Design of transparent electrodes
Whether palmtop, laptop or desktop, the basic idea layout of a computer has not changed for decades: the screen showing the data is separate from the chip processing the data. This could all change with the advent of new materials that can bring display and processing functions into intimate contact. These new materials are simultaneously semiconducting (like silicon chips) and transparent (like a display screen). This combination of properties is also needed in technologies such as LED’s, solar cells, smart windows and sensors.
Collaborating with partners in an EU-funded project called NATCO, we have developed a new material that is simultaneously semiconducting and transparent. In terms of its transparency across a wide range of wavelengths, the new Tyndall material is better than existing materials in its specific class of so-called p-type transparent conductive oxides (TCOs).
The new material was designed at Tyndall entirely through computer simulation, by initially simulating how different elements from the periodic table affect the transparency and conductivity of a known material, copper oxide, Cu2O. This approach yielded a set of design rules that were applied to other existing TCO materials in order to determine the optimum composition.
The first step was to identifying neutral Cu vacancies as a mechanism for p-type semiconducting behaviour in Cu2O. Upon formation of a neutral Cu vacancy, a delocalised hole state forms, with acceptor levels 0.2 eV above the valence band. Similar investigations of CuAlO2 and SrCu2O2 show that the same vacancy formation process holds for these TCOs and that acceptor (p-type) defects dominate the defect chemistry of these materials. This makes SrCu2O2 a good candidate since its band gap is also reasonably large, giving moderate transparency.
Therefore, we have looked at appropriate doping strategies to widen the band gap. We illustrate our DFT findings using the examples of Sr and Cd doping of Cu2O (see second illustration). Both dopants distort the structure of the oxide. However, Sr doping enhances the band gap, while Cd doping reduces the band gap. The density of states shows that Cd introduces electronic states in the band gap, while Sr does not and this is the origin of the differing effects of these dopants on the band gap. Having looked at a wide range of possible dopants, we found that the optimum new TCO material was barium-doped strontium copper oxide.