Following researches carried out during the evaluation of this project (2022-2023) and published in Langmuir (Maillet et al, 2022) and Science Advances (Cocusse et al. 2022), we have developed techniques of observation and quantification of the amount, state and location of bound water by NMR relaxometry [1-4] and MRI [1] which allow to characterize the transport and phase changes of bound water through the material structure. It is worth emphasizing that, over the world, we are the only ones to use such techniques to systematically characterize bound water transport.
With the help of these techniques, and by filling the pores (between the solid) with oil, we have been able to show for the first time that bound water can diffuse inside the solid structure without any exchange with vapor or water in the pores. This was done initially for a cellulose fiber network (Zou et al., Cellulose, 2023). This result was then generalized for bound water in wood [5]. This work provided the first direct measurements of the transport diffusion coefficient of bound water in the different directions of wood.
Since in general construction materials are used under hygroscopic conditions (no liquid water in the pores), the next step is to properly characterize the coupled transport of bound water and vapor in such materials. In that framework we developed a fundamental model [3] for the project, which takes into account both transport types, i.e. vapor diffusion, bound water diffusion strictly inside the solid (see above), and the exchanges between the two phases by absorption or desorption. We could show that, this model simplifies in a single diffusion model describing the total moisture transport, with a diffusion coefficient depending on the bound water fraction. The comparison with a series of MRI experiments of transfers from cellulose fiber samples at different porosities demonstrated the ability of the model to predict in details, i.e. at a local scale inside the medium, the evolution of the moisture content over time [1, 3].
We have worked in parallel at a better characterization and understanding of the thermal properties of these materials. We started by considering in more details the standard procedures and discovered that they should be adapted to prevent heat losses, very important with insulating materials of low conductivity, to significantly perturb them [6]. Relying on this new approach we are currently pursuing a systematic study of the thermal conductivity of different materials as a function of their physical characteristics such as the orientation of the fibers, their volume fraction, and their exact components (composite, mixtures). This contrasts with previous approaches which just observe and compare the properties of the different materials without further scientific analysis. Although not identified as a critical domain of research in the initial project (Description of action), this appeared of fundamental interest to progress in that field in order to provide as complete knowledge and technical bases for properly describing the thermal transfers in such materials.
Equipped with these different tools and knowledge, we have started to advance towards the modelling of the hygroscopic behavior of a biobased wall. This is based on a standard set of equations for describing the mass conservation and the energy transport, but now describing the physics in agreement with our clarifications of the mechanisms at a local scale (see above). The essential point is the fact that we properly take into account bound water transport and exchanges with vapor. Thus, for the first time, we were able to introduce in the mass conservation equation a model describing the full moisture transport (our model described above). Moreover we could also express in the model the variations of these parameters as a function of the temperature and humidity, thus fulling expressing the couplings between the equations. We emphasize that, also in contrast with previous models, there is no fitting or unknown parameters, all physical parameters are determined from independent straightforward measurements. The full model [7] has been set up and, through a similarity approach, its general predictions in terms of dominant mechanisms (heat or moisture transfers) and coupling effects, have been analyzed, providing for the first time general conclusions on the hygrothermal behavior of biobased materials (which contrasts with the multiple studies concerning one specific material under specific conditions, and in general focused on global trends, without definitive conclusions).
[1] Yan et al., Physical Review Applied 23 (1), 014073 (2025)
[2] Julien et al., Langmuir 40 (51), 26975-26987 (2024)
[3] Zou et al., PNAS nexus 3, pgad450 (2024)
[4] Yan et al., Wood Science and Technology 59, 29 (2025)
[5] Yan et al., in press in Physical Review Applied (2025) Editor’s suggestion
[6] Mourda et al., J. Building Engineering, 112996 (2025)
[7] Daunais et al., under revision for Int. J. Heat and Mass Transfers (2025)
