Improving the instrumental resolution of sensors based on localized surface plasmon resonance.

The colorimetric variations induced upon changes in interfacial refractive index of nanoscale noble metal structures exhibiting localized surface plasmon resonance (LSPR) provides a convenient means of label-free, affinity-based detection of biomolecular recognition reactions. However, despite being similar in nature to conventional SPR, LSPR has so far suffered from significantly lower data quality in terms of its signal-to-noise ratio (S/N) in typical biomolecular recognition analysis. In this work, generic data analysis algorithms and a simple experimental setup that provide a S/N upon protein binding that is comparable to that of state-of-the art SPR systems are presented. Specifically, it is demonstrated how temporal variations (rate approximately 0.5 Hz) in parameters proportional to the resonance peak position can be recorded simultaneously, yielding a peak position precision of <5 x 10(-4) nm and an extinction noise level of <5 x 10(-6) absorbance units (Abs). This, in turn, is shown to provide a S/N of approximately 2000 (equivalent to a detection limit of <0.1 ng/cm(2)) for typical protein binding reactions. Furthermore, the importance of utilizing changes in both peak position and magnitude is highlighted by comparing different LSPR active noble metal architectures that respond differently to bulk and interfacial refractive index changes.