Performance of Density Functionals and Semiempirical 3c Methods for Small Gold-Thiolate Clusters.

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Increasing computational studies on gold thiolate clusters necessitate performance evaluations of theoretical methods.
• Density Functional Approximations (DFAs) and 3-part corrected methods (3c-methods) are popular choices for such calculations.

Purpose of the Study:

• To compare the accuracy and efficiency of various DFAs and 3c-methods for gold thiolate clusters (Au(SCH3)n, n=1-3).
• To evaluate performance in geometry optimization and energy calculations using established reference methods.

Main Methods:

• Utilized the AuSR18 dataset comprising 18 isomers of gold thiolate clusters.
• Geometry optimization compared against RI-SCS-MP2.
• Energy evaluation benchmarked against DLPNO-CCSD(T).

Main Results:

• TPSS demonstrated accuracy, while mPWPW offered a balance of speed and accuracy.
• Hybrid range-separated DFAs proved superior for relative cluster energies; CAM-B3LYP excelled, B3LYP performed poorly.
• Some methods struggled to locate various minima, impacting versatility; 3c-methods were fast but less accurate for relative stability.

Conclusions:

• TPSS and specific hybrid range-separated DFAs are recommended for accurate gold thiolate cluster studies.
• Method selection should consider both accuracy in geometry/energetics and versatility in locating diverse isomers.
• CAM-B3LYP is a strong candidate for relative energy calculations, while LC-BLYP offers balance but lacks diversity.