*4.2. Planar Waveguide Immunosensor Design*

The PW structures were devised by standard microelectronic processes as a thin (200 nm) layer of Si3N<sup>4</sup> sandwiched between much thicker (3 µm) layers of SiO2. The 200 nm core layer thickness was required to accommodate a single mode electromagnetic wave propagating along the waveguide [25,26]. Due to the large difference in RIs between the Si3N<sup>4</sup> core (*n* = 2.01) and the SiO<sup>2</sup> cladding (*n* = 1.46), the light propagates at an angle of 47◦ (corresponding to an angle of total internal reflection) and with a consequent 3000 reflections/mm approximately. The use of a slant edge of the waveguide cut at 47◦ was an optimal light coupling solution, which provided sufficient light intensity propagating through the waveguide.

In the experimental PI set-up, 650 nm light from a laser diode was focused with a lens to a narrow (less than a millimetre) spot on a slant edge of the waveguide and collected on the other side with a CCD array. A polarising element in front of a CCD camera allows the visualisation of a phase shift between p- and s-components of polarised light. The reaction cell equipped with the inlet and outlet tubes is sealed against a sensing window, which was etched in the top SiO<sup>2</sup> layer. The surface of Si3N<sup>4</sup> in the sensing window could be coated with a biosensing layer. Any changes in RI and/or thickness of this sensing layer affect mostly the p-component of polarised light (while the s-component acts as a reference), thus resulting in a multi-periodic output signal.
