| Charge based detection of biological and chemical molecules at ultra low concentration using functionalized silicon nanowire (SiNW) field effect transistors (FET) have been successfully demonstrated by several research groups. In particular, SiNWs that are lithographically fabricated on silicon-on-insulator (SOI) substrate have often been chosen in favor of its complementary metal oxide semiconductor (CMOS) compatible processing techniques. Despite the promising empirical results reported in the field, consistency of sensor sensitivity, stability and device characteristics have been lacking. These performance discrepancies come from the highly complex nature of such a sensor system, which combines many physical, electro-chemical and biological factors, each having a distinct impact to device characteristics and sensing results. A comprehensive scheme is in need to disentangle the various contributing factors to the performance of such sensors and to suggest design rules for sensor fabrication and experimental setups. The object of this dissertation is to address this critical need and bridge the gap between the empirical results and the theoretical understanding of such sensor system. We present here a numerical simulation platform for quantitative device level analysis of SiNW bioFET sensor system through comprehensive modeling and vigorous computation. Simulation results are contrasted with experimental measurements obtained in our lab, and the good agreements between the two validates the approach. The platform is used to investigate the impact of biasing condition, reactive surface property geometrical scaling and process related defects such as trapped charge to device characteristics, sensor sensitivity and linearity for uniformly distributed ions or discrete biomolecules detection. Beyond basic sensor performance analysis, our method can also provide guidance to critical parameters in system design and experimental setup. Two studies are presented to showcase such capability: 1) In the study on impact of electrochemical property of sensor surface and surface functionalization techniques to sensor performance, our method is used to determine the optimal pH level for a given case of target protein and antibody pair involved; 2) In the stochastic treatment of doping level distribution induced performance variation, our method is used to determine the optimal biasing condition for overall sensitivity in a multi NW sensor system. |
