Interrogating helical nanorod self-assembly with fractionated cellulose nanocrystal suspensions

Like Comment
Read the paper

This work recently published in Communications Materials beautifully summarizes my PhD journey under the supervision of Prof. Jan Lagerwall. It builds on the knowledge generated during the four-year thesis, with the main focus on fractionation of cellulose nanocrystals (CNCs) and the new information that can be gained thanks to the fractionated CNC suspensions. CNCs can be described as rod-like nanoparticles of crystalline cellulose, which are extracted from plant materials, usually by acid treatment of purified cellulose. CNCs fresh from production are highly disperse, i.e., there is much variation in the length of the individual particles. Depending on the acids used, the CNC surface is covered with electrical charges, which make it easy to disperse CNCs in water. If the concentration of CNCs is high enough, the liquid changes character in a very fundamental way: the rods spontaneously organize along a helix to form what is called a cholesteric (or chiral nematic) liquid crystal phase. This is fascinating from a fundamental science perspective, but it also becomes practically useful, because after evaporating the water, the helical structure can be maintained in the solid film that is left behind. The helix period shrinks in this process to sub-micrometer values, and this gives rise to iridescent colours, as in peacock feathers or butterfly wings. This “structural color” arises when light interacts with materials that are periodic with a repeat distance similar to the light wavelength, according to a process called Bragg reflection.

Because of the high dispersity of CNCs, there has long been an interest in a good a method for fractionation in the CNC community, but until our work it was not quite clear how strong the impact of dispersity actually is. We actually use the liquid crystal formation itself to do the fractionation, because at a specific concentration range, the ordered cholesteric phase and the normal disordered liquid phase exist together, the disordered floating on top of the ordered phase, and because long CNCs are better at organizing than shorter ones, the cholesteric phase contains the longest rods while the disordered phase contains the shortest ones. The idea of fractionation by the separation of phases in this way is not ours. In fact, it was demonstrated for CNCs already in 1998 by the group of Derek Gray, who is one of the pioneers of CNC research. A more detailed experimental study on liquid crystal-driven fractionation of disperse nanorods was reported by the Windle group for carbon nanotubes (CNTs) in 2006. But while the principle was known, nobody had devised an efficient scheme of using it in practice.

I remember that I found the CNT paper at home and I made a small sketch on a piece of paper. I thought that the most efficient way to fractionate CNCs would be to use an initial parent CNC suspension that separated in equal volumes of disordered and ordered phase and by using a pipette, separate one phase from the other. Afterwards, either by evaporation or concentration of the phases, we could repeat the process couple of more times. This procedure worked well (although we now know that it is not the best procedure!) and the results were published in NPG Asia Materials in 2018. The main drawback of this method was the contamination of the cholesteric phase by the disordered phase during the extraction by pipette, so a better control would be good, and it would also be difficult to scale this approach to large volume. In another project, my supervisor Prof. Jan Lagerwall learnt from a post-doc with chemistry background, Dr. Anshul Sharma, about the use of a separatory funnel, which chemists use to separate reaction products from each other. Being a physicist, Jan did not know of this, but he realized it should be possible to use it also for physically separated phases, so he asked me to try it on the CNC suspension. First, I thought it didn’t work, because the volume was now so large that it was difficult to see the difference between the phases, but then I came up with the trick to illuminate the funnel from behind with my phone torch, and this immediately revealed that the cholesteric phase had indeed separated below the disordered one. Now we had a method that allowed clean separation of the two phases and was scalable to large volumes, so we were really excited about having a way of fractionating CNCs in large quantities!

Figure 1: (a) In cholesteric liquid crystals, the rod-like building blocks rotate in a helical fashion, with the period (pitch, p) known as the distance for which the rods make a complete turn. (b) Schematic illustration of a CNC rod in suspension, electrically stabilized with negative surface charges and Na+ as counterions. (c) Atomic Force Microscopy (AFM) image of a sample of CNCs deposited on a charged mica surface.

In our new paper, we present the fractionation using the separatory funnel, and at the same time we upgrade the strategy from the 2018 version, so we can now get many different fractions with stepwise increasing CNC length. The fractionation procedure is improved by varying the initial macroscopic volume fraction at which the phases were separated, which also reduces the multiple separation cycles from three to only one if only the longest rods are desired. This new method allows the easy access to several fractions of CNC samples with different rod sizes and much reduced dispersity. Because of this new strategy, we are able to present (for the first time­) the real effect of the rod size on the development of the helix in cholesteric liquid crystalline suspensions and on how this affects the later formation of colourful solid films. One very surprising result is that the helix pitch gets shorter the longer the CNC, which is actually opposite of what was generally assumed in the community. Another striking, and very important effect, is that the viscosity of cholesteric suspensions of long rods is much lower than that of short rods. This means that we get a uniformly organized helix much faster for long than for short rods, and we showed that this allows us to get uniform violet-reflecting films from long rods, when unfractionated CNC suspensions dried in the same way give non-uniform appearance. There are still many intriguing mysteries in the physics and chemistry of CNC suspensions, but we think our new results and the new method for large-scale fractionation of CNCs can break new ground both in the basic understanding and in new applications of these biomaterials, for instance in pigments that do not degrade with time and that are environmentally friendly.

For more details about our work, please check out our paper here: