Peierls Electron-Phonon Coupling

• Peierls electron–phonon coupling is an interaction where lattice bond distortions modulate electronic hopping, fostering dimerization and correlated hopping.
• It is characterized by momentum-selective phonon softening and the formation of light polarons and bipolarons, which differ from local Holstein modes.
• This coupling mechanism underlies unconventional superconductivity and charge-density-wave transitions, with its effects sensitive to band geometry and momentum dependence.

Peierls electron-phonon coupling denotes an interaction in which lattice distortions modify electronic motion through bonds rather than merely shifting on-site energies. In its microscopic Su-Schrieffer-Heeger form, the phonon coordinate multiplies a hopping operator, so the lattice couples to bond kinetic energy and naturally favors dimerization, correlated hopping, and unconventional effective interactions. In the charge-density-wave literature, the same Peierls framework also denotes an electron-phonon-driven instability whose ordering vector is selected not only by electronic susceptibility but, in realistic materials, by the full kernel $|g(q)|^2 \chi_0(q)$. Across one-dimensional chains, two-dimensional polaron problems, molecular crystals, perovskites, and quasi-one-dimensional CDW compounds, this coupling produces phenomena that differ sharply from density-displacement Holstein or Fröhlich models, including light polarons and bipolarons, momentum-selective phonon softening, and strong sensitivity to band geometry (Sous et al., 2018, Maschek et al., 2014, Yu et al., 2023).

1. Definition, terminology, and contrast with density-displacement coupling

The literature represented here uses “Peierls” in two closely related senses. First, in tight-binding models it refers to bond-modulated hopping, often called Peierls or SSH coupling. Second, in the theory of one-dimensional and quasi-one-dimensional charge-density waves, it denotes the electron-phonon mechanism behind the $2k_F$ lattice instability, including its generalization to realistic systems where the momentum dependence of the coupling itself becomes decisive (Sous et al., 2018, Maschek et al., 2014).

Form	Coupled electronic quantity	Canonical consequence
Peierls / SSH	Hopping or bond kinetic energy	Dimerization, bond order, correlated hopping
Holstein	Local density / on-site energy	Local attraction, density wave or small-polaron physics
Fröhlich-type	Nonlocal density-displacement	Long-range density-mediated attraction

In the SSH realization studied for a one-dimensional chain with Einstein phonons, the coupling is

$$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$

so the relative displacement $u_i-u_{i+1}$ modulates the bond $i\leftrightarrow i+1$. By contrast, the Holstein interaction takes the density-displacement form

$$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$

and a Fröhlich-type interaction remains density-displacement, but nonlocal in space (Sous et al., 2018).

In the original Peierls problem, a half-filled one-dimensional chain is unstable to dimerization: alternating short and long bonds break translation symmetry, double the unit cell, and open a gap at the Fermi surface. In the canonical extension of this picture to higher-dimensional metals, one often assumes weakly momentum-dependent electron-phonon coupling $g(q)\approx g_0$, so the ordering vector is dictated mainly by $\chi_0(q)$. Real materials need not obey that simplification; a more accurate statement is that the effective instability is governed by $|g(q)|^2 \chi_0(q)$, and momentum-dependent matrix elements can shift the CDW wave vector away from the best-nested vector (Maschek et al., 2014, Diego et al., 4 Feb 2026).

2. Microscopic Hamiltonians and effective interactions

A standard correlated Peierls model combines nearest-neighbor hopping, Coulomb terms, phonons, and bond coupling. In the two-electron SSH problem the Hamiltonian is written

$$\mathcal{H} = \mathcal{H}_{\rm e} + \mathcal{H}_{\rm ph} + \hat{V}_{\rm e\text{-}ph},$$

with

$2k_F$0

In the extended Hubbard-Peierls chain, the microscopic coupling is equivalently written as bond-length coupling,

$2k_F$1

which directly couples bond distortion to electronic kinetic energy (Sous et al., 2018, Pearson et al., 2010).

The decisive distinction from Holstein physics appears after phonons are eliminated. In the anti-adiabatic limit $2k_F$2, a single Peierls polaron acquires a formation energy and a dynamically generated next-nearest-neighbor hopping,

$2k_F$3

leading to

$2k_F$4

For two particles, the effective interaction is not a negative-$2k_F$5 density-density term. Instead, the low-energy Hamiltonian contains pair-hopping and exchange processes with amplitudes $2k_F$6, $2k_F$7, $2k_F$8, and $2k_F$9. The interaction vertex depends on total pair momentum as $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$0, rather than only on transferred momentum $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$1 (Sous et al., 2018).

This difference has direct many-body consequences. In the mixed Holstein-SSH chain, integrating out phonons gives retarded g-ology couplings with $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$2 for Holstein coupling and $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$3, $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$4 for SSH coupling. The total $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$5 can therefore be tuned close to zero while $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$6 remains attractive, opening a route to a spin-gapped metallic state generated by competing site and bond couplings (Hohenadler, 2016).

3. Polarons, bipolarons, and momentum-space reconstruction

Peierls coupling produces polarons and bipolarons with properties that are atypical for density-displacement models. In the one-dimensional SSH model with $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$7 and $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$8, a bipolaron exists for any $$\hat{V}_{\rm e\text{-}ph} = g \sum_{i,\sigma} \left( c_{i,\sigma}^\dagger c_{i+1,\sigma} + h.c. \right) \left( b_i^\dagger + b_i - b_{i+1}^\dagger - b_{i+1}\right),$$9, and for $u_i-u_{i+1}$0 its band is completely separated from the two-polaron continuum. The same calculations show that the bipolaron becomes strongly bound around $u_i-u_{i+1}$1, yet for $u_i-u_{i+1}$2 its effective mass satisfies $u_i-u_{i+1}$3, with $u_i-u_{i+1}$4. At $u_i-u_{i+1}$5, the paper finds $u_i-u_{i+1}$6, whereas a Holstein bipolaron at the same $u_i-u_{i+1}$7 and $u_i-u_{i+1}$8 has bandwidth $u_i-u_{i+1}$9 and is heavier by about two orders of magnitude (Sous et al., 2018).

The same SSH study identified two low-energy bipolaron bands at strong coupling. In the anti-adiabatic strong-coupling limit, the two-particle Hilbert space splits into even- and odd-separation sectors, each of which supports its own bound state. Bare hopping then mixes the sectors and generates two hybridized bipolaron bands with an avoided crossing and strongly non-parabolic dispersion. Because the binding originates from pair-hopping and exchange rather than static on-site attraction, the odd-sector component remains largely insensitive to strong on-site Hubbard repulsion, so the critical $i\leftrightarrow i+1$0 for bipolaron dissociation is much larger than in the Holstein case, even when a sizable nearest-neighbor repulsion $i\leftrightarrow i+1$1 or $i\leftrightarrow i+1$2 is included (Sous et al., 2018).

In two dimensions, Peierls/Su-Schrieffer-Heeger polarons remain qualitatively distinct from Holstein polarons, but the outcome depends on the microscopic realization of displacement-modulated hopping. In model A, where site phonons enter through relative bond displacements, the ground-state momentum undergoes a continuous transition from $i\leftrightarrow i+1$3 to finite diagonal momenta at $i\leftrightarrow i+1$4 for $i\leftrightarrow i+1$5 and $i\leftrightarrow i+1$6 for $i\leftrightarrow i+1$7. The transition is driven by a quadratic instability of the dispersion, so the effective mass diverges at the critical point and then decreases anisotropically. In model B, where phonons live directly on bonds, the ground-state momentum always remains at $i\leftrightarrow i+1$8, but the effective mass increases monotonically and levels off at strong coupling rather than growing exponentially (Zhang et al., 2021).

More complex lattices preserve the same qualitative lesson. In a BaBiO$i\leftrightarrow i+1$9-inspired perovskite model with Peierls coupling of oxygen motion to $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$0-$$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$1 and $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$2-$$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$3 hoppings, the single-polaron ground-state momentum exhibits sharp jumps between high-symmetry points. For $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$4, the transition is $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$5; for $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$6, the sequence can be $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$7. The paper concludes that such behavior cannot be mapped onto a one-band Holstein model over the full parameter space, because the Holstein polaron always keeps the same ground-state momentum (Yam et al., 2020).

At finite carrier number, the many-body consequence of these light bound states is a stable bipolaron liquid rather than generic clustering. In a one-dimensional chain with Peierls coupling and optical phonons, DMRG finds no phase separation in the dilute limit except at very large $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$8, where the linear approximation $$\hat{V}^{\rm Hol}_{\rm e\text{-}ph} = g_{\rm H} \sum_{i,\sigma} \hat n_{i,\sigma} (b_i^\dagger + b_i),$$9 becomes invalid. The same work shows that stability extends to finite carrier concentrations up to quarter filling, and that in the anti-adiabatic limit the critical coupling for phase separation diverges, $g(q)\approx g_0$0 (Nocera et al., 2020).

4. Peierls transitions, quantum fluctuations, and wave-vector selection

The ideal Peierls instability of a one-dimensional metal is tied to the divergence of the noninteracting susceptibility at $g(q)\approx g_0$1, which softens a phonon and condenses a distortion with that wave vector. Once lattice dynamics are quantized and Coulomb interactions are included, the transition is no longer captured by the adiabatic “any nonzero coupling orders” picture. In the extended Hubbard model with quantized phonons, DMRG finds a Berezinskii-Kosterlitz-Thouless transition at a nonzero critical coupling $g(q)\approx g_0$2 for nonvanishing electron-electron interaction. For fixed optical phonon gap $g(q)\approx g_0$3, $g(q)\approx g_0$4 increases continuously as the phonon spectrum is tuned from Einstein to Debye, showing that dispersive phonons destabilize the dimerized ground state relative to the dispersionless limit. Using Debye phonons, the same study places trans-polyacetylene close to the critical regime (Pearson et al., 2010).

Thermal and quantum lattice fluctuations modify the same transition even in simpler Peierls chains. In the spinless Holstein model at half filling, quantum Monte Carlo finds that the adiabatic regime exhibits pronounced phonon softening and a Peierls CDW insulator above a critical coupling, while the antiadiabatic regime remains a Tomonaga-Luttinger liquid with strong polaronic band renormalization. The specific heat and compressibility track the opening of the Peierls gap, the emergence of low-energy polaron excitations, and the renormalization of the phonon mode across the transition (Weber et al., 2018).

In real CDW materials, the central question is often not whether electron-phonon coupling matters, but how its momentum structure selects the ordered state. In TbTe$g(q)\approx g_0$5, high-resolution inelastic x-ray scattering and DFPT show that the soft mode minimum and largest linewidth occur near $g(q)\approx g_0$6, whereas the eJDOS and weak nesting maximum lie near $g(q)\approx g_0$7. The offset between linewidth and eJDOS demonstrates that $g(q)\approx g_0$8 is strongly momentum dependent and shifts the instability away from the nesting vector (Maschek et al., 2014).

A closely related conclusion emerges in quasi-one-dimensional ZrTe$g(q)\approx g_0$9. First-principles calculations show that the Fermi surface is reproduced correctly only when a Hubbard interaction is included on Te $\chi_0(q)$0 orbitals; this same correction is essential for the appearance of a soft harmonic phonon mode at $\chi_0(q)$1. The nesting function $\chi_0(q)$2 peaks at the CDW vector, but the phonon linewidth divided by the nesting factor remains sharply peaked as well, demonstrating that variations of the electron-phonon matrix elements with momentum dominate over purely electronic effects (Diego et al., 4 Feb 2026).

Metallic single-wall carbon nanotubes provide an instructive quasi-one-dimensional contrast. In a weak-coupling RG treatment with retardation, strong electron-phonon interactions produce phonon softening and a Peierls insulator phase in armchair nanotubes; the paper finds that this softening may occur for any intraband scattering phonon mode. For zigzag nanotubes, by contrast, the effect of electron-phonon coupling is negligible in the same framework, and no comparable Peierls lattice distortion is obtained (Okamoto et al., 2018).

5. Molecular solids, transport, and finite-temperature band renormalization

In organic molecular crystals, Peierls coupling is the natural language for intermolecular vibrations that modulate transfer integrals. In pentacene films on Bi(001), temperature-dependent ARPES yields an electron-phonon mass enhancement factor $\chi_0(q)$3 and an effective Einstein phonon energy $\chi_0(q)$4 meV. From the relation

$\chi_0(q)$5

the effective Peierls coupling is $\chi_0(q)$6. The same measurements show that the HOMO bandwidth narrows by $\chi_0(q)$7 between $\chi_0(q)$8 K and $\chi_0(q)$9 K, and the extracted $|g(q)|^2 \chi_0(q)$0 identifies low-energy intermolecular vibrations as the dominant modes behind the observed electron-phonon effects (0908.4258).

A nonperturbative analytical treatment of a Holstein-Peierls model with linear and quadratic coupling reaches the same conclusion from the band-structure side. In crystalline naphthalene, the transformed hopping matrix acquires a temperature-dependent renormalization factor,

$|g(q)|^2 \chi_0(q)$1

so band narrowing is controlled by both linear and quadratic local and nonlocal couplings. Applied to naphthalene, the formalism finds that quadratic coupling contributes substantially to the LUMO and HOMO bandwidth renormalization and that full local-plus-nonlocal quadratic coupling is needed for quantitative agreement (Risueño et al., 2018).

Transport calculations based on the Holstein-Peierls chain make the distinction between local and nonlocal coupling especially explicit. In a TD-DMRG study of current-current and heat-current correlation functions, nonlocal electron-phonon coupling strongly affects the conductivity through dynamic disorder, while having very weak influence on the Seebeck coefficient because of cancellation between the heat current-current correlation function and the current-current correlation function. The same study finds that doping concentration strongly affects both conductivity and Seebeck coefficient, and that the optimal doping ratio to reach the highest power factor is $|g(q)|^2 \chi_0(q)$2 fillings (Ge et al., 2023).

These results support a unified picture in which Peierls coupling in organic materials is not a minor correction to local polaron physics. It directly controls transfer-integral fluctuations, bandwidth narrowing, transient localization, and the crossover among hopping, phonon-assist current, band, and transient-localization transport regimes (0908.4258, Risueño et al., 2018, Ge et al., 2023).

6. Superconductivity, quantum geometry, and broader implications

The superconducting significance of Peierls coupling follows from the same microscopic distinction that separates it from Holstein and Fröhlich models. In the SSH bipolaron problem, the authors explicitly argue that the standard objection to phonon-mediated high-$|g(q)|^2 \chi_0(q)$3 superconductivity—namely, that strong coupling inevitably produces very heavy bipolarons—does not apply when the effective interaction is of pair-hopping form. Their conclusion is that light bipolarons should condense into a BEC-type superconductor with high $|g(q)|^2 \chi_0(q)$4, and that materials with predominantly Peierls electron-phonon coupling represent a distinct route to phonon-mediated superconductivity at high temperatures (Sous et al., 2018).

The many-body stability analysis strengthens that claim by removing a second standard obstacle. In one dimension, Peierls bipolaron liquids remain stable against phase separation over most of the physically valid parameter space, unlike the extended Holstein model where phase separation occurs at much smaller $|g(q)|^2 \chi_0(q)$5. This suggests that, in higher dimensions, a low-density liquid of light bipolarons need not be preempted by clustering, making a bipolaronic superconducting mechanism materially more plausible (Nocera et al., 2020). The two-dimensional PSSH polaron results reinforce the same theme: even when the ground-state momentum structure depends strongly on microscopic details, strong coupling does not force exponential mass growth, which is precisely the property needed if a bipolaron mechanism is to remain viable (Zhang et al., 2021).

A separate development generalizes the meaning of Peierls coupling in multiband systems. When hopping modulation is treated within a tight-binding framework that keeps Bloch wave functions explicitly, the electron-phonon matrix can be written in a Gaussian approximation as

$|g(q)|^2 \chi_0(q)$6

This decomposes the dimensionless coupling constant into an energetic part and a geometric part,

$|g(q)|^2 \chi_0(q)$7

with $|g(q)|^2 \chi_0(q)$8 controlled by the Fubini-Study metric or its orbital-selective version. In graphene, the geometric contribution accounts for approximately $|g(q)|^2 \chi_0(q)$9 of the total electron-phonon coupling constant; in MgB$$\mathcal{H} = \mathcal{H}_{\rm e} + \mathcal{H}_{\rm ph} + \hat{V}_{\rm e\text{-}ph},$$0, it accounts for approximately $$\mathcal{H} = \mathcal{H}_{\rm e} + \mathcal{H}_{\rm ph} + \hat{V}_{\rm e\text{-}ph},$$1. In both systems, the geometric contribution is bounded from below by topological invariants, and the paper concludes that nontrivial electron band geometry or topology might favor superconductivity with relatively high critical temperature (Yu et al., 2023).

Taken together, these results suggest a broader reinterpretation of Peierls electron-phonon coupling. In low-dimensional lattice models it generates bond-order instabilities, correlated hopping, and unusually light composite carriers. In realistic CDW compounds it enters through a sharply momentum-dependent kernel that can determine the ordering vector more strongly than bare nesting. In multiband superconductors it acquires a quantum-geometric component that can dominate the total coupling strength. The common thread is that Peierls coupling is not adequately described as a simple attractive interaction; it is a structured kinetic coupling whose consequences depend on bond geometry, symmetry, and the momentum dependence of the underlying electronic states (Sous et al., 2018, Diego et al., 4 Feb 2026, Yu et al., 2023).