pith. sign in

arxiv: 2605.05675 · v1 · submitted 2026-05-07 · ❄️ cond-mat.mtrl-sci · physics.chem-ph· physics.comp-ph

A Scalable Translationally Invariant Variational Theory of Ab Initio Polarons

Moritz K. A. Baumgarten , Hamlin Wu , Tong Jiang , Joonho Lee This is my paper

Pith reviewed 2026-05-08 08:55 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci physics.chem-phphysics.comp-ph
keywords polaronelectron-phonon couplingvariational wavefunctionthermodynamic limitab initioLiFTiO2Fröhlich model
0
0 comments X

The pith

Momentum-projected variational wavefunctions enable scalable ab initio polaron calculations in the thermodynamic limit.

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

The paper develops a variational method for calculating properties of polarons in real materials that maintains translational invariance without supercells. It combines momentum-projected Toyozawa-type wavefunctions with low-rank factorization of the electron-phonon kernel to achieve near-linear scaling with k-points while treating both delocalized and self-trapped carriers. Benchmarks on the Fröhlich model, LiF, and TiO2 produce accurate binding energies and band structures, with agreement to diagrammatic Monte Carlo in weak coupling but indication of bias in the Monte Carlo results for strong coupling. A reader would care because this supplies a practical, improvable route to polaron studies at the natural size limit of materials where charge transport and optical behavior are determined.

Core claim

The authors establish that their momentum-projected Toyozawa-type variational wavefunction, paired with low-rank factorization of the electron-phonon kernel, yields a translationally invariant and scalable approach to computing ab initio polaron properties. This produces reliable binding energies, thermodynamic-limit band structures, and real-space polaron sizes for materials including LiF and TiO2, and highlights a potential bias in diagrammatic Monte Carlo for the strong-coupling hole polaron in LiF.

What carries the argument

momentum-projected Toyozawa-type wavefunction with low-rank factorization of the electron-phonon kernel, which enforces translational invariance and near-linear scaling with k-points

If this is right

  • Polaron binding energies and band structures become available for real materials directly in the thermodynamic limit.
  • Both weak- and strong-coupling regimes are handled within the same variational framework without separate approximations.
  • Real-space measures of polaron extent provide direct, interpretable quantities for comparison with experiment.
  • The approach serves as a benchmark that can identify biases in other techniques such as diagrammatic Monte Carlo.
  • Systematic improvement follows from enlarging the variational space or refining the kernel factorization.

Where Pith is reading between the lines

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

  • The near-linear scaling suggests the method can be applied to larger or lower-symmetry unit cells that remain inaccessible to supercell techniques.
  • The noted discrepancy with diagrammatic Monte Carlo implies variational methods may become the preferred route for strong-coupling polarons where stochastic sampling faces bias issues.
  • Transparent real-space polaron sizes could be used to predict signatures in angle-resolved photoemission or optical spectra.
  • The same wavefunction form offers a natural starting point for computing polaron mobilities or finite-temperature effects in the same materials.

Load-bearing premise

The low-rank factorization of the electron-phonon kernel combined with the Toyozawa-type ansatz remains accurate and unbiased in the strong-coupling regime.

What would settle it

A converged reference calculation of the LiF hole-polaron binding energy via diagrammatic Monte Carlo or exact methods on extrapolatable clusters that differs substantially from the variational result would falsify the claim of unbiased accuracy in strong coupling.

Figures

Figures reproduced from arXiv: 2605.05675 by Hamlin Wu, Joonho Lee, Moritz K. A. Baumgarten, Tong Jiang.

Figure 1
Figure 1. Figure 1: FIG. 1. Ground state energy of the Fröhlich polaron for nu view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Size extrapolation for the small (a) hole polaron in LiF and (b) electron polaron in rutile, and large electron polarons view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Visualization of the density–displacement correlator for (a) the LiF hole polaron along the [010]-[001] plane, and for view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. (a) Band structure of the LiF electron polaron. The view at source ↗
read the original abstract

We introduce a scalable, translationally invariant variational theory for ab initio polarons that remains applicable across coupling regimes without resorting to supercells. Our approach combines a momentum-projected Toyozawa-type wavefunction with a low-rank factorization of the electron-phonon kernel, enabling near-linear scaling with the number of $\mathbf{k}$-points while capturing both delocalized and self-trapped carriers. Benchmarks for the Fr\"ohlich model, LiF, and anatase and rutile TiO$_2$ yield accurate polaron binding energies, thermodynamic-limit band structures, and transparent real-space measures of polaron extent. For LiF, comparison with first-principles diagrammatic Monte Carlo (DiagMC) reveals close agreement for the weak-coupling electron-polaron ground state and band structure. However, in the hole-polaron of LiF, which is in the strong-coupling regime, we found a significant bias in DiagMC results. These results establish momentum-projected variational wavefunctions as a systematically improvable route to thermodynamic limit studies of polarons in real materials.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces a scalable, translationally invariant variational method for ab initio polarons that combines a momentum-projected Toyozawa-type wavefunction with low-rank factorization of the electron-phonon kernel. This enables near-linear scaling with k-points and applicability across coupling strengths without supercells. Benchmarks on the Fröhlich model, LiF (electron and hole polarons), and TiO2 (anatase/rutile) report polaron binding energies, thermodynamic-limit band structures, and real-space extents, with close agreement to DiagMC for the weak-coupling LiF electron polaron but a claimed significant DiagMC bias for the strong-coupling hole polaron.

Significance. If the ansatz completeness and low-rank factorization accuracy hold, the approach offers a systematically improvable route to thermodynamic-limit polaron studies in real materials, addressing a key limitation of supercell-based methods. The variational character and explicit scaling claims are strengths, as is the provision of transparent real-space diagnostics.

major comments (2)
  1. [LiF hole polaron results section] § on LiF hole polaron results (near abstract claim of 'significant bias in DiagMC'): The attribution of the discrepancy solely to DiagMC bias is load-bearing for the claim of applicability 'across coupling regimes,' yet no explicit convergence data with respect to factorization rank, number of variational parameters, or additional correlation channels is provided to rule out ansatz incompleteness. Because the method is variational, any missing channels produce a systematic upper-bound error whose magnitude remains uncontrolled precisely where the discrepancy appears.
  2. [Method / low-rank factorization subsection] Low-rank factorization of the e-ph kernel (method section): The paper asserts that the factorization remains accurate and unbiased in the strong-coupling regime, but no independent validation (e.g., comparison to known strong-coupling analytic limits, larger-rank calculations, or energy lowering upon rank increase) is shown for the LiF hole case. This leaves open whether the reported energies are converged or limited by the rank truncation.
minor comments (2)
  1. [Wavefunction ansatz definition] The original Toyozawa ansatz reference is cited but the precise momentum-projection implementation and its relation to the low-rank kernel could be clarified with an explicit equation for the projected wavefunction.
  2. [Figures on real-space measures] Figure captions for real-space polaron extent plots should include the specific k-point sampling density used to reach the thermodynamic limit.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and insightful comments on our manuscript. We appreciate the opportunity to clarify and strengthen our presentation of the results, particularly regarding the convergence and validation of the method in the strong-coupling regime. Below, we provide point-by-point responses to the major comments.

read point-by-point responses

  1. Referee: [LiF hole polaron results section] § on LiF hole polaron results (near abstract claim of 'significant bias in DiagMC'): The attribution of the discrepancy solely to DiagMC bias is load-bearing for the claim of applicability 'across coupling regimes,' yet no explicit convergence data with respect to factorization rank, number of variational parameters, or additional correlation channels is provided to rule out ansatz incompleteness. Because the method is variational, any missing channels produce a systematic upper-bound error whose magnitude remains uncontrolled precisely where the discrepancy appears.

    Authors: We agree that explicit convergence studies are essential to robustly attribute the observed discrepancy to a bias in the DiagMC results rather than incompleteness in our variational ansatz. In the revised version of the manuscript, we will include additional data demonstrating the convergence of the LiF hole polaron binding energy with respect to the factorization rank and the number of variational parameters. We will also address the potential for additional correlation channels and their impact on the energy. These additions will provide better control over the variational upper bound and support our claim of applicability across coupling regimes, especially given the excellent agreement with DiagMC in the weak-coupling electron polaron case. revision: yes

  2. Referee: [Method / low-rank factorization subsection] Low-rank factorization of the e-ph kernel (method section): The paper asserts that the factorization remains accurate and unbiased in the strong-coupling regime, but no independent validation (e.g., comparison to known strong-coupling analytic limits, larger-rank calculations, or energy lowering upon rank increase) is shown for the LiF hole case. This leaves open whether the reported energies are converged or limited by the rank truncation.

    Authors: We acknowledge the need for more direct validation of the low-rank factorization in the strong-coupling regime for the LiF hole polaron. While analytic strong-coupling limits are not straightforward to apply to ab initio systems like LiF, we will add to the revised manuscript results from calculations with increased factorization ranks for the LiF hole polaron. These will show the energy lowering upon rank increase and demonstrate convergence, thereby confirming that the reported energies are not limited by the rank truncation and that the factorization remains accurate. revision: yes

Circularity Check

0 steps flagged

No circularity in variational ansatz or benchmarks

full rationale

The paper introduces an explicitly variational ansatz (momentum-projected Toyozawa-type wavefunction plus low-rank kernel factorization) whose energy is obtained by direct minimization; this construction does not reduce to its own inputs by definition, nor does it rename fitted parameters as predictions. Benchmarks are performed against independent external methods (DiagMC, Fröhlich model limits) rather than self-referential data. No load-bearing self-citation, uniqueness theorem imported from the authors' prior work, or ansatz smuggled via citation appears in the provided text; the discrepancy with DiagMC is presented as an empirical observation whose interpretation does not alter the internal derivation of the variational energies or band structures.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Review performed on abstract only; no explicit free parameters, axioms, or invented entities are stated. The approach rests on standard variational principles and electron-phonon coupling from prior literature.

pith-pipeline@v0.9.0 · 5506 in / 1111 out tokens · 63117 ms · 2026-05-08T08:55:05.971747+00:00 · methodology

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

64 extracted references · 64 canonical work pages

  1. [1]

    L. D. Landau, Phys. Z. Sowjet.3, 664 (1933)

    work page 1933

  2. [2]

    Giustino, Rev

    F. Giustino, Rev. Mod. Phys.89, 015003 (2017)

    work page 2017

  3. [3]

    Bardeen, L

    J. Bardeen, L. N. Cooper, and J. R. Schrieffer, Phys. Rev.108, 1175 (1957)

    work page 1957

  4. [4]

    Marsiglio, Ann

    F. Marsiglio, Ann. Phys. (N.Y.)417, 168102 (2020)

    work page 2020

  5. [5]

    H. Y. Fan, Phys. Rev.82, 900 (1951)

    work page 1951

  6. [6]

    A. B. Migdal, Sov. Phys. JETP7, 996 (1958)

    work page 1958

  7. [7]

    P. B. Allen and V. Heine, J. Phys. C: Solid State Phys. 9, 2305 (1976)

    work page 1976

  8. [8]

    P. B. Allen and M. Cardona, Phys. Rev. B23, 1495 (1981)

    work page 1981

  9. [9]

    S. M. Story, J. J. Kas, F. D. Vila, M. J. Verstraete, and J. J. Rehr, Phys. Rev. B90, 195135 (2014)

    work page 2014

  10. [10]

    Lihm and S

    J.-M. Lihm and S. Poncé, Phys. Rev. Lett.134, 186401 (2025)

    work page 2025

  11. [11]

    W. H. Sio, C. Verdi, S. Poncé, and F. Giustino, Phys. Rev. Lett.122, 246403 (2019)

    work page 2019

  12. [12]

    W. H. Sio, C. Verdi, S. Poncé, and F. Giustino, Phys. Rev. B99, 235139 (2019)

    work page 2019

  13. [13]

    Y. Luo, B. K. Chang, and M. Bernardi, Phys. Rev. B 105, 155132 (2022)

    work page 2022

  14. [14]

    A.Mahajan, P.J.Robinson, J.Lee, andD.R.Reichman, Phys. Rev. B112, 184310 (2025)

    work page 2025

  15. [15]

    P.J.Robinson, J.Lee, A.Mahajan, andD.R.Reichman, Phys. Rev. B111, 054304 (2025)

    work page 2025

  16. [16]

    Lafuente-Bartolome, C

    J. Lafuente-Bartolome, C. Lian, W. H. Sio, I. G. Gur- tubay, A. Eiguren, and F. Giustino, Phys. Rev. B106, 075119 (2022)

    work page 2022

  17. [17]

    Lafuente-Bartolome, C

    J. Lafuente-Bartolome, C. Lian, W. H. Sio, I. G. Gur- tubay, A. Eiguren, and F. Giustino, Phys. Rev. Lett. 129, 076402 (2022)

    work page 2022

  18. [18]

    Y. Luo, J. Park, and M. Bernardi, Nat. Phys.21, 1275 (2025)

    work page 2025

  19. [19]

    Carlström, Phys

    J. Carlström, Phys. Rev. B103, 195147 (2021)

    work page 2021

  20. [20]

    Toyozawa, Prog

    Y. Toyozawa, Prog. Theor. Phys.26, 29 (1961)

    work page 1961

  21. [21]

    Y. Zhao, J. Chem. Phys.158, 080901 (2023)

    work page 2023

  22. [22]

    Shih and T

    P. Shih and T. C. Berkelbach, J. Chem. Phys.160, 204705 (2024)

    work page 2024

  23. [23]

    Y. Luo, D. Desai, B. K. Chang, J. Park, and M. Bernardi, Phys. Rev. X14, 021023 (2024)

    work page 2024

  24. [24]

    Franchini, M

    C. Franchini, M. Reticcioli, M. Setvin, and U. Diebold, Nat. Rev. Mater.6, 560 (2021)

    work page 2021

  25. [25]

    L. D. Landau and S. I. Pekar, Zh. Eksp. Teor. Fiz.18, 419 (1948)

    work page 1948

  26. [26]

    R. E. Peierls and J. Yoccoz, Proc. Phys. Soc. A70, 381 (1957)

    work page 1957

  27. [27]

    T. Hahn, S. Klimin, J. Tempere, J. T. Devreese, and C. Franchini, Phys. Rev. B97, 134305 (2018)

    work page 2018

  28. [28]

    Nagy and P

    P. Nagy and P. Markoš, Phys. Status Solidi B151, 571 (1989)

    work page 1989

  29. [29]

    V. M. Buimistrov and S. I. Pekar, Sov. Phys. JETP6, 977 (1958)

    work page 1958

  30. [30]

    A. S. Davydov, J. Theor. Biol.38, 559 (1973)

    work page 1973

  31. [31]

    Y. Zhao, D. W. Brown, and K. Lindenberg, J. Chem. Phys.107, 3159 (1997)

    work page 1997

  32. [32]

    Griewank, Pesqui

    A. Griewank, Pesqui. Oper.34, 621 (2014)

    work page 2014

  33. [33]

    Giustino, M

    F. Giustino, M. L. Cohen, and S. G. Louie, Phys. Rev. B76, 165108 (2007)

    work page 2007

  34. [34]

    Vogl, Phys

    P. Vogl, Phys. Rev. B13, 694 (1976)

    work page 1976

  35. [35]

    Verdi and F

    C. Verdi and F. Giustino, Phys. Rev. Lett.115, 176401 (2015)

    work page 2015

  36. [36]

    Park, J.-J

    J. Park, J.-J. Zhou, V. A. Jhalani, C. E. Dreyer, and M. Bernardi, Phys. Rev. B102, 125203 (2020)

    work page 2020

  37. [37]

    Makov and M

    G. Makov and M. C. Payne, Phys. Rev. B51, 4014 (1995)

    work page 1995

  38. [38]

    J. Lee, S. Zhang, and D. R. Reichman, Phys. Rev. B 103, 115123 (2021)

    work page 2021

  39. [39]

    Dai and F

    Z. Dai and F. Giustino, Proc. Natl. Acad. Sci. U.S.A. 121, e2414203121 (2024)

    work page 2024

  40. [40]

    P. O. J. Scherer and S. F. Fischer, Chem. Phys.86, 269 (1984)

    work page 1984

  41. [41]

    A. H. Romero, D. W. Brown, and K. Lindenberg, Phys. Lett. A254, 287 (1999)

    work page 1999

  42. [42]

    Hoffmann and Z

    M. Hoffmann and Z. G. Soos, Phys. Rev. B66, 024305 (2002)

    work page 2002

  43. [43]

    F. Liu, Z. Yang, Y. Luo, S. Guo, C. Zhang, S. Choo, X. Xu, S. G. Jeong, J. S. Kumar, X. Wang, A. Mkhoyan, M. Bernardi, and B. Jalan, Rep. Prog. Phys.89, 028003 (2026)

    work page 2026

  44. [44]

    Epifanovsky, A

    E. Epifanovsky, A. T. B. Gilbert, X. Feng, J. Lee, Y. Mao, N. Mardirossian, P. Pokhilko, A. F. White, M. P. Coons, A. L. Dempwolff, Z. Gan, D. Hait, P. R. Horn, L. D. Jacobson, I. Kaliman, J. Kussmann, A. W. Lange, K. U. Lao, D. S. Levine, J. Liu, S. C. McKen- zie, A. F. Morrison, K. D. Nanda, F. Plasser, D. R. Rehn, M. L. Vidal, Z.-Q. You, Y. Zhu, B. Ala...

    work page 2021

  45. [45]

    Van Voorhis and M

    T. Van Voorhis and M. Head-Gordon, Mol. Phys.100, 1713 (2002)

    work page 2002

  46. [46]

    Giannozzi, S

    P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococ- cioni, I. Dabo, A. Dal Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. S...

    work page 2009

  47. [47]

    J.-J. Zhou, J. Park, I.-T. Lu, I. Maliyov, X. Tong, and M. Bernardi, Comput. Phys. Commun.264, 107970 (2021)

    work page 2021

  48. [48]

    Poncé, E

    S. Poncé, E. R. Margine, C. Verdi, and F. Giustino, Comput. Phys. Commun.209, 116 (2016)

    work page 2016

  49. [49]

    Sjakste, N

    J. Sjakste, N. Vast, M. Calandra, and F. Mauri, Phys. Rev. B92, 054307 (2015)

    work page 2015

  50. [50]

    Schlipf and F

    M. Schlipf and F. Gygi, Comput. Phys. Commun.196, 36 (2015)

    work page 2015

  51. [51]

    D. R. Hamann, Phys. Rev. B88, 085117 (2013)

    work page 2013

  52. [52]

    Pizzi, V

    G. Pizzi, V. Vitale, R. Arita, S. Blügel, F. Freimuth, G. Géranton, M. Gibertini, D. Gresch, C. Johnson, T. Koretsune, J. Ibañez-Azpiroz, H. Lee, J.-M. Lihm, D. Marchand, A. Marrazzo, Y. Mokrousov, J. I. Mustafa, Y. Nohara, Y. Nomura, L. Paulatto, S. Poncé, T. Pon- weiser, J. Qiao, F. Thöle, S. S. Tsirkin, M. Wierzbowska, N. Marzari, D. Vanderbilt, I. Sou...

    work page 2020

  53. [53]

    By localizing the electron at a specific origin, the LP state breaks the translational invariance of the underlying Hamiltonian

    The Broken Symmetry of Strong Coupling In the strong-coupling regime (α≫1), the standard variational approach is the Landau–Pekar (LP) product ansatz: |ΨLP⟩=|ψ e⟩ ⊗ |f⟩(A4) where|ψ e⟩is a localized electron state (often modeled as a Gaussian,ψe(r)∝e − 1 2 µr2 ), and|f⟩= exp( P q fqa† q − f ∗ qaq)|0⟩is a coherent state representing the lattice deformation....

    work page

  54. [54]

    Since the Hamiltonian is translationally invariant, states localized at different positionsRmust be degenerate

    The Peierls–Yoccoz Projection To restore translational symmetry, we employ the Peierls–Yoccoz (PY) projection operator. Since the Hamiltonian is translationally invariant, states localized at different positionsRmust be degenerate. We construct a momentum eigenstate by integrating over all continuous spatial translations of the localized LP wavepacket: |Ψ...

    work page

  55. [55]

    Variation After Projection (VAP) To correctly capture the crossover regime (α≈6), one must use the Variation After Projection (VAP) method, as opposed to the projection after variation method. In VAP, the exact expectation value of the Hamiltonian is evaluated using the fully projected state: EPY = ⟨ΨPY(0)|H|ΨPY(0)⟩ ⟨ΨPY(0)|ΨPY(0)⟩ (A7) 9 This yields a mu...

    work page

  56. [56]

    Analytic Matrix Elements The localized electronic trial state is constructed as a linear combination of Gaussians,ϕ(r) =P3 i=1 cie−µir2/2. Because the generator coordinate shiftRacts as a simple translation, the overlap and kinetic energy matrix elements between the localized state and its shifted counterpartϕR(r) =ϕ(r−R)can be evaluated completely analyt...

    work page

  57. [57]

    Numerical Quadrature for Phonon and Interaction Matrix Elements The localized trial state assumes the lattice relaxes into a coherent state characterized by the displacement ampli- tudesf q. This displacement is phenomenologically tied to the exact electronic density via a single momentum-space low-pass filter: fq =− Vq ℏωq ρq 1 1 +βq 2 ≡ − Vq ℏωq ρqgq,(A...

    work page

  58. [58]

    As the polaron localizes,µ→ ∞, causing the physical radius to shrink drastically

    Scale-Invariant Quadrature The final PY energy requires integrating over the generator coordinateR: EPY = R dR R2 exp(G(R)) [Tel(R) +Iel(R)Uph(R) +Vint(R)]R dR R2 exp(G(R))Iel(R) .(A22) 11 A naive integration overRandqfails in the strong-coupling limit (α≫1). As the polaron localizes,µ→ ∞, causing the physical radius to shrink drastically. This causes the...

    work page

  59. [59]

    For a non-magnetic crystal, dynamical matrices (and therefore phonon eigenmodes) obeyD(q) =D∗(−q)

    Alignment of Degenerate Modes We found that in order to converge our variational calculations efficiently and produce a reliable finite-size ex- trapolation degenerate modes need to be aligned betweenqand−qwhen generating the electron–phonon coupling elements. For a non-magnetic crystal, dynamical matrices (and therefore phonon eigenmodes) obeyD(q) =D∗(−q...

    work page

  60. [60]

    Re{Ank} εnk −E dD2(A, B) Dk + Re{Ank} X νq ωνq|Bνq|2Dk+q + Re (X mνq gnmν(k−q,q)A mk−qB∗ ν,−qDk−q + X mνq gmnν(k,q)A ∗ mk+qB∗ ν,−qDk )# ,(G2a) ∂EdD2 ∂Im{A nk} = 2⟨Ψ|Ψ⟩−1

    Gradient We derive the analytical gradients for the dD2 wavefunction below. The corresponding D2 expressions are obtained immediately by settingD k = 1. These derivatives assume a Hermitian electron–phonon coupling tensor. If the full tensorg mnν(k,q)is stored explicitly, hermiticity can be enforced directly. Otherwise, for example when working with singu...

    work page

  61. [61]

    εnk −E dD2(A, B)−Re{A nk} ∂EdD2(A, B) ∂Re{A nk} ! Dk + X νq ωνq|Bνq|2Dk+q + 2 Re (X ν′ gnnν′(k,0)B ∗ ν′0 Dk )# . (G4) ∂2EdD2 ∂Im{A nk}2 = 2⟨Ψ|Ψ⟩−1

    Diagonal Elements of Hessian For approximate second-order methods such as the GDM, the diagonal elements of the Hessian are useful for preconditioning. In the followings, we list necessary equations for implementing the diagonal elements of the Hessian. ∂2EdD2 ∂Re{A nk}2 = 2⟨Ψ|Ψ⟩−1 " εnk −E dD2(A, B)−Re{A nk} ∂EdD2(A, B) ∂Re{A nk} ! Dk + X νq ωνq|Bνq|2Dk+...

    work page

  62. [62]

    Phonon Occupations We compute the average number of phonons⟨Nph⟩in the ground state as well as the second central moment of the corresponding probability distribution: Nph =⟨Ψ|Ψ⟩ −1  X n eiK·Rne P νq |Bνq |2e−iq·Rn  X jk |Ajk|2e−ik·Rn    X ν′q′ |Bν′q′|2e−iq′·Rn     (H1) Nph − Nph⟩ 2 =⟨Ψ|Ψ⟩ −1   X n eiK·Rne P νq |Bνq |2e−iq·Rn  X jk |Ajk|2...

    work page

  63. [63]

    Electronic densities are shown in Fig

    Carrier Occupations To further characterize the ground states obtained, we compute the momentum-resolved electronic densities, ⟨c† nkcnk⟩=⟨Ψ|Ψ⟩ −1 |Ank|2Dk.(H4) Since ourk-mesh is uniformly discretized, we use cubic splines to subsample along the high-symmetry points within the unit cell. Electronic densities are shown in Fig. H2. 21 0 5 10 150.00 0.05 0....

    work page

  64. [64]

    Polaron Extent In order to make a prediction on the polaron size of a system we investigate the density-displacement correlation dn,κα(Rp) = *X Re ˆnnRe ˆuκα Re+Rp + ,(H5) where ˆnnRe =c † nRe cnRe = 1 Nk X k1k2j1j2 e−iRe·(k1−k2)U ∗ j1n(k1)Uj2n(k2)c† j1k1 cj2k2 (H6) ˆuκα Re+Rp = X νq s ℏ 2Mκωνq eν κα(q)eiq·(Re+Rp) √Nk bνq +b † ν−q .(H7) We can sum outRe, ...

    work page