Semiconductor nanowires Research Papers - Academia.edu (original) (raw)

"The sintering behavior of the interface between Al :Si(l%) alloy and polycrystalline Si (poly-Si) was studied as a function of the poly-Si implantation dose by combining RBS, SEM, TEM and X-ray microanalysis. Two different N-dopants were used: arsenic and phosphorus. The dopants were implanted in the poly-Si layer and thermal annealing was used to obtain dopant segregation towards the poly-Si interfacea.
After sir&ring, two main effects were detected: (1) Al-Si eutectic phase precipitates and Si crystallites are formed at the interface. (2) The density of precipitates is a function of the implantation dose. For doses above 1 x lOI5 at./cm2. segregated arsenic and phosphorus are found to completely inhibit this precipitation process, provided that the segregation peak of the dopant profile is preserved before metallization.
Several conclusions can be drawn: for surface concentrations higher than 8~10’~ at./cm3, arsenic and phosphorus inhibit the precipitation of the Al-Si eutectic phase, and thus inhibit interactions between the films at the interface. Moreover, argon gas, usedfor sputtering deposition of aluminum, segregated at the poly-Si/Al: Si(l%) interface and may also inhibit the metal-semiconductor interdiffusion.""

NOTE: As the first and principal author of this journal article, I need to add a comment about the atomistic model I proposed in Fig.2 of this paper, to explain the anomalously fast diffusion of metals in semiconductors at low temperatures. I expanded on it in my PhD thesis, but is not quite stated in the present paper:

So why would metal atoms like Au and Si so easily go in solution and diffuse in Si, SiO2, glass, etc...?

It turns out this phenomenon has been observed for decades. It was first reported in the articles cited in this paper by Kobayahsi and Hiraki, back in 1979 and 1982 , for in situ deposition of metallic layers on silicon (100) in ultra-high-vacuum.
They also measured by X-ray Photoelectron Spectroscopy (XPS) how electronic bonding changes as one deposits metals atom-by-atom, in increasingly thicker layers of atomic clusters, just as is done nowadays in nano-wires and quantum dots.

As one, then two, then three, then four monolayers of metallic atoms are deposited on CLEAN Si, the electronic bonding of the metals to the silicon surface first mimicks that
of Si(100). This is WHY metals such as Al and Au CAN be grown epitaxially on Si despite the lattice mismatch between close-packed metals and the open lattice structure of semiconductors like Si and GaAS. So both Aluminum and Gold can bond as an ordered sheet of atoms on the surface. The first ordered hetero-epitaxial metal layer skips one silicon atomic row about every 4 or 5 rows. This enables for the metal atoms align themselves in a commensurate manner with the cubic symmetry of the Si(100) surface atoms.

When about the fifth monolayer is deposited, electronic delocalization start occurring as they is now a longer range metallic environment in the film, atoms in the core of the film can be now surrounded by 1st, 2nd and 3rd neighbors metallic films, and start screening the effect of the Silicon.
Remember now how many atoms are in cluster/dot or what nano-wires cross-sections because the size of the clusters matters to determine the electronic bonding inside the metallic nano-cluster and at the nano-cluster/silicon or GaAs interface.

Around 4-5 monolayers of thickness, Kobayashi measured that, the energy of the XPS emission of electrons from Au changes, and that the energy of electrons emitted from the Au layer becomes closer to that from an electron emitted from an metallically bonded atoms as opposed to a more covalent bond.

Once a metallic layer reaches a thickness where the sea of electrons from metallic bonds take over, the electronic bond start to exhibit a true metallic character. Electronic become delocalized and free to move, and their "atoms" shrink to their ionic core. The effect is especially dramatic with Gold, with Z= 79, thus 79 electrons, of which the majority are outer shell electrons as opposed to inner shell electrons.

My PhD thesis [reference 13] was metal-semiconductor interdiffusion at low temperatures. I proposed in it his model, published here in short form, and shown in Fig.2. Fig.2. describes in a cartoon-form abut my then very poor scientific English gives far from a clear explanation.