LI ET AL. VOL. 5 ’ NO. 8 ’ 6661–6668 ’ 2011 www.acsnano.org 6661 August 04, 2011 C 2011 American Chemical Society Effect of Nanowire Number, Diameter, and Doping Density on Nano-FET Biosensor Sensitivity Jason Li, †,‡,^ Yanliang Zhang, †,^ Steve To, § Lidan You, †,‡, * and Yu Sun †,‡,§, * † Department of Mechanical and Industrial Engineering, ‡ Institute of Biomaterials and Biomedical Engineering, and § Department of Electrical and Computer Engineering, University of Toronto, Toronto ON M5S 3G8, Canada ^ These authors contributed equally to this work. N anowire field-effect transistors (nano- FETs) enable dynamic label-free de- tection of molecules with higher sen- sitivity and shorter detection times compared to conventional bioassays. Research efforts over the past decade have produced signifi- cant advances in nano-FET biosensor technol- ogy and resulted in highly sensitive proof- of-concept devices capable of detecting exceedingly low concentrations of proteins, 1À3 nucleic acids, 4,5 and viruses 6 in solution. In order to achieve high performance and con- sistency across devices, understanding sens- ing mechanisms and the effect of important parameters is important. A number of experi- mental studies have been reported, which sought to elucidate the sensing mechanism and the effect of various device parameters on nano-FET sensitivity including electrode material, 7 nanowire composition, 8,9 functio- nalization method, receptor size, 10,11 gate bias, 12À14 electrolyte ion concentration, 15,16 and analyte delivery methods. 17À19 How- ever, the influence of nanowire number, doping density, and diameter on nano-FET biosensor sensitivity remains to be experi- mentally quantified. Previous studies explored the influence of nanowire number, doping density, and dia- meter on device sensitivity in the context of nanowire electrical transport studies, 20À22 gas-phase chemical sensing, 8,23 aqueous sensing of pH and ionic species, 3,24 and nanoribbon FETs. 25,26 However, fundamental differences between these sensor systems and nanowire FET biosensors require that the effect of these parameters be experimen- tally examined under biomolecule sensing conditions to obtain quantitatively meaning- ful relationships. Several numerical models were also established for qualitatively pre- dicting nano-FET biosensor sensitivity de- pendence on nanowire diameter, doping concentration, and number. 12,27,28 Herein we experimentally determine the influence of nanowire number (the number of bridging nanowires incorporated into each device), nanowire doping density, and nanowire diameter on the sensitivity of silicon nanowire FET protein sensors using human immunoglobulin G (hIgG) as a model analyte. As existing large-scale nano-FET construction methods such as directed self-assembly, 29À31 contact print- ing, 32À34 flow alignment, 35 and dielectro- phoresis 36,37 of presynthesized nanostruc- tures are incapable of precisely controlling the diameter and/or number of nanowires incorporated into each device, we further present a unique fabrication method to achieve reliable nanowire number and dia- meter control through a combined use of * Address correspondence to sun@mie.utoronto.ca, youlidan@mie.utoronto.ca. Received for review June 14, 2011 and accepted July 26, 2011. Published online 10.1021/nn202182p ABSTRACT Semiconductive nanowire-based biosensors are capable of label-free detection of biological molecules. Nano-FET (field-effect transistor) biosensors exhibiting high sensitivities toward proteins, nucleic acids, and viruses have been demonstrated. Rational device design methodologies, particularly those based on theoretical predictions, were reported. However, few experimental studies have investigated the effect of nanowire diameter, doping density, and number on nano-FET sensitivity. In this study, we devised a fabrication process based on parallel approaches and nanomanipulation-based post-processing for constructing nano-FET biosensor devices with carefully controlled nanowire parameters (diameter, doping density, and number). We experimentally reveal the effect of these nanowire parameters on nano-FET biosensor sensitivity. The experimental findings quantitatively demonstrate that device sensitivity decreases with increasing number of nanowires (4 and 7 nanowire devices exhibited a ∼38 and ∼82% decrease in sensitivity as compared to a single-nanowire device), larger nanowire diameters (sensors with 81À100 and 101À120 nm nanowire diameters exhibited a ∼16 and ∼37% decrease in sensitivity compared to devices with nanowire diameters of 60À80 nm), and higher nanowire doping densities (∼69% decrease in sensitivity due to an increase in nanowire doping density from 10 17 to 10 19 atoms 3 cm À3 ). These results provide insight into the importance of controlling nanowire properties for maximizing sensitivity and minimizing performance variation across devices when designing and manufacturing nano-FET biosensors. KEYWORDS: nano-FET . biosensor . nanowire number . nanowire diameter . nanowire doping density . nanomanipulation . nano-FET fabrication ARTICLE