Theoretical model for the detection of charged proteins with a silicon-on-insulator sensor

For a bio-sensor device based on a silicon-on-insulator structure, we calculate the sensitivity to specific charge distributions in the electrolyte solution that arise from protein binding to the semiconductor surface. This surface is bio-functionalized with a lipid layer so that proteins can specifically bind to the headgroups of the lipids on the surface. We consider charged proteins such as the green fluorescent protein (GFP) and artificial proteins that consist of a variable number of a spartic acids. Specifically, we calculate self-consistently the spatial charge and electrostatic potential distributions for different ion concentrations in the electrolyte. We fully take in to account the quantum mechanical charge density in the semiconductor. We determine the potential change at the binding sites as a function of protein charge and ionic strength. Comparison with experiment is generally very good. Furthermore, we demonstrate the superiority of the full Poisson-Boltzmann equation by comparing its results to the simplified Debye-Huckel approximation.