Polymer Stabilized Cholesteric Liquid Crystal Siloxane for Temperature-Responsive Photonic Coatings

Temperature-responsive photonic coatings are appealing for a variety of applications, including smart windows. However, the fabrication of such reflective polymer coatings remains a challenge. In this work, we report the development of a temperature-responsive, infrared-reflective coating consisting of a polymer-stabilized cholesteric liquid crystal siloxane, applied by a simple bar coating method. First, a side-chain liquid crystal oligosiloxane containing acrylate, chiral and mesogenic moieties was successfully synthesized via multiple steps, including preparing precursors, hydrosilylation, deprotection, and esterification reactions. Products of all the steps were fully characterized revealing a chain extension during the deprotection step. Subsequently, the photonic coating was fabricated by bar-coating the cholesteric liquid crystal oligomer on glass, using a mediator liquid crystalline molecule. After the UV-curing and removal of the mediator, a transparent IR reflective polymer-stabilized cholesteric liquid crystal coating was obtained. Notably, this fully cured, partially crosslinked transparent polymer coating retained temperature responsiveness due to the presence of non-reactive liquid-crystal oligosiloxanes. Upon increasing the temperature from room temperature, the polymer-stabilized cholesteric liquid crystal coating showed a continuous blue-shift of the reflection band from 1400 nm to 800 nm, and the shift was fully reversible.


Synthesis of TP-s1
Monomers M1, M2, and M3 were put into a Schlenk flask in a molar ratio of 88 : 6 : 6, for a total of 11 mmol. 0.92 g OMHS was added to the flask, in which the amount of Si-H bonds was approximately 9 mmol. The components in the flask were flushed continuously with argon. 25 mL of anhydrous toluene was injected under argon atmosphere to dissolve the reactants. A catalytic amount of Karstedt's catalyst was then injected to start the reaction. The content of the flask was brought to reflux at 65 o C under argon atmosphere. The reaction was continued until sampling NMR showed no residual Si-H groups at δ=4.7. The finished reaction was cooled, and the product solution was precipitated in toluene/cold methanol (-20 o C) 3 times. After removal of the solvent in vacuum, the resulting solid was dissolved in toluene, and an adequate amount of silica gel metal scavengers (Si-Triamine, SilicaMetS, Silicycle Inc.) was added. This was followed by vigorous stirring overnight to remove the platinum catalyst from the product. The silica gel was removed through filtration and washed with toluene. The eluent was evaporated, yielding a white, viscous solid product. Yield: 3.01 g, 77.6 %. Navg calculated from NMR using a literature method[3]: 6.12.
Further analysis is shown in Session 2.1.

Effect of photoinitiator concentration on the crosslinking completion
We investigated the cross linked fraction of the polymer coating as a function of photo initiator concentration (0.05% -0.5%). These coatings were collected in 0.2 μm PTFE filters, and then filtrated by dichloromethane flushing. Proton NMR spectra were recorded of the eluents after evaporation of the solvent ( Figure S13a). When 0.05 wt% of photoinitiator was used, which corresponds to ~1wt% compared to the total weight of the acrylic mesogens, it was found that acrylate groups were poorly converted, as the acrylate signals were still prominently present in the spectrum. When the photoinitiator concentration was increased, the peaks became less prominent, and at 0.25 wt% (5 wt% compared to the total weight of the acrylate groups), the acrylate signals were no longer observable, suggesting high conversion of the acrylate groups.  Figure S14. Diagram of the fully cured TP coating with reflection center wavelength against temperature during multiple heating and cooling rounds. Duration of each test is different, depending on the time needed for its optical properties to reach equilibrium at one temperature. The methodology of the temperature dependent spectra measurements is described in Materials and Methods session.  Figure S15. Transmission spectra of the fully cured TP coating at room temperature (20°C) right after cooling (Test#14 of Figure S14), and after 48 hours of storage at 20°C. The slight difference in transmittance is probably caused by variations at different measured spots. Figure S16. The reflection band center -temperature relation of the fully cured coating (using 0.5 wt% photoinitiator, red dots), compared to the coating by the same procedure but without the curing step (black squares) and the polymer in an alignment cell (blue squares). The reflection band of the TP coating without curing moved out of spectrometer detection range near room temperature.  Figure S17. Transmission spectra of the fully cured TP coating taken during cooling from the isotropic temperature.

Kinetic behavior of the fully cured TP coatings.
We further analyzed the spectra to better understand the kinetic behavior of the temperature responsiveness. By heating the coating, the visible light transmittance dropped immediately, but slowly increased with time when maintaining the temperature ( Figure S18a). Meanwhile, the reflection band center moved rapidly with temperature during heating ( Figure S18b). These phenomena are due to scattering, which is caused by defects generation during mesogens rearrangement. The kinetics during cooling were different. Visible light transmittance remained high and unchanged with temperature ( Figure S19a), but the reflection band center redshifted slowly ( Figure S19b). Polymer distribution of the completely polymerized TP coatings.
The crosslinked network fraction of the photonic coatings was estimated by filtration experiments. The coating was collected and weighted in a Teflon filter with a pore size of 0.2 μm and then filtrated with dichloromethane. The remaining solid in the filter was the insoluble part of the coating, and was therefore considered to be crosslinked polymers. 4 wt% of the original weight of the fully polymerized coating was left in the filter, while a control experiment of an uncured coating (made without using the photoinitiator and processed without the curing step) yielded 0 wt%. However, this 4 wt% does not represent all the reacted TP. The eluent from the filtration experiment contains a substantial amount of high Mw fraction as shown by GPC chromatography ( Figure S21), with molecular weights as high as 200 kDa. We assume this fraction contains larger macromolecules consisting of multiple linked TP molecules that can pass the filter pores. TP not cured Completely polymerized coating (the eluent from filtration test) Figure S21. GPC spectra showing the molecular weight distributions of the completely polymerized TP coatings, compared to the original TP. Normalized by the intensity maximum of the signals.