I obtained by degree in chemistry (MChem) from the university of Manchester in 2010. Toward the end of my degree I developed an interest in the processes that occur at an atomic level. This lead to an interest in the use of compuational methods and molecular modeling methods to understand what goes on at the atomic scale. My final year project involved the use of molecular dynamics within the AMBER code to predict and understand the partition coefficient of various organic solutes under the supervision of Dr. Richard Henchman.
My research focusses on the physics and chemistry of defects, and how these defects can affect the properties of electronic devices. My recent work involves studying technologically relevant oxides and their interfaces between semiconductor materials.
Ensuring reliability of electronic devices requires an understanding of the atomic processes occuring at interfaces. Defects can have a profound effect on the electrical properties of devices and identifying them is an important issue to produce reliable devices.
1) Electron trapping in SiO2
The mechanisms of electron and hole trapping in SiO2 and the nature of trapping sites are important for our understanding of a wide range of physical phenomena, such as radiation-induced damage and electrical breakdown, and for applications in fiber optics and microelectronics. In particular, electron trapping is known to have dramatic effects on the performance and reliability of electronic devices employing SiO2 as gate insulator, providing a direct contribution to the electric field at the surface of the semiconductor channel. Hole trapping has been relatively well understood in SiO2, however identifying sites responsible for electron trapping in silica, bulk and surface, has proved particularly challenging. Little is still known regarding the possibility of intrinsic electron trapping in the a-SiO2 network.
Ab initio calculations demonstrate that electrons can be trapped even in an idealised a-SiO2 matrix forming deep electron states in the gap. These electron traps are not dissimilar, both in geometry and in electronic structure, to electrons trapped by Ge impurities in α-quartz. These states can be responsible for the electron trapping observed at interfaces of SiO2-based MOS devices and should be present in bulk silica samples. Molecular dynamics simulations can be used to estimate the concentration of electron trapping sites.
2) Formation mechanisms for SiO2 defects and their subsequent charge trapping reactions
Ever since the identification of the paramagnetic E′ centre in SiO2 as an unpaired electron localised in an sp3 hybrid orbital of an Si atom backbonded to three oxygen atoms, a number of attempts has been made at explaining the optical and electronic properties of SiO2 in the presence of E′ centres. The irradiation or hole injection induces trapping of positive charge in thin layers of a-SiO2 grown on silicon surfaces by thermal oxidation. This effect has been correlated with paramagnetic E′ centre signals, leading to the initial assignment of the E′ centre as a hole trapped at a neutral oxygen vacancy. However, this model fails to account for a number of observations, such as the positive charge trapping without generation of E′ centres, the formation of high density of E′ centres without the corresponding density of positive charge, and the absence of correlation between the decrease of the E′ centre density and the density of positive charge upon post-irradiation electron injection in SiO2.
More recent experiments reveal that the paramagnetic state of the E′ centre is not always correlated with the entity bearing the positive charge. It has been suggested that the positive charge is protonic in origin, a hypothesis later corroborated by a number of experimental results. Consequently the O3≡Si-H entity in a-SiO2 has been suggested as a possible E′ precursor, where upon hole trapping hydrogen dissociates in the form of a proton leaving behind a neutral paramagnetic E′ centre.
Molecular dynamics simulations can be used to construct amorphous SiO2 models which contain Si-H bonds. Subsequent ab initio calculations reveal that hole injection leads to the dissociation of the Si-H bond, producing a 3-coordinated Si centre and a proton bound to a bridging oxygen. After hole injection, this process can occur spontaneously or with a barrier up to 0.5 eV, depending on the spatial location of the Si-H moiety in the amorphous structure.
Silicon based MOSFET's, widely used in highly integrated electronic devices, contain a thin layer of SiO2 grown on silicon making a Si/SiO2 interface. Scaling of transistors, a method of increasing computational power, has meant that the physics of the interface is becoming ever-more dominant. Understanding how the atomic structure of the interface correlates with physical and electrical properties will allow us to create more reliable devices in the future.
We are looking at ways to construct reliable models of the Si/SiO2 interface. Using a combination of classical and ab initio simulations can be used to construct interface models. We are interested in the use of charge variable potentials, such as ReaxFF. Ab initio calculations are performed to further optimise the structures and study the electronic structures of the models to compare with experimental data. Defect models that exist in bulk a-SiO2 provide a starting point for studying the defects that can exist at the Si/SiO2 interface.
NB: To obtain email addresses please replace [*] with @
University College London - Gower Street - London - WC1E 6BT - +44 (0)20 7679 2000 - Copyright © 1999-2013 UCL