Slashdot Mirror
Scientists Slow the Speed of Light
lightbox32 sends news that scientists have found a way to slow individual photons within a beam of light. Their work was published today in Science Express (abstract, pre-print). The researchers liken a light beam to a team of cyclists — while the group as a whole moves at a constant speed, individual riders may occasionally drop back or move forward. They decided to focus on the individual photons, rather than measuring the beam as a whole. The researchers imposed a particular pattern on a photon, then raced it against another photon, and found that the two arrived at their destination at slightly different times.
The work demonstrates that, after passing the light beam through a mask, photons move more slowly through space. Crucially, this is very different to the slowing effect of passing light through a medium such as glass or water, where the light is only slowed during the time it is passing through the material—it returns to the speed of light after it comes out the other side. The effect of passing the light through the mask is to limit the top speed at which the photons can travel.
This is incredibly cool. Previous work has managed to fully stop light, but this is quite a finding (that light can travel slower through a vacuum).
The old stuff, from Wiki:
In 1998, Danish physicist Lene Vestergaard Hau led a combined team from Harvard University and the Rowland Institute for Science which succeeded in slowing a beam of light to about 17 meters per second,[1] and researchers at UC Berkeley slowed the speed of light traveling through a semiconductor to 9.7 kilometers per second in 2004. Hau later succeeded in stopping light completely, and developed methods by which it can be stopped and later restarted.
However, now we can alter the structure of the beam of light and measure a slowdown (from the abstract):
Our work highlights that, even in free space, the invariance of the speed of light only applies to plane waves. Introducing spatial structure to an optical beam, even for a single photon, reduces the group velocity of the light by a readily measurable amount.
Details from the pre-print:
We use an ultraviolet laser incident upon a beta-barium borate (BBO) crystal to produce photon pairs with central wavelength at 710 nm. The photons, called signal and idler, pass through an interference filter of spectral bandwidth 10 nm and are collected by polarization-maintaining, single-mode fibers. One fiber is mounted on an axial translation stage to control the path length (Fig. 2A). The idler photon goes through polarization maintaining fibers before being fed to the input port of a fiber-coupled beam splitter (Fig. 2B) (17). Instead of going straight to the other beam splitter input, the signal photon is propagated through a free-space section (Fig. 2C). This consists of fiber-coupling optics to collimate the light and two spatial light modulators (SLMs). SLMs are pixelated, liquid-crystal devices that can be encoded to act as diffractive optical elements implementing axicons, lenses and similar optical components. The first SLM can be programmed to act as a simple diffraction grating such that the light remains collimated in the intervening space, or programmed to act as an element to structure the beam (e.g. axicons or lenses with focal length ). The second SLM, placed at a distance 2, reverses this structuring so that the light can be coupled back into the single-mode fiber that feeds to the other input port of the beam splitter. The output ports of the fiber-coupled beam splitter are connected to single-photon detectors, which in turn feed a gated counter (Fig. 2D). The coincident count rate is then recorded as a function of path difference between the signal and idler arms. The position of the HOM dip is recorded as a function of the spatial shaping of the signal photon.
I read TFA and could not specifically find where they showed they adjusted the speed and not just added an initial delay. They ran it through a mask, then onto a ~1 meter long "race track" to compare. I really wanted a clear explanation that they ran the test over 2 lengths to factor out any static delay caused by the propagation through the mask itself.
If they indeed can do this, I would have like to have seen a demonstrate interference pattern showing the beat note between the normal beam and the "slowed" beam. It should be roughly as simply as using a beam splitter, one though their mask, then back into a beam combiner. If coherent laser light is pump in the slower photons should create an interference pattern along the length of the beam that any crummy detector should be able to pick up.
Instead they compared time of arrival over a single distance (as best I can tell from TFA), which is subject to systematic offsets, such as the fixed delay to get through the mask.
Smells like a **detector effect**, like photo-electric emission ... a quantized absorber making a continuous wave appear as a "particle". The honest Scotsmen induced distortion of the continuous wavefront thus making the "front-edge" non-detectable ... like tickling a rhinoz *zz with a thrush-feather not a hemp cannon-swab.