Reduced Density Matrix Functional Theory Across Molecules and Periodic Systems: Screening and Coupled Optimization

Shuxin Pei, Neil Qiang Su · 2026

Read Reduced Density Matrix Functional Theory Across Molecules and Periodic Systems: Screening and Coupled Optimization by Shuxin Pei, Neil Qiang Su (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2608-00381.txt.

Introduction

Pei and Su advance RDMFT for molecules and periodic solids. RDMFT treats strong electron correlation in a rigorous frame. Use in solids was long blocked by two snags. Good transferable functionals were lacking. Tuning orbital occupations was slow as well. They show both snags can fall together.

Their build tunes natural orbitals and occupation numbers at once. That coupled scheme needs far fewer steps than split routes. Short-range screening makes the Power functional transferable across molecules and solids. Screening also reshapes the tuning landscape toward robust step sizes. Convergence speeds up sharply as a result. The screened functional beats semilocal and hybrid density functionals on surface reaction barriers.

Catalog record

CREATORShuxin Pei, Neil Qiang Su
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE93 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2608.00381

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Reduced Density Matrix Functional Theory Across Molecules and Periodic Systems: Screening and Coupled Optimization Shuxin Pei, Neil Qiang Su Abstract Reduced density matrix functional theory (RDMFT) provides a rigorous framework for treating strong electron correlation, yet its application to periodic systems has long been hindered by two fundamental challenges: the lack of transferable approximate functionals and the poor efficiency of orbital-occupation optimization. Here we show that these two longstanding obstacles can be overcome simultaneously. We develop a periodic implementation of RDMFT based on a coupled optimization framework that enables the efficient simultaneous optimization of natural orbitals and occupation numbers under periodic boundary conditions. Using this implementation, we demonstrate that physically motivated short-range screening transforms the Power functional into a transferable functional applicable to both molecules and periodic solids. Remarkably, short-range screening is found to play a dual role: besides substantially improving energetic accuracy, it fundamentally reshapes the optimization landscape, producing robust optimization step sizes and dramatically accelerating convergence. The coupled optimization framework consistently requires substantially fewer optimization iterations than conventional decoupled optimization for periodic calculations, while the screened ω \omega P22 functional outperforms semilocal and hybrid density functionals in predicting representative surface reaction barriers. These results establish a computationally efficient periodic RDMFT framework and identify short-range screening as a promising design principle for developing transferable one-body reduced-density-matrix functionals. I Introduction Over the past few decades, Kohn-Sham density functional theory (KS-DFT) [ 1 , 2 , 3 , 4 ] has become the workhorse for electronic-structure calculations of molecules and solids because of its excellent balance between computational efficiency and predictive accuracy. Nevertheless, standard KS-DFT remains fundamentally limited in systems where strong electronic correlation plays a decisive role. Typical examples include reaction energies and activation barriers [ 5 , 6 , 7 , 8 , 9 ] , molecular dissociation [ 10 , 11 , 12 , 13 ] , and Mott insulating states [ 14 , 15 , 16 ] , where the underlying single-determinant description becomes inadequate. Reduced density matrix functional theory (RDMFT) [ 17 , 18 , 19 ] provides a rigorous alternative framework by adopting the one-body reduced density matrix (1-RDM), or equivalently its spectral representation in terms of natural orbitals (NOs) and their occupation numbers (ONs) [ 20 ] , as the fundamental variable. This formulation enables an explicit description of fractional orbital occupations arising from both static and dynamic electron correlation, thereby naturally incorporating multireference character without relying on a fictitious non-interacting Kohn-Sham reference system. Consequently, RDMFT offers an attractive framework for describing electron correlation and has demonstrated particular promise for systems dominated by strong static correlation, where conventional KS-DFT often fails [ 21 , 22 , 23 , 24 , 25 ] . Despite its theoretical elegance, the practical applications of RDMFT remain relatively limited and have been focused primarily on finite systems, particularly challenging problems such as molecular dissociation [ 21 , 23 , 24 , 25 ] . Applications to extended systems are considerably fewer, with pioneering work largely confined to studies by Gross and co-workers. Using the Power functional, they successfully reproduced the insulating nature of prototypical Mott insulators and accurately predicted the fundamental gaps of conventional semiconductors [ 26 ] . The same functional was subsequently shown to reproduce spectral properties in good agreement with experimental measurements and high-level many-body calculations [ 27 , 28 ] , while yielding a physical picture of metal-insulator transitions consistent with dynamical mean-field theory [ 29 ] . These representative successes highlight the considerable yet largely untapped potential of RDMFT for electronic-structure calculations of extended systems. More generally, the widespread adoption of RDMFT has been hindered by the limited availability of approximate functionals that are simultaneously accurate and transferable. In particular, developing a single functional capable of reliably describing the ground-state energies of both molecules and solids remains an outstanding challenge, paralleling the long-standing quest for universally accurate density functionals in KS-DFT. Recently, introducing physically motivated short-range screening into the Power functional led to the development of the screened functional \omega P22 [ 25 , 30 ] , which substantially improved the accuracy for molecular systems, especially for chemically relevant ground-state energetics such as reaction barriers [ 30 ] . These developments naturally raise a fundamental question: can short-range screening serve as a general design principle for constructing transferable 1-RDM functionals that simultaneously improve the performance for periodic solids and molecular systems? Besides the quality of approximate functionals, another major factor limiting the widespread application of RDMFT is the efficiency of its numerical implementation. Conventional RDMFT calculations employ a decoupled optimization procedure, in which the total-energy functional is minimized successively with respect to the NOs and ONs. Such a nested optimization is computationally demanding and often exhibits slow convergence, severely limiting the applicability of RDMFT to large and complex systems. Recently, by combining unitary transformations of the NOs [ 31 , 32 ] , which automatically preserve orbital orthonormality, with the explicit-by-implicit (EBI) optimization method [ 33 , 34 ] , which rigorously enforces the ensemble- N -representability constraints on the ONs [ 20 , 19 , 35 ] , an efficient coupled optimization (CO) framework was developed in which the NOs and ONs are optimized simultaneously at every iteration [ 36 , 37 ] . Extensive benchmark calculations for molecular systems demonstrated that this approach substantially accelerates convergence while consistently achieving lower variational energies. However, extending the CO framework to periodic systems is far from straightforward because periodic RDMFT calculations involve Bloch orbitals, Brillouin-zone sampling, and the infinite repetition of electron-pair interactions under periodic boundary conditions. This naturally raises a second fundamental question: can the excellent efficiency and robustness of the CO framework be retained when extending RDMFT from finite molecules to periodic systems? Motivated by these two fundamental questions, in this work we develop a periodic implementation of RDMFT within both the CO and conventional decoupled optimization (deCO) frameworks, and systematically examine the performance of both the original Power functional and its screened counterpart, \omega P22. This enables us to investigate whether short-range screening constitutes a transferable design principle for 1-RDM functionals applicable to both molecular and periodic systems, and whether the efficiency and robustness of the CO framework can be retained for periodic RDMFT calculations. Together, these developments establish an accurate, transferable, and computationally efficient framework for RDMFT calculations of periodic systems. The remainder of this paper is organized as follows. Section II presents the theoretical formulation of periodic RDMFT, including the derivation of the first- and second-order derivatives of the total-energy functional with respect to the optimization variables, together with the CO algorithm under periodic boundary conditions. Computational details are provided in Sec. III . Section IV presents and discusses the performance of the proposed method for a series of representative periodic systems, including its energetic accuracy, convergence behavior, and application to the reaction barrier of a prototypical heterogeneous catalytic reaction. Finally, concluding remarks are given in Sec. V . II Theory Throughout this paper, the indices \mu,\nu,\kappa,\lambda refer to atomic orbitals (AOs), whereas i,j,a,b,p,q represent crystalline/molecular orbitals (MOs). II.1 RDMFT for periodic systems In RDMFT, the 1-RDM is defined as \gamma(\mathbf{x},\mathbf{x}^{\prime})=N\int d\mathbf{x}_{2}\dots d\mathbf{x}_{N}\Psi(\mathbf{x},\mathbf{x}_{2}\dots\mathbf{x}_{N})\Psi^{*}(\mathbf{x}^{\prime},\mathbf{x}_{2}\dots\mathbf{x}_{N}), (1) which serves as the fundamental variable of the total energy functional. Here, \Psi represents the

Read the complete text