| dc.description.abstract | Preventive action by early detection of disease has been identified as one of the best means of improving health care. This has created a demand for high-sensitivity biosensors. By detecting low levels of disease specific molecules in human samples of blood, saliva, urine or spinal fluids, biosensors can be used to discover illnesses at a stage where they are still harmless and treatable. There is also demand for automated desktop appliances or hand held devices, as these can replace labor intensive analysis, lower costs and improve efficiency.
This thesis covers the development of a novel type of single particle detector that potentially fulfills all of the above demands. Through simulations, fabrication, and optical characterization, I have shown how free-standing dielectric membranes with a well-designed pattern of holes can be used to detect single particles trapped in the holes. The particles are detected with the help of narrowband illumination. In combination with chemical surface functionalization, the detector can potentially arrange for specific capture and detection of particles in the form of proteins, deoxyribonucleic acid (DNA), ribonucleic acid (RNA) or viruses. I estimate the detection limit of the present detector to be particles with a radius of 26 nm, corresponding to the size of a single virus.
The pattern is etched in the dielectric membrane, forming a square lattice of through holes. The permittivity hence varies periodically in two dimensions in the plane of the membrane. Such structures are called 2D photonic crystals (PCs). In general, they possess a number of useful optical properties. The structures that I have designed and fabricated, operate as narrowband filters in the visible range, while supporting resonantly enhanced fields in the vicinity of the membrane. This is achieved by coupling to so-called guided-resonance modes, which are optical modes that concentrate their power in the vicinity of the membrane, similar to fully guided modes. They are different from fully guided modes in the way that they can be coupled to by plane waves, incident on the membrane plane. Fabricated PCs are made in a three layered thin film stack of Si3N4/SiO2/Si3N4, with a total thickness of 150 nm. Lattice periods are in the order of 500 nm, and the holes have a radius in the order of 100 nm.
I have also designed an imaging system based on a standard optical microscope, where particles in the membrane appear as bright spots in the microscope image. Supported by simulation and experimental results, I have developed a model that explains this effect: Particles trapped in my crystals, are detected as a result of enhanced Rayleigh scattering. The intensity of the signal they produce is proportional to the square of their volume, and to the square of the amplitude of the field where they are located. This has motivated a study on how the resonantly enhanced field can be maximized, i.e. a study on high-Q guided-resonance modes.
In theory, high-Q modes can be achieved by decreasing the scattering strength of the PC lattice, for example, by decreasing the hole radius. However, in general this only holds for infinite structures. In my research collaboration, an experiment has been designed and carried out to verify this fact. The experimental results, supported by my simulations, show how the Q-factor of guided resonance modes is fundamentally limited by lattice size. Edge related losses may entail a need for an impractically large number of periods in the lattice, in order for high-Q optical modes to be observable.
I have designed methods that suppress these edge losses, resulting in a PC bound by in-plane Bragg mirrors. I show, both in simulations and experiments, how Bragg mirrors can be exploited to reduce edge related loss. In addition to presenting a way around the fundamental limitation on Q-factors for guided resonances in finite PCs, the new design gives an intuitive demonstration of the physical nature of guided resonance modes. | nb_NO |
