A Potential Replacement to Phenol–Formaldehyde-Based Adhesives: A Study of Plywood Panels Manufactured with Bio-Based Wood Protein and Nanolignin Adhesives

Abstract

Plywood production relies on phenol–formaldehyde (PF), which is why bio-based wood adhesives (BBWAs) were developed as potential replacements, showing promising results in several tests performed. A control sample (PLY-C) with PF and two samples (PLY-1 and PLY-2) with BBWA were manufactured, on which physical and mechanical properties, adhesive bonding morphology, formaldehyde emissions, and accelerated UV aging were evaluated. The adhesive penetration results, into the wood cells, were according to the viscosity of each adhesive. About the mechanical properties, the sample PLY-2 presented the same MOE and tensile strength as the sample PLY-C and reached 87% of the sample PLY-C MOR in the parallel direction. On the other hand, the sample PLY-1 presented the same behavior in the Janka hardness test as the sample PLY-C. All the samples subjected to shear strength tests met the requirement, and the samples PLY-1 and PLY-2 reached 68% and 80% of the PLY-C sample, respectively. The samples manufactured with BBWA presented a decrease in formaldehyde emissions by 88% and they were less susceptible to color change than the control sample under UV aging. According to the results obtained, it is concluded that plywood manufactured with BBWA might be a considerable replacement for plywood manufactured with PF adhesives at a laboratory scale.

1. Introduction

A plywood board is made of wood veneers whose distribution ensures that the fibers are oriented perpendicular to the fibers of the following veneer. Internationally, the requirements for plywood are regulated by the UNE-EN 636:2012+A1:2015 standard [1], which establishes structural or non-structural use requirements in dry, humid, or outdoor environments.

Given plywood versatility, it is widely used in moldings, structures between different floors, finishes, and roofing. An added value to this material is its insulating and acoustic conditioning properties, minimum edging, and easy bending [2].

Plywood production uses phenol–formaldehyde resins for veneer bonding, which are characterized by maintaining dimensional and mechanical properties under humid conditions [3]. These adhesives can release formaldehyde in the form of gas or vapor into the air, a compound that has been considered carcinogenic by the National Cancer Institute [4].

Currently, several studies have shown that the use of phenol–formaldehyde in the production of panels has disadvantages such as high energy consumption, high temperatures for the curing process (130 to 160 °C), phenol is expensive, its raw material is petroleum, and its high toxicity, which causes harm to both people and the environment [5].

As a solution to the problem of using phenol–formaldehyde resins for wood-based panel manufacture, different wood bio-based adhesives have been proposed with which different investigations have reached good results in the adhesive bonding strength of the boards. These adhesives are being manufactured to replace partially or completely formaldehyde-based adhesives. The latest research on this subject is as follows:

The researchers Núñez-Decap et al. [6], in 2023, manufactured two formulations of wood protein-based adhesive for particleboards to compare them with phenol–formaldehyde in different mechanical and formaldehyde emission tests. The authors concluded that particleboards manufactured with bio-based wood adhesive demonstrated comparable performance compared to particleboards manufactured with urea–formaldehyde adhesive.

Asafu-Adjaye et al. [7], in 2022, investigated the effect of bio-oil substitution in epoxy resin as an adhesive in oriented strand board production. The researchers could successfully replace epoxy resin with bio-oil. And they also found that 20% of the bio-oil improved the wet properties of OSB and the results were comparable to pMDI.

Nicolao et al. [8], in 2022, used a soy protein-based adhesive to study the relation between the quality and the morphology of the bonding of plywood panels manufactured with northeast Argentinian woods.

Pradyawong et al. [9], in 2019, developed a wood adhesive using soy protein and oxidized lignin with laccase enzymes in the presence of TEMPO, which refined the protein–lignin bonding. These researchers found that the manufacture of soy protein, lignin, and laccase wood adhesives is simple and only uses bio-materials. These bio-based-wood adhesives can be a good alternative to be used in the plywood manufacturing industry. Furthermore, this bio-based wood adhesive improved the wet shear test compared to the other combinations without these three elements working together.

Bai et al. [10], in 2019, developed a novel wood adhesive based on yeast hydrolysate incorporating ethylene glycol diglycidyl ether (EGDE). The influence of EGDE on the adhesive properties and the mechanism of the adhesive curing was also studied and improved the shear strength results compared to the samples without EGDE.

Chen et al. [11], in 2019, studied the preparation and characterization of a nanolignin and phenol formaldehyde resin by replacing phenol partially with lignin nanoparticles to manufacture a wood adhesive. The researchers found that when partially replacing a PF resin with nanolignin, the results of the bond strength test and the formaldehyde emissions passed the requirements of the standards. Besides, the curing temperature was reduced.

Esfandiyari et al. [12], in 2019, investigated the possibility of making lignin–glyoxal resins as wood adhesives in the replacement of aldehyde and urea-based adhesives in the production of plywood panels. The researchers found that a modified lignin was more reactive than an unmodified lignin. Also, when combining the lignin with glyoxal better results were obtained, and better results when obtained when increasing the adhesive quantity.

Núñez-Decap et al. [13], in 2018, worked on yeast protein extract adhesives for wood and tried them on particleboards, which showed reduced formaldehyde emissions compared to a control sample manufactured with a commercial wood adhesive. Besides, the particleboards manufactured with yeast-protein-based adhesive presented similar physical and mechanical properties compared to particleboards manufactured with urea–formaldehyde adhesives.

Li et al. [14], in 2018, developed a lignin-epoxy resin derived from biomass as an alternative to formaldehyde-based wood adhesives because of formaldehyde’s harmful effect on people’s health. The adhesive formulated without formaldehyde achieved the objective, and this adhesive presented higher results of bond strength compared to the PF and passed the standard requirement.

Zhang et al. [15], in 2017, improved the performance of four formulations of wood soy flour-based adhesive with and without different proportions of lignin (LB) and polyamidoamine-epichlorohydrin-based resin (PAE). The researchers obtained better results in dry and wet shear strength when modifying the soy flour adhesive with LB and PAE because these components created a strong network.

Pradyawong [16], in 2017, studied the relation between protein and lignin considering different sizes of lignin particles to understand the adhesion properties of soy protein adhesives enhanced by biomass lignin. The researchers concluded that the adhesion of lignin improved the wet shear strength of the soy protein-adhesives with the optimum ratio of soy protein and lignin found which was 10:2.

Zhu et al. [17], in 2017, studied the improvement of bio-based wood adhesive through the enhancement of the interaction of camelia protein and depolymerized lignin through an oxidation process to create a more environmentally friendly wood adhesive. These researchers found that the interaction of camelia protein with oxidated lignin improved water resistance and when irradiated with ultrasound presented an even better intermolecular bonding.

Núñez-Decap et al. [18], in 2016, studied the evaluation of a single-cell protein from three different types of yeast to develop wood adhesives for the wood industry. The authors compared the results of the yeast protein-based adhesives with soy protein-based adhesives and urea–formaldehyde. The authors concluded that the yeast-protein wood adhesives are adequate to manufacture bio-based wood adhesives for interior use, with similar results compared to urea–formaldehyde adhesives. Besides, they do not have formaldehyde in their composition.

Luo et al. [19], in 2015, created an eco-friendly wood adhesive from soy protein and lignin to improve the shear strength property in humid conditions of a soy-based adhesive by adding a lignin-based resin. They probed different ratios of lignin between 5 and 25%, obtaining considerably better results mainly in wet shear strength when using 10% of lignin.

Then, the main problem addressed by this study is the use of phenol–formaldehyde (PF) adhesives in the plywood industry. In this sense, an important development opportunity is generated in the use of bio-based adhesives based on protein and nanolignin to exploit and enhance the physical, mechanical, thermal, and emission properties of wood-based panels. This new adhesive is expected to match the strength properties of the adhesive bonding, considerably reducing the emissions of formaldehyde released into the environment while providing added value to the boards for their diversification in the market.

2. Materials and Methods

2.1. Materials

For the elaboration of plywood (PLY) panels with a phenol–formaldehyde adhesive and with bio-based adhesives, it was utilized Pinus radiata D. DON veneers (7%–12% of moisture content), with dimensions of 500 mm length × 500 mm width × 2.6- and 3.2-mm thickness. The Laboratory of Engineering Products and Adhesives based on Wood (PRODIMA-LAB) of the Universidad del Bío-Bío, Chile, provided the wood veneers.

The adhesive systems utilized in the fabrication of the plywood panels were a phenol–formaldehyde-based adhesive (PF) and two protein and lignin bio-based wood adhesives (BBWAs), formulated as explained in the studies by Núñez-Decap et al. [6,18]. The PRODIMA-LAB also provided the adhesives used.

2.2. Manufacture of Plywood Panels Using Commercial Adhesive and Bio-Based Wood Adhesives

It was necessary to classify the available veneers (198) to select the better ones (75). A visual classification under wood roughness and knot defects criteria was performed, determining 3 grades: rough, medium, and clear. In addition, it was measured four different veneer thicknesses to obtain a thickness average, the veneer weight, and the moisture content of four different positions to obtain a more trustful value. The classification made is presented in

Table 1. Summary of veneers classification.

Thickness (mm)	Moisture Content (%)	Grade of Visual Classification	Defects	Veneers Quantity
2.6	6.6	Clear	Knots	121
3.2	6.7	Clear	knots	77

.

To the adhesion of the veneer, it was applied an adhesive grammage of 200 g/m² over one of the veneer faces.

After the classification of the veneers, the best 75 veneers were selected and the configuration of the plywood panels was made. Fifteen panels were configured with 5 veneers each, and the distribution of the veneers is shown in Figure 1.

The panels were fabricated under the conditions mentioned in

Table 2. Plywood manufacturing conditions.

Pressing Parameters	PLY-C	PLY-1	PLY-2
Grammage (g/m2)	200	200	200
Gluing time (min)	5	5	5
Hot press temperature (°C)	130	150	150
Press factor (mm/min)	1	0.6	0.6
Specific pressure stage 1 (bar)	12	12	12
Specific pressure stage 2 (bar)	9	5	5
Specific pressure stage 3 (bar)	5	2	2
Press time stage 1 (s)	675	675	675
Press time stage 2 (s)	112.5	325	325
Press time stage 3 (s)	112.5	500	500

and then they were hot pressed with different pressing cycles depending on the type of adhesive (Figure 2). The fabrication process was executed under laboratory conditions (temperature: 20 °C and relative humidity: 55%) and after the pressing cycles, the plywood panels were also kept under laboratory conditions for 7 days.

Three different samples of plywood panels were manufactured at a laboratory scale: control sample PF-PLY (PLY-C), BBWA1-PLY sample (PLY-1), and BBWA2-PLY (PLY-2). Five replicates per sample were manufactured, with dimensions of 500 cm length × 500 cm width × 14.2 mm thickness and an adhesive grammage of 200 g/m². The adhesive grammage was chosen according to the current parameters used in plywood manufacture in Chilean industries.

Then, they were formatted in dimensions of 450 mm length × 450 mm width with a squaring saw. Then, the necessary specimens were cut to determine the physical, mechanical, and thermal properties, as well as the formaldehyde emissions, by the respective specifications of the European and American standards.

2.3. Physical–Chemical Properties of Wood Adhesives

The viscosity of the adhesives was measured following the parameters mentioned in the ASTM D1084-16R21 [20]. The parameters used for each adhesive were a spindle Nº6, 150 rpm to PF adhesive; and a spindle Nº4, 200 rpm to BBWA1 and BBWA2 adhesive, in a viscosimeter Brookfield (model LVF).

The pH of each adhesive was measured according to ASTM E70-24 [21] with a Hanna Instruments pH-meter model 211. The solid content of the adhesives was measured according to ASTM D1490-01 [22].

2.4. Evaluation of Adhesive Bonding Morphology

To evaluate the adhesive bonding morphology, the samples were prepared under an internal procedure of the PRODIMA-LAB.

Firstly, they were cut into 10 mm³ cubes, then cut using a horizontal sliding microtome Leica SM2000R (Leica Microsystems Ltd., Shanghai, China), obtaining 2 slices, of approximately 40 µm thick each one, per sample, which were placed on an object holder. The slices were subjected to a dehydration process using technical absolute alcohol diluted in concentrations of 30%, 80%, and 95%. Subsequently, all the samples were protected with a drop of xylol, then one slide of each sample was dyed with a drop of 1% safranin, and finally, a drop of epoxy adhesive for microscopy was added to all the samples to cover them with an object cover and then reviewed in an advanced optical microscope LEICA DM2000 LED (Leica Microsystems Ltd., Shanghai, China) under 20× magnification. The stained samples were observed with fluorescence and without fluorescence.

Then, the evaluation of the adhesive bonding morphology was performed according to the procedure of Gavrilović-Grmuša et al. [23] and Vidal-Vega et al. [24] calculating the average penetration (AP), measured in µm, of the adhesive in the wood, according to Equation (1).

$$AP={\displaystyle \frac{{\Sigma Y}_i}{n}},$$

(1)

The parameters $Y_i$ and n correspond to the adhesive line interphase (µm) and the number of measurements, respectively, which are shown in Figure 3.

2.5. Physical Properties

The density was evaluated on 6 specimens per sample following the parameters mentioned in the European standard UNE-EN 323 [25]. The density profile of the sample specimens was measured directly through the thickness of 2 specimens per sample, with the dimensions of 50 mm length × 50 mm width, with an incremental step of 0.1 mm. This measurement was performed with an Amersham plc AMCK6693 gamma-ray densimeter (Amersham plc, Buckinghamshire, UK).

The moisture content was evaluated on 4 specimens per sample, following the parameters mentioned in the European standard UNE-EN 322 [26]. Finally, the thickness swelling and water absorption, for 24 and 48 h under water immersion, were measured in 8 specimens per sample, under the conditions of the European standard EN 317 [27].

2.6. Mechanical Properties

2.6.1. Static Bending Test

A three-point static bending test was carried out to determine the stiffness (MOE) and strength (MOR) of 12 specimens per sample, 6 with surface fiber orientation parallel to the load and 6 with surface fiber orientation perpendicular to the load, with dimensions 450 mm length and 50 mm width. A constant speed load of 10 mm/min was applied to the specimens until failure. The duration of the load applied during the test did not exceed 90 s, as stipulated in the European standard UNE-EN 310 [28]. The plywood samples were classified according to the standard UNE-EN 636 [29].

2.6.2. Bonding Quality Shear Test at Dry Condition

The shear test of the plywood samples was conducted on 10 samples per sample, with the dimensions of 150 mm length, 20 mm width, 10 mm thickness, and 10 mm shear length. The load was applied at a constant rate of motion so that rupture occurred within 30 ± 10 s, in accordance with the European standard UNE-EN 314 [30]. The plywood samples were classified according to the standard UNE-EN 636 [29].

2.6.3. Tensile Strength

The tensile strength test of the plywood samples was performed on 12 specimens of each sample, 6 with surface fiber orientation parallel to the load and 6 with surface fiber orientation perpendicular to the load. The standard establishes that the specimens must measure 406 mm in length, 25 mm in width, and a thickness in the critical zone of 6 ± 0.02 mm. The speed of advance was 0.9 mm/min, following the parameters mentioned in the American standard ASTM D3500 method A [31].

2.6.4. Hardness Test

The Janka hardness test was conducted on 6 specimens per sample with the dimensions of 100 mm length and 50 mm width. A compression load was applied at a speed of 6 mm/min to a depth of 5.65 mm with a sphere diameter of 11.3 mm and a specimen center distance of 25 mm following the parameters mentioned in the American standard ASTM D1037-12 [32].

A universal testing machine, Instron EMIC 23-100 (INSTRON, São José dos Pinhais, PR, Brazil), equipped with the BlueHill2 v. 4.23 software was used to measure the mechanical properties.

2.7. Measurement of Formaldehyde Emissions

Formaldehyde emissions were measured on 3 specimens per sample by MASISA S.S. using a dynamic microchamber (DMC) following the parameters mentioned in the American standard ASTM D6007 [33]. The results were compared with the European standard UNE-EN 636 [29] and the regulation TSCA-VI [34]. The plywood samples were classified according to the standard UNE-EN 636 [29].

2.8. Effects of UVAccelerated Aging

An accelerated UV aging was performed on 3 specimens per sample, with dimensions of 300 mm length and 50 mm width, following the parameters mentioned in the American standard ASTM G154 [35], and the methodology and equipment used in the research of Núñez-Decap et al. in 2023 [6].

2.9. Analysis of Data

Based on the obtained results, an ANOVA was conducted, accepting differences between the means at a significance level of p < 0.05. When needed, an LSD test was performed using the Statgraphics Centurion 19 software.

The results of the data analysis are presented as a footnote in each relevant table.

3. Results and Discussion

3.1. Physical–Chemical Properties of Wood Adhesives

For the characterization of each adhesive, it was measured the viscosity, pH at 25 °C, and the solid content which results are shown in

Table 3. Results of characterization of adhesives.

ID	Viscosity (cP)	Ph 25 °C	Solid Content (%)
PLY-C	4827 a ± 29.37	12.53 a ± 0.03	42.30 a ± 0.06
PLY-1	2520 b ± 52.30	5.43 b ± 0.00	59.10 b ± 0.00
PLY-2	582 c ± 6.08	5.90 c ± 0.00	57.60 c ± 0.00

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a, ^b and ^c) are not statistically dissimilar by the LSD test at 95% probability.

.

The physical and chemical properties of the wood adhesives contribute to the quality of the final product and to obtaining the right amount of penetration into the wood to have a satisfactory mechanical interlocking via the lumens [36,37].

3.2. Evaluation of Adhesive Bonding Morphology

The samples manufactured with bio-based adhesives (PLY-1 and PLY-2) stained with safranin could be observed under the microscope with blue light fluorescence, highlighting the adhesive bondline. In contrast, the phenol–formaldehyde adhesive of the control sample (PLY-C) did not fluoresce with safranin adhesion, so it was observed without fluorescence under white light.

The results of the evaluation of the adhesive bonding morphology are presented in Figure 4 and

Table 4. Average penetration of the adhesives used into each plywood sample.

ID	AP (µm)
PLY-C	73.14 a ± 19.72
PLY-1	80.35 a ± 55.21
PLY-2	105.60 b ± 38.81

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a and ^b) are not statistically dissimilar by the LSD test at 95% probability.

.

The results show that the average penetration (AP) of the PLY-C and PLY-1 samples are statistically dissimilar to the PLY-2 sample, which presented the highest AP. However, the results of all three samples are consistent with the average penetration results obtained by other authors mentioned in Table 5.

The findings of the average penetration of the samples are related to the viscosity of each adhesive tested, as the lower molar masses portions (lower viscous portions) of adhesive can easily penetrate in radiata pine wood, but higher molar masses remain at the wood surface and give a more stable bonding and higher cohesion properties [23,38,39,40].

Table 5. Adhesive penetration measurements presented by other researchers.

Reference	Type of Adhesive	Type of Wood Composite/Joint	AP (µm)
[41]	Melamine Urea–Formaldehyde	Two-ply pine parallel Plywood	186–283
[42]	HDPE film	Plywood	80–300
[43]	Melamine-modified Urea–Formaldehyde	Laminated joint	36–89
[44]	Urea–Formaldehyde	Laminated joint	150–274
[23]	Urea–Formaldehyde	Laminated joint	120–207

Table 5. Adhesive penetration measurements presented by other researchers.

Reference	Type of Adhesive	Type of Wood Composite/Joint	AP (µm)
[41]	Melamine Urea–Formaldehyde	Two-ply pine parallel Plywood	186–283
[42]	HDPE film	Plywood	80–300
[43]	Melamine-modified Urea–Formaldehyde	Laminated joint	36–89
[44]	Urea–Formaldehyde	Laminated joint	150–274
[23]	Urea–Formaldehyde	Laminated joint	120–207

3.3. Physical Properties

The plywood samples glued with BBWA showed a decrease in density compared to the reference panel PLY-C (

Table 6. Results of physical properties.

ID	Density (kg/m3)	Moisture Content (%)	Thickness Swelling (%)
			24 h	48 h
PLY-C	597 a ± 24.64	12 a ± 0.30	5.43 a ± 0.42	5.88 a ± 1.01
PLY-1	507 b ± 23.62	8 b ± 0.12	88.79 b ± 31.71	88.92 b ± 31.34
PLY-2	493 b ± 44.27	7 c ± 0.46	6.29 c ± 0.65	7.27 c ± 0.93

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a, ^b and ^c) are not statistically dissimilar by the LSD test at 95% probability.

). The range of decrease was between 15.08% and 17.42% for the PLY-1 sample with a density of 507 kg/m³ and the PLY-2 sample with a density of 493 kg/m³, respectively, but there were no statistically significant differences between the PLY samples. Besides, the density profile (Figure 5) shows the difference in density along the manufactured panels in the different samples, it can be seen in the glue lines that the density of the adhesives in each panel is higher than the density of the veneers. In addition, in the density profile, there can also be seen a difference between the samples, where the adhesive of the sample PLY-C presented the highest density value and the adhesive of the sample PLY-1 presented the lowest density value. The difference in the density of the samples may affect or be related to the mechanical properties results, as it happened in other investigations [6,45,46]. Apart from the density values, there can also be a difference in the thickness of the panels, which can be associated with the “spring-back” phenomenon, an irreversible thickness swelling that occurs after the releasing of stresses accompanied by some loss of glue bonds [47].

The moisture content (MC) value of the PLY-C sample was higher than the MC value of the PLY-1 and PLY-2 samples. Anyways, all of the MC results of the samples varied between 7% and 12%, which is the expected range for conditioning.

All the samples subjected to the swelling test obtained statistically different results from each other. The PLY-C sample obtained the best results. The sample PLY-2 achieved 84% and 76% in 24 and 48 h of underwater immersion, respectively, but in the case of the sample PLY-1, most of the specimens did not resist underwater immersion.

3.4. Mechanical Properties

3.4.1. Three-Point Bending Test Properties

The results from the three-point bending test are presented in

Table 7. Results of the strength and stiffness bending properties in the parallel and perpendicular directions of the panel fiber.

ID	Panel Fiber Direction	MOE (MPa)	MOR (MPa)	Density (kg/m3)	Moisture Content (%)
PLY-C	Parallel	6878 a ± 635	75.54 a ± 5.45	588 ± 8	10.7 ± 0.35
	Perpendicular	2786 c ± 276	35.28 cd ± 6.10	573 ± 20	10.4 ± 0.31
PLY-1	Parallel	5470 b ± 1837	44.77 c ± 13.44	526 ± 10	6.0 ± 0.34
	Perpendicular	2580 c ± 193	32.19 d ± 5.54	523 ± 11	6.4 ± 0.62
PLY-2	Parallel	6591 a ± 583	65.66 b ± 10.85	526 ± 10	6.0 ± 0.54
	Perpendicular	2454 c ± 147	31.28 d ± 4.66	487 ± 24	6.7 ± 0.47

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a, ^b, ^c and ^d) are not statistically dissimilar by the LSD test at 95% probability.

. Overall, the best results in stiffness (MOE) and strength (MOR) were obtained by the PLY-C and PLY-2 samples.

Regarding the MOE results (Figure 6a), the sample PLY-2 (MOE: 6591 MPa), in a parallel direction, presents similar results in comparison to the PLY-C sample (MOE: 6878 MPa), while the sample PLY-1 (MOE: 5470 MPa) is statistically significantly different to the others, but reached a 79.52% of the PLY-C sample.

About the MOR results (Figure 6b) in parallel direction, the sample PLY-2 (MOR: 65.66 MPa) reached an 86.92% of the PLY-C (MOR: 75.54 MPa) sample, while the sample PLY-1 (MOR: 44.77 MPa) reached an 59.27% of the MOR value of the PLY-C sample, and there were statistically significant differences between the three samples. In the perpendicular direction, all the samples tested presented similar results about MOE and MOR properties and additionally, they did not present statistically significant differences between them.

The results are consistent with the results obtained using the same BBWA but in particleboard, where the best results were performed by the samples bonded with the control and P2 adhesives [6]. In the investigations of high-density fiberboards bonded with starch-based wood adhesives and of plywood bonded with chitosan as a BBWA, they also obtained comparable results with the control adhesive (UF) [48,49]. This analysis means that plywood panels manufactured with BBWA might be a suitable replacement for plywood panels manufactured with formaldehyde-based adhesives.

The load–displacement curves (Figure 7) were plotted to visualize the bending performance of the specimens of each sample while the load was increasing until they withstood their maximum. It can be seen that in the parallel direction, the samples PLY-C and PLY-2 demonstrated achieving a higher maximum load and a stiffer behavior than the sample PLY-1; on the other hand, in the perpendicular direction, all three samples achieved a similar maximum load, but the sample PLY-C demonstrated a ductile behavior compared to the PLY-1 and PLY-2 samples, which showed a fragile behavior.

Finally, according to the stiffness and strength results and the standard UNE-EN 636, the samples can be classified as follows: PLY-C: F 50/20 E 60/25; PLY-1: F 25/20 E 50/25; PLY-2: F 40/20 E 60/20.

3.4.2. Shear Strength

A dry shear strength test at dry conditions was performed to analyze the bonding quality of the samples. The results are presented in

Table 8. Shear strength results of each sample and the comparison with its requirement.

ID	Shear Strength (MPa)		Wood Failure (%)		Standard Requirement
	Result	Requirement	Result	Requirement	Accept/Fail
PLY-C	2.5 a ± 0.43	≥1	87 a ± 13.75	-	Accept
PLY-1	1.7 b ± 0.30	≥1	53 b ± 21.10	-	Accept
PLY-2	2.0 b ± 0.50	≥1	40 b ± 23.34	-	Accept

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a and ^b) are not statistically dissimilar by the LSD test at 95% probability.

, compared to the requirements of the standard UNE-EN 314 [30], and classified according to the standard UNE-EN 636 [29].

The shear strength values fluctuated between 1.7 (MPa) and 2.5 (MPa), reaching the samples manufactured with BBWA 68% (PLY-1), and 80% (PLY-2) of the control sample. The results show that there are statistically significant differences between the plywood samples glued with BBWA (PLY-1 and PLY-2) and the control sample (PLY-C), presenting the control sample with the highest shear strength. Nevertheless, all three samples are accepted by the standard, they presented higher results than the standard requirements and with the bonding quality of the adhesives used, the plywood samples are classified as class 1: dry interior.

The results obtained are not directly related to the average penetration of the adhesive into the wood cells, which was unexpected. Still, as mentioned in the research of Vidal-Vega et al. [24], the adhesive penetration must be enough. This means overpenetration can negatively affect the bondline quality. However, the bondline quality depends on other wood properties such as density, physical and chemical properties, adhesive characteristics, and pressing cycle. On the other hand, the wood species used is also an important factor, because the porosity of a wood species affects how much of the adhesive soaks into the wood, and different wood species have different strengths, which influence how strong the bond is expected to be [50].

3.4.3. Tensile Strength

The results of tensile strength (TS) are presented in Figure 8 and

Table 9. Tensile strength results of each sample in the parallel and perpendicular direction of the wood fiber.

ID	Wood Fiber Direction	Tensile Strength (MPa)	Density (kg/m3)
PLY-C	Parallel	45.87 d ± 3.6	583 ± 18.17
	Perpendicular	32.73 abc ± 12.5	528 ± 52.61
PLY-1	Parallel	30.89 ab ± 4.0	508 ± 24.55
	Perpendicular	24.36 a ± 6.0	510 ± 45.15
PLY-2	Parallel	41.31 cd ± 5.2	558 ± 24.34
	Perpendicular	33.01 bc ± 8.5	508 ± 39.78

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a, ^b, ^c, and ^d) are not statistically dissimilar by the LSD test at 95% probability.

. Figure 8 shows that in the parallel direction, the sample PLY-C (TS: 45.87 MPa) did not present a statistically significant difference in comparison to the PLY-2 sample (TS: 41.31 MPa), but the PLY-1 sample (TS: 30.89 MPa) presented statistically significant differences to the samples PLY-C and PLY-2, even then, the PLY-1 sample reached 67.34% of the control sample. On the other hand, in the perpendicular direction, the control sample did not present statistically significant differences compared to the PLY-1 and PLY-2 samples, but between the samples glued with BBWA, there were statistically significant differences presenting the PLY-2 sample the highest value of the TS property.

These results can be compared with the study of the authors Núñez-Decap et al. [6], who used the same bio-based wood adhesives, but for particleboard manufacture and they obtained a statistically identical performance for the PLY-C and PLY-1 samples, while in the case of the current study, a comparable performance to the PLY-C sample was achieved by the PLY-2 sample. In summary, the adhesives studied in this research can be compared with commercial adhesives, but their application will depend on the type of wood composite panel to be manufactured.

3.4.4. Janka Hardness

The results of Janka hardness (JH) of the samples in Figure 9 show that between the samples PLY-C (JH: 2694 N) and PLY-1 (JH: 2595 N), there is no statistically significant difference, but there is when compared to the PLY-2 sample (JH: 2356 N), which reached 87.45% of the maximum load that the control sample can withstand.

The authors Núñez-Decap et al. [51] obtained lower values of the hardness of their control sample (JH: 1870 N) of a seven-layered plywood in comparison to all the samples of this study. Hence, the samples manufactured with BBWA in the current study had a better performance than a seven-layer plywood panel manufactured with phenol–formaldehyde. This finding can be associated with several factors: the higher density of the plywood panels in the current study, the hardness of the wood veneers [52], or the control pressure in the second and third stages (Figure 2) in the pressing process of the control sample manufacture [53].

As hardness is commonly used to determine the suitability of wood for flooring applications [54], it can be said that the sample PLY-1 can be a suitable replacement for the PLY-C sample for flooring applications.

Besides the maximum load in the hardness test withstood by the samples, it is important to notice that this mechanical property is related to the samples’ density. From the statistical analysis, it is indicated a moderately strong relationship between the variables, which means that the maximum load that a sample can withstand depends on its density.

3.5. Measurement of Formaldehyde Emissions

The results of the measurement of formaldehyde emissions are presented at 2 h and 7 days, in

Table 10. Formaldehyde emissions at 2 h and 7 days for plywood samples.

ID	2 h (ppm)	7 Days (ppm)
PLY-C	0.045	0.009
PLY-1	0.005	0.001
PLY-2	0.005	0.001

.

The results show that the samples manufactured with bio-based wood adhesives presented a decrease in formaldehyde emissions by about 88% compared to the control sample for both the 2 h and the 7 days. The formaldehyde emissions of the samples PLY-1 and PLY-2 correspond only to the volatile organic compounds (VOCs) from the wood itself [55].

The results obtained of every sample met the emission regulation (0.05 ppm of formaldehyde for plywood) TSCA-VI [34] and the European emission standard (0.1 ppm of formaldehyde for plywood), so the plywood samples are classified as E1 by the UNE-EN 636 [29]. Apart from the standards, the findings are also consistent with the results of other authors [6,11,13] who have also worked with bio-based wood adhesives and have succeeded in reducing formaldehyde emissions.

3.6. Effects of UVAccelerated Aging

The results of the accelerated UV aging can be easily observed in Figure 10, where a noticeable change in the surface color of the samples can be seen. The photo discoloration caused by the UV radiation occurred because of complex chemical reactions, causing the degradation of amorphous carbohydrates and the condensation of lignin present in the wood [56,57,58].

The results, shown in

Table 11. Color change artificial weathering of the samples.

ID	ΔΕ Average
PLY-C	14.64 a ± 1.14
PLY-1	12.70 b ± 2.24
PLY-2	14.22 a ± 0.69

The results presented correspond to the mean ± St. Dev. The means in the columns followed by the same letters (^a and ^b) are not statistically dissimilar by the LSD test at 95% probability.

, revealed that the specimens of the PLY-1 sample were 11% and 13% less susceptible to color change than the PLY-2 and PLY-C samples, respectively. Likewise, there were statistically significant differences between the sample PLY-1 and the samples PLY-C and PLY-2.

Apart from the color change after the UV treatment, it can be noticed that in the sample PLY-2, the adhesive has come to the surface of the veneer, which can be associated with the low viscosity and the higher adhesive average penetration of the adhesive BBWA2 compared to the others.

4. Conclusions

According to the physical, mechanical, formaldehyde emissions, and UV-treatment tests performed, it is concluded that plywood panels manufactured with bio-based wood adhesives might be a considerable replacement for plywood panels manufactured with formaldehyde-based wood adhesives at a laboratory scale.

Between the results were found that the adhesive BBWA2 provided to the composite panels statistically similar results to the phenol–formaldehyde adhesive in the bending performance (MOE in the parallel direction, MOE, and MOR in the perpendicular direction), in the tensile strength test, and the evaluation of accelerated UV aging. In the case of the MOR in parallel direction and the Janka hardness test, the sample PLY-2 reached 86.92% of the control samples’ MOR and 87.45% of the maximum load withstood by the control sample in the hardness test, respectively.

On the other hand, the adhesive BBWA1 provided to the plywood panels statistically similar results to the PF adhesive in the bending performance (MOE and MOR in the perpendicular direction) and the Janka hardness test, while in the case of the bending properties in the parallel direction, the sample PLY-1 reached 79.53% and 59.27% of the control sample MOE and MOR, respectively. About the shear strength test, all three samples met the standard requirement, and the samples PLY-1 and PLY-2 reached 68% and 80% of the control sample, respectively. In the tensile strength test, the sample PLY-1 reached 67.34% of the control sample result. When the UV aging was applied, the sample PLY-1 was 13% less susceptible to color change than the control sample.

In addition to the similar mechanical performance of the samples tested, the plywood panels manufactured with bio-based wood adhesives presented a decrease in formaldehyde emissions by 88% compared to the control sample.

With the results obtained, it is suitable to inform that plywood panels made with BBWA might replace commercial plywood panels even under UV aging conditions and when formaldehyde emissions must be reduced. The type of BBWA will depend on the use of the panel and they must have an indoor application. Tests at a pilot scale are recommended.