---
title: 'Eliashberg Effect: Retardation in Superconductivity'
url: https://www.emergentmind.com/topics/eliashberg-effect
type: topic
---

# Eliashberg Effect: Retardation in Superconductivity

Searching arXiv for recent and foundational papers on the Eliashberg effect and Eliashberg theory usage across contexts.
Using the available environment to query arXiv metadata for the cited topic.
```python
import urllib.parse, urllib.request, ssl, re
query = urllib.parse.quote('all:"Eliashberg effect" OR ti:Eliashberg')
url = f'http://export.arxiv.org/api/query?search_query={query}&start=0&max_results=5'
try:
    data = urllib.request.urlopen(url, timeout=20).read().decode('utf-8', errors='ignore')
    print(data[:1200])
except Exception as e:
    print("ERR", e)
```
```python
import urllib.parse, urllib.request
query = urllib.parse.quote('all:"Eliashberg effect" OR ti:Eliashberg')
url = f'https://export.arxiv.org/api/query?search_query={query}&start=0&max_results=5'
try:
    data = urllib.request.urlopen(url, timeout=20).read().decode('utf-8', errors='ignore')
    print(data[:1500])
except Exception as e:
    print("ERR", repr(e))
```
```python
import urllib.parse, urllib.request, socket
socket.setdefaulttimeout(10)
query = urllib.parse.quote('all:"Eliashberg effect" OR ti:Eliashberg')
url = f'https://export.arxiv.org/api/query?search_query={query}&start=0&max_results=3'
print(url)
try:
    with urllib.request.urlopen(url) as r:
        b = r.read(500)
        print("bytes", len(b), b[:200])
except Exception as e:
    print(type(e).__name__, e)
```
Attempting a direct arXiv API lookup via the container.
to=container.exec  彩神争霸是不是0code='{"cmd":["bash","-lc","python - <<\'PY\'\nimport urllib.parse, urllib.request, socket\nsocket.setdefaulttimeout(10)\nquery = urllib.parse.quote(\'all:\"Eliashberg effect\" OR ti:Eliashberg\')\nurl = f\'https://export.arxiv.org/api/query?search_query={query}&start=0&max_results=3\'\nprint(url)\ntry:\n    with urllib.request.urlopen(url) as r:\n        data = r.read().decode(\'utf-8\', errors=\'ignore\')\n        print(data[:1200])\nexcept Exception as e:\n    print(type(e).__name__, e)\nPY"]}'}
The **Eliashberg effect** is a term used in more than one, related sense in the condensed-matter literature. In its narrow historical usage, it denotes the **microwave-induced enhancement of superconductivity** in a dirty superconductor through nonequilibrium quasiparticle redistribution. In a broader and now more common usage, it refers to the set of **retardation-driven, frequency-dependent strong-coupling phenomena** described by Eliashberg theory: dynamical self-energy renormalization, a frequency-dependent gap function, and boson-spectrum-dependent modifications of superconducting and metallic properties. Modern work has extended this framework from conventional phonon superconductors to spin-fluctuation systems, multiband materials, quantum-critical metals, non-Fermi liquids, cavity-driven nonequilibrium states, and first-principles \(GW_0\)-Eliashberg calculations [1911.04959], [1911.05065], [2404.11820].

## 1. Terminological scope and historical meanings

The phrase “Eliashberg effect” does not designate a single phenomenon across all subfields. The literature represented here uses it in at least three technically distinct but conceptually connected ways.

| Usage | Meaning | Representative source |
|---|---|---|
| Historical nonequilibrium usage | Microwave-induced enhancement of superconductivity via quasiparticle redistribution near \(T_c\) | [1911.04959] |
| Conventional strong-coupling usage | Retardation-induced, frequency-dependent corrections to BCS superconductivity | [1911.05065] |
| Extended modern usage | Eliashberg-type self-consistent dynamics for unconventional, quantum-critical, or non-Fermi-liquid systems | [2404.11820] |

In the historical sense, Eliashberg’s 1970 result concerns a weak ac field that produces a stationary nonequilibrium quasiparticle distribution favorable for pairing. In the broader strong-coupling sense, the defining feature is not microwave irradiation but the replacement of a static pairing interaction by a **dynamical bosonic kernel**, so that both the fermionic self-energy and the superconducting order parameter become frequency dependent [1911.04959], [1911.05065].

This broader usage is now standard in discussions of conventional superconductivity and is increasingly applied to unconventional settings. Reviews of pairing in non-Fermi liquids and quantum-critical metals explicitly treat Eliashberg theory as a general self-consistent framework for fermions coupled to bosonic modes, no longer restricted to ordinary phonons [1912.07646], [2506.11952]. A plausible implication is that “Eliashberg effect” now functions less as a single named effect than as a family of dynamical renormalization phenomena organized by Eliashberg equations.

## 2. Formal structure of Eliashberg theory

In the comparative analysis of electron-phonon and electron-electron quantum-critical problems, Eliashberg theory is defined as the **one-loop self-consistent treatment of fermions coupled to a bosonic mode** in which vertex corrections are neglected. Its formal structure is
\[
G^{-1}(k)=G_0^{-1}(k)+i\Sigma(k),\qquad \chi^{-1}(q)=\chi_0^{-1}(q)+\Pi(q),
\]
with coupled equations
\[
\Sigma(k)= i\tilde g^2\!\int\!\frac{d^2q\,d\Omega_n}{(2\pi)^3}\,G(k+q)\chi(q),\qquad
\Pi(q)=2\tilde g^2\!\int\!\frac{d^2k\,d\omega_m}{(2\pi)^3}\,G(k+q)G(k),
\]
and the defining approximation is that the full interaction vertex is replaced by the bare one [2404.11820].

In superconducting applications, the same logic appears in Nambu form through the renormalization function \(Z\), the anomalous self-energy \(\phi\), and the gap function
\[
\Delta(i\omega_m)=\frac{\phi(i\omega_m)}{Z(i\omega_m)}.
\]
For isotropic finite-temperature phonon superconductivity, the imaginary-axis equations take the standard form
\[
Z(i\omega_m)=1+{\pi T_c\over \omega_m}\sum_{m'}\lambda(i\omega_m-i\omega_{m'}){\omega_{m'}\over \sqrt{\omega_{m'}^2+\Delta^2(\omega_{m'})}},
\]
\[
Z(i\omega_m)\Delta(i\omega_m)=\pi T_c\sum_{m'}\left[\lambda(i\omega_m-i\omega_{m'})-u\,\theta\!\left(\frac{W}{2}-|\omega_{m'}|\right)\right]
\frac{\Delta(i\omega_{m'})}{\sqrt{\omega_{m'}^2+\Delta^2(\omega_{m'})}},
\]
where \(\lambda(z)\) is determined by the Eliashberg spectral function \(\alpha^2F(\nu)\) and \(u\) represents the Coulomb term [1911.05065].

The physical content of these equations is the retention of **retardation**. In BCS theory, pairing is effectively instantaneous; in Eliashberg theory, the interaction depends on frequency transfer, so the normal self-energy, mass renormalization, and anomalous sector are all dynamical. That dynamical structure is the common core linking conventional electron-phonon superconductors, spin-fluctuation models, and quantum-critical metals [1807.04907], [1911.05065].

## 3. Retardation, spectral functions, and deviations from BCS

For conventional superconductors, the central Eliashberg quantity is the electron-phonon spectral function \(\alpha^2F(\omega)\). The coupling constant is
\[
\lambda = 2\int_0^\infty \frac{\alpha^2F(\omega)}{\omega}\, d\omega,
\]
and \(T_c\) depends not only on \(\lambda\) but on the detailed shape of \(\alpha^2F(\omega)\), on logarithmic and higher phonon moments such as \(\omega_{\log}\) and \(\bar{\omega}_2\), and on the Coulomb pseudopotential \(\mu^*\) [2106.05235]. This is the standard strong-coupling sense of the Eliashberg effect: superconductivity is controlled by a retarded, structured bosonic spectrum rather than by a single static attraction.

The weak-coupling limit of Eliashberg theory is not identical to BCS. For an Einstein phonon spectrum, the weak-coupling critical temperature becomes
\[
T_c = \frac{1.13}{\sqrt{e}}\,\omega_E\, \exp\!\left(-\frac{1+\lambda}{\lambda}\right),
\]
rather than the naive BCS prefactor, and the Matsubara-axis gap function approaches the universal form
\[
\Delta_0(\omega_m)=\frac{1}{1+\bar{\omega}_m^2}.
\]
On the real axis, the zero-temperature gap edge acquires the same \(1/\sqrt{e}\) correction,
\[
\Delta_0 = \frac{2\omega_E}{\sqrt{e}\,\exp\!\left(-\frac{1+\lambda}{\lambda}\right)},
\]
while the ratio \(2\Delta_0/(k_B T_c)\) tends back to its BCS value because the same prefactor appears in both scales [1807.04907], [1912.09460]. This directly shows that retardation survives even as \(\lambda\to 0\).

Several modern developments modify the kernel without abandoning Eliashberg structure. A machine-learned formula for \(T_c\), trained on 2874 spectra and numerical isotropic Eliashberg solutions, preserves the McMillan/Allen-Dynes exponential structure but improves accuracy for multimodal spectra and higher-\(T_c\) hydrides, reducing percent RMSE from \(14.4\%\) to \(8.4\%\) for non-hydrides, from \(45.1\%\) to \(9.2\%\) for artificial Gaussians, and from \(36.6\%\) to \(5.8\%\) for hydrides when the learned prefactor \(f_\omega\) is used [2106.05235]. Another extension incorporates nonlinear \(1\)-electron–\(2\)-phonon coupling and leaves the Migdal-Eliashberg equations formally unchanged except for a generalized \(\alpha^2F\); in the toy model discussed there, increasing the nonlinear coupling raises \(T_c\) monotonically [2503.04560].

A further implication is that the “effect” is increasingly tied to the structure of the interaction kernel rather than to any specific boson species. Whether the bosonic mode is a phonon, a spin resonance, or a collective critical fluctuation, the same dynamical logic persists.

## 4. Nonequilibrium Eliashberg effect: microwaves and cavities

In the narrow historical sense, the Eliashberg effect is the **microwave enhancement of superconductivity** in a dirty superconductor. The mechanism is nonequilibrium quasiparticle redistribution: a weak ac field depletes occupation near the gap edge, where quasiparticles are most pair breaking, and can thereby increase the order parameter \(\Delta\) and the critical current. The original regime is
\[
\gamma_\text{in}\ll (\hbar\omega,\Delta)\ll k_B T,
\]
with slow inelastic relaxation, weak microwave power, and temperatures near \(T_c\) [1911.04959].

The generalized theory valid at arbitrary \(T\), \(\omega\), dc supercurrent, and \(\gamma_\text{in}\) separates the nonequilibrium correction into a **spectral** contribution and a **kinetic** contribution,
\[
\mathcal{F}_\textrm{neq} = \mathcal{F}_\textrm{neq}^\textrm{sp} + \mathcal{F}_\textrm{neq}^\textrm{kin}.
\]
The kinetic term is the Eliashberg enhancement mechanism proper; the spectral term represents direct depairing. Near \(T_c\), the lower threshold frequency is
\[
\hbar\omega_\text{min}(T)\approx \sqrt{ \frac{2\pi \gamma_\text{in}\Delta_0(T)}{\ln[\Delta_0(T)/\gamma_\text{in}]} },
\]
and the generalized treatment gives the more precise minimum
\[
\hbar\omega_\text{min,min} = 3.23\,\gamma_\text{in}.
\]
The enhancement region is finite in the \((\omega,T)\) plane, roughly
\[
T \gtrsim 0.5\,T_c,\qquad \hbar\omega \lesssim 3\,k_B T_c,
\]
and disappears at low \(T\) or sufficiently high \(\omega\), where direct depairing and heating dominate [1911.04959].

The cavity quantum generalization replaces a coherent microwave drive by a **quantized photon bath**. In this setting the gap is modified not by changing the microscopic pairing interaction \(g\) but by changing the quasiparticle distribution \(n(E)\) entering the gap equation,
\[
\frac{1}{g}=\int \frac{dE}{E}\,\nu_{\rm qp}(E)\,[1-2n(E)].
\]
The cavity-induced fluctuation correction has the form
\[
iS^{\mathrm{fluc}}=\frac{\nu\Delta^q}{2}\int_0^\infty d\omega\, J(\omega)\,[N(\omega)-\mathcal B(\omega)]\,Y(\omega),
\]
and the steady-state redistribution is obtained from a kinetic equation with a cavity collision integral and inelastic relaxation time \(\tau_{\rm in}\) [1805.01482]. The same basic logic extends further in driven-dissipative cavity settings, where superconducting and charge-density-wave transitions acquire a non-equilibrium cubic term in the effective Landau functional and can become discontinuous with bistability [2509.07865].

## 5. Spin fluctuations, multiband systems, and unconventional superconductors

In unconventional superconductors, Eliashberg theory is often reinterpreted with a bosonic spectrum built from **spin fluctuations** rather than phonons. In optimally doped Bi-2212, a \(d\)-wave Eliashberg analysis replaces \(\alpha^2F(\omega)\) by a spin-fluctuation spectral function
\[
F(\omega,T)=H(T)F(\omega)^{Res}+BG(T)F(\omega)^{BG},
\]
with a sharp resonance near \(\Omega_{Res}\approx 40\,\mathrm{meV}\). The resonance amplitude decreases as
\[
H(T)=H_0\left(1-\frac{T}{90\,\text{K}}\right)^{1/2},
\]
vanishing at \(T_c=90\) K. In that treatment, \(\Delta(0)=38\) meV and
\[
\frac{2\Delta(0)}{k_B T_c}\approx 9.3,
\]
whereas holding the glue spectrum fixed at its \(T=0\) form yields \(T_c\approx 153\) K and \(2\Delta(0)/(k_B T_c)\approx 5.7\). The paper therefore attributes the observed large gap ratio to the temperature dependence of the magnetic resonance itself [1205.2381].

A related feedback structure appears in iron pnictides. In a three-band \(s_\pm\) Eliashberg model for LaFeAsO\(_{1-x}\)F\(_x\), SmFeAsO\(_{1-x}\)F\(_x\), Ba\(_{1-x}\)K\(_x\)Fe\(_2\)As\(_2\), and Ba(Fe\(_x\)Co\(_{1-x}\))\(_2\)As\(_2\), the spin resonance energy is taken to soften with temperature according to
\[
\Omega_0(T)=\Omega_0 \tanh\!\left(1.76\sqrt{\frac{T_c^*}{T}-1}\right),
\]
so that the condensate feeds back on the boson that mediates pairing. This allows the model to reproduce both gap magnitudes and experimental \(T_c\) with \(\lambda_{tot}\approx 1.7-2.0\) [1101.3473].

The same formalism has been applied to more specific multiband systems. For LiFeAs, a four-band \(s_\pm\) Eliashberg model requires an additional intraband term \(\lambda_{11}\sim 0.9\) together with \(\lambda_{tot}^{sf}\sim 1.5\) to reproduce the largest gap and \(H_{c2}\); the paper explicitly notes that this effective intraband term may be a fictitious consequence of Migdal-theorem violation in the low-Fermi-energy first band [1304.3638]. For bulk FeSe, a full-bandwidth anisotropic Eliashberg calculation with RPA spin and charge fluctuations yields \(T_c\approx 6\) K and a maximum gap \(\sim 1.4\) meV, consistent with experiment, whereas the same spin-fluctuation framework gives only \(T_c\le 11\) K and a \(d\)-wave gap in monolayer FeSe/SrTiO\(_3\), implying that spin fluctuations alone cannot explain the observed high-\(T_c\) state there [2004.01539].

Beyond iron-based materials, Eliashberg formulations have also been used for interfacial magnon-mediated superconductivity, where a full-frequency treatment substantially lowers the \(p\)-wave \(T_c\) relative to weak-coupling estimates because the effective cutoff on the magnon propagator is much smaller than the magnon bandwidth; the \(d\)-wave phase near half filling is less strongly affected because it relies less on long-wavelength magnons [2105.02235].

## 6. Quantum-critical applicability, validity limits, and contemporary extensions

A central modern issue is when Eliashberg theory remains controlled once one leaves the conventional phonon problem. In the comparative analysis of electron-phonon and nematic quantum-critical systems, the control parameter for phonons is
\[
\lambda_E=\frac{g^2}{E_F\omega_D}=\lambda\frac{\omega_D}{E_F}.
\]
Vertex corrections scale as \(\Delta g/g\sim \lambda_E\) for generic phonon momentum, and the two-loop self-energy satisfies
\[
\Sigma^{(2)}(\omega_m)\sim \left(\lambda_E |\log\lambda_E|\right)\Sigma^{(E)}(\omega_m).
\]
Hence the expansion is parametrically valid when \(\lambda_E\ll1\). However, as the phonon softens, \(\lambda_E\) grows, and Eliashberg theory breaks down at finite distance before \(\omega_D\to 0\), signaling a crossover toward a different, polaronic description [2404.11820].

Near an Ising-nematic or Ising-ferromagnetic quantum-critical point, by contrast, the bare boson is not parametrically slower than the fermions. After Landau damping, the propagator becomes
\[
\chi(q)=\frac{\chi_0}{\xi^{-2}+q^2+\alpha |\Omega_n|/|q|},
\]
and the relevant small parameter is
\[
\lambda_E^*=\frac{g^*}{E_F}.
\]
In the Fermi-liquid and Fermi-gas regimes, vertex corrections are small; in the quantum-critical regime they are not parametrically small, but the explicit calculation gives
\[
\Sigma^{(2)}(\omega_m)\simeq 0.038\,\Sigma^{(E)}(\omega_m),
\]
so the theory remains numerically accurate all the way to the QCP. The paper’s striking conclusion is that Eliashberg theory is, in this sense, **better behaved near the nematic QCP than in the soft-phonon problem** [2404.11820].

This broadened applicability is reinforced by reviews of pairing in non-Fermi liquids, which identify controlled limits where Eliashberg equations become asymptotically exact even in the absence of coherent quasiparticles. One route is a Fermi surface coupled to gapless bosons in a double expansion in small \(\epsilon=z_b-2\) and large \(N\); another is a large-\(N\) SYK-like setting with local frustrated interactions, where the Eliashberg equations arise as exact saddle-point equations [1912.07646]. In the Yukawa-coupled SYK framework, the exact large-\(N\) equations
\[
\Sigma(i\omega_m) = -\overline g^2 T \sum_{m'}D(i\omega_m - i\omega_{m'})G(i\omega_{m'}),
\]
\[
\Phi(i\omega_m) = (1-\alpha)\overline g^2 T \sum_{m'}D(i\omega_m - i\omega_{m'})F(i\omega_{m'}),
\]
\[
\Pi(i\nu_m) =  -2\overline g^2 T \sum_{m'} \left[G(i\omega_{m'} + i \nu_m) G(i\omega_{m'}) - (1-\alpha) F(i\omega_{m'} + i \nu_m) F(i\omega_{m'})\right]
\]
define a quantum-critical Eliashberg theory in which superconductivity emerges from an incoherent non-Fermi-liquid state [2506.11952].

Contemporary extensions continue to preserve the Eliashberg logic while changing the microscopic content. In twisted graphene, band-off-diagonal pairing has been treated within Eliashberg theory, where the leading-order interaction from intervalley phonons or intervalley-coherent fluctuations forbids admixture of an intraband component while permitting even- and odd-frequency mixing [2501.12435]. In electrostatically doped films, a proximity Eliashberg theory couples a field-modified surface layer to the bulk and concludes that the best route to \(T_c\) enhancement is a very thin film with a strong field-induced increase of the electron-phonon coupling [1704.08159]. At the first-principles level, a \(GW_0\)-Eliashberg benchmark for the uniform electron gas shows that full momentum- and frequency-dependent screened Coulomb and phonon interactions, together with the normal-state self-energy, are essential; neglecting the normal-state Coulomb self-energy can overestimate \(T_c\) by roughly an order of magnitude in some regimes [2508.13779].

Taken together, these developments suggest a unifying view. The Eliashberg effect is best understood not as a single isolated result, but as the recurring consequence of **retarded boson-mediated interactions treated through self-consistent, frequency-dependent fermionic and bosonic self-energies**. In some settings it appears as a specific nonequilibrium enhancement mechanism; in others it is the organizing principle for strong-coupling superconductivity, quantum-critical pairing, or multiband dynamical order.

Source: https://www.emergentmind.com/topics/eliashberg-effect