About the PI
John Head received his B.Sc. from University College, London in 1973 and his Ph.D. from the University of British Columbia in 1979. He then held postdoctoral positions at the University of Arkansas, the University of Guelph and the Quantum Theory Project at the University of Florida. He joined the faculty at Hawaii in 1985. His research interests are in theoretical chemistry applied to understanding the chemical interactions taking place at surfaces. More recently he has been developing theoretical methods for investigating nanocluster materials
- McBride, K. L.; Head, J. D. DFT investigation of MoS2 nanoclusters used as desulfurization catalysts Int. J. Quant. Chem. 2009, 109, 3570.
- Shiraishi, Yukihide; Robinson, David; Ge, Yingbin; Head, John D. Low-Energy Structures of Ligand Passivated Si Nanoclusters: Theoretical Investigation of Si2L4 and Si10L16 (L = H, CH3, OH, and F). Journal of Physical Chemistry C 2008, 112, 1819-1824.
- Yingbin Ge and John D. Head Global Optimization of H-Passivated Si Clusters at the Ab Initio Level via the GAM1 Semiempirical Method. J. Phys. Chem. B 2004, 108, 6025.
Theoretical Investigation of Silicon Nanoparticles
Silicon is an inexpensive and environmentally friendly semiconductor making it the central element used in most modern electronic devices. Currently there is tremendous interest in developing a silicon based optical device which could be directly integrated with silicon based microelectronic circuits. Unfortunately, bulk crystalline Si has inefficient light emitting capabilities due to it being an indirect gap semiconductor. In the early 1990’s porous Si, fabricated by electrochemically etching Si wafers in HF, was found to exhibit a surprisingly bright luminescence. Chemical characterization showed porous Si to essentially correspond to an array of Si nanocrystallites with dimensions of order of a few nanometers. Further studies have now synthesized ligand passivated Si nanoparticles which also exhibit bright luminescence. The nanometer dimensions of these particles suggest the bright luminescence and emitted wavelengths are a consequence of quantum confinement effects although the detailed origin of the emission is still being debated by researchers. The main emphasis of the research in my group is to theoretically investigate the geometric structures of the atoms composing passivated Si nanoparticles using quantum chemistry calculations. Calculations are needed because experimentally it is very difficult to elucidate the exact stoichiometry, let alone the structure, for a single nanoparticle. Using our theoretically determined atomic arrangements, we perform electronic structure calculations to better understand the origin of the optical properties in passivated Si nanoparticles. The overall goal is to perform quantum chemistry calculations which can then be used as a guide to tailoring the passivated Si nanoparticle’s optical properties for use in a real practical optical device.
Nanoparticle Structure Determination using Global Optimization Methods
These days most quantum chemistry computer programs, such as Gaussian or GAMESS, include local energy optimization routines for finding the equilibrium structures of molecules. At an equilibrium structure the energy E of a molecule has the derivative dE/dq=0 for each of the 3N nuclear coordinates q. The different possible stable isomers for a particular stoichiometry correspond to different local minima on the molecular potential energy hypersurface E(q). Chemical reactions where a reactant structure R changes to a product structure P can be interpreted in terms of moving from the R local minimum to the P local minimum via some transition state structure on the potential energy hypersurface. The most stable isomer for a particular stoichiometry is the local minima with lowest molecular energy E and is called the global minimum structure. Optimization routines in quantum chemistry codes usually find the local minimum (and equilibrium structure) associated with the catchment region of the starting guess structure and do not find the structure’s global minimum. This is a good thing because it enables quantum chemistry calculations to determine the structure and properties for different species present in a chemical reaction.
However, in passivated Si nanoclusters, such as SixHy, the number of local minima grow exponentially with increasing nanocluster sizes and we need to know what are the possible low energy structures for SixHy. Thus in our lab we have been developing cluster global optimization techniques to find the lowest energy structure for passivated Si nanoclusters. Our global optimization strategy is based on a genetic algorithm (GA) which is inspired by the Darwinian evolution process. The GA works by randomly selecting and mating the more fit individuals in a generation to produce the next generation of offspring, where the fitness is some measure of the energetic stability for an individual cluster structure. The global minimum is eventually located because some of the new cluster conformations created by the GA have lower energies than the structures in previous generations. A good mating operator causes good structural features in a cluster to be passed to the next generation while maintaining structural diversity in the overall population. Genetic algorithms are increasingly being used in a number of global optimization problems in chemistry ranging from crystal structure prediction and protein folding in biomolecules to parameter development for empirical and semi-empirical quantum mechanical calculations. Most of the cluster global optimization work so far has been on clusters with unpassivated surfaces since this avoids needing to develop special theoretical techniques to treat the surface atoms. Our major accomplishment has been to develop a GA which can be applied to the covalent networks found in SixHy clusters, where so far we have performed calculations with L = H and F.
Some of the globally optimized SixHy clusters we have obtained so far are shown below:
Figure 1. Ab Initio global minima determined for (a) Si10H16, (b) Si10H14, © Si14H20, (d) Si14H18, (e) Si18H24 and (f) Si18H22.
These results show that when the Si cluster core is passivated with enough H atoms, as in Si10H16, Si14H20 and Si18H24, the core adopts a bulk Si like structure. However, in the clusters Si10H14, Si14H18 and Si18H22, with 2 less H atoms the global minima have very different Si core structures.
In the SixHy global optimization studies the H atom serves as a prototypical passivating ligand which simplifies performing the calculations. Recently we have performed global optimization calculations with F as the passivating ligand. As illustrated below we find that specific SixLy stoichiometries have quite different lowest energy Si core structures when passivated by F atoms.
Figure 2. Ab Initio global minima for Si7L14, Si8L14, Si10L16 and Si10L14 where L is H or F in the right and left columns.
In the SixHy clusters, we find the H atoms tend to spread out evenly over the Si atoms, whereas for SixFy the F atoms favor forming SiF3 groups. We are currently developing strategies to globally optimize Si nanoclusters with OH, CH3 and OCH3 as the passivating ligand L.
Other Research Interests
I am also interested in performing quantum chemistry calculations related to adsorption of molecules on surfaces. The usual approach is to model the extended surface by a cluster of atoms with the same geometric arrangement of atoms at the active site of interest on the extended surface. We then perform quantum chemistry calculations using the GAMESS or Gaussian computer packages to investigate various properties such as the geometric structure at the preferred adsorption site for an adsorbate, the strength of the chemisorption interaction and identify any characteristic vibrational information.