Sunday, December 23, 2007

The Quantum Hall Effect

One of the basic rules of electricity says that in order to maintain an electrical current I in a wire, a voltage U is necessary that is proportional to the current. This is Ohm's law, U = RI, where the constant of proportionality is the resistance, R. If the wire with the current is exposed to a magnetic field, the Lorentz force acts on the wire, normal to both the direction of the current and the magnetic field. If the current flows through a sheet, as a result of the Lorentz force on the electrons in the current, a voltage can be measured that is oriented along the direction of the Lorentz force. This is the Hall voltage, named after the American physicist Edwin Herbert Hall, who was the first to observe this phenomenon. The Hall effect is used today, for example, to measure magnetic fields.

Ohm's law and the Hall effect can be understood in terms of classical physics: electrons travelling through the wire bump into the atoms of the material of the wire, which causes the resistance. However, as we know for example from the study of the electronic band structure of materials, electric current is carried by particles subject to the rules of quantum mechanics. And if the temperature of the conducting material is low enough, thus reducing the strength of thermal effects, quantum effects can become apparent, leading to such striking phenomena as superconductivity. In a superconductor, charges set in motion do not scatter, since quantum mechanics doesn't allow it, and hence, there is no resistance.

Another quite strange quantum phenomenon can be observed if the current is constrained to a very thin sheet, so that it's essentially two-dimensional, and if a strong magnetic field is applied normal to the sheet. In this case, there is a special class of states for the electrons called Landau levels, and depending of the occupancy of the Landau levels, a new macroscopic phenomenon similar to superconductivity can set in: the Quantum Hall Effect.

[Source: K. v. Klitzing, G. Dorda, M. Pepper: New Method for High-Accuracy Determination of the Fine-Structure Constant Based on Quantized Hall Resistance, Phys. Rev. Lett. 45 (1980) 494-497, Figure 1.]

The experimental data shown in the plot stem from the experiment in which the Quantum Hall effect was discovered. In the experiment, shown schematically in the inset, a thin sheet structure of semiconducting material, carrying a constant current of 1 µA, was exposed to a strong magnetic field at the temperature of liquid helium, and the voltage drop along the current and the Hall voltage normal to the current and the magnetic field have been measured. These voltages are plotted on the vertical axis, and the curves are labelled as UH for the Hall voltage, and Upp. The horizontal axis shows another voltage, the so-called gate voltage Vg, which squeezes the electrons in a thin, two-dimensional sheet, and thus determines the occupancy of the Landau levels. Now, with varying gate voltage, a curious phenomenon can be observed in the Hall and standard voltage: the standard voltage drops to zero several times, meaning that the current flows without resistance, as in a superconductor, and at the same time, the Hall voltage takes on constant values. There is series of step-like plateaux in the Hall voltage, where it stays constant.

Now, what looks like quite an esoteric effect - take some contrived semiconductor structure, cool it down to a few Kelvin, put it in the strongest magnetic field your high-field lab can provide - has an extremely interesting twist to it:

The Hall resistance, the quotient of Hall voltage and the current, in the step-like structures comes in a regular series, R = RK/n, where n is an integer, and RK, the von Klitzing constant, is RK = 25.812807557 kΩ. And the really fascinating thing is that as a consequence of the theory of Landau levels, the von Klitzing constant does not depend on the details of the material, but is universal: It is given by R = h/e² = μ0 c/2α, where h is Planck's constant, e is the charge of the electron, and α = 1/137 is the fine structure constant, the coupling constant of electrodynamics.

Thus, the quantum Hall effect can be used, for example, to set a standard for electrical resistance, or as a means to measure the fine structure constant. Klaus von Klitzing, the German physicist who discovered the effect in February 1980 and immediately grasped these implications, was awarded the Nobel Prize in Physics 1985 for the discovery of the quantized Hall effect.

Klaus von Klitzing's Nobel Lecture The Quantized Hall Effect gives a good introduction into theory and experiment of the Quantum Hall Effect (PDF file).

More details can be found in the proceedings of the Poincaré Seminar of November 13, 2004, dedicated to the Quantum Hall Effect. Among the talks:

Klaus von Klitzing: 25 Years of Quantum Hall Effect (QHE) A Personal View on the Discovery, Physics and Applications of this Quantum Effect (PDF file)

Benoît Douçot and Vincent Pasquier: Physics in a Strong Magnetic Field (PDF file), about Landau levels etc...

Beat Jeckelmann and Blaise Jeanneret: The Quantum Hall Effect as an Electrical Resistance Standard (PDF file), about the implications of the Quantum Hall Effect for metrology.

This post is part of our 2007 advent calendar A Plottl A Day.


  1. Something I've been wondering about for a long time of and on; when and how did 'normal' mean perpendicular?

  2. Sorry that should be "wondering off and on about for a long time"...

    one of cats started to play with computer screen;loved the cursor

  3. maybe a cat ate the ortho? Don't know. I guess it's because the normal vector of a plane defines that plane, so it's kind of normal to use it? Or so. If you think that doesn't make sense, I learned recently that 'can' means sometimes actually the same as 'cannot'. That's one of the things I've been wondering off and on about...



  4. Hello Michael,
    Line 30 :
    " Norma, Ein winckelmess. Gnomon. "
    Ex oriente lux :=)

  5. From the OED:

    [--classical Latin norm{amac}lis right-angled, in post-classical Latin also conforming to or governed by a rule (4th-5th cent.) -- norma NORMA n. + -{amac}lis -AL suffix1. Cf. French normal (1450-65 in Middle French in an isolated attestation in verbe normal (cf. sense A. 1), then from mid 18th cent., earliest in ligne normale (1753; cf. sense A. 5), and subsequently in more general senses ‘which serves as a model’ (1793 in école normale, 1803 in more general use), ‘ordinary, regular’ (1830s); cf. earlier anormal ANORMAL adj.), Italian normale according to the norm, routine, predictable, common, boring (1683 in sense ‘perpendicular, orthogonal’, 1831 in sense ‘customary, expected’), Portuguese normal (1844), Spanish normal (1855).


    Yes, that was hard to read, but basically, normal is from the latin for perpendicular. The other meanings flow from that one.

  6. 25.8128075578K von Klitzing constant
    25.8238916125K 60(pi)(137)

    Recipocal alpha is really 137.035999, giving less of a coincidence. One could use correct 1/alpha and replace 60 with


    thereby offering a brillant new insight into the least publishable bit.

  7. Frohe Weihnachten Euch beide!

    I had my headaches with the German word "senkrecht" which would normally translate to vertical, but the Germans also use it for naming a perfect 90deg angle.

    Am I right?



  8. Ich glaub 'senkrecht' geht zurück auf 'lotrecht' und das 'Lot' zum bestimmen vertikaler Achsen, soll heissen 'lotrecht' is 90° zu horizontal.

    Gleichfalls frohe Weihnachten und schöne Feiertage!


  9. Hello Bee, Stefan, your colleagues, and all the clever commenters here (even you, Al! ;-) )
    Please take a peek at the following URL, it should be fun (assuming it works):
    Holiday e-card for all of you!

  10. Bee, Stefan,
    Frohe Weihnachten, und danke für den tollen Adventskalender :)


  11. What gets me is the (mis)use of the word "significant."

    If researchers are just barely able to detect some causal effect or difference in a statistical sense, they say that it is "significant result" or a "significant difference." Well that is a valid statistical usage.

    And then the press reports it in mass media, and the vast majority of the public interprets the word "significant" the way "significant" is used in common language, and they are thereby mislead into believing that some large difference or large effect was measured. And so, for example, they rush out to buy blueberries or kumquats or pine bark or some such for their "significant" health effects.

    I see that over and over again, particularly in medical research

    I suspect the press may know better, but then unethically allow the confusion to occur anyway because when something can be reported as "significant" it attracts attention and sales. The profit motive often over-rides ethics and all else. A dark side of Capitalism.

    So often what is statistically "significant" in medical research is such a minute difference or effect as to be totally meaningless in a practical sense.

    ps I wonder if normal came to mean perpendicular from its being first used in relation to "normal force" .. which, with the most common force of gravity, is "normally" perpendicular to the
    surface between two objects? Just wondering ... a good question, Michael.

  12. I always assumed that normal was just short for orthonormal, where orthogonal means perpendicular and normal has the more appropriate meaning of length 1....


COMMENTS ON THIS BLOG ARE PERMANENTLY CLOSED. You can join the discussion on Patreon.

Note: Only a member of this blog may post a comment.