Polarised Raman spectroscopy as a quantitative probe of interfacial molecular orientation

Abstract

Raman scattering is a ubiquitous phenomenon that can be used to great effect to study molecules near interfaces. It has traditionally been used as an analytical tool to identify materials, but by using polarised light, the degree of order within that material can be assessed simultaneously. This thesis seeks to enhance this technique by accurately quantifying interfacial molecular orientation from peak intensities in polarised Raman spectra. This requires a joint modelling and experimental approach.

The experimental system, previously developed in our group, obtains surface selectivity through total internal reflection (TIR) of an incident laser beam at the interface under investigation. The evanescent wave generated by TIR causes Raman scattering by the molecules of interest. This system enables investigation of molecular layers at solid-air, solid-liquid and solid-solid interfaces.

A numerical model is constructed to predict Raman scattering intensities based on a generalised experimental geometry, the Raman tensor of the vibrational mode under investigation and the orientation of the scattering molecule. A local field correction is implemented for incident as well as emitted radiation. The scattered intensity is calculated with Lorentz reciprocity and integration over the microscope objective that collects the Raman signal. The modelling outcomes are fitted to experimental Raman scattering intensities to deduce molecular orientation. The electrodynamic model of the scattering process is complemented with Raman tensors, polarisabilities and molecular radii obtained by ab initio computation.

The novel methodology is validated with isotropic scatterers and a supported monolayer of zinc arachidate. Analysis of Raman spectra of zinc arachidate in a contact under static load reveals a variation in alkyl chain tilt of (4.8±0.5)° per 100 MPa around (27±4)° at 500 MPa. The exact tilt angle depends on the intensity and fitting metrics used.

The model further allows quantitative interpretation of Raman spectra as well as optimisation of experimental design. Limitations as well as future applications of this approach are discussed.