Comment of the Day: Theoretical Materials Edition

Image for article titled Comment of the Day: Theoretical Materials Edition

In today's comments, we looked at some behavioral glitches in the animal kingdom, read physics journals with the hope of seeing the TARDIS in there (and it was!), and learned some fascinating new details about the world of theoretical materials science.


We were joined for a Q&A today by Craig J. Fennie, a theoretical materials scientist at Cornell who recently won a MacArthur Genius grant. In between answering the age old question new nanofab facility or backyard synchrotron ("synchrotron hands down!") and sharing some sample questions with which to impress the thermoelectric engineers in our lives, Fennie answered a question from commenter AverageDrafter on whether it was theoretically possible to create atoms with customizable electron configurations:

If I literally answer your question I would say, 'not really,' the energy to do such a thing are much much greater than could be used in a practical material (this is the realm of nuclear physics). But this is a great opportunity to mention the beauty of materials physics and chemistry.

Materials are made of atoms, which we kind of understand, but when we put these atoms together to make a material they take on a life of their own, which can't be anticipated with understanding the atom. (Think about a classical example, rubbernecking, we all know how to describe how a car works and can write an algorithm to drive a single car on a road today. But put many many cars on the road, new rules of driving emerge, one would never guess based on how a car works that if there is an accident on the other side of the road that that would affect how you drive). This is what we call "emergent properties". In a material, one known emergent property is something just like you asked about and is the basis of much research today.

Image: Vichy Deal / Shutterstock



I don't think rubbernecking is a good example of an emergent property. It doesn't come from the cars but rather from the people driving them. Of course you can't predict what traffic will do based on the designs of the vehicles. To predict that you have to understand human psychology.

Take those exact same vehicles and put space aliens behind the wheel (assuming the aliens are able to work the controls) and the behavior of traffic could be entirely different.

Plus, the whole "can't be predicted from knowledge of the components" thing could just reflect the limits of our imagination and/or analytic techniques. "We can't figure it out" rather than "it can't be figured out".