Artificial intelligence algorithms developed to find engineering materials
Researchers at Stanford University are leveraging machine learning to help find engineering materials more efficiently.
Imagine being an engineer who has solved an age-old challenge, but there’s one problem: You lack the perfect material for the task and you’re not sure such a material exists. Even if it did, the realm of possible candidates could number in the thousands or millions.
Stanford’s Evan Reed, associate professor of materials science and engineering, and two graduate students, Austin Sendek and Gowoon Cheon, have found a solution. The trio have developed artificial intelligence algorithms that can help engineers find the perfect material for important applications.
“These algorithms are like search engines for materials,” Reed said.
In three recent papers Reed, Sendek, and Cheon describe how they use machine learning–computational systems that teach themselves–to analyze the physical and chemical properties of thousands of materials. The machine learning systems create algorithms that comb through these materials to find those that mathematically match the requirements for the task at hand. Experimental scientists can use these computer-generated hunches to create and test materials under real-world conditions. This is much faster than past hit-or-miss methods.
In Reed’s paper, he explored a category of materials known as metal-organic frameworks (MOFs). MOFs are crystals that are structurally strong, porous at the nanoscale and inexpensive to make. So far they have been used to capture hydrogen and carbon gases within the pores of solid materials. But researchers have become interested in using MOFs for fuel cells and thermoelectric devices, and these new uses would require electrical conductivity. Unfortunately, very few known MOFs are electrically conductive.
Reed screened almost 3,000 metal-organic frameworks to winnow to the six best candidates. His team confirmed the electrical conductivity of these materials through detailed calculations that used standard computational methods. This offered experimentalists a new palette of MOF possibilities.
Machine learning for finding solid-state batteries
Sendek’s paper described efforts to identify safer lithium-based solid materials for solid-state lithium-ion (Li-Ion) batteries. This is an alternative to commercial lithium-ion batteries with liquid electrolytes, which have caused some smartphone and electric car fires.
For decades, the search for solid Li-Ion electrolyte materials proceeded mostly on a trial-and-error basis. Sendek used a machine learning-based approach to survey 12,000 possible candidates and identify 21 materials that were likely to exhibit a number of important properties for solid electrolytes, including high lithium-ion conductivity. After a year of evaluation, he found 10 of the 21 materials exhibited the conductivity values the machine learning model predicted in a simulation. This was a significantly higher success rate than the intuition-guided guess-and-check searches of the past.
In fact, Sendek turned the search into a bit of a competition. He pitted his algorithm against a team of six doctoral students who performed a similar hunt for fast lithium-ion conductors.
His model proved twice as accurate and over a thousand times faster than the human team.
“I liken our approach to facial recognition software for materials,” Sendek said.
Predicting the existence of 2-D materials
Cheon’s paper described using machine learning to predict the existence of two-dimensional materials – substances so thin their thickness is measured in layers of atoms. Graphene, an atomic layer of graphite, was the first 2-D material discovered. The scientists who found it in 2004 shared the 2010 Nobel Prize in physics for revealing its amazing properties. Electronics researchers hope to use 2-D materials to create atomically thin circuits and devices. Only a few dozen compounds have been discovered, however.
Cheon used a physics-based machine learning model and studied 16 million chemical compounds to predict which might be exfoliated into 2-D layers. Her model added more than a thousand candidates to the field. Most have never been synthesized, providing experimentalists with a roadmap in the search for ultrathin materials, Cheon said.
Artificial intelligence and machine learning benefits
Reed said the process is useful because few materials have ever been measured, probed, and tested, much less cataloged in a database. Machine learning can expedite the discovery process.
The process is complex but the logic is simple: Molecules and crystalline structures have physical and chemical properties. Computational systems are adept at mathematical analysis and can express these properties.
“Even though we lack lab test data on most materials, we can still predict something about their likely properties computationally,” Reed said.
– Edited by CFE Media. See more Control Engineering info management stories.