Probe-assisted Doping : Modelling and Validation
Open AccessAtomic force microscope (AFM) is widely used as a material property measurement tool which delivers high resolution topographic/tomographic images to reveal mechanical and electrical properties of materials in the nanoscale domain. The nanoscale tip utilized in the fundamental scanning mechanism of AFM can be used as an instrument of sophisticated material modification. With theoretical and empirical understandings of tip-material interaction, AFM can yield a high-resolution surface and subsurface modifications with nanoscale control.This work demonstrates localized probe-assisted doping (PAD), a modification to the technique of AFM nano-indentation, by achieving nanoscale doping of aluminum into silicon. Subsequent application of rapid thermal annealing generates controlled doping profile and nanoscale aluminum-silicon junctions. The mechanics of PAD was investigated with analytical and computational modelling using two major approaches based on the following theoretical considerations: 1) Bonded steered molecular dynamics simulations adopting Dreiding Force Fields for preliminary observation of the potential energy profile of aluminum and silicon atoms during atomic diffusion of an aluminum atom in the silicon crystal lattice and 2) Non-bonded molecular dynamics simulation adopting Lennard-Jones (L-J) and Modified Embedded Atomic Method (MEAM) potentials for nanoscale estimation of the potential energy profile, crystal deformity, and temperature-induced aluminum insertion and diffusion into the silicon crystal lattice. The simulation results were empirically validated by comparing the calculated parameters to corresponding measurements from PAD experiments. In the molecular dynamics simulations, the atomic diffusion of aluminum into the silicon crystal lattice with PAD is observed during tip indentation. The resulting potential energy profiles show identical changes in the aluminum and silicon with respect to tip motion. Furthermore, the tip indentation results in irreversible plastic deformation of the aluminum layer due to the force dispersion in the x/y directions that results in nanoscale pile-ups of aluminum atoms surrounding the contact area. A comparison between the potential energy of silicon before and after tip impact suggests the deformity of the silicon is minimized when controlling the indentation depth. The energy conversion of the tip potential energy at the contact area indicates that tip indentation generates enough thermo-kinetic energy to cause insertion and diffusion of aluminum atoms into the silicon at the nanoscale area of impact. From PAD experiments, localized doping results in changes to the electrostatic potential (CPD) with nanoscale spatial resolution. In contrast to the literature on nano-mechanical defects of silicon substrates, thermal post-processing of PAD samples electrically activate the aluminum dopants in the silicon crystals. Unfortunately, thermal activation also promoted the diffusion of the dopants and geometric distortions in the doping area. Optimization efforts show the “active” dopant concentration depended primarily on the temperature of the thermal annealing process rather than tip dwell time during PAD. In this research, we discovered the energy transition mechanism from tip potential energy to thermo-kinetic energy of diffusing aluminum atoms using the molecular dynamics simulations, and the surface potential energy changes in the PAD processed area due to thermal annealing-induced aluminum atom diffusion by using SKPM. The insertion and recombination of aluminum atoms in the silicon lattice results in the minimal crystal deformity when controlling the tip loading force, limiting the final indentation depth and optimizing the temperature of post-process thermal annealing. Furthermore, the difference between the results of simulations and the PAD process can be reduced by the following: 1) The Monte Carlo simulation with the optimal kinetic and temperature parameters should be implemented to the aluminum diffusion estimation and, 2) The empirical contact response model with the optimal mechanical parameters for the tip geometry and the elastic and plastic properties of thin film is required to improve the resolution and the reliability of the PAD technique.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.