This figure shows the crystal structure of a monolayer of aresenic sulfide, which has special properties for making non-volatile memory devices and piezoelectric sensors. Credit: Center for Computational Materials

The Most Flexible 2D Material Ever Found

Researchers from UT Austin’s Center for Computational Materials have used supercomputers to discover 2D materials with unique properties, bringing next generation flexible electronic devices one step closer.

The scientists from the Oden Institute for Computational Science and Engineering have found a new two-dimensional (2D) material that exhibits ferroelectric, piezoelectric properties. These atomically-thin materials are ferroelectric because of their capacity for switchable, spontaneous electric polarization; and piezoelectric because that electric current is the result of applying pressure and heat. They are also the most flexible 2D materials to be discovered.

The group of crystal structures, known as arsenic chalcogenides, are two dimensional. In the world of materials science, size matters. The super thin structure makes them ideal for use in the miniaturization of next generation flexible electronic devices. These super-thin materials are all around us, but were only discovered in recent years.

“The advantage of 2D ferroelectric materials is that they can be made atomically thin, meaning they are measured at the sub-nanometer scale,” said Weiwei Gao, co-author of the study and a postdoctoral fellow in the Center for Computational Materials.

Working alongside Center Director and UT Physics Professor, James R. Chelikowsky, the Oden Institute researchers revealed their findings in a paper entitled, Prediction of Intrinsic Ferroelectricity and Large Piezoelectricity in Monolayer Arsenic Chalcogenides, published in a recent edition of the journal Nano Letters.

Materials scientists are aware of several ferroelectric materials in the three-dimensional (3D) world. Less is known about atomically-thin two-dimensional materials with similar properties.

Graphene is considered the world's first known 2D material. In 2004, scientists peeled flakes of the material from 3D bulk graphite — found in pencil leads, lubricants, and tennis rackets — using sticky tape. This material has the highest known thermal and electrical conductivity. It is also stronger than steel, but is light, flexible, and transparent.

Most people would imagine the lab setting for a materials scientist being filled with test tubes, petri dishes, and devices for controlling temperature, pressure, exposure to light, and so on. However, computational scientists bypass the physical experimentation aspect of their role. Instead, they rely upon complex mathematical tools powered by supercomputers to simulate experiments with different compounds. They can accurately identify and record the properties of new materials using the computational model alone.

This is how Gao and Chelikowsky conducted their research on arsenic chalcogenides. And, the ferroelectric and piezoelectric properties they found make this discovery potentially very significant.

“Two-dimensional materials with ferroelectricity have applications in state-of-the-art memory devices, piezoelectric sensors, and nonlinear optical devices,” said Gao. In addition, their unmatched flexibility means they could be used in soft polymers or plastics opening up economically viable options for wearable devices and other flexible electronics.

“There are a lot of examples of non-centrosymmetric 2D materials showing spontaneous electric polarization” Gao added. “But not all of them are switchable. In fact, only a few 2D ferroelectric materials have ever been confirmed in experiments.”

The research was conducted using the high-performance computing capabilities at the Texas Advanced Computing Center (TACC), also at UT Austin. The researchers used the Stampede2 supercomputer to perform their calculations. “Stampede2 is an indispensable facility for our research,” Chelikowsky said.

The TACC supercomputer has various “queues” suited for different types of calculation jobs. In particular, the ’development‘ queue is very efficient for debugging and benchmarking methods before production runs.

Chelikowsky concluded: “We performed all our computer simulations on this system. It’s the fastest and most user-friendly supercomputer cluster we’ve used.”

By John Holden

Posted: March 14, 2021