Algorithmic approach for designing plasmonic nanotweezers.

We use machine learning (simulated annealing) to design plasmonic nanoapertures that function as optical nanotweezers. The nanoapertures have irregular shapes that are chosen by our algorithm. We present electromagnetic simulations that show that these produce stronger field enhancements and extraction energies than nanoapertures comprising double nanoholes with the same gap geometry. We show that performance is further improved by etching one or more rings into the gold surrounding the nanoaperture. Lastly, we provide a direct comparison between our design and work that is representative of the state of the art in plasmonic nanotweezers at the time of writing.