Computational studies of conductivity in wide-band-gap semiconductors and oxides

The ability to control conductivity is essential for design and fabrication of (opto)electronic devices. Such conductivity control has traditionally been very difficult in wide-band-gap semiconductors, and native point defects have often been invoked to explain these problems. State-of-the-art first-principles calculations based on density functional theory have been used to elucidate these issues. Approaches for overcoming the 'band-gap problem', including the LDA + U method, allow more accurate comparisons and predictions of defect levels. The methodology is illustrated with the case of native point defects in zinc oxide. Computations reveal that the prevailing n-type conductivity cannot be attributed to native defects; it must thus be caused by impurities that are unintentionally incorporated. Hydrogen is shown to be an excellent candidate for such an impurity.