The dependence of the output power on the delay indicated that the generated pulse length was less than a picosecond, which suggested an extremely high direct-modulation rate. Moreover, finer interferometric measurements of the spectral composition of the radiation allowed the authors to establish that the generated pulse was even shorter; on a subpicosecond scale. Thus, this spaser has more than a terahertz in the direct modulation bandwidth — a record-setting achievement.
Presently, fundamental physics of plasmonic nanolasers is rather well understood and widely investigated, both theoretically and experimentally. Although there is still a lot to study, especially with respect to new materials and designs, the attention will naturally shift towards applications, as the recent implementation of spasers in the detection of explosives suggests10. In this vein, the type of ultrafast nanolaser demonstrated by Sidiropoulos et al.2 will potentially allow for ultrawideband optical communications, as it can be efficiently loaded by an optical plasmonic nanofibre. And by directly modulating them using transistor currents, ultrafast on-chip communications in processors with a terahertz speed may also be possible12. Another feasible application is the ultrafast spectroscopy of biological objects, as the sizes of many macromolecules, their complexes and cell organelles are on the same order of magnitude as the transverse size of the spasing mode2.
Avenues for further improvements in spaser operating properties and applications are ultimately determined by the underlying physics. The spaser speed, for example, is related to the modal volume11, which is why smaller means faster. In this respect, the ultrafast behaviour demonstrated by Sidiropoulos et al.2, though unprecedented, is not a fundamental limit. A smaller spaser could even be faster, although further miniaturization may not be possible with the current design. This is most likely due to two peculiarities: first, it works very close to the plasmon frequency, where the losses are relatively large; second, it is based on a continuous metal and is probably leaking its energy into surface plasmon polariton waves, which further increases the losses. So this device may not generate when significantly reduced in size.
Spasers with a nanoparticle as the plasmonic resonator, on the other hand, are known to operate on much smaller scales4, so these may be the design choice for the faster spasers of the future. How the rate of electron–hole relaxation in the gain medium limits the spaser’s direct modulation speed is yet to be studied both theoretically and experimentally. But the achievement reported by Sidiropoulos et al.2 is an excellent launch pad for future progress.
Nature PhysicsVolume: 10,Pages:799–800Year published: (2014)DOI: doi:10.1038/nphys3127
28 September 2014