Room temperature plasmonic nanowire laser near the surface plasmon frequency

Summary form only given. The reduction in the size of semiconductor lasers has opened a path for future on-chip light sources for communications and sensing applications [1]. However, their minimum size is limited by the classical diffraction limit of light. Research has focused on surface plasmon based lasers to circumvent this limit, due to their nature of mixing light and electrons at the metal surface and thus generating intense optical localization [2]. Hybrid plasmonic nanowire lasers are of great interest as they allow mode sizes as small as λ²/400, are simple to fabricate, and allow a strong field confinement [3]. Until now, lasing from sub wavelength semiconductor nano wires has only been achieved at cryogenic temperatures. In this work, we present the first room temperature hybrid-plasmonic semiconductor nanowire laser operating at UV frequencies. We show that ZnO nanowires sitting on a Silver substrate, emitting at 385 nm at room temperature, generate lasing surface plasmons very close the surface plasmon frequency.Experiments are conducted on ZnO nanowires separated from a Silver substrate by a thin LiF gap layer of about 10 nm thickness. The nanowires were optically pumped (100 fs pulses @ 800 KHz) at 355 nm and the band edge luminescence was monitored spectrally at room temperature. We have observed plasmonic laser emission occurring for pump powers moderately larger than non-metal samples in the GWcm-2 range (Fig. 1b). Evidence for the lasing of plasmonic modes arises from a range of tests. Firstly, the optical modes of ZnO nanowires on glass substrates do not lase for diameters smaller than around 170 nm [4], whereas their plasmonic counterparts operated for diameters as small ~100 nm. We also verify surface plasmon lasing by polarization measurements of the emitted light. We have also observed a blueshift in laser emission at threshold with decreasing nanowire diameter, which we associate with both higher loss and slower group ve- ocity. We believe this could be evidence for operation near the surface plasmon frequency, where the group velocity is expected to tend zero or even can become negative (Fig. 1a).These preliminary studies highlight the feasibility of surface plasmon lasers operating near the surface plasmon frequency, despite the extremely high losses associated with fast electron scattering of electrons metals [5].