A very large optical nonlinearity, which occurs when gallium is kept near its melting point, is making possible a new kind of broadband optical switch. Researchers at the University of Southampton (Southampton , UK) have recently filed a patent for devices based on this nonlinearity; devices that they say should make all sorts of optically addressed switching applications practical. So far, the switch has proven to be useful for pulses as short as 10 ns, for modulation rates up to 1 MHz, and for most visible and communications wavelengths. In addition, Southampton scientists have demonstrated that their gallium mirror is particularly suitable for use in Q-switched fiber lasers.
The new switch consists of the cleaved end of a silica fiber that is dipped in a bead of molten gallium1,2. The temperature of the gallium is controlled via a miniature Peltier cell, which in turn controls the reflectivity of the mirror formed at the gallium/silica interface (Figure 1a). The change is slow until the gallium changes from solid to liquid, at which point the reflectivity suddenly shoots up. This is due to the fact that the solid gallium is not very metallic in nature, while it is metallic in liquid form. Heating the gallium past its melting point results in a hysteresis loop, due to a phenomenon called overcooling.
Figure 1. a) By heating gallium past its melting point, the reflectivity of a gallium/silicon mirror can be changed dramatically. The hysteresis loop is caused by a phenomenon called overcooling. b) By using a pump beam, however, a large change in reflectivity can be produced before reaching the melting point. (Filled squares show the heating part of the cycle and empty circles the cooling part.)
Shown in the bottom part of figure 1b is what happens when an optical pump signal is used in conjunction with change in temperature. This experiment was done using the set-up shown in figure 2. Here, a pump (higher-power controller beam -- mW) and a probe (lower-power signal beam -- µW) are coupled together and fed into the fiber that is terminated by the gallium/silica mirror. The reflected beam is routed (via the optical circulator) through a spectral filter to isolate the modulated probe only. This beam can then be detected and its behavior examined.
Figure 2. Schematic for the pump-probe experiments with the gallium fiber-mirror switch.
At just about 1.5° C below gallium's melting point (30°C-Figure 1b), the modulation depth of the probe beam reaches a maximum and then drops sharply as the temperature rises further. This rise before the melting point is caused by the presence of the pump beam. Researchers say that the pump may be forcing the gallium into a metastable phase of matter, possibly another crystalline state or something between solid and liquid, by breaking the covalent bonds in the material's part-metal/part-molecular structure3. By keeping the temperature of gallium within this region, the effect of the pump beam on the mirror is highly nonlinear. With up to a 30 percent change in the intensity of a reflected probe beam, due to just a few milliwatts of optical pump, this makes the device a highly practical optical switch.
As well as investigating the performance of the device for communications applications, the Southampton team also demonstrated its use as a Q-switching mechanism for fiber lasers (Figure 3). Essentially, until the pump beam reaches a certain power (270 mW for an Er+:Yb3+ fiber laser cavity) the laser operates in continuous wave mode. Above that power, Q-switching automatically kicks in, with the repetition rate rising with the power of the pump. Above another limit (in this case 410 mW) the laser goes into a third regime, characterized by rate doubling behavior or irregular pulsing. Researchers have shown that this behavior is inherent to the gallium mirror, and is not caused by any other feature of the laser system, by replacing the mirror with a conventional reflector and experimenting over the entire temperature range.
Figure 3. Passively Q-swiched fiber laser a) using the gallium nonlinear mirror and b) using a conventional mirror as a control.
Research continues in order to more fully understand and exploit this new, and very large, source of optical nonlinearity.
1. P. J. Bennett, S. Dhanjal, P. Petropouls, D. J. Richardson, and N. I. Zheludev, A photonic switch based on a gigantic, reversible optical nonlinearity of liquefying gallium,Applied Physics Letters 73 (13), 28 September 1998.
2. V. Albanis, S. Dhanjal, P. Petropouls, D. J. Richardson, and N. I. Zheludev, Confining metallic gallium - new material structure for optical data processing at milliwatt power level,MRS Fall Meeting, Boston, MA, USA 30 November-4 December 1998.
3. V. Albanis, S. Dhanjal, V. Emelyanov, P. Petropouls, D. J. Richardson, and N. I. Zheludev, Light-induced structural phase transition in confining gallium and associated gigantic optical nonlinearity,MRS Fall Meeting, Boston, MA, USA 30 November - 4 December 1998.
Sunny Bains is a scientist and journalist based in San Francisco Bay area. www.sunnybains.com