Scanning tunneling microscopy and spectroscopy of molybdenum disulphide

Citation

Kramar, John Adam (1990) Scanning tunneling microscopy and spectroscopy of molybdenum disulphide. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/Q36A-AW73. https://resolver.caltech.edu/CaltechETD:etd-06132007-103520

Abstract

Scanning tunneling microscopy (STM) is a recently developed surface analysis technique that is capable of atomic resolution imaging in real space. In STM, a sharp electrically conducting tip is brought near the sample and a tunneling current is established between the two. For topographical imaging, the tip is then raster-scanned over the surface while a feedback control system maintains constant current by adjusting the position of the tip in the surface normal. The trajectory that the tip follows is displayed as the surface topography. The unique geometry of the STM has also allowed the electronic nature of surfaces to be probed with unprecedented spatial resolution. This can be implemented, for example, by posing the probe tip over a specific surface location and examining the current-versus-voltage characteristics. Careful consideration must be given in the design of an STM system. The crucial elements of vibration isolation and microscope rigidity must be optimized within the constraints of allowing coarse positioning of the tip and sample and permitting high-resolution scanning. A stable feedback control system must also be designed with flexibility to allow for different operating conditions. We have built an ultrahigh vacuum (UHV) STM that is similar to the familiar pocket STM design. The UHV system includes a separate sample preparation chamber and vacuum-transfer load lock to facilitate in studies of clean, carefully prepared surfaces. The instrument is interfaced with a microprocessor for control of scanning, data acquisition, and coarse tip-sample approach and positioning. A high-resolution graphics monitor is also included for displaying the topographic images and the current-voltage spectra during acquisition and for reviewing previously stored images. Studies of the basal cleavage plane of MoS2 have been performed with this instrument. Large area images up to 360 x 360 nm reveal a high degree of variability in surface morphology, ranging from atomically smooth planes, to islands or mounds ranging from 1 to 10 nm in diameter, to areas of complete surface roughness. Many unusual imaging phenomena were also observed in these scans, including bias-dependent images and surface modifications that were due to tip-sample interactions. Atomic-resolution images revealing the trigonal symmetry of the surface plane were obtained in both the constant-current and current-imaging modes on the smooth areas of the surface. Two distinct sites can be seen, corresponding to the known molybdenum and sulfur atomic positions. A simple description of the distance dependence of tunneling between the STM tip and an ideal semiconductor surface (no surface Fermi level pinning) is presented, based on conventional metal-insulator-semiconductor (MIS) theories. The current conduction mechanism involves thermionic emission over the semiconductor diffusion-potential barrier, which is a decreasing function of the tip-sample separation, followed by tunneling through the vacuum gap. The competition between the decreased vacuum-tunneling probability and the increased carrier population at the semiconductor surface for increasing separation gives rise to a predicted peak in the I-s curves at small separations, and a lowering of the apparent tunneling barrier height out to separations of more than 1 nm. The normally rectified current-voltage characteristics are also found to be a function of the tip-sample spacing, showing a weakening and then a reversal of rectification as the separation is increased. These predicted effects are substantiated by means of a detailed numerical calculation for the passivated n-type Si(111) surface. The current-voltage spectroscopy of MoS2, which is expected to behave as an ideal semiconductor, was examined. The surface-averaged spectra show a high degree of variability, with different data sets showing rectification in opposite polarities for the same physical tip and sample. The results are shown to be qualitatively different from known mechanisms for rectification in STM, including tip-curvature-induced field gradient effects and effects that are due to the separation dependence of rectification in ideal MIS structures. The results are best explained as doping inhomogeneities in our mineralogical samples.

Thesis Files