Controllable nanofabrication of aggregate-like nanoparticle substrates and evaluation for surface-enhanced Raman spectroscopy

The development of new and better substrates is a major focus of research aimed at improving the analytical capabilities of surface-enhanced Raman spectroscopy (SERS). Perhaps the most common type of SERS substrate, one consistently exhibiting large enhancements, is simple colloidal gold or silver nanoparticles in the 10-150 nm size range. The colloidal systems that are used most for ultrasensitive detection are generally aggregated clusters that possess "hot spot(s)" within some of the aggregates. A significant limitation of these synthetic substrates is that the "hot" aggregates are extremely difficult to create consistently or predict. Electron beam lithography (EBL) along with combinatorial spectral mapping can be used to overcome this limitation. Our previous work, and that of other researchers, invokes the special capabilities of EBL to design and fabricate periodic, highly ordered nanoparticle arrays for SERS. Building on this work, EBL, in conjunction with ancillary fabrication steps, can be used to create complex patterns that mimic random aggregates. These aggregates, unlike those created by colloidal deposition methods, can be uniquely reproduced within the resolution limits of EBL. In the work reported herein, we use a unique approach to create substrates containing a large number of randomly generated cells with different morphologies that are arrayed on silicon wafers. Instead of isolated metal nanoparticles, these structures resemble the aggregates of colloid. By spectral mapping, we investigate the SERS activity of the combinatorial arrays of cells using probe analytes. Two general categories of shapes are randomly designed in different sizes and densities into several hundred different 5 μm square cells. Following fabrication, it is shown that a SERS performance contrast of more than a factor of 44 is achieved among these cells and that the best performing cells can be cloned into uniformly high performing macropatterns of lithographically defined nanoaggregates (LDNAs). In this manner, extended LDNA surfaces with uniform 5 × 10 ⁸ enhancement factors are created. Furthermore, the LDNAs can be further dissected and studied in an effort to increase the SERS enhancement per unit geometric substrate area.