Holographic Insulator/Superconductor Model

Updated 5 March 2026

The model is a gauge/gravity duality construction using a gapped AdS soliton that transitions to a superconducting phase via scalar or vector field condensation.
Analytic methods, such as the matching technique, determine a critical chemical potential (around 4.5 for m²=0) and yield a mean-field critical exponent of 1/2.
Extensions involving p-wave, higher-derivative corrections, and dark sector couplings reveal rich phase diagrams with reentrant and nonlinear phenomena.

A holographic insulator/superconductor (I/S) model is a class of gauge/gravity duality constructions in which a gravitational system with a mass gap (the "insulator") exhibits a quantum phase transition to a gapless, symmetry-breaking state (the "superconductor") as a function of chemical potential, temperature, or external parameters. In such models, the insulator phase is realized by a gapped AdS soliton geometry, while the superconducting phase emerges from scalar or vector field condensation in the bulk. These models provide an analytic and numerical framework to study strongly coupled analogs of quantum phase transitions, capturing both static and dynamical features (e.g., optical conductivity, phase diagrams, critical exponents) and allowing precise exploration of universal and model-dependent properties in high-dimensional condensed-matter systems.

1. Model Construction and Bulk Setup

Holographic I/S models are based on $(d+1)$ -dimensional Einstein gravity coupled to gauge fields and matter, set in asymptotically AdS spacetimes with a solitonic deformation. The canonical $d=4$ and $d=5$ models use the following action in the probe (minimal coupling) limit: $S = \int d^{d+1}x\, \sqrt{-g} \left[ R + \frac{d(d-1)}{L^2} - \frac{1}{4}F_{\mu\nu}F^{\mu\nu} - |\nabla\psi - i q A\psi|^2 - m^2|\psi|^2 \right],$ where $F=dA$ is the Maxwell field and $\psi$ is a complex scalar. For p-wave and higher-order transitions, $SU(2)$ Yang–Mills or more complex vector fields are employed (Horowitz et al., 2010, Akhavan et al., 2010, Cai et al., 2013).

The soliton metric, which realizes a mass gap, is given in 5D as

$ds^2 = r^2( -dt^2 + dx^2 + dy^2 ) + \frac{dr^2}{f(r)} + f(r)\, d\chi^2,\quad f(r) = r^2 \left( 1 - \frac{r_0^4}{r^4} \right),$

with $\chi$ periodically identified to remove the tip conical singularity. In this background, excitations are gapped, and the boundary dual theory is an insulator.

2. Equations of Motion and Boundary Conditions

The standard ansatz for the condensate phase involves $\psi=\psi(r)$ and $A_t=\phi(r)$ (with other components vanishing). The equations of motion (EOM) in $d=4$ or $5$ reduce to two coupled ODEs: $\begin{aligned} &\psi'' + \left( \frac{f'}{f} + \frac{d-2}{r} \right)\psi' + \left( \frac{\phi^2}{r^{2}f} - \frac{m^2}{f} \right)\psi = 0,\ &\phi'' + \frac{d-2}{r} \phi' - \frac{2 \psi^2}{f} \phi = 0. \end{aligned}$ At the soliton tip $r=r_0$ , regularity conditions fix the expansions for $\psi$ and $\phi$ (finite $\phi(r_0)$ , Taylor series for $\psi(r)$ ). Near the AdS boundary ( $r\to\infty$ ), the asymptotic behavior is

$\psi(r) \sim \frac{\psi_-}{r^{\Delta_-}} + \frac{\psi_+}{r^{\Delta_+}},\quad \phi(r) \sim \mu - \frac{\rho}{r^{d-3}},$

with scaling exponents $\Delta_\pm$ determined by $m^2$ . Standard quantization corresponds to $\psi_-=0$ , identifying $\psi_+$ as the order parameter $\langle \mathcal{O}_+ \rangle$ (Parai et al., 2021).

3. Analytic Approaches: Matching Method and Thermodynamic Geometry

3.1. Matching Method

Parai, Ghorai, and Gangopadhyay constructed an analytic determination of the critical chemical potential $\mu_c$ via a local expansion ("matching method") about the tip and boundary. After transforming to $z=r_0/r\in[0,1]$ , the EOMs are expanded near $z=1$ and $z=0$ ; these expansions are then matched at an intermediate point $z=z_m$ . In the linearized regime (just below the critical point), this yields an algebraic equation for $\mu_c$ . For $m^2=0,\, z_m=1/2$ ,

$9\mu_c^4 - 160\mu_c^2 - 512 = 0 \implies \mu_c \simeq 4.533.$

A generalized formula for arbitrary $z_m=\lambda$ gives

$\mu_c = \frac{1}{\lambda} \sqrt{ m^2 + 4(\Delta_+ - 3)\lambda^{-(\Delta_+-1)} }.$

This matching approach reproduces $\mu_c$ to within $15-20\%$ of full numerics (Parai et al., 2021).

3.2. Thermodynamic Geometry

The free energy per unit volume, evaluated on-shell, provides a generating function for a Ruppeiner-like geometric analysis: $\frac{\Omega}{V_3} = -\mu\,\rho + c_4 \mu^4 + c_6 \mu^6 + \ldots,$ where coefficients $c_n$ are extracted analytically via the matched expansion. The scalar curvature $R$ of the metric $g_{ij} = -\partial^2 \Omega/\partial x^i \partial x^j$ (with $x^1=\mu, x^2=\rho$ ) diverges when $\det g_{ij}=0$ , providing an independent criterion for $\mu_c$ . For $m^2=0,\, \lambda=2$ , this gives $\mu_c\simeq 4.703$ (Parai et al., 2021).

4. Phase Structure, Order Parameter, and Critical Exponents

For $\mu<\mu_c$ , only the trivial solution $\psi=0$ exists (insulator). As $\mu$ increases past $\mu_c$ , a second-order quantum phase transition occurs, and a nontrivial condensate

$\langle \mathcal{O}_+ \rangle \sim C\sqrt{\mu - \mu_c}$

develops, with mean-field exponent $1/2$. For $m^2=0$ , the prefactor is $C\approx 6.86$ (Parai et al., 2021). This scaling is robust to model deformations, appearing universally in both analytic and numerical studies (Horowitz et al., 2010, Cai et al., 2011).

Backreaction, higher-order corrections, and the presence of additional gauge sectors (e.g., “dark matter” $U(1)$ ) can shift the order of transition to first order under specific parameter regimes (notably, at small scalar charge or large dark-matter coupling), as revealed by swallow-tail features in $F(\mu)$ and by discontinuities or kinks in the entanglement entropy (Peng et al., 2015, Peng et al., 2016). Retrograde (thermodynamically unstable) condensation is also possible (Peng et al., 2015).

5. Extensions and General Phenomenology

5.1. p-wave, Weyl, and Gauss-Bonnet Generalizations

Analogous constructions for vector (p-wave) order parameters employ $SU(2)$ Yang-Mills actions. The critical chemical potential is generically lower for p-wave (indicating easier condensation) and is sensitive to higher-derivative (Gauss-Bonnet, Weyl) corrections, which raise $\mu_c$ and suppress condensation, typically without affecting the critical exponent ($1/2$) (Pan et al., 2011, Zhao et al., 2012, Akhavan et al., 2010).

5.2. Multi-order, Mott, and Nonlinear Effects

Multi-scalar (“two-order parameter”) models exhibit a rich phase diagram, with both first- and second-order transitions between purely $\Psi_1$ , purely $\Psi_2$ , or mixed-condensate superconducting phases (Peng et al., 2016). Nonlinear gauge interactions (e.g., DBI, iDBI) generate Mott-like insulating behavior (DC conductivity strictly suppressed at zero temperature), and transitions to a superconducting dome as doping increases (Baggioli et al., 2016). The presence of lattice structures, disorder, or coupling to additional $U(1)$ sectors further enriches the phase structure with reentrance and anisotropic effects (Erdmenger et al., 2015, Ling et al., 2017).

5.3. Entanglement Entropy and Nonlocal Probes

Topological entanglement entropy and Wilson loops track the insulator/superconductor transition and provide nonlocal diagnostic of criticality, with entropy exhibiting nonmonotonic or kink/jump behavior at first- or second-order points (Cai et al., 2012, Yao et al., 2018). In s-wave models, entanglement entropy is universally nonmonotonic as a function of $\mu$ above the transition.

6. Complete Phase Diagrams and Reentrant Phenomena

The full $(T,\mu)$ phase diagram includes four phases—insulator, soliton superconductor, conductor (Reissner-Nordström AdS black hole), and black hole superconductor. The dominant phase is determined by comparison of the grand potential (free energy). Transitions between phases occur via lines of first or second order, with two triple points present for suitable parameter choices. Notably, reentrant transitions (superconductor $\rightarrow$ insulator upon lowering $T$ ) can appear for appropriately tuned scalar charge and backreaction, revealing nontrivial interplay between the four phases (Horowitz et al., 2010).

7. Universality, Numerical–Analytic Agreement, and Critical Behavior

The various analytic methods (matching, Sturm-Liouville, thermodynamic geometry) yield critical potentials and exponents in close agreement with numerics, typically overestimating $\mu_c$ by 15–20%, but capturing the phase-transition order and scaling (Parai et al., 2021, Cai et al., 2011). All known holographic I/S models at mean field (probe) level exhibit $1/2$ critical exponent for the condensate and a linear relation between charge density and $\mu-\mu_c$ near criticality, regardless of further model specifications.