Computed model of the structure of the defect that stores charge in silicon nitride, namely the neutral nitrogen vacancy. (The host structure is shown in a stick representation, while distorted Si atoms around the N-vacancy are shown as light grey balls).

Computed model of the structure of the defect that stores charge in silicon nitride, namely the neutral nitrogen vacancy. (The host structure is shown in a stick representation, while distorted Si atoms around the N-vacancy are shown as light grey balls).

• 
• 

First-principles calculations based on density functional theory have been successfully used to gain understanding of defects and impurities in semiconductors and insulators used for electronic and optoelectronic devices. 

In the example shown here, we find out what atomic arrangements allow charge to be stored in silicon nitride (Si₃N₄), which allows that material to be used to store data in SONOS and TANOS-type flash memories.

Crucial issues for the growth and fabrication of electronic devices are the control of defect formation and of doping level. This requires characterisation of the native defects, to understand the origin of intrinsic photoconductivity and the effect of doping with impurity elements.

We have calculated defect formation energies and thus defect concentrations. Additionally, the electrical behaviour of the defects is assessed by calculating the thermodynamic transition levels between charge states in the band gap. This allows us to characterise localised defect levels in terms of their energy depth within the band gap (shallow vs deep).