|
|
|
![]() |
Fig. 1: Adatom processes (sort red and blue) during heterogeneous condensation. |
Our actual research is focused on the basic adatom processes of diffusion and capture during heterogeneous condensation of metal and alloy particles from the vapour phase on substrates, e.g. of different material. Deposition experiments are performed in high vacuum and in ultrahigh vacuum environment. Nucleation densities, particle shapes and orientations, and compositions are quantitatively analyzed by transmission electron microscopy (TEM), electron diffraction (TED), and x-ray spectroscopy (EDX), respectively. In particular, a specially modified TEM allows us to perform deposition and reaction experiments in situ, such that we can observe and record kinetic processes directly and in real time (in-situ TEM).
An illustrative example is the nucleation and growth of gold on crystalline graphite (HOPG). At room temperature, particles grow with flat, dendritic shapes, with initial thickness of about 1 nm. Lateral growth of neighbouring aggregates is governed by competitive capture of gold atoms diffusing on the substrate, while the integrated contribution of direct impingement from the vapour beam is initially not significant, despite the rapidly increasing substrate coverage. In contrast, at elevated temperatures, more compact crystallites grow with pronounced (111) and (100) facets, and competitive capture does not play a role at this stage. It was also found that condensation of gold on graphite is far from complete at the early stages. At room temperature, the initial condensation coefficient is less than 0.05, and below 0.001 at above 300 oC. Thus, the adatom stay time on the substrate is initially limited by re-evaporation, but only later by capture.
The kinetics of aggregation is simulated on the basis of a model of adatom diffusion and capture, which, in fact, yields mean walk lengths of Au adatoms on graphite before desorption of about 400 nm at room temperature, which is well within the range of neighbouring aggregates, while this length is reduced to only about 4 nm at 350 oC, where particle distances are typically much larger at the early stages of growth. Detailed analysis of this data allowed us to estimate the energy of adatom adsorption on crystalline graphite to below 0.64 eV, and of surface diffusion to below 0.24 eV, respectively. These relatively low values reflect the rather weak bonding and high mobility of the gold atoms on the graphite surface.
Quite another case is palladium, which condenses virtually complete on graphite up to about 500 oC, indicating an adsorption energy well above 1.8 eV. This has interesting consequences for the deposition of a Pd-Au alloy, as the composition of the nuclei differs dramatically from that of the vapour beam, and, moreover, varies during growth up to a continuous film. In contrast, on amorphous carbon (a-C) substrates, the diffusion behaviour of Au and Pd shows peculiar differences, such that, in certain ranges of compositions, not even complete intermixing takes place in the particles. This results in the growth of two classes of particles, namely small ones, rich in Au, with high number density, and few isolated, but large crystallites rich in Pd. A similar behaviour has also been observed for Pd-Pt and Ag-Au alloy deposits on a-C, but not on crystalline graphite. Possible reasons are the differences in the metal-substrate binding energies, and/or the porous structure of a-C.
Nucleation and growth can strongly be influenced by the presence of active gases (e.g. O2, H2, CO) before and/or during deposition ("surfactants", "interfactants"). Pd, for example, is very sensitive in this respect. In a vacuum of around 10-8mbar, Pd particles grow on HOPG with more or less rounded shapes, and not very well oriented, even at elevated temperatures. In contrast, a partial pressure of oxygen of 5x10-6mbar during deposition causes almost perfect epitaxy, and pronounced facetting with mainly (111) and (100) planes (these are the planes with the lowest surface energies), in addition to a high mobility of even large crystallites.
![]() |
Fig. 2: Growth of Palladium crystallites on crystalline graphite at 480 oC under 5*10-6 mbar oxygen. The frame size is 330 nm x 290 nm. |
This is shown in the sequence of video images, recorded during growth in the TEM. The frame size is 330 nm x 290 nm. Most particles assume flat, facetted shapes at the nucleation stage, and exhibit rather fast lateral growth. Some particles exhibit peculiar, three-dimensional shapes from the beginning. Coalescence results in more rounded shapes, which become again facetted during further growth. It is known that oxygen adsorbs dissociatively on (rough) Pd surfaces (see also in the section "Catalysis"). Thus, the influence of atomic oxygen on surface energies and on the metal-support interaction is currently investigated. In general, it is thought that surfactants or interfactants may help to achieve deposits with extreme characteristics for special applications, for example, to grow a film of only a few atomic layers, which is also atomically flat. The opposite effect may also occur. As an example, the presence of an elevated partial pressure of CO during deposition of Pd particles on SiO2 has been found to result in a very low nucleation density, and growth of few, but large particles with rounded shapes.
Last modified: 27/11/2001