In our recent work, published in Phys.Rev.Lett. we have demonstrated that waveguide dispersion, combined with material dispersion, can be used to slow or even speed up the signals in nm-thick waveguides. OK, the slowing down is understandable. But - can you really speed up light? Can you make the signals propagate faster than the speed of light? The answer is... both. The notes and animations below illustrate how to achieve the control over the speed of signal propagation, how to manipulate their speed, and what exactly is meant by "slow" and "superluminal" light.

top

Velocities of light

The propagation of light in any material or device is typically described by a set of velocities. The most used ones are phase and group velocities. The latter describes the speed of propagation of the maximum of a wavepacket; the former - the speed of individual waves in the wavepacket. These two velocities may not only differ in "value", but also can be of the different sign. The new research direction in photonics studies these unique systems, also known as materials with negative refractive index.

A waveguide with anisotropic dielectric core is an example of a negative index material. Its properties are very interesting in several ways. First and foremost, such a waveguide does not have a cut-off radius. Which means that in contrast to conventional optical fibers, anisotropy-based waveguide can be used to compress light to arbitrary small spots via device known as photonic funnel. The performance of such a device and its comparison to the performance of a conventional optical fiber is shown in Fig.1.

Fig.1. Light propagation in conventional dielectric fiber (right) and in photonic funnel (left)

Note that in conventional fiber the light stops and reflects as soon as the radius decays to a certain point, known as cut-off radius. In anisotropy-based fiber the light continues to propagate to the nanoscale. Interestingly, the phase velocity in these materials is opposite to the group velocity. This is shown in Fig.2. Note that the pulse as the whole propagates "forward" while individual cycles in the pulse run backward.

Fig.2 Pulse propagation in anisotropy-based waveguide.
Note that pulse runs "upward" (positive group velocity),
while cycles within the pulse run "downward" (negative phase velocity)

top

Controlling the group velocity

In real waveguides both group and phase velocities can be controlled by changing waveguide size or varying material properties (for example, with external laser radiation). Even if the change in material properties is small, it will have a dramatic effect on the group velocity of signals. Now, since the material properties can be varied by external radiation, it is possible to illuminate only a portion of the waveguide, and modify the group velocities within that portion. Figures 3 and 4 show how the external light delays or speeds up the signals in the waveguide.

Fig.3. Slowing down the pulse by lighting the "grey" region of the waveguide.
Black line shows reference pulse (in the absence of external light),
red line shows the delayed signal.
Initially two signals are completely aligned with each other;
upon passage through grey region, the red signal is delayed.  

Fig.4. Speeding up the pulse by lighting the "grey" region of the waveguide.
Black line shows reference pulse (in the absence of external light),
green line shows the sped-up signal.
Initially two signals are completely aligned with each other;
upon passage through grey region, the green signal is advanced.

Note that the maximum of the green signal leaves the active (grey) region before entering (Fig.5). The pulse, is therefore transmitted through the active are with the velocity larger than the speed of light. Of course such a transmission does not violate any physics laws - the front of the pulse still enters the material before leaving it.

Fig.5. Superluminal light: The maximum of the green pulse leaves the active (grey) region before entering. 

top