MVO-10: A Gas-Phase Oxide Benchmark for Localization/Delocalization in Mixed-Valence Systems.

Area of Science:

• Computational chemistry
• Inorganic chemistry
• Quantum chemistry

Background:

• Mixed-valence (MV) oxo complexes present challenges in accurately describing charge and spin density localization versus delocalization.
• Environmental effects can complicate computational studies of MV systems, necessitating simpler gas-phase models.

Purpose of the Study:

• To screen various density-functional methods for their accuracy in modeling charge and spin density behavior in gas-phase MV oxo complexes.
• To provide benchmark coupled-cluster energies for evaluating the performance of different electronic structure methods.

Main Methods:

• Utilized benchmark coupled-cluster (CC) calculations (CCSDT(Q)/CBS and CCSD(T)/CBS) to establish reliable energy data.
• Screened a range of density-functional approximations, including generalized gradient approximations, global hybrids, range-separated hybrids, and local hybrids.
• Incorporated experimental data for V4O10- to validate computational findings for delocalized systems.

Main Results:

• No single density-functional method accurately describes both highly localized (e.g., Al2O4-) and highly delocalized (e.g., V4O10-) MV oxo complexes.
• Methods with large exact-exchange admixtures are better for localized systems, while lower admixtures are preferred for delocalized d1d0 systems.
• A local hybrid (Lh-SVWN), a global hybrid (MN15), and a range-separated hybrid (ωB97X-D) demonstrated the best overall performance.

Conclusions:

• The choice of electronic structure method is critical for accurately predicting charge and spin density behavior in MV oxo complexes.
• The screened methods offer a basis for selecting appropriate computational tools for future studies of MV systems.
• Vibrational data are provided for experimental comparison, aiding further validation of computational models.