Modern methods of detection of biomolecules for differential genome sequencing, protein recognition, etc. rely on variety of chemical and optical methods to signal the conjugation of target biomolecules with corresponding capture probes. Although these classical methods are widely used, extremely sophisticated and very reliable, and they form the basis of an industry with billions of dollars of revenue, the techniques are expensive and cumbersome. Therefore, replacement of the classical techniques with less expensive approaches that rely on electronic (rather than chemical) detection of biomolecules has been one of the grand challenges of biotechnology and electronics. The new techniques are based on the fact that biomolecules (e.g., DNA, cancer markers, etc.) have definite charge-states depending on the pH of the surrounding environment, therefore the conjugation of these molecules with capture probes would modulate the current flow between source and drain of a transistor, thereby flagging the conjugation and identifying the molecule (for capture probes with known sequences).
Insulated-gate field Effect transistors (ISFET, circa 1970) has been the earliest known examples of such electronic detection schemes. This generation of electronic detectors, however, failed to compete with chemical detection methods. It has suggested recently that a new generation of surround-gate FETs (e.g., Si-NW and CNT, etc.) will do better: indeed, there are many recent reports of extraordinary sensitivity, response time, and selectivity of these new sensors. Although it is broadly accepted that the Si-NW or nanocomposite sensors should have better sensitivity that those of ISFET and Chem-FETs, the origin of the extraordinary sensitivity remains poorly understood. The standard interpretation of better electrostatic coupling of reduced-geometry devices appears reasonable – but a closer analysis suggests that it would only explain a factor of 2-5 improvement in sensitivity, not 2-4 orders of magnitude improvement in sensitivity that have been observed in experiments.
In this talk, we will use classical diffusion-capture (D-C) model to suggest that it is the "geometry of diffusion" rather than "geometry of electrostatics" this is responsible for this remarkable improvement in sensor performance. We establish a scaling-law (based on the solution of the D-C model) to interpret experiments to date within a simple coherent framework. Our scaling laws resolve many classical puzzles and provide guidance of future design of sensors for improved sensitivity.
Researchers should cite this work as follows:
Muhammad A. Alam; Pradeep R. Nair (2006), "Geometry of Diffusion and the Performance Limits of Nanobiosensors," https://nanohub.org/resources/2048.
MSEE 239, Purdue University, West Lafayette, IN