A vivid rainbow of data from liquid crystals in silico

The liquid crystals that regulate the brightness in computer monitors, TVs, and countless other modern devices appear as translucent oil to the naked eye, but can show vibrant colors under polarized light. The fine optical textures seen under the microscope enable scientists to study the precise molecular details of liquid crystals.

However, computational studies, which are essential to understanding the complex physics of this phase of matter, have been lagging in predicting and gaining information from these rich color textures, relying on simple black-and-white images of liquid crystals.

Now, researchers at the University of Chicago Pritzker School of Molecular Engineering (PME) have developed a new software package that brings color to microscope images of liquid crystal structures, more fully explaining how different wavelengths of light interact with the liquid crystals under realistic imaging conditions.

The suite of tools, freely available to researchers online, will allow better comparison of theoretical models with experimental data on liquid crystals, and help with the development of new types of liquid crystals that can act as chemical or temperature sensors.

“We’re very excited to see how other researchers are going to use these tools to advance the science of liquid crystals,” said Juan de Pablo, Liew Family Professor of Molecular Engineering and senior author of the new work, published in Chemistry of Materials.

Liquid crystals are the basis for many emerging smart materials because their internal organization can change in response to the surrounding environment. One common example is liquid crystal displays (LCDs), where a change in electrical voltage can make liquid crystals quickly re-orient, allowing the light to either pass through or get blocked off, like opening or closing a door.

Liquid crystals are not only useful for display technologies but also serve as emerging materials for liquid crystal-based sensors to detect very small amounts of chemicals, such as microplastics and toxic chemicals in water.

Such advanced applications typically rely on observations with polarized optical microscopy, where the colors and patterns change according to external cues such as chemical or electric signals. Computer simulations provide physics insight into how liquid crystals pick up such signals and help scientists improve the material design rationally. But frequently, computer models rely on simple black-and-white images calculated for a single wavelength, making the comparisons between experiments and simulations difficult.

“You would see this beautiful color image during your experiment, and then the resulting simulation data would be a black-and-white or grayscale image that only partially captured what you saw,” explained PME postdoctoral researcher Viviana Palacio-Betancur, a co-first author of the new work.

Palacio-Betancur and colleagues, including co-first author and PME graduate student Chuqiao Elise Chen, integrated multiple fields of knowledge about the liquid crystals, optical microscopy, and color science to build their new approach.

“We were able to bring together some new areas of physics to finally reproduce accurate colors in our images of liquid crystals,” said Chen. “A big advantage of our package is also that it is very fast, which opens the door to new kinds of database analysis and machine learning efforts for liquid crystal design.”

When the team applied their software to generate images of liquid crystals under many conditions, they found that their resulting data closely matched experimental data from experiments the de Pablo group as well as the lab of Sanaz Sadati at the University of South Carolina. The researchers envision that this tool can be used by experimental researchers to gain information from the microscopy images from different liquid crystals and their dynamic changes.

“These images give us a new ability to make connections between what you do in simulations and what you see in experiments,” said PME postdoctoral researcher Pablo Zubieta-Rico. “That also makes it easier to push on the inverse design side of things; you can play with more things in simulations to make new liquid crystals with desired properties.”

The team plans to continue improving their software so that it can be used for more experimental settings and types of liquid crystals.

Citation: “LCPOM: Precise Reconstruction of Polarized Optical Microscopy Images of Liquid Crystals,” Chen et al, Chemistry of Materials, March 28, 2024. DOI: 10.1021/acs.chemmater.3c02425

Funding: This work was supported by the University of Chicago Materials Research Science and Engineering Center, the National Science Foundation, and a Fulbright PhD student scholarship.