Assemblies of silver nanocubes for highly sensitive SERS chemical vapor detection

Probably Ag nanocube (AgNC) aggregates within cylindrical pores (PAM-AgNC) can be employed as efficient nanostructures for highly efficient, robust, tunable, and reusable surface-enhanced Raman scattering (SERS) substrates for trace level org. vapor detection which is a challenging task in chem. detection. The authors demonstrate the ability to tune both the detection limit and the onset of signal satn. of the substrate by switching the adsorption behavior of AgNCs between highly aggregated and more disperse by varying the no. of adsorption-mediating polyelectrolyte bilayers on the pore walls of the membrane. The different AgNC distributions show large differences in the trace vapor detection limit of the common Raman marker benzenethiol (BT) and a widely used explosive binder N-Me-4-nitroaniline (MNA), demonstrating the importance of the large electromagnetic field enhancement assocd. with AgNC coupling. The SERS substrate with highly aggregated AgNCs within the cylindrical pores allows for consistent trace detection of mid ppb (∼500) for BT analyte, and a record limit of detection of low ppb (∼3) for MNA vapors with an estd. achievable limit of detection of ∼600 ppt. The dispersed AgNC aggregates do not sat. at higher ppb concns., providing an avenue to distinguish between higher ppb concns. and increase the effective range of the SERS substrate design. A comparison between the AgNC substrate and an electroless deposition substrate with Ag quasi-nanospheres (PAM-AgNS) also demonstrates a higher SERS activity, and better detection limit, by the nanocube aggregates. This is supported by FDTD electromagnetic simulations that suggest that the higher integrated electromagnetic field intensity of the hot spots and the large specific interfacial areas impart greatly improved SERS for the AgNCs. Also, the AgNC substrate can be reused multiple times without significant loss of SERS activity which opens up new avenues for in-field monitoring. [on SciFinder(R)]