Evanescent Waveguide Sensors for Biomedical Applications

The American Cancer Society reported that the major public health problem is cancer. Based on the report, colonic cancer is the major cancer type in the digestive system. In addition, the percentage of death rate due to this cancer is 50%. Surgery to remove the part of the colon infected by the cancer and then reconnect the healthy one is called anastomosis. However, after the anastomosis, there is sometimes leakage that gives an opportunity for bacteria inside the colon (E. coli) into leak to the abdominal cavity and cause infections, multiple organ failure or even mortality. The leakage is serious problem but unfortunately it is difficult to predict early. Consequently it is often too late to take action for the patients. The leakage can be examined by detecting bacteria in drain fluid. However by using conventional methods, it will take long time. In this thesis, a new approach to diagnose the leakage by detecting E. coli in drain fluid was demonstrated. The approach uses sensor based on evanescent waves during the propagation of light in a waveguide. The wave propagates on the waveguide surface and it is a very good candidate for this application because it is very sensitive to environmental changes. In general, there are three aspects to be considered to develop high quality sensors i.e fabrication, sensitivity, and selectivity. Simulation using effective index method (EIM) and calculation were conducted to estimate the optimum dimensions of the sensor so that it affects the existence of the evanescent wave positively and influences the sensitivity of the sensor. In addition, to enhance the sensitivity, the structure of the sensor was designed as freestanding so that both surfaces are able to react to the surrounding. Material development consists of thin film deposition of TiO2 using atomic layer deposition (ALD) and characterisation of the thin film: surface roughness, thin film stress and optical properties. Based on the characterisations of the TiO2-ALD, the material was found to have good properties as waveguide sensor material. Lithography-based micromachining by combining bulk and surface micromachining was used to fabricate the waveguide in the cleanroom and MEMS-lab. The advantage of the method is in the integration aspect, so that the device can be integrated with other platforms such as CMOS, optical systems, and microfluidic systems. Drain fluid is a highly complex fluid which consists of bacteria, plasma, fluids etcetera. The sensor must be able to detect the target of interest in the complex fluid. Consequently, selectivity is an important property in the development of the sensor. In order to enhance selectivity, the evanescentwaveguide sensor surfaces were functionalised using antibody for E. coli. In this treatment, self-assembly monolayers (SAM) and coupling agents (protein-A and EDC) were deposited on the surface prior to antibody deposition to increase bonding between the waveguide surface and antibodies. The sensor sensitivity was evaluated using isopropyl alcohol (IPA) by measuring the evaporation rate of several different concentrations of IPA in water. Functionalised sensor sensitivity was tested using solutions of E. coli K12 in LB medium. Furthermore, selectivity of the sensor was also examined in this experiment. Solutions that consist of constant E. coli and several different concentrations of Bacillus Subtilis in LB medium were used for this testing. The sensor responses show that the variation of Bacillus does not affect the output signal of the sensor. Different concentrations of drain fluid from patients were also used to evaluate the operation of the sensor. In the experiments using drain fluid, the different concentrations of drain fluid could be detected by the sensor. The results show that the sensor has good sensitivity and selectivity for chemical and biological applications. Finally, to evaluate the reusability of the sensor, the reproducibility and repeatability of the sensor was investigated. By using several samples and measuring the samples in two conditions (load and unload) several times, the reproducibility of the sensor was determined.MicroelectronicsElectrical Engineering, Mathematics and Computer Scienc