http://hdl.handle.net/10012/14271

Microwave technology presents tremendous potential as a remote-sensing technology for a wide range of applications spanning from life science research to food industries, pharmaceutical research, and new material discoveries. Integration of microwave sensing with microfluidics for sample processing makes it an ideal choice for point of care applications highly demanded in resourcelimited areas. The vast majority of the existing microwave sensors are manufactured using sophisticated soft lithography technology which has largely limited its development and applications. There is a large demand for developing new fabrication approaches for the feasibility of mass production at a reasonable cost. In this thesis, a new, yet simple method is developed to fabricate split ring resonator (SRR) based microwave sensors. A simple RLC model is used to characterize the resonant frequency of the SRR, and the equations for calculating the RLC’s resonant frequency is modified to predict the SRR’s resonant frequency base on its geometry. The design is also validated by comparing the simulation results obtained using the commercial software HFSS, and measurements from a real SRR developed sensor. The double ring structure was fabricated onto a printed circuit board by using the industrial photolithograph method. Coating with PDMS and epoxy layer as the passivation layer was tested and compared. Two testing approaches using the SRR sensor developed in this thesis are implemented in this thesis. Their performance for real-time sensing is characterized by applying it to differentiate chemical diary samples and other chemical solutions. In the dipping mode, the sensor is dipped in the material under test (MUT), and in the microfluidic channel mode, the sensor is integrated with a microchannel. The MUT is characterized by analyzing the spectrum data of the reflection coefficient as the function of frequencies. Experimental results indicate that this sensor is capable of differentiating various liquid samples such as DI water, ethanol, isopropanol, oil and salt solutions. Linear relationships between the resonant frequency and the concentrations of chemical composites are also observed in ethanol solutions (0-90%), and salt solutions (NaCl). This sensor is also used to differentiate various milk samples and milk dilutions and it is capable of distinguishing milks with different fat percentages and protein contents. A fully customized vector network analyzer (VNA) is also developed. The circuit structure is designed by referring the existing customized VNAs that were implemented in previous work by iv other lab colleagues. Modifications are made including replacement of the microwave source, using Arduino platform to perform controlling and data acquisition, addition of a harmonic filtering device, and development of a calibration algorithm. The device is validated by comparing its measuring result with a commercial VNA. The customized VNA is able to output a similar spectrum pattern as the commercial VNA, but with slightly shift of the peak frequency.

http://hdl.handle.net/10012/14271