The Hall Conductance
Field Note: Transverse Response and the Geometry of Bands
When you apply an electric field $\mathbf{E}$ perpendicular to a magnetic field $\mathbf{B}$ in a two-dimensional electron gas, the current flows not in the direction of $\mathbf{E}$ but transverse to it. The longitudinal conductivity vanishes (the electrons are locked into cyclotron orbits), and the transverse conductivity — the Hall conductance — takes on a value that is quantized, precisely, and universally:
$$\sigma_{xy} = \nu \frac{e^2}{h}$$
where $\nu$ is an integer. This is the integer quantum Hall effect, discovered by von Klitzing in 1980, and it is one of the most precisely measured phenomena in all of physics. The quantization is known to parts in $10^9$. The Hall conductance is a topological invariant. And that topological invariant is, at its core, an integral of the Berry curvature.
The TKNN Formula
Thouless, Kohmoto, Nightingale, and den Nijs (TKNN, 1982) showed that the Hall conductance of a two-dimensional crystal in a magnetic field is given by:
$$\sigma_{xy} = \frac{e^2}{h} C_1$$
where $C_1$ is a Chern number — an integer topological invariant — computed from the Berry curvature of the occupied Bloch bands:
$$C_1 = \frac{1}{2\pi} \int_{\text{BZ}} \Omega(\mathbf{k}) , d^2k$$
Here the parameter space is the Brillouin zone, a torus. The Berry curvature $\Omega(\mathbf{k})$ is the curvature of the $U(1)$ Berry connection associated with the occupied bands. The integral of this curvature over the Brillouin zone torus yields an integer. That integer is the Hall conductance, measured in units of $e^2/h$.
This is the deepest connection between Berry phase and measurable physics: the Hall conductance, an experimentally measurable transport coefficient, equals a topological invariant computed from the geometry of the quantum wavefunctions. There are no adjustable parameters. The quantization is topological — it cannot be changed by disorder, by sample geometry, by small perturbations. The Chern number is robust because it is an integer.
Semiclassical Dynamics and the Anomalous Velocity
The same Berry curvature appears in the semiclassical equations of motion for Bloch electrons. When you include the Berry curvature in the phase-space measure, the velocity of an electron acquires an additional transverse term:
$$\dot{\mathbf{r}} = \frac{1}{\hbar} \nabla_\mathbf{k} E_n(\mathbf{k}) - \dot{\mathbf{k}} \times \boldsymbol{\Omega}(\mathbf{k})$$
The second term — the "anomalous velocity" — is proportional to the Berry curvature and perpendicular to the applied force $\dot{\mathbf{k}} = -e\mathbf{E}$. This transverse velocity is precisely what generates the Hall current. In this picture, the Hall conductance is a Berry curvature effect in momentum space, and the quantization comes from the fact that the Berry curvature integrates to an integer over the Brillouin zone.
This semiclassical picture was developed by Sundaram and Niu (1999) and by Xiao, Chang, and Niu (2010), and it makes the geometric origin of the Hall effect transparent. The electrons move transversely because the Berry curvature deflects them. Not a force, not a curvature in real space, but a curvature in the geometry of the quantum state.
The Role of Time-Reversal Symmetry
Time-reversal symmetry forces the Berry curvature to be odd: $\Omega(-\mathbf{k}) = -\Omega(\mathbf{k})$. In a system with unbroken time-reversal symmetry, the integral of the Berry curvature over the Brillouin zone vanishes, and the Hall conductance is zero. To get a quantized Hall conductance, you must break time-reversal symmetry — which the external magnetic field does.
This explains why the quantum Hall effect requires a magnetic field. It also explains why the anomalous Hall effect in ferromagnets (where time-reversal is broken by magnetization, not by an external field) is generally not quantized — the Chern number need not be an integer without the periodicity of the magnetic Brillouin zone.
From Integer to Fractional
The integer quantum Hall effect is explained by Berry curvature integrated over single-particle Bloch bands. The fractional quantum Hall effect — where $\nu$ is a fraction like $1/3$ or $2/5$ — requires interaction. The Berry phase picture still applies, but the "bands" are now bands of a many-body ground state, and the Berry curvature lives in the space of many-body parameters. The Chern number is still an integer, but it counts the topological structure of the many-body wavefunction, not of single-particle bands.
The Berry connection provides the language that unifies these pictures. Whether you are talking about single-particle Bloch bands or interacting many-body states, the Hall conductance is always an integral of Berry curvature, and the quantization is always topological.
Summary
The Hall conductance is the experimental fingerprint of Berry curvature in momentum space. It is quantized because the curvature integrates to a Chern number. It is a transverse response because the curvature deflects electrons sideways. It is a topological invariant because the Chern number is an integer. It is, in every sense, a geometric phenomenon.
The Berry phase, discovered in the abstract language of adiabatic evolution, turned out to be the key to understanding one of the most precisely measured quantities in physics. That is the kind of story that makes geometric phases feel not like a mathematical curiosity but like a fundamental principle of nature.