Investigation of Thermo-Optical Phase Modulator for Chip Based Structured Illumination Microscopy

Abstract

Optical nanoscopy is an emerging field and enables super-resolution of biological cells. Among the existing nanoscopy techniques, structured illumination microscopy (SIM) is most promising candidate for live cell imaging due to its high-imaging speed, limited photo-toxicity and its suitability with most commonly used fluorophore. One of research focus of the optics group at the department is to develop photonic-chip based SIM. For SIM it is essential to perform phase stepping. This thesis involves developing a protocol for fabricating and investigation of on-chip polymer based phase modulator for SIM-on-chip. From our experimental result we can conclude that, fabrication of a polymer-based phase modulator is achievable using lift-off method. The entire fabrication process was optimized locally at UiT. The phase response of on-chip polymer based phase modulator is fast, has sufficient visibility and repeatable over time. I have also demonstrated three equidistant phase stepping as required for SIM imaging. This thermo-optics chip was designed with smaller interference angels (20° and 30°) to be able to capture the interference fringe patterns. We observed that a 60x, 1.2NA objective lens is needed to capture the interference fringes. The experimental results of fringe spacing show good agreement with the analytical results. Some limitation of using the polymer based thermos-optic phase modulation was also observed. One of them is the cross-heating. Experiment shows that for chosen chip continuous phase modulation time should be limited to 15 seconds, after 15 seconds phase shift measurements might not be reliable. After 15 seconds, both the arms of the waveguides are heated and the phase difference reduces. From the experimental results, required power for a π shift using 640 nm, TM mode is 185mW and for 561nm, TM mode it is 230mW. In TM mode phase starts to change with as little as 20mW power. The low power for TM mode also means that the chip will not be heated up so much, which is advantageous for live cell imaging. Whereas, For TE mode phase shift starts at a higher power (≈ 150mW) and total power required for one π shift is considerably more than TM mode. At higher power cross-heating happens much faster. From phase change according to the length experiment, we observed that for 1cm length arm, phase shift starts at ≈ 300mW, but the total power it takes for one π shift is approximately half of the power that is needed for 0.5 cm. Considering all these results, we can suggest that for an ideal SIM chip, one should use 640nm, TM mode in combination with shortest possible heat sensing length.