Home / Courses / GATE Physics / Practice / The quantum Hall effect and low-dimensional systems

The quantum Hall effect and low-dimensional systems

Solid state physics and devices · 73 questions and templates for GATE Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    On the plateau of the integer quantum Hall effect with filling factor ν, the Hall resistance is
    1. νh/e2
    2. h/2νe2
    3. h/νe2
    4. νe2/h
    Answer and solution

    Answer: C

    With ν Landau levels completely filled, σxy=νe2/h and σxx=0, so ρxy=1/σxy=h/νe2=25812.807 Ω/ν.

    Why people get this wrong: νe2/h is the Hall conductance, not resistance.

  2. MCQWarm-up

    While the Hall resistance of a 2D electron gas sits on a quantum Hall plateau, its longitudinal resistance ρxx is
    1. essentially zero
    2. equal to the Hall resistance
    3. very large and rising with B
    4. equal to h/e² exactly
    Answer and solution

    Answer: A

    On a plateau the Fermi level lies in localised states between Landau levels, so there are no extended states at the Fermi energy to scatter into: no dissipation, σxx=0. With σxy≠0 the tensor inversion makes ρxx=0 too. Measured values go below 10−10 Ω.

    Why people get this wrong: Zero resistance here is not superconductivity: σxx is also zero.

  3. NATWarm-up

    The von Klitzing constant RK=h/e2 is ______ Ω (nearest whole number). Use h=6.62607×10−34 J s and e=1.602177×10−19 C.
    Answer and solution

    Answer: 25813 Ω (±1)

    RK=6.62607×10−34/(1.602177×10−19)2=6.62607×10−34/2.56697×10−38=25812.8 Ω. Since 2019 h and e are exact, so RK is exact too, and quantum Hall devices realise the ohm.

    Why people get this wrong: h/2e2=12906 Ω is the resistance of one spin-degenerate channel (or the ν=2 plateau).

  4. MCQWarm-up

    A 2D electron of effective mass m∗ in a perpendicular field B has energies (ignoring spin)
    1. ℏωcn2, with ωc=eB/m∗
    2. ℏωc(n+12), with ωc=eB/2m∗
    3. continuous, from 0 upwards
    4. ℏωc(n+12), with ωc=eB/m∗
    Answer and solution

    Answer: D

    In the Landau gauge A→=(0,Bx,0), H=[px2+(py+eBx)2]/2m∗; with py=ℏk it is a harmonic oscillator in x centred at −ℏk/eB with frequency ωc=eB/m∗. So En=ℏωc(n+12), each level hugely degenerate (one state per k).

    Why people get this wrong: eB/2m is the Larmor frequency, half the cyclotron frequency.

  5. MCQWarm-up

    The density of states per unit area of a 2D free-electron gas (effective mass m∗, spin included), for energies above the band bottom, is
    1. proportional to E
    2. constant, m∗/πℏ2
    3. proportional to 1/E
    4. proportional to E
    Answer and solution

    Answer: B

    States fill a disc in k space: N/A=2×πk2/(2π)2=k2/2π, with E=ℏ2k2/2m∗: N/A=m∗E/πℏ2. Its derivative m∗/πℏ2 is constant, so each 2D subband adds a step.

    Why people get this wrong: E is 3D; 1/E is 1D.

  6. MCQWarm-up

    A ballistic narrow constriction carries a single spin-degenerate transverse mode with perfect transmission. Its conductance is
    1. e2/h
    2. zero, since there is no scattering
    3. 2e2/h
    4. infinite, since there is no scattering
    Answer and solution

    Answer: C

    Landauer: G=(2e2/h)∑nTn. One mode with T=1 gives 2e2/h=77.5 μS, a resistance of 12.9 kΩ even with no scattering: it is a contact resistance from fitting many reservoir modes into one channel.

    Why people get this wrong: A perfect wire still has finite resistance h/2e² per mode.

  7. MCQWarm-up

    The conductance quantum e2/h is about
    1. 25.8 kΩ
    2. 38.7 μS
    3. 77.5 μS
    4. 12.9 kΩ
    Answer and solution

    Answer: B

    e2/h=(1.602×10−19)2/6.626×10−34=3.87×10−5 S =38.7 μS. Its inverse, h/e2=25.8 kΩ, is a resistance. With spin degeneracy, 2e2/h=77.5 μS is the conductance of one open channel in a quantum point contact.

    Why people get this wrong: 25.8 kΩ is h/e2, a resistance, not a conductance.

  8. NATStandard

    Each Landau level of a 2D electron gas holds eB/h states per unit area for each spin direction. At B=10 T this is ______ ×1015 m−2 (two decimal places).
    Answer and solution

    Answer: 2.42 × 10¹⁵ m⁻² (±0.01)

    eB/h=1.602×10−19×10/6.626×10−34=2.42×1015 m−2: one state per flux quantum h/e threading the sample. This is the same as the number of B=0 states in an energy range ℏωc per spin, (m∗/2πℏ2)ℏωc.

    Why people get this wrong: The degeneracy does not depend on the effective mass.

  9. NATStandard

    A 2D electron gas has density 3.0×1015 m−2. In a perpendicular field of 5.0 T its filling factor ν=nsh/eB is ______ (two decimal places).
    Answer and solution

    Answer: 2.48 (±0.01)

    eB/h=1.209×1015 m−2 at 5 T. ν=3.0×1015/1.209×1015=2.48: two spin-resolved Landau levels full and the third half full.

    Why people get this wrong: Using ħ instead of h changes ν by 2π.

  10. NATStandard

    The magnetic length ℓB=ℏ/eB at B=9.0 T is ______ nm (two decimal places).
    Answer and solution

    Answer: 8.55 nm (±0.01)

    ℓB=1.0546×10−34/(1.602×10−19×9)=7.31×10−17=8.55×10−9 m, i.e. 25.7 nm/B[T]. A disc of area 2πℓB2 holds one flux quantum.

    Why people get this wrong: ℓB is independent of the effective mass; it depends only on B.

  11. NATStandard

    Electrons in GaAs have effective mass 0.067me. At B=10 T their cyclotron energy ℏωc is ______ meV (two decimal places).
    Answer and solution

    Answer: 17.28 meV (±0.05)

    ℏωc=ℏeB/m∗=1.0546×10−34×1.602×10−19×10/(0.067×9.109×10−31)=2.77×10−21 J =17.3 meV, i.e. 1.73 meV/T: about 200 K at 10 T, much larger than at liquid-helium temperatures, which is why GaAs is ideal for the QHE.

    Why people get this wrong: With the free-electron mass you get only 1.16 meV.

  12. MCQStandard

    Shubnikov-de Haas oscillations of the longitudinal resistance of a 2D electron gas are periodic in
    1. B
    2. 1/B
    3. B2
    4. B
    Answer and solution

    Answer: B

    A minimum occurs each time the Fermi level lies between Landau levels, i.e. at integer ν=nsh/eB. Consecutive integers ν and ν+1 are equally spaced in 1/B: Δ(1/B)=e/nsh per spin-resolved level (2e/nsh for spin-degenerate levels).

    Why people get this wrong: The minima crowd together at low field when plotted against B; they are uniform only in 1/B.

  13. NATStandard

    A GaAs 2D electron gas (m∗=0.067me, spin-degenerate, one subband) has density 3.0×1015 m−2. Its Fermi energy measured from the subband bottom is ______ meV (two decimal places).
    Answer and solution

    Answer: 10.72 meV (±0.05)

    ns=(m∗/πℏ2)EF, so EF=πℏ2ns/m∗=π×(1.0546×10−34)2×3.0×1015/(6.10×10−32)=1.72×10−21 J =10.7 meV.

    Why people get this wrong: Forgetting spin gives 2πℏ2ns/m∗, twice as large.

  14. MCQStandard

    On the ν=2 plateau, the Hall resistance of a 2D electron gas is about
    1. 6453 Ω
    2. 25813 Ω
    3. 51626 Ω
    4. 12906 Ω
    Answer and solution

    Answer: D

    ρxy=h/νe2=25812.8/2=12906.4 Ω. It is the same in Si MOSFETs, GaAs heterostructures and graphene, to parts in 1010.

    Why people get this wrong: 6453 Ω is ν=4.

  15. MSQStandard

    Which statements about the integer quantum Hall effect are correct?
    1. The plateau values do not depend on the sample shape or the host material.
    2. It requires electron-electron interactions to appear.
    3. The longitudinal resistance vanishes on a plateau.
    4. Some disorder (localised states) is needed for the plateaus to have finite width.
    Answer and solution

    Answer: A, C, D

    The IQHE is a one-electron effect: Landau quantisation plus disorder. Localised states between the levels pin the Fermi energy while B changes, so σxy stays at νe2/h over a range of field and ρxx=0 there. The quantisation is topological, independent of shape, material and impurities. Interactions are needed for the fractional effect, not the integer one.

    Why people get this wrong: In a perfectly clean sample the plateaus would shrink to points.

  16. MCQStandard

    In an infinitely deep quantum well of width L, the confinement energies of the subbands scale as
    1. n/L
    2. n2/L
    3. n2/L2
    4. n/L2
    Answer and solution

    Answer: C

    En=π2ℏ2n22m∗L2: standing waves with k=nπ/L. Halving the width quadruples every level, which is how the colour of quantum-well lasers and dots is tuned.

    Why people get this wrong: Energy goes as k², and k as n/L.

  17. NATStandard

    An electron (m∗=0.067me) is confined in an infinitely deep GaAs well of width 10 nm. Its ground-state confinement energy is ______ meV (one decimal place).
    Answer and solution

    Answer: 56.1 meV (±0.2)

    E1=π2ℏ2/2m∗L2=9.870×1.1122×10−68/(2×6.103×10−32×10−16)=8.99×10−21 J =56.1 meV. With the free-electron mass it would be only 3.76 meV.

    Why people get this wrong: Using the free-electron mass underestimates the confinement energy fifteen-fold.

  18. NATStandard

    Electrons in GaAs (m∗=0.067me, me=9.109×10−31 kg) are in a field of 5.0 T. The free-space wavelength of radiation that drives their cyclotron resonance is ______ μm (to one decimal place). Use e=1.602×10−19 C and c=3.00×108 m/s.
    Answer and solution

    Answer: 143.6 μm (±0.5)

    fc=eB2πm∗=1.602×10−19×5.02π×0.067×9.109×10−31=2.09×1012 Hz. λ=c/fc=3.00×108/2.09×1012=1.436×10−4 m =143.6 μm, in the far infrared. Measuring it gives m∗ directly.

    Why people get this wrong: eB/m∗ is the angular frequency; divide by 2π before finding the wavelength.

  19. MCQStandard

    Since the 2019 revision of the SI, the von Klitzing constant RK=h/e2 is
    1. measured more precisely each year with better Hall bars
    2. exact, because h and e themselves now have fixed values
    3. defined to be exactly 25000 Ω
    4. no longer used, since resistance standards are made from wire
    Answer and solution

    Answer: B

    The 2019 SI fixes h=6.62607015×10−34 J s and e=1.602176634×10−19 C exactly, so h/e2=25812.807… Ω is exact. A quantum Hall device on a plateau now realises the ohm directly; its value is the same in every sample (GaAs, graphene) to parts in 1010.

    Why people get this wrong: The Hall plateaus are not tuned to round numbers; their value follows from the fixed constants.

  20. MCQExam level

    In a 2D conductor in a magnetic field, σxx=σyy and σxy=−σyx. On a quantum Hall plateau, σxx=0 while σxy≠0. Then the longitudinal resistivity ρxx is
    1. zero, because ρxx=σxx/(σxx2+σxy2)
    2. infinite, because ρxx=1/σxx
    3. equal to 1/σxy
    4. indeterminate (zero divided by zero)
    Answer and solution

    Answer: A

    Inverting the 2×2 tensor: ρxx=σxxσxx2+σxy2 and ρxy=−σyxσxx2+σxy2, which gives 1/σxy in magnitude when σxx=0. So a perfect insulator along x (σxx=0) is at the same time a perfect conductor (ρxx=0): the current flows at right angles to the field with no dissipation.

    Why people get this wrong: ρxx=1/σxx holds only when B=0.

  21. NATExam level

    Shubnikov-de Haas minima of a spin-degenerate 2D electron gas are spaced by Δ(1/B)=0.050 T−1. The electron density is ______ ×1015 m−2 (two decimal places).
    Answer and solution

    Answer: 9.67 × 10¹⁵ m⁻² (±0.03)

    Minima occur when an integer number N of spin-degenerate levels is full: ns=N⋅2eBN/h, so 1/BN=2eN/hns and Δ(1/B)=2e/hns. Then ns=2e/(hΔ)=2×1.602×10−19/(6.626×10−34×0.050)=9.67×1015 m−2.

    Why people get this wrong: If the spin splitting is resolved, the period halves; here the levels are spin-degenerate, hence the factor 2.

  22. MCQExam level

    The elementary excitations (quasiparticles) of the Laughlin state at ν=1/3 carry an electric charge of
    1. e
    2. e/3
    3. 3e
    4. 2e
    Answer and solution

    Answer: B

    Laughlin's thought experiment: inserting one flux quantum h/e adiabatically at a point pushes out a charge νe=e/3, leaving a quasihole of charge e/3. Shot-noise experiments confirmed fractional charge e/3.

    Why people get this wrong: Charge 2e belongs to Cooper pairs, not to quantum Hall quasiparticles.

  23. NATExam level

    A 2D electron gas has density 2.4×1015 m−2. The magnetic field at which the filling factor is exactly 1 is ______ T (two decimal places).
    Answer and solution

    Answer: 9.93 T (±0.02)

    ν=nsh/eB=1 gives B=nsh/e=2.4×1015×4.136×10−15 T m2=9.93 T. (h/e=4.136×10−15 Wb is the flux quantum for single electrons.)

    Why people get this wrong: h/2e is the superconducting flux quantum; here each electron needs h/e.

  24. MCQExam level

    The fractional quantum Hall effect (plateaus at ν = 1/3, 2/5, 2/3, ...) arises from
    1. disorder alone, as in the integer effect
    2. spin-orbit coupling in the host crystal
    3. Coulomb interactions among electrons in a partly filled Landau level
    4. the periodic potential of the crystal lattice
    Answer and solution

    Answer: C

    In a partly filled lowest Landau level all single-particle states have the same kinetic energy, so the many-body ground state is decided by the Coulomb repulsion alone. At special fractions it forms an incompressible liquid (Laughlin's ∏(zi−zj)3e−∑|z|2/4ℓB2 for 1/3) with a gap, which gives plateaus.

    Why people get this wrong: Disorder is needed to give plateaus width, but the gap itself comes from interactions.

  25. MCQExam level

    In monolayer graphene (Dirac electrons with 4-fold spin and valley degeneracy), the quantum Hall plateaus appear at σxy equal to
    1. ±2,±6,±10,… in units of e2/h
    2. ±4,±8,±12,… in units of e2/h
    3. 0,±1,±2,… in units of e2/h
    4. ±1,±3,±5,… in units of e2/h
    Answer and solution

    Answer: A

    Dirac Landau levels are En=±vF2eℏB|n|, including a level exactly at E=0 shared equally by electrons and holes. Each level holds 4 states per flux quantum, and the zero level is half-filled at neutrality, so σxy=±4(N+12)e2/h=±2,±6,±10,…: the "half-integer" quantum Hall effect, a signature of the Berry phase π.

    Why people get this wrong: ±4,±8,… would be an ordinary 4-fold degenerate 2D gas without the zero-energy level.

  26. NATExam level

    Graphene's Landau levels are En=±vF2eℏB|n| with vF=1.0×106 m/s. At B=10 T, the n=1 level lies ______ meV above the Dirac point (nearest whole number).
    Answer and solution

    Answer: 115 meV (±1)

    2eℏB=2×1.602×10−19×1.0546×10−34×10=1.838×10−26 kg m/s. Times vF: 1.838×10−20 J =115 meV, far more than in GaAs at the same field, which is why graphene shows the QHE even at room temperature.

    Why people get this wrong: Dirac levels go as B and n, not linearly like ℏωc(n+12).

  27. MCQExam level

    In the quantum Hall regime, the current is carried by edge states. Which property explains the absence of backscattering?
    1. Edge states have infinite effective mass.
    2. On a given edge all edge states move in the same direction, so an electron has no backward state to scatter into nearby.
    3. Edge states exist only at zero temperature.
    4. Edge states carry no charge.
    Answer and solution

    Answer: B

    Near the edge the confining potential bends the Landau levels up, and where they cross the Fermi energy they form channels with group velocity E→×B→/B2 along the edge. The direction is fixed by B→: opposite edges carry opposite directions. To reverse, an electron would have to cross the whole sample, which is exponentially suppressed. Each channel gives e2/h.

    Why people get this wrong: Chirality, not the absence of impurities, is what protects the edge channels.

  28. MSQExam level

    Which statements about the density of states g(E) of free electrons in low-dimensional systems are correct?
    1. In 3D, g(E)∝E.
    2. In a quantum wire (1D), each subband gives g∝(E−En)−1/2.
    3. In a quantum dot (0D), the spectrum is a continuous band.
    4. In a quantum well, g(E) is a staircase with a step at each subband edge.
    Answer and solution

    Answer: A, B, D

    In d dimensions N(E)∝kd∝Ed/2, so g∝Ed/2−1: E (3D), constant (2D), E−1/2 (1D). Confinement in the other directions makes each subband add its own g from its edge En: steps in a well, inverse-square-root spikes in a wire. In a dot every direction is confined, so the spectrum is discrete: delta functions, like an atom.

    Why people get this wrong: A quantum dot is an artificial atom: discrete levels, not a band.

  29. NATExam level

    An infinitely deep GaAs well (m∗=0.067me) is 20 nm wide. Its electrons fill states up to 60 meV above the bottom of the well. How many subbands are occupied?
    Answer and solution

    Answer: 2

    En=n2E1 with E1=π2ℏ2/2m∗L2=14.0 meV for L=20 nm (a quarter of the 10 nm value, 56.1 meV). E2=56.1 meV <60 meV and E3=126 meV >60 meV, so 2 subbands hold electrons.

    Why people get this wrong: The subband energies go as n², so the third subband is far above the second.

  30. NATExam level

    A quantum point contact transmits 3 spin-degenerate modes perfectly. Its resistance is ______ kΩ (two decimal places).
    Answer and solution

    Answer: 4.3 kΩ (±0.01)

    G=3×2e2/h=6e2/h, so R=h/6e2=25812.8/6=4302 Ω=4.30 kΩ. Narrowing the gate removes modes one at a time, and G falls in steps of 2e2/h.

    Why people get this wrong: Each spin-degenerate mode gives 2e²/h, not e²/h.

← All topics · Timed mock test →