How lignin sticks to cellulose-insights from atomic force microscopy enhanced by machine-learning analysis and molecular dynamics simulations

Advancing plant cell wall modelling: Atomistic insights into cellulose, disordered cellulose, and hemicelluloses - A review

The complexity of plant cell walls on different hierarchical levels still impedes the detailed understanding of biosynthetic pathways, interferes with processing in industry and finally limits applicability of cellulose materials. While there exist many challenges to readily accessing these hierarchies at (sub-) angström resolution, the development of advanced computational methods has the potential to unravel important questions in this field. Here, we summarize the contributions of molecular dynamics simulations in advancing the understanding of the physico-chemical properties of natural fibres. We aim to present a comprehensive view of the advancements and insights gained from molecular dynamics simulations in the field of carbohydrate polymers research. The review holds immense value as a vital reference for researchers seeking to undertake atomistic simulations of plant cell wall constituents. Its significance extends beyond the realm of molecular modeling and chemistry, as it offers a pathway to develop a more profound comprehension of plant cell wall chemistry, interactions, and behavior. By delving into these fundamental aspects, the review provides invaluable insights into future perspectives for exploration. Researchers within the molecular modeling and carbohydrates community can greatly benefit from this resource, enabling them to make significant strides in unraveling the intricacies of plant cell wall dynamics.

Performance Assessment of Universal Machine Learning Interatomic Potentials: Challenges and Directions for Materials' Surfaces

Machine learning interatomic potentials (MLIPs) are one of the main techniques in the materials science toolbox, able to bridge ab initio accuracy with the computational efficiency of classical force fields. This allows simulations ranging from atoms, molecules, and biosystems, to solid and bulk materials, surfaces, nanomaterials, and their interfaces and complex interactions. A recent class of advanced MLIPs, which use equivariant representations and deep graph neural networks, is known as universal models. These models are proposed as foundation models suitable for any system, covering most elements from the periodic table. Current universal MLIPs (UIPs) have been trained with the largest consistent data set available nowadays. However, these are composed mostly of bulk materials' DFT calculations. In this article, we assess the universality of all openly available UIPs, namely MACE, CHGNet, and M3GNet, in a representative task of generalization: calculation of surface energies. We find that the out-of-the-box foundation models have significant shortcomings in this task, with errors correlated to the total energy of surface simulations, having an out-of-domain distance from the training data set. Our results show that while UIPs are an efficient starting point for fine-tuning specialized models, we envision the potential of increasing the coverage of the materials space toward universal training data sets for MLIPs.

Degradation selectivity for bamboo fiber and parenchyma lignin-carbohydrates complexes (LCC) esters

The degradation of lignin-carbohydrate complex (LCC) esters has been proven to be crucial for the selective separation of lignocellulosic components. This study utilized Raman microspectroscopy to image the preferential degradation of lignin and LCC esters from the bamboo wall during successive NaOH (0.2 to 5.0 % w/w), H2SO4 (1 to 8 % v/v), and NaClO2 (5 to 20 min) treatments. Raman imaging showed that lignin and LCC esters were selectively removed from the middle lamella of fibers and the secondary wall of parenchyma during NaOH and NaClO2 treatments. In contrast, H2SO4 primarily caused the simultaneous removal of lignin and LCC esters from the fiber wall under harsh conditions (8 %), while the middle lamella of parenchyma was less affected, both morphologically and topochemically. Raman spectral analysis indicated that the band intensity at 1605 cm-1 for lignin and at 1173 cm-1 for LCC esters decreased by >87.0 % in the highly lignified parenchyma secondary wall after a 5.0 % NaOH treatment, while the decrease was <67 % in the fiber wall. Interestingly, a strong linear correlation was observed between LCC esters and carbohydrates in the parenchyma (R2 > 0.912). These findings provide important insights into the graded and classified utilization of bamboo resources.