Electric-Field-Induced Superconductivity

• Electric-field-induced superconductivity is a phenomenon where an external electric field triggers superconducting states through mechanisms like electrostatic carrier accumulation, subband quantization, and electrochemical transformation.
• The topic features experimental and theoretical methodologies using field-effect devices, ionic liquids, and gate-controlled setups to modulate superconducting phase transitions and critical temperatures.
• It has significant implications for designing low-dimensional and multifunctional superconducting systems, with research continually addressing phase competition, pairing symmetry shifts, and interface effects.

Electric-field-induced superconductivity denotes superconducting behavior that is created, stabilized, or reconfigured by an external electric field rather than by conventional chemical substitution or pressure. In the literature this expression covers several distinct regimes: electrostatic carrier accumulation in field-effect or electric-double-layer transistor geometries, electrochemically assisted transformations that permanently change stoichiometry, electric-field-driven band-structure and symmetry engineering in correlated-electron models, and gate-controlled switching between superconducting, metallic, insulating, and topological phases in low-dimensional materials (Mizohata et al., 2013, Zhang et al., 2018, Shiogai et al., 2015, Choi et al., 2024).

1. Conceptual scope and principal regimes

A basic distinction is between electrostatic and electrochemical control. In electrostatic surface superconductivity, a gate field accumulates carriers within a screened surface layer, and superconductivity disappears when the field is removed. In electrochemical regimes, the field drives ion motion or redox processes, so the superconducting state can become permanent and bulk-like even after the gate voltage is released. A second distinction concerns what the field acts on most strongly: carrier density, subband structure, interlayer or sublattice polarization, or a competing ordered state such as antiferromagnetism or spin-density-wave order (Zhang et al., 2018, Mizohata et al., 2013, Haraldsen et al., 2011).

This diversity is visible already at the model level. In AA-stacked bilayer graphene, a perpendicular electric field is represented by a layer-dependent potential and drives a dominant chiral $d+id$ pairing tendency by increasing the density of states near the Fermi energy and suppressing antiferromagnetic spin correlations (Fang et al., 2019). In doped silicene, the field-induced staggered potential $\Delta$ breaks sublattice symmetry and yields a quantum phase transition from singlet chiral $d+id'$ to triplet $f$-wave superconductivity while enhancing $T_c$ through density-of-states effects (Zhang et al., 2013). In bilayer octagraphene, an interlayer potential difference $V$ weakens Fermi-surface nesting, raises the critical interaction for SDW order, and stabilizes unconventional $s^{\pm}$ pairing in the weak-coupling regime (Yao et al., 3 Jul 2025).

2. Electrostatic surface accumulation and confined superconductivity

The canonical experimental platform is the electric double-layer transistor. In layered nitrides and related systems, the electric double layer at an ionic-liquid/solid interface can reach capacitances as large as $10\text{–}100\ \mu{\rm F/cm^2}$, enabling surface charge densities $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$, substantially larger than in conventional solid dielectrics (Zhang et al., 2018). In oxide surface superconductivity, the same logic produces a quasi-2D accumulation layer confined within a few to a few tens of nanometers. For SrTiO$_3$, a representative analysis used a surface electric field $\Delta$0 and sheet density $\Delta$1; solving the Bogoliubov–de Gennes equations showed that confinement quantizes the $\Delta$2-motion into subbands and produces a multiple-gap superconducting state with in-gap states even for isotropic s-wave pairing (Mizohata et al., 2013).

This subband physics alters the spectroscopy and the depth dependence of superconductivity. Lower subbands are localized closer to the surface and feel a larger effective $\Delta$3, whereas higher subbands extend deeper and acquire smaller gaps. The resulting local density of states therefore contains a spread of gap edges rather than a single BCS edge, and the integrated density of states can exhibit low-energy weight that superficially resembles nodal or anisotropic superconductivity even though the pairing interaction is isotropic s-wave (Mizohata et al., 2013).

At complex oxide interfaces the electric field also enters at the Ginzburg–Landau level. For LaAlO$\Delta$4/SrTiO$\Delta$5, inversion-symmetry breaking at the interface allows a coupling $\Delta$6, while the same gate field changes the interfacial carrier density through STO’s nonlinear dielectric response. By relating measured capacitance to the density of states and then to a bulk-STO electron-phonon pairing scale, a gate-tunable $\Delta$7 dome with a peak near $\Delta$8 K was estimated (Haraldsen et al., 2011).

Surface electrostatic superconductivity is not restricted to single crystals. In bulk polycrystalline MoS$\Delta$9, electric double-layer doping of a millimeter-thick pellet produced an insulator-to-metal transition and then a low-temperature superconducting resistance drop localized at the surface. The onset temperature $d+id'$0 increased strongly with conductance and then saturated at about $d+id'$1 K, although no zero-resistance state or magnetic-field-derived gap parameters were obtained in that work (Shimazu et al., 3 Sep 2025).

3. Electrochemical routes and permanent superconducting states

A central development was the demonstration that an electric field can do more than accumulate carriers transiently: it can permanently transform a layered insulator into a bulk superconductor. In HfNCl and ZrNCl, an ionic-liquid EDLT converts the parent wide-gap insulators into superconductors with $d+id'$2 K for HfNCl single crystals and $d+id'$3 K for ZrNCl (Zhang et al., 2018). The outcome depends sharply on the temperature at which the gate voltage is applied. At $d+id'$4 K, applying positive and negative $d+id'$5 yields a reversible insulator–superconductor cycle attributed to short-range motion of partial Cl$d+id'$6 ions. At $d+id'$7 K and above, the same device enters an electrochemical regime in which partial Cl deintercalation produces irreversible superconductivity that persists after the gate voltage is removed and after thermal cycling (Zhang et al., 2018).

The irreversible state is bulk-like rather than a thin surface sheath. In HfNCl, a superconducting transition with onset around $d+id'$8 K and zero resistance at $d+id'$9 K is accompanied by a clear diamagnetic transition in $f$0. In ZrNCl, permanent superconductivity near $f$1 K appears only after high-temperature gating; after low-temperature gating, no diamagnetic transition is visible down to $f$2 K (Zhang et al., 2018). X-ray diffraction showed no shift of the $f$3 peaks, arguing against intercalation of large ionic-liquid species and supporting the interpretation in terms of subtle Cl deintercalation rather than foreign-ion insertion.

This electrochemical regime establishes a direct connection between field-effect doping and synthesis. The field ceases to function merely as a capacitor and becomes a route to chemical transformation, producing HfNCl$f$4 or ZrNCl$f$5 with superconducting properties close to chemically doped analogues (Zhang et al., 2018). A plausible implication is that other layered materials with weakly bound anion or cation sublattices may admit analogous gate-driven synthesis pathways.

4. Correlated-electron and band-engineering mechanisms

In lattice models for graphene-derived and related systems, the electric field commonly appears as an interlayer or sublattice potential that reshapes flat bands, density of states, and spin fluctuations. In AA-stacked bilayer graphene, constrained-path auxiliary-field quantum Monte Carlo on the half-filled Hubbard model with interlayer potential difference $f$6 found that the dominant interacting pairing channel is nearest-neighbor chiral $f$7. The effective long-range pairing correlation $f$8 increases with both $f$9 and $T_c$0, while intralayer antiferromagnetic nearest-neighbor spin correlations are suppressed as the field grows (Fang et al., 2019).

Doped silicene provides a different mechanism rooted in its low-buckled lattice. A perpendicular field induces a staggered potential $T_c$1 because the two sublattices sit at different heights, thereby flattening the relevant bands and selecting an effectively single-sublattice low-energy sector. Random-phase-approximation calculations then predict a field-driven change from singlet chiral $T_c$2 to triplet $T_c$3-wave pairing, with the enhancement of $T_c$4 tied to the increased density of states and to ferromagnetic-like intra-sublattice spin correlations at low doping (Zhang et al., 2013).

Bilayer octagraphene illustrates the same logic in a nesting-driven setting. There the perpendicular field is an interlayer potential $T_c$5 that enlarges band splitting, shrinks inner Fermi pockets, expands outer pockets, and weakens the nesting responsible for SDW order. The RPA phase competition yields a leading $T_c$6 $T_c$7 superconducting channel in the paramagnetic regime below the critical interaction $T_c$8, with a pairing eigenvalue $T_c$9 at $V$0 eV and $V$1 eV, while the subleading $V$2 channel has $V$3 (Yao et al., 3 Jul 2025).

5. Conventional thin films, field-modulated $V$4, and supercurrent suppression

In conventional metallic superconductors the electric field usually modulates superconductivity rather than creating it from an insulating state. A microscopic example is indium thin films, where a static field $V$5 induced a critical-temperature shift of order $V$6 K around bulk $V$7 K. Proximity-effect Eliashberg theory with ab-initio input reproduced these decades-old experiments by treating the field-affected region as a thin surface layer of thickness $V$8 nm, of the same order as the calculated Thomas–Fermi screening length $V$9 nm (Ummarino et al., 2019).

More recent microscopic theory for metallic films emphasizes confinement and screening. A phonon-mediated thin-film model that includes quantum-confinement effects on $s^{\pm}$0, the Fermi energy, Thomas–Fermi screening in the electron-phonon matrix element, and the thickness dependence of the Coulomb pseudopotential predicts a critical electric field $s^{\pm}$1 above which superconductivity is suppressed. In that theory, $s^{\pm}$2 is lower for thinner films, in agreement with supercurrent field-effect experiments (Zaccone et al., 2023).

A complementary Ginzburg–Landau analysis attributes the field effect to a gap-dependent permittivity $s^{\pm}$3. Using $s^{\pm}$4 nm and $s^{\pm}$5 nm, it reproduces supercurrent quenching in thin films with critical fields of order $s^{\pm}$6 V/m, a very small $s^{\pm}$7 mK, and the observed thickness dependence: strong suppression only when the film thickness is of order $s^{\pm}$8 (Amoretti et al., 2022). By contrast, a time-dependent Ginzburg–Landau treatment of low-carrier-density systems with preformed pairs argues that a penetrant electric field can shift the effective GL coefficient so that pairs above $s^{\pm}$9 Bose-condense, thereby increasing the critical temperature and inducing superconductivity in systems that are insulating in the normal state, superconducting semiconductors at low carrier concentration, or strongly interacting ultracold fermions (Karchev et al., 2018).

6. Representative material platforms and experimentally realized phases

Several material systems combine large gate tunability with unusually high or unconventional superconducting responses. In hydrogenated diamond (111), first-principles calculations that include the self-consistent field effect on structure, bands, phonons, and electron-phonon coupling predict a multiband phonon-mediated superconducting state at hole densities up to $10\text{–}100\ \mu{\rm F/cm^2}$0. At that density the full Brillouin-zone calculation gives $10\text{–}100\ \mu{\rm F/cm^2}$1 and $10\text{–}100\ \mu{\rm F/cm^2}$2 K, with superconductivity supported mainly by in-plane diamond phonons and significant intra- and interband scattering (Romanin et al., 2019).

In ultrathin FeSe, the electric field can cooperate with top-down thickness reduction. Electrochemically etched FeSe transistors on SrTiO$10\text{–}100\ \mu{\rm F/cm^2}$3 and MgO display high-$10\text{–}100\ \mu{\rm F/cm^2}$4 superconductivity around $10\text{–}100\ \mu{\rm F/cm^2}$5 K under a gate voltage of $10\text{–}100\ \mu{\rm F/cm^2}$6 V, with $10\text{–}100\ \mu{\rm F/cm^2}$7 K on SrTiO$10\text{–}100\ \mu{\rm F/cm^2}$8 and $10\text{–}100\ \mu{\rm F/cm^2}$9 K on MgO (Shiogai et al., 2015). The field expands the observable thickness window for the high-$$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$0 state up to about $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$1 unit cells and enables an electrostatically controlled insulator–superconductor transition at fixed thickness, implying that field-induced reconstruction of both conduction- and valence-band sectors is central to the FeSe problem (Shiogai et al., 2015).

Flat-band rhombohedral graphene extends electric-field control into topological regimes. Rhombohedral tetralayer graphene aligned to hBN hosts a zero-field quantized anomalous Hall state at $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$2 and a superconducting state at large negative filling. In one device the superconducting pocket had $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$3 mK, $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$4 mT, and $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$5 nA, while the QAH state exhibited $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$6, $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$7, and nonvolatile gate-controlled switching of chirality (Choi et al., 2024). Thermodynamic compressibility further revealed a fractional Chern insulator at $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$8, and adding a TMD layer nucleated an additional superconducting pocket while leaving the $$n_{\rm 2D}\sim 10^{14}\text{–}10^{15}\ {\rm cm^{-2}}$$9 QAH topology intact (Choi et al., 2024). This suggests an intrinsically low-disorder route to proximity between superconductivity and chiral or fractionally charged edge modes.

7. Phase competition, evidentiary standards, and unresolved issues

A recurring theoretical theme is the competition among superconducting, metallic, and insulating surface states under a screened field. Self-consistent BdG calculations for the one-dimensional attractive Hubbard model find that a screened surface potential can drive the surface of a bulk superconductor through a direct superconducting–insulating transition at $_3$0, while at finite temperature the sequence becomes superconducting $_3$1 metallic $_3$2 insulating. The calculated phase diagram in temperature and field is in qualitative agreement with transport phase diagrams reported for (Li,Fe)OHFeSe thin flakes (Yin et al., 2023).

The strongest experimental controversies concern claims that lack definitive superconducting signatures. In hydrogenated graphitic fibers, increasing DC current density produces linear-in-$_3$3 resistivity, plateaus near $_3$4 K, and nonlinear random-resistor-network transport that the authors interpret as compatible with electric-field-induced superconductivity. However, the same work explicitly reports neither zero resistance nor a Meissner effect, so the evidence remains transport-based and nonconclusive (Gheorghiu et al., 2020). Such cases underscore that “electric-field-induced superconductivity” can denote anything from a rigorously established bulk superconducting phase with diamagnetic shielding to a field-tuned transport anomaly compatible with pairing.

The central unresolved issue is therefore not whether electric fields matter, but how they matter in any given platform. Depending on the material, the field may act mainly by electrostatic carrier accumulation, by subband quantization, by interlayer or sublattice polarization, by suppression of competing magnetism, by electrochemical stoichiometric change, or by modification of screening and Coulomb repulsion (Zhang et al., 2018, Mizohata et al., 2013, Fang et al., 2019, Zaccone et al., 2023). This suggests that future progress will depend less on a single universal mechanism than on controlled separation of these channels within the same device architecture, especially in systems where superconductivity coexists with topological or magnetic order.