Improving the Limit of Detection of Nanoscale Sensors by Directed Binding to High-Sensitivity Areas

Journal article, 2010

The revelation of protein protein-interactions is one of the main preoccupations in the field of proteomics. Nanoplasmonics has emerged as an attractive surface-based technique because of its ability to sense protein binding under physiological conditions in a label-free manner. Here, we use short-range ordered holes with a diameter of similar to 150 nm and a depth of similar to 50 nm as a nanoplasmonic template. A similar to 40 nm high cylindrical region of Au is exposed on the walls of the holes only, while the rest of the surface consists of TiO2. Since the sensitivity is confined to the nanometric holes, the use of two different materials for the sensor substrate offers the opportunity to selectively bind proteins to the most sensitive Au regions on the sensor surface. This was realized by applying material-selective poly(ethylene glycol)-based surface chemistry, restricting NeutrAvidin binding to surface-immobilized biotin on the Au areas only. We show that under mass-transport limited conditions (low nM bulk concentrations), the initial time-resolved response of uptake could be increased by a factor of almost 20 compared with the case where proteins were allowed to bind on the entire sensor surface and stress the generic relevance of this concept for nanoscale sensors. In the scope of further optimizing the limit of detection (LOD) of the sensor structure, we present finite-element (FE) simulations to unravel spatially resolved binding rates. These revealed that the binding rates in the holes occur in a highly inhomogeneous manner with highest binding rates observed at the upper rim of the holes and the lowest rates observed at the bottom of the holes. By assuming a plasmonic field distribution with enhanced sensitivity at the Au-TiO(2)interface, the FE simulations reproduced the experimental findings qualitatively.