Optical trapping of small structures is a powerful tool for the manipulation and investigation of colloidal and particulate materials. The tight focus excitation requirements of optical trapping are well suited to confocal Raman microscopy. In this work, an inverted confocal Raman microscope is developed for studies of chemical reactions on single, optically trapped particles and applied to reactions used in solid-phase peptide synthesis. Optical trapping and levitation allow a particle to be moved away from the coverslip and into solution, avoiding fluorescence interference from the coverslip. More importantly, diffusion of reagents into the particle is not inhibited by a surface, so that reaction conditions mimic those of particles dispersed in solution. Optical trapping and levitation also maintain optical alignment, since the particle is centered laterally along the optical axis and within the focal plane of the objective, where both optical forces and light collection are maximized. Hour-long observations of chemical reactions on individual, trapped silica particles are reported. Using two-dimensional least-squares analysis methods, the Raman spectra collected during the course of a reaction can be resolved into component contributions. The resolved spectra of the time-varying species can be observed, as they bind to or cleave from the particle surface.

Quantitative, high resolution (0.1 cm -1 ) infrared spectra have been acquired for a number of pressure broadened (101.3 KPa N 2 ), vapor phase chemicals including: Sarin (GB), Soman (GD), Tabun (GA), Cyclosarin (GF), VX, nitrogen mustard (HN3), sulfur mustard (HD) and Lewisite (L). The spectra are acquired using a heated, flow-through White cell of 5.6 m optical path length. 1 Each reported spectrum represents a statistical fit to Beer's law, which allows for a rigorous calculation of uncertainty in the absorption coefficients. As part of an ongoing collaboration with the National Institute of Standards and Technology (NIST), cross-laboratory validation is a critical aspect of this work. 2,3 In order to identify possible errors in the Dugway flow-through system, quantitative spectra of isopropyl alcohol from both NIST and Pacific Northwest National Laboratory (PNNL) are compared to similar data taken at the Dugway Proving Ground (DPG).

Solutions of organic compounds are proposed as viable high-pass, Rayleigh rejection filters for ultraviolet resonance Raman spectroscopy. The steep transmittance curves of these solutions effectively reject elastically scattered light in this region while passing Raman-shifted frequencies. The materials used in the filters are readily available and inexpensive, and the solutions are easily prepared. Filters for four lines in the range of 288 nm to 342 nm from a Raman-shifted 3rd and 4th harmonic of a Nd:YAG laser are presented, although the principle of preparing similar liquid filters can be applied to virtually any near-UV wavelength. The use of these filter solutions in conjunction with a single monochromator was found to significantly reduce levels of elastically scattered light without sacrifice of optical throughput; Raman scattering at frequency shifts within 200 cm−1 of the Rayleigh line could be observed, and the transmittance at shifts >1000 cm−1 was ≥80%. The Rayleighline rejection efficiencies for the filters in this study are modest (102–103) compared with those for filters employed in the visible region; but they can be easily boosted by increasing the chromophore concentration or filter pathlength with a trade-off of throughput for Raman scattering at small wavenumber shifts.

Resolution of transient excited-state Raman scattering from ground-state and solvent bands is a challenging spectroscopic measurement since excited-state spectral features are often of low intensity, overlapping the dominant ground-state and solvent bands. The Raman spectra of these intermediates can be resolved, however, by acquiring time-resolved data and using multidimensional data analysis methods. In the absence of a physical model describing the kinetic behavior of a reaction, resolution of the pure-component spectra from these data can be accomplished using self-modeling curve resolution, a factor analysis technique that relies on the correlation in the data along a changing composition dimension to resolve the component spectra. A two-laser UV pump-probe resonance-enhanced Raman instrument was utilized to monitor the kinetics of amine quenching of excited-triplet states of benzophenone. The formation and decay of transient intermediates were monitored over time, from 15 ns to 100 micros. Factor analysis of the time-resolved spectral data identified three significant components in the data. The time-resolved intensities at each Raman wavenumber shift were projected onto the three significant eigenvectors, and least-squares criteria were developed to find the common plane in the space of the eigenvectors that includes the observed data. Within that plane, the three pure-component spectra were resolved using geometric criteria of convex hull analysis. The resolved spectra were found to arise from benzophenone excited-triplet states, diphenylketyl radicals, and the solvent and ground-state benzophenone.

Analysis of transient excited-state Raman spectra is a challenging spectroscopic measurement since transient spectral features are often overlapped with dominant ground-state and solvent bands. In the previous manuscript, resolution of component Raman spectra from the time-resolved amine quenching of excited-triplet benzophenone was accomplished using self-modeling curve resolution, a model-free factor analysis technique that relies on correlation in the data along a changing composition dimension. The results are consistent with the production of diphenylketyl radicals by H-atom abstraction from the amine and subsequent free-radical decay by recombination reactions. A kinetic model for this chemistry is developed in the present work, based on the observed Raman scattering data and the structures of product species confirmed by mass spectral analysis. The model is applied to the analysis of the time-dependent Raman scattering data using multidimensional least-squares methods, and it yielded well-resolved spectra of benzophenone excited-triplet states, diphenyl ketyl radical, and the solvent and ground-state precursors. The best-fit kinetic parameters agree well with the time-dependent triplet-state and ketyl-radical concentration profiles.