Superfluid Weight Tensor

• Superfluid weight tensor is defined as the observable measuring superconducting phase stiffness, capturing both conventional (band curvature) and geometric (quantum metric) contributions.
• It decomposes into conventional terms that vanish in flat bands and geometric terms that dominate in topological or strongly correlated systems.
• Its derivation via Kubo formulas and symmetry constraints underpins predictions of measurable quantities like London penetration depth and BKT transition temperatures.

The superfluid weight tensor is a fundamental observable in superconductivity, characterizing the phase stiffness of a superconductor and determining its response to electromagnetic fields and phase gradients. It plays a central role in phenomena such as the Meissner effect, London penetration depth, and the Berezinskii-Kosterlitz-Thouless (BKT) transition in two-dimensional systems. The superfluid weight incorporates both conventional contributions associated with band dispersion and geometric contributions arising from the quantum geometry of electronic or bosonic Bloch states, including the quantum metric and, in nontrivial cases, non-Abelian generalizations. Recent work firmly establishes its decomposition, symmetry constraints, and unique physical consequences in both conventional and unconventional superconductors, multiorbital and flat-band systems, as well as quasicrystals.

1. Definition and Physical Interpretation

The superfluid weight tensor, commonly denoted $D_{ij}$, quantifies the second derivative of the equilibrium free energy $F$ of a superconductor with respect to a uniform vector potential $\mathbf{A}$ (or equivalently a phase twist or flux insertion) at zero field:

$$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$

where $V$ is the system volume. Equivalently, in linear response,

$\langle J_i \rangle = D_{ij} A_j$

links the induced supercurrent to the applied vector potential. In the London gauge, $D_{ij}$ directly determines the tensorial inverse squared penetration depth: $(\lambda^{-2})_{ij} = \mu_0 D_{ij}$, controlling the Meissner screening of magnetic fields (Peotta, 2022, Hiorth et al., 11 Mar 2026).

Alternative, but equivalent, formulations use the Kubo formula: $D_{ij}$ is the static, long-wavelength limit of the current-current response kernel,

$D_{ij} = K_{ij}(\mathbf{q} \to 0, \omega=0)$

with $F$0 comprising diamagnetic and paramagnetic terms, distinguished by their microscopic origins (Liang et al., 2016, Porlles et al., 23 May 2025).

In bosonic systems such as flat-band Bose-Einstein condensates (BECs), analogous definitions arise from imposing a long-wavelength phase twist and expanding the free energy, including necessary corrections for condensate-density and chemical-potential changes induced by the twist (Julku et al., 2022).

2. Decomposition: Conventional and Geometric Contributions

A general multiband formalism shows that at the mean-field level, $F$1 decomposes as: $F$2 where:

• $F$3 (conventional term) is controlled by the band curvature (Fermi velocity and its derivatives) and vanishes if the bands are perfectly flat,
• $F$4 (geometric term) arises from quantum geometric properties of the Bloch states—specifically the quantum metric tensor—and can remain finite even for exactly flat bands provided the quantum metric is nonzero (Liang et al., 2016, Herzog-Arbeitman et al., 2021).

The microscopic origin is elucidated by expanding the current operator matrix elements in the Bloch basis. The diagonal (intraband) part corresponds to $F$5, while the off-diagonal (interband) matrix elements yield $F$6. Closed-form expressions include

$F$7

and

$F$8

where $F$9 are components of the (non-Abelian) quantum metric tensor (Hiorth et al., 11 Mar 2026, Chen et al., 28 Jan 2025).

In systems with time-reversal symmetry, the geometric term can be expressed as an integral over the trace of the quantum metric: $\mathbf{A}$0 with $\mathbf{A}$1 the (non-Abelian) real quantum metric, and $\mathbf{A}$2 the Fermi distribution (Chen et al., 28 Jan 2025).

This decomposition persists for bosonic BECs, where $\mathbf{A}$3 (with $\mathbf{A}$4 the Fubini-Study metric at the condensate momentum) (Julku et al., 2022), as well as in superconductors with unconventional (momentum-dependent) pairing, although additional functional terms appear in the latter (Buthenhoff et al., 14 May 2025).

3. Quantum Geometry, Topological Bounds, and Flat Bands

The geometric contribution is fundamentally tied to the quantum metric of Bloch bands. In paradigmatic flat-band systems with uniform pairing and strong isolation from other bands, the geometric term dominates and can be written as: $\mathbf{A}$5 where $\mathbf{A}$6 depends on microscopic details (Herzog-Arbeitman et al., 2021, Xie et al., 2019).

Several rigorous lower bounds have been established:

• In 2D, the integrated quantum metric is bounded below by the modulus of the Berry curvature, and, for topologically nontrivial bands, by the Wilson-loop winding number or Chern number: $\mathbf{A}$7 leading to a nonzero lower bound on $\mathbf{A}$8, and thus on the BKT transition temperature $\mathbf{A}$9 (Xie et al., 2019, Herzog-Arbeitman et al., 2021, Jiang et al., 2024). Bands with obstructed Wannier centers—i.e., Wannier functions displaced from atomic sites—can be strictly identified using real space invariants (RSIs) and are guaranteed a positive superfluid weight even when Berry curvature vanishes identically (Herzog-Arbeitman et al., 2021).

For composite bands (multiple nearly flat and isolated bands), the geometric contribution includes both intra- and inter-band quantum metric terms. Provided the total Chern number is nonzero, a topological lower bound on $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$0 persists. Detailed perturbative expressions for these "composite quantum metrics" and their dependence on band representatives are available (Jiang et al., 2024).

Notably, in systems where one or more of isolation, flatness, or symmetry conditions are relaxed (e.g., multiorbital, spin-polarized, or three-dimensional systems), these bounds can be lost, and $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$1 can vanish or become negative within certain parameter regimes (Brzezicki et al., 2023).

4. Anisotropy, Symmetry, and Extensions

The superfluid weight tensor is intrinsically anisotropic in systems lacking full rotational or crystalline symmetry, with $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$2 elements reflecting the underlying symmetry group. For instance, orthorhombic or lower-symmetry crystals possess distinct diagonal and, potentially, off-diagonal components. The tensor structure is inherited from both the microscopic band structure and the symmetry of the superconducting order parameter (Hiorth et al., 11 Mar 2026, Julku et al., 2022).

Pairing symmetry profoundly influences the tensorial structure. For $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$3-wave (on-site) pairing, especially in the presence of $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$4 symmetry, $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$5 is isotropic, while in nematic (nearest-neighbor) or unconventional (e.g., chiral $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$6-wave) pairing, anisotropy and nonvanishing off-diagonal elements emerge—an effect directly observable in the kinetic inductance or penetration depth anisotropy (Julku et al., 2019, Buthenhoff et al., 14 May 2025).

For systems with degenerate bands, the appropriate quantum geometric tensor is non-Abelian, with all physically relevant quantities depending only on the trace of the real (symmetric) part. A crucial result is that a vanishing net Chern number does not guarantee a vanishing geometric contribution; in realistic systems, the metric can be responsible for a significant fraction of $$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$7, as seen in DFT calculations for MoS$$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$8 and TiSe$$D_{ij} = \frac{1}{V} \frac{\partial^2 F}{\partial A_i \partial A_j}\biggr|_{\mathbf{A}=0}$$9 (Chen et al., 28 Jan 2025).

5. Inhomogeneity, Quasicrystals, and Real-Space Formulation

Spatial inhomogeneity (e.g., vortex lattices, pinning, strong disorder, quasicrystal structures) requires more sophisticated formalisms, as translational symmetry is lost and momentum-space geometry is not directly accessible. In these situations, the superfluid weight retains a definition in terms of boundary-twist or real-space Kubo formulas: $V$0 where $V$1 are fluxes inserted through the holes of a torus or ring (Sun et al., 28 Jul 2025, Saito et al., 8 Jun 2026). A key development is the introduction of a "flux-space quantum metric", defined via derivatives with respect to fluxes: $V$2 with $V$3 the many-body ground state at flux $V$4.

In the flat-band, weak-coupling limit, the conventional component vanishes and the geometric term, proportional to the averaged local Wannier spread or its fluctuations, dictates the entire superfluid weight (Sun et al., 28 Jul 2025).

For quasicrystals hosting critical states, which are neither extended nor exponentially localized, the geometric contribution to $V$5 dominates at all relevant couplings. Real-space Kubo and momentum-space approaches reveal that $V$6 strictly traces the critical nature of the wavefunctions. This is an essential demonstration that geometric superfluidity extends far beyond periodic crystals (Saito et al., 8 Jun 2026).

6. Experimental Manifestations and Numerical Evaluation

The superfluid weight determines macroscopic observables such as the London penetration depth $V$7 and the BKT transition temperature in two dimensions. Explicitly,

$V$8

(for isotropic $V$9), and for a 2D superconductor,

$\langle J_i \rangle = D_{ij} A_j$0

(Hiorth et al., 11 Mar 2026, Xie et al., 2019, Julku et al., 2019). Experimentally, in conventional superconductors with wide bands such as Al, Pb, Nb, the geometric contribution is negligible (3–4 orders of magnitude smaller than the band-curvature term), and numerical evaluation using density functional theory band structures achieves agreement within 10–20% of measured London depths (Hiorth et al., 11 Mar 2026). In topological or flat-band candidates (e.g., twisted bilayer graphene, copper-doped apatite, $\langle J_i \rangle = D_{ij} A_j$1–𝒯$\langle J_i \rangle = D_{ij} A_j$2 lattice), the geometric term dominates or contributes significantly, with its magnitude tunable via band-structure engineering, electron filling, and interaction strength (Mojarro et al., 10 Dec 2025, Brzezicki et al., 2023).

For detailed numerical evaluation, dense $\langle J_i \rangle = D_{ij} A_j$3-grid sampling, analytic derivatives (e.g., via Wannier interpolation), and careful separation of band curvature and geometric contributions are standard. Real-space “superfluid-weight markers” can now be constructed to capture local diamagnetic responses even in the presence of strong disorder or spatial inhomogeneity, revealing direct physical signatures such as turbulence in the local supercurrent and modifications to the penetration depth (Porlles et al., 23 May 2025).

7. Generalizations, Theoretical Developments, and Open Questions

The functional approach provides a general framework for arbitrary pairing symmetries. In unconventional superconductors, a nontrivial functional term appears that encodes the response of the gap function to gauge fields. This is associated with Wilczek-Zee connections and two-point fidelity tensors, only vanishing for $\langle J_i \rangle = D_{ij} A_j$4-wave pairing. In realistic topological materials the geometric superfluid weight can thus be governed by non-Abelian quantum geometry and topological invariants. In composite or multi-orbital systems, the lattice geometric (orbital-position dependent) term cancels in the physical observable, ensuring gauge invariance and independence from microscopic model details (Buthenhoff et al., 14 May 2025, Jiang et al., 2024).

The true minimal bounds, universality, and the fate of the geometric superfluid weight under strong correlation, disorder, or topological transitions remain active topics of investigation. The generality of the real-space quantum metric approach enables studies in amorphous, fractal, and other non-crystalline media, offering a unified description of geometric effects in superfluidity.

Contribution	Dependency (Summary)	Dominance Regime
Conventional	Band curvature ($\langle J_i \rangle = D_{ij} A_j$5)	Dispersive/wide-band, conventional superconductors
Geometric	Quantum metric of Bloch states	Flat/topological bands, quasicrystals, critical states
Functional	Gap function and Wilczek-Zee connection	Unconventional ($\langle J_i \rangle = D_{ij} A_j$6-dependent) pairing, topological bands

In summary, the superfluid weight tensor synthesizes a comprehensive, symmetry-respecting, and quantum-geometric account of superconducting phase stiffness, providing a bridge between condensed-matter topology, strong correlations, and observable electromagnetic responses. Its refined understanding, particularly the interplay of geometric, topological, and conventional effects, underpins ongoing developments in materials discovery and theoretical superconductivity (Liang et al., 2016, Hiorth et al., 11 Mar 2026, Herzog-Arbeitman et al., 2021, Sun et al., 28 Jul 2025, Jiang et al., 2024, Chen et al., 28 Jan 2025, Porlles et al., 23 May 2025).