Enhancement of Biological Durability and Fire Safety in Wood Modified with Maleic Anhydride and Sodium Hypophosphite

Abstract

Scots pine (Pinus sylvestris L.) sapwood was modified using maleic anhydride (MA) and sodium hypophosphite (SHP) to improve its durability against wood-deteriorating fungi, mechanical strength, and fire retardancy (thermal stability). The modification significantly reduced mass loss caused by wood-decaying fungi (Trametes versicolor, Rhodonia placenta, and soft rot fungi) due to the formation of cross-links between wood, MA, and SHP, which limited the moisture uptake and altered the chemical structure of wood. On the other hand, the modification did not provide improved resistance to fungi growth on the wood surface, which indicated that the modification had little impact on the accessibility of nutrients on the surface. A bending test showed that the modulus of elasticity (MOE) was not affected by the treatment, whilst the modulus of rupture (MOR) decreased to half the value of untreated wood. Thermal resistance was improved, as demonstrated by micro-scale combustion calorimeter testing, where the total heat release was halved, and the residue percentage nearly doubled. These results indicate that phosphonate protects the modified wood via the formation of a protective char layer on the surface and the formation of radical moieties. Based on the results, wood modified with MA and SHP shows potential for possible use in outdoor, non-loadbearing structures.

1. Introduction

Wood construction has grown in popularity over the past few decades, particularly due to society’s focus on sustainability. To meet the requirements of modern construction standards, it is essential to protect wood from moisture, biological decay, and fire. Wood modification has become an integral part of wood protection strategies, typically involving the application of a chemical, biological, or physical agent to the material. Active modifications result in a change to the chemical nature of the material, while passive modifications result in a change in properties without altering the chemistry of the material [1]. Active chemical modification of wood, which alters the chemical nature of wood by covalently bonding a chemical group to reactive part of the cell wall polymers [2], can provide sustainable wooden products by the chemical reaction between a reagent and accessible hydroxyl groups within wood components. The chemical bonding between the reagent and wood reduces the number of accessible hydroxyl groups and increases the bulking of the cell wall, resulting in a reduction in moisture uptake, improved dimensional stability, enhancement of the durability of wood, and limited risk of the chemical agents leaching out from the wood [3,4]. Whilst the major focus of wood modification has been around treatments, such as acetylation [5], furfurylation [6], DMDHEU [7], resin impregnation [8], and more recently, citric acid/sorbitol [9], a variety of other treatments have been considered, including isocyanates [10], citric acid [11], and mixed anhydride systems [12]. Thus, chemical modification has been recognized as an important alternative to wood preservatives which rely on the toxicity of the added chemicals [1], with each treatment offering a degree of protection against biological decay and/or dimensional instability.

As well as providing protection against biological decay and dimensional instability, protecting wood from fire has become a critical concern in timber engineering. In recent years, new additives have been introduced to enhance fire retardancy and smoke suppression, whilst minimizing toxic byproducts. One group, organophosphorus compounds containing P-C bonds have, in recent years, been developed as flame-retardant additives on various materials, such as cotton [13]. Organophosphorus-based flame retardants are known to be stable both thermally and hydrolytically, versatile in their flame-retardant action, and have the possibility to exhibit both condensed and gas-phase activities [14]. When applied to wood, phosphorus-based fire retardants suppress the smoldering combustion reaction of wood and can promote a change in the thermal-decomposition pathway of the wood material, thereby reducing flammable volatiles and increasing the yield of char residue [15].

Previous studies into the use of maleic anhydride have considered its application to solid wood [16], although the majority of studies have focused on its use as a compatibilizer in the preparation of wood composite products with polypropylene [17]. Wood modification using a combined treatment of maleic anhydride (MA) and sodium hypophosphite (SHP), based on their ability to react with cotton [18], has demonstrated that this modification can enhance dimensional stability [19] and reduce water uptake and the affinity of the wood cell wall to water [20], through the formation of phosphonate ester cross-linking [21]. However, it has not yet been revealed how this modification process alters wood properties. Since the wood–water relation affects wood characteristics, such as dimensional stability, mechanical properties, and biological resistance [22], it is likely that the modification with MA and SHP changes the properties of treated wood. Furthermore, the phosphonate ester, which can be found in wood modified with MA and SHP [21], has the possibility to behave as a fire retardant. Hence, the objective of this study was to determine the influence of modification with MA and SHP on durability against wood-deteriorating fungi, mechanical strength, and thermal stability.

2. Materials and Methods

2.1. Preparation of Specimens

Scots pine (Pinus sylvestris L.) sawn timber from northern Sweden was used in this study. The specimens were prepared for various tests, as outlined in Table 1. All specimens were oven-dried at 103 °C to a constant weight prior to modification. Four groups were prepared, namely untreated specimens (R) as a reference, untreated and leached specimens (RL), specimens modified with MA and SHP (T), and modified and leached specimens (TL).

Specimens were impregnated with a 3.5 mol/L maleic anhydride (MA, CAS No. 108-31-6, Sigma Aldrich, Saint Louis, MO, USA) solution in acetone at 12 bar for 2 h, followed by heating at 115 °C for 2 h. The esterified specimens were then impregnated in an aqueous solution of sodium hypophosphite monohydrate (SHP, CAS No. 7681-53-0, Alfa Aesar, Ward Hill, MA, USA) (0.5 mol/L) for 2 h under reduced pressure, followed by heating at 170 °C for 6 h. The impregnation time for SHP was extended compared to previous studies [19,20] due to the dimensions of the specimens. Further details of the specimen preparation can be found in Kim et al. [19].

Half of the untreated and treated specimens were subjected to a leaching test based on a modified EN 84 standard [23] test, i.e., the specimens were first impregnated in de-ionized water for 1 h under reduced pressure, followed by immersion in water for ten days, with the water being replaced every 24 h.

Before testing, the oven-dried weights of each specimen, both after modification and following the leaching test, were measured to calculate the weight percentage gain (WPG) and initial dry mass.

Table 1. Dimensions and number of specimens per each treatment group (untreated, untreated, and leached, modified, and modified and leached) for each performance test.

Test	Specimen Dimension (T × R × L, mm)	Replicates per Treatment Group
Mini block test [24]	10 × 5 × 30	30 (10 per fungal species, 10 control)
Terrestrial microcosm test (EN 807)	10 × 5 × 100	40 (10 per harvest)
Laboratory mold test	50 × 4 × 50	5
Outdoor mold test	50 × 4 × 50	20 (10 per direction)
Bending test	10 × 10 × 200	10

Table 1. Dimensions and number of specimens per each treatment group (untreated, untreated, and leached, modified, and modified and leached) for each performance test.

Test	Specimen Dimension (T × R × L, mm)	Replicates per Treatment Group
Mini block test [24]	10 × 5 × 30	30 (10 per fungal species, 10 control)
Terrestrial microcosm test (EN 807)	10 × 5 × 100	40 (10 per harvest)
Laboratory mold test	50 × 4 × 50	5
Outdoor mold test	50 × 4 × 50	20 (10 per direction)
Bending test	10 × 10 × 200	10

2.2. Mini-Block Test—White- and Brown Rot

Specimens were tested against two commonly-studied fungal species, namely the white-rot fungus Trametes versicolor (strain CTB 863 A) and the brown-rot fungus Rhodonia placenta (strain FPRL 280), in a mini-block test devised from the EN113 standard [24,25]. We selected these two fungal species for our study to provide examples of one brown-rot and one white-rot fungus. Both species are regarded as model fungi and listed as mandatory in the European standard EN 113-2 [26]. The specimens were autoclaved at 120 °C for 20 min for sterilization and then placed on sterile plastic nets in petri dishes filled with malt agar where the test fungi had been cultivated. For Rhodonia placenta, agar from Sigma-Aldrich (Saint Louis, MO, USA) and Bacto™ malt extract (Thermo Fisher Scientific, Waltham, MA, USA) were used, while agar from Scharlau and malt from VWR (Radnor, PA, USA) were used for Trametes versicolor.

The test was conducted at 22 °C and 70% relative humidity (RH) for 8 weeks. To account for any potential changes in mass loss of the treated samples due to factors other than attack by the test fungi, control specimens according to the EN 113-1 standard [25] were included, i.e., wood samples in uninoculated culture dishes.

To determine mass loss, all specimens were oven-dried before and after the test at 103 °C until a constant mass was reached. The mass loss by fungal decay was calculated as follows:

$$Mass loss={\displaystyle \frac{m_0-m_1}{m_0}},$$

(1)

where m₀ is the oven-dried mass of specimen prior to the decay test and m₁ is the oven-dried mass after the test.

2.3. Terrestrial Microcosms (ENV 807)—Soft Rot

The resistance against soft rot was determined according to the European standard ENV 807 [27]. Two parts of compost soil (Nelson Garden, Tingsryd, Sweden) were manually mixed with one part of sand. Prior to the test, the water-holding capacity of the soil was determined according to the standard. A lidded plastic container was filled up approximately halfway with soil. Water was added to retain 95% of the water-holding capacity. The specimens were randomly arranged in the soil within a plastic grid. To maintain the moisture content of the soil during the experiment, the weight of the container was monitored regularly. For the first 24 weeks, ten specimens from each group were harvested and measured every 8 weeks. The final (fourth) harvest and measurement were conducted after 52 weeks.

2.4. Mold Growth in the Laboratory

A mixed spore suspension was prepared using a culture comprising Penicillium spp. (2021-142-A-1 NIBIO), Penicillium spp. (2015-21/1/3 NIBIO), Alternaria alternata (DSM 62010), Ulocladium atrum (DSM 63068), Cladosporium cladosporioides (DSM 62121), Aureobasidium pullulans (DSM 2404), and Sydowia polyspora (BAM 31 (DSM 3498)). Prior to preparing the mixed spore suspension, a monoculture of each fungus was cultivated on malt agar in a petri dish. To collect spores of each individual mold, 5 mL of deionized water was added to each mold fungi culture, and the spores were harvested with a sterile Drigalski spatula and mixed in a beaker. The resulting suspension contained 296 × 104 spores per mL. The number of spores were counted using Burkard hemocytometer.

For the growth medium, agar from Sigma-Aldrich and Bacto™ malt extract (Thermo Fisher Scientific, Waltham, MA, USA) were used. Wood specimens were sterilized by autoclaving at 120 °C for 20 min and placed on the plastic net in petri dishes filled with agar malt. 500 µL of suspension was pipetted to the front face of the specimen and spread with a Drigalski spatula. The specimens were placed at 22 °C and 70% RH for 60 days while being observed and graded visually every week to one of the respective five ratings (0: no growth, 1: up to 10%, 2: 10%–30%, 3: 30%–50%, 4: 50%–100%) according to the EN 15457 standard [28].

2.5. Outdoor Weathering Test

The specimens were placed outdoors on racks facing south at a 45° inclination and vertically in the northern direction in Skellefteå, Sweden (64.744453° N, 20.955569° E). The outdoor test was performed between September 2023 and June 2024.

2.6. Bending Test

All specimens were conditioned at 20 °C and 65% RH until equilibrium was reached prior to the bending test. The modulus of rupture (MOR) and the local modulus of elasticity (MOE) in bending were tested according to the EN 408 standard [29] using a four-point bending test with universal testing machine (MTS System corporation, Eden Prairie, MN, USA) in a laboratory climatized at 20 °C and 65% RH. The span was 180 mm, the distance between the load points was 60 mm, the loading rate was 0.03 mm/s, and the load cell was 10 kN.

2.7. Thermogravimetric Analysis

Thermogravimetric analysis (TGA) (PerkinElmer TGA 4000, Waltham, MA, USA) was used to analyze the thermal behavior of the specimens. Approximately 8 mg of specimen from the radial surface of the specimens were collected using a microtome steel blade (Leica DB80 LX, Nussloch, Germany), loaded in alumina crucible and heated from 30 °C to 800 °C at a rate of 20 °C/min under a nitrogen purge gas at a flow rate of 120 mL/min. Each treatment group was tested in triplicate. The first-order derivative of the TGA curve was calculated using Origin 2021 software (OriginLab, Northampton, MA, USA).

2.8. Microscale Combustion Calorimeter (MCC)

The microscale combustion calorimeter (MCC) (Fire Testing Technology Ltd., East Grinstead, UK) was used to measure key parameters, such as heat release rate (HRR), peak heat release rate (pHRR), total heat release (THR), and heat release capacity (HRC), in accordance with the ASTM D7309 Method A standard [30]. Approximately 8 mg of specimen was loaded into an alumina crucible and heated up to 750 °C at a heating rate of 1 K/s under N₂ flow (80 cm³/min) in the pyrolysis chamber. The evolved gases were mixed with oxygen gas (20 cm³/min) before entering the combustion chamber, which was at a constant temperature of 900 °C. The residue at the end of the test was weighed to calculate the char yield. The tests were conducted for each sample and the average results were reported.

2.9. Statistical Analysis

To determine the significance of difference between treatment groups, all results were analyzed using analysis of variance (ANOVA). A probability of 0.05 was used as the threshold for type-I error.

3. Results and Discussion

The weight percent gain (WPG) of Scots pine sapwood modified with MA and SHP was 12.5 ± 1.5%. After leaching according to the EN 84 (CEN 2020), the WPG was 11.9 ± 0.2%, which did not bring significant difference. This is consistent with findings from a previous study using the same method [19]. Throughout the study, leaching test according to the EN 84 standard [23] did not result in significant changes to the tested wood properties.

3.1. Resistance Against Wood-Decaying Fungi

The mass loss observed in the mini-block test is presented in Figure 1. The modified wood exhibited significantly improved resistance against both Trametes versicolor (white rot) and Rhodonia placenta (brown rot), with mass loss below 1% for both treatments. In contrast, the mass loss of the untreated samples exposed to R. placenta was in the expected range (minimum 20%), while the mass loss of untreated samples decayed by the white-rot fungus T. versicolor was slightly lower than expected (minimum 15%–20%) according to EN 113 [25]; however, this result aligns with previous study [31]. This discrepancy may be attributed to the shorter incubation time (8 weeks) and to the fact that Scots pine is not a preferred substrate for T. versicolor.

The mass loss in the untreated samples was significantly higher than those of the treated samples, confirming the effectiveness of the modification. The check test specimens (i.e., wood exposed to malt agar without fungal inoculum) did not lose any mass, which meant the chemical agents used in wood modification did not leach out during incubation. These findings align with previous studies on wood modification with MA alone, where a 5.5% weight gain following reaction with MA on Japanese red cedar (Cryptomeria japonica D. Don) reduced mass loss from T. versicolor exposure to 1.8% [32].

Furthermore, the modified wood exhibited good resistance against soft rot fungi (Figure 2). During the first 60 days, the mass was not significantly different between sample groups. However, as the test progressed, degradation in the untreated reference specimens (groups R and RL) accelerated, whereas the treated specimens (groups T and TL) remained more resistant. After 52 weeks of incubation, the modified samples exhibited less than 4% mass loss.

The modification of wood with MA and SHP involves the penetration of chemical reagents into the cell wall and their reaction with wood components [21]. This was observed in earlier work as a bulking effect [19] and the formation of C-P bonds after the treatment [21]. The modified cell wall can hinder effective contact between fungi and the modification site of substrate by occupying space in cell wall [33]. As a result of chemical alteration of wood polymers, the modified wood may no longer be a viable nutrient source for fungi, given their substrate-specific nature [33,34,35]. Furthermore, a previous study on wood–water interaction in wood modified with MA and SHP reported a reduction in the moisture content in the cell wall [20], which can further inhibit fungal attacks on modified wood [2,33,35].

3.2. Resistance Against Surface Fungi

The mold growth on the specimens was studied using a mixture of common mold species. Initially, mold growth began at the edge of the specimen and gradually spread across the surface. At the beginning of the incubation, hyphae seemed to grow faster at the edge of R and TL specimens than on the RL and T specimens. However, the growth on the surface did not show significant differences between groups (Figure 3). After 25 days of incubation, dark fungal clusters became visible on the surface of untreated wood, with continued expansion as the test progressed.

On modified specimens, the hyphal growth of Penicillium spp. was easily distinguishable due to its color contrast against the darker wood. In contrast, dark-colored molds were less visible on the modified specimens due to their appearance. Nonetheless, some evenly distributed discoloring fungi were still observed on the surface of modified specimens (Figure 4). However, unlike the untreated specimens, modified specimens did not develop dense clusters of discoloring fungi.

Overall, the modification with MA and SHP was not highly effective in preventing mold colonization (Figure 3), which could be attributed to the different modes of decay, as surface molds are known to degrade cell walls to a lesser extent, instead relying on nutrients easily available on wood surface (e.g., extractives and free sugars) [36]. Furthermore, the mold fungi can still penetrate the modified wood and continue growing [35]. Based on mold experiments, it appears that MA and SHP do not alter the nutrients upon which surface molds feed.

3.3. Outdoor Performance

The outdoor test yielded similar results to the laboratory test, with mold growth observed on all specimens. This indicates that the modification did not provide protection against mold growth outdoor conditions. The color of untreated specimens became darker and turned greyer due to UV degradation and mold growth, whereas the modified specimens, which were initially darker, became lighter over time (Figure 5). This type of color change in the specimens was not observed during the laboratory mold growth test, suggesting that the discoloration in the outdoor test was primarily influenced by weathering.

3.4. Mechanical Properties in Bending Test

Modification with MA and SHP reduced the modulus of rupture (MOR) while the modulus of elasticity (MOE) did not change significantly (

Table 2. Modulus of elasticity (MOE) and modulus of rupture (MOR) of Scots pine sapwood (untreated (R), untreated leached (RL), modified (T), and modified and leached (TL)). Standard deviations are shown in parentheses.

Groups	MOE (GPa)	MOR (MPa)
R	11.4 (1.99)	83.8 (10.82)
RL	10.2 (1.87)	71.5 (10.32)
T	9.9 (2.41)	35.1 (13.83)
TL	9.7 (1.98)	36.2 (14.1)

). The lower MOR of the modified samples may be attributed to the degradation of hemicelluloses during the modification process, as hemicelluloses are less heat-resistant compared to lignin and cellulose [37,38,39].

3.5. Thermal Behavior

Thermogravimetric analysis results for the samples are shown in Figure 6. Both untreated and treated specimens exhibited an initial mass loss due to the evaporation of physically bonded water (<105 °C). The main decomposition began at approximately 250 °C. For untreated wood, rapid thermal decomposition of cellulose and hemicellulose occurs between 250 °C and 380 °C, involving the cleavage of glycosidic linkages, dehydration, decarbonylation, and decarboxylation [40,41]. Lignin decomposes over a broad temperature range, starting at 200 °C and continuing up to 900 °C, with primary decomposition occurring beyond 500 °C due to the higher thermal stability of its aromatic structures compared to polysaccharides [41]. However, between 380 °C and 600 °C, the bonds between monolignols are cleaved, and monomeric phenols are vaporized [42].

However, the modified specimens showed a different mass loss behavior. The highest mass loss rate of modified specimens occurred between 300 °C and 350 °C, which is lower than that of the untreated specimens. Within this range, the mass loss of the modified specimens was faster than that of the untreated specimens, possibly due to the cleavage of bonds between the wood and phosphonate, as well as the degradation of the material. The loss of side chains from the phosphorus ester in this temperature range has been observed in previous studies [43,44]. At elevated temperatures, however, the modified specimens showed a lower mass loss with higher residue (31%) than untreated specimens (13%) at 800 °C, indicating an alteration in the thermal degradation pathway.

The combustion behavior, analyzed by microscale cone calorimeter (MCC), showed results consistent with those obtained from thermogravimetric analysis (TGA) (Figure 7,

Table 3. Microscale combustion calorimeter test results of total heat release (THR), peak heat release (pHRR), and pHRR temperature, mass residue, and heat release capacity (HRC). Standard deviations shown in parentheses.

Parameters	R	T	TL
THR (KJ/g)	11.18 (0.1)	6.1 (0.1)	6.6 (0.2)
pHRR temperature (°C)	385.1 (2.6)	329.9 (1.1)	324.2 (3.1)
pHRR (W/g)	145.4 (4.6)	85.68 (5.2)	94.35 (6.5)
Residue (%)	16.9 (1.9)	29.4 (2.6)	27.5 (1.9)
HRC (J/g∙K)	148.4 (4.1)	87.3 (5.4)	96.5 (6.8)

). The peak heat release rate (pHRR) of modified specimens occurred at a lower temperature compared to untreated wood, and a shoulder was observed between 370 °C and 500 °C, suggesting a multi-step decomposition process [45]. The modified wood showed lower pHRR values as well as total heat release (THR). Furthermore, the HRC of modified wood was significantly lower than the untreated wood, which indicates the fire safety of wood can be enhanced by modification with MA and SHP (Table 3).

The presence of a cross-linked phosphonate structure in modified wood is central to these observations. At relatively low temperatures, phosphonate decomposes, releasing phosphorus-containing acids (e.g., phosphoric acid), which catalyze dehydration reactions and promote char formation. This is consistent with the earlier decomposition observed in TGA and the increased percentage residue in both TGA and MCC (Table 3). Additionally, the dehydration process releases water, which may dilute the fuel in the gas phase, potentially reducing combustion intensity. At higher temperatures, the C-P-O bonds in phosphonates may break, generating phosphorus radicals (e.g., PO• or PO₂•) that interfere with flame propagation by quenching combustion-related radicals (e.g., •OH, •H) [43,44,46].

4. Conclusions

This study demonstrated that modifying Scots pine sapwood with maleic anhydride (MA) and sodium hypophosphite (SHP) significantly enhanced its resistance to fungal decay, resulting in less than 5% mass loss while preserving its structural integrity after leaching. The improved durability is attributed to cross-linking, which reduces moisture uptake and alters the wood chemistry.

However, the modification did not prevent surface mold growth in both laboratory and outdoor conditions. Additionally, UV exposure led to gradual color lightening. When it comes to mechanical properties, the MOE remained largely unchanged, while the MOR decreased to half of untreated wood. Thermal analysis of modified wood showed nearly doubled residue (%) and two-thirds of HRC compared to untreated wood, suggesting potential fire resistance benefits.

Overall, wood modification with MA and SHP can be a promising method for enhancing wood’s durability and thermal stability without enhancing its mechanical properties. The modified wood can, therefore, be suitable for outdoor non-loadbearing structures.