Low cost solar energy

Materials World magazine
29 Apr 2014

Dr Vincent Barrioz of the Low Carbon Research Institute, Wales, explains how a new solar photovoltaic processing technique could help the country to deliver its low-carbon agenda.

High-cost materials and processing have traditionally priced solar out of an increasingly competitive energy market. But a new R&D project could turn that around. Researchers at the Low Carbon Research Institute (LCRI), in Wales, are developing innovations in thin-film solar photovoltaics (PV) as part of a new Welsh Government-funded project aimed at making PV solar energy more affordable. As part of the project, the team is developing a new deposition process to improve the production and efficiency of PV cells.

A thin film PV cell comprises a substrate that is coated with very thin layers of semiconductor materials, which enables the cell to convert solar radiation into electricity. The cells need to be lightweight, durable and easily produced at the lowest possible cost. One of several methods used to coat the substrate is metalorganic chemical vapour deposition (MOCVD). Based on this, the Solar Photovoltaic Academic Research Consortium (SPARC) project is developing an in-line system to reduce the amount of materials used to coat the substrates, speed up production and, in turn, reduce costs, while maintaining and improving the efficiency of the cells.

Over the last five years, the production scale-up and the global deployment of PV have been substantial thanks to a variety of Government subsidies, such as the Feed in Tariffs in the UK. The SPARC project is part of the LCRI’s Convergence Energy Programme, which is supported by the Welsh Government’s European Regional Development Fund (ERDF). In 2010, LCRI secured £19m to help Wales and its industry partners to lead the way in research to cut carbon emissions, as part of the EDRF’s Convergence programme. Solar was an obvious area on which to focus, as the PV community has always been driven to improve conversion efficiency towards the theoretical limit, which is around 30% for a single junction cell.

Ultimately, PV has to compete on a commercial basis with other electricity generation sources. As such, the amortised cost of electricity per kWh has to be reduced to reach grid parity. The price of PV has come down dramatically over the last three years, with increased production volume and continual improvement in PV module and inverter efficiency. New technology is struggling to compete on price but has better long-term potential as these production volumes increase.

Making the most of the sun
The conversion efficiency of solar PV is generally measured under standard conditions, where the device is kept at a constant temperature and is illuminated with artificial solar radiation that matches the sun’s intensity and spectrum at air mass 1.5. The maximum electrical power is calculated from the current–voltage measurements and is divided by the power from the incoming artificial solar radiation to yield the efficiency.

The record efficiency for a single junction device currently stands at 28% – not far off the maximum theoretical for epitaxial gallium arsenide – while that for crystalline silicon is also good, at 25%. Thin-film PV has traditionally been lagging further behind, but with the recent world record efficiency standing at 20.4% (held by US manufacturer First Solar, for cadmium telluride [CdTe]), thin-film PV efficiency is getting closer to that for polycrystalline silicon solar cells.

Thin film PV devices are generally deposited in vacuum-based processes, such as close-space sublimation, sputtering or plasmaenhanced CVD. However, by lowering energy use during production of thin-film PV modules, production cost and carbon footprint will also be reduced as a result. This can be achieved by any of two methods – either by going to lower processing temperatures or by adopting a non-vacuum processing method.

The chamberless secret
The SPARC project focuses on the development of a non-vacuum MOCVD process, which uses a low deposition temperature in comparison to other technology used for thin-film PV production. MOCVD allows conformal films to be deposited over non-planar surfaces, and is generally associated with batch process technology for complex opto- and micro-electronic structures. These are mainly used for III-V materials, with applications including mobile phones, lasers and high-brightness LEDs.

The SPARC team has developed a new, chamberless version of MOCVD, which does away with the vacuum chamber and delivers the gas mixtures for each film to be deposited through a novel injector arrangement onto a continuously moving substrate. In this process, metal-organic chemical precursors, having sufficiently high vapour pressure, are used as the starting materials. A carrier gas is used to controllably transport the vapour and precise mixtures into the reaction zone, where surface (and gas-phase) reactions onto a hot surface effectively form the desired compounds with controlled doping or alloying.

Moving MOCVD, which is usually an enclosed process, to a chamberless system, was not without its challenges. Ensuring safe and contained deposition was key, and so extensive computational flow dynamic modelling was used during the design and development of the coating head arrangement. The containment was verified, and real-time monitoring was made possible by placing a residual gas analyser in the exhaust of the deposition zone.

An industry game-changer?
Conversion efficiency of CdTe devices deposited by MOCVD has been making rapid progress over the past few years, and currently stands at 16%. Early results indicate that this new chamberless system can provide more control and better material uniformity than the batch process alternative. This is expected to translate into higher efficiency with higher quality material and improved design of the device structure. As chamberless MOCVD is able to deposit structures as a continuous process and is not limited to batch processes, it has the benefits of being both flexible and scalable. Furthermore, as the process operates at atmospheric pressure, it can be manufactured using roll-to-roll processes.

While the width of the process is currently limited to 15cm (compared to 5cm in the batch process), the current design of the coating head can be upgraded to a width of 30cm. Expanding the width to one metre or more would require some engineering scaleup, but is not impossible. This would come down to demand, and so while it could be used for large-scale PV production, it would require significant investment.

However, there is still work to be done. At present, the chamberless MOCVD system is limited to use on substrates that can withstand temperatures of up to around 450°C. Reducing the required processing temperature could open opportunities with plastic PV devices, and is something the SPARC team is hoping to achieve further down the road. It is also working on lightweight, flexible PV for space applications, by depositing the films onto ultra-thin, space-qualified coverglass – a system that will be ideally placed for producing large volumes of thin-film PV modules.