Search

International audience; The potential of muirapiranga (Brosimun sp.) as a substitute material for violin bows was estimated in terms of vibrational properties, and the infl uence of extractives on the vibrational properties was examined. The loss tangent of muirapiranga was somewhat higher and the specifi c dynamic Young's modulus was rather lower than the respective values for pernambuco, which is regarded as the best material for violin bows. Therefore, muirapiranga is a poorer bow material in terms of vibrational properties. Impregnation of crude extractives from muirapiranga decreased the loss tangent of other wood specimens. The main compounds of the extractives were identifi ed as xanthyletin and luvangetin. Impregnations of isolated xanthyletin and commercially available methoxsalen, which was tested as an analogue of luvangetin, markedly decreased the loss tangent of other wood specimens. Methoxsalen and xanthyletin differ from conventional loss tangent-decreasing substances, namely protosappanin B and hematoxylin, in terms of water insolubility and the absence of hydroxyl groups. From the similarity in molecular characteristics of loss tangent-decreasing substances found so far, restriction of molecular motion due to an impregnated substance in the wood matrix is suggested as one loss tangent-decreasing mechanism.

Ezomatsu wood blocks were impregnated with potassium acetate (KAc) and then exposed to acetic anhydride vapor at 25°C and 120°C. The KAc-impregnated wood was rapidly acetylated at 120°C, and only 6 min was needed to achieve 20% weight percent gain (WPG). The WPG increased with increasing catalyst loading (CL), but it turned to decrease above 20% CL probably because the diffusion of acetic anhydride vapor was hindered by excess KAc depositing in the cell lumina. Thus, careful control of CL is necessary in the vapor-phase acetylation. KAc was also effective in catalyzing the vapor-phase acetylation at 25°C: the KAc-impregnated wood attained 20% WPG within 7 days, whereas the WPG did not exceed 10% even after 1 month in the uncatalyzed system. Irrespective of treatment methods, the hygroscopicity of wood was reduced and its dimensional stability was improved with an increase of WPG. These results confirm that the use of KAc simplifies the acetylation process at room temperature with minimal loss of acetic anhydride.

Ezomatsu wood blocks were acetylated in a mixture of acetic anhydride and acetic acid containing excess potassium acetate (KAc). The mixture method enabled rapid acetylation at 120°C: a 20% weight gain (weight percent gain; WPG) was achieved within 30 min while the WPG did not exceed 18% after 120 min of conventional uncatalyzed acetylation. At 40°C, however, a satisfactory WPG was not achieved with the mixture method because both the wood swelling and KAc concentration in the reagent solution were limited at that temperature. In addition, the antiswelling efficiency attained by the mixture method was irregularly low, probably because of nonuniform reaction involving shrinkage of the cell lumina. These results suggest that the mixture method is not advantageous for low-temperature acetylation, whereas it enables simple and rapid acetylation at high temperature.

To analyze the effects of lignin on the destabilization of wood due to quenching, we examined the dielectric properties of untreated and delignified wood before and after quenching at 20°C from 50 Hz to 100 MHz. For untreated wood, the inflection points of log e′ and log σ vs log f and the peak of log(tan δ) vs log f were attributed to interfacial polarization before quenching, and the location of the inflection point shifted to a higher frequency with increasing moisture content because of changes in the water cluster. After quenching, the inflection points of log e′ and log σ and the peak of log(tan δ ) shifted to higher frequency; however, the values of log e′, log σ recovered to those before quenching with the passage of time. For delignified wood, dielectric relaxation was observed at a higher frequency than for untreated wood irrespective of quenching. It was inferred that the mobility of water molecules was influenced by the cluster surroundings because of increased number of adsorption sites in hemicellulose. Moreover, after quenching, the recovery process did not change greatly over time; it was shown that the matrix structure was affected more by quenching with the loss of lignin.

To clarify the role of lignin in the affinities of acetylated wood for organic solvents, the effects of delignification on the solvent adsorption of acetylated wood were investigated. Acetylated wood meals rapidly adsorbed organic solvents that were hardly adsorbed by unmodified wood. For nonpolar and low-polarity organic solvents, a clear positive correlation was observed between the amount of adsorption and the lignin content. This indicated that acetylated lignin was responsible for the excellent affinities of acetylated wood for hydrophobic organic solvents. On the other hand, for lower alcohols and water, the amount of adsorption reduced with an increase in the lignin content. It was suggested that the adsorption of such polar solvents was dominated by insufficiently acetylated hydrophilic polysaccharides.

In order to investigate the effects of chemical components and matrix structure on the destabilization of quenched wood, we examined the physical and mechanical properties of steam-treated wood, hemicellulose-extracted wood, and delignified wood, which were treated at different levels. For steam-treated and hemicellulose-extracted wood,the relative relaxation modulus of the quenched sample was lower than that of the respective control sample. For delignified wood, the relative relaxation modulus fell with weight loss and reached a minimum value at a certain weight loss, and subsequently increased significantly. The hygroscopicity of all treated samples changed slightly by steaming, whereas increased with removing the component. More-over, the average volumetric swelling per 1% MC at 100% relative humidity (RH) was less than at 75% RH and 93% RH for component-removed wood. It was clear that a void structure existed. As a result, the destabilization evaluated by the fluidity (1 - Et/E0) of steam-treated wood was influenced by the amount of adsorbed water. For component-removed wood, destabilization increased temporarily at lower weight loss because of nonuniform cohesive structure. At high weight loss, destabilization will decreased because capillary-condensed water gathered in the voids and obstructed the motion of adsorbed water. However, the destabilization of all treated wood changed less than that of chemically modified wood.

In order to examine the contribution of wood components to the acetylation of wood, we acetylated wood meal that had been partially delignified. The results were analyzed in terms of the reaction kinetics. The first-order rate equation was successfully adjusted to the weight gain data. The rate constant for acetylation initially increased with progress of lignin elimination and then turned to decrease; the apparent activation energy showed the reverse tendency and ranged from about 90 to 130 kJ/mol. These results suggest that lignin elimination brings not only separation of lignin but also drastic change of the chemical and/or physical structure in the residual lignin, and this affects the reactivity of wood meal as a whole. The ultimate weight gain estimated by the regression of the rate equation showed a minimum when lignin was moderately eliminated, which was explained in terms of enhanced reactivity of lignin and lower accessibility for holocellulose than predicted. The equilibrium moisture content had a maximum when lignin was moderately eliminated. This tendency is the opposite of that observed for the ultimate weight gain, and suggests that the sites for acetylation do not always correspond to those for moisture adsorption.

In order to understand the mechanism of destabilization occurring when wood is quenched, we applied chemical modifications, and controlled the number of moisture adsorption sites in wood. The degree of destabilization was evaluated according to the fluidity (1-E t/E 0), increase in fluidity, and relative fluidity in relation to the nonmodified wood, and was discussed by comparing these quantities with the hygroscopicity or swelling of wood. We found that destabilization of chemically modified wood was lower than that in nonmodified wood, and the amount of adsorbed water controlled the magnitude of flow of wood. Moreover, according to the analysis of water state by the Hailwood-Horrobin equation, it was shown that the function of dissolved water to the fluidity is almost identical for both chemical modifications, whereas hydrated water has more effect on acetylated wood than on formaldehydetreated wood. We speculate that the motion of water molecules due to quenching accompanied with the redistribution of energy resulting from the exchange of their potential energy and movement to attain a new balance, and the introduced acetyl groups and cross-linking restrict the water molecule movement.

The amount of boron contained in the xylem of Japanese persimmon (Diospyros kaki Thunb.) and ebony (Diospyros ebenum Koen) was determined by inductively coupled plasma (ICP) emission spectrometry and a modified curcumin—acetic acid method. The boron content was compared between the heartwood and sapwood of ebony as well as between a blackened portion (“kurogaki”) and normal portion of Japanese persimmon. The kurogaki contained a higher level of boron than the normal portion of the same individual, although the boron content varied among individuals. Moreover, the boron content of the heartwood of ebony was much higher than that of the sapwood. These results suggest the participation of boron in the blackening of Japanese persimmon. Because both kurogaki and heartwood of ebony are durable to fungal attack, the blackening of Diospyros genus appears to be related to the formation of defensive substances in which boron seems to take part. The convenient curcumin—acetic acid method is an alternative to the ICP method with comparable accuracy.

The vapor phase reaction of wood with maleic anhydride (MA) was investigated from the aspect of the mechanism of dimensional stabilization. Notably the existence of cross-links was examined by detailed analyses of dimensional stability and related properties, diffuse reflectance infrared Fourier transform (DRIFT) spectra, and changes in mechanical properties such as creep property and vibrational property. Higher reaction temperature resulted in less leaching of reagent. Also a peak in DRIFT spectra at 1730 cm−1 showed the esterification of wood components with MA, while that at around 1780 cm−1, which became remarkable with increasing reaction temperature, suggested the formation of cross-linking. The loss tangent decreased and the creep deformation was restrained for the specimens treated at high temperature. From these results it is plausible that MA mainly forms monoester with wood components at lower temperature; however, at elevated temperature cross-linking appears in addition to formation of the monoester.

Influences of moisture content on the catalysis of sulfur dioxide and physical properties of reaction products were examined for a vapor-phase treatment of spruce wood with formaldehyde. The reaction rate was strongly dependent on the amounts of water and sulfur dioxide in the reaction system, and this was consistent with a proposal that the hydroxymethylsulfonic acid (HOCH2SO3H) formed from sulfur dioxide, water, and formaldehyde acts as a catalyst for the reaction. However, not all water molecules contributed to the formation of HOCH2SO3H, because some were adsorbed by the wood components. When the initial moisture content was high, polymeric cross-linking was likely because the ultimate weight gain was much greater than the value estimated based on monomeric cross-links. The lower values of antiswelling efficiency and higher equilibrium moisture content at a specific level of weight gain also suggested the existence of polymeric cross-links. However, the difference in the length of cross-links did not significantly influence the mechanical properties.

Factors influencing the rate of acetylation were examined based on the swelling of wood in the reaction solution and the dimensions of the wood sample. The activation energy of acetylation was also estimated. In a swelling test, it was found that wood swells thoroughly in acetic anhydride even without pyridine above 60°C. Therefore, pyridine may facilitate the acetylation process as a catalyst and not as a swelling agent. The weight gain, x (%), attained at reaction time t (h), for various compositions of acetylation solution or dimensions of wood sample were analyzed by applying an original rate equation [x = a × (1 − e−kt )1/n ], where a is the ultimate weight gain (%), k is the rate constant (h−1), and n is a measure of the hindrance against the diffusion of reagent. The optimum volume fraction of pyridine in the pyridine-catalyzed acetylation was about 0.2. Accompanied by a rise in pyridine content, the reaction showed increased diffusion-controlled behavior. The rate constant, which is not affected by the dimensions of the wood sample, was estimated from which an activation energy of about 130 kJ/mol was calculated.

A modified rate equation, in which the diffusion of reagent was taken into consideration, was proposed and applied to results reported previously for the acetylation of wood. The equation was fitted more successfully to the prolonged as well as the early stage of the reaction than a common first-order rate equation. Although the conclusions were virtually unchanged, the reliability was enhanced because reaction parameters such as the rate constant, ultimate extent of the reaction, and activation energy calculated for the modified equation were based on much more data than those reported previously. The equation proposed here may be applicable not only to acetylation but to other chemical modifications of cellulosic materials in heterogeneous systems.

Mechanical property changes due to the moisture content (MC) and/or temperature changes were examined for 15 Indonesian wood species. A static bending test was carried out at 20°C, 65% relative humidity (air-dry), and water-saturated at 20°C (wet-20) and 80°C (wet-80). For individual test conditions, modulus of elasticity (MOE) and modulus of rupture (MOR) increased linearly with specific gravity regardless of wood species; however, maximum deflection did not correlate with specific gravity for any MC or temperature conditions. The relative values of MOE and MOR measured in wet-20 to air-dry conditions were variously affected from slightly to strongly depending on the wood species. However, the relative values always decreased markedly when saturated in water at 80°C, regardless of wood species. The relative MOE, MOR, and maximum deflection values due to the change in MC or MC and temperature combined were independent of specific gravity but may be dependent on wood type: softwood or hardwood.

Acetylation of wood was carried out in acetic anhydride only, acetic anhydride/xylene 1 : 1 (v/v), and acetic anhydride/pyridine 4 : 1 (v/v) solutions. The antishrink efficiency (ASE), hygroscopic properties, vibrational properties, and bending strength were compared among the three reaction solutions. The ASE was a simple function of weight gain (WG); the equilibrium moisture content at a given WG differed among the reaction solutions. Based on this fact and the results of repeated water soaking and oven-drying tests, it was found that the bulking effect was a major factor, and that decreased hygroscopicity contributes only slightly to the dimensional stabilization by acetylation. The difference in equilibrium moisture content among reaction solutions appears more significant in block samples than wood meal, probably due to the fiber-to-fiber bonds in the former. The tendencies for change in the specific Young’s modulus and the loss tangent differed among reaction solutions, whereas in the static bending test the difference was not marked.

The dependence of the reaction parameters of acetylation on the reaction mixture was compared among uncatalyzed, acetic anhydride-xylene mixed, and acetic anhydride-pyridine mixed solutions. Wood meal and blocks were used to examine the effect of sample size. A first-order rate equation was applied to the data, and a rate constant and leveling off value of weight gain (WG) were estimated. The rate-determining step was examined from the viewpoint of activation energy. The results were as follows: (1) Regarding the magnitude of the rate constant, the order was pyridine system > uncatalyzed system > xylene system. (2) The ultimate value of WG was lower in the uncatalyzed and xylene systems than the pyridine system, probably because of the swelling ability of pyridine. (3) The activation energies of acetylation estimated for wood meal were 120, 135, and 110 kJ/mol for the uncatalyzed, xylene, and pyridine systems, respectively. (4) The characteristics of the diffusion-controlled reaction became marked when the acetylation was carried out in the pyridine system, at elevated temperature, and for wood blocks. Under these conditions, the supply of reagent to the reaction site might not be sufficient to fuel the reaction.

The heartwood of the Japanese persimmon tree (Diospyros kaki) becomes black on rare occasions and has been highly esteemed as a substitute for ebony. We attempted to clarify how the physical, mechanical, chemical, and biodegradation properties differ between sapwood and blackened heartwood. The specific gravity, equilibrium moisture content, modulus of rupture, and modulus of elasticity in the blackened heartwood were higher, and the loss tangent was lower, than those in sapwood. Furthermore, the blackened portion was more resistant to fungal and termite attacks. A section of heartwood was dark-brown, and the specific gravity and mechanical properties of this portion were slightly lower than those in sapwood. The dark-brown portion was speculated to be a sign or interrupted state of fungal attack. The blackening substance was bound tightly to cell wall components and could not be extracted with any of the organic solvents used. The findings of trace element analysis using inductively coupled plasma-mass spectrometry showed that the boron content was markedly high in the blackened portion. The findings obtained here suggest a role of boron in the antifungal properties and the blackening phenomenon of Japanese persimmon.

Summary To clarify tangential and radial dimensional changes in hinoki (Chamaecyparis obtusa) wood by ten kinds of treatments and successive water soaking, a vector expression on the tangential-radial coordinates was proposed and a new measure (ASE′) to evaluate the dimensional stability of wood was defined. The ASE′ is expressed by 100(1 − Vs′ /Vu ) where Vu is the volumetric swelling coefficient of untreated wood and Vs′ is that of treated wood calculated on the basis of its oven-dried volume before treatment. The results from formaldehyde treated, heat treated and steam treated wood showed large positive ASE′ values and the ratio of radial swelling to tangential swelling (swelling anisotropy) of these treated woods decreased with increasing weight percent gain or weight loss. The results from acetylated wood, wood impregnated with polyethylene glycol and wood treated with a low molecular weight phenol-formaldehyde resin showed either small positive or negative ASE′ values, and those of wood treated with propylene oxide showed large negative values. The swelling anisotropy of treated woods accompanied by cell wall bulking was larger than that of untreated wood. The swelling anisotropy decreased when the cell lumen was filled with a hydrophobic resin (WPC-MMA), and increased when the cell lumen was filled with a hydrophilic resin (WPC-HEMA).

Cellulose isolated from wood is more susceptible than cotton cellulose to homogeneous hydrolysis in phosphoric acid. The influence of carboxyl group introduction at the C6 position on the hydrolysis rate of cellulose in 82.5% phosphoric acid was studied as a model of the oxidation of cellulose during pulping. The rate constant of hydrolysis for dissolving pulp was larger than that of cotton cellulose at temperatures of 25°–35°C. Mercerized cotton cellulose was partially oxidized regioselectively at the C6 position by a free radical system using 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO). The oxidized cellulose was obtained at a range of 1.7–12.7 mEq carboxyl content per 100 g of cellulose. The hydrolysis rate of the oxidized cellulose samples accelerated with increasing carboxyl content in the samples.

Heartwood of pernambuco (Guilandina echinata Spreng, synCaesalpinia echinata Lam.), which has been used as material for violin bows, was extracted by soaking in water, and the obtained extractives were analyzed. The main components of the extractives were identified to be protosappanin B and brazilin. In particular, protosappanin B occupied about 40% of the pernambuco extractives. The loss tangent (tanδ) of spruce wood impregnated with protosappanin B decreased markedly, the same as that of specimens impregnated with extractives before being purified. It is expected that protosappanin B can make a contribution to the decrease in tanδ due to impregnation with extractives.

To clarify the lowering mechanism of loss tangen (tanδ) of sitka spruce (Picea sitchensis Carr.) wood impregnated with extractives of pernambuco (Guilandina echinata Spreng. synCaesalpinia echinata Lam.), we examined the vibrational properties of the impregnated wood in relation to the adsorbed water. The results obtained were as follows: (1) The equilibrium moisture content (EMC) of impregnated sitka spruce decreased to some extent compared with untreated wood. (2) Frequency dependencies of tanδ a about 400–8000Hz showed that impregnated wood has much lower tan δ than untreated wood at around 9% mois ture content (MC), except for the high-frequency region. At high relative humidity, impregnated wood has a minimum tanδ (at around 4000Hz); and at other frequency ranges the tanδ of impregnated wood did not differ considerably from that of untreated wood. (3) The apparent activation energy of the mechanical relaxation process (ΔE) concerned with adsorbed water molecules was higher for impregnated specimens than for untreated ones at moderately high relative humidity, whereas at high relative humidity the difference was not observed. Based on these results, it is thought that the tanδ of impregnated wood decreased at low rela tive humidity because of the formation of direct hydrogen bonds between impregnated extractives and wood components. However, when the specimen is at higher relative

Sitka spruce (Picea sitchensis Carr.) was treated with water-soluble extractive components of pernambuco (Guilandina echinata Spreng. syn Caesalpinia echinata Lam.) by two methods: impregnation under evacuation using an aspirator and repetitive surface application using a brush. The influence of these treatments on the vibrational properties were examined. The loss tangent (tan δ) of the impregnated specimen decreased, up to nearly a half of its original value, with increasing weight gain. It is suggested that the decrease in tan δ results from impregnation of the extractive components into the amorphous region of cell walls, forming secondary bonds between matrix substances. The surface application of the extractive components, on the other hand, hardly brought about the desirable change in vibrational properties.

The factors that cause weather-induced deterioration of wood surfaces were determined by chemical and spectroscopic analyses. Albizzia (Paraserianthes falcata Becker.) and sugi (Cryptomeria japonica D. Don) were exposed to two temperate conditions of natural weathering with and without rainfall and to accelerated conditions of artificial weathering coupled with ultraviolet (UV) light irradiation and water flashing. Infrared spectroscopic analysis showed that the oxidative reaction of lignin was observed under all conditions of weathering for both wood species. However, a marked decrease in lignin and hemicellulose content were recognized when albizzia woods were exposed to weathering with water. Lignin content in the softwood sugi did not decrease as much as in albizzia even in the presence of water, but the modification of lignin macromolecules was assumed to be accelerated by water, as seen by electron spin resonance spectroscopy. These results showed that the presence of water promotes the weathering deterioration of wood under UV irradiation.

The impregnation of various simple phenolic and natural polycyclic compounds into wood was investigated from the viewpoints of vibrational property and dimensional stabilizing effect. When simple phenolic compounds were impregnated, the loss tangent (tan δ) in the longitudinal direction increased linearly with increasing weight gain. Meanwhile, among the natural polycyclic compounds hematoxylin decreased the tan δ drastically by impregnation. It was suggested that the five hydroxyl groups and the pyran ring oxygen in the hematoxylin molecule contribute to formation of the crosslinkage-type hydrogen bonds between wood components. The rigidity of hematoxylin molecules may also be important. By impregnation of about 10% catechol, resorcinol, and saligenin, a 40% level of antiswelling efficiency (ASE) was attained, although a significant dimensional stabilizing effect was not observed after impregnation of natural polycyclic compounds.

Participation of lignin in the reaction between vapor-phase formaldehyde and wood was examined by using gradually delignified wood meal. A fi rst-order rate equation was successfully applied to the weight gain data. From the estimated reaction parameters such as rate constant, k, and ultimate weight gain, a, the reactivity toward formaldehyde was discussed among wood components, and compared with that for acetylation. k decreased monotonously with progress of the elimination of lignin, suggesting that the reaction rate of lignin is dominant over that of whole wood, and the decrease in the ratio of lignin retarded the reaction of wood as a whole. On the other hand, a increased with decreasing lignin content. This may be attributable to the enhanced reactivity of the remaining lignin due to some structural changes and to the increase in the number of reactive sites in polysaccharides as a result of their exposure accompanying the elimination of lignin. The dependencies of k and a on the lignin content were not similar to the case for acetylation, probably because of the difference in the reaction phase. In vapor-phase formaldehyde treatment, the remaining lignin reacts as it is, whereas in liquid-phase acetylation it would undergo rearrangement or swelling of the structure in the reaction solution.

Pernambuco (Guilandina echinata Spreng. syn Caesalpinia echinata Lam.), which has been used for a material of violin bows, was compared with substitutable wood species from the viewpoint of processing properties and durability. The properties required for bow material are discussed. The shearing strength of pernambuco was higher than that of the other wood species at the same specific gravity. High shearing strength seems to be effective for preventing damage to the head (top part) of the bow. The presence of extractives affects the thermal softening of wood material, but ease of handling and permanent retention of form are not particularly superior for pernambuco, although it does have a high extractives content. A peculiarly low loss tangent (tan S) of pernambuco can probably be attributed to the large amount of extractives, rather than the mean microfibril angle.

Pernambuco (Guilandina echinata Spreng. syn Caesalpinia echinata Lam.), which has been used for the production of violin bows, was compared with alternative wood species from some physical and mechanical aspects and the reasons for why it is suitable for bows were discussed. A low value of loss tangent (tanδ) may be helpful for bow material, as the tanδ of pernambuco was much lower than that of the other wood species. On the other hand, the most suitable Young's modulus could not be clarified, but the lower limit of Young's modulus needed for the bow must exist. Low tan δ of pernambuco seems to be due to a high content of extractives. However, not only the quantity but also the chemical component and/or the location of extractives in the wood structure may have a different effect on the tanδ. From the result that the wood with a higher extractive content showed a lower softening temperature, it can be concluded that the extractive in pernambuco is probably influential in the thermal bending of the bow. Redness (a*) of pernambuco has a correlation with tan δ. Indentation hardness of pernambuco was somewhat higher at the same level of specific gravity than that of the other wood species, but hardness would not be an inevitable property for the bow.

The influence of varnishing for violins on the vibrational properties of wood strips was investigated during the drying process by vaporization of solvent. The changes of weight, specific dynamic Young's modulus (E/r), and loss factor (tan δ) by varnishing with natural varnish (shellac) and artificial ones (cellulose nitrate: CN) were measured and their subsequent changes were traced for 1 to 2 years. For both varnishes, though the remarkable weight decrease just after varnishing ceased after 3 to 5 months, a slight weight decrease continued even after then. The change of E/r became almost undetectable within several months, whereas the decrease of tan δ continued for a longer period. Acoustic converting efficiency (ACE) fell down drastically just after varnishing, but it recovered extending for 1 year. The recovery of ACE was more largely for shellac varnished specimen than the CN lacquered one. The frequency dependencies of E/r and tan δ changed by applying the CN lacquer, whereas those by shellac varnish did not change significantly. These results suggested the superiority of shellac varnish to CN lacquer

The isotherms for untreated and chemically modified wood specimens were analyzed using the Hailwood-Horrobin equation. The enthalpy (ΔH), entropy (ΔS) and free energy (ΔG) changes during moisture adsorption were estimated. The ΔH, ΔS and ΔG values of untreated wood were compared with those found in literature as well as those of specimens treated with formaldehyde (formalization), acetic anhydride, glyoxal, glutaraldehyde and dimethylol dihydroxy ethyleneurea (DMDHEU). The ΔH of the untreated specimen was almost constant up to ca. 3% of moisture content differing from the results reported by others. However, the bonding energy of hydrated water molecules, which represented a large portion of the adsorbed water in the low moisture content range, should not depend on the moisture content, so that the result was undoubtedly reasonable. The decrease of the absolute ΔH for formalization and acetylation indicates the remarkable loss of the hydrated water as a result of those reactions. A peak of the absolute ΔS found at ca. 3% of moisture content except for acetylation may reflect the rearrangement of adsorption sites associated with taking up a little amount of water. In the acetylation, such rearrangement could not occur because of the introduction of bulky hydrophobic groups. Remarkable increases of both absolute ΔH and ΔS observed in glyoxal treated specimens will be due to the ester groups introduced. The ΔG values of all treated specimens became less negative than that of untreated. This indicates the formation of some bond which restricts the swelling of wood structure.