Density-functional tight-binding (DFTB) [1] is an approximation to density functional theory (DFT) and as such inherits a number of its strengths and almost all of its weaknesses. In this presentation, starting with recent DFTB successes [2], I will illustrate the adverse effects of self-interaction error, the DFTB two-center repulsive potential, and the minimum basis set treatment on graphene fluorination [3], oxide ion migration in metal oxides [4], and on stabilities of molecules composed of electronegative elements, respectively. Possible solutions to all three problems will be discussed.
[1] a) Seifert, G. et al., WIREs Comput. Mol. Sci. 2012, 2, 456-465; b) Gauss, M. et al., WIREs Comput. Mol. Sci. 2014, 4, 49-64.
[2] a) Gruden, M. et al., J. Comput. Chem. 2017, DOI: 10.1002/jcc.24866, b) Lee, K. H. et al., unpublished
[3] Hutama, A. S. et al., J. Phys. Chem. C 2017, 121, 14888-14898.
[4] Hutama, A. S. et al., AMTC Letters 2016, 5, 78-79.