Compute me the perfect alloy

IOM3
,
1 Feb 2014

Do you need a less costly substitute for palladium? Do you want to blend previously immiscible compounds? If so, software developed at Duke University, USA, could help you. Eoin Redahan reports.

Stefano Curtarolo is searching for cheapium – a byword for cheap materials that mimic the attributes of their expensive counterparts. To do this, the Duke University researcher, based in North Carolina, USA, has used knowledge of how atoms interact to create a computer algorithm that trawls through data to generate models of chemical compounds. So far, the programme has sifted through 40,000 calculations to identify 37 new binary alloys in the platinum group – alloys that could be used to provide cheaper catalytic convertors in cars and synthetic rubbers.

Curtarolo explains, ‘Researchers mix different elements. They melt them, cool them and see how they are arranged. What if they were to search the spectrum? What if they got all the possible mixtures and cooked them, cooled them and put them on X-rays to check their stabilities? By having this information already, they can rapidly accelerate the process.’

According to Curtarolo, the technology could also be used to modify the composition of alloys that never quite worked. ‘If we make them using nanograins, we can make them stable. Instead of making an alloy with big grains, we make it with small grains. That way, we can make a nanograin alloy with completely different properties. We’re looking to publish this work in a couple of years.’ He adds that the research could be particularly useful for high-temperature alloys. ‘There is a class of tungsten alloys that does not mix in a big batch. But if you make them as a nanopowder, you can stabilise them. These alloys have very good impact and temperature resistance.’

For now, Curtarolo is honing the metal systems code that has eaten into his time since 1999. His programme has received interest from the steel and space exploration industries and has been mooted for use in everything from palladium substitutes to additive manufacturing on Mars. However, the programme is not a commercial venture. Rather, it is being developed as a time-saving tool for scientists. Curtarolo says, ‘In science, the biggest bottleneck is the availability of good researchers. This programme frees researchers from running calculations. The calculation is now done automatically so we can focus on the scientific part.’