Self-Organizing Cellulose Nanorods: From the fundamental physical chemistry of self-assembly to the preparation of functional films

Cellulose nanocrystals (CNCs), nanorods isolated by acid hydrolysis from cellulose sources, be-
long to a selective type of functional biomaterials. The intriguing ability of these nanoparticles
to self-organize and develop a chiral nematic liquid crystal phase when suspended in aqueous
suspensions, is increasing interest regardless of the diverse range of research fields. Unfortu-
nately (or fortunately, for this thesis), pristine CNCs are always disperse, with great variations
in rod length within a single sample. Of particular interest is the fractionation of CNC rods by
separation of the coexisting phases: isotropic phase from the liquid crystalline (LC) part. Since
the aspect ratio is considered to be the critical parameter that dictates the particle fraction
at which cholesteric-isotropic phase separation starts, it is expected that the high aspect ratio
rods will separate from low aspect ratio rods, and this is indeed what I found in this thesis.
By a systematic repetition of separation of phases, I could reach a quality of separation of long
from short rods that is remarkable. The fractionation procedure was then improved by varying
the equilibrium phase volume fraction at which the phases were separated, reducing with
this new procedure the multiple separations from five cycles to only one. The onset of liquid
crystallinity was drastically reduced in the long rod fraction and the decrease in the threshold
for complete liquid crystallinity was even stronger.
The mass fraction threshold at which gelation of the CNC suspension is triggered is not at all
affected by the fractionation. Since gelation is a percolation phenomenon, the expectation was
that also the onset of gelation would move to lower mass fractions, but this remained at about
the same value. Together with the shift to lower mass fractions of the cholesteric liquid crystal
phase formation we have thus opened access to a whole new range of the equilibrium phase
diagram, where the full sample is cholesteric yet not gelled.
I demonstrate that the critical parameter for inducing gelation is in fact not the fraction of
CNC, but the concentration of counterions in the solution. This suggests that the gelation is
more complex than direct percolation between individual CNC rods, and instead is related to
loss of colloidal stability due to reduced electrostatic screening.
I also show that the behavior of key parameters, such as the period of the helical modulation,
so-called pitch, that is characteristic of the cholesteric phase, is very different in the range of
phase coexistence compared to the range of complete liquid crystallinity. In addition, I found
that the dependence of the pitch on CNC mass fraction has less to do with the size of the nanorods but rather than with the variation of effective volume fraction as a result of more rods in the suspension or higher counterion concentration. I corroborate this hypothesis by adding different amounts of salt to CNC suspensions of varying mass fraction such that the ion concentration is held constant, thereby tuning the pitch to the same value throughout the suspensions.
In films prepared by drying CNC suspensions, the pitch can go down to a few hundred nanome-
ters, resulting in circularly polarized colorful Bragg re ection of visible light. By working with
the long-rod fraction we can absolutely obtain a highly-ordered monodomain structure that results in uniform color of films, with only one circular polarization re ected, as should be the case.
While the study is carried out on CNCs, the implications go far beyond this particular nanoma-
terial, revealing new challenges and opportunities in general liquid crystal and colloid physics, as
well as in strategic research where fractionation and the drying of initially disperse populations
of nanorods is desirable.