Heat treatment at 160-220 C cuts the equilibrium moisture of Chinese fir by up to about 40%: dynamic vapour sorption results

Introduction

Thermal modification, in which sawn wood is heated to roughly 160 to 220 degrees Celsius with limited oxygen, is one of the best-established ways to improve the dimensional stability of softwood boards for flooring, cladding and furniture. Much of the benefit is attributed to lower hygroscopicity, meaning the wood takes up less moisture from the surrounding air. Yet how exactly the sorption behaviour changes across the whole humidity range, and how it differs between taking up and releasing moisture, is less often described in detail.

A paper in the journal Forests, published in March 2024 by four authors from the Pan Tianshou College of Architecture, Art and Design at Ningbo University and the Research Institute of Wood Industry of the Chinese Academy of Forestry in Beijing, examines this for Chinese fir (Cunninghamia lanceolata), a major plantation softwood in China. The authors assess how heat treatment between 160 and 220 degrees Celsius changes the equilibrium moisture content (EMC), the adsorption and desorption behaviour and the sorption hysteresis. They also fit a standard sorption model (GAB) and estimate the internal specific surface area available for water. The stated purpose is to support process optimisation for products with improved dimensional stability.

For a sawmill or a drying and treatment plant, the interest lies in the size and temperature-dependence of the effect and in what the study does not cover. It is a small laboratory study on one species, with a single treatment duration. The figures below are those reported by the authors; the portal has not repeated the measurements.

Research methodology

Specimens. The authors used defect-free specimens of 150 x 15 x 15 mm (longitudinal, tangential and radial) cut from the same growth rings of air-dried heartwood of Chinese fir. Only pure earlywood was used. Five groups were prepared: an untreated control and four heat-treated groups named HT160, HT180, HT200 and HT220.

Heat treatment. Treatment took place in airtight equipment with an oxygen content of 2 percent or less, at 160, 180, 200 and 220 degrees Celsius for 2 hours each. After treatment the specimens cooled to room temperature inside the equipment and were stored over phosphorus pentoxide. The paper refers to an earlier publication from the same group for details of the procedure; those details were not reproduced in the text read by the portal.

Sorption measurement. Moisture sorption was measured by dynamic vapour sorption (DVS) with a DVS Intrinsic instrument (Surface Measurement Systems, London) with a balance accuracy of 0.1 microgram. The initial sample mass was about 25 mg, the temperature 25 degrees Celsius, and relative humidity (RH) was controlled with nitrogen and water vapour. The mass was recorded every 5 seconds, and each humidity step ended when the mass changed by less than 0.002 percent within 10 minutes. Humidity was raised in steps of 10 percent from 0 to 90 percent RH and in steps of 5 percent from 90 to 95 percent, and then lowered again to give the desorption branch.

Hysteresis and model. The hysteresis ratio (HR) was defined as the difference between desorption and adsorption EMC at the same humidity, divided by the adsorption EMC. The GAB (Guggenheim-Anderson-de Boer) model was fitted over 0 to 95 percent RH, giving the monolayer moisture content (Mm) and two constants, C and K. The internal specific surface area (S_GAB) was calculated from Mm using a water-molecule area of 0.114 square nanometres.

Statistics. One-way ANOVA and Duncan's multiple-comparison test at the 5 percent level were applied at 30, 60 and 95 percent RH. The degrees of freedom in the ANOVA table imply 15 specimens in five groups, probably three per group, but the paper does not state the number of replicates explicitly. No named test standard (ISO or ASTM) is cited for the sorption measurements.

Key findings

Equilibrium moisture content falls with treatment temperature. The EMC values given in Table 2 of the paper were:

By the portal's arithmetic from this table, treatment at 220 degrees Celsius lowered EMC relative to the control by about 37 percent at 30 percent RH, 39 percent at 60 percent RH and 40 percent at 95 percent RH; at 160 degrees the reductions were about 26 percent, 19 percent and 14 percent. The paper's own text gives reduction percentages that the portal did not fully check against its calculation (for 30 percent RH the portal's figure for HT220 is lower than a figure of about 40 percent that appeared in the text extraction), so readers needing exact percentages should recalculate from the table.

Significance. According to the ANOVA results, F-values were 42.531 at 30 percent RH, 211.933 at 60 percent RH and 572.804 at 95 percent RH, all significant at the 5 percent level. At 30 percent RH the HT160 and HT180 groups were not significantly different from the control, whereas HT200 and HT220 were; at 60 and 95 percent RH every treated group differed significantly from the control. The effect of heat treatment is therefore clearest at higher humidity and at higher temperatures.

Isotherm shape. All adsorption isotherms were of Type II in the IUPAC classification. The desorption curves of heat-treated wood, especially HT220, were more linear than the sigmoidal curve of the control. No obvious capillary condensation was seen above 90 percent RH under the test conditions. According to the authors, monolayer sorption dominates at low humidity and multilayer sorption as humidity rises, and the control showed larger EMC increments, particularly near 10 percent RH.

Hysteresis. The hysteresis ratio increased with heat-treatment temperature, which the authors link to changes in the cell-wall chemistry and microstructure. The text gives no numeric hysteresis values; they are shown only in a figure that could not be read.

GAB parameters and surface area. The fitted GAB model described the data well (R-squared between 0.997 and 0.999). The monolayer moisture content fell from 5.73 percent in the control to 4.47, 3.96, 3.58 and 3.51 percent for HT160 to HT220. The corresponding internal specific surface areas were 219, 171, 152, 137 and 134 square metres per gram. Expressed against the control, the treated samples retained about 78, 69, 63 and 61 percent of the surface area. The portal recalculated these values from the reported Mm and found them consistent with the table. The authors note that the trend in surface area parallels the trend in EMC. The text of the paper describes the K and C ranges the opposite way round to Table 3; the table values (K about 0.76 to 0.80, C about 6.5 to 7.9) are the ones used here.

Interpretation by the authors. The lower change in EMC with humidity in heat-treated wood is presented as a likely reason for its better dimensional stability. The explanation for the reduced hygroscopicity is discussed on the basis of literature (changes in hemicellulose, lignin and crystallinity), not measured in this study.

Practical applications for sawmills

Portal interpretation (these are not claims made by the paper):

• Temperature is the main lever, with diminishing returns at the top. In this data set, EMC at 60 percent RH fell from 9.15 percent to 7.44, 6.74, 6.10 and 5.57 percent over the four temperature steps, and the monolayer moisture content and surface area changed less between 200 and 220 degrees than between the control and 160 degrees. A producer choosing a treatment temperature will be balancing the moisture benefit against other property changes that this paper did not measure, such as strength and colour.
• Specifying treated product by moisture behaviour. The tabulated EMC values give a rough indication of the equilibrium moisture a customer could expect for treated Chinese fir indoors (30 to 60 percent RH) compared with untreated wood. The data come from small earlywood samples at 25 degrees Celsius, so they should not be transferred directly to full-size boards or other species.
• Moisture-content targets for finished product. Because treated wood equilibrates at a lower moisture content, a mill that treats and then dispatches product may need to adjust the moisture content it quotes or the conditioning it applies. The paper does not discuss this.
• Hysteresis and conditioning. The increase in sorption hysteresis with treatment temperature suggests that treated wood may release moisture differently from the way it takes it up. Whether this matters for kiln schedules or for conditioning of treated boards is not examined in the paper.
• A research reference, not a process specification. One treatment duration, one species and one measurement temperature mean the results can inform expectations and discussion with a laboratory, but not a production recipe.

Limitations

The authors state that the mechanisms behind the reduced hygroscopicity still need to be explored, and that accessible hydroxyl content alone may not explain the change, citing contradictory literature. Further limitations, stated or evident from the paper:

• Only one treatment time (2 hours) was used, so the combined effect of temperature and duration is not shown.
• Only Chinese fir, and only earlywood, was tested, and sorption was measured at 25 degrees Celsius only.
• The humidity step changed from 10 to 5 percent above 90 percent RH, which the authors note affects the increments and decrements in the EMC change figure.
• The analysis of EMC increments and decrements is qualitative and relies on figures.
• Chemical and structural changes (hemicellulose, lignin, crystallinity) were not measured in this study; they are discussed from literature.
• The number of replicates is not stated explicitly, and the small implied sample (about three per group) limits the strength of the statistics.
• The paper contains some inconsistencies between text and tables (the roles of K and C, and percentage reductions), which reduce clarity.
• Figures 1 to 3 (adsorption and desorption isotherms, EMC change by humidity step, hysteresis ratio by sample) are graphics that could not be read in the text version used for this summary, so statements depending on them are taken from the authors' text and not independently checked. Tables 1 to 3 were readable.

Conclusions

For Chinese fir earlywood treated at 160 to 220 degrees Celsius for 2 hours in a low-oxygen atmosphere, equilibrium moisture content at 30, 60 and 95 percent RH decreased with rising treatment temperature, with the strongest effect at 200 to 220 degrees under high humidity. The sorption hysteresis increased with temperature, and the estimated internal specific surface area fell from 219 to 134 square metres per gram. The authors conclude that the smaller change in moisture content with humidity is a likely reason for the better dimensional stability of heat-treated wood. The paper provides a clear laboratory description of how the sorption behaviour shifts with temperature but, by its own account, leaves the underlying chemical mechanisms open.

Original research

• Authors: Yulei Gao, Zhu Li, Liyuan Zhao, Jianxiong Lyu
• Title: Effect of Heat Treatment on Hygroscopicity of Chinese Fir (Cunninghamia lanceolata [Lamb.] Hook.) Wood
• Journal: Forests, volume 15, article 630
• Year: 2024 (published 29 March 2024)
• DOI: 10.3390/f15040630
• Affiliations: Pan Tianshou College of Architecture, Art and Design, Ningbo University, Ningbo, China; Research Institute of Wood Industry, Chinese Academy of Forestry, Beijing, China
• Licence: open access, CC BY 4.0, as stated in the article PDF

This article is an independent summary prepared by the portal; the original paper should be consulted for full methods, tables and figures.