New mechanism for microcavity lasers at room temperature
Traditional lasers produce emission by stimulating electrons to drop from a higher to a lower energy level. Initiating this ‘coherent’ lasing requires more electrons in the upper level to start with (inversion). In other words, the power required to turn on a laser scales with the number of electron states or the volume of the active region. Applications such as optical chip interconnects or dense optical switching require many very low power lasers, and low-threshold lasers are highly desirable.
So far, creating low-power devices has meant making them tiny, to minimize the energy required to put all the electrons into their upper state. At the same time, since the optical gain is now small, the mirrors on either side of the cavity (inside which lasing photons must accumulate) have to be extremely reflective. This is the regime in which so-called vertical-cavity surface-emitting lasers operate. But it is a technological challenge, particularly for systems like gallium nitride (GaN), which emit in the blue/UV region. A novel approach that enables photons and carriers to merge results in a coherent emission regime unconstrained by this inversion requirement.
Our polariton laser is a new kind of coherent light emitter that operates in a semiconductor microcavity. An active medium is sandwiched between two high-reflectivity mirrors that trap light with a specific wavelength resonant within the micron-scale size of the cavity (see Figure 1). Photons bouncing back and forth between the mirrors continuously excite bound electron–hole pairs (excitons) in the semiconductor and are re-emitted. If the circulating cavity photons are reabsorbed by their parent excitons before anything else interacts with them, new quasi-particles are formed—polaritons—which exhibit a half-light, half-matter nature different from their original constituents.
The strong coupling between excitons and photons leads to very peculiar properties. In 2000, our group showed that polaritons in gallium arsenide–based microcavities have the largest optical gain in any system, along with huge optical nonlinearities that can be exploited for optical switching devices.1,2 Moreover, we demonstrated that polaritons essentially behave like bosons, that is, they like to congregate in the same quantum state, rather than keeping apart as fermions do (because fermions cannot occupy the same quantum state). We used this principle to build the world's first micron-scale optical parametric oscillator, in which the input pump energy is converted into two output waves of different frequencies, in compliance with the laws of energy and momentum conservation (see Figure 2).
These devices only worked at very low temperatures (T), however, because excitons in GaAs tend to dissociate at T≥50K, hindering industrial application on a large scale. Hence, for the last five years, we and collaborators in the European Union have been developing the rapid-emitting GaN material system, which we predicted should allow operation above 300K.3,4 We have now produced the first semiconductor microcavity laser with coherent polaritonic effects up to room temperature and emission wavelengths lying in the UV region (∼350nm). This paves the way to many applications, for example, in high-density memory storage. The success is based on advances in growing high-quality cavity mirrors from GaN alloys developed at the Swiss Federal Institute of Technology Lausanne, Switzerland, by our collaborators, under the leadership of Nicolas Grandjean.
In a polariton laser, the processes of stimulation and emission (which are inherently linked in a conventional laser) are separated into two stages: the excited semiconductor states now experience stimulated scattering into the lowest polariton state, accumulating at the bottom of the lower polariton branch (see Figure 2). From there, they are emitted into photons as they leak out of the slightly imperfect cavity mirrors. This polariton laser is different from a normal laser because the lasing states do not need to be inverted (i.e., more than 50% excited) to get the laser to turn on. Hence, the pumping threshold for polariton lasers is no longer related to the volume of the active medium, as in a conventional laser, and can be much lower; indeed, we have now made a GaN-based polariton laser which has a 300K threshold 10 times lower than any previous best GaN-based laser.
Since polaritons are bosons, these results have implications for making room-temperature solid-state Bose–Einstein condensates, where polaritons condense into a macroscopic coherent state potentially above 300K. Such systems would allow construction of interferometer chips for high-precision measurements. But much work remains before these possibilities can be achieved: the GaN material is still under development, and electrical pumping needs to be implemented (although in principle there are no roadblocks to doing it). Challenges notwithstanding, the consistent progress seen over the last seven years indicates the rich potential in basic science and novel technologies awaiting polariton devices.
This work was partially supported by Clermont2 (grant EU RTN- 503-677), the StimScat project (grant EU FP6-517769), and the UK Engineering and Physical Sciences Research Council NanoPhotonics Portfolio.
