Hacker News Reader

Comment by wolfi1 | Hacker News Reader

about DFT not predicting it: could it be that they used a non-relativistic model?

Given the energy scales involved, it's more likely that the issue lies with the approximations DFT makes to quantum mechanics, rather than the approximations quantum mechanics makes to quantum field theory.

I agree. There are some important relativistic correction for heavy atoms like Gold where the inner electrons are "rotating" too fast near the nuclei, but for Boron it's not necessary to do that correction.

but, don't you need it already for C60?

As far as I know, no relativistic corrections in C60. They are "slow moving" electrons, you can get away with non-relativistic QM. So I guess it's not necessary to make relativistic corrections in B80.

It's very different in a heavy atom like Gold. The charge of the nuclei is x10 of the charge of the Carbon nuclei, so the inner electrons are x10 closer and the "speed" is x10 faster. [1] I don't remember the numbers, but some google search claims 2.7%c for Carbon and 58%c for Gold. You definitively can ignore the relativistic corrections at 2.7%c, there is a square somewhere in the formula and the corrections is like ~(1-(2.7/100)^2).

As the G...GP says, the problem are the weird parts of Quantum Mechanics that are ignored to get a fast method to get an approximation in DFT.

[1] There are some technical details to define "speed", but just squint your eyes and ignore the cries of physicist.

Reply on news.ycombinator.com