Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 1911.04709

Excited state orbital optimization via minimizing the square of the gradient: General approach and application to singly and doubly excited states via density functional theory

Diptarka Hait·Martin Head-Gordon
View original source ↗Read source document ↗Export metadata ↓

Abstract

We present a general approach to converge excited state solutions to any quantum chemistry orbital optimization process, without the risk of variational collapse. The resulting Square Gradient Minimization (SGM) approach only requires analytic energy/Lagrangian orbital gradients and merely costs 3 times as much as ground state orbital optimization (per iteration), when implemented via a finite difference approach. SGM is applied to both single determinant $Δ$SCF and spin-purified Restricted Open-Shell Kohn-Sham (ROKS) approaches to study the accuracy of orbital optimized DFT excited states. It is found that SGM can converge challenging states where the Maximum Overlap Method (MOM) or analogues either collapse to the ground state or fail to converge. We also report that $Δ$SCF/ROKS predict highly accurate excitation energies for doubly excited states (which are inaccessible via TDDFT). Singly excited states obtained via ROKS are also found to be quite accurate, especially for Rydberg states that frustrate (semi)local TDDFT. Our results suggest that orbital optimized excited state DFT methods can be used to push past the limitations of TDDFT to doubly excited, charge-transfer or Rydberg states, making them a useful tool for the practical quantum chemist's toolbox for studying excited states in large systems.

Explore related subjects

physics.chem-phquant-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Diptarka Hait, Martin Head-Gordon. 2019-11-12. Excited state orbital optimization via minimizing the square of the gradient: General approach and application to singly and doubly excited states via density functional theory. https://doi.org/10.1021/acs.jctc.9b01127

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Intermolecular Interactions between Polyethylene, Water, and Potential Antistatic and Slip Additives: a Molecular Dynamics Study

Additives are essential to enhance or modify the properties of plastics for target applications. However, finding appropriate additives may be challenging, since we lack knowledge on their interactions with plastics and moisture, and the interplay between them. In this work, we study stearoyl diethanolamine as well as two amphiphilic molecules as potential new additives for their antistatic or slip properties in polyethylene by means of atomistic molecular dynamics simulations. We reveal that additive/water interactions and relative solubility are strongly determined by their relative ratio. The polyethylene model thin film adopts a crystalline core and an amorphous-like surface, with polymer chain terminations predominantly located at the surface of the slab. Water forms a layer on top of the polymer surface or droplets when its concentration is lowered, but it never enters the polymer matrix. All additives interact with water mainly by their polar heads, with water acting as a hydrogen bond acceptor or donor depending on the additive. The additives studied exhibit remarkably different structures when they are mixed with the polymer: two of them enter the polymer matrix to various degrees, either by intercalating their chains with the polyethylene ones or by forming micellar-like structures, while the third one stays at the surface. When water is incorporated into the system, the structure of some of the additive/polyethylene systems changes. The magnitude and nature of these changes depend on the relative concentrations of all species and on the nature of the additive. We propose that one of our two modeled molecules could have promising properties as a slip agent, as its behavior in the PE matrix resembles that of the industrial slip agent erucamide.

physics.chem-ph↗
arXiv2026

All-electron Dynamical Bethe-Salpeter Equation for Extended Systems with Atom-centered Orbital Basis

Solving Bethe-Salpeter equation (BSE) for the two-particle Green's function is the most widely used approach for taking into account the particle-hole (exciton) interaction in electronic excitation in the context of the many-body theory based on Green's function. In BSE calculations, the static approximation to the screened Coulomb interaction kernel is commonly employed. However, when the excitonic character is significant as typically indicated by a large exciton binding energy, dynamical screening effects become non-negligible, rendering the static approximation questionable. Because of the large computational cost due to the dense Brillouin zone integration necessary for convergence, solving the dynamical BSE for extended systems remains a significant challenge, especially when combined with GW calculation for the calculation of quasi-particle energies. In this work, we formulate the plane-wave based effective dielectric function method [Zhang, et al., Phys. Rev. B 107, 235205 (2023)] for the dynamical BSE calculation using atom-centered orbitals as basis functions. We implement this approach in our recently developed all-electron numerical atom-centered orbital (NAO) implementation of BSE@GW [Zhou, et. al. J. Chem. Theory Comput. 21, 291 (2025)] for extended systems. We validate our all-electron NAO-based implementation of the dynamical BSE method, and we then discuss its realistic application to molecular crystal of naphthalene by performing the dynamical BSE@G0W0 calculation.

physics.chem-ph↗
arXiv2026

Fixed-Dimensional Latent Flow for Generating Variable-Size 3D Molecules

In molecular discovery, molecule size is coupled to composition, structure, and other target properties. Yet most 3D generators require molecule size to be specified before generation. Here, we introduce Equivariant-Free Transformer-Autoencoded Latent Flow Matching, a two-stage generative framework that relies entirely on a single fixed-dimensional molecule-level latent representation to generate variable-size molecules. The second-stage flow matching model samples this latent vector, and an autoregressive Transformer decoder then determines molecule size while generating atom types, coordinates, and chemically informative states. Canonical atom ordering and rigid-pose alignment enable standard Transformers without equivariant layers, while joint decoding of molecular geometry and an enriched chemical state enables reliable, deterministic, chemistry-guided graph recovery without requiring a learned dense pairwise bond decoder. The same fixed-dimensional latent supports unconditional and property-conditioned flow matching, while optional property supervision adds an internal ranking readout, with no separate predictor or reference calculations. On PCQM4Mv2, EF-TALFM achieves the highest fraction of molecules that are unique, training-set novel, pass sanitization and PoseBusters sanity checks, 89.4\%, compared with 75.6\% for UAE-3D and 69.8\% for FlowMol. EF-TALFM also achieves higher measured computational throughput for training and sampling. Across ten target HOMO--LUMO gaps, internal ranking doubles the density functional theory (DFT)-verified hit rate within $0.1\,\mathrm{eV}$, while preserving 97\% novelty among unique verified hits. These results demonstrate that fixed-dimensional molecule-level generation followed by symmetry-resolved autoregressive realization provides a practical architecture for open-ended and property-directed 3D molecular design.

physics.chem-ph↗