Quantum Hall Effect

Definition

The quantum Hall effect is the quantization of the Hall conductance of a two-dimensional electron gas in a strong magnetic field:

σ_xy = ν · e²/h        ν = 1, 2, 3, ...

Equivalently, the transverse (Hall) resistance is locked to

R_xy = R_K / ν     where     R_K = h/e² = 25812.807 Ω

The remarkable part is not that resistance is quantized — it is how perfectly. As the magnetic field is swept, the Hall resistance does not rise as a smooth ramp. It climbs in a staircase of flat plateaus, and on each plateau the value is reproducible to better than one part in 10⁹ — independent of the sample's shape, its material, or how dirty it is. At the same time, the ordinary (longitudinal) resistance collapses to essentially zero. Current flows, but no energy is lost.

How it works

Three ingredients combine to produce the effect.

1. Confine electrons to a plane. In a semiconductor heterostructure such as GaAs/AlGaAs, electrons are trapped in a sheet only a few nanometers thick. They move freely in two dimensions but cannot move in the third — a 2D electron gas.

2. Turn on a strong perpendicular field. A magnetic field B bends every electron's path into a tiny circle at the cyclotron frequency ω_c = eB/m*. Quantum mechanically, those circular orbits are quantized into discrete energy levels — Landau levels:

E_n = ℏ·ω_c·(n + ½)        ω_c = eB/m*

Each Landau level is enormously degenerate. The number of states it can hold per unit area is

n_B = eB/h        (states per m² per Landau level)

so the capacity of each level grows linearly with B. The continuous spread of allowed energies that an ordinary metal has is replaced by a comb of sharp, hugely degenerate spikes separated by gaps of size ℏω_c.

3. Fill an integer number of levels. Define the filling factor ν = n / n_B = n·h/(eB), where n is the electron density. When ν is exactly an integer, an integer number of Landau levels is completely full and the rest are empty. The Fermi energy then sits in the gap between levels: the bulk has no available states to scatter into, so it becomes an insulator. As B is swept, ν passes through integer values, and at each one the system parks on a plateau.

The edge does the conducting. Even when the bulk is insulating, the boundary of the sample is not. At the edge the confining potential bends the Landau levels upward, forcing them to cross the Fermi energy. Those crossings create one-dimensional chiral edge channels that wrap around the sample and carry current in a single direction. There is no counter-propagating channel on the same edge, so an electron cannot backscatter — it has nowhere to scatter to. The result is dissipationless transport: R_xx → 0 while R_xy is pinned at R_K/ν.

Why it is exact. The deep reason the plateau value does not drift is topological. Thouless, Kohmoto, Nightingale, and den Nijs showed in 1982 that the Hall conductance of a filled band equals e²/h times an integer — the TKNN invariant, or Chern number — computed from the global geometry of the electron wavefunctions across the Brillouin zone. A topological integer cannot change without closing the energy gap. Smoothly deforming the sample, adding disorder, or changing the material leaves the Chern number untouched. That is why labs on different continents, using different crystals, measure the same plateau to nine decimal places.

Worked example — reading a plateau

Take the ν = 2 plateau, the workhorse of resistance metrology. The Hall resistance is

R_xy = R_K / ν = 25812.807 Ω / 2 = 12906.4035 Ω

Now ask: at what field does ν = 2 occur for a typical GaAs sample with electron density n = 3 × 10¹⁵ m⁻²? Using ν = n·h/(eB):

B = n·h / (ν·e)
  = (3×10¹⁵ × 6.626×10⁻³⁴) / (2 × 1.602×10⁻¹⁹)
  ≈ 6.2 T

So at about 6 tesla, two Landau levels are full and the device sits on the 12906 Ω plateau. Push the field to about 12 T and you reach ν = 1 at R_K = 25812.807 Ω. Drop it to about 3 T and you reach ν = 4 at 6453 Ω. The plateaus are wide because disorder localizes the states between levels — those localized states absorb extra electrons without changing the Hall conductance, holding the plateau steady over a finite range of B.

The temperature condition is just as concrete. The cyclotron gap at 6 T in GaAs (m* ≈ 0.067 m_e) is ℏω_c ≈ ℏeB/m* ≈ 1.7 × 10⁻²¹ J ≈ 11 meV, which corresponds to about 125 K. To resolve the plateau cleanly you want k_B·T well below that gap, which is why metrology labs operate near 1.5 K — there ℏω_c/(k_B·T) ≈ 80, and thermal smearing is negligible.

Variants and regimes

Regime	Filling ν	Origin	Distinguishing feature
Integer QHE	1, 2, 3, …	Single-particle Landau filling + disorder	Plateaus at R_K/ν; the metrology standard
Fractional QHE	1/3, 2/5, 2/3, …	Electron–electron interactions (Laughlin liquid)	Fractionally charged anyonic quasiparticles
ν = 5/2 state	5/2	Paired composite fermions	Candidate non-Abelian anyons for topological qubits
Graphene (monolayer)	±2, ±6, ±10, …	Relativistic Dirac Landau levels + degeneracy	Half-integer sequence; survives to room temperature
Quantum spin Hall	—	Spin–orbit coupling, no external B	Counter-propagating spin-polarized edge channels
Quantum anomalous Hall	1	Intrinsic magnetization in a topological insulator	Quantized Hall conductance at B = 0

The integer and fractional effects share the same staircase signature but differ profoundly in mechanism. The integer effect is essentially single-particle physics dressed by disorder; the fractional effect is a genuinely correlated quantum liquid in which the electrons reorganize into composite particles. The quantum anomalous and quantum spin Hall effects extend the idea to zero external field, the foundation of modern topological-materials research.

The resistance standard

Because R_xy = R_K/ν depends only on h and e, the quantum Hall effect gives a reproducible unit of resistance that needs no physical artifact. Before 1990, the ohm was maintained with banks of wire-wound resistors that drifted over time. From January 1990, national metrology institutes adopted a conventional value R_K-90 = 25812.807 Ω as the reference. The 2019 SI redefinition went further: it fixed Planck's constant h and the elementary charge e to exact numerical values, which makes R_K = h/e² an exact constant (25812.80745…Ω). The ohm is now realized directly by measuring a quantum Hall plateau — typically ν = 2 at about 12906 Ω — and the same redefinition lets the volt (via the Josephson effect) and the ampere be tied to the same fundamental constants. A single chip in a dilution refrigerator can anchor the unit of resistance for an entire country.

Common pitfalls and misconceptions

• "The plateau value depends on the material." It does not. R_K/ν is the same in GaAs, silicon MOSFETs, and graphene to within the experimental uncertainty. Universality is the whole point — and the consequence of topology.
• "Quantization comes from clean, perfect samples." Counterintuitively, disorder is essential for the integer effect. Without localized states between Landau levels, the plateaus would be infinitely narrow and unobservable. Disorder broadens them into the wide flats you measure.
• "It's just the classical Hall effect at low temperature." The classical Hall effect gives R_xy = B/(n·e), a straight line in B with no plateaus. The quantum version replaces that line with a staircase. They agree only on average, where the plateau centers track the classical slope.
• "R_xx = 0 means the sample is a superconductor." No. The bulk is an insulator; current is carried by edge channels that cannot backscatter. There are no Cooper pairs and no Meissner effect — the zero resistance has a topological, not a pairing, origin.
• "Integer and fractional effects are the same physics with different numbers." The integer effect is single-particle; the fractional effect is a strongly correlated many-body state. Conflating them hides the most interesting physics — fractional charge and anyonic statistics.
• "You need a perfect 2D sheet of exactly zero thickness." Real 2D electron gases are a few nanometers thick. What matters is that motion perpendicular to the plane is frozen into its quantum ground state, so the system behaves two-dimensionally for the relevant energies.

Applications

• The SI ohm. Every calibrated resistor in the world traces back, through a chain of comparisons, to a quantum Hall plateau. This is the flagship application.
• Fundamental-constant metrology. Combining the quantum Hall effect (R_K = h/e²) with the Josephson effect (K_J = 2e/h) lets metrologists measure h and e independently — a key input to the redefined SI.
• Materials characterization. Quantum Hall measurements reveal carrier density, mobility, and effective mass in new 2D materials, and confirm whether a material is topologically nontrivial.
• Topological quantum computing. The ν = 5/2 fractional state may host non-Abelian anyons, whose braiding could store and process quantum information immune to local noise.
• Graphene devices. Room-temperature quantum Hall plateaus in graphene make portable, less cryogenically demanding resistance standards conceivable.
• Probing fundamental symmetry. Precision comparisons of R_K across labs test whether the fine-structure constant α = e²/(2·ε₀·h·c) is truly universal, since R_K and α are directly related.

Precision analysis — where the nine digits come from

It is worth being explicit about why the quantum Hall effect is so much more precise than a typical physics measurement. The conductance of a single chiral edge channel is, by the Landauer formula, exactly e²/h — each one-dimensional mode carries one quantum of conductance. With ν fully transmitting channels and no backscattering, the total Hall conductance is exactly ν·e²/h. Two facts protect that value:

• No backscattering. Edge channels are chiral; an electron at one edge cannot turn around without crossing the insulating bulk to the opposite edge, which is exponentially suppressed in a macroscopic sample. So the transmission probability is unity to extraordinary accuracy.
• Topological protection. The integer ν is a Chern number. It is mathematically forbidden from changing continuously, so smooth perturbations — disorder, geometry, weak interactions — cannot shift it. Errors enter only through processes that close the Landau gap, and those are thermally activated, scaling like exp(−Δ/2k_B·T).

This is why the dominant correction is set by the ratio of the Landau gap to temperature. With ℏω_c/(k_B·T) ≈ 80 at metrology conditions, the activated correction exp(−Δ/2k_B·T) is fantastically small, leaving residual deviations at the 10⁻⁹ level — limited in practice by contact resistance and instrumentation, not by physics. No fitting, no calibration to a more fundamental standard: the number falls out of h and e directly. That is what makes the quantum Hall effect unique among condensed-matter phenomena — it is a place where an abstract topological invariant becomes a voltmeter reading good to nine figures.