Simon D. Elliott

Chiral Shells and Achiral Cores in CdS Quantum Dots

We report and explain circular dichroism in semiconductor quantum dots. CdS nanocrystals capped with penicillamine enantiomers were prepared and found to be both highly luminescent and optically active. No new features in circular dichroism were observed as the nanocrystal grew larger. Density functional calculations reveal that penicillamine strongly distorts surface Cd, transmitting an enantiomeric structure to the surface layers and associated electronic states. The quantum dot core is found to remain undistorted and achiral.

The p-type conduction mechanism in Cu2O: a first principles study

Materials based on Cu2O are potential p-type transparent semiconducting oxides. Developing an understanding of the mechanism leading to p-type behaviour is important. An accepted origin is the formation of Cu vacancies. However, the way in which this mechanism leads to p-type properties needs to be investigated. This paper presents a first principles analysis of the origin of p-type semiconducting behaviour in Cu2O with 1.5 and 3% Cu vacancy concentrations. Plane wave density functional theory (DFT) with the Perdew-Burke Ernzerhof (PBE) exchange-correlation functional is applied. In order to investigate the applicability of DFT, we firstly show that CuO, with 50% Cu vacancies cannot be described with DFT and in order to obtain a consistent description of CuO, the DFT + U approach is applied. The resulting electronic structure is consistent with experiment, with a spin moment of 0.64 mu(B) and an indirect band gap of 1.48 eV for U = 7 eV. However, for a 3% Cu vacancy concentration in Cu2O, the DFT and DFT + U descriptions of Cu vacancies are similar, indicating that DFT is suitable for a small concentration of Cu vacancies; the formation energy of a Cu vacancy is no larger than 1.7 eV. Formation of Cu vacancies produces delocalised hole states with hole effective masses consistent with the semiconducting nature of Cu2O. These results demonstrate that the p-type semiconducting properties observed for Cu2O are explained by a small concentration of Cu vacancies.

Clusters of aluminium, a density functional study

Density functional calculations have been performed to study the structure and energetics of mainly neutral Aln with n up to 153. Clusters up to n=15 have been investigated systematically, in part by simulated annealing techniques. No pronounced islands of stability are found in this range although Al7 and Al13 are of special importance. The treatment of selected larger clusters with different packings and cluster shapes leads to the following conclusions. Icosahedral packing is favoured only for n=13; around n=55, decahedral packing is most stable, whereas fcc is definitely preferred, energetically, for n>80. Among clusters of comparable size and packing those with dominantly (111) surfaces are found to be most stable. These trends are rationalized by considering the average number of next neighbours and the internal strain arising from a mismatch of bond distances. Extrapolation of the computed cluster binding energies yields a cohesive energy of bulk Al of 3.3–3.4 eV, in close agreement with experiment (3.36 eV).

Simulating the atomic layer deposition of alumina from first principles

Atomic Layer Deposition (ALD) is a chemical vapour deposition technique that is suitable for the controlled growth of thin, conformal oxide films. Insulating layers of alumina (Al2O3) are fabricated by ALD for electroluminescent flat-screen-displays, memory capacitors and read/write heads. Successful precursors for alumina ALD are trimethylaluminium (TMA) and water. Various models for the mechanism of alumina ALD have been proposed but definitive evidence for the surface intermediates is lacking. We therefore use density functional theory to investigate the atomic-scale structure and reactivity of bare and hydroxylated alumina surfaces with respect to TMA. The competition between dissociative chemisorption and molecular desorption is quantified. The calculations show that TMA adsorbs and dissociates at both O and OH sites on the surface, but that these lead to different ALD efficiencies. In this way we clarify how the degree of surface hydroxylation and hence the substrate temperature affects the ALD growth rate.

Atomic-scale simulation of ALD chemistry

Published papers on atomic-scale simulation of the atomic layer deposition (ALD) process are reviewed. The main topic is reaction mechanism, considering the elementary steps of precursor adsorption, ligand elimination and film densification, as well as reactions with substrates (particularly Si and SiO2) and CVD-like decomposition at the surface. Density functional theory is the first principles method generally applied to these mechanistic questions. The most popular subject for modelling is the ALD of oxides and nitrides, particularly the high-k dielectrics HfO2, ZrO2?and Al2O3, due to their importance in semiconductor processing.

Mechanism, products, and growth rate of atomic layer deposition of noble metals.

We present mechanisms for atomic layer deposition of Ru, Rh, Pd, Os, Ir, or Pt metal from homoleptic precursors and oxygen. The novel mechanistic feature is that combustion of ligands produces transient hydroxyl groups on the surface, which can undergo Brønsted-type elimination of a further ligand or water from the surface. Each ligand therefore releases one electron for reduction of the metal. The growth reaction may be described as oxide-catalyzed redox decomposition of the precursor. To validate the mechanism against experiment, we derive analytical expressions for product ratios and the growth rate in terms of saturating coverages.

Effect of Reaction Mechanism on Precursor Exposure Time in Atomic Layer Deposition of Silicon Oxide and Silicon Nitride

Atomic layer deposition (ALD) of highly conformal, silicon-based dielectric thin films has become necessary because of the continuing decrease in feature size in microelectronic devices. The ALD of oxides and nitrides is usually thought to be mechanistically similar, but plasma-enhanced ALD of silicon nitride is found to be problematic, while that of silicon oxide is straightforward. To find why, the ALD of silicon nitride and silicon oxide dielectric films was studied by applying ab initio methods to theoretical models for proposed surface reaction mechanisms. The thermodynamic energies for the elimination of functional groups from different silicon precursors reacting with simple model molecules were calculated using density functional theory (DFT), explaining the lower reactivity of precursors toward the deposition of silicon nitride relative to silicon oxide seen in experiments, but not explaining the trends between precursors. Using more realistic cluster models of amine and hydroxyl covered surfaces, the structures and energies were calculated of reaction pathways for chemisorption of different silicon precursors via functional group elimination, with more success. DFT calculations identified the initial physisorption step as crucial toward deposition and this step was thus used to predict the ALD reactivity of a range of amino-silane precursors, yielding good agreement with experiment. The retention of hydrogen within silicon nitride films but not in silicon oxide observed in FTIR spectra was accounted for by the theoretical calculations and helped verify the application of the model.

Atomistic kinetic Monte Carlo study of atomic layer deposition derived from density functional theory

To describe the atomic layer deposition (ALD) reactions of HfO2 from Hf(N(CH3)2)4 and H2O, a three‐dimensional on‐lattice kinetic Monte‐Carlo model is developed. In this model, all atomistic reaction pathways in density functional theory (DFT) are implemented as reaction events on the lattice. This contains all steps, from the early stage of adsorption of each ALD precursor, kinetics of the surface protons, interaction between the remaining precursors (steric effect), influence of remaining fragments on adsorption sites (blocking), densification of each ALD precursor, migration of each ALD precursors, and cooperation between the remaining precursors to adsorb H2O (cooperative effect). The essential chemistry of the ALD reactions depends on the local environment at the surface. The coordination number and a neighbor list are used to implement the dependencies. The validity and necessity of the proposed reaction pathways are statistically established at the mesoscale. The formation of one monolayer of precursor fragments is shown at the end of the metal pulse. Adsorption and dissociation of the H2O precursor onto that layer is described, leading to the delivery of oxygen and protons to the surface during the H2O pulse. Through these processes, the remaining precursor fragments desorb from the surface, leaving the surface with bulk‐like and OH‐terminated HfO2, ready for the next cycle. The migration of the low coordinated remaining precursor fragments is also proposed. This process introduces a slow reordering motion (crawling) at the mesoscale, leading to the smooth and conformal thin film that is characteristic of ALD.

Mechanism for zirconium oxide atomic layer deposition using bis(methylcyclopentadienyl)methoxymethyl zirconium

The mechanism for zirconium oxide atomic layer deposition using bis(methylcyclopentadienyl)methoxymethyl zirconium and H2O was examined using ab initio calculations of hydrolysis energies to predict the order of ligand loss. These predictions were tested using in situ mass spectrometric measurements which revealed that the methyl ligand, and 65% of the methylcyclopentadienyl ligands are lost during the zirconium precursor adsorption. The remaining 35% of the methylcyclopentadienyl ligands and the methoxy ligand are lost during the subsequent H2O exposure. These measurements agree very well with the predictions, demonstrating that thermodynamic calculations are a simple and accurate predictor for the reactivities of these compounds.

New rigid backbone conjugated organic polymers with large fluorescence quantum yields

The effect on optical emission of the presence of carbon–carbon triple bond linkages in conjugated aromatic copolymers is studied and leads to fluorescence quantum yields of up to 50%, substantially higher than most conventional conjugated polymers.