PTB News 1/2004 (111 kB) PTB News 1/2004, English edition Issue May 2004Optical communication systems are on the advance. In this context one goal is to achieve a completely optical – as well as parallel – processing of binary information. Technical approaches based on non-linear semiconductor optics have so far failed due to charge carrier diffusion. The latter represents no problem if one uses “spatial solitons” to transfer information.
In the field of optics solitons are primarily known as non-diverging (self-focusing) pulses in fibre optic cables. The pulses do not diverge because the particular non-linear optical properties of the fibre material compensate the dispersion. The “spatial” analogon is when the non-linearity balances the diffraction. Such “spatial” solitons have the properties of bi-stability and transversal mobility. Thus they can be used as optical carriers of information but also as inertia-free probes for microscopy. Whereas such spatial solitons were hitherto demonstrated in slow non-linear materials only, PTB has now successfully created spatial solitons in semiconductor microwave resonators.
High switching velocities and favourable integration with other semiconductor components are now feasible. One application which especially exploits the mobility of the solitons is a “photonic buffer”. The latter is generally needed to deal with problems that require temporary storage (“buffering”) of optical information, particularly, however, to synchronize optical communication networks and for optical “package” transportation.
The figure depicts the first step in making a “photonic buffer”. A quasi-one-dimensional region of a semiconductor cavity resonator is irradiated. On the left-hand side solitons are written successively. They drift with a varible velocity to the right-hand side where they are read out. The drift velocity of the solitons determines the buffer time of the optical information. It can be varied continuously via light field parameters or through an applied electrical field.
The power needed to maintain a soliton in passive resonators is ca. 1 mW, still quite high. By employing an active resonator (VCSEL, vertical cavity surface emitting laser) the power has already been reduced to 80 µW. Even 10 µW appear feasible. This opens the door for two-dimensional applications such as switching matrices. This work was carried out in joint cooperation with further research groups in Germany, France, and the United Kingdom. It was supported by the ESPRIT Programme of the European Union.
C. O. Weiss
phone:+49 (0) 531 592-4400