Effect of impregnation on mechanosorption in wood and paper studied by dynamic mechanical analysis

The results of dynamic mechanical measurements in dual cantilever and uniaxial tension of untreated Scots pine veneer and unbleached sulphate paper under stepwise changes in the relative humidity of the surrounding atmosphere between 5% and 85% are compared to corresponding results for acetylated samples and samples treated by impregnation with poly(ethylene oxide) (PEO), glycerol or melamine formaldehyde resin. The treatments resulted in significant reductions in the transient mechanical losses associated with stepwise humidity changes. Although acetylation of wood and impregnation with PEO or glycerol both result in greatly improved dimensional stability with respect to humidity changes, the two types of treatment appear to be fundamentally different in terms of the models used here to interpret the results. We find that it is helpful to consider the dynamic process of moisture sorption/ desorption in these samples in terms of two coupled processes: the diffusion of moisture into or out of the sample, with a characteristic diffusion time given approximately by t* = (pi/16)h(2)/D where h is the sample thickness and D a characteristic diffusion constant for the moisture in the material, and the attachment/detachment of (clusters of) water molecules to binding sites in the samples, apparently with characteristic "chemical" relaxation times of the order of seconds or minutes. Clearly, when the sample thickness is quite small, 0.8 mm in our case, and the diffusion coefficient of the order of 50 x 10(-8) cm(2)/s, then the characteristic diffusion time is about 40 minutes, already for these small dimensions more than an order of magnitude larger than the characteristic "chemical" relaxation time. By varying the frequency of the dynamic mechanical measurements these characteristic processes for untreated and treated samples can be probed in some detail.