ACCDB: A collection of chemistry databases for broad computational purposes

Abstract

The importance of databases of reliable and accurate data in chemistry has substantially increased in the past two decades. Their main usage is to parametrize electronic structure theory methods, and to assess their capabilities and accuracy for a broad set of chemical problems. The collection we present here—ACCDB—includes data from 16 different research groups, for a total of 44,931 unique reference data points, all at a level of theory significantly higher than density functional theory, and covering most of the periodic table. It is composed of five databases taken from literature (GMTKN, MGCDB84, Minnesota2015, DP284, and W4‐17), two newly developed reaction energy databases (W4‐17‐RE and MN‐RE), and a new collection of databases containing transition metals. A set of expandable software tools for its manipulation is also presented here for the first time, as well as a case study where ACCDB is used for benchmarking commercial CPUs for chemistry calculations. © 2018 Wiley Periodicals, Inc.

Nonempirical (double‐hybrid) density functionals applied to atomic excitation energies: A systematic basis set investigation

“…Excited‐state studies continue to be a topic of the most interest, not only because the underlying theory intrinsically presents challenges for its implementation in common codes but also due to the interplay and large number of factors affecting the final results in molecular and real systems. Whereas ground‐state properties still receive much more attention in the ongoing development of density functional theory (DFT) (eg, for built‐in datasets and the associated benchmarking of density functional approximations) the applications to atoms are recently emerging as an alternative for the benchmarking of DFT for excited states too . This is facilitated by the reasonable computational cost of atomic calculations compared to more complex systems, together with the lack of geometry‐induced and environmental issues often difficulting the adequate comparison between various theoretical methods.…”

Double hybrid DFT calculations with Slater type orbitals

“…While there is considerable overlap between MGCDB84 and GMTKN55, the reference data in the latter one is generally newer and often more accurate. [ 226 ] In example, MGCDB84 contains much data from GMTKN30, [ 162 ] which has been updated in GMTKN55. [ 57,226 ] Consequently, for test sets contained in both databases, we always compared our results to the reference values presented in the latter one and in total we selected 45 out of the 55 subsets in GMTKN55 (two of them, (MCONF, BUT14DIOL) with reduced size [ 265 ] ) to benchmark all DFAs in Table 1 and included three more into our subcategory of large molecules.…”

“…For all test sets contained in GMTKN55, the structures and reference energies as available on the dedicated website [ 286 ] have been used. For all other datasets we employed the structures and reference energies from the ACCDB [ 226 ] database.…”

CLB18: A New Structural Database with Unusual Carbon–Carbon Long Bonds

“…We did not include the GFN-xTB1 and GFN-xTB2 methods because their application to very small molecules such as those in BH76 is not recommended, and also because a reliable transition-state search algorithm is not implemented in the programs that we have available for these calculations. While several barrier heights databases are included in major database collections, 27,92,[106][107][108][109][110] the geometries of the transition structures of most of them are reported at the DFT level and cannot be used for our current purpose (see Table S8 in the Supporting Information). The BH76 database includes very small molecules exclusively, with transition states optimized at the QCISD/MG3 level of theory.…”