Speaker
Description
The air-water interface is one of the most prevalent interfaces on Earth and is central to a vast range of natural and industrial processes. The sheer presence of the interface induces significant changes in its properties such as density and dielectric function, as well as the distribution of molecular orientations and interconnectivity of the H-bond network. The widespread importance of these perturbations has led the air-water interface to be the subject of intensive research efforts spanning several decades. Nevertheless, many of its fundamental aspects, such as the thickness of its structural anisotropy, remain contentious, with no direct experimental measurements being available. A key technique in these interfacial investigations is sum-frequency generation (SFG) spectroscopy which gains insight into molecular structure and H-bonding through the characteristic vibrational line-shapes. Nevertheless, SFG only yields spatially integrated responses, meaning no depth information is accessible, and the interpretation of any resulting spectra is highly challenging. For this reason, experimental results are often compared to predictions from molecular dynamics (MD) simulations which yield a depth-dependent view specific structural motifs at the interface. Such comparisons, however, necessitate the agreement between the observed and simulated spectra. While this is generally the case for the O-H stretching vibrations, substantial differences are observed for the H-O-H bending mode. This disagreement raises critical questions over much of our current understanding about the structure of aqueous interfaces.
In this work, we directly address these questions using a novel depth-resolved vibrational spectroscopy which involves the simultaneous measurement of phase-resolved sum- and difference frequency generation (SFG and DFG) signals, enabling precise depth-profiling at the sub-nanometre scale. Firstly, by probing the O-H stretching region, we obtain a direct measure of the decay length of the structural anisotropy, comparing the results to predictions from MD simulations. Secondly, by probing the H-O-H bending region, we show that the observed discrepancy between experiment and simulation arises from a dominant bulk quadrupolar contribution that is not included in simulations. Through further measurements at charged interfaces, the experimental spectrum is broken down into its different components and the interfacial dipolar signal predicted by theory is retrieved, thus consolidating many years of simulations on aqueous interfaces. Analysis of the purely dipolar contribution then gives new insight into the dominant factors governing the molecular structure at the interface.
