Natural Science Forum / Physics / General Physics / August 2007

Crueltyfree: Counting photons without killing them
  http://www.sciencenews.org/articles/20070825/fob6.asp

  Davide Castelvecchi

  Disclaimer: No particles were harmed in the making of this
  experiment. Physicists have found a way to count photons as they zip
  along, without destroying them. The researchers say that the
  technique will enable scientists to probe quantum effects that so far
  have been the subject only of speculation.

  In physics labs, detecting light has long been synonymous with
  absorbing photons. Typically, the photons cease to exist and the
  light's energy transforms into an electrical signal. Physicists can
  count single photons--but they haven't been able to count them
  and keep them.

  "Up to now, when you measure light, it's a destructive process," says
  Serge Haroche of the École Normale Supérieure of Paris. Now, Haroche
  and his colleagues have shown how to count photons nondestructively
  while they bounce back and forth between two mirrors.

  Haroche's team began by introducing small numbers of photons into the
  space between two niobium-coated screens. Kept at less than 1 kelvin,
  the niobium became superconducting, which made the screens into
  virtually perfect mirrors. The photons could bounce back and forth up
  to a billion times, lingering inside their hall of mirrors for more
  than a tenth of a second.

  The team then shot rubidium atoms one by one across the photons'
  path. The atoms were in a highly excited state in which their
  electrons were especially sensitive to the photons' electric fields.
  The electrons responded with a shift in the timing of their orbits,
  essentially acting as the hands of microscopic clocks. The amount of
  shift was proportional to the number of photons between the two
  mirrors.

  Quantum uncertainty dictates that the number of photons could not be
  well defined at the start of the experiment. Measuring the influence
  of the photons on a single rubidium atom yielded only incomplete
  information about the number of photons. But after the researchers
  had shot about 100 atoms through the chamber--gaining information
  and reducing uncertainty at each step--the number of photons
  converged to a definite value. Subsequent measurements confirmed that
  count. So far, the team has managed to count up to seven photons,
  Haroche says.

  While the photons didn't die, their lives would never be the same. In
  any experiment, measuring one physical quantity with increasing
  precision leads to increased fuzziness in a related quantity. In this
  case, obtaining a precise count of the photons came at the expense of
  losing knowledge about the relative timing, or phase, of the photons'
  wavelike fluctuations. The findings appear in the Aug. 23 Nature.

  David Hume of the National Institute of Standards and Technology in
  Boulder, Colo., says that the results are "an elegant demonstration
  of the measurement process in quantum mechanics."

  The experiment highlights a little-known aspect of quantum physics:
  When quantities go from a fuzzy state to one with a precise value,
  the transition can take place in small increments. In that way,
  measurements can extract partial information (SN: 5/12/07, p. 292).
  Haroche says that his team's setup could be a means for testing new
  quantum phenomena in which photons occupy multiple states
  simultaneously. "Quantum physics textbooks are illustrated by thought
  experiments," Haroche says. "Now we are doing those experiments."

  References:

  Guerline, C. . . . and S. Haroche 2007. Progressive field-state
  collapse and quantum non-demolition photon counting. Nature 448(Aug.
  23):889-893. Abstract available at
  http://dx.doi.org/10.1038/nature06057.

  Further Readings:

  Castelvecchi, D. 2007. Degrees of quantumness: Shades of gray in
  particle-wave duality. Science News 171(May 12):292. Available at
  http://www.sciencenews.org/articles/20070512/fob3.asp.

  Hume, D.B., T. Rosenband, and D.J. Wineland. Preprint. High-fidelity,
  adaptive qubit measurements through repetitive information transfer.
  Abstract and preprint available at http://arxiv.org/abs/0705.1870.

  Sources:

  Serge Haroche
  Laboratoire Kastler Brossel
  École Normale Supérieure
  CNRS
  Université Pierre et Marie Curie
  24 rue Lhomond
  75231 Paris Cedex 05
  France

  David B. Hume
  National Institute of Standards and Technology
  Time and Frequency Division 847
  325 Broadway
  Boulder, CO 80305

  Luis Orozco
  Joint Quantum Institute
  Department of Physics
  University of Maryland, College Park
  College Park, MD 20742

  David J. Wineland
  National Institute of Standards and Technology
  Time and Frequency Division 847
  325 Broadway
  Boulder, CO 80305-3328

  From Science News, Vol. 172, No. 8, Aug. 25, 2007, p. 117.