Using super-resolution imaging to investigate the coupling dynamics of single emitters to plasmonic nanoantennas

Abstract

In the first study (Chapter 2), the plasmonic properties of the structures are explored as the different geometrical parameters, namely sidelength, gap size, tip sharpness, and thickness, are tweaked. We demonstrate the dependence each parameter has on the electromagnetic field strength as well as the position of the localized surface plasmon resonance (LSPR) peak. We show a wide tunable range within the optical spectrum, allowing us to tune the resonance of the structures in an out of the resonance of the emitters. We then fabricate the structures and characterize them both experimentally and computationally.

Lastly, to increase the efficiency and accuracy of our data analysis, we developed a convolutional neural network to analyze single frames mirroring our experiments. We explore different network architectures and tune the associated hyperparameters, developing different neural networks for different applications. The neural network designed for photoluminescence background subtraction performs better than our standard In-House Matlab code using temporal background subtraction and detects molecules at a higher rate. We benchmark the efficacy of our networks against simulated data sets and then apply the neural networks to analyze real experimental data.

The remaining Chapters are purely computational and focus on optimization of the antenna itself or the mechanism by which emitters are delivered to the nanogap region. In Chapter 4, we consider the effect heat generation in the structures has on the surrounding fluid flow and show that by introducing an a.c. electric field into the fluid, micron per second flows can be achieved pointing towards the plasmonic hotspots. In Chapter 5, we explore different antenna geometries by use of a simulated annealing algorithm, finding feasible to fabricate structures that have thirty percent increased electromagnetic field strengths and optical trapping forces.

In Chapter 3, we controllably and reliably place quantum emitters into the gaps of our structures, using single-molecule super-resolution imaging and tracking to investigate the trapping dynamics of single emitters interacting with single nanocavities. We show that molecules trapped in the nanogaps of bowtie antennas have increased photostability and track lengths compared to molecules farther away from the structure. For resonant antennas, these effects are magnified compared to the off-resonant case due to stronger optical trapping effects. Using finite-element method simulations, we calculate the expected fluorescence enhancement values based on the increased radiative rate and quantum efficiency of the emitter in the vicinity of the structures, as well as consider the effects of mislocalization on the true positions of the emitters. These results open the way to using plasmonic optical trapping to reliably couple single emitters to single nanoantennas, enabling further studies of these coupled systems.

The central aim of this thesis is to explore the various ways in which the coupling between plasmonic nanostructures and single-molecule emitters can be tuned. We look at Cy5 fluorophores coupled to gold bowtie nanoantennas (two opposing equilateral triangles with a small nanogap) using super-resolution single-molecule microscopy and demonstrate optical trapping and increased photostability. Having tunability and control of the coupling strength between single emitters and nanocavities has many applications ranging from biosensing to quantum optical computing. Throughout this work, I explore the extent to which we can control this coupling, gaining a deeper knowledge of the physics at play and finding new avenues to explore both computationally and experimentally.