Measuring Higher-Level QED

A new experiment at Livermore National Lab has made the best measurement yet of a complicated correction to the simplest quantum description of how atoms behave. Livermore researchers did this by measuring the Lamb shift, a subtle shifting of quantum energy levels, including a first measurement of "two-loop" contributions, in a plasma of highly charged uranium ions.

Quantum electrodynamics (QED), the modern theory of the electromagnetic force (the development of which earned Richard Feynman, Sin-Itiro Tomonaga, and Julian Schwinger a Nobel Prize in 1965), was an improvement over early quantum mechanics since it took into account that electrons inside atoms don't merely interact with the nucleus; they also interact with the vacuum. To be more precise, in determining various quantum energy levels available to the electron in a simple hydrogen atom, QED accounts for occurrences in which an electron will spontaneously emit, and shortly thereafter reabsorb, a photon.

If one portrayed such an event in graphlike form, on a Feynman diagram, the photon would be depicted as a squiggly line leaving and rejoining the line depicting the electron moving through time. Conversely, a photon can spontaneously rematerialize as an electron-positron pair of virtual particles, providing that these particles very quickly recombine into a photon.

These are examples of "single-loop" processes, since the Feynman diagram features a loop where the virtual particle or particles pass into and back out of existence. One can imagine additional, higher-order processes, depicted by loops within loops, which play a lesser but still considerable role in the overall sense of what an electron or photon "is." Such hidden processes (hidden in the sense that they cannot be directly observable in the lab) can be probed, however, by looking at the alteration, or Lamb shift, in the spectrum of light emitted by atoms.

For hydrogen atoms, containing but a single electron and a proton for a nucleus, the Lamb shift can be measured to an accuracy of a few parts in a million, and theoretical and experimental values agree very well. One would like also to measure the Lamb shift (and hence test basic QED precepts) for other elements. One would like also to measure separately the contribution of higher-order contributions to the Lamb shift.

In hydrogen, two-loop and other higher contributions play a very small role in the Lamb shift. Furthermore, uncertainty in the size of the proton limits any effort to measure two-loop effects. This is not true for a uranium atom in which nearly all the electrons have been stripped away. With a much larger nucleus, the proton-size issue is much reduced, and the electric fields holding electrons inside the atom are a million times stronger (10¹⁷ volts per meter) than in hydrogen. Thus, QED can be tested under extreme conditions.

The Livermore physicists (contact Peter Peiersdorfer, beiersdorfer@llnl.gov) study uranium atoms that have been stripped of all but three electrons. These lithiumlike uranium ions, held in a trap, are then carefully observed to search for the Lamb discrepancy from simple quantum predictions as to the frequencies of light emitted by the excited ions. In hydrogen atoms, the two-loop corrections constitute only a few parts per million, but in uranium atoms they contribute about one third of one percent of the Lamb shift. In this way, the Livermore team has measured this higher-level QED term for the first time, with an accuracy of about one part in ten, or 10 percent.

Beiersdorfer et al., Physical Review Letters, 2 December 2005
For a background article on the Livermore electron-beam ion trap, or EBIT, see the October 1994 issue of Physics Today