(ARROW) Anti-Resonant-Reflecting-Optical-Waveguide bio-sensors

Top down image of an ARROW bio-sensor. Inset: SEM images of SC & LC waveguide cross section, Example of time domain signals from fluorescent beads

Our silicon bio-sensor platform is based on Anti-Resonant Reflecting Optical Waveguides (ARROW). Orthogonally intersecting SiO2 based planar solid core (SC) and liquid core (LC) waveguides are fabricated using standard microfabrication techniques (by our collaborators) on top of an alternating layer of high and low index dielectric thin films on a silicon wafer. The index and the thickness of the layers are designed so as to reflect a broad spectrum of light. This enables light guiding in both solid core and liquid core channel. The SC waveguides are excited by butt-coupling a fiber which is coupled to a laser. Fluorescently tagged bio markers are flowed through the LC waveguide. They get excited when they reach the SC-LC intersection. The fluorescence signal is guided and collected through the collection waveguide. Photon counts from the signal is detected using off chip optical components and an Avalanche Photo Detector (APD). Thus detecting time domain signals from individual particles.

SEM image of ARROW bio-sensor with MMI waveguide(Scale bar: 100um). Top down fluorescent image of MMI pattern on LC waveguide. Color coding scheme for 7X multiplexed detection

To have multiplexed detection capabilities the device is also integrated with a wider multi-mode interference (MMI) waveguide. They are designed such that they excite the LC waveguide with specific number of spots for a specific wavelength of light (Ex: 6 spots-756nm, 7 spots-633nm, 8 spots-556nm). The MMI waveguide is excited simultaneously with all three lasers. Different target particles tagged with dyes corresponding to the excitation lasers are introduced in the LC waveguide. Particles tagged with a certain fluorescent dye, let’s say red generate time domain signals specific to the excitation wavelength of the dye i.e. a fluorescence signal with 7 peaks. Using simple peak counting algorithms one can detect multiple targets simultaneously. Please click on the figures below for details.