Talking about density functional theory

Thread Content

Talking about density functional theory 1) People always think that only crystals are related to density functional theory. In fact, without density functional theory, HF (or SCF or MO) would not have become feasible for computational use. For example, the basis set of the Gaussian function is an embodiment of density functional theory. Therefore, the variable of the Gaussian function is only three-dimensional ； 2) HF (or SCF or MO) early adopted the core ideas of density functional theory. However, it lags behind crystals in terms of computation, thanks to the large amount of post-HF processing. Thanks to the continuous fitting of the exchange-correlation function (Exc), density functional theory is widely applied to HF (or SCF or MO) ； 3) Density functional theory has evolved to TDDFT, which enables the calculation of excited states; as a result, the advantages of HF (or SCF or MO) have been **reduced**. It is currently recognized that the DFT method has more potential than post-HF ； 4) It is easy to use DFT for HF (or SCF or MO), because once the wave function is obtained, the density function can be derived naturally based on the probability assumptions of that wave function. The density function and the wave function are of the same type: they are both expressions (with many terms), only with fewer variables. If the wave function is converted into a density function, spin exchange will no longer exist. Therefore, although the Gaussian function uses a Slater function, it does not experience spin exchange repetition (since it has been transformed into a density function using DFT). 5) The density function and the wave function are at the same level. Therefore, their algorithms are different. But they can be used interchangeably; it’s just a matter of precision. I ask all the teachers to offer their criticism and suggestions.