Pith. sign in

REVIEW 4 major objections 5 minor 93 references

Analytical Excited-State Gradients and Derivative Couplings in TDDFT with Minimal Auxiliary Basis Set Approximation and GPU Acceleration

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper extends the TDDFT-ris approximation to analytical excited-state gradients and derivative couplings, reporting two- to three-fold speedups on medium-sized molecules while pinpointing near-degenerate state couplings as the main acc

desk verdict Useful, honestly benchmarked implementation of TDDFT-ris gradients and couplings, but the abstract oversells speedups and reliability. read the letter →

arxiv 2511.18233 v2 pith:AMABGGNQ submitted 2025-11-23 physics.chem-ph

Zhichen Pu , Xiaojie Wu , Yuanheng Wang , Cheng Fan , Wen Yan , Zehao Zhou , Yi Qin Gao , Qiming Sun This is my paper · ORCID
classification physics.chem-ph
keywords TDDFT-risexcited-stategradientsderivativecouplingsnonadiabaticresolutionofidentityminimalauxiliarybasisGPUaccelerationTamm-Dancoffapproximation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper extends an existing fast approximation for TDDFT excitation energies, known as TDDFT-ris, to the analytical derivatives that simulations actually need: excited-state nuclear gradients, ground-to-excited derivative couplings, and excited-to-excited derivative couplings. The central claim is that the same minimal auxiliary basis that accelerates excitation energies can be used in the derivative formulas, preserving accuracy for most geometries while reducing cost by roughly two- to three-fold for medium-sized organic molecules. The paper shows the method works well for excited-state geometry optimization, emission-energy estimates, and minimum-energy crossing point searches, but explicitly warns that derivative couplings between nearly degenerate states are the weak spot because small errors in the energy gap sitting in the denominator of the coupling formula get amplified. It also reports that ground-to-excited couplings gain no speedup, since those formulas contain no approximated terms. If right, the contribution is a practical, moderate acceleration rather than a drastic one, with a clear caveat about where it fails.

What carries the argument

The central object is the TDDFT-ris approximation: the Coulomb and exchange parts of the coupling matrix are rebuilt from three-center and two-center integrals over a minimal atom-centered auxiliary basis, plus the elimination of the pure exchange-correlation kernel term. The derivative couplings between excited states are then expressed through generating functions whose two-electron parts use the same ris integrals, with an explicit (EJ - EI)^-1 prefactor. The argument runs through a Lagrangian formulation with Z-vector equations for the ground-state orbital response; the Z-vector solve and the XC terms dominate the cost, which is why the overall speedup is only moderate despite the drasti

What would settle it

Compare the S1–S2 derivative coupling vector from TDDFT-ris and from standard TDDFT for a small molecule such as cyclopentadiene, ethene, or furan at a geometry where the exact energy gap is below 0.01 eV, and also compute a high-level wavefunction reference; if the ris vector magnitude is governed by its own gap rather than the true gap, so that the error grows roughly as 1/gap, the method's reliability claim for near-degenerate couplings is refuted.

Watch

Extended reading notes

Core claim

The paper derives and implements analytical excited-state gradients and derivative couplings in which the four-center two-electron integrals of the linear-response kernel are replaced by a resolution-of-identity (RI) approximation using a minimal auxiliary basis consisting of one Gaussian per angular momentum component (up to l = 2) on each atom. Because the pure exchange-correlation kernel is dropped and the integrals are shrunk, the cost of the response-related terms falls sharply. The authors' benchmarks show a two- to three-fold speedup for S1 gradients and S1–S2 derivative couplings on medium-sized molecules (1.5- to 1.6-fold when ground-state integrals are evaluated exactly), no speedu

Load-bearing premise

The load-bearing premise is that auxiliary-basis exponents fitted to reproduce excitation energies also preserve the energy-gap denominators that enter the derivative-coupling formula; the paper's cytosine benchmark shows this premise fails in near-degeneracy.

Editorial extensions

If this is right

  • Excited-state geometry optimizations and emission energies from the ris method should track standard TDDFT closely for most molecules, with gradient-norm deviations on the order of 10^-3 Hartree/Bohr and small geometry differences.
  • For S0–S1 derivative couplings the method is essentially indistinguishable from standard TDDFT, with mean absolute deviations around 10^-4 a0^-1, so simulations needing only ground-to-excited couplings gain no speedup but also no accuracy loss from the approximation itself.
  • For S1–S2 couplings, the method is trustworthy only when the two states are not near-degenerate: the numerator (transition-vector) errors remain small, but denominator gap errors dominate and can produce order-of-magnitude deviations.
  • The minimum-energy crossing point search on furan suggests the crossing geometry and the existence of a conical intersection are qualitatively conserved, even though the absolute crossing energy shifts by about 0.15 eV and the branching-plane topography becomes more isotropic.
  • Because the Z-vector solve is the dominant cost and is not accelerated by the ris approximation, the two- to three-fold speedup is the realistic ceiling for the current formulation in dynamics-focused calculations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • This suggests that energy-only benchmarks cannot certify derivative-coupling accuracy: a method can reproduce excitation energies well yet badly miss couplings whenever a gap near degeneracy is distorted.
  • An obvious extension, tested in the paper but left for future work, is applying the same ris approximation to the Z-vector solver; the paper reports four- to five-fold further speedups with small gradient deviations, but this would break energy-derivative consistency and needs careful validation for dynamics.
  • A pragmatic screening strategy would be to run the ris method alongside the cheaper energy-level calculation to identify geometries where the S1/S2 gap is below, say, 0.05 eV, and switch to full TDDFT for those cases.
  • The authors suggest spin-flip TDDFT as a fix for ground-to-excited couplings; one could equally imagine a hybrid that keeps the ris integrals but uses a corrected energy gap from a more accurate electronic-structure calculation to repair the near-degenerate derivative couplings.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper derives and implements analytical excited-state gradients and derivative couplings for the TDDFT-ris method, a minimal-auxiliary-basis RI approximation to the TDDFT linear-response kernel, within the GPU4PySCF framework. The formal derivation (Sec. 2) follows the standard TDDFT Lagrangian formalism and replaces the relevant two-electron integral terms with the ris-approximated integrals; the ground-to-excited-state coupling is identical to standard TDDFT. The implementation is validated by finite differences (Appendix B), with gradient MADs on the order of 10^-6 to 10^-5 Hartree/Bohr and derivative-coupling errors of 10^-7 to 10^-4 a0^-1. Benchmarks on medium-sized organic molecules (Sec. 4) show modest speedups for some quantities, while application tests (Sec. 5) reveal significant outliers in emission energies and derivative couplings. The abstract claims a two- to three-fold speedup and general reliability, which are not fully supported by the data in the body.

Significance. If the claims were fully supported, this would be a valuable contribution: analytical gradients and derivative couplings for the efficient TDDFT-ris approximation, with an open-source GPU implementation, would enable larger-scale excited-state geometry optimization and nonadiabatic dynamics. The finite-difference validation is a clear strength, and the public code availability is a positive for reproducibility. However, the paper's own benchmarks substantially temper the headline claims: the speedup is not uniform across derivative types or integral evaluation modes, and the accuracy has several outliers—most notably cytosine, where the emission energy is off by 0.74 eV and the S1/S2 derivative coupling has a MAD of 21.7 a0^-1. The conclusion itself recommends against black-box use and requires per-system validation, a caveat that should be reflected in the abstract. The central implementation claim is sound, but the performance and reliability claims need revision.

major comments (4)
  1. [Abstract; §4.1, Fig. 1, Table 2] The abstract states "a two- to three-fold speedup for both gradients and derivative couplings." Section 4.1 reports no speedup for g_{01} (ground-to-excited derivative coupling) and, for g_{12} and g1, speedups of 2.3/2.4 only when density fitting is used; with exact integrals the speedups drop to 1.6 and 1.5. The conclusion repeats the two-to-three-fold figure without this qualification. Please revise the abstract and conclusion to state the speedup as modest (≈1.5–2.5×) for excited-state gradients and S1–S2 couplings, and to note that g_{01} is not accelerated.
  2. [§5.1, Table 4] The text claims "a close agreement in the emission energies of S1 between the two approaches is observed across the entire set" and that the data points "lie closely along the diagonal." This is contradicted by the cytosine row: TDA 2.97 eV vs. TDA-ris 3.71 eV (0.74 eV error), and geometry RMS of 0.043 Bohr, an order of magnitude larger than most molecules in the table. This outlier is not discussed in the text, and it directly undermines the "reliable approximations for most cases" claim in the abstract. The authors should either reinterpret cytosine as an exception and state that, or restrict the claim.
  3. [§5.2 vs. §5.4, Fig. 8] Section 5.2 concludes that g_{01} couplings "exhibit excellent agreement with the reference TDA values." Section 5.4, however, reports that the g_{01} coupling in ethene shows "poor linear correlation" with discrepancies in components as large as 0.0200 a0^-1, and attributes this to transition-vector differences. Since g_{01} does not involve the energy denominator (EJ−EI)^−1, this error is not a near-degeneracy effect. This directly contradicts the abstract's claim that "noticeable errors mainly occurring in derivative couplings between nearly degenerate states." The abstract's caveat is too narrow.
  4. [§5.2, Eqs. (17)/(26); Ref. 87] The derivative-coupling accuracy for g_{12} depends on the energy gap denominator (EJ−EI)^−1, but the ris auxiliary exponents are fitted to reproduce excitation energies, not energy differences. The paper's own cytosine case shows this clearly: TDA S1/S2 gap is 0.002 eV while TD-ris gives 0.031 eV, and the resulting g12 MAD is 21.7 a0^-1 even though the numerator error is only 1.2×10^-4 a0^-1. The paper identifies near-degeneracy as a risk, but does not acknowledge that the user cannot know a priori whether a system is near-degenerate. The conclusion's recommendation of per-system validation should be reflected in the abstract's reliability statement, rather than the unqualified "most cases."
minor comments (5)
  1. [Table 6] There are two identical "pyridine" entries; the second is likely "pyrimidine." Please correct.
  2. [Figure 3] The linear regression line is shown but the slope, intercept, and R² values are not reported. These would help quantify "close agreement" and make the cytosine outlier visible in the fit.
  3. [Table 4] The column headings under "Gradients consistency" ("TDA-ris" and "TDA") would be clearer if the caption explained that each column gives the norm of the contrast gradient evaluated at the geometry optimized by the other method. Currently the meaning is implicit.
  4. [Eq. (27)] The derivative notation in the second term, "(M−1)AB(B|sr)ξ", is ambiguous. Use (M−1)AB ∂(B|sr)/∂ξ or add parentheses for clarity.
  5. [§2.2, Eq. (20)] The neglect of the XC kernel fxc is stated, but its effect on derivative properties is not discussed. A sentence noting that this is inherited from the TDDFT-ris energy formalism and is expected to be small for the tested functionals would be useful.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: derivative implementation is checked by finite differences; accuracy claims are benchmarks against standard TDA, not derived from fitted inputs.

full rationale

The paper's new content is an analytical-gradient and derivative-coupling implementation for the TDDFT-ris model. The only fitted quantities entering the model are the minimal auxiliary-basis exponents, explicitly stated in Section 4 to have been optimized in Ref. 87 to reproduce excitation energies; the present paper adopts them without refitting. No result here reduces to those fits by construction. Excitation energies enter the derivative-coupling expression through (E_J - E_I)^-1, but the paper does not present the energy gap as a prediction derived from the new derivatives; instead it benchmarks the resulting couplings against standard TDA and explicitly attributes large g_12 errors to energy-gap discrepancies (Section 5.2, cytosine example). The implementation's correctness is validated by finite differences in Appendix B (Tables 6-7), confirming that the analytic gradients and couplings are consistent derivatives of the ris energy functional rather than a renaming of inputs. The self-citations to Refs. 86/87 supply the ris method and its auxiliary-basis parameterization, but they are not used as the evidence for derivative accuracy, which rests on finite-difference validation and comparison to standard TDA. The acknowledged failures for near-degenerate states, state reordering, and the recommendation to validate per system (Sections 5.2 and 6) are honest limitation statements, not evidence that a claimed derivation was assumed. Thus no circular step of any enumerated kind is present; the inherited-accuracy concern is a correctness/robustness risk, not circularity.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

No new physical entities are introduced. The method inherits the fitted auxiliary-basis exponents from Ref. 87; the derivative formulas are standard TDDFT with the ris-replaced integrals.

free parameters (1)
  • Auxiliary basis exponents (J-fit s,p; K-fit s) = Not given; adopted from Ref. 87
    Exponents of the minimal auxiliary GTOs, optimized to minimize excitation-energy deviations from standard TDDFT for PBE0/def2-TZVP. Directly control the accuracy of the approximated integrals used in the derivative equations.
assumptions (6)
  • domain assumption Standard linear-response TDDFT (Casida) and Lagrangian derivative formalism of Ref. 38.
    The entire derivative framework, including the Z-vector equation and generating functions, is taken from Li et al.; the paper does not re-derive it.
  • ad hoc to paper Neglect of the XC kernel fxc in the TDDFT-ris coupling matrix (Eq. 20).
    The ris approximation drops the exchange-correlation kernel entirely, which is not an exact limit; it is justified only by benchmarks.
  • ad hoc to paper Minimal auxiliary basis: one GTO per angular momentum per atom.
    The choice of a single s/p (J) and s (K) GTO per atom is a heuristic model, not derived from first principles.
  • domain assumption Neglect of quadratic-response term in derivative-coupling generating function (following Ref. 74).
    The paper states 'It should be noted that the quadratic-response term is neglected, following the recommendation in Ref. 74.' This is an approximation inherited from prior work.
  • domain assumption Adiabatic approximation of the XC functional in TDDFT.
    Standard in TDDFT; frequency-independent kernel.
  • domain assumption Closed-shell, spin-restricted reference state.
    The implemented equations assume a closed-shell ground state; all benchmarks use this.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Analytical Excited-State Gradients and Derivative Couplings in TDDFT with Minimal Auxiliary Basis Set Approximation and GPU Acceleration." pith.science (2026). https://pith.science/paper/AMABGGNQ

@misc{pith2026251118233,
  author       = {Pith},
  title        = {Pith review of: Analytical Excited-State Gradients and Derivative Couplings in TDDFT with Minimal Auxiliary Basis Set Approximation and GPU Acceleration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AMABGGNQ}},
  note         = {Machine review of arXiv:2511.18233}
}
read the original abstract

Calculating excited-state gradients and derivative couplings using time-dependent density functional theory (TDDFT) remains a computationally demanding task. An efficient variant, TDDFT with resolution of the identity and a minimal auxiliary basis (TDDFT-ris), has been developed to accelerate excitation energy calculations. However, the formulation and implementation of analytical derivatives for this method have not yet been reported. In this work, we present an implementation of analytical excited-state gradients and derivative couplings within the TDDFT-ris framework. Benchmark calculations on medium-sized organic molecules demonstrate a two- to three-fold speedup for both gradients and derivative couplings compared to standard TDDFT. The accuracy of the TDDFT-ris approach is assessed for gradient-dependent applications, including geometry optimizations, emission energy calculations, and the localization of minimum-energy crossing points. Overall, the TDDFT-ris method provides reliable approximations for most cases, with noticeable errors mainly occurring in derivative couplings between nearly degenerate states.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

93 extracted references · 4 canonical work pages

  1. [1]

    Toward reliable density functional methods without adjustable parameters: The pbe0 model.The Journal of chemical physics, 110(13):6158–6170, 1999

    Carlo Adamo and Vincenzo Barone. Toward reliable density functional methods without adjustable parameters: The pbe0 model.The Journal of chemical physics, 110(13):6158–6170, 1999

    1999

  2. [2]

    Christoph Bannwarth and Stefan Grimme. A simplified time-dependent density functional theory approach for electronic ultraviolet and circular dichroism spectra of very large molecules.Computational and Theoretical Chemistry, 1040:45–53, 2014

    2014

  3. [3]

    Gfn2-xtb—an accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions

    Christoph Bannwarth, Sebastian Ehlert, and Stefan Grimme. Gfn2-xtb—an accurate and broadly parametrized self-consistent tight-binding quantum chemical method with multipole electrostatics and density-dependent dispersion contributions. Journal of chemical theory and computation, 15(3):1652–1671, 2019

    2019

  4. [4]

    A new time dependent density functional algorithm for large systems and plasmons in metal clusters.The Journal of Chemical Physics, 143(2), 2015

    Oscar Baseggio, Giovanna Fronzoni, and Mauro Stener. A new time dependent density functional algorithm for large systems and plasmons in metal clusters.The Journal of Chemical Physics, 143(2), 2015

    2015

  5. [5]

    A direct method for the location of the lowest energy point on a potential surface crossing.Chemical physics letters, 223(3):269–274, 1994

    Michael J Bearpark, Michael A Robb, and H Bernhard Schlegel. A direct method for the location of the lowest energy point on a potential surface crossing.Chemical physics letters, 223(3):269–274, 1994

    1994

  6. [6]

    Time-dependent density functional response theory for molecules

    Mark E Casida. Time-dependent density functional response theory for molecules. InRecent AdvancesIn Density Functional Methods: (Part I), pages 155–192. World Scientific, 1995

    1995

  7. [7]

    Analytic gradients for selected configuration interaction

    Jeremy P Coe. Analytic gradients for selected configuration interaction. Journal of Chemical Theory and Computation, 19(3):874–886, 2023

    2023

  8. [8]

    Jeremy P Coe. Analytic non-adiabatic couplings for selected configuration interaction via approximate degenerate coupled perturbed hartree–fock.Journal of Chemical Theory and Computation, 19(22):8053–8065, 2023

    2023

Show all 93 references
  1. [9]

    Ab initio nonadiabatic quantum molecular dynamics.Chemical reviews, 118(7):3305–3336, 2018

    Basile FE Curchod and Todd J Martínez. Ab initio nonadiabatic quantum molecular dynamics.Chemical reviews, 118(7):3305–3336, 2018

    2018

  2. [10]

    Damped linear response tddft with range-separated functionals and density fitting.The Journal of Physical Chemistry A, 129(40):9453–9463, 2025

    Pierpaolo D’Antoni, Daniele Toffoli, and Mauro Stener. Damped linear response tddft with range-separated functionals and density fitting.The Journal of Physical Chemistry A, 129(40):9453–9463, 2025. doi: 10.1021/acs. jpca.5c03798. URLhttps://doi.org/10.1021/acs.jpca.5c03798. P...

    2025 doi

  3. [11]

    The exact integral simplified time-dependent density functional theory (xstd-dft)

    Marc de Wergifosse and Stefan Grimme. The exact integral simplified time-dependent density functional theory (xstd-dft). The Journal of Chemical Physics, 160(20), 2024

    2024

  4. [12]

    Gonzalo Diaz Miron, Carlos R Lien-Medrano, Debarshi Banerjee, Marta Monti, Bálint Aradi, Michael A Sentef, Thomas A Niehaus, and Ali Hassanali. Non-adiabatic couplings in surface hopping with tight binding density functional theory: The case of molecular motors.Journal of Chem...

    2024

  5. [13]

    Shirin Faraji, Spiridoula Matsika, and Anna I Krylov. Calculations of non-adiabatic couplings within equation-of- motion coupled-cluster framework: Theory, implementation, and validation against multi-reference methods.The Journal of chemical physics, 148(4), 2018

    2018

  6. [14]

    Ignacio Fdez Galvan, Mickael G Delcey, Thomas Bondo Pedersen, Francesco Aquilante, and Roland Lindh. Analytical state-average complete-active-space self-consistent field nonadiabatic coupling vectors: Implementation with density-fitted two-electron integrals and application to...

    2016

  7. [15]

    Approximate analytical gradients and nonadiabatic couplings for the state-average density matrix renormalization group self-consistent-field method

    Leon Freitag, Yingjin Ma, Alberto Baiardi, Stefan Knecht, and Markus Reiher. Approximate analytical gradients and nonadiabatic couplings for the state-average density matrix renormalization group self-consistent-field method. Journal of chemical theory and computation, 15(12):...

    2019

  8. [16]

    On the density matrix based approach to time-dependent density functional response theory.The Journal of Chemical Physics, 114(14):5982–5992, 2001

    Filipp Furche. On the density matrix based approach to time-dependent density functional response theory.The Journal of Chemical Physics, 114(14):5982–5992, 2001

    2001

  9. [17]

    Adiabatic time-dependent density functional methods for excited state properties

    Filipp Furche and Reinhart Ahlrichs. Adiabatic time-dependent density functional methods for excited state properties. The Journal of chemical physics, 117(16):7433–7447, 2002

    2002

  10. [18]

    Analytic gradients for the coupled-cluster singles, doubles, and triples (ccsdt) model

    Jürgen Gauss and John F Stanton. Analytic gradients for the coupled-cluster singles, doubles, and triples (ccsdt) model. The Journal of chemical physics, 116(5):1773–1782, 2002. 21

    2002

  11. [19]

    Giulia Giannone and Fabio Della Sala. Minimal auxiliary basis set for time-dependent density functional theory and comparison with tight-binding approximations: Application to silver nanoparticles.The Journal of Chemical Physics, 153(8), 2020

    2020

  12. [20]

    A simplified tamm-dancoff density functional approach for the electronic excitation spectra of very large molecules.The Journal of chemical physics, 138(24), 2013

    Stefan Grimme. A simplified tamm-dancoff density functional approach for the electronic excitation spectra of very large molecules.The Journal of chemical physics, 138(24), 2013

    2013

  13. [21]

    On the evaluation of analytic energy derivatives for correlated wave functions

    Nicholas C Handy and Henry F Schaefer III. On the evaluation of analytic energy derivatives for correlated wave functions. The Journal of chemical physics, 81(11):5031–5033, 1984

    1984

  14. [22]

    Analytical excited state gradients for time-dependent density functional theory plus tight binding (tddft+ tb).The Journal of Chemical Physics, 158(22), 2023

    Shana Havenridge, Robert Rüger, and Christine M Aikens. Analytical excited state gradients for time-dependent density functional theory plus tight binding (tddft+ tb).The Journal of Chemical Physics, 158(22), 2023

    2023

  15. [23]

    Spin-flip tddft for photochemistry

    John M Herbert and Aniket Mandal. Spin-flip tddft for photochemistry. InTime-dependent density functional theory, pages 361–404. Jenny Stanford Publishing, 2022

    2022

  16. [24]

    John M Herbert, Xing Zhang, Adrian F Morrison, and Jie Liu. Beyond time-dependent density functional theory using only single excitations: Methods for computational studies of excited states in complex systems.Accounts of chemical research, 49(5):931–941, 2016

    2016

  17. [25]

    Time-dependent density functional theory within the tamm–dancoff approximation

    So Hirata and Martin Head-Gordon. Time-dependent density functional theory within the tamm–dancoff approximation. Chemical Physics Letters, 314(3-4):291–299, 1999

    1999

  18. [26]

    Hohenberg and W

    P. Hohenberg and W. Kohn. Inhomogeneous electron gas.Phys. Rev., 136:B864–B871, Nov 1964. doi: 10.1103/ PhysRev.136.B864. URLhttps://link.aps.org/doi/10.1103/PhysRev.136.B864

    1964 doi

  19. [27]

    Pedagogical overview of the fewest switches surface hopping method.ACS omega, 7(50):45810–45824, 2022

    Amber Jain and Aarti Sindhu. Pedagogical overview of the fewest switches surface hopping method.ACS omega, 7(50):45810–45824, 2022

    2022

  20. [28]

    Kohn–sham time-dependent density functional theory with tamm–dancoff approximation on massively parallel gpus.npj Computational Materials, 9(1):81, 2023

    Inkoo Kim, Daun Jeong, Won-Joon Son, Hyung-Jin Kim, Young Min Rhee, Yongsik Jung, Hyeonho Choi, Jinkyu Yim, Inkook Jang, and Dae Sin Kim. Kohn–sham time-dependent density functional theory with tamm–dancoff approximation on massively parallel gpus.npj Computational Materials, ...

    2023 doi

  21. [29]

    Weisburn, Alexandra Alexiu, Troy Van Voorhis, Young Min Rhee, Won-Joon Son, Hyung-Jin Kim, Jinkyu Yim, Sungmin Kim, Yeonchoo Cho, I nkook Jang, Seungmin Lee, and Dae Sin Kim

    Inkoo Kim, Daun Jeong, Leah P. Weisburn, Alexandra Alexiu, Troy Van Voorhis, Young Min Rhee, Won-Joon Son, Hyung-Jin Kim, Jinkyu Yim, Sungmin Kim, Yeonchoo Cho, I nkook Jang, Seungmin Lee, and Dae Sin Kim. Very-large-scale gpu-accelerated nuclear gradient of time-dependent den...

    2024 doi

  22. [30]

    Self-consistent equations including exchange and correlation effects.Phys

    Walter Kohn and Lu Jeu Sham. Self-consistent equations including exchange and correlation effects.Phys. Rev., 140(4A):A1133, 1965

    1965

  23. [31]

    Oliveira, and Miguel A.L

    Susi Lehtola, Conrad Steigemann, Micael J.T. Oliveira, and Miguel A.L. Marques. Recent developments in libxc — a comprehensive library of functionals for density functional theory.SoftwareX, 7:1–5, 2018. ISSN 2352-7110. doi: https://doi.org/10.1016/j.softx.2017.11.002. URL htt...

    2018 doi

  24. [32]

    Recent developments in libxc—a comprehensive library of functionals for density functional theory.SoftwareX, 7:1–5, 2018

    Susi Lehtola, Conrad Steigemann, Micael JT Oliveira, and Miguel AL Marques. Recent developments in libxc—a comprehensive library of functionals for density functional theory.SoftwareX, 7:1–5, 2018

    2018

  25. [33]

    On the evaluation of nonadiabatic coupling matrix elements using sa-mcscf/ci wave functions and analytic gradient methods

    Byron H Lengsfield, Paul Saxe, and David R Yarkony. On the evaluation of nonadiabatic coupling matrix elements using sa-mcscf/ci wave functions and analytic gradient methods. i.The Journal of chemical physics, 81(10): 4549–4553, 1984

    1984

  26. [34]

    Conical intersections and double excitations in time-dependent density functional theory.Molecular Physics, 104(5-7):1039–1051, 2006

    Benjamin G Levine, Chaehyuk Ko, Jason Quenneville, and Todd J MartÍnez. Conical intersections and double excitations in time-dependent density functional theory.Molecular Physics, 104(5-7):1039–1051, 2006

    2006

  27. [35]

    Noncollinear and spin-flip tddft in multicollinear approach

    Hao Li, Zhichen Pu, Qiming Sun, Yi Qin Gao, and Yunlong Xiao. Noncollinear and spin-flip tddft in multicollinear approach. Journal of Chemical Theory and Computation, 19(8):2270–2281, 2023

    2023

  28. [36]

    Introducing gpu acceleration into the python-based simulations of chemistry framework.The Journal of Physical Chemistry A, 129(5):1459–1468, 2025

    Rui Li, Qiming Sun, Xing Zhang, and Garnet Kin-Lic Chan. Introducing gpu acceleration into the python-based simulations of chemistry framework.The Journal of Physical Chemistry A, 129(5):1459–1468, 2025

    2025

  29. [37]

    Zhendong Li and Wenjian Liu. First-order nonadiabatic coupling matrix elements between excited states: A lagrangian formulation at the cis, rpa, td-hf, and td-dft levels.The Journal of chemical physics, 141(1), 2014. 22

    2014

  30. [38]

    Zhendong Li, Bingbing Suo, and Wenjian Liu. First order nonadiabatic coupling matrix elements between excited states: Implementation and application at the td-dft and pp-tda levels.The Journal of chemical physics, 141(24), 2014

    2014

  31. [39]

    Analytic evaluation of nonadiabatic coupling terms at the mr-ci level

    Hans Lischka, Michal Dallos, Péter G Szalay, David R Yarkony, and Ron Shepard. Analytic evaluation of nonadiabatic coupling terms at the mr-ci level. i. formalism.The Journal of chemical physics, 120(16):7322–7329, 2004

    2004

  32. [40]

    Surface hopping dynamics including intersystem crossing using the algebraic diagrammatic construction method.The Journal of Chemical Physics, 147(18), 2017

    Sebastian Mai, Felix Plasser, Mathias Pabst, Frank Neese, Andreas Köhn, and Leticia González. Surface hopping dynamics including intersystem crossing using the algebraic diagrammatic construction method.The Journal of Chemical Physics, 147(18), 2017

    2017

  33. [41]

    Ultrafast photochemistry and electron diffraction for cyclobutanone in the s2 state: Surface hopping with time-dependent density functional theory

    Ericka Roy Miller, Sean J Hoehn, Abhijith Kumar, Dehua Jiang, and Shane M Parker. Ultrafast photochemistry and electron diffraction for cyclobutanone in the s2 state: Surface hopping with time-dependent density functional theory. The Journal of Chemical Physics, 161(3), 2024

    2024

  34. [42]

    Machine learning for nonadiabatic molecular dynamics: best practices and recent progress.Chemical Science, 2025

    Carolin Müller, Štěpán Sršeň, Brigitta Bachmair, Rachel Crespo-Otero, Jingbai Li, Sascha Mausenberger, Max Pinheiro Jr, Graham Worth, Steven A Lopez, and Julia Westermayr. Machine learning for nonadiabatic molecular dynamics: best practices and recent progress.Chemical Science, 2025

    2025

  35. [43]

    Approximate time-dependent density functional theory.Journal of Molecular Structure: THEOCHEM, 914(1-3):38–49, 2009

    Thomas A Niehaus. Approximate time-dependent density functional theory.Journal of Molecular Structure: THEOCHEM, 914(1-3):38–49, 2009

    2009

  36. [44]

    Thomas A Niehaus. Ground-to-excited derivative couplings for the density functional-based tight-binding method: semi-local and long-range corrected formulations.Theoretical Chemistry Accounts, 140(4):34, 2021

    2021

  37. [45]

    Exact non-adiabatic coupling vectors for the time-dependent density functional based tight-binding method

    Thomas A Niehaus. Exact non-adiabatic coupling vectors for the time-dependent density functional based tight-binding method. The Journal of Chemical Physics, 158(5), 2023

    2023

  38. [46]

    Tight-binding approach to time-dependent density-functional response theory.Physical Review B, 63(8):085108, 2001

    Thomas A Niehaus, S Suhai, F Della Sala, P Lugli, Marcus Elstner, Gotthard Seifert, and Th Frauenheim. Tight-binding approach to time-dependent density-functional response theory.Physical Review B, 63(8):085108, 2001

    2001

  39. [47]

    First-order derivative couplings between excited states from adiabatic tddft response theory.The Journal of chemical physics, 142(6), 2015

    Qi Ou, Gregory D Bellchambers, Filipp Furche, and Joseph E Subotnik. First-order derivative couplings between excited states from adiabatic tddft response theory.The Journal of chemical physics, 142(6), 2015

    2015

  40. [48]

    Surface hopping with cumulative probabilities: Even sampling and improved reproducibility

    Shane M Parker and Colin J Schiltz. Surface hopping with cumulative probabilities: Even sampling and improved reproducibility. The Journal of Chemical Physics, 153(17), 2020

    2020

  41. [49]

    Treatment of nonadiabatic dynamics by on-the-fly trajectory surface hopping dynamics.Acta Phys.-Chim

    Jiawei Peng, Yu Xie, Deping Hu, Likai Du, and Zhenggang Lan. Treatment of nonadiabatic dynamics by on-the-fly trajectory surface hopping dynamics.Acta Phys.-Chim. Sin, 35(1):28–48, 2019

    2019

  42. [50]

    Best practices for nonadiabatic molecular dynamics simulations.arXiv preprint arXiv:2508.05263, 2025

    Antonio Prlj, Jack T Taylor, Jiří Janoš, Petr Slavíček, Federica Agostini, and Basile FE Curchod. Best practices for nonadiabatic molecular dynamics simulations.arXiv preprint arXiv:2508.05263, 2025

    2025

  43. [51]

    Enhancing pyscf-based quantum chemistry simulations with modern hardware, algorithms, and python tools.arXiv preprint arXiv:2506.06661, 2025

    Zhichen Pu and Qiming Sun. Enhancing pyscf-based quantum chemistry simulations with modern hardware, algorithms, and python tools.arXiv preprint arXiv:2506.06661, 2025

    2025 arXiv

  44. [52]

    Noncollinear density functional theory.Physical Review Research, 5(1):013036, 2023

    Zhichen Pu, Hao Li, Ning Zhang, Hong Jiang, Yiqin Gao, Yunlong Xiao, Qiming Sun, Yong Zhang, and Sihong Shao. Noncollinear density functional theory.Physical Review Research, 5(1):013036, 2023

    2023

  45. [53]

    Dmitrij Rappoport and Filipp Furche. Analytical time-dependent density functional derivative methods within the ri-j approximation, an approach to excited states of large molecules.The Journal of chemical physics, 122(6), 2005

    2005

  46. [54]

    Dirk R Rehn and Andreas Dreuw. Analytic nuclear gradients of the algebraic-diagrammatic construction scheme for the polarization propagator up to third order of perturbation theory.The Journal of Chemical Physics, 150 (17), 2019

    2019

  47. [55]

    Turbo charging time-dependent density-functional theory with lanczos chains.The Journal of Chemical Physics, 128(15), 2008

    Dario Rocca, Ralph Gebauer, Yousef Saad, and Stefano Baroni. Turbo charging time-dependent density-functional theory with lanczos chains.The Journal of Chemical Physics, 128(15), 2008

    2008

  48. [56]

    Robert Rüger, Erik Van Lenthe, Thomas Heine, and Lucas Visscher. Tight-binding approximations to time- dependent density functional theory—a fast approach for the calculation of electronically excited states.The Journal of chemical physics, 144(18), 2016. 23

    2016

  49. [57]

    Density-functional theory for time-dependent systems.Physical review letters, 52(12):997, 1984

    Erich Runge and Eberhard KU Gross. Density-functional theory for time-dependent systems.Physical review letters, 52(12):997, 1984

    1984

  50. [58]

    Ryabinkin, Jayashree Nagesh, and Artur F

    Ilya G. Ryabinkin, Jayashree Nagesh, and Artur F. Izmaylov. Fast numerical evaluation of time-derivative nonadiabatic couplings for mixed quantum-classical methods.J. Phys. Chem. Lett., 6(21):4200–4203, November

  51. [59]

    Ibele, Richard Bleyer, and Leticia González

    Eduarda Sangiogo-Gil, Lea M. Ibele, Richard Bleyer, and Leticia González. Velocity rescaling in surface hopping based on atomic contributions to electronic transitions.J. Chem. Theory Comput., 21(17):8278–8290, September

  52. [60]

    Giovanni Scalmani, Michael J Frisch, Benedetta Mennucci, Jacopo Tomasi, Roberto Cammi, and Vincenzo Barone. Geometries and properties of excited states in the gas phase and in solution: Theory and application of a time-dependent density functional theory polarizable continuum ...

    2006

  53. [61]

    Benchmarks for electronically excited states: Caspt2, cc2, ccsd, and cc3.The Journal of chemical physics, 128(13), 2008

    Marko Schreiber, Mario R Silva-Junior, Stephan Sauer, and Walter Thiel. Benchmarks for electronically excited states: Caspt2, cc2, ccsd, and cc3.The Journal of chemical physics, 128(13), 2008

    2008

  54. [62]

    Robert Send and Filipp Furche. First-order nonadiabatic couplings from time-dependent hybrid density functional response theory: Consistent formalism, implementation, and performance.The Journal of chemical physics, 132 (4), 2010

    2010

  55. [63]

    Analytic energy derivatives for ionized states described by the equation-of- motion coupled cluster method.The Journal of chemical physics, 101(10):8938–8944, 1994

    John F Stanton and Jürgen Gauss. Analytic energy derivatives for ionized states described by the equation-of- motion coupled cluster method.The Journal of chemical physics, 101(10):8938–8944, 1994

    1994

  56. [64]

    The requisite electronic structure theory to describe photoexcited nonadiabatic dynamics: Nonadiabatic derivative couplings and diabatic electronic couplings

    Joseph E Subotnik, Ethan C Alguire, Qi Ou, Brian R Landry, and Shervin Fatehi. The requisite electronic structure theory to describe photoexcited nonadiabatic dynamics: Nonadiabatic derivative couplings and diabatic electronic couplings. Accounts of chemical research, 48(5):13...

    2015

  57. [65]

    Understanding the surface hopping view of electronic transitions and decoherence.Annual review of physical chemistry, 67(1): 387–417, 2016

    Joseph E Subotnik, Amber Jain, Brian Landry, Andrew Petit, Wenjun Ouyang, and Nicole Bellonzi. Understanding the surface hopping view of electronic transitions and decoherence.Annual review of physical chemistry, 67(1): 387–417, 2016

    2016

  58. [66]

    Recent developments in the pyscf program package.The Journal of chemical physics, 153(2), 2020

    Qiming Sun, Xing Zhang, Samragni Banerjee, Peng Bao, Marc Barbry, Nick S Blunt, Nikolay A Bogdanov, George H Booth, Jia Chen, Zhi-Hao Cui, et al. Recent developments in the pyscf program package.The Journal of chemical physics, 153(2), 2020

    2020

  59. [67]

    Analytic evaluation of the nonadiabatic coupling vector between excited states using equation-of-motion coupled-cluster theory.The Journal of chemical physics, 131(12), 2009

    Attila Tajti and Péter G Szalay. Analytic evaluation of the nonadiabatic coupling vector between excited states using equation-of-motion coupled-cluster theory.The Journal of chemical physics, 131(12), 2009

    2009

  60. [68]

    Trajectory surface hopping within linear response time-dependent density-functional theory.Physical review letters, 98(2):023001, 2007

    Enrico Tapavicza, Ivano Tavernelli, and Ursula Rothlisberger. Trajectory surface hopping within linear response time-dependent density-functional theory.Physical review letters, 98(2):023001, 2007

    2007

  61. [69]

    Molecular dynamics with electronic transitions.The Journal of Chemical Physics, 93(2):1061–1071, 1990

    John C Tully. Molecular dynamics with electronic transitions.The Journal of Chemical Physics, 93(2):1061–1071, 1990

    1990

  62. [70]

    Perspective: Nonadiabatic dynamics theory.The Journal of chemical physics, 137(22), 2012

    John C Tully. Perspective: Nonadiabatic dynamics theory.The Journal of chemical physics, 137(22), 2012

    2012

  63. [71]

    Samat Tussupbayev, Niranjan Govind, Kenneth Lopata, and Christopher J Cramer. Comparison of real-time and linear-response time-dependent density functional theories for molecular chromophores ranging from sparse to high densities of states.Journal of chemical theory and comput...

    2015

  64. [72]

    Hiroki Uratani, Toshiki Morioka, Takeshi Yoshikawa, and Hiromi Nakai. Fast nonadiabatic molecular dynamics via spin-flip time-dependent density-functional tight-binding approach: Application to nonradiative relaxation of tetraphenylethylene with locked aromatic rings.Journal o...

    2020

  65. [73]

    Efficient approach to time-dependent density-functional perturbation theory for optical spectroscopy.Physical review letters, 96(11):113001, 2006

    Brent Walker, A Marco Saitta, Ralph Gebauer, and Stefano Baroni. Efficient approach to time-dependent density-functional perturbation theory for optical spectroscopy.Physical review letters, 96(11):113001, 2006

    2006

  66. [74]

    Nac-tddft: Time-dependent density functional theory for nonadiabatic couplings

    Zikuan Wang, Chenyu Wu, and Wenjian Liu. Nac-tddft: Time-dependent density functional theory for nonadiabatic couplings. Accounts of Chemical Research, 54(17):3288–3297, 2021

    2021

  67. [75]

    Hartree–fock exchange fitting basis sets for h to rn.Journal of computational chemistry, 29(2): 167–175, 2008

    Florian Weigend. Hartree–fock exchange fitting basis sets for h to rn.Journal of computational chemistry, 29(2): 167–175, 2008. 24

    2008

  68. [76]

    Florian Weigend and Reinhart Ahlrichs. Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for h to rn: Design and assessment of accuracy.Physical Chemistry Chemical Physics, 7(18): 3297–3305, 2005

    2005

  69. [77]

    On the quantum correction for thermodynamic equilibrium.Physical review, 40(5):749, 1932

    Eugene Wigner. On the quantum correction for thermodynamic equilibrium.Physical review, 40(5):749, 1932

    1932

  70. [78]

    from fragments to molecule

    Fangqin Wu, Wenjian Liu, Yong Zhang, and Zhendong Li. Linear-scaling time-dependent density functional theory based on the idea of “from fragments to molecule”.Journal of Chemical Theory and Computation, 7(11): 3643–3660, 2011

    2011

  71. [79]

    Enhancing gpu-acceleration in the python-based simulations of chemistry frameworks.Wiley Interdisciplinary Reviews: Computational Molecular Science, 15(2):e70008, 2025

    Xiaojie Wu, Qiming Sun, Zhichen Pu, Tianze Zheng, Wenzhi Ma, Wen Yan, Yu Xia, Zhengxiao Wu, Mian Huo, Xiang Li, et al. Enhancing gpu-acceleration in the python-based simulations of chemistry frameworks.Wiley Interdisciplinary Reviews: Computational Molecular Science, 15(2):e70...

    2025

  72. [80]

    Nonadiabatic molecular dynamics simulations based on time-dependent density functional tight-binding method

    Xiaoyan Wu, Shizheng Wen, Huajing Song, Thomas Frauenheim, Sergei Tretiak, ChiYung Yam, and Yu Zhang. Nonadiabatic molecular dynamics simulations based on time-dependent density functional tight-binding method. The Journal of Chemical Physics, 157(8):084114, 08 2022. ISSN 0021...

    2022 doi

  73. [81]

    Nonadiabatic molecular dynamics simulations based on time-dependent density functional tight-binding method

    Xiaoyan Wu, Shizheng Wen, Huajing Song, Thomas Frauenheim, Sergei Tretiak, ChiYung Yam, and Yu Zhang. Nonadiabatic molecular dynamics simulations based on time-dependent density functional tight-binding method. The Journal of Chemical Physics, 157(8), 2022

    2022

  74. [82]

    Noncollinear spin-flip tddft for potential energy surface crossings: Conical intersections and spin crossings.Journal of chemical theory and computation, 0(0):null, 0

    Xiaoyu Zhang, Tai Wang, Yi Qin Gao, and Yunlong Xiao. Noncollinear spin-flip tddft for potential energy surface crossings: Conical intersections and spin crossings.Journal of chemical theory and computation, 0(0):null, 0. doi: 10.1021/acs.jctc.5c01272. URLhttps://doi.org/10.10...

    doi

  75. [83]

    Analytic derivative couplings for spin-flip configuration interaction singles and spin-flip time-dependent density functional theory.The Journal of Chemical Physics, 141(6), 2014

    Xing Zhang and John M Herbert. Analytic derivative couplings for spin-flip configuration interaction singles and spin-flip time-dependent density functional theory.The Journal of Chemical Physics, 141(6), 2014

    2014

  76. [84]

    Xing Zhang and John M Herbert. Analytic derivative couplings in time-dependent density functional theory: Quadratic response theory versus pseudo-wavefunction approach.The Journal of chemical physics, 142(6), 2015

    2015

  77. [85]

    Xing Zhang and John M Herbert. Nonadiabatic dynamics with spin-flip vs linear-response time-dependent density functional theory: A case study for the protonated schiff base c5h6nh2+.The Journal of Chemical Physics, 155 (12), 2021

    2021

  78. [86]

    Zehao Zhou and Shane M Parker. Converging time-dependent density functional theory calculations in five iterations with minimal auxiliary preconditioning.Journal of Chemical Theory and Computation, 20(15):6738– 6746, 2024

    2024

  79. [87]

    Minimal auxiliary basis set approach for the electronic excitation spectra of organic molecules.The Journal of Physical Chemistry Letters, 14(7):1968–1976, 2023

    Zehao Zhou, Fabio Della Sala, and Shane M Parker. Minimal auxiliary basis set approach for the electronic excitation spectra of organic molecules.The Journal of Physical Chemistry Letters, 14(7):1968–1976, 2023

    1968

  80. [88]

    Linear-scaling time-dependent density-functional theory in the linear response formalism.The Journal of Chemical Physics, 139(6), 2013

    Tim J Zuehlsdorff, Nicholas DM Hine, James S Spencer, Nicholas M Harrison, D Jason Riley, and Peter D Haynes. Linear-scaling time-dependent density-functional theory in the linear response formalism.The Journal of Chemical Physics, 139(6), 2013. 25 Appendix A Appendix: Detaile...

    2013

  81. [91]

    The results demonstrate excellent agreement for bothgξ 01 and gξ

  82. [92]

    Deviations from the numerical reference consistently fall within the range of10 −7 to10 −4 a−1 0 , which strongly validates the present implementation

    This high level of accuracy is consistent for calculations employing either the exact two-electron integrals or the density fitting approximation. Deviations from the numerical reference consistently fall within the range of10 −7 to10 −4 a−1 0 , which strongly validates the pr...

  83. [93]

    RMS(g ξ 12) acetamide2.22×10 −6 4.35×10 −6 5.50×10 −6 1.40×10 −5 acetone4.98×10 −6 7.83×10 −6 1.13×10 −5 1.81×10 −5 cyclopropene1.56×10 −6 3.08×10 −6 2.21×10 −5 4.70×10 −5 ethene8.57×10 −8 1.81×10 −7 5.85×10 −7 9.07×10 −7 formamide3.90×10 −5 1.07×10 −4 1.23×10 −5 2.82×10 −5 pr...

  84. [2015]

    URLhttps://doi.org/10.1021/acs.jpclett.5b02062

    doi: 10.1021/acs.jpclett.5b02062. URLhttps://doi.org/10.1021/acs.jpclett.5b02062

    doi

  85. [2025]

    doi: 10.1021/acs.jctc.5c00737

    ISSN 1549-9618. doi: 10.1021/acs.jctc.5c00737. URLhttps://doi.org/10.1021/acs.jctc.5c00737

    doi

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.