Assessment of density functional methods with correct asymptotic behavior.

Area of Science:

• Computational chemistry
• Quantum chemistry
• Theoretical chemistry

Background:

• Long-range corrected (LC) hybrid functionals and asymptotically corrected (AC) model potentials are density functional methods designed for accurate asymptotic behavior.
• These methods are crucial for molecular properties sensitive to the exchange-correlation potential's asymptote, like highest occupied molecular orbital energies and Rydberg excitation energies.

Purpose of the Study:

• To comprehensively compare the performance of LC hybrid functionals and AC model potentials against other density functional schemes.
• To evaluate these methods across a broad spectrum of chemical applications, including asymptote problems, self-interaction errors, energy gaps, and charge transfer.

Main Methods:

• Investigated the performance of LC hybrid functionals and AC model potentials.
• Expanded existing EA115 and FG115 databases to include 16 additional molecules.
• Developed a new AE113 database for atomization energies of 113 molecules.

Main Results:

• LC hybrid functionals consistently demonstrated superior performance compared to AC model potential schemes.
• The study utilized and expanded established molecular databases (EA115, FG115) and introduced a new one (AE113).

Conclusions:

• LC hybrid functionals are more accurate than AC model potentials for a wide range of chemical applications.
• The developed and expanded databases provide valuable resources for benchmarking and improving density functional methods.