Scientists at USC Viterbi School of Engineering have developed a machine learning model using which they can study how light can be used to control the behavior of materials. Their experiments investigate how light polarises materials like Lead Titanate into a Vortex-shaped polarisation pattern that improves its energy efficiency.
Humans have always explored the properties of materials by subjecting them to heat, pressure and friction. That’s how pre-historic humans discovered fire and created glass by heating quartz. As material science evolved, we started using electricity, magnetic fields and light to control materials.
Particularly, using light to control materials opens up interesting possibilities in modern electronics for the creation of next-generation gadgets that are more energy efficient and perform better. This is because light interacts with the materials on such extraordinarily rapid time scales that it is able to uncover features of the material that heating or stretching cannot. These features are typically referred to as hidden “quantum phases”, since light interacts with the material at a quantum-mechanical or atomic level. This, however, makes modelling incredibly challenging, limiting our ability to find and design next-generation materials.
The new ML framework developed by USC Viterbi researchers can be used to study how light can control materials at a scale that’s never seen before. By leveraging the power of ML, they were able perform simulations of light control of materials with over a billion atoms — 10 million times greater than conventional methods. They experimented on lead titanate, a ferroelectric material that has an inherent electronic polarization. It’s this polarization that makes it ideal for use in sensors, energy storage, and memory.
Lead titanate has recently gained prominence because it allows researchers to create complex vortex-like patterns in its electronic polarization. These vortices have been discovered to be extremely stable and efficient in these materials, which is why they are being researched for next-generation energy storage and memory applications. The researchers at USC Viterbi wanted to see if these complicated patterns could be manipulated by light.
“We wanted to view these large-scale structures with highly accurate simulation methods that use things like quantum mechanics,” said Thomas Linker, PhD student. “But that’s really difficult and very expensive, so we developed a multi-scaled framework where we train a machine learning model to learn a simpler representation of the light matter interaction. Thus, we can simulate much faster.”
“Without machine learning, it would have been impossible to design this kind of simulation,” said the Senior Lecturer of Chemical Engineering and Materials Science, Ken-ichi Nomura. “By training the machine learning model to learn how the material behaves in response to a strong laser, we can perform our simulation on supercomputers.”
With their framework, the researchers found a new type of phase that was induced by light matter interaction in lead titanate. “If we shine laser (light), we can create a string pattern in the polarization that is topologically different than the original vortex pattern,” said Nomura.
