Accurate and Numerically Efficient r²SCAN Meta-Generalized Gradient Approximation.

Area of Science:

• Computational chemistry
• Materials science
• Condensed matter physics

Background:

• The SCAN (Strongly Constrained and Appropriately Normed) functional is a meta-generalized gradient approximation (mGGA) known for its accuracy.
• The rSCAN functional, a regularized version of SCAN, offers improved numerical performance but breaks exact constraints.
• Restoring these constraints is crucial for reliable electronic structure calculations.

Purpose of the Study:

• To develop a new mGGA functional that combines the numerical advantages of rSCAN with the constraint adherence of SCAN.
• To improve the accuracy and transferability of density functional approximations for electronic structure.
• To address the limitations of existing functionals in accurately describing chemical systems.

Main Methods:

• Construction of a new meta-generalized gradient approximation (mGGA) functional.
• Restoration of exact constraint adherence to the rSCAN functional.
• Systematic testing and validation of the new functional's performance.

Main Results:

• The new functional successfully restores exact constraint adherence to rSCAN.
• The developed functional maintains the numerical performance benefits of rSCAN.
• The new functional recovers the transferable accuracy of the original SCAN functional.

Conclusions:

• The newly developed mGGA functional offers a superior balance of numerical stability and theoretical accuracy.
• This advancement provides a more reliable tool for electronic structure calculations in various scientific disciplines.
• The research contributes to the ongoing development of more accurate and robust density functional approximations.