The journey of scientific discovery often starts with a simple, "Hmm, I wonder if this could work?" For this paper, that question quickly snowballed into a kaleidoscope of challenges and surprises. Our goal was to try and understand how chirality transfer between a dopant and host LC works, and how to control and lock-in the vibrant, dynamic colors of cholesteric liquid crystals (CLCs) by harnessing the bistability of hydrazone-based photoswitches—a tool developed in the Aprahamian group.[1] These colorful CLCs are unique because it is their helical structures, and associated pitches, that allow them to reflect specific wavelengths of light.[2] So, naturally, the idea of modulating this helical pitch (and thus tuning the color) with triptycene-hydrazone chiral dopants seemed like an irresistible. Why triptycene, you ask? Well, it was a nudge from none other than Prof. Timothy Swager from MIT, who casually dropped the idea during a Telluride meeting. While his group has extensively worked with triptycenes in the context of aligning LCs,[3] developing them into chiral dopants was a whole new ballgame. The challenge? I had never synthesized chiral enantiomers before. Piece of cake, right? Well, not quite. I dove in, starting with the synthesis of triptycene-based diastereomers using chiral synthons (Fig. 1a). It turns out, separating these diastereomers via column chromatography was like trying to find a needle in a haystack—with one hand tied behind my back. So, I turned to my trusty prep thin layer chromatography (TLC) plates, running them over and over, waiting for that glorious separation to show up. It was a tedious process and not exactly scalable, but eventually, I had enough to move forward for the preliminary results and convincing my advisor, Prof. Ivan Aprahamian, to shell out a few extra dollars for the separation of racemic 2,6-diamino triptycene using a preparative chiral column (Fig. 1b-d). After that small financial victory, we had a solid collection of dopants in hand, ready to explore the triptycene-hydrazone based dopants (Fig. 2a), which exhibit multistable states because of the bistability of hydrazone photoswitches. These states arise from the varied isomeric ratios obtained under different irradiation wavelengths (Fig. 2b), offering an exciting potential for the precise control over material properties.

Fig. 1. Synthesis and characterization of the 2,6-diamino triptycene enantiomers (R, R)-8 and (S, S)-8.  a via diastereomeric separation method initially (not scalable). b via scalable chiral column (CHIRACEL OD-H) method. c Chiral HPLC chromatogram. d Circular dichroism spectra.

And then—Eureka! For the first time, we were able to obtain visible light reflection from doped LCs using hydrazone photoswitches. Cue the confetti! The subsequent photomodulation unlocked a beautiful spectrum of colors, which could be dynamically controlled and locked-in (Fig. 3). When I say locked-in, I mean trapped for far longer than what other photoswitchable dopants can handle.

Fig. 2. Chemical structures and photoswitching behavior of the chiral dopants. a Chemical structure of dopants (S, S)-1 to (S, S)-7 and their Z to E photoisomerization. b The UV-Vis spectra showing the photoisomerization process of the representative chiral dopant (S, S)-1 as a function of different irradiation wavelengths.

The challenge, however, was finding the right dopant that could not only integrate seamlessly with the CLCs but also reflect a wide range of the electromagnetic spectrum. With a range of triptycene-hydrazone derivatives in hand having different terminal groups (e.g., long alkyl chain to short alkyl chain, and everything from non-polar to polar groups) we eventually identified a dopant that had all the properties we were hoping for. Also, through the structure-property studies, we were able to answer the simple yet profound question: How does chirality transfer from a dopant to a LC host at the molecular level? While the concept of chirality transfer is well-documented, the exact molecular interactions driving this process remain largely unexplored.

Fig. 3. Reflection properties of (S, S)-1 doped in 5CB as a function of different irradiation times using 442 nm light. a Images of reflected colors. b Transmittance spectra. 

One particularly finding was related to how the helical twisting power (β) values of our dopants varied—especially between the E and Z isomers. For every dopant, the E isomer showed higher β values, indicating that it interacts better with the LC host than its Z counterpart. This was a revelation! The E isomer’s flexibility helped it cozy up to the LC host, while the Z isomer, stiffened by intramolecular hydrogen bonding, was not quite as sociable. Another observation was that swapping a long decyloxy chain for a shorter methoxy group resulted in the E isomer β value to have a nosedive. We figured that the loss of dispersion interactions between the dopant's alkyl tail and the LC host’s alkyl chains was the reason for this effect. But here’s where things got interesting: our studies threw a wrench into the long-held belief that only highly polarizable or rigid dopants give you those coveted high β values. Surprise! Some of our flexible, electron-rich dopants (methoxy and dimethylamino groups) worked wonders, while electron-withdrawing groups like nitro or fluorine were a little shy. Yet, the real wildcard? A dopant with a cyano group, which bonded just as well as the electron-rich dopants thanks to its structural resemblance to the LC host.

Apart from understanding the chirality transfer among the LC mesogens, all these dopants allowed us to dynamically tune and lock-in the reflected visible colors. The next question was how can we push the boundaries of what was possible with these dopants? Fortunately, the collaboration between the Aprahamian and Lippert research groups allowed us to use the latter’s Digital Light Processing (DLP) micropatterning techniques (Fig. 4a),[4] in creating sophisticated patterns. This is where Josh came in. First, since the final color of the system could be chosen with either the wavelength or exposure time of incident light, he had to choose which method to use to produce multi-colored images. Ultimately, he decided on using blue light with variable time, as the blue LED was already equipped in the DLP Microscope lined up well with the 442 nm light used in the switching studies. From there, he chose some simple patterns, projected them for different times, and produced our first multi-colored images. Interestingly, by controlling the illumination times of patterned blue LED light, he achieved a resolution of 76 μm (Fig. 4b), enabling the creation of highly detailed images.

Fig. 4. DLP image imprinting and resolution determination. a LC cell mounted on DLP projector for image imprinting e.g. “The Scream”. b Microscopic images of the line patterns obtained by projecting the LC cell with patterned blue light (0.475 W cm^-2) for a 10 s exposure time. Each of the above images (total 6) consists of 12 lines giving n = 72 to give a minimum resolution of 76.07 µm with StDev of 18.39 µm.

Once it was established that we could make some simple shapes, the hard part was finding images to use for the paper and photographing the end results. While we could get some truly phenomenal results with the system with resolutions down to the tens of microns, no matter how good the images looked in person, the reflected colors of the CLCs were hard to photograph. Although Josh tried using higher end cameras on tripods and a variety of lighting setups, at the end of the day, the only real solution to capturing images was to put the cell on a black piece of paper and spend several minutes moving the camera around to find a good angle. These photography restrictions, combined with Josh’s personal artistic proclivities, lead to our selection of the impressionist and proto-impressionist pieces “Starry Night” and “The Scream,” respectively. Although we aren’t exactly artists (notice our publication in Nat. Chem. and not The Burlington Magazine!), we converted said paintings into virtual photomasks via discerning which details qualified for the “foreground” and “background” of the painting and almost completely repainted them digitally. While every painting we tried was patterned successfully, including pieces by Monet, Mondrian, and Goncharova, the camera’s eye favored the impressionist pieces’ bold curls and low density of detail.

This work has not only contributed to our understanding of chirality transfer in CLCs but also explored the unique optical properties of CLCs in a creative and dynamic way. We both are excited to share these findings with the scientific community through Nature Chemistry.

References

1. Shao, B. & Aprahamian, I. Hydrazones as new molecular tools. Chem 6, 2162-2173 (2020).

2. Krishna, B. H. & Quan, L. Light-Driven Liquid Crystalline Materials: From Photo-Induced Phase Transitions and Property Modulations to Applications. Chem. Rev. 116, 15089−15166 (2016).

3. Long, T. M. & Swager, T. M. Minimization of free volume: Alignment of triptycenes in liquid crystals and stretched polymers. Adv. Mater. 13, 601-604 (2001).

4. Haris, U., Plank, J. T., Li, B., Page, Z. A. & Lippert, A. R. Visible light chemical micropatterning using a digital light processing fluorescence microscope. ACS Central Science 8, 67-76 (2022).