Abstract:We provide a rigorous derivation of a class of double-hybrid approximations, combining Hartree-Fock exchange and second-order Moller-Plesset correlation with a semilocal exchange-correlation density functional. These double-hybrid approximations contain only one empirical parameter and use a density-scaled correlation energy functional. Neglecting density scaling leads to a one-parameter version of the standard double-hybrid approximations. We assess the performance of these double-hybrid schemes on representative test sets of atomization energies and reaction barrier heights, and we compare to other hybrid approximations, including range-separated hybrids. Our best one-parameter double-hybrid approximation, called 1DH-BLYP, roughly reproduces the two parameters of the standard B2-PLYP or B2GP-PLYP double-hybrid approximations, which shows that these methods are not only empirically close to an optimum for general chemical applications but are also theoretically supported.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to provide a strict theoretical basis for double - hybrid density - functional theory (DH - DFT). Specifically, the author aims to derive a class of double - hybrid approximation methods containing only one empirical parameter by combining Hartree - Fock exchange energy and second - order Møller - Plesset correlation energy with semi - local exchange - correlation density functionals, and evaluate their performance on test sets such as atomization energies and reaction barrier heights. ### Research Background Density - functional theory (DFT) is a powerful electronic structure calculation method, widely used in the study of atoms, molecules, and solids. DFT in the Kohn - Sham (KS) form improves the accuracy of exchange - correlation energy through a series of approximations, such as local density approximation (LDA), generalized gradient approximation (GGA), and hybrid approximation (introducing Hartree - Fock exchange energy). However, although the existing double - hybrid approximation methods perform excellently, they lack a strict theoretical basis. ### Main Problems 1. **Insufficient Theoretical Basis**: Although the existing double - hybrid approximation methods perform well in practical applications, their theoretical basis is not rigorous enough, especially with more empirical steps involved. 2. **Selection of Optimization Parameters**: Standard double - hybrid methods usually rely on two empirical parameters (\(a_x\) and \(a_c\)), and how to reduce the number of these empirical parameters through strict theoretical derivation is an important issue. 3. **Performance Evaluation**: It is necessary to evaluate the performance of the newly proposed double - hybrid method on different test sets, especially its advantages compared with existing methods. ### Solutions The author solves the above problems in the following ways: 1. **Strict Derivation**: Using the multi - determinant - expanded Kohn - Sham scheme, a double - hybrid approximation method containing only one empirical parameter is proposed. This method obtains an appropriate correlation energy functional by uniformly scaling the density in coordinates, and naturally includes the MP2 correlation energy expression without ignoring single - excitation contributions. 2. **Performance Evaluation**: The newly proposed double - hybrid method is tested on representative atomization energy and reaction barrier height test sets and compared with other hybrid approximation methods. ### Results - The newly proposed one - parameter double - hybrid method (such as 1DH - BLYP) shows results close to chemical accuracy on some test sets. - Compared with the standard double - hybrid method (such as B2 - PLYP), the new method is not only more theoretically rigorous but also shows similar or better performance in practical applications. - For different benchmark test sets (such as AE6 and BH6), the optimal parameter values are different, but overall the new method shows good accuracy. ### Summary This paper provides new insights and support for double - hybrid density - functional theory through strict theoretical derivation and performance evaluation, further promoting the development of this field.

Double-hybrid density-functional theory made rigorous

Double-hybrid density-functional theory applied to molecular crystals

A general range-separated double-hybrid density-functional theory

Non-empirical Double-Hybrid Density Functionals as Reliable Tools for Electronic Structure Calculations

Consistent Improvement on Both Exchange and Correlation within Hybrid Multiconfiguration Pair-density Functional Theory: tB4LYP

A double-hybrid density functional based on good local physics with outstanding performance on the GMTKN55 database

Doubly Hybrid Density Functional For Accurate Description Of Thermochemistry, Thermochemical Kinetics And Nonbonded Interactions

Doubly Hybrid Density Functionals That Correctly Describe Both Density and Energy for Atoms.

Doubly Hybrid Density Functional for Accurate Descriptions of Nonbond Interactions, Thermochemistry, and Thermochemical Kinetics

Empirical Double‐Hybrid Density Functional Theory: A ‘Third Way’ in Between WFT and DFT

Range-separated double-hybrid density-functional theory applied to periodic systems

Adaptive hybrid density functionals

Symmetry Dilemma of Doubly Hybrid Density Functionals for Equilibrium Molecular Property Calculations.

Benchmarking Density Functionals on Structural Parameters of Small-/Medium-Sized Organic Molecules.

Toward the Construction of Parameter-Free Doubly Hybrid Density Functionals

Doubly Hybrid Functionals Close to Chemical Accuracy for Both Finite and Extended Systems: Implementation and Test of XYG3 and XYGJ-OS.

Spin-component-scaled double hybrids: An extensive search for the best fifth-rung functionals blending DFT and perturbation theory

Long-Range Corrected Hybrid Density Functionals with Improved Dispersion Corrections

A General and Transferable Local Hybrid Functional for Electronic Structure Theory and Many-Fermion Approaches

Pure density functional for strong correlations and the thermodynamic limit from machine learning

Doubly Hybrid Density Functional Xdh-Pbe0 From A Parameter-Free Global Hybrid Model Pbe0