Development of semiconductor nanowire based solar cells

Solar energy conversion is an attractive process among other clean and renewable energy sources and is today highly demanded in research strategies worldwide. Recent theoretical studies have indicated that coaxial semiconductor nanowires having a radial p -n junction with the wire axis parallel to the light direction could provide large improvement of carrier collection and overall efficiency relative to the conventional planer p-n junction solar cell. The claimed improved efficiency is because of the fact that in the case of nanowires the light absorption and the charge carrier collection directions are decoupled. For the solar cell application group III-V semiconductor material such as gallium arsenide (GaAs) is preferred over silicon (Si) because of its bandgap is direct and nearly ideal for single junction cell. Due to the tiny diameter of the GaAs nanowires, integration of this group III-V semiconductor material with Si has also become possible and the lattice mismatch between GaAs and Si is no longer an issue. This integration would not be possible for planer structure and the interface between GaAs and Si would be associated with defects such as dislocations. Since 2010 we have been able to GaAs nanowires epitaxially on a Si substrate by Ga-assiste d vapor-liquid-solid mechanism using molecular beam epitaxy. In this Ga-assisted mechanism no foreign catalyst particle, e.g. gold, was used which might otherwise get incorporated in Si creating mid-bandgap traps. These nanowires are covered with an alumi num gallium arsenide (AlGaAs) shell to passivate the surface states of bare GaAs core and to form a radial shell. To tune the electrical properties we also successfully doped the NWs, and then we made a p-n junction radial core-shell structure for a NW so lar cell device. We already have achieved the growth of the nanowires with controlled density and interwire separation on a small scale by pre-defining nanohole arrays on silicon oxide covered Si substrate by electron beam lithography. Recently, using nan oimprint lithography we also have been able to pattern the Si substrate and grow the nanowires on a large scale. A yield of about 80% of such vertical GaAs nanowires was achieved which required a systematic and clean substrate preparation method together with a fine tuning of the nanowire growth parameters. Radial p-i-n junction position controlled nanowires with different doping and their layer thickness are successfully grown. The device fabrication and their solar cell characterizations are under proce ss. Popular science description of or recent our discovery of epitaxial growth of GaAs nanowires on graphene: Both graphene and semiconductor nanowires are today getting prime attention in nanoscience and nanotechnology due to their unique properties. NTNU and the newly created start-up company CrayoNano AS introduce a route how to combine them epitaxially into novel functional hybrid structures. This is an important scientific breakthrough, since earlier attempts to grow epitaxial semiconductor thin films, like e.g. silicon and GaAs, on graphitic substrates have not yet been successful. The research group led by Professor Helge Weman at the Norwegian University of Science and Technology has demonstrated a model for how semiconductors can be epitaxia lly grown on graphene and other graphitic substrates, by making use of the nanowire?s unique geometry and crystal symmetry. The model is general, and thus applicable for all semiconductor materials. The research group has also experimentally verified the model by epitaxially growing uniform self-catalyzed GaAs nanowires on graphite and graphene by molecular beam epitaxy. The nanowires were found to have a uniform size and cross-section, with a strong binding to the graphitic surface. Since the nanowires a re self-catalyzed (i.e. grown under gallium droplets rather than grown under e.g. gold droplets), the approach avoids any foreign elements that could affect the active semiconductor in subsequent device processing or operation. Furthermore, the group has fabricated photodetectors from single GaAs nanowires grown on graphitic substrates with a high responsivity. This reveals that semiconductor nanowires epitaxially grown on graphene have a potential in e.g. future flexible nano-electronic and nano-optoele ctronic hybrid device systems where graphene can function both as a transparent electrode as well as an epitaxial "substrate-free" growth template for the active semiconductor material. For example, vertical semiconductor nanowires grown on free-standing graphene can function as a template in order to fabricate flexible and transparent electrodes for nanowire solar cells and light emitting diodes.

Our proposed core/shell nanowire (NW) solar cell will consist of two main layers: a core GaAsSb NW and a radial GaAs shell. Using this material system a type-II bandgap alignment with GaAs is achieved, where electrons are confined in the GaAs shell and ho les in the GaAsSb core. When light is absorbed by the core/shell NW, electron-hole pairs are created. One key advantage of this type-II system is that the electron and hole are separated very efficiently even without any doping. Another advantage is that the electrons and holes only have to move across a very short distance (<100 nm) and therefore one can reduce the electron-hole recombination rate. Good optical absorption is achieved since the light is mainly absorbed along the direction of the NW. In pr inciple both the core (GaAsSb) and shell (GaAs) can absorb light and final dimensions will be optimized in terms of overall solar cell efficiency. Compared with Si based NW solar cells, GaAs has the additional advantage that it has a direct band-gap and t hus absorbs light much better than silicon. The thickness of material required to absorb 90 % of the power due to above-band-gap photons is 140 times smaller in GaAs than in Si (less material means lower cost). This "optical thickness" is only 900 nm for GaAs compared to 125 micrometer for Si. Even with 140 times shorter NWs, the solar cell efficiency for GaAs NWs is still expected to be a factor of two higher than for Si NWs, partly due to a better optimized bandgap to the solar radiation spectrum. In this proof-of-principle NW solar cell we target an efficiency of 10% by 2012, a factor of three higher than present state-of-the-art. Clean, renewable and high efficiency energy sources are today highly demanded in research strategies worldwide in order t o limit the green house effect on the climate and ensure the ever increasing energy need in the world. If successful our solar cell has the potential to be a good source of cheap renewable energy in the future.