Improved third-order Møller-Plesset perturbation theory.

Area of Science:

• Quantum Chemistry
• Theoretical Chemistry
• Computational Chemistry

Background:

• Standard Møller-Plesset (MP) perturbation theories often exhibit limitations in accurately describing electron correlation.
• Previous work has focused on improving MP2 by scaling spin components, necessitating further development for higher-order methods.

Purpose of the Study:

• To develop a modified third-order Møller-Plesset perturbation theory incorporating spin-component scaling.
• To assess the performance of the new spin-component-scaled third-order Møller-Plesset (SCS-MP3) method against established quantum chemical approaches.

Main Methods:

• Developed a modified third-order Møller-Plesset perturbation theory by partitioning total correlation energy into parallel- and antiparallel-spin electron pairs.
• Introduced the spin-component-scaled (SCS) approximation to the MP3 method, termed SCS-MP3.
• Evaluated SCS-MP3 using benchmark datasets including isogyric reaction energies, atomization energies, and stretched geometries.

Main Results:

• SCS-MP3 significantly outperforms standard MP2 and MP3 methods and is competitive with or superior to the more computationally expensive QCISD(T) method.
• The SCS correction uniformly improves results across various chemical problems, unlike the standard MP series.
• Dramatic enhancements were observed for atomization energies, stretched geometries, and transition metal compounds, with results often matching QCISD(T) accuracy.

Conclusions:

• SCS-MP3 is a robust and accurate quantum chemical method for general use, offering significant improvements over existing methods.
• The method demonstrates remarkable uniformity in improving results for diverse correlation problems, highlighting its broad applicability.
• SCS-MP3 shows particular promise for challenging systems like transition metal compounds and complex molecular structures.