The
efficiency of signal enhancement for surface-enhanced Raman
scattering (SERS) spectroscopy is a crucial parameter for the design
and development of nanoparticle-based sensing applications. However,
screening SERS capabilities of anisotropic nanoparticles by iterative
synthesis is time consuming, and their prediction still suffers from
the complex nonlinear relationships of morphological and electromagnetic
properties. We present an approach to use a macroscopic gradient array
of substrate-supported nanoparticles for rapid screening of their
SERS efficiencies. The gradient represents a “plasmonic library”
of colloids synthesized by two-step post-modification of a monolayer
of randomly close-packed gold nanospheres covered with poly(N-isopropylacrylamide) shells. A first chemical overgrowth
process yields a continuous gradient of seed particles, with diameters
ranging between 10 and 60 nm. Subsequently, the seeds are further
grown into nanostars (NSs) with spiky tips, which improve their SERS-enhancing
capabilities. Raman mapping along the gradient provides rapid and
reliable quantification of the specific SERS efficiencies for the
whole library, as well as correlation of their optical and structural
properties. By ensuring that the number and density of particles in
the Raman excitation volume remain constant, the most appropriate
synthetic conditions for efficient SERS can be readily identified
on a single screening. As a proof of application, we screened the
SERS performance of a library of NSs and applied the selected best
candidates for the detection of the bacterial biomarker pyocyanin.
This bacterial quorum sensing signaling molecule was quantitatively
detected within a linear dynamic range between 10–7 and 10–5 M, suitable for clinical applications.