT-IR spectra as PF resin. This indicates that the BPF resins display a similar molecular structure as the PF resin (Cui et al., 2016). According to FT-IR analysis, it can be claimed that bio-based resol resins were successfully produced by polymerization of phenol, bio-oil and formaldehyde.

Bond performance 3.3. Svojstva lijepljenog spoja
The shear strength and wood failure of the samples bonded with the BPF, lab. PF and commercial PF resins are given in Table 4. Under dry conditions, the percentage of wood failure was almost 100 %, except for 30 wt% modified BPF resin that was a bit lower (97 %). According to Table 4, the bonding performances of all resins are satisfying. The highest shear strength was observed for the control samples bonded with the lab. PF resin. As the amount of the bio-oil was increased up to 30 wt%, the shear strength of the samples decreased from 12.08 to 11.76 N/mm 2 . The bonding performance was not negatively affected by the combination of bio-oil under dry condition. According to TS EN 12765 standard, the relevant performance requirements for bonded samples under dry condition must be at least 10 N/mm 2 . Relating to the standard, all samples bonded with BPF resins obtained the requirements for durability class C1.
The shear strength and wood failure values of all samples after 24 h soaking in water (pre-treatment 2) are presented in Table 4. The shear strength of wood bonded with the reference PF resins was higher than that of BPF resins. The shear strength was gradually decreased with increasing concentration of the bio-oil. For pre-treatment 2, the reference resins had the highest percentage of wood failure (100 % -97 %). All the BPF samples showed a lower percentage of wood failure (especially samples bonded with 30 wt% BPF resin), The BPF resins containing 10-30 wt% bio-oil fulfilled the requirements for durability class C2 (EN 12765).
For pre-treatment 3, similar trends were observed as for pre-treatment 2. The strength values of the samples decreased after pre-treatment 3 as compared with pre-treatment 1 and 2. The bond line was more or less negatively affected by the presence of the bio-oil. As   the amount of bio-oil in the PF resin was increased by 30 wt%, the shear strength of the samples after pretreatment 3 decreased from 5.78 to 4.43 N/mm 2 . However, these samples exceeded the minimum requirements for the durability class C3 (EN 12765). Cheng et al. (2012) reported that the methylolated bio-oil-phenol formaldehyde (MBPF) adhesives were more sensitive to water than the PF adhesive. They could be used as adhesives for the production of interior wood products. It can be concluded that renewable resources such as bio-oil have potential as a substitute for PF resin.

ZAKLJUČAK
In this work, bio-based PF resins were produced by substitution of phenol with the bio-oil at varying ratios (from 10 wt% to 30 wt%) in the synthesis of phenol-formaldehyde resins under alkaline condition. The following conclusions can be obtained: The bio-oil produced from woody biomass was composed mainly of phenols, aldehydes, ketones, alcohols and, benzenes. The phenols accounted for the largest amount of compound in the bio-oil.
The viscosities of the BPF resins were higher than those of reference resins and gradually increased with the increasing addition of bio-oil.
The laboratory PF resin and all of the BPF resins had a similar trend of thermal behavior.
FTIR spectra demonstrated that bio-based resol resins were successfully produced by polymerization of phenol, bio-oil, and formaldehyde.
The shear strength values of the BPF resins were similar to the commercial PF and the lab. PF resins under dry condition. Under wet conditions, the shear strength and wood failure values of the samples bonded with BPF resins exhibited poorer performance as compared to samples bonded with the reference resins. However, the BPF resins fulfilled the requirements for C1, C2, and C3 specified in EN 12765.