Materials are the building blocks of every man-made thing, and the way that glass, metals, ceramics, polymers, adhesives, composites and other materials are engineered is being transformed. Consumers will never see it – but it will touch their lives.
Consider glass, invented in Mesopotamia in 4,000 BC. Despite 6,000 years of experience with the substance, full knowledge of how to create different types of glass for specific purposes has remained elusive. Unlike metals or ceramics, which have consistent patterns of atoms, different types of glass have atomic variations that make their properties difficult to predict. For centuries, scientists could only invent new types of glass by combining silicon with other substances, heating it to great temperatures and testing the result.
“That was what we called ‘cook and look’,” said John Mauro, research manager of the glass research group at Corning Incorporated in Corning, New York (USA). “The difference today is that we can use scientific principles.”
Companies such as Corning are revolutionizing age-old materials by combining advances in physical knowledge with software and massive computing power to understand and manipulate materials at the atomic and sub-atomic level. In the case of Corning’s trademarked Gorilla Glass, a break-resistant glass that protects the screens of mobile telephones and tablets worldwide, scientists built a computer model of how the atoms were connected to each other to determine how the formula could be tweaked. “The key breakthrough was to allow the glass structure to dynamically adjust in response to external stress,” Mauro said. Thanks to predictive computer modeling, “we are seeing a new renaissance in glass science and glass technology.”
But glass isn’t the only material getting a makeover. Consumer electronics, which traditionally have been stiff to protect internal components, are becoming wearable thanks to polymer research emerging from Akron University and Case Western University in Ohio (USA).
“If they want a certain kind of molecule, the fastest way to get it is through computational design and experimentation.”SANJAY MEHTA
Manager of Computational Modeling, Air Products
Polymers – large molecules consisting of many repeating sub-units – are the building blocks of many electronic components. About 70 companies in northeastern Ohio are using the universities’ research to create “flexible” electronics. Researchers “use computers and software to model a lot of the interactions because it’s about functionalizing the polymers,” said Tim Fahey, director of cluster acceleration for flexible electronics at Nortech, an economic development organization based in Cleveland. “It’s about how you tailor their structure at the nano level and make them do cool things.”
Some of those ‘cool things’ are evident in recently released wearable electronics, including electronics woven directly into fabrics. “We’re entering the next wave of electronics when you move from mobile to wearable,” Fahey said. “When you do that, things have to be soft because they will be next to the skin.”
Electronic devices also are becoming smaller and more powerful, since major semiconductor makers asked Air Products of Allentown, Pennsylvania (USA) to devise new chemicals that could deposit sub-nanometer circuits on chips, plus chemicals that could clean the smaller circuits.
Sanjay Mehta, who manages Air Products’ computational modeling center, worked with his team to model the new molecules. The molecules worked first in computer simulations and then in the real world. “If they want a certain kind of molecule, the fastest way to get it is through computational design and experimentation,” Mehta said. “The efficiency is just immense. Instead of doing experiments in the lab, we are doing experiments in the computer.”
The US has the lead in commercializing new materials, Mehta said, because of “high-throughput” computing power created at the request of the US government. “As far as industrial adaptation and making these new molecules work, US companies are definitely in the lead,” Mehta said. Computer speeds for such complex computations are measured by the number of floating-point operations per second (FLOPS). Air Products is using server farms that operate in terms of teraflops, or 10 to the power of 12 (1 followed by 12 zeroes) – equal to 1 trillion operations per second.
With help from predictive computer modeling the materials industry, once a backwater, is bursting with innovation – and the world’s consumers will reap the benefits. ◆