Reports: G10

46079-G10 Molecular Dynamics Simulations of Time-Dependent and Transport Processes

Traian Dumitrica, University of Minnesota

In the first second year of this project our goal was to apply the developed molecular dynamics code that allows us to perform simulations of the nanomechanical response of nano-objects (quantum dots, nanowires, and nanotubes) including in situations where nonequilibrium electronic distributions are encountered. These situations are encountered when matter is subjected to ultrafast laser pulses or to external applied voltages. This goal was accomplished previously: we amended the well-established ground-state molecular dynamics formalism, as implemented in the tight-binding molecular dynamics code Trocadero, to account for the fraction of electrons populating the excited states. This new computational capability has several distinct features. Specifically, molecular dynamics can be now performed in the presence of an external electromagnetic field, which was coupled by us to the atomic orbitals in a way that includes the wave polarization. Moreover, the resulting ion dynamics is highly accurate because it includes the correction due to the motion of atomic-orbital basis and because of transferable non-orthogonal tight-binding description of the solid.

Earlier this year, we have investigated with thigh-binding molecular dynamics the remarkable physical properties of carbon nanotubes. These properties originate in their objective atomic structure where each carbon atom sees precisely the same environment up to rotation and translation. Modulating these properties is highly desirable for various applications and systematic ways to manipulate the perfect arrangement of hexagonal rings are needed. A wealth of experimental data shows that the near sublimation thermal agitation does not necessarily destroy these structures. Instead, it can have a positive effect, especially when combining the significant random agitation of the atoms with a coherent component caused by an externally applied deformation. For example, recent experiments on superplasticity obtained that nanotubes under tensile load can undergo large elongation and thinning without abandoning their perfection. Other theoretical studies indicated that superplasticity relies on primary microscopic mechanisms, like a mass-conserving glide along a helical slip path, as well as on a nearly axial kink propagation with dimers directly breaking out of the lattice. Remarkably, each mass-conserving glide step lowers diameter and changes its index from (n,m) to (n,m-1) or (n-1,m). Plasticity under bending was also described in terms of kink motion along a helical path. What other primary transformations can be induced by external deformation on the hot CNT lattice? To address this question we considered plasticity under another fundamental type of deformation—torsion. Due to recent experimental advances, it is now possible to probe CNTs as torsional springs. The popular atomistic modeling tools are unsuitable for modeling this type of deformation. Relying on our theoretical innovations, we described the nanotube torsional response and predict the possibility of a new mass-conserving nearly axial glide. Such glide cannot be promoted by pure tension.

More recently, we have performed a study for the optical properties of Si quantum dots. Until recently, the relationship between the intrinsic symmetry and the optical response of Si quantum dots has been overlooked. Relying on the general concept of symmetry lowering, our recent calculations indicated new effective routes to modulate the optic response without affecting the stability of these tiny structures.

Advances in synthetic methods have made possible efficient production of Si dots with various sizes, shapes, and core structures. The exhibited photoluminescence shows a tremendous potential for applications in the energy area. However, an important challenge is adjusting the optic response by manipulating the electronic states around the last occupied – first empty levels. In crystals, this is usually achieved by doping but this route proves difficult at the nanoscale.

We have focused on highly symmetrical dots, sometimes called “artificial atoms”. As in atoms, high symmetry in Si dots brings electronic degeneracies and large level spacings, and enforces strict selection rules for the optical transitions between levels. Many transitions are forbidden. For instance, the energy spacing between last occupied - first empty levels is generally different from the first possible excitation. Of course, no symmetry implies no degeneracy and all transitions would be allowed.

The question addressed was whether it is possible to alter the atom-like electronic levels of such dots without considering the unlikely endohedral doping. In atoms, splitting the degenerate energy levels is usually accomplished by breaking the symmetry with the help of an external magnetic field. In Si dots, the researchers demonstrated via density functional theory calculations that symmetry lowering and level splitting could be readily accomplished in new ways, such as introducing a slight structural imperfection vis-à-vis the spherical shape, applying mechanical squeezing, and contaminating the surface with Na atoms.

The continual development of nanotechnology will provide a greater range of highly symmetrical Si quantum dots. The uncovered connection between symmetry and electronic states makes these structures very exciting for both fundamental and applied research. Higher-level calculations are under way to more precisely quantify the energy of emitted light. In optoelectronics, symmetry lowering could become a useful strategy for manipulating the optic response.