To accurately characterize molecular interactions, determinations should be done in a label-free manner. [1,2] Furthermore, in order to have the capability to fully evaluate the molecular diversity that exists in nature, it would be most advantageous to make these determinations with micro-totalanalysis systems (m-TAS) [3,4] configured for high throughput. [5,6] Yet, conventional detection methods are difficult to implement in m-TAS and are often volume-or sensitivitylimited, both impediments to performing systems-biology analyses.
Investigators have developed various detection methods for use with microfluidic devices that have promise for labelfree nanoscale detection. These include the nanoelectrode, [7] the porous Si (p-Si) sensor, [8,9] the surface plasmon resonance (SPR) detector, [10][11][12][13] and backscattering interferometry. [14][15][16] However, nanoelectrodes foul easily in real-world applications and require multistep manufacturing procedures for the integration into microfluidic chips. The SPR and p-Si methods are sensitive (detection limits of % 10-50 nm for protein interactions [8,9,17] ) and capable of sensing label-free biochemical interactions, yet neither technique is inherently compatible with m-TAS. Integration or immobilization of p-Si into the fluidic network and long solute-diffusion times hinder the use of this method in m-TAS. Since SPR relies on the excitation of plasmons-collective oscillations of free electrons that occur predominantly in metals-SPR surfaces are coated with a thin metal layer (for example, gold). This makes integration of SPR sensors into plastics challenging and relatively expensive as a result of the deposition process. Additionally, the