Performance Comparisons of Conventional and Line-Focused Surface Raman Spectrometers

Abstract

Four Raman spectrometer configurations were compared for obtaining spectra of surface monolayers, including a new design that employed a line rather than point focus. Each spectrometer used a 514.5 nm laser and charge-coupled device (CCD) detector, but they differed in collection efficiency and sampling optics. Previously defined figures of merit for Raman signal and signal-to-noise ratio (SNR) were determined for each spectrometer, to aid quantitative comparison. The figure of merit for SNR, FSNR, is demonstrated to be useful for comparisons because it permits prediction of SNR for a given spectrometer, sample, and measurement conditions. A rigorous definition of FSNR is based on power density and takes into account the laser damage threshold of the sample. A simpler but less rigorous definition is based on laser power at the sample rather than power density and may be more useful to users who rarely determine the laser spot size. A new spectrometer design employing line focusing and collection is presented, with f/2 optics and a 6 mm slit image at the CCD. A proprietary aberration correction prevents slit image curvature common to most spectrographs with low f/#, and permits full height binning of the CCD. The line-focused spectrometer yielded an SNR and FSNR which are comparable to those for a point focus using the same collection optics and slightly lower than those for the most efficient spectrograph examined. However, the line focus permitted much lower power densities to be employed, or yielded much larger signal for the same power density at a point focus. In quantitative terms, the new line-focused design yielded an SNR which is 67 times that of the best point-focused system, for the same sample, measurement time, and laser power density.