The power density of optical excitation in microfluidic-photonic-integrated flow cytometers is typically provided from
an integrated waveguide and the beam is therefore divergent within the microchannel due to the NA of the waveguide; a
detrimental effect on detection capabilities as excitation is not uniform throughout the channel and will generate a long
pulse for excitation. Through integration of a lens system specially designed and simulated to collect and reshape 100%
of input power, the excitation power within the microchannel has been controlled to form an optimal spot size within the
microchannel. The device was formed via a one-shot processing method where designs are patterned into a SU-8 layer
on a Pyrex substrate. A poly(dimethylsiloxane) (PDMS) layer was used to seal the device and serve as an upper
cladding for integrated waveguides. Spot sizes were improved from an unfocused width of 86um to less than 40um.
Power densities were controlled throughout the width of the channel - an improvement for flow cytometry applications.
The aim of this paper is to improve the functionality and efficiency of a microfluidic device by optically
simulating all of the components of the device and then identifying and optimizing the areas of the device
where the optics and the microfluidic components overlap. Design of the device will incorporate current
advanced methods, such as a state-of-the-art one-shot manufacturing processand noise reduction techniques
utilizing 8 parallel waveguides. A suitable material to build this device is the polymer SU-8, which has a
refractive index of 1.56 while borofloat glass, refractive index of 1.47, is used as a substrate, and an optical
adhesive (Norland Optical Adhesive 74, index of 1.52) is used to seal the device (and serve as the
waveguide cladding) by filling the gap in between waveguides reducing scattering losses and confining
higher modes. As such, simulations take into account all these parameters. Design of a photomask will
take into account three main sources of loss due to the integration of optics and microfluidics: wall
thickness, channel thickness, and input angle. Simulations have yielded behaviours and values for these
parameters. Wall thickness should be limited to 200um thick which will yield a -5.54dB attenuation (-
2.77dB at the particle) on the input (due to high angular propagation of higher order modes). Input angle of
the waveguides (which is crucial to the elimination of signal noise on the output) has been found to reduce
output signal noise to -9.60dB at an input angle of 74 degrees to the channel.