Periodic arrays of Ta nanocolumns, 200nm wide and 600nm tall, were grown by glancing angle sputter deposition onto self-assembled close-packed arrays of 260-nm-diameter silica spheres. Each sphere leads to the development of a single Ta column. As growth progresses, roughening of the column top surfaces causes branching of some columns into subcolumns. The measured fraction of branched columns fb decreases with increasing growth temperature, from 30% at 200°C to 4% at 700°C. This is attributed to the increased adatom mobility at elevated temperatures, leading to a larger average separation of growth mounds and, in turn, lower nucleation probabilities for subcolumns. Branching into 3 and 4 subcolumns exhibits probabilities proportional to fb2 and fb3, respectively. A fit of the data with a simple nucleation model provides an effective activation energy for Ta surface diffusion of 2.0eV.

No abstract

Arrays of Ta nanorods were grown by glancing angle deposition ͑GLAD͒ onto honeycomb Cr nanodot patterns that were evaporated onto Si substrates through a self-assembled monolayer of SiO 2 nanospheres. Statistical size analyses from arrays of rods with variable average length l and width w, with 210 nmഛ l ഛ 650 nm and 109 nmഛ w ഛ 304 nm, show that the distribution in w broadens with increasing l and decreasing w, but remains approximately constant with a fixed l / w ratio. This is attributed to an intercolumnar growth competition that exacerbates nanorod size fluctuations but scales with rod size. These results suggest that the overall nanostructure shapes during low-temperature GLAD are independent of material-specific length scales and are, therefore, completely controlled by the geometric shadowing.

No abstract

No abstract