A couple of physics stories in the last few days have caught my attention for reasons that can be lumped together under the Vizzini Effect– that is, they say things that involve unconventional uses of common words. Take, for example, the Physics World story Physicists distinguish between the indistinguishable, which starts off:

Spurred on by their work on building one of the world’s most accurate atomic clocks from strontium-87 atoms, researchers in the US have now discovered that “forbidden” collisions can occur between these atoms.

Strontium-87 atoms belong to a class of objects known as fermions and according to quantum physics; fermions cannot occupy the same energy state and location in space at the same time. Fermions in identical energy states are therefore not meant to collide. Collisions perturb the internal energy levels of the atoms, and therefore such fermions should have very stable energy levels.

This is up my alley, having studied such collisions as a grad student. A violation of the prohibition against spin-polarized collisions in fermions would be big news.

The explanation, when it comes, is nothing so dramatic:

Gretchen Campbell of JILA explains that at the beginning of the measurement, all of the atoms are in the same atomic state. However, during the transition from the ground state to the excited state, using a laser pulse, light-atom interactions are not uniform across the entire atomic sample. This means that different atoms are excited at slightly different rates. These atoms are therefore no longer identical.

There’s nothing here that goes against known phyisics– in fact, what they measured is exactly what I saw in my thesis, namely, that if your state preparation isn’t perfect, you can have collisions between atoms because they’re not in exactly the same state. Each atom is in a superposition of two states, and the part of one atom that is in one state can collide with the part of another atom that is in the other state.

(It’s also slightly misleading to say that the atoms are “no longer identical.” In fact, they are still identical particles, and still subject to the requirements of Fermi statistics, that they be in an overall antisymmetric wavefunction. The effect is somewhat similar to what you would get if the particles were distinguishable, so this is a common shorthand used in the field, but there are some subtle but important differences if you want to get technical.)

The other article is from the Arxiv Blog, Physicists propose new kind of quantum tunneling, subtitled “Quantum tunnelling of a new, third kind could finally put string theory to the test.” This jumped out at me because I wasn’t aware that there were two kinds of tunneling, let alone a third type.

When I refer to “quantum tunneling,” I mean the process whereby a quantum particle– an electron, say– headed toward a barrier passes through the barrier as if it weren’t there, appearing on the other side with no loss of energy. So what’s the second type? It turns out to be a completely different process:

In recent years, physicists have explored the possibility of a second type of tunneling which happens in an entirely different way. This relies on the idea that a quantum particle can change into another quantum particle and back again with a certain probability. The tunnelling occurs when the particle changes from one that interacts strongly with a barrier and so cannot pass though it, into a particle that does not interact with the barrier and so passes through with ease.

For example, quantum particles that change briefly into particles of dark matter should pass through barriers that should otherwise be impassable. This kind of “shining-a-light-through-a-wall” experiment is a serious contender in the search for dark matter.

This is superficially similar to the normal process of tunneling, in that both involve particles showing up on the far side of a barrier, but the rest of the process is completely different. They’re different enough, in fact, that I’m not sure there’s any benefit to giving them such similar names. And the “third type” mentioned in the post seems to be a straightforward variant of the “second,” with the incident particle turning into a pair of exotic particles, rather than just one.

This strikes me as fundamentally the same sort of thing as Max Tegmark’s numbering of “multiverses,” in which he lumps together a bunch of unrelated phenomena into a large meta-category. I don’t think there’s any significant gain to lumping these things together, and there’s a significant chance of introducing confusion about what’s really going on.

I should note that I’m not complaining about the physics content, here, just the presentation. In fact, the strontium experiments described in the first story sound way cool, and my complaints about the phrasing should not be taken as questioning Jun Ye or Gretchen Campbell, both of whom are scary smart. I haven’t read their paper (it’s paywalled), but I’d be shocked if it had any physics mistakes.

This is also not intended as a look-at-how-worthless-journalists-are post (something I’ve complained about other people doing). In fact, the Physics World piece is about as good a write-up as you can get of such an experiment. My complaints are highly technical in nature, and wouldn’t occur to a lot of other physicists, let alone lay readers.

Comments

  1. #1 Alex
    April 20, 2009

    In condensed matter physics, I wonder if you could get a quasiparticle to tunnel through a barrier by changing to a different type of excitation and then changing back. Less exotic than pair production via E=mc^2, so not as cool for pop science journalists, but possibly more practical in device physics.

  2. #2 Anonymous Coward
    April 20, 2009

    > Each atom is in a superposition of two states, and the part of one atom that is in one state can collide with the part of another atom that is in the other state.

    The discussion that preceded this statement was beautiful, but the statement itself is slightly misleading: if all the atoms were prepared in identical superpositions of |g> and |e>, they wouldn’t collide with each other.

    > It’s also slightly misleading to say that the atoms are “no longer identical.”

    Why? Sure, the electrons and nuclei are still all indistinguishable. But if we’re talking about the “atoms”, then if two strontium atoms are prepared in orthogonal internal states they should be treated as distinguishable particles.

    (Assuming that we’re at sufficiently low density/energy/etc. that it’s a good approximation to consider them as individual composite particles and not worry about antisymmetrizing all the individual electrons. Which is certainly the case here.)

  3. #3 Jonathan Vos Post
    April 20, 2009

    Elsewhere in the “multiverses” Max Tegmark is grousing about Nobel laureate Chad Orzel’s metaphysical stance. Yeah, right.

  4. #4 Chad Orzel
    April 21, 2009

    The discussion that preceded this statement was beautiful, but the statement itself is slightly misleading: if all the atoms were prepared in identical superpositions of |g> and |e>, they wouldn’t collide with each other.

    What they’re dealing with in the relevant experiment is almost certainly a mixed state, not a superposition.

    Sure, the electrons and nuclei are still all indistinguishable. But if we’re talking about the “atoms”, then if two strontium atoms are prepared in orthogonal internal states they should be treated as distinguishable particles.

    The collision wavefunction still needs to be antisymmetric overall– that’s what it means to have colliding fermions, after all. Which means, loosely speaking, that you aren’t allowed to say that Atom 1 is in State A and Atom 2 is in State B, particularly after the collision.

    The effect isn’t as dramatic as the suppression in a perfectly polarized sample, but if you want to understand the process in detail, you need the colliding pair wavefunction to be antisymmetric for fermions, and symmetric for bosons. Which changes the distribution of possible states a bit, and thus the overall collision rate.

  5. #5 LtStorm
    April 21, 2009

    The first part of this post reminds me of a news article that appeared on MSNBC a year or so ago; Electron filmed for first time ever! It got a good laugh out of everyone in the chemistry department I was doing research under as an undergrad.

    The physical chemist in the department looked up the paper MSNBC mis-reported on for us, and it turned out to actually be much more interesting, but not in violation of any sort of physics.

  6. #6 Anonymous Coward
    April 21, 2009

    >The collision wavefunction still needs to be antisymmetric overall– that’s what it means to have colliding fermions, after all. Which means, loosely speaking, that you aren’t allowed to say that Atom 1 is in State A and Atom 2 is in State B, particularly after the collision.

    I stand corrected. You’re totally right.

    But just to see where my intuition was breaking down: do you think it would be accurate to say that if the collisions were purely elastic (or that elastic scattering was the dominant relevant collisional process), then it would be OK to treat two strontium atoms in orthogonal states as distinguishable particles? Or even in that case would there be consequences of identical particle symmetry?