Wood has many appealing characteristics, including a beautiful appearance, good strength, low density, and excellent insulating qualities. It readily absorbs moisture from its surroundings, creating swelling, and subsequent desorption causes shrinkage.
Unfortunately, once the forces exceed the wood’s fracture strength, it will creep and break in reaction to these wetting–drying loads. Furthermore, it tends to develop longer and deeper cracks at later stages.
These defects also contribute to increased water uptake, thus producing optimal moisture conditions for wood-decay fungi to attack. Wood modification can overcome various drawbacks of wood, including poor dimensional stability, poor biological durability, low surface hardness and wear, fire retardancy, high equilibrium moisture content, and poor weathering stability.
According to statistics collected, thermal modification (TM) processes dominate the global commercial production of treated wood, producing 1.11 million cubic metres per year. Acetylation accounts for 120 thousand cubic metres per year of commercialised processes, while furfurylation accounts for 45 thousand cubic metres per year.
Another 330 thousand cubic metres per year of global production is predicted for various methods, primarily based on resin-based systems (e.g., Impreg-and Compreg-based processes). Oy Lunawood (Thermowood process) is the largest TM manufacturer, with an annual manufacturing capacity of 155 thousand cubic metres.
The Estonian company Thermory AS (Wood treatment technology process) also has a high capacity of approximately 90 thousand cubic metres per year spread across five production facilities in Estonia, Finland, and Belarus.
TM in nitrogen has not grown much in terms of market share, and data on current production volumes are unavailable. Overall, the commercial amount of modified wood is still dwarfed by that of preservative-treated wood (about 21 million cubic metres in the US alone each year and 6.5 million cubic metres in Europe).
Numerous wood-based TM processes have entered the European market, mostly varying in terms of pressure conditions and oxygen exclusion.
Thermal Modification Technologies
There are two types of TM technologies: open-system and closed-system technologies. The majority of open systems are upgraded classic drying ovens.
A pyrolysis technique at high temperature and normal atmospheric pressure is used to perform TM on wood. The world’s largest TM wood factories, including the International ThermoWood Association, treat wood in a water steam environment at atmospheric pressure.
Closed systems hydrolyse wood materials at high pressure but at a lower temperature, so the pressure serves as a temperature substitute.
The fundamental advantage of TM in steam at higher pressure over atmospheric pressure treatment is lower maximum treatment temperatures and, as a result, cheaper energy costs.
Another advantage is that the entire TM time from heating, incineration, and cooling to room temperature is reduced. The total processing time in a closed-system process is 22–30 hours until the material has completely cooled. ThermoWood requires up to 70 hours to complete the process.
The Rétification process is distinguished from other thermal modification procedures by the use of nitrogen as the modification medium. In specific equipment, pre-dried (Wrel = 12%) wood is slowly heated to 210–240 deg C in a nitrogen environment with an oxygen content below two percent. The maximum pressure in the system during the TM period is questionable.
There is no mention of it in the literature. Wood Treatment Technology recently launched a new closed-system TM procedure in a nitrogen atmosphere. The TM process under pressure in nitrogen has a higher heat transfer capacity and requires no energy to convert water to steam.
The procedure is carried out at higher pressures (up to 18 bar), with the elevated pressure acting as a temperature replacement. As a result, the maximum modification temperature and processing cycle time can be reduced to 8–12 hours.
In terms of manufacturing costs and productivity, the closed-type process in nitrogen appears to be a superior technology that necessitates more extensive scientific research. The scientific literature contains a wealth of information about the TM of several wood species in a nitrogen environment at atmospheric pressure.
However, insufficient information on the TM process in a nitrogen environment at elevated pressure and the qualities of the TM wood is known dating back to the 1970s and 1980s.
The dimensional stabilisation of wood by the Feuchte–Wärme–Druck process was evaluated in a 1.8 cubic metres volume pilot-scale reactor at 180–200 deg C in an inert gas environment of 8–10 bar. Heat/pressure treatment of beech, birch, poplar, pine, and spruce wood, fibre, particle board, and plywood is demonstrated.
After treatment, there is a 50–80 percent reduction in swelling and shrinkage, an improvement in resistance to fungi and insects, and an improvement in the remaining physical and mechanical wood qualities.
When the maximum temperature is held, it appears that the pressure in the reactor remains constant. However, the entire process time (5–15 hours) is ambiguous and not specified in the temperature and time plan, and the heating/cooling speed is suspect.
Weathering stability, equilibrium moisture content (EMC), and swelling of European spruce (Picea abies) and beech (Fagus sylvatica) wood were investigated after TM in nitrogen at 175 deg C for 2 hours and 185 deg C for 3 hours at 10 bar pressure.
The data are difficult to compare because EMC (at relative humidity (RH) 30, 65, 80, and 100%) and swelling are presented as percentage reductions.
According to the data provided, the approximate EMC values for TM spruce are six to eight percent (65% of RH) and 17–20 percent (100% of RH). The untreated control EMC values were 12–13 percent and 30 percent, respectively.
TM beech wood EMC was five to seven percent (65% RH) and 13–23 percent (100% RH), whereas untreated control EMC was 10 percent and 30 percent, respectively. In comparison to other specimens, the EMC of TM beech wood at 185 deg C was the lowest.
TM in a nitrogen environment was generally conducted in small-sized laboratory reactors for one to eight hours, with a total processing period of 20 to 50 hours, over a wide maximum temperature range of 130 to 260 deg C in a variety of experiments.
Larch (Larix gmelinii) and red oak (Quecus rubra) were heated in a tub furnace in a nitrogen environment, with the ambient temperature of wood samples set at 200, 250, 300, and 400 deg C. Wood’s water absorption (WU) dropped as the TM temperature increased, and it reduced drastically over 300 deg C.
The pressure in the system during the modification phase is not mentioned in the publications; however, based on the maximum pressure listed in the equipment characteristics, it did not exceed two bar.
The effects of TM parameters on the mechanical characteristics, colour, chemical composition, and dimension stability of black poplar (Populous nigra L.) wood in a nitrogen environment have been widely researched in the temperature range of 160–220 deg C.
Mass loss (ML) occurs after TM in nitrogen; the chemical composition changes; the density, compressive strength, modulus of rupture, modulus of elasticity, and EMC decrease; while dimensional stability improves.
The characteristics of black poplar wood are more affected by an increase in TM temperature. There is no indication of pressure throughout the TM process.
However, the operating pressures of the TM chamber (0.25 cubic metres) are mentioned differently in the experimental sections (−1 to 1 atm and −1 to 5 atm).
It is unclear whether the TM is operating in a nitrogen flow or whether the system is filled with nitrogen and closed during the modification phase. It is mentioned that the TM chamber has forced air circulation.
We investigated the TM of silver birch (Betula pendula) and Scots pine (Pinus sylvestris) wood in this study, with a particular focus on enhancing the dimensional stability and minimising the hydrophilicity.
Birch and pine wood were chosen as the most frequent deciduous and coniferous wood species in the Republic of Latvia, as they offer a wide range of applications in the national economy.
The study’s goal was to examine the correlations between TM parameters in a closed nitrogen atmosphere under pressure and TM wood water and humidity uptake. The information available on the process parameters is insufficient.
As a result, a thorough analysis of this TM approach is an obvious next phase to confirm, supplement, or update the existing knowledge.
Thermal Modification Process
For each TM, silver birch (Betula pendula) and Scots pine (Pinus sylvestris) wood boards with dimensions of 1000 × 100 × 25 mm (longitudinal × tangential × radial) were used (20 for each treatment).
The boards used were of the finest grade, with no evident material faults (knots, grain slop, resin pocked, bark pocked, reaction wood, wanes, blue stains, decays, bug holes, shakes, distortions, and so on). All boards were conditioned in a standard climate (temperature 20 ± 2 deg C, relative humidity 65 ± 5%) prior to TM.
Sapwood and heartwood were not removed from the Scots pine wood, and randomly selected boards within specified density limitations were utilised for TM.
The wood moisture content was evaluated using ISO 13061-1:2014 and the density was determined using ISO 13061-2:2014. Before TM, the average densities of silver birch and Scots pine were 627± 14 kg × m−3 and 581 ± 10 kg × m−3, respectively.
TM was performed in a 340 L stainless steel pilot-scale chamber constructed by Wood Treatment Technology (Grinsted, Denmark). The jacket’s temperature was maintained stable with circulating hot mineral oil.
The modification chamber is designed to operate at pressures ranging from 0.1 bar to 20 bar, with temperatures reaching 190 deg C.
A program controls the equipment automatically, allowing it to switch between manual and automatic modes during the TM process. Prior to heating, the samples were stored in the autoclave for 30 min under a 0.2 bar vacuum to remove oxygen.
Nitrogen was fed into the autoclave from a nitrogen gas cylinder after the vacuum step to create the required initial pressure (3–6 bar). At the start, a small amount of water (1–1.5 L) was pumped into the autoclave to generate a small amount of steam to catalyse the hydrolysis of hemicelluloses.
Until the pressure-release stage, the TM system remained sealed and static (no mixing was used). The TM parameters for silver birch. Tmax, time at Tmax, and initial nitrogen pressure were varied from 160 to 170 deg C, 30 to 120 min, and three to six bar, respectively.
The TM parameters for Scots pine. Tmax, time at Tmax, and initial nitrogen pressure were varied from 160 to 180 deg C, 30 to 180 min, and four to six bar, respectively.
From ambient temperature to 100 deg C, the heating rate of the modification chamber was 0.36–0.42 deg C/min, and from 100 deg C to Tmax, it was 0.24–0.32 deg C/min.
The temperature was maintained after heating (30–180 min). The pressure in the autoclave grew during the TM process, reaching a maximum in the Tmax stage.
The chamber was then cooled by pumping out a mixture of nitrogen and wood thermal destruction products until atmospheric pressure was reached. The modification chamber door was partially opened to allow for additional cooling. The modification chamber was cooled at a rate of 0.30–0.35deg C/min.
The overall time for the TM process, including heating, holding temperature, and cooling, ranged from 16 to 19 hours, depending on the maximum TM temperature.
The mass loss (ML) was calculated by weighing wood boards (20 for each treatment) before and after the TM in a nitrogen atmosphere. The ML was calculated for each treatment as a percentage of the original mass of the fully dry wood.
The anti-swelling efficiency (ASE) of 15 specimens with dimensions of 20 × 20 × 20 mm3 (L × T × R) was examined under cyclic conditions. Before each cycle, specimens were conditioned in a normal climate (temperature 20 ± 2 deg C, relative humidity 65 ± 5%).
Five full-cell saturation and oven-drying cycles were performed. The ASE was calculated using the volumetric swelling (VS) coefficients of the treated specimens (St) in comparison to the untreated controls (Su).
The soaking phase consisted of vacuum water impregnation (0.2 bar, 120 min) followed by 24 hours of water storage at 22 deg C. The drying phase included a 72h step at 20 deg C, followed by eight hours at 45 deg C, eight hours at 60 deg C, eight hours at 80 deg C, and lastly, eight hours at 103 deg C.
The ASE estimate was based on the average measurements of radial, tangential, and axial swelling. The ASE was determined for each treatment.
Moisture uptake from the environment with constant humidity was measured for 15 TM wood specimens (dimensions 20 × 20 × 20 mm3) at relative humidity (RH) of 65, 75, and 98% at 20 ± 2 deg C. After being completely dried, the wood specimens were placed in a conditioning room with an RH of 65%.
Wood conditioning at higher RH conditions was performed in desiccators by placing specimens over a saturated NaCl salt solution (RH of 75%) and CuSO4 × 5H2O salt solution (RH of 98%).
Specimens were stored at successively increasing RHs of 65, 75, and 98%. When the mass of the wood samples remained constant across three weightings at 48-hour intervals (up to 30 days), the equilibrium moisture content (EMC) and dimensional changes were determined.
Mass Loss & Physical Changes
ML is used as a criterion to characterise the TM process’s intensity and the degree of wood degradation. ML is typically created due to changes in the chemical structure of wood.
The degradation of hemicelluloses, which are extremely vulnerable to thermal destruction, is responsible for the majority of the ML. A higher ML after TM often correlates with better biological durability and dimensional stability of wood. However, it has a disadvantage in terms of mechanical strength. The ML after TM for birch wood ranged from 5.9 to 12 percent.
These results are related to the initial birch wood densities of 627 ± 14 kg × m−3. The maximum average ML (10.1%–12.0%) was obtained for birch wood after TM at 160 deg C for 120 min and 170 deg C.
The lowest average ML was obtained after TM at 160 deg C for 60–90 min. When all other process parameters were held constant, a rise in initial pressure resulted in a higher ML after TM. It should be emphasised, however, that the ML values produced using TM had large error margins.
The birch wood board dimensions were substantially reduced after TM in both the radial and tangential axes. The tangential direction of TM birch wood showed a greater reduction (4.4–6.4%) than the radial (2.9–4.7%).
The most significant changes in both the tangential and radial directions of birch wood were observed after TM at 170 deg C and a time at Tmax of 120 min, while the least change occurred after TM at 160 deg C and shorter treatment times.
This could be explained by the fact that the wood structure varies in different directions, which affects the swelling capability of the wood cell wall. After TM, radial and tangential direction reduction resulted in total birch board volumetric changes ranging from 7.1 to 10.2 percent.
The ML after TM for pine wood was 2.6–9 percent. These results are related to the initial pine wood densities of 581 ± 10 kg × m−3. After TM, pine wood had a lower ML (2.6%–9.0%) than birch wood (5.9%–12.0%). The maximum ML of pine wood (7.6%–9.0%) was obtained at 180 deg C and 170 deg C with a duration at Tmax of 120 min. For all exposition times at Tmax, the lowest ML was obtained at 160 deg C.
An increase in the initial pressure from four to six bar at 170 deg C had no noticeable effect on ML after TM. In comparison to the radial direction (1.5%–2.5%), TM pine wood also showed a larger reduction in the tangential direction (2.9%–4.5%).
Following TM, the pine wood showed the greatest tangential and radial direction changes at 170 deg C with an initial pressure of four bar and at 180 deg C. All TM at 160 deg C resulted in the lowest dimensional changes. After TM, the volumetric changes in pine wood were 4.4–6.9 percent.
After 1.5–3.5 hours of TM at 185 deg C in a nitrogen flow at 10 bar pressure, the ML of beech wood was 11–17 percent and eight percent for pine wood.
After TM in an oven at 220 deg C for 20 hours under a nitrogen environment, the ML of short- and long-rotation teak (Tectona grandis L.f) was 13.4 and 9.2 percent, respectively.
Mimosa scabrella and Pinus oocarpa wood samples were TM in a nitrogen atmosphere at 180, 200, and 220 deg C for one hour. After TM, the ML for Mimosa scarabella wood was 4.3, 7.0, and 8.9 percent, and for Pinus oocarpa it was 4.5, 4.9, and 6.3 percent, respectively.
After being exposed to nitrogen for two ,four, six and eight hours, black poplar (Populus nigra L.) TM yielded 0.9 percent ML at 160 deg C, 0.8–1.1 percent at 170 deg C, 0.8–3.0 percent at 180 deg C, 0.8–3.5 percent at 190 deg C, 4.0–7.1 percent at 200 deg C, and 6.8–14.0 percent at 220 deg C. Temperature had a higher impact on ML than treatment time.
Our findings reveal that TM in a closed system under pressure in a nitrogen flow induces a similar ML at lower Tmax (170–180 deg C) as TM in an open system in a nitrogen flow at Tmax above 200 deg C.
Anti-Swelling Efficiency & Cell Wall Water Capacity
ASE is one of the most important indicators of modified wood materials. It demonstrates the material’s resistance to dimensional change under simulated, cyclic conditions of complete soaking and drying.
Almost all TM wood is used to make cladding and decking. Dimensional stability is critical in such applications due to the influence of cyclic weather conditions (moistening and drying, heating and cooling, etc.).
As a result, many cycling tests were carried out in which wood was subjected to several soaking and drying cycles.
Birch wood after TM had ASE values ranging from 22 to 69 percent. ASE decreased after the second and third soaking–drying cycles, but did not change significantly after the fourth and fifth cycles.
After the first cycle, ASE was 40–69 percent, while after the fifth cycle, it was down to 22–63 percent. This decrease is caused by the leaching of thermal destruction products from TM wood, which makes its structure more accessible to water. The sample ML after each cycle also supports this statement (data not shown).
The soaking water was light brown after each cycle, suggesting the absence of degradation products. The primary ML can be attributed to the degradation of hemicellulose and extractives, which are very vulnerable to thermal destruction.
Birch wood had the lowest ASE values (22–23% after the fifth cycle) after TM. Increasing the initial pressure from 3 to 4 bar at 160 deg C raised ASE from 24 to 33 percent.
Simultaneously, changing the initial pressure from four to five bar significantly reduced the ASE values after the fifth cycle from 42 to 26 percent.
At the maximum TM time of 120 min, an increase in initial pressure from four to six bar reduced ASE values after the fifth cycle from 45 to 41 percent. At 170 deg C, raising the initial pressure from three to four bar enhanced ASE from 38 to 47 percent.
However, initial pressure increase from four to six bar reduced the ASE values from 47 to 23 percent after the fifth cycle.
TM birch wood’s cell wall total water capacity (CWTWC) was reduced to 10–29 percent compared to untreated birch wood (32–35%). The CWTWC for all specimens increased slightly during five soaking–drying cycles. This can be explained by the leaching of natural extractives from untreated wood and thermal destruction products from the TM wood.
The majority of the extractives were soaked out within the first, second, and third cycles, resulting in minor CWTWC variations after the fourth cycle. After TM at 160–180 deg C in saturated steam in a closed system under pressure, a similar trend was found with aspen (Populus tremula), birch (Betula pendula), and grey alder (Alnus incana) wood.
Birch wood had the lowest CWTWC (10–14%) after TM, which had the highest ASE (63%). The CWTWC for the other treatments ranged from 13 to 20 percent after the first cycle to 21–29 percent after the fifth cycle.
The maximum CWTWC was observed after TM, having the lowest ASE (22% after the fifth cycle). Our findings revealed a clear link between ASE and CWTWC alterations after TM.
Equilibrium Moisture Content & Dimensional Changes
EMC is an indicator of a material’s theoretical ability to absorb moisture from the air. Prolonged EMC exceeding 20 percent is regarded as the threshold above which a favourable environment for the development of microorganisms (rot fungus, blue stain, mould) in wood products has been produced. All wood modification techniques, in general, attempt to limit water and moisture absorption.
When compared to untreated wood, TM birch wood’s EMC was 51–64 percent lower and VS was 48–67 percent lower, with substantial reductions in all RHs.
The EMC of untreated birch wood at 65, 75, and 98 percent RH was 9.6, 12.9, and 17.4 percent, respectively. At 65, 75, and 98 percent RH, the EMC of TM birch wood dropped to 3.7–4.7 percent, 4.6-5.6 percent, and 6.6–7.9 percent, respectively.
TM wood absorbs much less moisture in high-humidity conditions (RH of 98%). As a result, TM wood has a substantially lower theoretical probability of infection and microbial growth than untreated wood.
The VS of untreated birch wood at 65, 75, and 98 percent RH was 4.8, 6.5, and 9.6 percent, respectively. TM birch wood’s VS at 65, 75, and 98 percent RH was reduced to 1.6–2.5 percent, 2.1–3.1 percent, and 3.3–4.6 percent, respectively.
The lowest EMC and VS values at all RH levels were obtained after TM. Differences in EMC and VS between the other TM regimes were not statistically significant, because the error limits overlapped.
After TM in a nitrogen atmosphere at 180, 200, and 220 deg C for one hour, the EMC (65% RH) of Mimosa scabrella and Pinus oocarpa wood was 12.0, 11.6, and 11.3 percent and 12.7, 12.9, and 11.4 percent, respectively.
The EMC values for untreated wood were 13.5 and 14.7 percent, respectively. The EMC reduction was only 1.5–3.3 percent. The EMC (98% of RH) of black poplar (Populus nigra L.) was reduced from 27 percent for untreated wood to 26, 23, and 16 percent for wood TM in nitrogen at 160, 190 and 220 deg C, respectively.
The EMC of beech wood (Fagus sylvatica) after TM in nitrogen at 175 deg C for two hours and 185 deg C for three hours at 10 bar pressure was five to seven percent (65% of RH) and 13–23 percent (100% of RH), respectively, whereas the untreated wood’s EMC was 10 percent and 30 percent, respectively. TM beech wood had the lowest EMC when heated to 185 deg C compared to other specimens.
After TM at 190 deg C for 2.5 hours in a nitrogen environment at a constant 10 bar pressure, the EMC of beech wood was 13 percent (90% RH), while that of untreated wood was 26 percent.
Swelling in the tangential, radial, and axial directions of TM beech was 4.3, 2.1, and 0.15 percent, respectively, whereas in untreated beech it was 12, 6, and 0.3 percent, respectively. This is consistent with our findings that EMC and swelling of TM wood in nitrogen under pressure can be reduced more than twice compared to untreated wood.
Conclusions
The results clearly revealed that TM in nitrogen in closed, pressurised process can improve the dimensional stability of birch and pine wood, as well as reduce moisture and water absorption. Birch wood was less resistant to TM because its ML (5.9%–12%) was higher than that of pine wood (2.6%–9%).
Our findings reveal that TM in a closed system under pressure in nitrogen induces similar ML at 170–180 deg C to TM in an open system in a nitrogen flow at Tmax above 200 deg C.
TM caused a shrinkage in the tangential, radial, and total volume of birch and pine wood. Birch wood after TM had ASE values ranging from 22 to 69 percent, while pine wood’s ASE was 27 to 58 percent.
The ASE values for both wood species tended to decrease within five soaking–drying cycles. Because of the leaching of thermal degradation products from TM wood, the CWTWC increased and the wood structure became more susceptible to water.
According to the ASE data, birch wood TM in nitrogen with an initial pressure of four bar appeared to be more suitable for achieving improved dimensional stability than three, five, or six bar.
For pine wood, the difference between four and six bar initial nitrogen pressure at 170 deg C was minimal. As a consequence, the optimal option for TM at 160 and 180 deg C was determined to be five bar initial pressure.
When compared to untreated wood, TM birch wood’s EMC was 51–64 percent lower and its VS was 48–67 percent lower, but TM pine wood’s EMC was 29–56 percent and its VS was 29–61 percent lower.
EMC and VS were greatly reduced at all RHs (65, 75, and 98%). When compared to the data available in the literature, the EMC values for birch and pine wood were significantly lower, suggesting the efficiency of the TM process.
Increases in TM process parameters (Tmax, time at Tmax, and initial pressure) resulted in decreased EMC for pine wood.
