Simple scaling of the empirical London dispersion corrections in DFT geometry optimizations of prototypical nonlinear optical crystals

The quantum treatment of electrons at the Kohn-Sham Density Functional Theory (DFT) level of approximation is used in order to optimize the structure of four organic and organo-metallic crystalline systems, starting from experimentally determined structures, allowing subsequent quantum-level theoretical analysis, the accent here being put on crystals having nonlinear optical (NLO) properties. By employing DFT, one must select an exchange-correlation functional (XCF), that is, a mathematical description of the electron interactions. Among most of the XCFs, a well-known general discrepancy is the lack of London-type interactions description. One of the most popular correction schemes consists of adding a posteriori the contribution of these interactions by using empirical expressions. Here, the approach relies on the simple D* scheme, a variation of the D2 correction originally elaborated by Grimme for molecular clusters, where the D2 correction is scaled to account for the interactions in the crystal solid state. The performance of this simple D* scheme is demonstrated and the most suitable scaling factors for four prototypical organic and organo-metallic NLO crystals are determined for four representative XCFs: zero (no correction) for PBESol, PBESol0, and omega B97X, and 0.7 for B3LYP.