1. Introduction
Wood has long been associated as an important natural resource, given its use by humanity to produce such diverse items as art and furniture and its use in construction. The use of wood is seen as a necessity for sustainable construction in our modern society [
1]. As greater importance is placed upon the performance of wood as a material, it has become necessary to alter some of the inherently undesirable properties limiting the longevity of service performance. Of particular importance are issues relating to limitations related to stability in service, susceptibility to fungal decay or weathering, etc., [
2,
3]. Thus, ways of improving factors such as moisture sensitivity, low dimensional stability, low hardness and wear resistance, low resistance to bio-deterioration against fungi, termites, marine borers and low resistance to UV radiation have become important parameters in wood treatment technologies and in particular wood modification treatments.
Nowadays, wood modification is defined as a process adopted to improve the physical, mechanical or aesthetic properties of sawn timber, veneer or wood particles used in the production of wood composites. This process produces a material that can be disposed of at the end of a product’s life cycle without presenting any environmental hazards greater than those that are associated with the disposal or combustion of unmodified wood. Recent publications [
4,
5] indicated how wood modification is becoming more established across Europe at both the commercial large and the niche manufacturing scale. A range of methods have become available, with the focus on thermal [
6,
7,
8,
9] and chemical treatments [
10,
11,
12,
13] being the most studied. The subject has attracted several reviews, e.g., [
14,
15,
16], and is hereby introduced in more detail.
The thermal modification of wood is now an accepted means of treating wood, with several commercialised processes. Depending on the process, humidity, temperature and time [
17], a range of reactions may occur, such as hydrolysis, oxidation and decarboxylation reactions, along with physical changes. The results of such treatments are end products presenting improved physical characteristics, such as lower hygroscopicity, better dimensional stability and durability, though mechanical properties are often significantly reduced, with dynamic mechanical properties (such as impact bending, fatigue) typically being more affected than static properties. Improvements in stability and durability are of importance where wood is exposed to different chemicals or biological agents such as fungi and bacteria, or to frequent use under natural environmental conditions.
Another possible type of wood modification is the reaction with different chemicals—chemical modification—where chemical moieties are covalently bonded to the wood cell wall polymers. Different chemical modifications [
11,
12,
13,
18,
19,
20] may be used to alter the properties of wood [
20], particularly for improving the dimensional stability, decay resistance and water performance of wood, and to improve mechanical properties compared to those from typical thermal modification processes [
21]. Most modification methods involve the hydroxyl groups in the cell wall, which are partially substituted, and the cell wall of the wood is bulked with the bonded chemicals. The substitution of the hydroxyl groups reduces the number of primary sorption sites (typically the OH groups), while the bulking reduces the volume in the wood cell wall which is available to water molecules. The most common chemical modification process involved reaction with anhydrides, with a multitude of examples listed in the literature. However, it is the acetylation process using acetic anhydride that has been the most studied, resulting in its commercialisation [
2]. Apart from anhydrides, many other chemicals have been used to try to improve the properties of wooden material. For example, treatments with dimethyloldihydroxyethylene urea (DMDHEU), melamine resin, silane or silicon polymers were found to improve the mechanical strength of wood [
3].
More recently, the use of Maillard reactions for wood treatment proved to be a promising modification method to improve some of the wood properties. This type of reaction is well-known in food chemistry, where it is responsible for the browning in many foods during baking [
22,
23]. The essence of the reaction is that a reducing sugar condenses with a compound possessing a free amino group to give a condensation product [
24]. Subsequently, a range of reactions takes place, including cyclisations, dehydrations, retroaldolisations, rearrangements, isomerisations and further condensations, which ultimately lead to the formation of polymers and co-polymers, known as melanoidins [
24]. The composition of its chemical structure is relatively unknown due to the complexity of the products that are generated in the reaction [
25]. The advantage of this reaction is that it is an aqueous process and initiated by heat only, making it relatively straightforward to apply to wood in a commercial process. In addition, the reaction does not require the use of strong acids or bases, which could degrade the wood structure.
In earlier experiments [
26], the influence of the Maillard reaction on European beech and Scots pine was investigated. The wood was impregnated with an amine (glucosamine, lysine or glycine), sugars (glucose or xylose) and an extra reagent to improve the reaction (magnesium chloride, maleic acid or citric acid). The results showed that when lysine, glucose or citric acid were reacted, a high weight percentage gain (WPG) was obtained (18% for beech and 40% for pine). After a leaching procedure, the WPG for beech was 11% and 25% for pine, respectively. This preliminary screening reaction has shown that the Maillard reaction offers potential as a potential new wood modification system, though a reduction in leaching would help maintain the modification benefits.
In this context, a recent work of Hauptmann and co-workers [
3] considered the use of tricine for the modification of wood. This is capable of binding with reduced sugars, though studies have been limited to a maximum temperature of 103 °C. They observed increased hardness and the tensile strength of their modified wood species. As indicated earlier, Maillard type reactions are a cascade of reactions (known also as nonenzymatic browning). The initial step consists in a reaction between a reducing end of a saccharide and an amino acid [
3,
26]. This reaction takes place usually during the heat processing of food with a low relative humidity [
27] and it is considered to cause the brown stains during wood kiln drying with temperatures higher than 80 °C due to the presence of natural amino acids and reduced sugars [
3]. The most efficient characterisation method to identify the modifications appearing in the wood structure during heating and to identify the possible interaction and bonds formed between the wood components and the reagents used in chemical modification is infrared spectroscopy. This technique proved to be an efficient tool to identify small modifications in the wood structure appearing during different treatments [
17,
28,
29] such as decay, i.e., [
30,
31] or photodegradation, i.e., [
32,
33]. Moreover, due to the use of the small amount of sample and little or no processing, this method is considered by many to be non-destructive.
In this study, the combined effects of chemical treatment and the thermal modification are investigated in terms of the use of bicine [2-(Bis (2-hydroxyethyl) amino) acetic acid] and tricine [N-(2-Hydroxy-1,1-bis (hydroxymethyl)ethyl) glycine] (
Figure 1). The concept of the Maillard reaction depends upon the availability of an active site on the nitrogen within the amino group (i.e., a N-H group). Whilst this is not present in bicine, it was postulated within this work that the compound may undergo side group displacement, allowing the general concepts within the Maillard reaction to proceed. Some of the resulting properties for the treatment of beech are considered (mechanical, physical and bioresistivity), along with infrared spectroscopic analysis in an attempt to demonstrate actual chemical bonding.
2. Materials and Methods
In this study, the potential of combining chemical and thermal treatments for European beech (Fagus sylvatica L.) and Norway spruce (Picea abies (L.) H. Karst) was determined. For both species, material sourced and felled in Slovenia was used. For each treatment set, separate sets without visible growth features and varying ring widths, mainly in semi-radial orientation, were prepared from heartwood sections. Thus, only mature wood was used for these experiments. The density ranges of the samples were within the typical oven-dry ranges commonly noted in the scientific literature. Thus, bicine (Bi, CAS Number 150-25-4, Fisher Scientific, Waltham, MA, USA) and tricine (Tri, CAS Number 5704-04-1, Fisher Scientific, Waltham, MA, USA) were impregnated into wood specimens, using a 10% solution (weight:volume, w:v) of the respective compound per impregnation.
2.1. Specimen Treatment
Beech (hereafter referred to as B) and spruce (hereafter referred to as S) specimens of varying dimensions depending on subsequent testing were prepared and subjected to the following treatment regimes (
Table 1).
For the impregnation procedure, the solutions were introduced using a vacuum- pressure impregnation (VPI) process according to the full-cell process in a laboratory impregnation setup (Kambič, Semič, Slovenia). It consisted of 30 min of vacuum (1.0 × 104 Pa), 40 min of pressure (10 × 105 Pa) and 10 min of vacuum (1.5 × 104 Pa). Reagent uptakes were subsequently determined gravimetrically. The impregnated specimens were conditioned (23 °C; 65% relative humidity (RH)) for 2 weeks prior to thermal modification. Non-impregnated and conditioned specimens served as controls.
The thermal modification (HT) was performed according to a modified Silvapro
® commercial procedure [
34] limited to 165 °C, due to the risk of thermal degradation of the chemicals if exposed to temperatures above 180 °C. Control specimens were only heated up to 100 °C during the drying procedure in atmospheric conditions. The time of thermal modification at the target temperature was 3 h and mass loss (ML) of the specimens after thermal modification was determined gravimetrically. The HT specimens were stored in the laboratory for 4 weeks (23 °C; 65% RH) before subsequent testing.
2.2. Physical Tests
2.2.1. Colour Analysis
Colour was determined on the semi-radial surfaces of a selection of specimens. The colour measurements were performed according to the CIE L*a*b* system, a method created by the Commission International de l’Eclairage. The CIE L*a*b* system is characterised by three parameters: L*, a* and b*. The L* axis represents the lightness, which varies from a hundred (white) to zero (black), representing the achromatic axis of greys, whereas a* and b* are the chromaticity coordinates. A positive value of a* denotes a redder colour on a green–red scale, whereas a positive value of b* denotes a more yellow colour on a blue–yellow scale. Together, those three components form a three-dimensional colour space. Colour measurements of in-service testing were performed several times a year with a portable Colour Measuring Device (EasyCo 566, Erichsen, Hemer, Germany) and expressed in the CIE L*a*b* system. This device enables contact-free precise colour measurement. The diameter of the measurement spot is 20 mm. However, laboratory test specimens were scanned and processed with Corel Photo-Paint 8 software. Corel Photo-Paint was used as colour analysis as this technique provides the colour of the whole surface and not of individual spots. Total colour difference ∆E* (Equation (1)) from a reference colour (L*0, a*0, b*0) to a target colour (L*1, a*1, b*1) in the CIE Lab space is calculated by determining the Euclidean distance between two colours given by:
2.2.2. Contact Angle Measurements
Contact angles were detected using a Theta optic tensiometer (Biolin Scientific Oy, Espoo, Finland) and OneAttesion 2.4 (r4931) software (Biolin Scientific, Espoo, Finland). Five replicates of each treatment were used, to which were added 2 droplets of water (4 μL each). The samples used were 25 × 15 × 50 mm in size, having been thoroughly dried (103 °C, 24 h). Subsequent changes in contact angle were determined using a 7.6-megapixel camera, with software constantly monitoring subsequent wetting over a period of 60 s.
2.2.3. Dynamic Vapour Sorption (DVS) Analysis
Samples for dynamic vapour sorption (DVS) analysis were milled in a Retsch SM 2000 cutting mill (Retsch GmbH, Haan, Germany) with a Conidur® perforation sieve with 1.0 mm perforations. Thus, several samples were milled together to create an average mix of fibres representing each treatment. The milled wood samples were conditioned at 20 ± 0.2 °C and 1 ± 1% RH through blowing with dry air. Analysis of the wood samples was performed using a DVS apparatus (DVS Intrinsic, Surface Measurement Systems Ltd., London, UK). A small amount (approximately 400 mg) of pre-conditioned wood chips was placed on the sample holder, which was suspended in a microbalance within a sealed thermostatically controlled chamber, in which a constant flow of dry compressed air was passed over the sample at a flow rate of 200 cm3/s and a temperature of 25 ± 0.2 °C. The schedule for DVS had two steps: 0% and 95% RH. The DVS maintained a given RH until the weight change of the sample was less than 0.002%/min for at least 10 min. The running times, target RH, actual RH and sample weights were recorded three times per min throughout the isotherm run. After the DVS analysis, equilibrium moisture content at 95% RH (EMC95% RH) was determined along with the hysteresis values, determined as the difference between desorption and adsorption values at a given RH.
2.3. Mechanical Tests
2.3.1. Mechanical Performance Tests
Modulus of elasticity (MOE) and modulus of rupture (MOR) were determined according to EN 310 [
35] with a static three-point bending test on a Zwick Z005 universal testing machine (Zwick-Roell, Ulm, Germany). In total, 60 specimens (10 replicates for each tested material) with dimensions 360 × 20 × 20 mm were prepared and oven dried (103 °C) until a constant mass was achieved. They were dried in order to eliminate the influence of different moisture contents between control and modified wood. The specimens were tested for MOE and MOR immediately after drying.
Compressive strength was determined according to the ASTM D1037-99 standard [
36] on a Zwick Z100 universal testing machine (Zwick-Roell, Ulm, Germany). In total, ten specimens for each test group with dimension 50 × 20 × 20 mm were prepared and oven dried until a constant mass was achieved. The specimens were tested for compressive strength immediately after conditioning under standard conditions. After the test, the compressive strength (F
m) was calculated.
The Brinell hardness (HB) was determined by a standard test method according to EN 1534 [
37]. The penetration depth (h) of an iron sphere (D = 10 mm) was used for 5 replicates, each undergoing 4 individual measurements, in calculations at load F = 1000 N to determine the Brinell hardness according to:
2.3.2. High Energy Multiple Impact (HEMI) Test
The development and optimisation of the high energy multiple impact (HEMI) test has been previously described by Rapp et al. [
38]. For testing modified and non-modified control specimens in a high-energy multiple impact (HEMI) test, specimens of 20 × 20 × 10 mm were split in four specimens of 5 × 20 × 10 mm (radial × tangential × longitudinal). Five times, n = 20 samples of 5 × 20 × 10 mm were submitted to the HEMI tests. HEMI tests were performed in a heavy-impact ball mill (Herzog HSM 100-H; Herzog Maschinenfabrik, Osnabrück, Germany). In short, 20 oven-dried (103 °C) samples were placed in the bowl (140 mm in diameter) of the mill, together with one steel ball of 35 mm in diameter plus three balls of 12 mm and 6 mm in diameter, respectively. For crushing the specimens, the bowl was shaken for 60 s at a rotary frequency of 23.3 s
−1 and a stroke of 12 mm. The fragments of the 20 specimens were fractionated on a slit sieve according to EN ISO 5223 [
39] with a slit width of 1 mm using an orbital shaker at an amplitude of 25 mm and a rotary frequency of 200 min
−1 for 2 min. The degree of integrity (I), fine percentage (F) and resistance to impact milling (RIM) were calculated following Equations (3)–(5), where m
20 is the oven-dry mass of the 20 biggest fragments, m
all is the oven-dry mass of all fragments and m
fragments<1mm is the oven-dry mass of fragments smaller than 1 mm. The RIM was calculated according to the optimised method [
38], which contains a three-fold weighting of the fine fraction (3 × F), with the values 300 and 400 guaranteeing to achieve a maximum RIM value of 100%.
2.4. Chemical Tests
2.4.1. Volatile Organic Compound (VOC) Analysis
In order to aid FTIR analysis and attempt to identify Maillard-type reaction products, it was decided to undertake gas chromatography–mass spectrometry (GCMS) analysis of samples. Thus, subsamples of the materials were cut and immediately placed into a minichamber (Markes International, Llantrissant CF72 8XL, UK). A flow of nitrogen at 2 mL/min was passed over the samples and through Tenax columns to collect any volatile organic compounds (VOCs) released. Samples were kept in the chamber for 2, 8 or 16 h. Collected VOCs were eluted form the Tenax columns using methanol and analysed using an electron impact (EI) capillary GCMS (Glarus 680 gas chromatograph, Perkin Elmer, using an Agilent VF5-MS column (30 m × 0.25 mm × 0.25 μm), coupled with a Clarus 600 C mass spectrometer, Perkin Elmer). GC conditions were an initial temperature of 60 °C for 1 min, followed by a temperature increase of 6.0 °C/min to a maximum temperature of 300 °C, with the final temperature being held for a further 10 min. Mass spectra were collected between elution times of 3–51 min from the gas chromatograph over an electron impact mass range of 40.00 to 600.00.
2.4.2. Infrared Spectroscopy
The infrared spectra of the reference and treated wood specimens were recorded in potassium bromide (KBr) pellets on a Bruker ALPHA FT-IR spectrometer with 4 cm−1 resolution. The concentration of the sample was a constant of 2 mg/200 mg KBr. Processing of the spectra was performed using the Grams 9.1 program (Thermo Fisher Scientific, Waltham, MA, USA).
2.5. Effects against Biological Deterioration Tests
In order to ascertain any biological effects of the treatments, the following aspects were evaluated within this section.
2.5.1. Resistance to Fungal Decay
In order to determine any potential effectiveness of treatment against wood-destroying fungi, tests were carried out according to EN113 [
40] once samples had undergone artificial ageing to remove possible extractives and other components that could act like surfactants. A similar process occurs in nature with the first extensive rain period. The European EN 84 standard [
41] describes a method for artificial ageing (leaching) of wood before testing the biological effectiveness. This standard was designed to simulate extensive leaching by natural precipitation. The first step was impregnation with demineralised water. The samples were stacked in a container, weighed down and vacuum impregnated (4 kPa) with demineralised water for 20 min and soaked for an additional 2 h. Samples were then immersed in water for 14 days and during this period water was replaced nine times. After the ageing process was completed, samples were dried in ambient conditions for 2 weeks prior to subsequent tests.
The decay test was performed according to the modified EN 113 [
40] on treated and untreated beech and spruce samples. Disposable Petri dishes containing 20 mL of 4% potato dextrose agar (PDA, Difco, NJ, USA) were inoculated with 3 different fungi: one white rot fungi (
Trametes versicolor (L.) Lloyd (ZIM L057)) and two brown rot fungi:
Gloeophyllum trabeum (Pers.) Murrill (ZIM L018) and
Poria monticola Murrill (ZIM L033). The fungal isolates originate from the fungal collection of the Biotechnical Faculty, University of Ljubljana, Slovenia. A plastic mesh was used to avoid direct contact between the samples and the medium. The assembled test dishes were then incubated at 25 °C and 80% RH for 12 weeks.
Samples of dimensions 8 mm × 25 mm × 25 mm were prepared and 5 replicates per fungus species were used for each type of treatment. The untreated beech and spruce wood samples served as reference wood species to assess the validity of the test. After incubation, the fungal mycelium was removed and the samples weighed for moisture content. After 24 h of drying at 103 °C, mass loss was determined gravimetrically.
2.5.2. Efficacy against Subterranean Termites
Subterranean termites belonging to the species
Reticulitermes grassei Clément were captured in a pine forest in Sesimbra, Setúbal district of Portugal and were brought to the laboratory and kept in a conditioned room at 24 ± 1 °C and 8 ± 5% RH. Groups of 150 workers of termites were established in 200 mL glass jars with moistened sand (Fontainebleau sand and water; 4:1
v/
v) as substrate. Three replicates (30 × 10 × 10 mm) per treatment were then placed in contact with the termites and the test run for four weeks at the described conditions. Maritime pine test specimens with the same dimensions were also included as internal virulence controls [
42,
43]. The initial moisture content of the blocks was measured in sets of three additional replicates per treatment and these values were used to determine the theoretical initial dry mass (IDM) of the exposed specimens (in all tests conducted). At the end of the trial, the final moisture content was recorded and the mass loss was obtained according to:
where FDM is the oven-dry mass of the block at the end of the test. The survival of the termites was also recorded and all wood blocks were graded in terms of termite attack using the scale: 0 = no damage; 1 = attempted attack; 2 = slight damage; 3 = superficial and inner damage; 4 = heavy inner damage.
2.6. Statistical Analysis
Where applicable, a multivariate analysis of variance (ANOVA) was performed to determine any significant affect as a result of specific treatments and resulting tests. Significance established at the p < 0.05. Tukey’s honest significance test was applied to find means that are significantly different from each other. Additionally, statistical analysis was undertaken using IBM SPSS Statistics V26.
4. Conclusions
The use of bicine and tricine as part of an enhanced thermal modification process has been considered. The results herein were part of establishing the feasibility of such combined chemical/thermal modification processes for treating wood. However, the thermal stability of the selected compounds resulted in the need for a reduced thermal modification temperature, which would have expected impacts on the effectiveness of the thermal modification on its own.
Studies using colour, FT-IR spectroscopy and VOC analysis seem to suggest there is some level of chemical interaction between the treatments and the wood resulting from the trials undertaken, particularly through the presence of compounds such as tetraacetyl-d-xylonic nitrile, though no clear evidence of exactly what the mechanism is has been determined to date. The use of principal component analysis with FT-IR studies confirmed specific peaks were a direct result of chemical reactions.
The hypothesis of combining chemical and thermal reactions of wood resulting in equivalent or better mechanical properties were not borne out in these studies, with three-point bending and compression tests being reduced. There were slight improvements in some Brinell hardness data, though this may be a direct result of the increased density resulting from the uptake of bicine or tricine. The lack of improvements may also be a direct result of the thermal instability of the bicine and tricine.
The results from fungal and termite attack of samples suggested that significant improvements were possible for experiments where the EN84 weathering had not been undertaken. Whilst moisture content levels for fungal decay tests on non-weathered samples suggested the possibility of water logging occurring, corresponding samples tested under standard hysteresis conditions did not show such significant uptakes of moisture. When samples underwent EN84 weathering prior to testing, the beneficial effects were not noted. This indicated that any benefits were lost when the bicine or tricine was leached from the samples. Thus, it would seem logical that further means of entrapment within the wood cell wall structure needs to be considered in future work.
Despite the inconclusive results from this study, there are sufficient indications to suggest some degree of reaction is occurring. Part of the issue with the process undertaken in this study was the use of an open system, whereby any potential intermediate moieties resulting from Maillard-type reactions were volatilised before subsequent reactions could occur. This could potentially be overcome through the use of a closed system reactor, thereby allowing these intermediate groups to undergo further reactions. In addition, the risk of thermal degradation of the key reagents in this study (bicine and tricine) at temperatures typically employed for more effective thermal modification (between 180 and 210 °C) may limit the degree of reactivity encountered in this study, even though reduced heat treatment temperatures (160 °C) were employed. Further consideration into the use of reagents that are more stable at elevated temperatures may enable the full effect of Maillard-type reactions to be explored in more detail.