Dielectric-sphere-based microsystem for optical super-resolution imaging

Well-established imaging techniques proved that features below the diffraction limit can be observed optically using so-called super-resolution microscopies, which overcome Abbe's resolution limit. In traditional far-field microscopy, the introduction of fluorescent samples and engineered light paths was key for this breakthrough. In parallel, near-field techniques with similar performance were developed, but they suffered from a limited field-of-view. The merge of the two approaches was already demonstrated ~15 years ago, when micrometer-sized dielectric objects positioned on a sample were found to be able to image the sample with super-resolution. By observing the sample through the micro-object with a classical optical microscope, the latter could capture a virtual image showing sub-diffraction details. Although this way the near-field information transfer into the far-field by the micro-object was proven and found to be key for enabling super-resolution imaging, the limited field-of-view, as determined by the size of the micro-object, remained an issue. In this dissertation, a novel method is presented that provides a microscopy technique capable of achieving super-resolution without field-of-view restrictions. Based on previous studies, dielectric microspheres were chosen for this imaging technique. First, the working principle of these microspheres was explored by investigating both the illumination and the reflected light path. These findings provided a better understanding on the phenomena working behind microsphere-assisted imaging and allowed to create an engineering toolbox that can be used to design microsphere-based optical systems. This was followed by an investigation on microfabrication techniques, in order to create a microchip that can serve as a bridge between a single microsphere and the macro-sized-components of a classical optical microscope. The resulting chip was later embedded in a custom fixation system that allowed scanning of this microsphere over the sample, while keeping its position fixed compared to the microscope objective. The microscope mounted camera recorded pictures during the scan, which were used to generate a large field-of-view super-resolution image by stitching. After initial successes, the setup was improved in terms of robustness and application range. The new version allowed field-of-view in the millimeter range, while it could be operated in both oil- and water-immersion. Parallel imaging with an array of microspheres was also implemented, which further enhanced the imaging speed. The algorithmic background (including an automated scanning and image reconstruction protocol) of this microscopy method was developed in-house. Its validation showed superior performance compared to existing software. Future developments (e.g. employment of 3D-printing for mass-production, imaging in vivo biological samples, metrology applications) are envisioned. The findings presented here may pave the road for an easy-to-use, generalized super-resolution imaging system.