BERC-E 4/11: Using theory to predict new energy materials

This week in the BERC Engineering subgroup we took a look at a new paper that goes over some cool examples of when theoretical models successfully predicted new materials as being useful for energy applications before they were experimentally synthesized and measured. Some cool examples from the list were things like new lithium-ion battery materials, hydrogen production and storage materials, solar cell materials and new thermoelectric materials for waste heat-to-energy conversion.

The work horse of the theoretical technique used here is a method called Density Functional Theory (DFT); an approach that uses fundamental quantum mechanics to compute material properties by using some of the world’s biggest and fastest super-computers. While the calculations involve a lot of scientific assumptions, the amount of information that correlates well with experimental data is fairly striking. Probably the biggest take away for non-theorists to realize is that predicted trends from DFT calculations tend to agree really well with experiments. This fact is important because a lot of people in the scientific community tend to discount theoretical methods when experimental values aren’t exactly what is predicted (i.e. things like band gaps). This isn’t what we should be looking for when trying to discover new materials – it is more important to look at trends that emerge from the computed data.

For example, if I want to figure out how I can change the conductivity of Gallium-(Group V) materials by intentionally adding point defects I might run some DFT calculations which would show me how vacancies play less of an important role as I move down group V on the periodic table and this fact would likely be replicated in experiment. (I should say that this isn’t the only way to do DFT calculations – it is definitely possible to get much more reliable experimental numbers by using more challenging assumptions. These just require a longer amount of time needed to complete the calculations)

This approach of screening large amounts of DFT calculations to search for new material candidates for energy applications has been growing in popularity in the past 10 years – and is likely to keep growing as computational power continues to increase. Organizations like the Materials Project have been leading the effort by running DFT jobs on all of the known inorganic materials and then making the results openly available for anyone to look at and survey (it just requires an email address to sign in and start playing around with the data). Many research groups have already been leveraging this large data set to search for new materials and this trend is likely to continue.