A scanning angle total internal reflection (SATIR) Raman spectrometer has been developed for measuring interfacial phenomena with chemical specificity and high axial resolution perpendicular to the interface. The instrument platform is an inverted optical microscope with added automated variable angle optics to control the angle of an incident laser on a prism/sample interface. These optics include two motorized translation stages, the first containing a focusing lens and the second a variable angle galvanometer mirror. The movement of all instrument components is coordinated to ensure that the same sample location and area are probed at each angle. At angles greater than the critical angle, an evanescent wave capable of producing Raman scatter is generated in the sample. The Raman scatter is collected by a microscope objective and directed to a dispersive spectrometer and charge-coupled device detector. In addition to the collected Raman scatter, light reflected from the prism/sample interface is collected to provide calibration parameters that enable modeling the distance over which the Raman scatter is collected for depth profiling measurements. The developed instrument has an incident angle range of 25.5 degrees-75.5 degrees, with a 0.05 degrees angle resolution. Raman scatter can be collected from a ZnSe/organic interface over a range of roughly 35-180 nm. Far from the critical angle, the achieved axial resolution perpendicular to the focal plane is approximately 34 nm. This is roughly a 30-fold improvement relative to confocal Raman microscopy.

Raman spectra were collected from a 1.25 M aqueous pyridine solution, 100-nm polystyrene film or a trimethyl(phenyl)silane monolayer at a plasmon waveguide interface under total internal reflection (TIR). The plasmon waveguide resonance (PWR) interface consisted of a sapphire prism/49 to 50 nm Au/548 to 630 nm SiO(2) and a monolayer, thin film or aqueous analyte. The Raman peak area as a function of incident angle was measured using a 785-nm excitation wavelength, and was compared to the Raman peak area obtained at a sapphire or sapphire/50 nm Au interface. In contrast to measurements at a bare sapphire prism, increased surface sensitivity and signal were obtained from the PWR interface. In contrast to measurements at a bare Au film where only p-polarized incident light generates an enhanced interfacial electric field, plasmon waveguide interfaces enable excitation with orthogonal polarizations using s- or p-polarized incident light. The Raman scatter from a monolayer was recorded at the PWR interface with a signal-to-noise ratio of 5.6 when averaging 3 accumulations with 3 min acquisition times using nonresonant excitation, whereas no signal was recorded from a monolayer at the sapphire interface. The reflected light from the interface enabled the identification of the incident angle where the maximum Raman scatter was produced, and the Raman signal generated at the plasmon waveguide interface was modeled by the enhanced interfacial mean square electric field relative to the incident field. In comparison to the techniques on which this work was based (i.e., PWR spectroscopy, TIR Raman spectroscopy at the prism interface, and surface plasmon resonance (SPR) Raman spectroscopy at the prism/Au interface), chemical specificity was added to PWR spectroscopy, a signal enhancement mechanism was introduced for TIR Raman spectroscopy, and polarization control of the interfacial electric field was added to SPR Raman spectroscopy.

Total internal reflection (TIR) Raman and reflectivity spectra were collected for nonresonant analytes as a function of incident angle at sapphire or sapphire/smooth 50 nm gold interfaces using 785 nm excitation. For both interfaces, the Raman signal as a function of incident angle is well-modeled by the calculated interfacial mean square electric field (MSEF) relative to the incident field times the thickness of the layer being probed in the Raman measurement (DRS). The Raman scatter was reproducibly enhanced at the interface containing a gold film relative to the sapphire interface by a factor of 4.3−4.6 for aqueous pyridine or 2.2−3.7 for neat nitrobenzene, depending on the analyzed vibrational mode. The mechanism for the increased Raman signal is the enhanced MSEF at incident angles where propagating surface plasmons are excited in the metal film. The background from the TIR prism was reduced by 89−95% with the addition of the gold film, and the percent relative uncertainty in peak area was reduced from 15 to 1.7% for the 1347 cm−1 mode of nitrobenzene. Single monolayers of benzenethiol (S/N = 6.8) and 4-mercaptopyridine (S/N = 16.5) on gold films were measured by TIR Raman spectroscopy with 785 nm excitation (210 mW) without resonant enhancement in 1 min.