UC Davis

Epitaxial growth of metal structures on semiconductor or other metal surfaces has been studied quite extensively in surface physics and materials science due to its potential applications towards nanotechnology as well as for fundamental research purposes. In particular, the Pb/Ge(111) system, as well as the Pb/Si(111) system, has gained considerable attention due to its being a prototypical 2D metal-semiconductor interface with an abrupt boundary between the overlayer and substrate. The work done in this dissertation was inspired by previous work done on Pb/Si(111) at low temperatures by Tringides et al. in which Pb, upon reaching a certain critical coverage, exhibited explosive nucleation of islands with a certain height selection as well as unusually fast diffusion rate during growth, a rare phenomenon called collective diffusion. Since silicon and germanium are both Group IV semiconductors with identical crystal structures, this study sought to perform Pb deposition on Ge(111) to see how the system would behave (diffusion rate, temperature dependence, etc.) under similar conditions. Low Energy Electron Microscopy (LEEM) was the primary technique used to study the evolution of the Pb/Ge(111) system in real time, with some supporting data obtained from Scanning Tunneling Microscopy. In the first part of this dissertation, defects on the clean faces of germanium were studied under various cleaning parameters to determine a better set of parameters beyond the accepted ones to more consistently produce surfaces suitable for experiments. It was found that performing multiple sequences of sputter-anneal cycles using different Ar+ ion bombardment energies rather than a single energy resulted in eliminating more unwanted defects by targeting specific features at each bombardment energy. Next, Pb was deposited on Ge(111) over multiple experiments, with each experiment performed at a fixed temperature between -30°C and room temperature. Below saturation coverages, Pb was found to form a single, amorphous wetting layer without any remarkable properties. Upon reaching a critical coverage, measured to be 1.33±0.07ML with respect to Ge(111), an explosive nucleation of Pb islands was observed. The average size of the islands and the island number density showed a relationship with the temperature, with lower temperatures corresponding with smaller average island size and larger island number density. Quantitative analysis of the islands during the very early stages of nucleation revealed that each island formed by the movement of hundreds of thousands of atoms over distances on the order of a hundred nanometers in just a few seconds. It was found that part of the island nucleation was fueled by relieving compression in the wetting layer, as the 1.33ML critical coverage amounted to about 2% compression of the Pb atoms beyond the bulk Pb(111) value. Further analysis of the island growth with additional deposition beyond the nucleation phase was performed with assistance from our collaborators at the University of Central Florida (V. Stroup, T. Panagiotakopoulos, A. Childs, D. Le, T.S. Rahman), and the island growth rate was found to be linear with coverage, which is the rate for collective diffusion as compared to the t1/2 dependence of classical diffusion. Our collaborators also provided simulations of chemical potential and the binding sites of the Pb/Ge(111) system as theoretical explanations for the experimental observations. To verify the heights of the Pb islands, quantum size contrast in the LEEM intensity of the islands was measured with respect to the start voltage of the electron beam used to illuminate the surface (IV curve). A Kronig-Penney model was used to simulate islands of various thicknesses, from which the reflectivity coefficient was calculated for each electron energy to produce simulated IV curves. The simulated curves were then matched with experimental IV curves until a proper fit was found. Using this technique, an island grown at -24°C was found to have a height of 7 Pb layers, while an island grown at +3°C was found to be a hybrid of 10 and 11 layers due to its growing over a step edge. Lastly, Pb was deposited on the other germanium faces (100 and 110) in an effort to see how the different surface symmetry would affect island growth under similar conditions. Pb/Ge(100) exhibited very similar behavior, with explosive nucleation of Pb islands upon reaching a critical coverage, which was unsurprising given the highly-symmetric surface structure of Ge(100), similar to the three-fold symmetry of the Ge(111) surface. The critical coverage appeared to be dependent on the density of surface defects on the substrate, with the critical coverage measured up to 2.53±0.03ML with respect to Ge(100) for one experiment, which is an exceptionally high value compared to Pb/Ge(111). Pb/Ge(110) was found to produce thin, 1-dimensional islands that grew along the [11 ̅0] axis and did not exhibit the same height selection as the other germanium faces. This is similar Ag/Ge(110), which was found to exhibit 1-dimensional growth due to the substrate’s rectangular symmetry. For Pb, the critical coverage was measured to be 1.06±0.06ML with respect to Ge(110), and the islands grew linearly with time, exhibiting collective diffusion. For both Pb/Ge(100) and Pb/Ge(110), classical, 3-dimensional islands were observed coexisting with the nonclassical islands in a narrow range of temperatures between 0°C and room temperature. 0°C was found to be the boundary between the classical regimes in which thermal diffusion is the dominating force, and nonclassical regimes, with only the nonclassical islands nucleating below this temperature, both types of islands nucleating in the coexistence region, and only the classical islands nucleating above room temperature. The classical Pb/Ge(110) islands formed in triangular and trapezoidal shapes that showed preference in how they oriented themselves with respect to the Ge(110) surface, and this behavior was explained by the different stability of the island sides, due to different arrangements of atoms forming the sides, compared to the atoms on the substrate.