Multi-parametric high-resolution MRI for the characterisation of water and PEG solutions in wood

Abstract Waterlogged wooden artifacts are often unstable when exposed to open air, and impregnation with polyethylene glycol (PEG) remains one of the most widely adopted conservation methods to ensure their long-term stability. Despite its widespread use, relatively few studies have focused on the interaction between PEG, water and wood. Motivated by the development and testing of new NMR-based imaging (MRI) protocols for analysing water-PEG-wood systems, NMR diffusion and relaxation experiments were first conducted on PEG-water solutions at increasing concentrations. High-field (9.4T) micro-MRI (µ-MRI) was then employed to examine the penetration and interaction of these solutions within chestnut wood, adapting diffusion models originally developed for medical diagnostics. NMR spectroscopy revealed clear concentration-dependent variations in relaxation and diffusion parameters, indicating a progressive reduction in molecular mobility with increasing PEG content. High-resolution µ-MRI weighted by T 2 *, T 2 , T 1 , and diffusion (quantifying the diffusion coefficient D and the kurtosis parameter K ) allowed non-destructive observation of the wood microstructure and the quantification of relaxation ( T 2 , T 1 ) and diffusion ( D , K ) values in the earlywood and latewood regions during treatment. Increasing PEG concentration resulted in decreasing T 1 and D values, while T 2 and K exhibited contrasting behaviors associated with changes in local molecular dynamics and microstructural heterogeneity. Notably, the kurtosis parameter proved sensitive to microstructural differences between earlywood and latewood and to the progressive evolution of the system during treatment. Overall, the study suggests the potential of µ-MRI approach to capture the increasing complexity of the water-PEG-wood system, supporting its applicability for non-invasive monitoring of consolidation treatments in small samples of waterlogged archaeological wood.