Nanoparticles are very interesting building blocks. Their large surface-to-bulk ratio gives them different properties from those of larger particles. Controlling their assembly can greatly affect macroscopic material properties. This often happens in nature, resulting in macroscopic materials with properties far better than those of similar human-made materials. However, in this fast-growing research field, we may soon compete with nature in certain areas. This thesis demonstrates that the distribution and orientation of nanocellulose particles can be controlled, which is crucial for many applications.
Nanocellulose is an interesting nanoparticle, for example, because of its high strength, low thermal expansion, and high crystallinity. Nanocellulose particles are called nanofibrillated cellulose (NFC) or cellulose nanocrystals (CNCs). NFC is obtained from wood by mechanically shearing apart fibrils from the fiber wall and to obtain CNCs, parts of the cellulose are broken down by hydrolytic acidic reactions, most commonly, prior to homogenization. NFC particles are longer and less crystalline than are CNCs, but both are similar in width. The particles attract each other in aqueous dispersions and have a high aspect ratio and, thus, a large tendency to aggregate. The rate at which this occurs is typically reduced by charging the particles, generating an electrostatic repulsion between them.
To fully utilize the many interesting properties of nanocellulose, the aggregation and orientation of the particles have to be controlled; examining this delicate task is the objective of this thesis. The limits for particle stability and aggregation are examined in papers 2–3 (as well as in this thesis) and orientation of the particles is investigated in papers 3–5. In addition, the liberation of the nanoparticles from different types of wood fibers is studied in papers 1 and 2.
It was found that the liberation yield improved with increased fiber charge. In addition, the charge of the fibrils is higher than the charge of the original fibers, indicating that the fibrils were liberated from highly charged parts of the fibers and that the low-charge fraction was removed during processing.
Aggregation was both theoretically predicted and experimentally studied. A theoretical model was formulated based on Derjaguin–Landau–Verwey–Overbeek theory, which is intended to predict the influence of salt, pH, and particle charge on the colloidal stability of the NFC. To predict the experimental trends, specific interactions between salt counterions and the particles charges had to be included in the model, which greatly increased the effect of salt on the NFC stability. Below the particle overlap concentration, instability induced by pH or salt created small sedimenting flocs, whereas above the overlap concentration the system gelled. Increasing the particle concentration further also gels the system.
Orientation of nanocellulose was first achieved by shearing, salt- or acid-induced NFC gels. This oriented the fibrils and increased the gel modulus in the direction of shear. The orientation persisted after the shear strain was released and did not cause breakdown of the macroscopic gel. The orientation is probably due to rotation in the interfibril crosslinks, which is possible because the crosslinks are physical, not covalent.
Second, orientation was also induced by elongational flow. Shear and acceleration forces were combined to align fibrils in the direction of the flow. The orientation was then frozen by gelation (adding salt or reducing the pH). Drying the gel threads created filaments of aligned fibrils with a higher specific strength than that of steel.
Finally, CNC particles could be aligned on flat surfaces. The particles were first forced to align due to geometrical constraints in grooves on a nanowrinkled surface. The CNCs were then transferred to a flat surface using a contact-printing process. This created surfaces with lines of highly aligned CNCs, where the line–line spacing was controlled with nanometer precision.
Stockholm: KTH Royal Institute of Technology, 2013. , vi, 49 p.