Resonant Raman spectroscopy on metal surfaces and interfaces

The project aims to explore the physical mechanisms of resonant Raman spectroscopy and its potential for vibrational spectroscopy of metal surfaces and interfaces. In Raman spec- troscopy the frequency of light scattered inelastically from the sample is analysed. The energy difference of incident and scattered light allows to draw conclusions on the elementary process which has been excited, e.g., a vibrational mode (phonon). The experiments in the frame of the project will be done on copper single-crystals, especially on Cu(110). The surface phonons of Cu(110) were intensively studied by complementary techniques and are well understood. In the study of the lattice dynamical properties of solids, Raman spectroscopy offers the advantage of a much higher resolution compared to other techniques. In addition, the scattering signal will arise exclusively from the surface region, since for metals like copper with one atom per unit cell in the bulk no Raman-allowed optical modes exist due to symmetry reasons. Since the penetration depth of the incident light is much larger compared to the thickness of the surface layer, usually in Raman experiments the largest part of the signal arises from scattering on phonons in the volume of the crystal. In addition, the Raman intensity is smaller for metals as compared to, e.g., semiconducting materials, since in metals the lifetime of the electron-hole pairs excited by the incident light and which then interact with the crystal lattice creating a phonon is very short. The Raman scattering intensity can be strongly enhanced when the exciting energy is chosen close to an allowed electronic transition of the material (resonant Raman scattering). In order to enhance the scattering intensity, the surface state electronic transition of Cu(110) is resonantly excited. In addition to enhancement of the otherwise very small scattering probability, this also provides enhanced surface sensitivity. Preliminary experiments for pristine Cu(110) show the existence of a Raman-active surface phonon resonance with a surprisingly large scattering efficiency. Furthermore, surface phonons were observed on oxygen modified Cu(110). The research aims to address and investigate the exact scattering mechanism which leads to the pronounced surface phonon resonance on Cu(110). To this end, the influence of adsorbed species on the phonon spectrum of the pristine surface will be studied. On the one hand, small molecules like water which interact only weakly with the surface will be used as adsorbate, on the other hand, oxygen will be used which strongly interacts with Cu(110) and even leads to a reconstruction of the surface.

This study focused on investigating surface vibrations on Cu(110) metal surfaces using resonant Raman spectroscopy. Raman spectroscopy is a technique that measures the scattering of light from a substance, which can interact with vibrations, phonons, or other excitations in the system, altering the energy of the photons. By analyzing the energy difference between the incident and scattered light, valuable information about the excited elementary process, such as a vibrational mode (phonon), can be obtained. Although Raman spectroscopy can be used to measure vibrations at surfaces and interfaces, the signal is often very weak and overshadowed by the bulk signal. To overcome this limitation, resonant Raman spectroscopy is commonly employed to amplify the surface signal. This method involves selecting an excitation energy that is close to an electronic surface transition of the material, which significantly enhances the Raman intensity of a surface vibrational mode. In our initial experiments, we discovered a surface vibrational resonance on Cu(110) that produces a strong Raman signal and does not involve bulk phonons. Our project aimed to investigate the physical factors that contribute to this enhanced Raman signal and determine if it can be used for high-resolution vibrational spectroscopy on metal surfaces with and without adsorbates. To achieve this, we conducted Raman measurements on Cu(110) using different polarizations and excitation energies, ranging from 1.8 to 3 eV. These experiments revealed an increase in Raman scattering when the excitation energy matched the electronic transition between surface states on Cu(110) at 2.1 eV. We used density functional theory calculations to confirm this relationship and examine how the excitation of the surface phonon resonance affects the electronic structure of the surface. The calculations clearly showed that the observed Raman signal is a result of the coupling between this specific mode and the electronic states on the surface. Additionally, we performed lattice dynamics calculations to gain a more detailed understanding of the scattering process and the spectral characteristics of the Raman intensity. By combining the results of the Raman experiments, electronic band structure analysis, and lattice dynamics calculations, we obtained a comprehensive understanding of the interaction between surface phonons and surface-localized electronic states.