Pith. sign in

REVIEW 2 major objections 4 minor 300 references

Iterating the static part of the Dyson equation turns a one-shot GW density matrix into a systematically improved one, equivalent to the RPA relaxed density matrix for Hartree-Fock starts.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 22:33 UTC pith:EUBLHFCZ

load-bearing objection Solid derivation of an iterated-Dyson GW density matrix with a clean RPA Z-vector equivalence for HF starts, but the benchmark's improvement claim is reference-sensitive and the paper honestly reports that itself. the 2 major comments →

arxiv 2607.15695 v1 pith:EUBLHFCZ submitted 2026-07-17 physics.chem-ph physics.comp-ph

GW reduced density matrix from iterated linearized Dyson equation

Fabien Bruneval , Erik Verzijl , Arno F\"orster , Mauricio Rodriguez-Mayorga This is my paper
classification physics.chem-ph physics.comp-ph
keywords GW approximationdensity matrixDyson equationrandom phase approximationZ-vectororbital relaxationmolecular benchmarkcoupled-cluster reference
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper establishes that iterating the linearized Dyson equation with the static (frequency-independent) part of the self-energy produces a corrected GW density matrix, γidGW, in which only the occupied-virtual block is relaxed. For a Hartree-Fock starting point, this iterated density matrix is exactly the analytical derivative of the random-phase-approximation (RPA) total energy—the same object as the Z-vector relaxed density matrix. For generalized Kohn-Sham starting points the two objects differ, and the discrepancy grows as the fraction of exact exchange decreases. Benchmarking 34 small molecules against coupled-cluster densities, the paper finds that γidGW systematically improves on the one-shot GW density matrix whenever the starting point has a large exact-exchange content, though MP2 remains competitive or better for several molecules. The reference level matters: switching the coupled-cluster reference from CCSD to CCSDT changes the method ranking.

Core claim

The central discovery is that iterating the static part of the self-energy—the Hartree and exchange response to the density-matrix change—can be folded into a linear system for the occupied-virtual block of the density-matrix correction. The operator of this linear system is the Hartree-Fock orbital-stability (electronic Hessian) matrix, and its right-hand side is built from the one-shot GW correction plus static exchange-correlation terms. When the starting mean field is Hartree-Fock, the solution coincides with the Z-vector relaxed density matrix of the RPA energy functional; with a generalized Kohn-Sham start it does not. The paper reports that the resulting density matrix improves over o

What carries the argument

The central object is the linear system in Eqs. (21)–(23): the unknown is the occupied-virtual block of the density-matrix correction, ΔγidGW, and the matrix Aia,jb = (εa − εi)δijδab + ⟨ij||ab⟩ + ⟨ib||aj⟩ is the Hartree-Fock electronic Hessian. The right-hand side Y combines the one-shot GW correction ΔγGW with integrals built from its occupied-occupied and virtual-virtual blocks. Solving this system performs the orbital relaxation that the static self-energy induces; the occupied-occupied and virtual-virtual blocks remain equal to the one-shot GW values, so electron number is conserved.

Load-bearing premise

The benchmark's central conclusion—that iterating the Dyson equation improves densities—assumes the CCSDT/cc-pVQZ densities are effectively exact references; the paper itself notes that switching from CCSD to CCSDT changes the method ranking, so if CCSDT is not converged (basis or core correlation) or if the few outliers dominate, the improvement claim could reverse.

What would settle it

Compute γidGW and the full configuration interaction (FCI) density for a small molecule with a near-degenerate HOMO-LUMO gap (e.g., stretched Be2 or twisted ethylene); if the linear system's A matrix has a negative eigenvalue, or if γidGW departs from the FCI density more than one-shot γGW does, the central improvement claim fails. More narrowly, take the 34-molecule benchmark, replace the CCSDT reference with FCI for a handful of the molecules, and check whether γidGW still beats γGW there.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • For Hartree-Fock starting points, γidGW is the analytical derivative of the RPA total energy, giving a variational, parameter-free route to relaxed one-electron properties within RPA.
  • For generalized Kohn-Sham starts, γidGW is not a relaxed density matrix of any known energy functional; the deviation from the Z-vector result grows as exact exchange decreases.
  • On the 34-molecule benchmark, γidGW improves the electron density and kinetic energy over the one-shot γGW for every molecule except Li2, provided the starting point has high exact-exchange content (α ≳ 0.5).
  • Optimal starting points for γidGW sit around PBEh(0.6–0.7), at which the density errors reach MP2 quality.
  • Low exact-exchange starting points can make the iterated procedure diverge or produce large density errors, traced to small HOMO-LUMO gaps and the ill-conditioning of the linear system.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Because the correction only rotates occupied-virtual pairs, γidGW can be read as a cheap orbital-relaxation step; one testable extension is to use γidGW as a starting density for a quasiparticle self-consistent GW loop, which might accelerate convergence.
  • The A matrix is the HF orbital-stability matrix; its smallest eigenvalue could serve as a diagnostic for whether the iterated correction is trustworthy—near-zero or negative eigenvalues would flag unstable starting points, which the paper already observes at low α.
  • The benchmark's dependence on the coupled-cluster reference (CCSD vs CCSDT) suggests that any future comparison of density-matrix methods should use CCSDT or better references; a natural test is to recompute the ranking with CCSDT(Q) or FCI for a few of the 34 molecules.
  • The numerical-integration route (Section III C) works even when the linear system cannot be formed explicitly; this opens the door to applying the same iterated-Dyson idea to self-energies whose static part is not a simple Hartree-Fock-like term, such as self-consistent GW or vertex-corrected schemes, though the linear-system equivalence would be lost.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper derives an iterated linearized Dyson equation for the GW one-body reduced density matrix. Updating only the static part of the self-energy with the final density matrix leaves the occupied-occupied and virtual-virtual blocks unchanged, while the occupied-virtual block is determined by a linear system (Eqs. 21-23). For a Hartree-Fock starting point the resulting density matrix is shown to be equivalent to the RPA Z-vector/relaxed density matrix; for gKS starting points it differs. The paper benchmarks the iterated GW density matrix against CCSDT/cc-pVQZ reference densities for 34 molecules, reporting that it improves upon the non-iterated GW density for high exact-exchange starting points, while cautioning that low exact-exchange leads to instability and that the CCSD vs CCSDT reference choice changes method rankings.

Significance. The theoretical derivation is a valuable, self-contained contribution: it provides a closed-form, non-iterative route to a relaxed-type density matrix from a GW self-energy, with a clear connection to the RPA Z-vector equations. If the benchmark conclusions are robust, the method offers a low-cost improvement to one-shot GW densities and one-electron properties. The paper is also commendably transparent about the numerical instability for low exact-exchange starts and about the sensitivity of the benchmark to the coupled-cluster reference level. The implementation appears reproducible from the equations given; the analytic linear system and numerical integration agree where tested.

major comments (2)
  1. [Section IV.B, Fig. 5] The central numerical claim that gamma_idGW 'systematically improves upon' gamma_GW is established only against CCSDT/cc-pVQZ. The paper itself states that replacing CCSD by CCSDT 'shakes up the ranking of the methods' and Fig. 4 shows MP2 is often closer to CCSDT than gamma_idGW for Si-containing molecules. Without checks of basis-set convergence (cc-pVQZ vs cc-pV5Z) or higher excitations (CCSDT(Q)) for at least a subset of molecules, the aggregate improvement could depend on the reference. Please add such tests or explicitly qualify the conclusion as 'relative to CCSDT at cc-pVQZ' and discuss the sensitivity.
  2. [Abstract and Section V] The abstract concludes that the iterated Dyson equation 'indeed produces improved density matrices for molecular systems' without the qualifier in Section V that this holds 'for mean-field starting points with a large content of exact-exchange.' Given Fig. 5 and Section IV.A, where low exact-exchange content leads to large errors and convergence failures, the unqualified statement overstates the domain of validity. Revise the abstract and conclusion to match the body's more careful claim.
minor comments (4)
  1. [Appendix A, Eq. (A3)] The linear system in Eq. (A3) appears to have a typo: the last factor should be DeltaDelta-gamma_idGW_jb, not DeltaDelta-gamma_idGW_ia. As written, the index ia is summed over twice (on A_ia,jb and on the unknown), which is inconsistent.
  2. [Section II.A, Eq. (4)] The notation for the self-energy in Eq. (4) is dense; it may help to explicitly state that the term in curly braces is the full self-energy Sigma(omega), with M(omega) the dynamic part. This is clear from context but would aid readability.
  3. [Section IV.B, Fig. 5] The use of PBEh(alpha) with a tuned alpha in Fig. 5 is post hoc. The paper does not present it as a prediction, but it would be helpful to state explicitly that the optimal-alpha values are obtained from the same benchmark and therefore are not predictive for new systems.
  4. [General] PT2@HF and MP2@HF are used in Fig. 4 but not defined in the text; a brief definition would help readers unfamiliar with the distinction.

Circularity Check

0 steps flagged

No significant circularity: the iterated-Dyson density matrix is derived from stated self-energy definitions and benchmarked against external CCSDT references.

full rationale

The central derivation is self-contained. Starting from the linearized Dyson equation (Eq. 13) and the static self-energy definition (Eqs. 14-16), the paper obtains an implicit equation for Δγ^idGW (Eq. 20), which is rearranged into the linear system A·X=Y (Eqs. 21-23). The unknown X is the solution of that system, not a fitted or pre-inserted result. No parameter is fitted to the benchmark data and then called a prediction. The occupied-occupied and virtual-virtual blocks are inherited from the earlier γ^GW expressions, but those expressions are rederived in Eqs. (12) from the contour integral (6) and the GW self-energy (8)-(9), so the reliance on Refs. 16-18 is recap rather than load-bearing circularity. The claimed equivalence with the RPA Z-vector relaxed density matrix is checked against the published equations of Burow et al. (Ref. 34) and against a finite-difference dipole test (Fig. 2), i.e., an external and internally consistent verification. The numerical benchmark uses CCSDT/cc-pVQZ as an independent reference; the paper's own observation that CCSD would change the method ranking (Section IV.B) is a reference-sensitivity caveat about the strength of the performance claim, not a circularity in the derivation. Thus there is no step where a 'prediction' reduces to its input by construction.

Axiom & Free-Parameter Ledger

1 free parameters · 4 axioms · 0 invented entities

No new particles or entities are introduced. The derivations rely on standard many-body perturbation theory, the linearized-Dyson truncation, and a CCSDT reference. The only fitted/tuned parameter is the exact-exchange fraction α of the gKS starting functional, optimized against the benchmark in Fig. 5.

free parameters (1)
  • PBEh(α) exact-exchange fraction α (benchmark tuning) = varied 0–1; optimal α≈0.4 (GW) and 0.6–0.7 (idGW) from CCSDT benchmark
    The starting-point functional is scanned and optimal fractions are selected post hoc by minimizing χ1 against CCSDT (Section IVB, Fig. 5); these optimized α values are not derived.
axioms (4)
  • domain assumption The linearized Dyson equation truncated to first order in the self-energy with G_gKS propagators is adequate for density matrices
    Used throughout Sections II and III (Eqs. 4, 13, 17); no error bound is given for neglected higher-order terms.
  • domain assumption In the iterated step, only the static Hartree/exchange part of the self-energy is updated with the final density matrix; the dynamical mass operator M(ω) and mean-field orbital energies are kept fixed
    Section IIIA, Eq. (13); this distinguishes 'iterated' from fully self-consistent GW and is the definition of the proposed approximation.
  • domain assumption The density-matrix correction respects real-orbital symmetry Δγ_bj=Δγ_jb and preserves electron number through the trace of the correction
    Used to derive Eq. (21) from Eq. (20) and to claim trace conservation from Eqs. (18a, 18b).
  • domain assumption CCSDT/cc-pVQZ density is an accurate reference for the 34-molecule benchmark
    All χ1 and kinetic-energy errors are measured against CCSDT (Section IVB, Figs. 4–6); the paper itself shows CCSD would rank methods differently.

pith-pipeline@v1.3.0-alltime-deepseek · 12304 in / 18658 out tokens · 154014 ms · 2026-08-01T22:33:16.479172+00:00 · methodology

0 comments
read the original abstract

Iterating the Dyson equation with the static part of the self-energy leads to a concise and possibly improved expression of the one-body reduced density matrix from any self-energy approximation. Here we apply the procedure to Hedin's $GW$ approximation. The non-iterated $GW$ based density matrix was already known to yield accurate density matrices for molecular systems. We show that the Dyson-equation-based procedure is equivalent to the so-called variational Z-vector approach applied to the Random-Phase approximation energy functional, but only in the case of a Hartree-Fock mean-field starting point. When a generalized Kohn-Sham scheme is employed instead, the two approaches differ. By comparing the density matrix for a benchmark set of 34 small molecules to coupled-cluster reference values, we conclude that the iterated Dyson equation indeed produces improved density matrices for molecular systems. Interestingly, we observe that the excitation rank of the reference coupled-cluster matters much and that the inclusion of triple excitations (CCSDT) quantitatively changes the conclusions of the benchmark as compared to single and double excitations coupled-cluster (CCSD).

Figures

Figures reproduced from arXiv: 2607.15695 by Arno F\"orster, Erik Verzijl, Fabien Bruneval, Mauricio Rodriguez-Mayorga.

Figure 1
Figure 1. Figure 1: The contour can only be closed in the upper half-plane because of the equal-time limit. Hereafter, we write γ gKS pq for the density matrix obtained from the gKS single-particle Green’s function GgKS . We write the corresponding linearized Dyson equation: Gpq(ω) ≈ G gKS pp (ω)δpq + G gKS pp (ω) ·  ⟨p|Σx[γ gKS] − vxc[γ gKS]|q⟩ + Mpq(ω) [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Static dipole along the CO molecule axis in atomic [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Convergence rate and stability for a benchmark set [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Deviation from coupled-cluster CCSDT for the electronic density as measured by the [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Deviation from coupled-cluster CCSDT for the elec [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: In this plot, we did not reproduce the HF ki [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Deviation against coupled-cluster CCSDT for the kinetic energy for a benchmark of 34 small molecules. The mean [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

300 extracted references · 146 canonical work pages

  1. [1]

    Journal of Chemical Theory and Computation , publisher=

    Connection between GW and extended coupled cluster , author=. Journal of Chemical Theory and Computation , publisher=. 2026 , DOI=

    2026

  2. [2]

    Advances in Quantum Chemistry , volume=

    The GW approximation: A quantum chemistry perspective , author=. Advances in Quantum Chemistry , volume=. 2024 , publisher=

    2024

  3. [3]

    Nuclear Physics , volume=

    Stability conditions and nuclear rotations in the Hartree-Fock theory , author=. Nuclear Physics , volume=. 1960 , publisher=

    1960

  4. [4]

    Physical Review B , volume=

    Modified Broyden’s method for accelerating convergence in self-consistent calculations , author=. Physical Review B , volume=. 1988 , publisher=

    1988

  5. [5]

    Baerends, Evert Jan and Aguirre, Nestor F and Austin, Nick D and Autschbach, Jochen and Bickelhaupt, F Matthias and Bulo, Rosa and Cappelli, Chiara and Duin, Adri C T Van and Egidi, Franco and. J. Chem. Phys. , pages =. doi:10.1063/5.0258496 , file =

    doi:10.1063/5.0258496

  6. [6]

    Ramberger, Benjamin and Sukurma, Zoran and Sch. J. Chem. Phys. , number =. doi:10.1063/1.5128415 , eprint =

    doi:10.1063/1.5128415

  7. [7]

    Physical Review B , volume=

    GW density matrix for estimation of self-consistent GW total energies in solids , author=. Physical Review B , volume=. 2023 , publisher=

    2023

  8. [8]

    Corrections to molecular one-electron properties using M

    Amos, Roger D , journal=. Corrections to molecular one-electron properties using M. 1980 , publisher=

    1980

  9. [9]

    The Journal of Chemical Physics , volume=

    Massively parallel implementation of gradients within the random phase approximation: Application to the polymorphs of benzene , author=. The Journal of Chemical Physics , volume=. 2024 , pages=

    2024

  10. [10]

    The Journal of Chemical Physics , volume=

    Communication: Analytic gradients in the random-phase approximation , author=. The Journal of Chemical Physics , volume=. 2013 , publisher=

    2013

  11. [11]

    and Cheng, Lan and Harding, Michael E

    Matthews, Devin A. and Cheng, Lan and Harding, Michael E. and Lipparini, Filippo and Stopkowicz, Stella and Jagau, Thomas-C. and Szalay, Péter G. and Gauss, Jürgen and Stanton, John F. , title =. 2020 , month =. doi:10.1063/5.0004837 , url =

    doi:10.1063/5.0004837 2020

  12. [12]

    Fully self-consistent finite-temperature GW in Gaussian Bloch orbitals for solids , author =. Phys. Rev. B , volume =. 2022 , month =. doi:10.1103/PhysRevB.106.235104 , url =

    doi:10.1103/physrevb.106.235104 2022

  13. [13]

    RPA, an Accurate and Fast Method for the Computation of Static Nonlinear Optical Properties , journal = JCTC, volume =

    Besal. RPA, an Accurate and Fast Method for the Computation of Static Nonlinear Optical Properties , journal = JCTC, volume =. 2023 , doi =

    2023

  14. [14]

    Analytical First-Order Molecular Properties and Forces within the Adiabatic Connection Random Phase Approximation , journal = JCTC, volume =

    Burow, Asbj. Analytical First-Order Molecular Properties and Forces within the Adiabatic Connection Random Phase Approximation , journal = JCTC, volume =. 2014 , doi =

    2014

  15. [15]

    2026 , eprint=

    Connection between GW and Extended Coupled Cluster , author=. 2026 , eprint=

    2026

  16. [16]

    , TITLE=

    Bruneval, Fabien and Dattani, Nike and van Setten, Michiel J. , TITLE=. Front. Chem. , VOLUME=. 2021 , URL=. doi:10.3389/fchem.2021.749779 , ISSN=

    arXiv 2021

  17. [17]

    and Angonoa, G

    Schirmer, J. and Angonoa, G. , title =. 1989 , month =. doi:10.1063/1.457081 , url =

    work page doi:10.1063/1.457081 1989

  18. [18]

    Birgitta and Zgid, Dominika , title =

    Wen, Ming and Abraham, Vibin and Harsha, Gaurav and Shee, Avijit and Whaley, K. Birgitta and Zgid, Dominika , title =. Journal of Chemical Theory and Computation , volume =. 2024 , doi =

    2024

  19. [19]

    Ground-state properties of simple elements from GW calculations , author =. Phys. Rev. B , volume =. 2009 , month =. doi:10.1103/PhysRevB.80.041103 , url =

    work page doi:10.1103/physrevb.80.041103 2009

  20. [20]

    Self-consistent calculation of total energies of the electron gas using many-body perturbation theory , author =. Phys. Rev. B , volume =. 2001 , month =. doi:10.1103/PhysRevB.63.075112 , url =

    work page doi:10.1103/physrevb.63.075112 2001

  21. [21]

    Many-Body

    Garc\'. Many-Body. Phys. Rev. Lett. , volume =. 2002 , month =. doi:10.1103/PhysRevLett.88.056406 , url =

    work page doi:10.1103/physrevlett.88.056406 2002

  22. [22]

    Reliability and performances of real-time time-dependent auxiliary density functional theory , journal=

    Tandiana, Rika and Clavaguéra, Carine and Hasnaoui, Karim and Pedroza-Montero and Jesús Naín and de la Lande, Aurélien , year=. Reliability and performances of real-time time-dependent auxiliary density functional theory , journal=

  23. [23]

    doi:10.1088/0031-9155/60/8/r155 , url =

    Wayne D Newhauser and Rui Zhang , title =. doi:10.1088/0031-9155/60/8/r155 , url =

    work page doi:10.1088/0031-9155/60/8/r155

  24. [24]

    and Da Pieve, Fabiana and Artacho, Emilio , title =

    Koval, Natalia E. and Da Pieve, Fabiana and Artacho, Emilio , title =. Royal Society Open Science , volume =. 2020 , doi =

    2020

  25. [25]

    Correa , keywords =

    Alfredo A. Correa , keywords =. Calculating electronic stopping power in materials from first principles , journal = CMS, volume =. 2018 , issn =. doi:https://doi.org/10.1016/j.commatsci.2018.03.064 , url =

    work page doi:10.1016/j.commatsci.2018.03.064 2018

  26. [26]

    Nuclear stopping power of antiprotons , author =. Phys. Rev. A , volume =. 2017 , month =. doi:10.1103/PhysRevA.96.042717 , url =

    work page doi:10.1103/physreva.96.042717 2017

  27. [27]

    and Tran, Fabien and Marques, Miguel A

    Borlido, Pedro and Aull, Thorsten and Huran, Ahmad W. and Tran, Fabien and Marques, Miguel A. L. and Botti, Silvana , title =. 2019 , doi =

    2019

  28. [28]

    and Yao,Yi and Kanai,Yosuke , title =

    Shepard,Christopher and Zhou,Ruiyi and Yost,Dillon C. and Yao,Yi and Kanai,Yosuke , title =. 2021 , doi =

    2021

  29. [29]

    Nonperturbative treatment of multielectron processes in ion-molecule scattering: Application to He

    Sisourat, Nicolas and Pilskog, Ingjald and Dubois, Alain , journal =. Nonperturbative treatment of multielectron processes in ion-molecule scattering: Application to He. 2011 , month =. doi:10.1103/PhysRevA.84.052722 , url =

    work page doi:10.1103/physreva.84.052722 2011

  30. [30]

    2003 , note =

    Anatomy of the Barkas effect , journal = NIMB, volume =. 2003 , note =. doi:https://doi.org/10.1016/S0168-583X(03)01485-X , url =

    work page doi:10.1016/s0168-583x(03)01485-x 2003

  31. [31]

    Vanishing gap in LiF for electronic excitations by slow antiprotons , author =. Phys. Rev. B , volume =. 2009 , month =. doi:10.1103/PhysRevB.79.125107 , url =

    work page doi:10.1103/physrevb.79.125107 2009

  32. [32]

    Quantitative electronic stopping power from localized basis set , author =. Phys. Rev. B , volume =. 2020 , month =. doi:10.1103/PhysRevB.101.035136 , url =

    work page doi:10.1103/physrevb.101.035136 2020

  33. [33]

    K -Shell Core-Electron Excitations in Electronic Stopping of Protons in Water from First Principles , author =. Phys. Rev. Lett. , volume =. 2019 , month =. doi:10.1103/PhysRevLett.123.066401 , url =

    work page doi:10.1103/physrevlett.123.066401 2019

  34. [34]

    Electron Elevator: Excitations across the Band Gap via a Dynamical Gap State , author =. Phys. Rev. Lett. , volume =. 2016 , month =. doi:10.1103/PhysRevLett.116.043201 , url =

    work page doi:10.1103/physrevlett.116.043201 2016

  35. [35]

    Mass-Ratio Method Applied to the Measurement of L -Meson Masses and the Energy Balance in Pion Decay , author =. Phys. Rev. , volume =. 1956 , month =. doi:10.1103/PhysRev.101.778 , url =

    work page doi:10.1103/physrev.101.778 1956

  36. [36]

    Barkas effect in electronic stopping power: Rigorous evaluation for the harmonic oscillator , author =. Phys. Rev. A , volume =. 1989 , month =. doi:10.1103/PhysRevA.40.101 , url =

    work page doi:10.1103/physreva.40.101 1989

  37. [37]

    Electronic stopping power in a narrow band gap semiconductor from first principles , author =. Phys. Rev. B , volume =. 2015 , month =. doi:10.1103/PhysRevB.91.125203 , url =

    work page doi:10.1103/physrevb.91.125203 2015

  38. [38]

    1967 , issn =

    Electronic spectrum of crystalline lithium fluoride , journal =. 1967 , issn =. doi:https://doi.org/10.1016/0022-3697(67)90280-6 , url =

    work page doi:10.1016/0022-3697(67)90280-6 1967

  39. [39]

    Bohr , title =

    N. Bohr , title =. Phil. Mag. , volume =. 1913 , publisher =. doi:10.1080/14786440108634305 , URL =

    work page doi:10.1080/14786440108634305 1913

  40. [40]

    Stopping Power in Insulators and Metals without Charge Exchange , author =. Phys. Rev. Lett. , volume =. 2004 , month =. doi:10.1103/PhysRevLett.93.042502 , url =

    work page doi:10.1103/physrevlett.93.042502 2004

  41. [41]

    Threshold in the Stopping of Slow Protons Scattered from the Surface of a Wide-Band-Gap Insulator , author =. Phys. Rev. Lett. , volume =. 1998 , month =. doi:10.1103/PhysRevLett.81.4831 , url =

    work page doi:10.1103/physrevlett.81.4831 1998

  42. [42]

    Absence of a ``Threshold Effect'' in the Energy Loss of Slow Protons Traversing Large-Band-Gap Insulators , author =. Phys. Rev. Lett. , volume =. 1997 , month =. doi:10.1103/PhysRevLett.79.4112 , url =

    work page doi:10.1103/physrevlett.79.4112 1997

  43. [43]

    Apparent Velocity Threshold in the Electronic Stopping of Slow Hydrogen Ions in LiF , author =. Phys. Rev. Lett. , volume =. 2005 , month =. doi:10.1103/PhysRevLett.95.113201 , url =

    work page doi:10.1103/physrevlett.95.113201 2005

  44. [44]

    Vanishing Electronic Energy Loss of Very Slow Light Ions in Insulators with Large Band Gaps , author =. Phys. Rev. Lett. , volume =. 2009 , month =. doi:10.1103/PhysRevLett.103.113201 , url =

    work page doi:10.1103/physrevlett.103.113201 2009

  45. [45]

    and Rinke, Patrick , title =

    Golze, Dorothea and Wilhelm, Jan and van Setten, Michiel J. and Rinke, Patrick , title =. Journal of Chemical Theory and Computation , volume =. 2018 , doi =

    2018

  46. [46]

    and Costa, Ramon and Vi

    van Setten, Michiel J. and Costa, Ramon and Vi. Assessing GW Approaches for Predicting Core Level Binding Energies , journal=JCTC, year=. doi:10.1021/acs.jctc.7b01192 , url=

    doi:10.1021/acs.jctc.7b01192

  47. [47]

    Weigend, Florian and Ahlrichs, Reinhart , title =. Phys. Chem. Chem. Phys. , volume =. 2005 , doi =

    2005

  48. [48]

    2019 , month=

    Bruneval, Fabien , title=. 2019 , month=. doi:10.1021/acs.jctc.9b00333 , url=

    doi:10.1021/acs.jctc.9b00333 2019

  49. [49]

    Focal Point Analysis of the Singlet-Triplet Energy Gap of Octacene and Larger Acenes , journal=JPCA, year=

    Hajgat. Focal Point Analysis of the Singlet-Triplet Energy Gap of Octacene and Larger Acenes , journal=JPCA, year=. doi:10.1021/jp2043043 , url=

    work page doi:10.1021/jp2043043

  50. [50]

    Host dependence of the electron affinity of molecular dopants

    Li, Jing and Duchemin, Ivan and Roscioni, Otello Maria and Friederich, Pascal and Anderson, Marie and Da Como, Enrico and Kociok-Köhn, Gabriele and Wenzel, Wolfgang and Zannoni, Claudio and Beljonne, David and Blase, Xavier and D ' Avino, Gabriele. Host dependence of the electron affinity of molecular dopants. Mater. Horiz. 2019. doi:10.1039/C8MH00921J

    work page doi:10.1039/c8mh00921j 2019

  51. [51]

    , title =

    Krylov, Anna I. , title =. Annual Review of Physical Chemistry , volume =. 2008 , doi =

    2008

  52. [52]

    New approach to the one-particle Green's function for finite Fermi systems , author =. Phys. Rev. A , volume =. 1983 , month =. doi:10.1103/PhysRevA.28.1237 , url =

    work page doi:10.1103/physreva.28.1237 1983

  53. [53]

    J. V. Ortiz , title =. Computational Chemistry: Reviews of Current Trends , chapter =. doi:10.1142/9789812812148_0001 , year=

    work page doi:10.1142/9789812812148_0001

  54. [54]

    and Marshall, Michael S

    Richard, Ryan M. and Marshall, Michael S. and Dolgounitcheva, O. and Ortiz, J. V. and Br. Accurate Ionization Potentials and Electron Affinities of Acceptor Molecules I. Reference Data at the CCSD(T) Complete Basis Set Limit , journal=JCTC, year=. doi:10.1021/acs.jctc.5b00875 , url=

    doi:10.1021/acs.jctc.5b00875

  55. [55]

    Generalized Neural-Network Representation of High-Dimensional Potential-Energy Surfaces , author =. Phys. Rev. Lett. , volume =. 2007 , month =. doi:10.1103/PhysRevLett.98.146401 , url =

    doi:10.1103/physrevlett.98.146401 2007

  56. [56]

    doi:10.1088/2515-7639/ab13bb , year = 2019, month =

    Claudia Draxl and Matthias Scheffler , title =. doi:10.1088/2515-7639/ab13bb , year = 2019, month =

    doi:10.1088/2515-7639/ab13bb 2019

  57. [57]

    Big Data of Materials Science: Critical Role of the Descriptor , author =. Phys. Rev. Lett. , volume =. 2015 , month =. doi:10.1103/PhysRevLett.114.105503 , url =

    work page doi:10.1103/physrevlett.114.105503 2015

  58. [58]

    Jain, Anubhav and Ong, Shyue Ping and Hautier, Geoffroy and Chen, Wei and Richards, William Davidson and Dacek, Stephen and Cholia, Shreyas and Gunter, Dan and Skinner, David and Ceder, Gerbrand and Persson, Kristin a. , doi =. APL Materials , number =

  59. [59]

    Goryaeva and Jean-Bernard Maillet and Mihai-Cosmin Marinica

    Alexandra M. Goryaeva and Jean-Bernard Maillet and Mihai-Cosmin Marinica. Towards better efficiency of interatomic linear machine learning potentials. 2019. doi:https://doi.org/10.1016/j.commatsci.2019.04.043

    work page doi:10.1016/j.commatsci.2019.04.043 2019

  60. [60]

    Machine learning unifies the modeling of materials and molecules , volume =

    Bart. Machine learning unifies the modeling of materials and molecules , volume =. 2017 , doi =. https://advances.sciencemag.org/content/3/12/e1701816.full.pdf , journal =

    2017

  61. [61]

    and Boukouvalas, Zois and Butrico, Mark S

    Elton, Daniel C. and Boukouvalas, Zois and Butrico, Mark S. and Fuge, Mark D. and Chung, Peter W. , title=. Scientific Reports , year=. doi:10.1038/s41598-018-27344-x , url=

    doi:10.1038/s41598-018-27344-x

  62. [62]

    and Cs \' a nyi, G \' a bor and Ceriotti, Michele

    De, Sandip and Bart \' o k, Albert P. and Cs \' a nyi, G \' a bor and Ceriotti, Michele. Comparing molecules and solids across structural and alchemical space. Phys. Chem. Chem. Phys. 2016. doi:10.1039/C6CP00415F

    doi:10.1039/c6cp00415f 2016

  63. [63]

    Unifying machine learning and quantum chemistry with a deep neural network for molecular wavefunctions , journal=

    Sch. Unifying machine learning and quantum chemistry with a deep neural network for molecular wavefunctions , journal=. 2019 , volume=. doi:10.1038/s41467-019-12875-2 , url=

    doi:10.1038/s41467-019-12875-2 2019

  64. [64]

    and Reuter, Karsten and Rinke, Patrick and Oberhofer, Harald , title=

    Stuke, Annika and Kunkel, Christian and Golze, Dorothea and Todorovic, Milica and Margraf, Johannes T. and Reuter, Karsten and Rinke, Patrick and Oberhofer, Harald , title=. Scientific Data , year=. doi:10.1038/s41597-020-0385-y , url=

    doi:10.1038/s41597-020-0385-y

  65. [65]

    Journal of Chemical Theory and Computation , volume =

    Loos, Pierre-François and Pradines, Barthélémy and Scemama, Anthony and Giner, Emmanuel and Toulouse, Julien , title =. Journal of Chemical Theory and Computation , volume =. 2020 , doi =

    2020

  66. [66]

    Toward GW Calculations on Thousands of Atoms , journal =

    Wilhelm, Jan and Golze, Dorothea and Talirz, Leopold and Hutter, J. Toward GW Calculations on Thousands of Atoms , journal =. 2018 , doi =

    2018

  67. [67]

    Frontiers in Chemistry , VOLUME=

    Golze, Dorothea and Dvorak, Marc and Rinke, Patrick , TITLE=. Frontiers in Chemistry , VOLUME=. 2019 , URL=. doi:10.3389/fchem.2019.00377 , ISSN=

    arXiv 2019

  68. [68]

    Effect of ladder diagrams on optical absorption spectra in a quasiparticle self-consistent

    Cunningham, Brian and Gr\"uning, Myrta and Azarhoosh, Pooya and Pashov, Dimitar and van Schilfgaarde, Mark , journal =. Effect of ladder diagrams on optical absorption spectra in a quasiparticle self-consistent. 2018 , month =. doi:10.1103/PhysRevMaterials.2.034603 , url =

    work page doi:10.1103/physrevmaterials.2.034603 2018

  69. [69]

    and Harding, M

    Kaplan, F. and Harding, M. E. and Seiler, C. and Weigend, F. and Evers, F. and van Setten, M. J. , title =. Journal of Chemical Theory and Computation , volume =. 2016 , doi =

    2016

  70. [70]

    and VandeVondele, Joost , title =

    Rybkin, Vladimir V. and VandeVondele, Joost , title =. Journal of Chemical Theory and Computation , volume =. 2016 , doi =

    2016

  71. [71]

    Distributed memory parallel implementation of energies and gradients for second-order Møller–Plesset perturbation theory with the resolution-of-the-identity approximation

    H \" a ttig, Christof and Hellweg, Arnim and Köhn, Andreas. Distributed memory parallel implementation of energies and gradients for second-order Møller–Plesset perturbation theory with the resolution-of-the-identity approximation. Phys. Chem. Chem. Phys. 2006. doi:10.1039/B515355G

    work page doi:10.1039/b515355g 2006

  72. [72]

    Beyond the Random-Phase Approximation for the Electron Correlation Energy: The Importance of Single Excitations , author =. Phys. Rev. Lett. , volume =. 2011 , month =. doi:10.1103/PhysRevLett.106.153003 , url =

    work page doi:10.1103/physrevlett.106.153003 2011

  73. [73]

    Random phase approximation up to the melting point: Impact of anharmonicity and nonlocal many-body effects on the thermodynamics of Au , author =. Phys. Rev. B , volume =. 2015 , month =. doi:10.1103/PhysRevB.91.201103 , url =

    work page doi:10.1103/physrevb.91.201103 2015

  74. [74]

    Melting Si: Beyond Density Functional Theory , author =. Phys. Rev. Lett. , volume =. 2018 , month =. doi:10.1103/PhysRevLett.121.195701 , url =

    work page doi:10.1103/physrevlett.121.195701 2018

  75. [75]

    Ab initio Calculation of van der Waals Bonded Molecular Crystals , author =. Phys. Rev. Lett. , volume =. 2009 , month =. doi:10.1103/PhysRevLett.102.206411 , url =

    work page doi:10.1103/physrevlett.102.206411 2009

  76. [76]

    and Henderson,Thomas M

    Scuseria,Gustavo E. and Henderson,Thomas M. and Sorensen,Danny C. , title =. 2008 , doi =

    2008

  77. [77]

    Static correlation and electron localization in molecular dimers from the self-consistent RPA and GW approximation , author =. Phys. Rev. B , volume =. 2015 , month =. doi:10.1103/PhysRevB.91.165110 , url =

    work page doi:10.1103/physrevb.91.165110 2015

  78. [78]

    Wiley Interdisciplinary Reviews: Computational Molecular Science , volume =

    Reining, Lucia , title =. Wiley Interdisciplinary Reviews: Computational Molecular Science , volume =. doi:10.1002/wcms.1344 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1002/wcms.1344 , year =

    doi:10.1002/wcms.1344

  79. [79]

    Improved density matrices for accurate molecular ionization potentials , author =. Phys. Rev. B , volume =. 2019 , month =. doi:10.1103/PhysRevB.99.041118 , url =

    doi:10.1103/physrevb.99.041118 2019

  80. [80]

    and Schaefer,Henry F

    Handy,Nicholas C. and Schaefer,Henry F. , title =. The Journal of Chemical Physics , volume =. 1984 , doi =

    1984

Showing first 80 references.