A rapid and low-cost route for characterizing heterogeneous fibrillar cellulose dispersions to support papermaking process control

Abstract Microfibrillated cellulose (MFC), cellulose nanofibers (CNF), and their mixtures exhibit different optical responses under flow, but simple methods for tracking these differences in process-relevant conditions are limited. Here, we evaluate a compact polarized-flow imaging setup based on LED illumination, linear polarizers, a camera, and a simple flow cell, combined with principal component analysis (PCA) of image sequences. The aim is not to extract physical material parameters from the images, but to determine whether different suspension states can be reproducibly discriminated in a simple optical measurement. For two sample sets, varying either the MFC:CNF ratio or the CNF concentration produced distinct and repeatable clusters in PCA space. MFC-rich suspensions formed broader clusters consistent with greater optical heterogeneity, whereas CNF-rich suspensions formed tighter clusters. Rheological measurements were used as an independent comparison and showed that optically distinct states were also mechanically different. In a holdout demonstration, an unseen 1:1 mixture was projected to an intermediate region between reference clusters, and small intensity perturbations shifted the projected cluster position. These results show that simple polarized imaging can support state discrimination of MFC/CNF suspensions and may serve as a basis for future at-line monitoring or deviation detection in papermaking.