Group postdoc Stephen Church has just had a paper accepted into Advanced Optical Materials. In this work, he reports the study of thousands of individual nanowire lasers, using advanced statistical methods to identify a route to optimization.
Photonic integrated circuitry is rapidly emerging as an alternative to electronics for specialist applications in machine learning, neuromorphic computing, and photonic quantum computing. While many of the key components have been reported, there is still a significant lack of efficient, easy to manipulate on-chip light sources, because the most common platform for photonics is silicon: an indirect semiconductor which cannot efficiency emit light.
Nanowire lasers provide one route to integrable on-chip light emission, as they can be either transferred or directly grown onto photonic circuits. In a single piece of semiconductor crystal they can act as a complete laser; however, because lasers are highly non-linear and hence challenging to control, small variations in quality or shape can lead to large changes in performance.
In this study, Stephen has studied over 5000 nanowire lasers using a range of measurements - lifetime to understand material quality, laser linewidth to measure reflectivity, emission shape to understand carrier dynamics. Taken together, these measurements allow correlations to be understood and allow us to see past wire-to-wire variation to understand trends.
In a key result, we show that, unexpectedly, variation in material quality of the emission region dominated other effects arising from reflectivity. This finding directly guides future efforts to optimize material properties rather than shape in the first instance.