| dc.description.abstract | This dissertation is composed of two projects that explored two new platforms for measuring atomic interactions using simpler designs than in the literature. The first project of this dissertation designed a platform that enables the measurement of Lennard-Jones interaction between two solid surfaces in the form of Atomic Force Microscope (AFM) probe, using different techniques from Micro electrical Mechanical Systems (MEMS). MEMS by definition implies a mechanical and electrical parts of a system. There are many defects and imperfections that emerges on both sides of the system. On the mechanical side, one of the most common imperfections is residual stress, where most fabrication recipes are designed to eliminate it. Residual stress on films causes curvature (manifested as buckling, bending, etc.) for structures that are meant to be straight. On the electrical side, fringing field is considered very complicated to model, and too small to experimentally detect and separate from the main direct electrostatic field; hence, mostly it gets ignored in modelling. This project will try to make a benefit of these two unwanted phenomena combined (residual stress and fringing field) to make a new design for an Atomic Force Microscope (AFM) probe (tip). The tip behavior is first analyzed and modeled statically using COMSOL software, then dynamically using Mathematica software. Both models were combined and compared with the experimental results obtained by an optical profilometer, scanning electron microscope, and a vibrometer. It was found that the model gave good predictions of the experimental behaviors, except with higher displacement amplitude of the model than that of experiment. The reason is due to the purposeful curvature of the probe (cantilever) induced by residual stress, which caused some parts of the probe not to be on the same level with the electrode; hence, weakened its actual response experimentally. Since use of correction factors to account for fringing field is nothing new, a correction reduction factor was introduced to lower the model response to match that of the experiment. The results show that the structure of the actuator (parallel plate or a single comb finger) is not of importance in modeling fringing field, as we have applied literature force modeled for non-curved parallel plate capacitors for our curved comb-finger structure and got identical response to our comb-finger derived new force with a matter of just a correction factor (i.e. free parameter). We have also shown that the curvature equation is unnecessary in the model, and the behavior of the curved probe can be modeled as a straight one.
The second project of this dissertation is another simple design for enhancing light-matter interaction between a single laser beam and an atomic gas (cesium) in what is known as cavity Quantum Electrodynamics (QED). Increasing the interaction between light and matter is inspired by the desire to unravel more understanding about the nature of both interacting entities: light and matter. This can be enabled by engineering necessary platforms where such maximally interacting light and matter can be realized. Usually there are two ways to increase such interaction: 1) increase transverse confinement, and 2) increase the interaction time (in addition to increasing the number of atoms). Each of these two ways is done in a separate platform design. This second project proposes a new platform that can have both ways: increasing both transverse confinement and interaction time by using the hollow core of photonic crystal fiber as the interaction host (hence blocking light from propagating transversally by the photonic bandgap effect), while the light will be bounced back and forth against the atomic gas, not by the conventional Fabry-Perot cavity, but instead by inscribing a Bragg grating mirror on the walls of the hollow core (hence, increase interaction time). The unblocked hollow core will allow easier atomic gas insertion. Different mirror inscription methods were studied, and the best method was employed using a photoresist-assisted layer, instead of direction inscription on the core silicon wall. Initial numerical modeling was done using Lumerical software that gave the Bragg parameters corresponding to the best Bragg mirror reflection which was up to 99.99% reflectivity from only about 300 Bragg periods (shorter mirror) corresponding to only ~100 µm penetration depth. Moreover, since the hollow core photonic crystal fiber is of a high cost, an injection port was designed and built to enable low fiber material loss caused by conventional injection. | en |
