2018

Inelastic electron tunneling spectroscopy (IETS) combined with scanning tunneling microscopy (STM) allows the acquisition of vibrational signals at surfaces. In STM-IETS, a tunneling electron may excite a vibration, and opens an inelastic channel in parallel with the elastic one, giving rise to a change in conductivity of the STM junction. Until recently, the application of STM-IETS was limited to the localized vibrations of single atoms and molecules adsorbed on surfaces. The theory of the STM-IETS spectrum in such cases has been established. For the collective lattice dynamics, i.e., phonons, however, features of STM-IETS spectrum have not been understood well, though in principle STM-IETS should also be capable of detecting phonons. In this review, we present STM-IETS investigations for surface and interface phonons and provide a theoretical analysis. We take surface phonons on Cu(110) and interfacial phonons relevant to graphene on SiC substrate as illustrative examples. In the former, we provide a theoretical formalism about the inelastic phonon excitations by tunneling electrons based on the nonequilibrium Green's function (NEGF) technique applied to a model Hamiltonian constructed in momentum space for both electrons and phonons. In the latter case, we discuss the experimentally observed spatial dependence of the STM-IETS spectrum and link it to local excitations of interfacial phonons based on ab-initio STM-IETS simulation.

Self-assembled molecular (SAM) structures have been intensively used in molecular electronics and spintronics. However, detailed nature of the interfaces between molecular layers and extended metallic contacts used to bias the real devices remains unclear. Buried interfaces greatly restrict application of standard techniques such as Raman or scanning electron microscopies. Here we introduce low-frequency noise spectroscopy as a tool to characterize buried molecular-metal interfaces. We take advantage of vibrational heating of the molecules with incomplete contacts to the interface. Electrons, being the main spin and charge carriers propagating through the interfaces involving SAMs, interact inelastically with the nuclei and excite quantum molecular vibrations (phonons). Our detailed investigation of both conductance and conductance fluctuations in magnetic tunnel junctions with few nm Perylenetetracarboxylic dianhydride (PTCDA) allows to map vibrational heating at specific biases taking place in hot spots such as where SAM layers make unstable contact to the metallic electrodes. We follow this effect as a function of PTCDA thickness and find the highest molecular-metal order for the lowest (3-5 monolayers) barriers. Moreover, we show experimentally that the low-frequency noise depends on the relative alignment of the electrodes well beyond expectations from fluctuation-dissipation theorem. In combination with modeling, we interpret this effect as due to a magnetic-state dependent molecular vibrational heating at the interfaces driven by the spin-polarized current.

The linewidths of the $\pi$ and $\sigma$ bands originating from the electron-phonon coupling in graphene are analyzed based on model calculations and experimental angle-resolved photoemission spectroscopy (ARPES) data. We find evidence for crucial contributions to the lifetime broadening from interband scattering $\pi\rightarrow\sigma$ and $\sigma\rightarrow\pi$ respectively, driven by the out-of-plane ZA acoustic phonons. The essential features of the calculated $\sigma$ band linewidths are in agreement with recent published ARPES data [F. Mazzola et al., Phys. Rev. B. 95, 075430 (2017)] and of the $\pi$ band linewidth with ARPES data presented here.

Electronic conduction through molecular junctions depends critically on the electronic state at the anchor site, suggesting that local reactions on the electrodes may play an important role in determining the transport properties. However, single-molecule junctions have never been studied with the chemical states of the electrodes controlled down to the atomic scale. Here, we study the effect of surface adsorbates on the molecular junction conductance by using a scanning tunneling microscope (STM) combined with density functional theory (DFT) and nonequilibrium Green's function (NEGF) calculations. By vertical control of a STM tip over a phenoxy (PhO) molecule on Cu(110), we can lift and release the molecule against the tip, and thus reproducibly control a molecular junction. Using this model system, we investigate how the conductance changes as the molecule is brought to the vicinity of oxygen atoms or a hydroxyl group chemisorbed on the surface. This proximity effect of surface adsorbates on the molecular conductance is simulated by DFT-NEGF calculations.