Whereas experimental investigation of materials is performed at finite, usually room temperauter, and a special effort is required to bring the experimental conditions close to absolute zero, in theoretical investigation of materials the situation is exactly the opposite: most calculations are performed at T = 0, and the temperature dependence is introduced as a correction, if it is not ignored at all. Particularly, when exploring the electronic structure of a material within density functional theory (DFT), usually one uses exchange-correlation (xc) approximations, which have been originally derived with zero temperature in mind. This approach is justified as long as the temperature is relatively low, compared to the electronic excitation energies. Whereas this is often the case, there definitely exist situations where this is not true.

 One such situation happens in a solid at high temperatures, up to melting, with T ~ 10²...10⁴ K, where the energy differences between competing crystallographic phases and the liquid phase are comparable to the order of magnitude of electronic excitations in the solid. Additional examples where temepreature-dependent treatment in DFT is important are matter in the cores of some planets, including the Earth, and chemical processes in interstellar gas clouds heated by X-ray radiation.

Therefore, accurate assessment of the temperature dependence in DFT calculations is relevant in many scenarios. This requires development of special approximations to exchange and correlation energies that take temperature dependence into account explicitly. Development and application of such approximations will be of interest to our group in the near future.  

...back to Research