Factor analysis of the time series of SERS spectra reveals water arrangement and surface plasmon changes in Ag nanoparticle systems.

The enhancement of Raman signals of molecules localized in the vicinity of plasmonic nanoparticles, known as surface-enhanced Raman scattering (SERS) effect, is strongly influenced by the selected excitation wavelength. The optimal excitation wavelength in SERS measurements is given by the position of the surface plasmon extinction (SPE) band of the studied system. Even a small change of the SPE band intensity, position and/or shape during the measurement may influence the SERS signal significantly. In this work, we prepared several systems of Ag nanoparticles, which were used for the demonstration how the information about SPE changes can be obtained by multivariate statistical analysis (factor analysis; FA) from SERS spectral sets, and employed in more precise and more comprehensive interpretation of the results. In non-aggregated Ag colloidal systems measured at the excitation wavelength of 445 nm, SPE band changes could be monitored by the analysis of water stretching vibration together with the vibrations in the fingerprint region. The FA of the water stretching band region was shown to provide unique information on both arrangement and disarrangement of water molecules in the vicinity of Ag NPs during the time evolution of these SERS active systems. In addition, the FA of the fingerprint region helped to monitor a rapid metalation of meso-tetrakis(N-methyl-4-pyridyl)porphine in etched SERS systems with Ag ions released from the NPs surface. In aggregated Ag colloidal systems measured at the excitation wavelength of 785 nm, the FA of SERS spectral sets enabled us to reveal the contribution of the 2nd electromagnetic enhancement to the overall SERS signal. The reliability of our conclusions was verified by comparing the results obtained from FA of SERS spectral sets with the data obtained from the parallel SPE measurements of the studied systems.