1 Introduction

In construction, the twentieth century was the point at which wood was largely replaced by steel and concrete (Pramreiter et al. 2023). However, nowadays there has been a change and more attention is being focused on developing innovative materials based on natural resources. Wood is considered as a main example of renewable, sustainable and easily workable material and the development of wood-based composites allowed the elimination of some of its significant drawbacks such as anisotropy, the occurrence of natural defects or susceptibility to biotic and abiotic factors (Reh et al. 2022; Bekhta and Krystofiak 2023). In the case of structural applications, this type of composite includes, among others, plywood, oriented strand board (OSB), fibreboard and particleboard. However, their application in the construction sector is regulated by fire safety standards due to the risks associated with the ignitability of the unprotected composites and their ability to contribute to the spread of a fire (Thomas et al. 2021). Therefore, it is necessary to implement fire protection measures and in the case of wood-based materials, the impregnation of wood before gluing seems to be a particularly reasonable solution allowing for the protection throughout the entire cross-section, not only on the surface of the resultant materials (Grexa et al. 1999; Kawalerczyk et al. 2022).

The number of studies carried out so far on the production of wood-based materials from impregnated wood is very scarce compared to the number of studies on fire-resistant coatings. The vast majority of them concern the use of inorganic salt solutions (Popescu and Pfriem 2020). In the case of plywood and laminated veneer lumber (LVL) with increased fire resistance, the possibility of protecting veneers with, for example, mixture of ammonium phosphate, ammonium sulphate and ammonium bromide (Bryn et al. 2016), mixture of diammonium phosphate and ammonium sulphate, mixture of sodium dichromate, ferrous sulphate and ammonium chloride (Bekhta et al. 2016), mixture of potassium carbonate and urea (Kawalerczyk et al. 2019, 2023c), sodium silicate (Hautamäki et al. 2020), melamine solution and amino trimethylene phosphonic acid (Lu et al. 2022), guanidine phosphate (Yan et al. 2023), zinc chloride (Özçifçi and Okçu 2008), borax, boric acid (Kol et al. 2010), and mixture of diammonium phosphate, borax and boric acid (Ozcifci et al. 2007), was determined. In the case of impregnation of wood particles intended to produce particleboard, the number of scientific papers is even more limited. Investigations performed so far included such substances as: commercial fire retardant Burnblock® (Medved et al. 2019), borax, boric acid (Nagieb et al. 2011), ammonium phosphate, zinc borate (Mamatha et al. 2017), mixture of boric acid and disodium octaborate (Laufenberg et al. 1986), and diammonium phosphate (Du and Song 2014). However, despite a significant improvement in fire properties of both veneer-based materials and particleboards, the majority of studies also showed a significant reduction in the strength of manufactured materials. Due to the fact that reduction in mechanical properties was noted for composites bonded with the most common formaldehyde-based resins such as: urea–formaldehyde (UF) resin, phenol–formaldehyde (PF) resin and melamine-urea–formaldehyde (MUF) resin, which according to Antov et al. (2021) still dominate the market, the search for an adhesive suitable for gluing protected wood was started. Based on the outcomes of research on the production of particleboard from wood particles impregnated with the mixture of potassium carbonate and urea, it was found that PF/pMDI hybrid resins can be used for this purpose (Kawalerczyk et al. 2023a, 2023b). This is an interesting solution from the practical perspective, because the use of this type of binding agent usually allows to meet the requirements for materials used outdoors, which is crucial, for example, in construction (Dukarska et al. 2017).

According to Beck (2020), polyesterification of chemicals in wood as the chemical modification technique is gaining attention over the past decade. An interesting example of substances that can be used for this purpose are sorbitol and citric acid (SCA). It is a low priced, widely available, bio-based and non-toxic solution which can contribute to, for example, increase the density, improvement in wood’s dimensional stability, resistance against fungal decay, resistance against subterranean termites and resistance against marine environment (Kusnierek et al. 2024; Larnøy et al. 2018; Treu et al. 2020a, 2020b). Moreover, research performed by Mubarok et al. (2020) showed that SCA-modified beech wood was characterized by lower thermal decomposition temperature which, according to the authors, indicates the possible fire-retardant properties of SCA.

Therefore, considering the promising results in terms of using PF resin modified with polymeric 4,4’-methylene diphenyl diisocyanate (PF/pMDI) for fire retardant-treated wood, the many advantageous effects of polyesterification on the properties of wood and the fact that the effect of using SCA-modified wood on the properties of particleboard has not been studied so far, the aim of this research was to determine the properties of particleboard made of wood particles polyesterified with sorbitol and citric acid.

2 Materials and methods

2.1 Materials

Pine (Pinus sylvestris L.) wood particles were supplied by the local manufacturer of wood-based materials. Citric acid (> 99% pure) and D-sorbitol (> 98% pure) were purchased from Merck (Poznań, Poland). Phenol–formaldehyde adhesive, with a viscosity of 618 mPa·s, gel time at 130 °C of 249 s, solid content of 53%, pH of 10.5 was provided by Silekol (Kędzierzyn-Koźle, Poland). pMDI obtained from Bayer AG (Leverkusen, Germany) was characterized by the following features: a solid content of 100%, an NCO content of 32%, chlorine hydrolytic of 96 mg/kg and a viscosity of 220 mPa·s.

2.2 Polyesterification of wood particles

To determine the average dimensions of wood particles, the sieve analysis was performed. These particles were passed through the set of flat sieves with the following square mesh dimensions: 0.315, 1.0, 1.5, 2.0, 2.5, 4.0, 5.0 and 6.3 mm. The analysis showed that more than 80% of wood particles were representing fractions in the range between 1.5 and 2.5 mm. Before the treatment, they were dried at 100 °C to a moisture content (MC) of 2 ± 2%.

The mixture of sorbitol and citric acid for impregnation was prepared according to the method described by Beck (2020). The stock solution (pH of 2) was made by mixing citric acid and sorbitol in a 3:1 molar ratio and dissolving powders in a distilled water to reach the concentration of 56%. Then, the obtained solution was diluted with distilled water to the concentrations of 25% and 35% which were used for the treatment. Wood particles were placed in 45-L HDPE containers filled with the impregnating solution and soaked for 3 h. The use of vacuum impregnation was omitted due to the very small dimensions of the particles, which make the impregnation by soaking effective enough for this form of wood. After soaking, they were left to drain for 30 min and then cured for 18 h at 140 °C to complete the polyesterification (Larnøy et al. 2018). To determine the weight percentage gain (WPG), 40 pieces of wood particles were placed in the containers wrapped in a mesh that allowed the free flow of impregnating solution. WPG (%) was calculated based on the following Eq. 1:

$$WPG=\left(\frac{{M}_{t}-{M}_{0}}{{M}_{0}}\right)\times 100$$
(1)

where Mt is the oven-dry mass of treated particles and M0 is the oven-dry mass of untreated particles.

2.3 Particleboard manufacturing and testing

The list of variants of manufactured particleboards differing in the concentration of SCA and loading of pMDI is presented in Table 1.

Table 1 List of variants of particleboard
Full size table

To prepare the hybrid adhesive, the assumed amount of pMDI was added to PF resin and the obtained mixture was stirred manually until the proper homogenization was achieved. The gluing of wood particles was performed in slow-speed gluing device equipped with the LFG-5 pneumatic system (Devilbiss, Warsaw, Poland). The gluing degree, calculated as the ratio of dry mass of adhesive to the dry mass of wood, was 10%. Two single-layer particleboards with the dimensions of 580 × 670 mm, assumed thickness of 10 mm and assumed density of 650 kg/m3 were produced for each variant using Simpelkamp hydraulic press (Krafeld, Germany). The final thickness of the boards was achieved by using steel spacers placed in a set intended for pressing in a shelf press. The following pressing parameters were applied: pressing time of 25 s/mm of final board thickness (total pressing time of 250 s), pressing temperature of 180 °C and maximal unit pressure of 2.5 N/mm2. The press closing time was approx. 20 s for each variant of a mat. The edges of approx. 15 mm, usually characterized by a slightly lower density were calibrated after pressing.

Particleboards were conditioned for 7 days at a relative humidity of 65 ± 5% and temperature of 20 ± 2 °C prior to testing of their physical and mechanical properties. Density profile of the samples (50 × 50 mm) was investigated using the density profiler working on X-ray system GreCon DAX 6000 (Fagus GreCon, Charlotte, USA) with a measurement rate of 0.05 mm/s and measurement resolution of 0.02 mm. The presented profiles were averaged based on the measurements of five samples from each variant. Density was determined according to EN 323 (1999), internal bond (IB) according to EN 319 (1993), modulus of elasticity (MOE) and bending strength (MOR) according to EN 310 (1993). Furthermore, to assess the resistance to water, internal bond after boiling (V100) was determined according to EN 1087 (1995), and thickness swelling after 24 h of soaking in water according to EN 317 (1998). Each test was performed using ten samples, half from the first board and half from the second one.

Although the main objective of the research was to investigate the effect of wood polyesterification on the physical and mechanical properties of the particleboards, the observations of Mubarok et al. (2020) led to preliminary assessment of the boards’ flammability. A mass burning rate was determined to investigate the burning behavior of the samples. This method was previously used to determine flammability and mass loss of, e.g., spruce wood (Zachar et al. 2012) particleboards with the addition of waste rubber (Mancel et al. 2022) and particleboards made of impregnated wood particles (Kawalerczyk et al. 2023b). A sample with a dimension of 50 × 70 mm was placed in a metal holder at 30 mm from the radiating heat source with a heat flux of 30 kW/m2 and an input of 1000 W. The weight loss was measured by electronic balance connected to the holder, for 600 s in 1 s intervals using five samples of each variant. The schematic presentation of testing device is presented in the paper of Zachar et al. (2012) and the appearance of the sample during the test is shown in Fig. 1. The burning rate (ϑ) was calculated according to the following Eq. 2:

$$\vartheta = \frac{\delta \left(\tau \right)- \delta (\tau +\Delta \tau )}{\Delta \tau }$$
(2)

where δ(τ) is the sample mass in the time, τ (g); δ(τ + Δτ) is the sample mass in a time τ + Δτ (g) and Δτ is a time interval in which the mass value was recorded (s).

Fig. 1
figure 1

Appearance of a burning sample during the test

Full size image

The flammability of particleboards in a hot air furnace was determined according to ISO 871 (2010) in a Setchkin oven. This method was previously used to determine the flash point temperature of wood (Zachar et al. 2017, 2021) and wood-based materials such as plywood, particleboard, fiberboard and OSB board (Zachar et al. 2014), and these results are crucial for the analysis of first phase of fire development. The lowest air temperature at which the samples with a dimension of 10 × 10 mm and a total mass of 1 ± 0.03 g ignited within 600 s was considered the flash point temperature. The temperature in the hot air furnace was measured with thermocouples and recorded with Ahlborn Almemo 2290–8710 V7 datalogger (Ahlborn Mess- und Regelungstechnik GmbH, Holzkirchen, Germany). The test was repeated five times for each variant.

2.4 Statistical analysis

Analysis of variance (ANOVA) was carried out to analyze the results of tests. To distinguish homogeneous groups and assess the significance of observed changes, a HSD Tukey test at the assumed significance level of α = 0.05 was performed using Statistica 13.3 software.

3 Results and discussion

The appearance of the particleboards is shown in Fig. 2. A noticeable difference in the color of wood particles can be observed, and moreover, the higher the concentration of SCA was, the darker the particles became. According to Lin et al. (2022), the more brownish color of SCA-treated wood results from the natural color of citric acid-sorbitol polymer and progressing thermal degradation of wood during curing at 140 °C. The color change was expected based on the previous work, however, it is not crucial in the case of wood-based materials for outdoor applications because wood usually darkens over the time anyway due to the exposure to natural weathering (Rüther and Jelle 2013).

Fig. 2
figure 2

Appearance of the samples of manufactured particleboards

Full size image

Figure 3 summarizes the density profiles of the manufactured boards. They can be affected by numerous factors associated with the production technology such as press closing time or pressing parameters and factors associated with the wood itself such as its moisture content or species (Suo and Bowyer 1994; Medved 2007). The correct course of density profile is crucial for shaping the assumed physical and mechanical parameters of wood-based materials (Laskowska 2024). For manufactured particleboards, the most common U-shaped pattern was observed for each variant. Moreover, no evident differences were observed between the density profiles for untreated particleboards and those impregnated with 25% SCA solution. The course of density profiles for particleboards made of particles treated with a 35% solution, on the other hand, indicated a slightly higher density and uniformity. Markedly, no effect of the pMDI addition was found.

Fig. 3
figure 3

Density profiles of manufactured particleboards: a made of untreated particles; b made of particles impregnated with 25% solution; c made of particles impregnated with 35% solution

Full size image

Figure 4 presents the results of the density measurements. Particleboards made of untreated wood demonstrated values close to the assumed one, and the introduction of pMDI did not influence the outcomes. It was also found that the treatment with SCA with a concentration of 25% did not lead to a change in density and the observed values remained very close to the assumption as well. However, the treatment of wood particles with a 35% solution resulted in a statistically significant increase in density of the particleboards. The obtained values increased by approx. 5%, and manufactured particleboards were characterized by a density of approx. 680 kg/m3. Despite a significant increase in the density of treated particles shown by a relatively high WPG value, which for 25% was approx. 41.6% and for 35% was approx. 59.7%, this phenomenon does not appear to be the direct cause of the increase in density of the manufactured boards. In the case of particleboard production, the particles’ mass is determined based on material calculations and is usually constant for given parameters. However, when the particles treated with 35% SCA solution were used, it was observed that the mat was characterized by a considerably higher consolidation before pressing. Highly consolidated mat maintains its dimensions better during the pressing operation, especially during its first stage when the press is closing to reach the targeted thickness, compared to a mat made of more loose particles. As a consequence, there was no shift in the linear dimensions of the mat from the assumed value, which in turn may contribute to a slightly increased density, in this case by 5%. This effect is also indicated by the course of the density profile. A more consolidated mat allows to produce a board with a more compact structure after pressing. In the case of particleboard made of wood particles impregnated with 35% SCA solution, the density profiles revealed that the surface layer demonstrated higher density and the course in the middle section was more uniform and had fewer fluctuations (Fig. 3), which in turn can result in higher values of the overall density (Korai 2022). Density of the surface layers can often be an indicator of consolidation because these particles are the first to fall off when closing the press and considering the small share of these layers in the entire volume of the board, it may cause such small changes of approximately 5%. Moreover, the adhesive properties of the mixture used for polyesterification may have contributed to the increase in the consolidation of the mat. The greater particle adhesion resulting from the interactions between the wood and the citric acid could also have improved the dimensional stability of the mat during press closing. The lack of changes in the case of 25% solution may be caused by a too low concentration of citric acid, which did not allow for achieving such a statistically significant effect on density.. Furthermore, as a future work, it would be interesting to investigate how the applied treatment of the particles affects the heat transfer rate during pressing because it could also potentially influence the uniformity of the density profile and consequently, the results of average density as well (Boruszewski et al. 2016).

Fig. 4
figure 4

Density of particleboards

Full size image

The results of internal bond, which is the basic parameter indicating the adhesive behavior in the particleboard, are presented in Fig. 5. It was found that the effect of treatment varied depending on the concentration of SCA. Based on the comparison of boards without the addition of pMDI, it was found that the use of wood particles treated with 25% solution resulted in a reduction in internal bond by approx. 32%. The negative effect could result from changes in the surface characteristic of particles after the impregnation which were not compensated due to too low citric acid concentration. However, the increase in the concentration to 35% resulted in an increase in internal bond by approx. 20% compared to boards labeled as 0–0. The positive effect observed in the case of higher concentration could be caused using sufficient amount of citric acid. There were numerous studies regarding the use of citric acid as a main green binder and cross-catalyst, cross-linking agent or dispersing agent for different binders in lignocellulosic panels production (Lee et al. 2020). Moreover, Ando and Umemura (2020) concluded that citric acid-based adhesives work as a covalent bonding type of adhesive in wood-based materials manufacturing. There was even a study on using a mixture of sorbitol and citric acid as a main binder in the production of particleboard (Lin et al. 2022). However, it seems that despite good properties in dry conditions, low water resistance limits the potential use of these materials for indoor applications. Another factor that had a positive effect on the internal bond of particleboards is the introduction of pMDI. Although the 5% addition did not cause statistically significant changes, the addition of 15 and 25% resulted in a considerable increase in IB. In the case of such a small loading of pMDI (5%), it can be assumed that the dominant reaction while curing was the condensation of the PF resin, and consequently, the number of created highly durable urethane linkages was insufficient to cause statistically significant changes. In the case of treatment with SCA at a concentration of 25%, the addition of pMDI (10 and 15%) allowed to reach the IB of reference board (0–0). There are many previously-described mechanisms of improvement in bonding strength by the addition of pMDI such as: formation of cross-linking and/or linear structures due to reactions between pMDI and water stored in wood tissue (Bao et al. 2003), improvement in morphology of cured bond lines (Zheng et al. 2004) and the formation of polyurethane structures resulting from reactions with functional groups of basic wood components (Wang et al. 2007). Since there was no statistically significant difference between 15 and 25% and considering a toxicity and high price of pMDI, the share should be optimized to 15%. Overall, the particleboard made of wood particles impregnated with 35% SCA solution, bonded with PF/pMDI adhesive with a pMDI content of 15% was found to be the most advantageous variant.

Fig. 5
figure 5

Internal bond of manufactured particleboards

Full size image

Figures 6 and 7 present the results of bending strength and modulus of elasticity. In the case of both parameters, the same tendency was shown, indicating differences due to both the concentration of the SCA used and the addition of pMDI.

Fig. 6
figure 6

Bending strength (MOR) of manufactured particleboards

Full size image
Fig. 7
figure 7

Modulus of elasticity (MOE) of manufactured particleboards

Full size image

A comparison of particleboards bonded with pure PF resin showed that the use of particles impregnated with 25% SCA resulted in a noticeable reduction in bending strength and modulus of elasticity by 16 and 15%, respectively. However, the increase in the concentration to 35% allowed to achieve the same level of flexural characteristics as the reference board. The reason for the lack of strength reduction in the case of solution with higher concentration was probably the improved bonding quality of particleboard shown in the results of IB. Moreover, research by Mubarok et al. (2020) showed that beech wood treated with 30% SCA and cured at 140 °C (similar treatment conditions) was characterized by an average MOE of 15,200 MPa, while for untreated wood it was only 13,988 MPa, which could also affect the results. Furthermore, the higher board density in the case of variants impregnated with 35% solution could have contributed to the improved MOR and MOE to some extent as well. The outcomes also revealed a positive effect of pMDI addition to PF resin in the amounts of 15 and 25%. For particles treated with the solution of lower concentration, the modification of PF adhesive with isocyanate (15 and 25%) allowed to achieve the same level of strength as the reference boards. The highest values of both MOR and MOE were obtained for variants made of untreated particles and particles treated with 35% solution, bonded with hybrid adhesive containing 15 and 25% pMDI.

Results of thickness swelling analysis is shown in Fig. 8. It was found that both the treatment of wood and the modification of the resin had a significant effect on the obtained results.

Fig. 8
figure 8

Thickness swelling of manufactured particleboards

Full size image

When analysing the effect of impregnation, it was found that boards made from particles treated with 25% solution showed TS at the same level as boards made from untreated particles. Furthermore, the increase in the concentration of the impregnate solution resulted in a decrease in TS by 18%. This is consistent with the findings of Kurkowiak et al. (2023a), who studied the water uptake of SCA-treated pine wood. According to the authors, applied SCA fills the cell wall with polyesters which leads to formation of cross-linked network reducing water uptake, even in the capillaries along the cross section. Moreover, as the weight percentage gain increased, the effect on wood-water interactions was clearer which is also consistent with the obtained results. The addition of pMDI in the amounts of 15 and 25% had a positive effect on TS values as well. This could be caused by a general increase in the strength of bonds in particleboard, which according to Beech (1975), can reduce the particleboard’s swelling. Moreover, the reason may be ongoing reactions between pMDI and hydroxyl groups of wood components leading to decreased water accessibility (Iswanto et al. 2019). With TS values dropping below 9%, variants labeled as 35–15 and 35–25 turned out to be the most favourable. It is particularly promising considering that no additional hydrophobic agent was added during the production of boards.

Figure 9 presents the results of internal bond after boiling. The outcomes showed the same tendency as in the case of IB results. In the case of impregnation with 25% solution, the internal bond after boiling decreased. This was most likely due to the lower bonding quality which was demonstrated during the internal bond analysis. However, the increase in the concentration of impregnate solution to 35% led to the significant increase in V100 values, which were even higher than those of reference boards. The reason was most likely a combination of improved bonding quality and reduced water uptake of wood shown in the swelling test. Less swelling during pre-treatment of samples before the test probably led to reduction of internal stresses, which consequently could contribute to the improvement in V100 value. Moreover, a positive effect of pMDI introduction was observed as well. It was probably caused by previously mentioned increase in the strength and water resistance of hybrid resin’s bond lines compared to pure PF resin. Overall, the combination of particles’ treatment with 35% solution and the addition of pMDI in the amounts of 15 and 25% made it possible to manufacture particleboard that was more resistant to external conditions compared to the reference board.

Fig. 9
figure 9

Internal bond after boiling of manufactured particleboards

Full size image

The best results were obtained in the case of using untreated particles and those treated with 35% SCA solution, bonded with the hybrid resin containing 15 and 25% pMDI. Based on both the very good strength characteristics and the water resistance, it was determined that these boards can be classified as P5 (load-bearing boards for use in humid conditions) according to EN 312 (2010). Furthermore, variants made of untreated particles and particles treated with 35% solution bonded with pure PF resin and PF resin with the addition of 5% pMDI and variants impregnated with 25% solution bonded with hybrid resin containing 15 and 25% pMDI reached P4 class (load-bearing boards for use in dry conditions). The worst results were noted for variants treated with 25% solution bonded with pure PF resin and PF resin with the addition of 5% pMDI, and in this case, mainly due to low internal bond values, only class P2 was obtained (boards for interior applications, including furniture, for the use in dry conditions). Considering that the assumption was to produce materials that can be used outdoors, for example in structural applications, this type of classification does not meet the expectations. In practice, this type of boards for furniture production used indoors are usually produced using urea–formaldehyde resin, and the use of PF resin for this purpose, especially mixed with pMDI, seems to be unjustified for many reasons such as, e.g., high price, high curing temperatures and a severe toxicity of monomeric phenol (Sedliacik et al. 2010; Karthäuser et al. 2023).

Table 2 presents the results of ANOVA performed based on the results of strength parameters and water resistance of the particleboards. A statistically significant effect of both the SCA concentration and share of pMDI was observed for all properties besides density. In this case, only the concentration of the impregnate solution was significant. Interestingly, no interactions between variable factors were noted in the case of strength parameters. However, the effect of interaction of SCA concentration and pMDI loading was only observed for TS and V100 outcomes. According to the results presented in Fig. 10, the simultaneous use of high concentration of SCA (35%) and high share of pMDI (15 and 25%) guarantees a significant increase in dimensional stability and water resistance of the boards.

Table 2 Results of ANOVA performed based on the results of strength parameters and water resistance of particleboards
Full size table
Fig. 10
figure 10

Interaction effects between concentration of SCA and share of pMDI on a thickness swelling; b internal bond after boiling

Full size image

The curves of percentage mass loss under the exposure to a heat source and other determined flammable properties are presented in Fig. 11 and Table 3, respectively.

Fig. 11
figure 11

Percentage mass loss of particleboards exposed to the heat source

Full size image
Table 3 Flammable properties of manufactured particleboards
Full size table

The results of performed analyses showed that the boards made of particles treated with 35% SCA solution have the lowest values of final mass loss after 600 s of experiment. When 25% solution was used, the results differed significantly from board made of untreated particles bonded only with pure PF resin. Furthermore, the analysis of the ignition times of the samples did not show any clear tendency due to both polyesterification and resin modification. However, as stated by Park and Lee (2008), the ignition of wood-based materials is a complex phenomenon to analyze because of the inhomogeneity of particleboard and the associated occurrence of void spaces, density dispersion, uneven surface structure etc. The burning rate of particleboard is also a complex phenomenon to analyze, especially considering the ongoing charring process (Quintiere and Quintiere 1992). On the other hand, the flash point temperature varied significantly depending on the assumed variables. Considering that most of the results were inconclusive at first glance, the ANOVA analysis was performed to assess the effect of individual variables, and the results are presented in Table 4.

Table 4 Results of ANOVA performed based on the results of flammable properties
Full size table

Based on the obtained statistical parameters, a significant effect of SCA concentration on each of the analyzed parameters was found, and it was particularly noticeable as confirmed by the high values of SS, MS, F and p-value. Based on thermogravimetric analysis performed by Mubarok et al. (2020), it was found that wood treated with 30% SCA cured at 140 °C demonstrated reduction in mass loss by approx. 20% compared to control samples. According to the authors, the combination of hydrolysis of the glycosidic bonds of wood polysaccharides by acidic SCA, progressing esterification and thermal degradation during the curing process could promote charring and reduce mass loss. This method of particle treatment seems to be less effective in terms of fire protection than, for example, the mixture of potassium carbonate and urea used in a previous study (Kawalerczyk et al. 2023b). However, recent findings of Kurkowiak et al. (2023b) indicate that combining SCA with a commercial phosphorous-based fire retardant could lead to further improvement in fire resistance of the particleboard. In turn, the share of pMDI in hybrid adhesive showed a statistically significant effect on the mass loss, time to ignition and flash point temperature. Although there are indications in the literature regarding, for example, densification of bond lines by eliminating microcracks due to the addition of pMDI (Zheng et al. 2004) or increased thermal stability of cured hybrid resins (Xu et al. 2010; Kawalerczyk et al. 2023a), which could potentially contribute to the improvement in flammability to some extent, this effect cannot be clearly explained based on current knowledge without the additional comprehensive analysis. Moreover, in the case of time to ignition and flash point temperature results, the interactions between SCA concentration and pMDI loading were noted. It means that a simultaneous increase in the concentration of SCA and the share of pMDI causes a gradual increase in the ignition characteristics of the boards.

4 Conclusion

Based on the conducted research, it was found that polyesterification of pine wood particles intended to produce particleboards had a significant effect on their physical and mechanical properties and the effect varied depending on the assumed variables. Both the concentration of the sorbitol and citric acid mixture and the share of pMDI in the PF/pMDI resin influenced the characteristics of the produced materials. The use of a 35% impregnate solution caused the improvement in internal bond, dimensional stability, water resistance and did not cause any changes in bending strength and modulus of elasticity. Reducing the concentration to 25% resulted in a decrease in the strength characteristics and water resistance of the boards and no longer caused a positive impact on the swelling of the boards. The use of hybrid resins with the addition of 15 and 25% pMDI had a positive effect on all the tested physical and mechanical properties of the boards. However, the use of 5% pMDI did not lead to any significant changes in their properties. Moreover, the applied treatment of wood particles contributed to the improvement of the flammable properties of the produced boards, and as the concentration of SCA increased, the visibility of the effect also increased. Outcomes of the study also indicate that the increase in loading of pMDI may have an impact on the ignitability of the board, however, further studies are planned to explain this effect in detail. Boards made of particles treated with 35% SCA solution, bonded with a hybrid resin containing 15% pMDI were considered the best variant. As a future work, it would be interesting to investigate the effect of polyesterification on the resistance of particleboard against wood-decaying fungi.