Nanoscale Insight in Moisture Adsorption in Lignin, Amorphous Cellulose, and Their Mixtures Investigated with Monte Carlo and Molecular Dynamics Simulation

Second to cellulose, lignin is the most abundant renewable polymer on the planet. Majorly found in plant cell wall composite, cross-linked to hemicellulose, and acting as the cohesive matrix that surrounds the wood holocellulose and provides increased hydrophobicity and chemical stability to the composite, Lignin is an exceptional source for different monomers and novel materials, biofuel feedstock and chemicals. In addition, Due to its less susceptibility to biological attack and high compatibility with holocellulose, synthesized lignin-based oligomers and lignin nanoparticles have been considered as potential consolidation material for archaeological wood. Due to the highly heterogeneous structure of lignin and complications of its extraction and separation in experimental approaches, relatively little evidence exists on lignin and cellulose interaction and its response to water adsorption on the molecular level. In this study, a molecular model is used to investigate hygromechanical properties of lignocellulose mixtures and to provide insight into the role of lignin and its derivatives as consolidation agents in wood composites with the aim of finding a mixture model that accurately describes the behavior of lignocellulose mixtures. Molecular models studied include an amorphous cellulose model composed of cellobiose units, various models for lignin from linear lignin composed of guaiacyl (G-type) and syringyl (S-type) units together with more complex models of softwood (spruce) and hardwood (birch) built using a stochastic method to accommodate more complex arrangements of units and linkages and mixtures of amorphous cellulose and uncondensed G-type lignin of different mass ratio. Hydrated samples equilibrated at relative humidity ranging from fully dry to saturation pressure are prepared by employing an iterative hybrid all-atom molecular dynamics and grand canonical Monte-Carlo (GCMC) simulation. The hybrid GCMC/MD technique provides an excellent tool to capture the highly coupled nature of sorption-induced swelling in the lignocellulose mixture by enabling us to capture the sorption by applying relative humidity in GCMC and allowing the resultant swelling during relaxation in the MD stage of the simulation. By applying the mechanical tests to hydrated samples role of lignin in adsorption-induced mechanical softening in wood-polymer is also studied. MD results for different lignin components show linear G-type lignin as the least hygroscopic and the linear S-type lignin, with the higher number of methoxy group, as the most hygroscopic lignin model examined. Amorphous cellulose shows considerably higher adsorption and swelling compared to all lignin models due to its hygroscopic nature and the high number of hydroxyl groups hosting many sorption sites. Looking at mixtures of the two components, as the lignin content increase, The sorption isotherms, swelling curves and mechanical data reveal reduced moisture adsorption and swelling together with mechanical softening signifying the effect of the polymer interaction, which is further characterized by measuring porosity pore size distribution and hydrogen bonding network. The rule of mixture is introduced as an analysis tool to elicit the role of polymer interphase. Comparing the mixture rule to MD data reveals the mixture rule as an acceptable but overestimating approximation of moisture content and swelling, the difference is then taken into account as the properties of cellulose-lignin interphase. The underlying molecular mechanism describing the lignocellulose hygromechanical response is compared to what we learned in our recent studies regarding molecular phenomena involved in polyethylene glycol consolidation of wood in order to unravel molecular characteristics of an inspired wood cell wall consolidant. Although lignin induces less softening in a mixture with amorphous cellulose, the moisture reduction and anti-swelling effects are less significant.