Chemical and biochemical sensors based on silicon nanowire field-effect transistor arrays

Field-effect transistors (FETs) made from semiconducting nanowires have great potential as electronic biochemical sensors if they can be integrated as an array in a CMOS-compatible architecture together with microfluidic channels and interfacing electronics. Such nanoscale electronic transducers based on ion-sensitive field-effect transistors could be mass fabricated at reasonable costs. This, in combination with their small size, makes them ideal for personalized medicine and for future implanted sensing devices.
The sensing principle is based on adsorption of charged species on the sensor surface, leading to a change in surface potential and subsequently a change in current in the FET channel. Thereby, the high-impedance input signal is transformed into a low-impedance output signal, which is an advantage against classic ion-selective electrodes. The potential for downscaling and integration for the simultaneous detection of multiple parameters make silicon nanowire FETs a promising platform to meet the demand for cheap, multifunctional and scalable sensors. Even though many promising results on chemical and biochemical sensing have been achieved so far, a detailed understanding of the electrolyte surface interaction is still missing. Inconsistent outcomes regarding the effect of electrolyte concentrations and electrical noise, suggest that further quantitative studies are needed. The aspect of the size compatibility between the sensor unit and the analyte species is often emphasized to favor nanoscale FETs. Another aspect, often mentioned, is the surface to volume ratio. Hence smaller sensing units should enhance the sensitivity of the sensor, allowing the detection at ultra-low concentrations or a small number of molecules. Furthermore, the capacitances decrease for smaller sensing units, which could lead to faster response times. However, other aspects such as the intrinsic electronic noise, the analyte diffusion time and surface reaction kinetics have to be considered for the development of an applicable sensor.
This thesis was part of a research project aimed at developing a modular, scalable and integrateable sensor platform for the electronic detection of analytes in solution. The main focus lies on the sensor-solution interface and thus the thesis quantitatively compares the experimental data with analytical models.
In this work we have established a versatile sensor platform based on silicon nanowire arrays. The sensor functionality was changed by surface modification for the detection of various analytes such as pH, alkaline ions and even FimH proteins. We achieved an ideal pH sensor with a response close to the Nernst limit. Full surface passivation for protons was accomplished for the implementation of a nanoscale reference electrode. Using the differential signal from differently functionalized silicon nanowires we could detect sodium and potassium ions selectively. Ultimately we present the detection of protein-ligand interactions of the physiologically relevant protein FimH. An extended site binding model was derived to calculate the theoretical limits and assess the properties of the surface groups by evaluating the experimental results.