Phonon-Induced Band Renormalization

• Phonon-induced band renormalization is a phenomenon where electron self-energies from lattice vibrations shift electronic energies, alter dispersions, and modify quasiparticle lifetimes.
• The topic is examined using various frameworks—ranging from Allen-Heine-Cardona to self-consistent Green’s function approaches—addressing both weak- and strong-coupling regimes.
• Experimental methods like ARPES reveal manifestations such as kinks, replica bands, and pseudogaps that validate the theory’s predictions on energy shifts and spectral modifications.

Searching arXiv for recent and foundational papers on phonon-induced band renormalization. Phonon-induced band renormalization is the modification of electronic energies, dispersions, spectral weights, lifetimes, and, in some formulations, wave functions by coupling to lattice vibrations. In crystalline solids this coupling appears through an electron self-energy and produces phenomena that range from zero-point and finite-temperature band-gap shifts to quasiparticle kinks, replica bands, pseudogaps, linewidth broadening, and polaronic self-trapping. Across contemporary work, the subject is treated within perturbative Allen-Heine-Cardona and Williams-Lax formalisms, fully dynamical Fan-Migdal calculations, generalized Debye-Waller theories, and self-consistent Green’s-function approaches that connect weak-coupling band renormalization to strong-coupling polaron physics (Lihm et al., 2020, Zacharias et al., 2016, Lafuente-Bartolome et al., 2022).

1. Formal definition and self-energy structure

At the formal level, phonon-induced band renormalization is encoded in the electron self-energy $\Sigma$, commonly decomposed into Fan and Debye-Waller contributions,

$$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$

so that the electronic Hamiltonian is replaced by an effective, frequency-dependent operator whose diagonalization yields renormalized quasiparticle energies and, when off-diagonal terms are retained, renormalized wave functions (Lihm et al., 2020). In this language, the real part of the self-energy shifts dispersions and band edges, while the imaginary part controls scattering rates, linewidths, and carrier lifetimes; this separation is used explicitly in angle-resolved analyses of graphene, fullerides, and photoexcited transition-metal dichalcogenides (TMDCs) (Pletikosić et al., 2012).

For band-edge problems, the renormalized quasiparticle energy is often written as a bare eigenvalue plus the self-energy evaluated at the electronic energy. In monolayer WSe$_2$, for example, the phonon-induced correction is expressed as

$$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$

and the temperature dependence follows from Bose-populated phonon branches within an Allen-Heine-Cardona-type treatment (Mishra et al., 2018). In crystalline naphthalene, the same structure is used with explicit Fan and Debye-Waller terms to obtain both band-gap shifts and electronic lifetimes near the band edges (Brown-Altvater et al., 2020).

A central refinement is the distinction between adiabatic and non-adiabatic treatments. In the adiabatic approximation the phonon frequency is neglected in the Fan denominator, whereas non-adiabatic formulations retain $\omega_{\mathbf{q}j}$ and thereby regularize long-wavelength divergences in infrared-active materials. Large-scale first-principles work shows that, for light-element materials, zero-point band-gap renormalization is often larger than $0.3$ eV and up to $0.7$ eV, and that agreement with experiment for infrared-active systems is obtained only when non-adiabatic effects are included (Miglio et al., 2020).

2. Theoretical frameworks and computational schemes

The modern literature organizes phonon-induced band renormalization into several partially overlapping frameworks. Perturbative theories remain the standard starting point, but substantial recent work has clarified where they fail and how they may be generalized.

Framework	Principal content	Regime or implication
Allen-Heine-Cardona	Fan + Debye-Waller band-energy shifts	Weak-coupling, delocalized quasiparticles
Williams-Lax	Thermal average over nuclear wave functions	Optical spectra and direct/indirect gaps
Generalized self-energy matrix	Includes off-diagonal Debye-Waller terms	Wave-function renormalization, topological transitions
Self-consistent polaronic Green’s function	Adds localization and self-trapping	Strong-coupling and small-polaron limits

Within Williams-Lax theory, temperature-dependent optical spectra and band gaps are written as Gaussian averages over normal coordinates. A notable practical development is the deterministic “one-shot” supercell procedure, in which a single optimal displacement pattern reproduces the thermal average in the large-supercell limit. This method was shown to capture zero-point and temperature-dependent gap renormalization in direct- and indirect-gap semiconductors and to include phonon-assisted optical absorption in indirect-gap materials (Zacharias et al., 2016). Closely related DFTB work on diamond and silicon compared stochastic Monte Carlo sampling, the deterministic one-shot method, and classical Born-Oppenheimer molecular dynamics, showing that the quantum treatments yield band gaps in good agreement with experiment while molecular-dynamics snapshots inadequately describe the low-temperature renormalization because nuclear quantum effects are absent (Heide et al., 2024).

Beyond diagonal energy shifts, generalized Allen-Heine-Cardona theory now treats renormalization of both electron energies and wave functions. The key advance is an operator generalization of the acoustic sum rule that yields diagonal and off-diagonal Debye-Waller matrix elements in a unified framework, allowing one to diagonalize the full self-energy matrix and follow wave-function hybridization across temperature-driven topological transitions (Lihm et al., 2020). This extension is essential when interband energy differences are comparable to the phonon-induced self-energy.

A further generalization addresses the breakdown of perturbation theory in systems that form localized polarons. The self-consistent many-body Green’s-function theory of "Unified approach to polarons and phonon-induced band structure renormalization" reduces to Allen-Heine theory in the weak-coupling limit and to total-energy calculations of self-trapped polarons in the strong-coupling limit (Lafuente-Bartolome et al., 2022). In that formulation, localization enters through an additional polaronic self-energy, so that renormalization is not restricted to small corrections around Bloch-like states.

3. Momentum-resolved manifestations: kinks, replicas, linewidths, and pseudogaps

Angle-resolved probes provide the clearest direct visualization of phonon-induced band renormalization. In these measurements the renormalization does not appear merely as a rigid band shift; it can instead manifest as localized slope changes, loss of coherent spectral weight, replica features, or strongly momentum-dependent linewidths.

A particularly strong-coupling example is TiSe$_2$. Time- and angle-resolved extreme-ultraviolet photoemission combined with ab initio electron-phonon calculations shows that, at room temperature, the normal phase is governed by strong electron-phonon-induced thermal fluctuations. In the moderate-coupling materials HfTe$_2$ and ZrTe$_2$, thermal fluctuations produce only small spectral broadening and minor band renormalization, whereas in TiSe$$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$0 soft-phonon-induced, momentum-dependent suppression of spectral weight produces pseudogaps extending up to $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$1 eV above the Fermi level. Additional signatures include faint replica Se-$$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$2 bands below the Ti-$$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$3 band at the M point, a kink in the Ti-$$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$4 band around $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$5 eV above $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$6 near $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$7, and loss of coherent spectral weight at $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$8 from $$\Sigma_{nn'}({\bf k}, \omega) = \Sigma^{\rm Fan}_{nn'}({\bf k}, \omega) + \Sigma^{\rm DW}_{nn'}({\bf k}),$$9 to $_2$0 eV. Standard DFT does not reproduce these features; full dynamical electron-phonon self-energies do (Fragkos et al., 30 Jun 2026).

In superconducting K$_2$1C$_2$2 films, high-resolution ARPES resolves a dispersive band crossing the Fermi level with an occupied bandwidth of about $_2$3 meV, multiple quasiparticle kinks at approximately $_2$4, $_2$5, and $_2$6 meV, and a replica band at about $_2$7 meV below the main band. These features are assigned to intermolecular and intramolecular phonon modes, including the Jahn-Teller active Hg(8) mode, and yield an electron-phonon coupling constant of about $_2$8, which dominates the quasiparticle mass renormalization (Zhou et al., 2023).

In potassium-doped graphene on Ir(111), the renormalization of the $_2$9 band was analyzed by a self-consistent reconstruction of the bare band from ARPES data. The resulting self-energy and spectral intensity indicate coupling to two distinct phonon excitations, at $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$0 meV and $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$1 meV, with a moderate coupling constant of about $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$2 (Pletikosić et al., 2012). This case is instructive because the bare band is not linear over the relevant window; the renormalization cannot therefore be extracted by assuming a simple linear reference dispersion.

An important misconception corrected by recent work is that dramatic phonon anomalies necessarily require Fermi-surface nesting. In YNi$$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$3B$$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$4C, inelastic neutron scattering, soft-x-ray ARPES, and ab initio calculations show that strong phonon broadening can occur in the absence of both Fermi-surface nesting and lattice anharmonicity when the electron-phonon matrix element is strongly enhanced for specific electron momenta $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$5. The linewidth is then governed by a $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$6-selective weighting of the electronic joint density of states rather than by nesting alone (Kurzhals et al., 2022).

4. Band-gap renormalization, zero-point motion, and temperature dependence

A large part of the literature addresses phonon-induced band renormalization in the narrower sense of direct and indirect gap shifts. Here the central quantities are zero-point renormalization, finite-temperature redshift, and the mode-resolved decomposition of the shift.

In monolayer WSe$$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$7, a fully ab initio treatment combining Kohn-Sham calculations, $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$8 quasiparticle corrections, and a Bethe-Salpeter equation with polaronic energies shows that renormalization of the bare electronic energies by phonon interactions and anharmonic lattice thermal expansion causes a decreasing band gap with increasing temperature. The zero-point energy renormalization is $$\Delta E_{n\mathbf{k}}^{ep}(T)=\langle n\mathbf{k}|\Sigma_{n\mathbf{k}}^{ep}(T)|n\mathbf{k}\rangle,$$9 meV and is found to be entirely due to the polaronic interaction with negligible contribution from lattice anharmonicities. All excitonic peaks redshift as temperature rises, and the in-plane torsional acoustic phonon branch contributes mostly to the A and B exciton build-up (Mishra et al., 2018).

The magnitude of the effect can be much larger in highly anharmonic polar compounds. In silver chalcohalide antiperovskites Ag$\omega_{\mathbf{q}j}$0XY, finite-temperature calculations based on density functional theory, tight binding, and anharmonic Fröhlich theory find a giant band-gap reduction of approximately $\omega_{\mathbf{q}j}$1–$\omega_{\mathbf{q}j}$2 near room temperature relative to the value calculated at $\omega_{\mathbf{q}j}$3 K. Low-energy optical polar phonon modes that break inversion symmetry and promote overlap between silver and chalcogen $\omega_{\mathbf{q}j}$4 orbitals in the conduction band are identified as the primary contributors. When temperature effects are included, the optical absorption coefficient increases by nearly an order of magnitude for visible-light frequencies (Benítez et al., 2024).

For organic crystals, crystalline naphthalene provides a representative case in which electron-phonon self-energies, $\omega_{\mathbf{q}j}$5 quasiparticle corrections, and transport observables can all be evaluated consistently. The zero-point renormalization of the DFT gap is $\omega_{\mathbf{q}j}$6 eV, with a further reduction at $\omega_{\mathbf{q}j}$7 K, and the phonon-renormalized $\omega_{\mathbf{q}j}$8 band structure predicts a fundamental band gap of $\omega_{\mathbf{q}j}$9 eV at room temperature, in good agreement with experiment. The same calculations yield temperature-dependent electron and hole mobilities through phonon-induced lifetimes (Brown-Altvater et al., 2020).

The low-temperature asymptotics have also been systematized. A general harmonic theory for phonon-renormalized properties shows that, for the class of observables that includes the electronic band gap, the leading low-temperature dependence is $0.3$0 in three dimensions, $0.3$1 in two dimensions, and $0.3$2 in one dimension (Monserrat et al., 2013). This result provides a rigorous replacement for empirical fitting forms when extrapolating zero-point corrections from finite-temperature data.

A related theoretical line uses low-energy Dirac models coupled to effective phonon fields. For monolayer TMDs, a large-$0.3$3 treatment of massive Dirac-like electrons yields a renormalized optical gap that remains nearly constant at low temperatures and then decreases linearly with $0.3$4 at higher temperatures; the reported reduction is approximately $0.3$5–$0.3$6 eV between about $0.3$7 K and $0.3$8 K, with crossover temperatures in the approximate range $0.3$9–$0.7$0 K for MoS$0.7$1, MoSe$0.7$2, WS$0.7$3, and WSe$0.7$4 (Bezerra et al., 2024). This suggests that low-energy field-theoretic models can reproduce the gross temperature dependence of optical gaps when fitted with reasonable ultraviolet cutoffs.

5. Strong-coupling, polarons, and interaction renormalization

In strong-coupling regimes, phonon-induced renormalization ceases to be a small correction to a fixed band structure. The appropriate language shifts toward polarons, self-trapping, and interaction renormalization, and the distinction between “band renormalization” and “state reconstruction” becomes less sharp.

The clearest formal statement of this crossover is provided by the self-consistent Green’s-function theory that unifies Allen-Heine band renormalization with self-trapped polaron energetics. In LiF, this theory predicts a zero-point band-gap renormalization of $0.7$5 eV, substantially larger than the $0.7$6 eV obtained from the Fan-Migdal plus Debye-Waller treatment, because the hole forms a small polaron that is not captured by perturbative approaches (Lafuente-Bartolome et al., 2022). The implication is not that Allen-Heine theory is invalid in general, but that it systematically underestimates renormalization when the carrier localizes.

A conceptually related result appears in the Anderson-Holstein impurity problem. There the electron-phonon interaction renormalizes the on-site Coulomb repulsion according to

$0.7$7

but this expected phonon-mediated reduction is completely suppressed when the phonons are slow compared to charge fluctuations, $0.7$8. The suppression is attributed to the oscillator’s inability to respond to rapid charge fluctuations, while fast phonons, $0.7$9, produce conventional polaronic behavior with renormalized onsite interactions and additional satellite features in the local spectral function (Lau et al., 13 Mar 2025). Although this is an impurity model rather than a crystalline band-structure calculation, it highlights a general lesson: the magnitude and even the sign of phonon-induced renormalization depend on dynamical timescale separation, not only on coupling strength.

TiSe$_2$0 offers a complementary solid-state realization of strong coupling without full self-trapping. There, strong intervalley $_2$1 scattering mediated by soft charge-density-wave phonons yields a hierarchy of electron-phonon coupling strengths,

$_2$2

and the normal phase is better described as a phonon-induced pseudogap phase than as a simple semimetal or semiconductor (Fragkos et al., 30 Jun 2026). This is a useful corrective to the common tendency to equate strong electron-phonon coupling solely with uniform mass enhancement or rigid gap shrinkage.

6. Extensions, special cases, and conceptual boundaries

Several developments extend phonon-induced band renormalization beyond the conventional problem of shifting a band edge. One extension concerns the renormalization of wave functions themselves. In BiTlSe$_2$3, the generalized self-energy framework yields temperature-dependent changes not only in band energies but also in hidden spin polarization across a topological transition, because off-diagonal Debye-Waller terms hybridize states whose separations are comparable to the phonon self-energy (Lihm et al., 2020). In such cases, a theory restricted to diagonal energy shifts gives incorrect band connectivity and misses wave-function observables.

A second extension arises in moiré materials, where lattice reconstruction and symmetry breaking feed back on both phonons and electrons. In reconstructed MoS$_2$4 moiré superlattices, atomic reconstruction and strain cause rapid evolution of the phonon spectrum, including anti-crossings of layer-breathing and folded transverse-acoustic modes and splitting of the intralayer $_2$5 mode into $_2$6 and $_2$7. The same reconstruction-induced symmetry breaking is identified as the microscopic origin of band splitting and subgap structure seen in optical spectroscopy, and the continuum model is explicitly formulated to treat electrons and phonons “on an equal footing” (Quan et al., 2020).

Mode-selective renormalization at electronic phase transitions constitutes another boundary case. In NaFe$_2$8Co$_2$9As, Raman scattering revealed a superconductivity-induced renormalization of the B$_2$0(Fe) phonon in the Co-doped compound: below $_2$1 K the mode hardened by about $_2$2 cm$_2$3 and narrowed by about the same amount. The effect could not be explained within single-band or simple multiband models and instead required interband scattering, spin-density-wave correlations, and sign-changing $_2$4 pairing (Um et al., 2014). This is not a generic band renormalization of all electronic states, but it demonstrates how phonon self-energies can diagnose correlated electronic restructuring.

Finally, adjacent literatures use closely related renormalization ideas for quantities other than Bloch-band energies. In optically driven InGaAs/GaAs quantum dots, fluctuations of a longitudinal-acoustic phonon bath renormalize the Rabi frequency according to

$_2$5

producing a phonon analogue of the Lamb shift (Ramsay et al., 2010). This is formally distinct from electronic band renormalization, yet it shares the same causal structure: a bosonic bath generates a complex self-energy whose real and imaginary parts renormalize coherent dynamics and induce damping. A plausible implication is that the conceptual reach of phonon-induced renormalization extends well beyond the equilibrium band picture, even when the primary observable is not a band edge or dispersion.

Across these settings, the core lesson is consistent. Phonons renormalize electronic structure through dynamical self-energies whose effects may appear as gap shifts, mass enhancement, linewidth broadening, pseudogaps, replica bands, wave-function hybridization, or state localization. Whether the outcome is perturbative and nearly rigid, or reconstructive and strongly momentum dependent, is controlled by the interplay of coupling strength, phonon frequency, dimensionality, symmetry, and the underlying electronic structure (Miglio et al., 2020, Lafuente-Bartolome et al., 2022).