Augmented Roothaan-Hall Hessian Applied to Spin-Restricted Open-Shell Density-Functional Theory
Yichi Zhang, Jun Yang · 2026
Read Augmented Roothaan-Hall Hessian Applied to Spin-Restricted Open-Shell Density-Functional Theory by Yichi Zhang, Jun Yang (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-2606-03709.txt.
Introduction
Zhang and Yang extend the augmented Roothaan-Hall Hessian method to open-shell density-functional theory. The target is self-consistent field optimization for spin-restricted open-shell wavefunctions. High-spin, low-spin, and two-determinant states are all covered. A detailed formulation of the method is given. The approach builds an effective Hessian in a systematic way.
Speed comes from that Hessian design, above all for quadratic-like energy surfaces. A universal energy form unifies closed-shell, unrestricted, and open-shell DFT codes. Grid-based integration fits inside the same frame. Iron-sulfur clusters with varied spin states test the method. These are famously hard SCF problems. The new method converges in far fewer iterations than L-BFGS and truncated Newton rivals. It also avoids higher-energy stationary points in two-determinant excited-state runs. A nickel-porphyrin spin-crossover study shows the method at work.
Catalog record
| CREATOR | Yichi Zhang, Jun Yang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.03709
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Augmented Roothaan-Hall Hessian Applied to Spin-Restricted Open-Shell Density-Functional Theory Yichi Zhang, Jun Yang Abstract We generalize the augmented Roothaan–Hall (ARH) Hessian formalism to the self-consistent field (SCF) optimization of spin-restricted open-shell (RO) wavefunctions, encompassing high-spin, low-spin, and two-determinant electronic states. A detailed ARH formulation is presented. We demonstrate that ARH is a highly efficient optimization algorithm for rapidly identifying accurate SCF minima, primarily owing to its systematic construction of an effective Hessian, particularly in the case of Euclidean quadratic energy functions. The ARH is built upon a universal energy formulation, including grid-based integration, for spin-restricted closed-shell, spin-unrestricted and RO density functional theory (DFT), thereby unifying and simplifying their numerical implementation. The performance of the present method is evaluated using two benchmarking studies. First, for a series of iron–sulfur clusters exhibiting different spin states, which represent notoriously challenging SCF problems, the ARH algorithm demonstrates superior convergence efficiency relative to L-BFGS and truncated Newton methods, requiring much fewer RO-SCF iterations to achieve convergence. Second, the ARH approach avoids convergence to higher-energy stationary points in two-determinant RO-SCF calculations for singlet excited states of selected photoactive compounds. Finally, an application of the ARH-based RO-SCF is illustrated by an investigation of the mechanistic origin of the spin-crossover phenomenon in Ni(II)-porphyrin complex utilized as a contrast agent. † † affiliation: Department of Chemistry, The University of Hong Kong, Hong Kong 999077, P.R. China † † affiliation: Laboratory for Synthetic Chemistry and Chemical Biology Limited, Hong Kong 999077, P.R. China † † email: [email protected] † † affiliation: Department of Chemistry, The University of Hong Kong, Hong Kong 999077, P.R. China † † affiliation: Laboratory for Synthetic Chemistry and Chemical Biology Limited, Hong Kong 999077, P.R. China † † affiliation: State Key Laboratory of Synthetic Chemistry and HKU-CAS Joint Laboratory on New Materials, The University of Hong Kong, Hong Kong 999077, P.R. China 1 1 Introduction The spin-restricted open-shell (RO) approach provides a wavefunction framework for describing electronic states exhibiting an imbalance of \alpha and \beta electron spins that are not perfectly paired, as an alternative to spin-unrestricted (U) wavefunctions. It yields a single set of spatial orbitals for both \alpha and \beta electrons, making a reasonable representation of high-spin electronic structures 57 ; 69 and a reference state for higher-level quantum chemistry methods such as the mixed-reference spin-flip time-dependent density functional theory (DFT) 43 ; 32 . Moreover, the wavefunction of two-determinant DFT (2D-DFT), which is calculated for spin-pure singlet excited states upon single-electron excitation, requires a single spin-adapted configuration state function that constitutes a pure singlet state ( S^{2}=0 ). 20 ; 40 ; 48 Despite its utility, RO self-consistent-field (SCF) wavefunctions are not in widespread use, as they are mathematically complex to solve, leading to difficulties in SCF convergence. The traditional approach for solving RO-SCF involves diagonalizing the composite Fock matrix and updating orbitals by direct inversion in the iterative subspace (DIIS). 57 ; 55 ; 56 ; 23 ; 26 A modern class of RO-SCF implementation leverages direct minimization on the flag manifold, 75 ; 52 where orbitals of separate spin components make up mutually orthogonal Fock subspaces. The mainstream algorithms of this class include the limited-memory Broyden–Fletcher–Goldfarb–Shanno (L-BFGS) 33 ; 76 ; 24 ; 9 and its variants for specific SCF problems 51 ; 70 ; 15 . Moreover, Newton’s method that is formally quadratically convergent for high-spin RO-SCF has been implemented in PySCF . 65 In this work, we aim to present a new scheme to solve RO-SCF, namely the augmented Roothaan-Hall (ARH) method, which was first reported for spin-restricted (R) closed-shell SCF 35 and later extended for unrestricted (U) and nuclear-electronic orbital SCF 19 as well as grand-canonical SCF 77 . Here we discuss the theoretical basis of ARH in a more general picture than its original closed-shell framework. 35 We demonstrate that ARH is more efficient than L-BFGS in the direct optimization of RO-SCF problems on the flag manifold, as the objective functions of the electronic energy are quadratic or approximately quadratic in the Euclidean space. The implementation of direct energy minimization requires an energy function and its orbital derivatives. As there are four types of spin-constrained SCF states, including R, U and RO of high spin (ROH) and low spin (ROL), it would be extremely tedious to implement their energy functions separately, especially the evaluation of their grid integration of exchange-correlation functionals. 54 ; 36 ; 63 In this work, we develop a universal formulation to encompass all four spin types from the tensorization viewpoint which adds the spin components as a new tensor dimension to all quantities, such as density and Fock matrices. The resulting compact and tidy expressions have enabled us to implement the RO-SCF energy function and orbital derivatives for both high spin and low spin states within a uniform implementation framework. Moreover, the 2D-DFT energy function and orbital derivatives can be simply derived based on the linear combination of those of the broken-symmetry singlet and triplet configurations. As such, an RO spin-pure diradical state can also be solved for by direct energy minimization including ARH, which will be discussed in details. The performance and applicability of our ARH-based RO-DFT formulation will be assessed on selected molecules for which traditional RO-DFT implementation is inefficient. For instance, it is a well-known fact that the SCF orbital optimization of iron-sulfur clusters often encounters the convergence difficulties. Here, our tests show that the ARH shortens the necessary iterations dramatically as compared with L-BFGS and truncated Newton’s method. Moreover, the ARH-based 2D-DFT carried out on selected photoactive compounds exhibits improved convergence behavior and yields excitation energies comparable with the experimental data. Lastly, we will showcase an application of the present method to Ni(II)-porphyrin complexes, a responsive contrasting agent for magnetic resonance imaging 14 , where the coordination-induced spin crossover mechanism is correctly identified. In view of those promising results, we recommend ARH as a superior solution to difficult RO-DFT problems. 2 2 A General Spin Formulation for SCF Implementation In most electronic structure implementations, the procedures for conducting spin-restricted closed-shell SCF, spin-restricted open-shell SCF, and spin-unrestricted SCF calculations are formulated and coded independently, including the corresponding two-electron integral contractions and exchange–correlation integrations. In this section, we present a unified theoretical framework for the SCF implementation that consistently encompasses all three spin formalisms. Within this framework, spin-restricted open-shell DFT for both high-spin and low-spin states is treated as a special case of the general formulation. The general SCF energy as a function of the occupied spatial molecular orbital coefficient matrix \mathbf{C} is given by \displaystyle E(\mathbf{C})=E^{\text{HF}}(\mathbf{C})+E^{\text{XC}}(\mathbf{C}), (1) where E^{\text{HF}}(\mathbf{C}) denotes the Hartree–Fock-like (HF) contribution and E^{\text{XC}}(\mathbf{C}) represents the exchange–correlation (XC) contribution. The coefficient matrix, which depends on the specific wavefunction form, can be decomposed into at most three column partitions corresponding to the spin components listed in table 1 , i.e., \mathbf{C}=\big[\mathbf{C}^{p}\ \mathbf{C}^{a}\ \mathbf{C}^{b}\big]. This coefficient matrix is directly related to the density matrix, which is defined as the outer product of the corresponding spin-block coefficient matrix, \mathbf{D}^{w}(\mathbf{C}^{w})=\mathbf{C}^{w}\mathbf{C}^{w\dagger}, (2) where the spin-component index w takes values in \{p,a,b\} . Note that, in contrast to the conventional definition used in quantum chemistry, our density matrix explicitly excludes occupation numbers. For convenience, one may also introduce a concatenated density matrix, \mathbf{D}=\big[\mathbf{D}^{p}\ \mathbf{D}^{a}\ \mathbf{D}^{b}\big],