Wood impregnation in relation to its mechanisms and properties enhancement

Abstract

The principle of wood impregnation entails treating wood with a monomer/impregnating agent that diffuses into the cell walls, often followed by polymerization to change desired properties. Numerous studies related to this matter have been reported and continue to attract more interest, as wood impregnation can significantly improve wood properties. These processes can be grouped into two approaches: active modification involves the chemical alteration of wood structure by cross-linking, and passive modification features filling of cell cavities and/or cell walls with impregnating agents without any chemical reaction taking place. Wood impregnations could have resulted in an increase in its weight gain due to impregnating agents filling its cavities. It will diminish the utilization of wood as an engineering material in selected application fields. Owing to the extensive literature available, this article summarizes the representative achievements of wood impregnation. The mechanisms, benefits, and drawbacks of various impregnating agents on wood properties, along with grouping the impregnating agents that cause greater or lesser weight gain of wood were analyzed, compared, and evaluated. Thus, according to the application state of wood impregnations, the problems existing in those processes and the developmental trends in the future are also discussed in this review.

Full Article

Wood Impregnation in Relation to Its Mechanisms and Properties Enhancement

Sarah Augustina,^a,* Wahyu Dwianto,^a,* Imam Wahyudi,^b Wasrin Syafii,^b Philippe Gérardin,^c and Sari Delviana Marbun ^a

The principle of wood impregnation entails treating wood with a monomer/impregnating agent that diffuses into the cell walls, often followed by polymerization to change desired properties. Numerous studies related to this matter have been reported and continue to attract more interest, as wood impregnation can significantly improve wood properties. These processes can be grouped into two approaches: active modification involves the chemical alteration of wood structure by cross-linking, and passive modification features filling of cell cavities and/or cell walls with impregnating agents without any chemical reaction taking place. Wood impregnations could have resulted in an increase in its weight gain due to impregnating agents filling its cavities. It will diminish the utilization of wood as an engineering material in selected application fields. Owing to the extensive literature available, this article summarizes the representative achievements of wood impregnation. The mechanisms, benefits, and drawbacks of various impregnating agents on wood properties, along with grouping the impregnating agents that cause greater or lesser weight gain of wood were analyzed, compared, and evaluated. Thus, according to the application state of wood impregnations, the problems existing in those processes and the developmental trends in the future are also discussed in this review.

DOI: 10.15376/biores.18.2.Augustina

Keywords: Impregnating agent; Weight gain mechanisms; Wood impregnation; Wood properties

Contact information: a: Research Center for Biomass and Bioproducts, National Research and Innovation Agency (BRIN), Jl. Raya Bogor Km. 46, Cibinong, Bogor, West Java, 16911, Indonesia; b: Department of Forest Products, Faculty of Forestry and Environment, IPB University, Bogor, West Java, 16680, Indonesia; c: Laboratory of Research and Wood Material (LERMAB), Universite de Lorraine, Boulevard des Aiguillettes, 54506 Vandoeuvre-lès-Nancy CEDEX, France;

*Corresponding authors: sarahaugustina@gmail.com; wahyudwianto@yahoo.com

GRAPHICAL ABSTRACT

INTRODUCTION

Wood is a heterogeneous biomaterial that is prone to undergoing dimensional changes in response to easily changing atmospheric conditions. In addition, wood is also susceptible to environmental and biological attacks, so its durability and mechanical strength are often insufficient for long-term end uses. This has motivated a great deal of effort from numerous researchers to find solutions to the issues raised by these drawbacks. Over the past decade, impregnation of wood has become one of the most discussed techniques and continued to attract more interest due to their good results.

The principle of many impregnation processes entails treating wood with a monomer/impregnating agent, which then diffuses into the cell walls, followed by polymerization to impart the desired properties. Both active modification by cross-linking and passive modification by bulking are included in this process (Rowell and Young 1981, Hill 2006). Hardwood and softwood have been impregnated with a variety of impregnating agents such as waxes and oils (Lesar and Humar 2011; Esteves et al. 2014; Ahmed et al. 2017; Belchinskaya et al. 2021), thermosetting resins, including epoxy, phenol formaldehyde, urea formaldehyde, etc. (Furuno et al. 2004; Gabrielli et al. 2010; Zhao et al. 2016; Biziks et al. 2018; Schwarzkopf et al. 2020; Wang and Zhao 2022), monomer/polymer, including methacrylate, acrylate, styrene, unsaturated polyester, followed by radiation or catalyst-thermal treatment (Chao et al. 2003; Hadi et al. 2013), noncyclic and cyclic anhydride (Li et al. 2001; He et al. 2020; Augustina et al. 2021), as well as organic acid (Fang et al. 2000; L’Hostis et al. 2017; Grosse et al. 2019; Lee et al. 2020; Augustina et al. 2022). These types of impregnating agents have been classified as either nonbonded-leachable (NBL), nonbonded-nonleachable (NBNL), or bonded-nonleachable (BNL), depending on the bonding efficiency of the impregnating agent to and from the wood cells (Rowell and Young 1981). For the purpose of understanding the effectiveness of impregnation process on dimensional stability (long-term or short-term application of product services), these categories can serve as the underlying mechanism.

In order to achieve adequate results of physical properties, particularly in terms of dimensional stability, the wood’s capillary system (void spaces/pores) must be impregnated with the impregnating agents prior to curing. The type and condition of wood, along with the properties of chemical, i.e., polarity and nature of each chemical, and impregnation conditions, including the processes employed, and the parameters of those processes, i.e., time, pressure, curing, etc., are undoubtedly essential for impregnation treatability (Sint et al. 2013; Augustina 2021). The permeability and penetrability of wood species are related to impregnation treatability, since the wood has a capillary structure that offers the primary path for liquid penetration into wood, i.e., cell wall and/or lumen penetration (Flynn 1995; Taghiyari 2012; Wen et al. 2014; Augustina 2019). The penetration depth of impregnating agents is also affected by the average molecular weight (MW) of impregnating agent. According to the findings of Imamura et al. (1998), the presence of impregnating agents in the wood cells was found to be affected by the MW, or size of the chemicals. Furthermore, cell wall-deposited impregnating agents would have a more profound effect on the wood’s properties than those located in the cell lumen. For this reason, it is essential that the impregnating solution has the capacity to swell the cell wall, allowing the penetration of impregnating agents.

As a result of impregnating agents filling the wood cavities, an impregnation process may have led to an increase in its weight gain (Matsuda 1987; Hill 2006). Consequently, it will be diminishing the utilization of wood, per unit of service, as an engineering material in certain application fields. Lightweight engineering materials with greater dimensional stability are needed in respected fields, including those that produce boats, windmills, small wind turbine blades, thermal insulation, and aircraft (Pourrajabian et al. 2019; Jang and Kang 2022). According to Sørensen et al. (2010), weight savings are critical for wind turbine companies, since the materials must have high stiffness, be fatigue resistant, have high dimensional stability, be damage tolerant, and be lightweight in order for the blades to meet the required design life. However, there is no guarantee that a higher increase in weight gain of impregnated wood means better wood properties (Dong et al. 2015; Ahmed et al. 2017). According to Rowell et al. (2005), at extremely high weight percent gain (WPG), the anti-swelling efficiency (ASE) began to decline to some extent. In propylene, butylene-oxide, and methyl-isocyanate-treated wood, the ASE increases as the WPG increases up to 25% to 30%, while it begins to decrease at higher WPGs. This may be due to the fact that at a WPG of 33%, small cracks begin to form in the tracheid walls, and at a WPG of 45%, a major crack occurs within cells.

Owing to the extensive literature available, this article summarizes the significant advances that have been made in the field of wood impregnation. The mechanisms, benefits, and drawbacks of various impregnating agents, along with grouping the impregnating agents that cause or less cause an increase in weight gain of wood, are analyzed, compared, and evaluated with respect to their effects on wood properties, especially dimensional stability. In light of this, the paper provides a review of the current state of wood impregnations, the problems that have arisen during these processes and the development trend in the future.

IMPREGNATION PROCESS

Wood impregnation is any method of altering the desired properties of wood by introducing a chemical, a mixture of chemicals, or an inert material (impregnating agents) into the wood cell wall (Hill 2006). These processes can be grouped into two categories: active and passive modification. The former modification involves chemical alteration of wood structure by cross-linking, whereas the latter modification involves the filling of cell cavities and/or cell walls with impregnating agents without any chemical reaction taking place. According to Hill (2006), the impregnating agents can be fixed using two main mechanisms. Monomer or oligomer impregnation, with subsequent polymerization within the cell wall, is one mechanism for fixing the impregnating agents; the other is diffusion of soluble material into the cell wall, followed by treatment in order to render the material insoluble (immobile/locked). Moreover, wood impregnation can be further subcategorized into diffusion, capillary, and pressure treatment (Bi et al. 2021). These processes are typically carried out by initial vacuum and pressure cycles at high pressure in the industrial scale. While, in the laboratory scale, one or more processes are usually used.

The most common method for controlling the quality of impregnation process is the increased mass brought on by the impregnating agents. Chemical uptake (CU) measurement is used to characterize the impregnation, whilst WPG measurement is used to characterize the quantity of cured impregnating agents present in the wood (Behr et al. 2018). Therefore, this section will discuss wood impregnation, which is mainly done by vacuum, pressure, and a combination of both processes, as well as the relationship between factors that affected the process including type of impregnating agents, impregnation parameters, and wood permeability with WPG and/or CU, as well as the changes in the wood properties especially dimensional stability.

Type of Impregnating Agents

As previously stated, the impregnation process can be divided into two categories: active modification by cross-linking and passive modification by bulking. Cross-linking, as described by Rowell and Young (1981), take place when two or more structural units of the wood cell wall are chemically bonded together (cross-linked), preventing the units from swelling/shrinking as a result of exposure to moisture. In the presence of suitable crosslinking agents, the hydroxyl groups on the same or different cellulose, hemicellulose, and lignin polymers are capable of forming cross-link with each other. Meanwhile, cell wall bulking is defined as an increase in wood volume after treatment that is proportional to the theoretical volume of chemical added. Bulked wood may experience negligible additional swelling/shrinking of the wood cell wall upon contact with water.

Impregnating materials can be divided into several categories: wax and oil, thermosetting resin, monomer/polymer, cyclic and noncyclic anhydrides, as well as organic acids. Different impregnating materials will have distinct reactions and phenomena when applied to wood. Based on the bonding effectiveness, these types of impregnating agents can be divided into three broad categories. The first classes of impregnating agents are termed nonbonded-leachable (NBL), and they include wax and oil, as well as monomer/polymer impregnating agents. The second one is nonbonded-nonleachable (NBNL) which can be represented by thermosetting resin treatment. The third class is bonded-nonleachable (BNL), which includes anhydride and organic acid treatment. The additive impregnates and forms covalent bonds with the wood cell walls, imparting long-term moisture resistance and dimensional stability.

Nonbonded Leachable (NBL)

As mentioned before, NBL classes include wax and oil, as well as monomer/ polymer as impregnating agents. Benefits of using waxes and oils include their good water repellency and wood protection, environmentally friendly materials derived from dried and/or extracted from plant parts, cost-effectiveness, abundance, and non-toxic nature (Chau et al. 2015; Chen et al. 2020). In terms of chemical structure, natural waxes can be described as esters of long-chain carboxylic acids and alcohols (Scholz et al. 2010a). It is possible to apply wax into wood either in a melted state while diluting it in organic solvents, in the form of emulsions, or as suspensions through impregnation or in conjunction with other treatments such as plasma, etc. The type of waxes can be divided into several categories, including natural waxes such as beeswax (Li et al. 2020) and carnauba wax (Chen et al. 2020); natural fossil waxes derived from petroleum or lignite such as paraffin (Esteves et al. 2014; Jiang et al. 2020) and montan wax; synthetic such as hydrocarbon or amide waxes (Scholz et al. 2010a). Meanwhile, the different kinds of oils that can be utilized as impregnating agents include linseed oil (Chen et al. 2020; Liu et al. 2020), tung oil (Ahmed et al. 2017; He et al. 2019), hemp oil (Baar et al. 2021), engine oil (Belchinskaya et al. 2021), becker oil (Ahmed et al. 2017), vegetable oil (Wei et al. 2022), and other plant oil-based substances (Vasiliauskiene et al. 2020).

Impregnation with waxes and oils has been studied extensively, with many different approaches taken by researchers. Methods range from the straightforward–dipping, soaking/immersing, vacuum-pressure–to the intricate ones by combining those processes with heat treatment, hot and cold soaking methods (physical impregnation), as well as mixing both of methods. For this specific sort of impregnating agent, three types of treatment exist: Type I, treatment decreases moisture uptake, but the swelling is relatively identical as untreated wood over time. This treatment enhances water repellency but not dimensional stability; as for Type II, treatment decreases swelling but not its moisture uptake. This treatment enhances dimensional stability but not water repellency; as for Type III, treatment decreases both moisture uptake and swelling. This treatment enhances water repellency and dimensional stability (Rowell and Bank 1985).

This sub-section focuses on waxes and oils impregnation with vacuum-pressure method. There are basically three major processes involved in wax impregnation: melting microcrystalline wax to a liquid state, impregnating hot wax fluid into wood cells; and then allowing the microcrystalline wax to solidify within the wood’s cell cavities and the intercellular space (Zhang et al. 2020). It was reported that carnauba wax (CW) could provide WPG ranging from 3.7% to 9.9% with water absorption (WA) reduced by 53% in comparison to that of untreated wood after soaking for 1 h, but the values increased to 129% after 192 h. Wang et al. (2017) studied paraffin wax emulsion (PWE) in different solid contents (1, 2, 4, and 8%) and particle sizes (535, 400, 320, 232, and 171 nm). They discovered that WPG increased with elevating solid contents and decreased with elevating particle sizes. During this process, the WPGs produced were 2.0% and 11.5% for 1% and 8% solid content of PWE, whereas for particle size of 535 and 171 nm were <2.0% and 3.5% in loblolly pine, respectively. Jiang et al. (2020) investigated various PWE concentrations (2, 5, and 8%) and treatment duration. It was found that WPG increased with elevating PWE concentrations, which were 14.5% and 4.8% for 8% and 2% concentration throughout the same treatment time, respectively. As a result, WA decreased as the concentration increased, reaching 40.9% and 52.6% for 8% and 2% concentrations throughout the same treatment time, respectively.

Waxes do not react chemically with the hydroxyl groups of wood; however, they could form a hydrophobic film-barrier on wood surface. Scholz et al. (2010b) and Humar et al. (2016) were persuaded that there is no interaction between the hydrophobic wax and hydrophilic cell wall with its hydroxyl groups. Furthermore, this compound intends to fill cell lumens, thus achieving a hydrophobic treatment. This compound will be deposited in wood capillaries, reducing water penetration through capillary action and ultimately limiting the dimension swelling. Scholz et al. (2010c) found that pore volume decreased from 65-68% to 12-13% and 53-58% to 7-9% for pine and beech treated samples. Nonetheless, there are still significant issues that limit its widespread use. Incomplete and unequal penetration, temperature instability, and insufficient resistance to high temperatures are the common issues (Bi et al. 2021). According to Chen et al. (2020), wax impregnation using CW has been demonstrated to be effective at inhibiting water absorption in short-term applications. During periods of immersion, the hydrophobic film-barrier may be eroded, allowing water molecules to swell and open up the inaccessible region of wood cell walls. This phenomenon is supported by Fig. 1. Increasing the relative humidity (RH) from 35% to 65% reduced the ASE of PWE treated wood significantly. It implies that as RH increases, the slope of moisture adsorption and T/R ratios increases, thus it will eventually reduce ASE (Chau et al. 2015).

Fig. 1. Relationship between weight percent gain and anti-swelling efficiency in different relative humidity conditions using paraffin wax emulsion (Data obtained from Esteves et al. 2014)

In addition, oils impregnation using vacuum-pressure method have been reported by other researchers (Table 1). Chen et al. (2020) and Liu et al. (2020), used linseed oil (LO) derived from seed of Linum usitatissimum as impregnating agents. Once entering the wood structure, LO can block the lumen and generate a stable LO film on the pore surface, thus limiting water uptake. Liu et al. (2020) showed that LO-treated wood could achieve a WPG of around 21.7% and result in lower tangential and radial swelling values. These values were reduced by 19.3% and 15.2%, respectively, when compared to that of untreated. Other researchers reported that the utilization of hemp oil and castor oil could produce WPG of 61.1% and 60.6%, respectively with lower volumetric swelling (VS) than that of untreated (He et al. 2019; Baar et al. 2021). Due to its larger molecular size, hemp oil also exhibits higher VS and WPG than that of castor oil (Wei et al. 2022) (Fig. 2).

Fig. 2. Relationship between weight percent gain and volumetric swelling on hemp oil and castor oil (Data obtained from He et al. 2019 and Wei et al. 2022)

As there is no chemical bonding between wood and oil, it might only reduce the water adsorption rate but not the final moisture content. Furthermore, the oxidative polymerization of LO requires a prolonged time, often allowing the oil to exude from the wood. Therefore, it was discovered that epoxidized LO (ELO) could accelerate oxidative polymerization. ELO-treated wood has shown considerable improvements in ASE, water repellency, biodegradation and leaching resistance compared to LO-treated wood. Chen et al. (2020) reported that ELO-treated wood could produce WPG around 31.0% with ASE around 38%. The key improvement mechanism of ELO-modified wood was discovered to be the covalent bonded to the wood cell wall (new hydroxyl and carbonyl groups generated) following the opening of the epoxy ring.

In most cases, oils impregnation led to increased WPG, which is mainly accumulated in the surface area and/or stayed in the cell lumens. This phenomenon was supported by Hill (2006), who noted the inability of triglycerides in vegetable oils to penetrate the cell walls. It is also agreed by Baar et al. (2021) and Olsson et al. (2001b) that hemp oil and linseed oil have too large molecule size and hydrophobic character to enter the cell wall during impregnation, and consequently uptake occurs through pore cavities such as the lumen of tracheid and ray parenchyma cells.

Table 1. Oil Impregnation on Wood using Vacuum-Pressure Method

Note: WPG = weight percent gain; ASE = anti-swelling efficiency; TS/RS = tangential and radial swelling; MA/WA = moisture and/or water absorption; VS = volumetric swelling, BC = bulking coefficient, NA = not available.

The introduction of monomers and the subsequent in situ polymerization of these monomers into a polymer chain is one approach to permanently encapsulating novel materials in wood. These compounds undergo in situ polymerization after being exposed to radiation (Hadi et al. 2019) or by radical initiator at elevated temperatures (Che et al. 2018). A wide variety of vinyl monomers that are commercially available on the market, including acrylonitrile, glycidyl methacrylate, methyl methacrylate, hydroxyethylene, ethylene glycol, dimethacrylate, butyl acrylate, butyl methacrylate, styrene, acrylamide, or acrylonitrile, have been studied by several researchers.

This type of modification is partially effective to improve wood dimensional stability. According to Chao and Lee (2003), styrene (ST) treated wood could give WPG values around 35.9% and 48.3% for 1- and 20-min vacuum conditions, respectively. After 24 h soaking, WA may be lowered from 50% (untreated) to ≈ 20%. Hadi et al. (2019) investigated the effects of methyl methacrylate (MMA) impregnation on different wood species. They observed that the WPG of MMA-treated wood was around 14.62 to 23.75%; hence, it could achieve a WA of around 24% to 56.7%. According to Meints et al. (2018), polyethylene glycol (PEG) treated wood resulted in higher WPG as concentration increased. The WPG was around 10%, 25%, and 42% for 15, 30, and 45% of concentration, respectively. Ding et al. (2012) found that poly-MMA could result in an ASE of around 67% with lower WA. Despite the fact that this treatment might provide better dimensional stability, especially ASE, it can only be applied for a brief period of time (short-term application). According to He et al. (2011), ST treated wood could provide 88.4% of ASE for almost 2 h water soaking process, then the values subsequently decreased to 23.9% after 168 h water soaking process. This phenomenon was confirmed by Meints et al. (2018). After undergoing leaching treatment, the WPG of PEG treated wood was reduced significantly. Leaching of PEG 400 was nearly complete after 14 days of immersion in water (99% mean loss), whereas leaching of PEG 1000 resulted in less mass loss (75%) throughout the same period of time. Furthermore, PEG may liquefy as a result of the enormous amount of water absorbed and exuded from treated wood in high-humidity environment.

In summary, due to the penetration constraints, polymerization of hydrophobic monomers or co-polymer, such as MMA, styrene copolymer and PEG, respectively could not guarantee sufficient dimensional stability, especially when used in longer term service. The utilization of hydrophobic monomers or polymers was unable to enter or penetrate the wood cell wall; instead, these substances could only be filled in the lumen due to hydrophilic nature of wood cell walls. Consequently, pre-treatments have been introduced into this process by methacrylation, a-bromoisobutyril bromide, and tosylation. These pre-treatments might involve modifying the samples with hydrophobic monomers or copolymers, which would result in an improvement of water repellence and dimensional stability (Ermeydan et al. 2014).

Nonbonded Nonleachable (NBNL)

As mentioned before, NBNL classes include thermosetting resin as impregnating agents. The term “thermosetting resin” refers to a specific kind of synthetic resin with high or low molecular weight oligomers. These oligomers are able to penetrate the cell wall and undergo subsequent polymerization, which is mainly polycondensation, to form the infusible and insoluble high molecular weight polymers. According to Stefanowski et al. (2018), there are two main types of thermosetting resin used for wood impregnation: formaldehyde-based resin, such as urea-formaldehyde (UF), phenol-formaldehyde (PF), melamine-formaldehyde (MF); and non-formaldehyde-based, such as methylene diphenyl di-isocyanate (MDI) and 1,3-methylol-4,5-dihydroxyethyleneurea (DMDHEU). In addition, tannin and furfuryl alcohol are included in this category but will not be discussed in this sub-section. A summary of thermosetting resin in both types based on Stefanowski et al. (2018) can be seen in Table 2. These resin types have led to a significant improvement on dimensional stability as an increase of WPG (Table 3). Pittman et al. (1994) investigated various types of MF resin. They discovered that the WPG and dimensional stability for MF-, plasticized MF- (PMF), methylated MF (MMF), and melamine-ammeline-formaldehyde (MAF)-treated wood increased significantly. The WPG ranged from 43 to 91%, whereas ASE was around 8% to 53%. According to Deka et al. (2000), high molecular weight (MW) of PF and MF resins resulted in WPG values around 12.1% to 40.8% with ASE around 29.3% to 70.6%. In addition, high MW PF yielded a greater increase in WPG and ASE than that of MF. This is likely because PF has a higher bulking coefficient than that of MF. According to Deka et al. (2000) the bulking coefficient (BC) ranged around 5.5% to 15.4% and 4.5% to 14.3% for PF and MF, respectively. UF impregnation could result in WPG above 110% and ASE less than 20%.

It is well known that thermosetting resin could be deposited in the lumen cells and function as bulking agent. Frihart (2004) proposed four different scenarios for resin penetration into the cell wall: first, resin simply occupies the free volume within the cell wall, thereby preventing shrinking and swelling; second, there is a mechanical interlocking effect due to cured resin extent from the lumen into the cell wall; third, the polymer network is composed of the cross-linked resin within the free volume of the cell wall; fourth, there is chemical crosslink formation with the cell wall polymeric components. This phenomenon is supported by the relationship between WPG and BC in various thermosetting resins (Fig. 3a). As can be seen in Fig. 3a, an increase in WPG may result in an increase in BC. Higher BC implies that a greater proportion of void cells covered by this compound, resulting in greater dimensional stability improvement. According to Fig. 3b, MUF may provide the highest WPG, followed by PF resins and PF oligomers. When compared to MUF and PF oligomers, the PF resin may provide better BC.

The use of high molecular weight (HMW) of resins could only provide gross penetration, indicating micrometer level penetration. According to Kamke and Lee (2007), gross penetration is caused by the passage of liquid resin into the porous structure of wood due to hydrodynamic flow and capillary action, mostly filling cell lumens. Hill (2006) stated that pre-polymerization of the resin before treatment led to a significant reduction in ASE. This was owing to the fact that bigger molecules had a restricted capacity to enter the cell wall. As seen in Fig. 3c, greater MW leads to a lower WPG and BC, while different phenomenon occurred for low molecular weight (LMW) resins.

Stamm and Seborg (1939) outlined three essential criteria for effective resin treatment of wood: 1. The size of the molecules needs to be sufficiently small enough to allow for penetration of the cell wall; 2. The resin molecules need to be soluble in polar solvents in order to allow for diffusion into microstructure of the cell wall; 3. The resin molecules need to possess sufficient polarity in order to exhibit high affinity with the cell wall. In light of these criteria, it is important to emphasize that the MW of resin needs to be small enough in order to accomplish cell wall penetration through the nanostructure of tissues. This was in agreement with Kamke and Lee (2007), who stated that the MW distribution, viscosity, solid content, and surface tension of the resin would directly affect its penetration. Thus, in turn, will contribute to the enhancement of wood properties, especially dimensional stability.

Table 2. Summary of Different Thermosetting Resin Type for Impregnation Process

Table 3. Thermosetting Resin Impregnation on Wood using Vacuum-Pressure Method

Note: WPG = weight percent gain; ASE = anti-swelling efficiency; TR/SR = tangential and radial swelling; WA = water absorption; BC = bulking coefficient, NA = not available, PF = phenol formaldehyde; LMW-PF = low molecular weight PF; MMF = methylated melamine formaldehyde, MUF = methylated urea formaldehyde. (*) marks representing solid content.

(a) (b) (c)

Fig. 3. Relationship between weight percent gain and bulking coefficient in different thermosetting resins (Table 3).

According to the findings of Klüppel and Mai (2013), LMW-PF treatment might offer a greater WPG (50%) than MMF treatment (40%) at a similar concentration of impregnating solution. Furuno et al. (2004) added that the amount of resin loading for LMW resin treated wood was found to be approximately 5.9% to 62.5% at a concentration of 1% to 15%, respectively. In addition, BC and WA were 12.6% and 37.1%, respectively, with ASE around 65% at 10% concentration. The lumina of tracheid and ray parenchyma cells contained very little or no phenolic resin when the concentration was between 1% and 5%. At 10% concentration, it was possible to observe a few dispersed cells that had phenolic filling their lumina. Even at concentration of 15%, the number of cell walls that had been filled with resin was extremely low and also scattered. This phenomenon can be observed in the photograph taken with a light microscope. The penetration mechanisms of LMW resins are also reported by Furuno et al. (2004). LMW resin is able to penetrate easily into the cell walls, and nearly all of it was found to be located in the wall polymer, with little or no presence resin in the cell lumina. When the concentration of resin is raised, any excess resin that is unable to enter the wall should remain in the lumina, coating and/or filling the space. The cells that had resin in their lumina were distributed sporadically throughout the tissue. As more resin penetrates the walls to form wall polymer, the walls will become bulkier and will absorb less water, leading to an increase in ASE as a direct consequence of this process. Girins et al. (2021) came to the conclusion that the MW of resin has a greater impact on enhancing dimensional stability than the solid content. Furthermore, it was also found that 30% WPG is necessary to penetrate wood cell wall and achieve ASE of 70%. Hill (2006) also mentioned that the improvement in dimensional stability has a significant correlation with the amount of mono-methylol phenol content that is present in the PF resin.

Bonded Nonleachable (BNL)

As mentioned before, BNL classes include anhydrides and organic acids as impregnating agents. Anhydride modification can be grouped into two categories, modification using cyclic anhydrides and non-cyclic anhydrides. Cyclic anhydride can be formed as succinic, alkenyl succinic, phthalic, maleic, glutaric, cyclohexane-1,2-dicarboxylic anhydride, etc. Meanwhile, non-cyclic anhydride, such as acetic, propionic, butyric, isobutyric, or hexanoic anhydride, may also be utilized (Hill 2006). These reactions can be carried out with or without catalyst at varying curing temperature (Li et al. 2001; He et al. 2020).

Among those anhydrides, cyclic anhydride can be represented by maleic anhydride (MAn). On the other hand, acetic anhydride (acetylation) has been the focus of substantial research for non-cyclic anhydride and is currently undergoing commercialization. Several studies have conclusively shown that the dimensional stability of anhydride-impregnated can significantly increase compared with untreated. He et al. (2020) investigated the impregnation of cyclic anhydride (maleic anhydride/MAn) at 110 °C for 8 h, which produced a high ASE with only 14.7% of WPG. Li et al. (2001) studied the impregnation of non-cyclic anhydride (acetic and propionic anhydride) with and without catalyst at a curing time 90 °C for 8 h in two different wood species. They discovered that acetic anhydride (acetylation) without catalyst can give higher WPG and ASE in hinoki, and yellow-poplar wood compared with propionic anhydride (propionylated). In the presence of catalyst (i.e., potassium acetate, sodium acetate, and sodium propionate), propionylated as well as acetylated wood of both hinoki and yellow poplar were significantly accelerated. Compared to acetylated wood, the propionylated wood exhibited a considerably greater increase in WPG and ASE. Research suggested that propionylation at 90 °C with catalysts could impart wood with a high dimensional stability. As shown in Fig. 4, WPGs of anhydride treated wood, particularly those treated with maleic, acetic, and propionic anhydride, did not exceed 25%, and the ASE values were around 80%. Hill (2006) concluded the amount of substitution for acetic or other linear chain anhydride rarely exceeds a WPG of 25% unless there has been cell wall damage. This demonstrates that anhydride can be an effective impregnating agent for less weight gain increases with higher dimensional stability.

Fig. 4. Relationship between weight percent gain and anti-swelling efficiency in different anhydride compounds (Data obtained from Li et al. 2001 and He et al. 2020).

The reaction between wood and cyclic anhydrides does not produce a by-product, and it results in the modified wood polymers with a covalently bonded carboxylic group. When the temperature is raised above 100°C, there is some formation of diester, which leads to cross-linking within the cell wall matrix (Hill 2006). On the other hand, when the reaction involving non-cyclic anhydride is taking place with the cell wall to form an ester bond, the by-product of this reaction is acetic acid. This phenomenon is in agreement with the findings of He et al. (2020) who stated that MAn can be partially replaced by hydroxyl groups and that the remaining hydroxyl groups are somewhat altered. One possible explanation for these alterations is that the carboxyl group formed as a by-product of MAn’s reaction with one single hydroxyl group. There is a different phenomenon reported by Li et al. (2001) with acetylation of wood. According to them, the swelling of the wood cell wall caused by the volume occupied by its compound and the blocking of the OH groups (bulking agents) is the primary mechanism of this process. Hill et al. (2005) investigated the cell wall micropore blocking for anhydride-modified wood. After the wood was treated, it was discovered that the degree of cell wall micropore accessibility was decreased significantly. This may be due to the chemically bonded adduct that occupy spaces in the cell wall. However, those mechanisms are still being debated between researchers. Hill (2006) stated that although cross-linking cannot be completely ruled out, hence bulking also contributed to the dimensional stabilization of wood.

Organic acids can be found in a variety of natural sources, including animals, plants, and microbes. They may be covalently bonded in groups such as amides, esters, and peptides and contain one or more carboxylic acid groups. Some commonly used organic acids include acetic, lactic, citric, malic, and ascorbic (Gurtler 2014). The effects of organic acids in wood get more attention since these acids may also play a role for wood dimensional stability.

Many studies have been using these acids in order to enhance the dimensional stability of wood. L’Hostis et al. (2017) conducted research on the topic of improving the characteristic of beech through in situ formation of polyesters of citric acid (CA) and tartaric acid (TA) in combination with glycerol at varying curing temperatures. It was found that CA and TA treatment of wood results in WPG values around 8.1% to 27.3% and 6.8% to 35.4%, respectively. These values were reduced with the increase of curing temperature. On the other hand, a different occurrence was observed with ASE, which intensified when the curing temperature increased. The ASE of CA- and TA-treated wood were 42.1% to 66.7% and 29.9% to 60.1%, respectively. In the study on the effects of lactic acid (LA) on the dimensional stability of wood, Grosse et al. (2019) found that the results were very encouraging. They observed that LA-treated wood cured at a variety of temperatures exhibited a similar phenomenon as previous studies. WPG and ASE of LA treated wood were 22.7% and 58.7%, as well as 20.3% and 61.4% at 140°C and 160°C, respectively. This could be owing to higher curing temperature, which leads to better fixation in the wood cell walls via ester bonds both in cases of organic acids treated wood. As shown in Fig. 5, WPGs of organic acid treated wood, especially using CA, TA, and LA, did not exceed 40% with ASE of 30% to 60%. This shows that organic acid can be an effective impregnating agent for less weight gain increases with higher dimensional stability, especially when proper curing temperature is applied.

Fig. 5. Relationship between weight percent gain and anti-swelling efficiency in different organic acid compounds (Data obtained from L’Hostis et al. 2017 and Grosse et al. 2019)

In general, the reaction mechanism that took place between wood and organic acid, in particular CA, involved a two-step esterification process. This began with the formation of anhydride, which was then followed by the reaction of cyclic anhydride with hydroxyl groups of wood; this sequence resulted in the formation of ester linkages (Lee et al. 2020). Fang et al. (2000) provided conclusive evidence that the cross-linking esterification reaction process takes place between polycarboxylic acid and wood. Instead of passing through a nucleophilic addition of the hydroxyl group and the carbonyl group, it was deduced that the crosslinking reaction went through cyclic anhydride intermediates.

As noted previously, the improved dimensional stability of modified wood has been found to be a function of WPG. From this section, it could be observed that impregnating agents in the first and second class (Fig. 1-3) could gain higher WPG than that of the third class (Fig. 4 and 5). However, the increasing of WPG in impregnated wood is not always correlated with an increase of wood properties, especially dimensional stability (Dong et al. 2015; Ahmed et al. 2017). According to Rowell (2005), when the WPG was exceptionally high, the ASE started to decrease, at least to some degree. In propylene, butylene-oxide, and methyl-isocyanate-treated wood, the ASE increases as the WPG increases up to 25% to 30%, while it begins to decrease at higher WPGs. This may be due to the fact that at a WPG of 33%, small cracks begin to form in the tracheid walls, and at a WPG of 45%, a major crack occurs within cells. A similar phenomenon also occurs in anhydride modification. Hill (2006) indicated that acetic or other linear chain anhydride rarely reaches a WPG of 25%, unless there is cell wall damage. In addition, oils impregnation increased WPG, and they predominantly accumulated on the surface area and/or remained in the cell lumens. Furthermore, an increase in WPG may cause microscopic cracks in the cell wall layers. According to Olsson et al. (2001a) a higher uptake between 75% and 105% could cause further damage due to the pressure gradient at the oil front. This process caused a change in the internal stress of the cell walls, which led to the formation of microcrack in the S1 layer. According to Rowell et al. (1978), succinic anhydride modified wood at 61% WPG exhibited extensive cell wall splitting. Hill and Jones (1996) concluded that dimensional stability does not increase much beyond a WPG of 30%.

Impregnation Parameters

Methods

Impregnation is accomplished by soaking or by vacuum-pressure to accelerate the impregnating agent into wood cell walls. The outcomes of these methods vary in terms of wood properties, especially dimensional stability. Hadi et al. (2019) compared the soaking and vacuum-pressure methods for treating three different kinds of wood species using MMA solution (Table 4). As shown in Fig. 6, the vacuum-pressure method resulted in higher WPG (14.6% to 23.7%) compared with the soaking method (8.7% to 13.4%).

Fig. 6. Comparison of weight percent gain produced from soaking and vacuum-pressure method in different wood species (Table 4)

Table 4. Comparison between Soaking and Vacuum-Pressure Method in Impregnation Process