A Verilog-A model for silicon nanowire biosensors : from theory to verification

Silicon nanowires offer great potential as highly sensitive biosensors. Since the signals they produce are quite weak and noisy, the use of integrated circuits is preferable to read out and digitize these signals as quickly as possible following the sensing event to take full advantage of the properties of the nanowires. In order to design optimized and tailored circuits, simulations involving the sensor itself in the design phase are needed. We propose here a Verilog-A model for silicon nanowire-based biosensors. The model can easily be applied using commercially available electronic design automation (EDA) tools that are commonly used for integrated circuit design and simulations. The model is quite general and comprehensive; it can be used to simulate different types of sensing events, while still being quite simple and undemanding in terms of computational power. The model is described in detail and verified with measurements from two different nanowire sensors featuring aluminum-oxide and hafnium-oxide coatings. Good agreement has been achieved in all cases, with errors never exceeding 21%.