Design of experiments to investigate multi-additive cellulose nanocrystal films

Abstract

Cellulose nanocrystal (CNC) suspensions can self-assemble into chiral nematic films upon the slow evaporation of water. These films are brittle, as indicated by their fracturing instead of plastically deforming once they are fully elastically deformed. This aspect can be mediated to some extent by plasticizing additives, such as glucose and glycerol, however, few reports consider more than one additive at a time or address the influence of additive content on the homogeneity of the self-assembled structure. In this work, design of experiments (DoE) was used to empirically model complex film compositions, attempting to relate additive concentrations in dilute suspension to film properties, and to understand whether outcome specific predictions are possible using this approach. We demonstrate that DoE can be used to predict film properties in multi-additive systems, without consideration given to the different phenomena that occur along the drying process or to the nature of the additives. Additionally, a homogeneity metric is introduced in relation to chiral nematic organization in CNC films, with most of the additive-containing compositions in this work found to reduce the homogeneity of the self-assembly relative to pure CNC films.

Keywords: additives; cellulose nanocrystals; design of experiments; films; machine learning; self-assembly.