The atoms-in-molecules (AIM) theory has been proposed as a method to understand chemical stability through stationary properties of the electron density. To assess the applicability of this method for establishing such correlations with various performance and vulnerability properties of energetic materials, we calculated the Kohn-Sham density functional theory (KS-DFT) wavefunctions and their subsequent AIM data for representative materials, including hexanitrobenzene (HNB), pentaerythritol tetranitrate (PETN), pentanitroaniline (PNA), 1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX), ethylenedinitramine (EDNA), 1,1-diamino-2,2-dinitroethylene (FOX-7), 3-nitro-1,2,4-triazol-5-one (NTO), nitroguanidine (NQ), 1,3,5-triamino-2,4,6-trinitrobenzene (TATB), and the TATB dimer using the B3LYP, PBE, and PW91 potentials as well as Hartree-Fock (HF). For the HNB and HMX molecules and the TATB dimer, the number of critical points in the low-density regions of the density gradient vector field varied, sometimes dramatically, with basis set and potential even at their individually optimized geometries. Adding ghost atoms in the low-density regions also affected the existence of critical points. The variation was seen in results generated with three separate AIM software packages, AIMPAC, AIMAll, and InteGriTy. This inconsistency implies that KS-DFT wave-functions can have significant variation in the topology of the electron density to such an extent that these calculations cannot be used to justify the existence or absence of low-density critical points. Therefore, predictions of the stability of a molecule based solely on properties of low-density bond critical points generated from a single DFT calculation are questionable.
Read full abstract