Core-shell doping of nanocrystals

A barrier to the widespread utilization of nanocrystal (NC) devices lies on the relatively poor electrical performance of nanoparticulate solids. The main challenges are to achieve effective doping and strong electronic connectivity across NCs. Dopant segregation, quantum confinement and incomplete screening are usually held responsible for difficulties in dopant introduction and activation. Routes to overcome these problems may involve shaping the chemistry of the NC surface to introduce a chemical barrier for dopant out-diffusion, or to modify the screening properties of the NCs to tune electrical levels. For instance, Ge has a lower valence band than Si, so that donors in a Si-NC core with a Ge shell should dump their electrons to the Ge-rich surface. This involves core-shell structuring of NCs. Although many of the results are yet to be obtained, important milestones were already attained during the visit (see below). With the established collaboration, we will investigate the electronic doping of group-IV and II-VI nanocrystals by means of density functional calculations. Core and shell atoms of the NCs will be replaced in NCs made of thousands of atoms and with sizes that span the values in real experiments. Most of the analisys will be carried out by calculating the local density of states, as well as ionization energies and electron affinities of doped NCs.


During the visit we carried out the following tasks:

1. Development of a code to produce core-shell NC structures, to saturate their surfaces with hydrogen (for group IV NCs) or with charge-compensated H atoms (for group II-VI NCs). The shape of the resulting NCs follows Wulff's construction rule for it depends on the relative surface energies of the facets. The code also allows for the pre-relaxation of the NCs with help of a valence-force field interatomic potential.

2. Development of a code to produce the local density of states (LDOS) out of the Mulliken bond population data from the AIMPRO code. The later is the density function code that we use to calculate the electronic structure of doped nanocrystals.

3. We started by producing small Ge-Si core-shell NCs (500-1000 atoms) and compared their respective LDOS for doped and undoped NCs. The dopant was phosphorous in a Ge core and its donor electron is expected to be transferred to the Si shell. For the smallest NCs we did not find such effect, but for the larger NCs we found a resonance between the P level and the Si LUMO located at the shell. This effect should be stronger for larger NCs and we will pursue that route. Another important route currently under investigation is boron doping in a Si core. The Si/Ge valence band has a larger misalignment than the conduction band, and charge transfer is expected to be stronger in this case.


Both groups established a common research area of interest that in the mid-term will lead to high-impact publications and to the joint application of funds for bilateral projects involving both groups.


This visit was also supported by the Portuguese Science Fundation (FCT) through Project No. PEst-C/CTM/LA0025/2011.

1 October 2013