Berry curvature: phases, Hall response and topology

Scope

Berry curvature describes local geometric twisting of a quantum-state family in parameter space. In a crystal, momentum-space curvature can contribute to electronic transport, but curvature, anomalous Hall conductivity and superconducting chirality are different objects. Specify whether the states belong to a normal-state Hamiltonian, a Bogoliubov–de Gennes Hamiltonian or a selected degenerate subspace before drawing a topological conclusion. Berry's original construction separates geometric and dynamical phases during adiabatic evolution. R1 R3

The inherited chapter includes the Xiao–Chang–Niu and Nagaosa reviews as broader reading. Here the defining phase equations are checked against Berry's original paper, semiclassical motion against Sundaram–Niu, and the metallic Hall interpretation against Haldane. Review citations are not substitutes for a completed audit of every underlying experiment. R1 R2 R3 R4 R6

Mathematical assumptions

For an isolated normalized Bloch state, use $A_i=i\langle u|\partial_i u\rangle$ and $\Omega_{ij}=\partial_iA_j-\partial_jA_i=-2\mathrm{Im}\,Q_{ij}$. Under $u\rightarrow e^{i\phi}u$, the connection changes as $A_i\rightarrow A_i-\partial_i\phi$, while curvature is unchanged. The closed-loop phase $\gamma=\oint A_i dk_i$ is defined modulo $2\pi$. In a smooth patch, Stokes' theorem relates it to the curvature flux. Nontrivial bundles can require multiple patches and transition functions; a globally smooth eigenvector should not be assumed. R1 R3

For a nondegenerate eigenstate of $H$, differentiating the eigenvalue equation gives $\langle m|\partial_i n\rangle=\langle m|\partial_iH|n\rangle/(E_n-E_m)$ for $m\ne n$. Consequently $\Omega_{n,ij}=-2\mathrm{Im}\sum_{m\ne n}\langle n|\partial_iH|m\rangle\langle m|\partial_jH|n\rangle/(E_n-E_m)^2$. Small denominators require resolving the neighboring bands and the validity of the isolated-band approximation. Exactly degenerate states require a subspace treatment, not arbitrary assignments to individual eigenvectors. Wilczek–Zee supplies the historical non-Abelian extension; its full derivation is not audited here. R1 R8

With electron charge $-e$, $e>0$, and zero magnetic field, the weak-field semiclassical equations are $\hbar\dot{\mathbf k}=-e\mathbf E$ and $\dot{\mathbf r}=\hbar^{-1}\nabla_{\mathbf k}\epsilon-\dot{\mathbf k}\times\boldsymbol\Omega$. For the clean intrinsic response, $\sigma_{xy}^{\rm int}=-(e^2/\hbar)\sum_n\int d^dk\,f_n\Omega_{n,xy}/(2\pi)^d$, with $j_x=\sigma_{xy}E_y$. The two-dimensional quantity is a sheet conductance; in three dimensions it has conductivity units. Haldane's checked manuscript uses the opposite connection sign, so its curvature and displayed Hall sign must be translated together. R3 R6

For a gapped two-dimensional occupied subspace, $C=(2\pi)^{-1}\int_{\rm BZ}\mathrm{Tr}\,\Omega_{xy}\,d^2k$ is an integer under the stated orientation. The corresponding clean filled-band result in this convention is $\sigma_{xy}=-Ce^2/h$. A metal can still contain an isolated band with an integer band Chern number; its partially occupied curvature integral is generally not that integer. Haldane expresses the nonquantized metallic contribution through Fermi-surface Berry phases, retaining a quantized ambiguity and Brillouin-zone boundary terms. It is not a license to discard all filled-band contributions. R3

As a bounded check, the spinor $u=(\cos(\theta/2),e^{i\varphi}\sin(\theta/2))$ taken around increasing $\varphi$ at fixed $\theta$ gives $\gamma=-\pi(1-\cos\theta)$ modulo $2\pi$. We compare cyclic overlap products with an independently integrated curvature flux, reverse the loop orientation and apply phase changes at every point. The energy-denominator formula and Hall-current sign are checked separately. These calculations validate conventions in a model, not a material's measured topology. R1 R6

Experimental criteria

An anomalous Hall measurement is not a unique image of a curvature hotspot. Onoda and colleagues' original impurity-transport study distinguishes skew-scattering and intrinsic regimes within its dilute-impurity approximation; that is not a universal material-independent scaling law. Separate magnetic state, occupation, disorder and the conductivity-to-resistivity conversion before fitting a mechanism. A finite normal-state Hall signal does not determine the topology of a superconducting gap. R3 R7

Mielke and colleagues studied KV₃Sb₅ using muon spin relaxation. The checked author manuscript finds enhanced relaxation around the approximately 80 K charge-order scale, while the selected zero-field relaxation stays approximately constant through the separate superconducting transition. It explicitly notes that zero-field results alone cannot establish the time-reversal-breaking interpretation because charge order can modify nuclear electric-field gradients. Longitudinal-field and higher-field controls form part of the argument. This is evidence about the charge-ordered material, not a direct reconstruction of Berry curvature or a demonstration of chiral superconductivity. The 80 K record is excluded from superconducting-temperature search. R5

Limits and disputes

Time reversal relates the curvature of symmetry-partner states with opposite sign at opposite momenta; inversion relates the corresponding curvature with the same sign. For spinful systems with both symmetries, treat the Kramers subspace appropriately. Cancellation of charge Hall response does not imply that all spin, valley or non-Abelian structure is trivial. Likewise, broken time reversal alone does not guarantee a nonzero measured Hall component: other symmetries or cancellations can forbid it. R3 R6 R8

This review visually checks Berry's original pages 46–47 and Haldane author v5 page 1, and reads selected Haldane, Sundaram–Niu and Mielke passages. The Haldane author copy prints an apparent repeated-momentum typo in its time-reversal equation; the transformation stated here follows the consistent symmetry argument rather than that literal expression. Full journal comparisons, non-Abelian derivations, disorder calculations and raw-data review remain open. Complete scientific acceptance has not been reached. R1 R3 R5 R6 R7 R8

Discovery implications

Use curvature maps together with band gaps, occupation and symmetry, and test response predictions against independent measurements. For superconducting topology, calculate the appropriate gapped BdG subspace and compare relevant boundary or thermal probes. Keep a charge-order signal, normal-state anomalous Hall effect and a new symmetry change below the superconducting transition as separate claims. A promising normal-state band structure is an input to this analysis, not its conclusion. R3 R5

Related topics

Compare quantum metric, chiral pairing, topological pairing and geometric superfluid weight.

Material cases

Values describe individual reports, not a ranking. Unknown means the checked source did not supply the field at this review scope.