Self-Assembled Nanoparticle Aggregates: Organizing Disorder for High Performance Surface-Enhanced Spectroscopy

The coherent oscillations of the surface electron gas, known as surface plasmons, in metal nanostructures can give rise to the localization of intense electromagnetic fields at the metal-dielectric interface. These strong fields are exploited in surface enhanced spectroscopies, such as Surface Enhanced Raman Scattering (SERS), for the detection and characterization of molecules at very low concentration. Still, the implementation of SERS-based biosensors requires a high level of reproducibility, combined with cheap and simple fabrication methods. For this purpose, SERS substrates based on self-assembled aggregates of commercial metallic nanoparticles (Nps) can meet all the above requests. Following this line, we report on a combined micro-Raman and Atomic Force Microscopy (AFM) analysis of the SERS efficiency of micrometric silver Np aggregates (enhancement factors up to 10(9)) obtained by self-assembly. Despite the intrinsic disordered nature of these Np clusters, we were able to sort out some general rules relating the specific aggregate morphology to its plasmonic response. We found strong evidences of cooperative effects among the NPs within the cluster and namely a clear dependence of the SERS-efficiency on both the cluster area (basically linear) and the number of stacked NPs layers. A cooperative action among the superimposed layers has been proved also by electromagnetic simulations performed on simplified nanostructures consisting of stacking planes of ordered Nps. Being clear the potentialities of these disordered self-assembled clusters, in terms of both easy fabrication and signal enhancement, we developed a specific nanofabrication protocol, based on electron beam lithography and molecular functionalization, that allowed for a fine control of the Np assemblies into designed shapes fixing their area and height. In particular, we fabricated 2D ordered arrays of disordered clusters choosing gold Nps owing to their high stability. AFM measurements confirmed the regularity in spacing and size (i.e. area and layer number) of the aggregates. Preliminary SERS measurements confirm the high signal enhancement and demonstrate a quite good reproducibility over large number of aggregates within 100x100 mu m(2) 2D super-structure. The availability of such a multisensor could allow a careful statistical analysis of the SERS response, thus leading to a reliable quantitative estimate of the presence of relevant molecular species even at ultra-low concentration.