Holzforschung, Vol. 64, pp. 315–323, 2010 • Copyright  by Walter de Gruyter • Berlin • New York. DOI 10.1515/HF.2010.044 2010/94 Article in press - uncorrected proof Water sorption in wood and modified wood at high values of relative humidity. Part I: Results for untreated, acetylated, and furfurylated Norway spruce Lisbeth G. Thygesen 1, *, Emil Tang Engelund 2,3 and Preben Hoffmeyer 3 1 Forest and Landscape, University of Copenhagen, Frederiksberg C, Denmark 2 Wood and Textile, Danish Technological Institute, Taastrup, Denmark 3 Department of Civil Engineering, Technical University of Denmark, Lyngby, Denmark *Corresponding author. Forest and Landscape, University of Copenhagen, Rolighedsvej 23, DK-1958 Frederiksberg C, Denmark E-mail: lgt@life.ku.dk Abstract Desorption isotherms at 208C for untreated, acetylated, and furfurylated Norway spruce w Picea abies (L.) Karst.x sap- wood were established in the 91.9–99.9% relative humidity (RH) range. Three methods were employed to secure various constant RH levels: saturated salt solutions, climate cham- bers, and the pressure plate technique. The curve form for the untreated samples did not show an upward bend, except perhaps above 99.5% RH, indicating that – contrary to what has hitherto been assumed – capillary condensation does not play a significant role for water sorption in wood below fiber saturation. Three additional results corroborate this conclu- sion: (1) calculation of the theoretical contribution of capil- lary condensation to the moisture content (MC) in wood based on idealized microstructural geometries by means of the Kelvin and Laplace equations resulted in very small con- tributions to the equilibrium moisture content (EMC), i.e., below 0.35% moisture at 99.9% RH. (2) The ratio between the EMC of acetylated and untreated samples did not show an increasing trend for increasing RH, as would have been the case if capillary condensation had taken place in both untreated and acetylated wood. (3) Low field time domain nuclear magnetic resonance results showed that only the relaxation curves from the furfurylated samples were affect- ed systematically by freezing, indicating that neither untreat- ed nor acetylated wood contained significant amounts of capillary condensed water. Keywords: acetylation; capillary condensation; furfuryla- tion; isotherm; sorption; wood-water relations. Introduction Moisture in wood has been studied for decades. Several explanations have been suggested regarding the way water molecules are bound in the wood structure as a function of the relative humidity (RH). One such explanation often cited relates the uppermost part of the sorption isotherm to capil- lary condensation in voids in the material (Sheppard 1933; Barkas 1937; Babbitt 1943; Stamm 1950; Spalt 1958; Koll- mann 1962; Simpson 1973). However, the question of cap- illary condensation at high levels of RH has not yet been resolved. The present study employs three techniques (satu- rated salt solutions, climate chambers, and the pressure plate technique) to examine the nature of moisture in wood in the RH range from 91.9% to 99.9%. Furthermore, two different types of chemically modified woods are included in the experiments as a tool to highlight the sorption behavior of untreated wood. Void radii and capillary condensation In the following, cavities, pores, voids, etc., in wood are categorized and termed according to their dimensions (Grif- fin 1977). This system complies better with wood anatomy than the IUPAC definition of pore size classes (Sing et al. 1985; Rouquerol et al. 1994). Macrovoids comprise the lumi- na of cells with radii from approximately 5 mm to 200 mm or greater. Microvoids include the pointed ends of lumens, pit apertures, pit-membrane voids, and other small voids with radii in the range from 5 nm to 5 mm. Nanovoids are found in the cell wall and range in size from 5 nm down to the level of single water molecules bound to hydrophilic sites of the wood polymer, at which level the concept of pore radii becomes meaningless. It is a generally accepted assumption that the dry cell wall is essentially non-porous (Griffin 1977; Fahle ´n and Salme ´n 2005; Salme ´n and Fahle ´n 2006) and that nanovoids emerge as a result of the presence of water. Meas- urements on different wood species in either green or satu- rated condition show a maximum nanovoid size of 1.8 nm for green spruce (Stone and Scallan 1968), 0.9–1.0 nm for green Sitka spruce (Tarkow et al. 1966), 0.8 nm for saturated sweetgum (Flournoy et al. 1991), and 2 nm for saturated Corsican pine (Hill et al. 2005). The Kelvin equation relates the relative humidity to the void radius that leads to capillary condensation within voids of these dimensions. Corresponding values of RH and void radius are given in Table 1. It has long been debated in what range the Kelvin equation is valid. Skaar (1988) argued that below 80% RH the concept of capillary condensation is questionable, as at this RH the corresponding pore radius of 4.9 nm was only approximately 15 times the diameter of a water molecule (0.3 nm). Measurements of menisci in mica capillaries of 5 nm diameters confirm that the Kelvin equa- tion is valid down to approximately 80% RH (Kohonen and