Cutting edge: Lijun Wang

An experiment with light speed showed that a pulse sent through a certain type of cell appeared to exit the cell before it entered.

Logic is the foundation of modern science and it requires that causality should never be violated. In other words, the cause must precede the effect - otherwise we face an unpleasant situation. If an effect can occur before its cause, then we can observe the effect and undo its cause. Do we still have the effect or not? Of course, this has never happened for any logically connected cause and effect. Mysterious events that plagued the minds of our ancestors with superstition were merely coincidences that were not logically related.

We know that the speed of light in a vacuum, c, imposes an absolute upper limit on all velocities. This is certainly true for any object with a mass, for before that object reaches the speed of c, its energy would have approached infinity. This simple fact in relativity has been succinctly summarised in a well-known statement: nothing can move faster than light.

However, there is an exception. The "group velocity" of light itself, a velocity at which a pulse of light in the form of a group of component waves moves, can exceed c. The group velocity can be very different from the simple wave velocity (also called the phase velocity) in a dispersive medium where light waves of different colours have slightly different velocities. In transparent materials such as air, glass and water, red light naturally propagates slightly faster than blue light, resulting in an effect called normal dispersion. This is the cause of the rainbow, in which white light separates into its components while travelling through water droplets in the atmosphere. The group velocity of light in these materials is slower than its phase velocity, which is already slower than c. The reverse can also occur when blue light moves faster than red light in materials that have "anomalous dispersion", resulting in a "superluminal" group velocity. These materials are naturally opaque, although they transmit the frequencies we use. A light pulse propagating through such a medium often shows only a very small transmission. Although its peak may emerge sooner than propagating through the same length in a vacuum, its observation has resulted in controversy in the past.

This is where we came in. By artificially creating a medium that has two amplifying resonances, we created a transparent material with anomalous dispersion. When a light pulse is propagated through the medium, we observed that the exit pulse showed essentially no change in its intensity or shape. The pulse propagating through the atomic medium in the experiment emerged from the exit side so much earlier than if it had gone through the same length in a vacuum that the peak of the pulse left the cell before entering it. The phenomenon is caused by the unusual interference effect of the various wave components of a light pulse inside an anomalously dispersive medium and is a direct result of the wave nature of light.

Finally, the observed superluminal - higher than c - group velocity of light is not at odds with causality or special relativity. Simply put, causality requires only that information cannot be transmitted faster than c and it remains the case for this experiment. Although the group velocity of light is different from "signal velocity", at which information is transmitted, the latter has not been well defined in accordance with laboratory practice. The study may yield a more practical way of thinking about light signals.

Lijun Wang is a research scientist at the NEC Research Institute in Princeton, New Jersey.