Growth model for metal films on oxide surfaces: Cu on ZnO(0001)-O

Abstract

The structural and electronic properties of Cu films vapor deposited on the oxygen terminated ZnO(OOOl)-0 surface at 130 K have been characterized using x-ray photoemission spectroscopy (XPS), He+ -ion-scattering spectroscopy, low-energy electron diffraction work-function and band-bending measurements, angular-resolved XPS, and CO and H 2 0 chemisorption. These results show that Cu is cationic at tiny coverages, but becomes nearly neutral at coverages beyond a few percent. The Cu clusters into two-dimensional (2D) metallic islands at these coverages. Further deposition of Cu leads to spreading of these 2D islands without forming thicker layers, until about 50% of the surface is covered. Thereafter, these Cu islands grow thicker without filling the gaps between the islands except at a rate much slower than the rate at which Cu is deposited into these clean spaces. The annealing behavior of these films has also been studied between 130 and 850 K. These results show that the Cu has a tendency to cluster into thick 3D islands which only cover a small fraction of the surface. We present a model here based on the energetics of the system which readily explains ~he apparent contradiction between this tendency for JD clustering, and the dynamical effect which leads to efficient wetting for coverages up to f monolayer at low temperatures. This model shows that a large fraction of the surface can first be covered by a 2D film even when the metal's self-adsorption energy significantly exceeds its adsorption energy on the oxide, provided the difference in these energies does not exceed the energy of 2D evaporation from kinks onto terraces. This model helps to explain a variety of confusing results in the growth of metal films on oxide surfaces.

Atomistic Theory of Ostwald Ripening and Disintegration of Supported Metal Particles under Reaction Conditions

“…This is likely when the metal− reactant bond energy exceeds the difference in energy between the metal−metal and metal−support bonds. 48,49 Since a fact of this could be used to regenerate or redisperse the low surface area catalysts, the formation of the metal−reactant complexes should be maximized. A rigorous thermodynamic study on this topic is described here.…”

“…It has also been suggested that reactants could change the wetting behavior of metal particles, causing them to spread out on supports when the adsorbate−metal bond energy exceeds the difference in energy between the metal−metal and metal−support bonding. 48,49 Moreover, the strong interaction between adsorbate and metal particle could weaken the metal−metal bond, 50 which would facilitate the detachment of the metal adatoms from small particles, and eventually promote sintering and disintegration.…”

Formation and morphological evolution of self-similar 3D nanostructures on weakly interacting substrates

“…Currently, the most detailed atomistic * kostas.sarakinos@liu.se description of far-from-equilibrium 3D island formation is based on homoepitaxial systems in which 3D islands (mounds) form by deposition onto existing small islands, followed by atomic-step descent limited by the Ehlrich-Schwöbel barrier [15][16][17][18][19]. However, for weakly interacting film/substrate systems-including Ag/SiO 2 [20][21][22][23][24], Pd/TiO 2 [25], Cu/ZnO [26,27], and Dy/graphene [4,28]-3D islands develop before the initially formed one-atom-high islands are large enough to efficiently capture vapor-phase deposition flux. Moreover, 3D island formation is also known to occur in the absence of deposition flux due to surface restructuring via dewetting [29].…”

“…A mechanism that can explain continued growth, beyond the second atomic layer, of 3D islands on weakly interacting substrates is assisted up-stepping [26,27,35], which assumes a reduction in the step ascent barrier, when adjacent layers are separated horizontally by only one atom, due to an attractive force from the upper step edge (see, for example, position x c in Fig. 1(b), for which the ascent barrier E * 1→2 is smaller than the corresponding value E 1→2 at position x b , but larger than E s→1 at x a ).…”

Controlling the growth of Bi(110) and Bi(111) films on an insulating substrate

“…It is well-known that metals on semiconductors and oxides usually show 3D growth modes [39] instead of atomically smooth (2D) films. However, this problem can be overcome by use of deposition at low temperatures [40,41] or surfactant-mediated growth [42,43], as it modifies the film kinetics.…”