Ultra-thin cells on steel

Solar panels have invaded the roofs of our homes. And they’ll stay there a long time even though new, more sustainable photovoltaic cells are being developed.

Electronic circuits, too, have invaded our daily life, but their small size usually means they’re overlooked: few people question their energy consumption or efficiency and probably even fewer their recyclability. At the UCLouvain Institute of Information and Communication Technologies, Electronics and Applied Mathematics (ICTEAM), however, research groups are studying the sustainability of electrical circuits, that is, all the processes in their manufacture, power consumption and recycling: every link is important and must be studied in order to make the whole and its use as environmentally friendly as possible. This is the mission of Prof. Denis Flandre, a member of Louvain4Energy, whose laboratory is part of the European Nanoelectronics Consortium on Sustainability (ENCOS). He wants electronics, specifically circuit energy supply, to become more respectful of the environment. ‘If a building has an energy consumption control system, that’s beneficial, but the system uses chemical batteries, some environmental benefits are lost owing to routine battery replacement, recycling problems, and so on. It’s better to try to recover the environmental energy.’

Environmental energy? This can be ambient heat, vibration, or even the electromagnetic energy that surrounds us. Many such possibilities are being explored, but none achieve the performance of the most abundant energy: light captured by photovoltaic cells. ‘Mind you,’ Prof. Flandre cautions, ‘not the silicon cells that have lain our roofs for decades. We need very small cells – a few cm2 at most – but still sufficiently powerful to provide enough energy for a set of electronic components.’ Let's imagine a building’s monitoring system: it’ll need many sensor nodes capable of measuring temperature, various forms of pollution, humidity, noise, the presence of people, not to mention components for storing and processing collected data and for wireless transmission. Each block will have to be optimised in order to consume as little as possible. If this is the case, we can imagine consumption of 100 microWatts (100,000 times less than an LED bulb). A priori, nothing more simple: in summer, with a beautiful blue sky, the sun provides a maximum of 1,000 Watts/m2! Except that the yield of good solar cells is only 20%, or 200 Watts/m2 in the best case; in winter, it's ten times less. Also, in the case of our sensor systems, cell exposure isn’t optimal throughout the day and, most often, the systems have to work with artificial light, which is much less intense (by a factor 100 more) than the sun. And above all, the maximum size of the cells you want to use isn’t a question of square meters but square centimetres. This requires technologies other than silicon.

New generation

The most widespread solar cells are in fact made of silicon-based semiconductors. They’re a first-generation, mature technology, although improvements can still be made. Silicon is produced from silica, the most abundant material of the earth's crust (sand), whose cost is obviously low. Silicon ‘cores’ are then cut into slices of 15 x 15 cm2, generally of a thickness of at least 100 to 150 microns (a hair is about 80 microns thick). They’re then assembled on a panel, or substrate. With a maximum energy efficiency of about 24%, they remain the most suitable cells for everyday use.

But there are other types of cells, called second-generation cells: very thin layers of metals are directly applied to a glass substrate. Two technologies predominate here, which are named after the applied metal layers, the so-called CdTe (cadmium and tellurium, whose toxicity raises questions) and the so-called CIGS (copper, indium, gallium, and selenium). The latter is the focus of Prof. Flandre and his team.

‘Standard CIGS cells achieve a maximum efficiency of around 20% in industrial applications’, Prof. Flandre says. ‘We’re therefore getting close to silicon cell performance, but this result should be improved. As for the film deposited on the glass substrate, it’s about two microns thick. It may not seem like much, but it's still too much!’ For a simple reason: indium, gallium and selenium are either not available, expensive or hard to recycle. Some research conducted by the Louvain4Energy laboratory and developed with the Arcelor-Mittal Metallurgical Research Center (CRM), in Liège, thus concerns a new alloy that would be composed of copper, zinc, tin, and sulphur (CZTS) instead of CIGS. These are widely available and potentially non-toxic materials, but their yields don’t exceed 12%. ‘So we have another project, which we’ve just started with the University of Mons and is funded by the Walloon Region, to test copper oxide, CuO, which is more than 99% recyclable. Mathematical modelling is promising, but so far actual yields fall far short of theoretical performance.’

Make it thinner

Another way to make these cells more sustainable in the short term is to keep the CIGS layer but reduce its thickness by a factor of three or four, that is, from two microns to 500 or 600 nanometres, in order to consume significantly fewer metals, less manufacturing energy, etc. ‘But there’s a problem’, Prof. Flandre says. ‘A two-micron layer of CIGS absorbs the solar spectrum from ultraviolet to near infrared. This will no longer be the case if we reduce its thickness, it won’t absorb long wavelengths, so energy efficiency will be lost.’ Overcoming this obstacle is the goal of Advanced Achitectures for Ultra-Thin High-Efficiency CIGS Solar Cells with High Manufacturability (ARCIGS-M), a European project in which UCLouvain researchers participate. One of their contributions to improving the CIGS technique was to place an aluminium layer (Al2O3) on the rear interface of the CIGS (see diagram).

For very thin cells to absorb the entire solar spectrum, researchers are now proposing to add another metal under this interface.

‘The idea,’ Prof. Flanders explains, ‘is to place a light reflector under the CIGS layer and build optical cavities in which the light will be trapped and reflected multiple times. The Al2O3 layer is not enough but we’ve found that if we add some metals such as silver, optical reflection increases. Electronic properties can also be improved and we hope to maintain a yield of about 20% while greatly reducing the amount of CIGS.’

From glass to steel

Another possible improvement is in the substrate. CIGS technology was originally designed for a glass substrate, which is very convenient for solid cells ranging from a few hundred cm2 up to one m2. Replacing the glass with steel, however, would have two advantages. The first is that it can be more environmentally friendly over a complete life cycle because it’s lighter to transport; the second is that printing metal films could be done almost continually. The roll-to-roll technique is well-established in the iron and steel industry: the cell’s metal layers would then be deposited via spraying on steel coils. This would integrate solar cells in the building material, as industrial building roofs are often made of steel. Arcelor-Mittal's CRM is therefore of great value for research on solar cells, a field that is a priori far removed from iron and steel.

Henri Dupuis

A glance at Denis Flandre's bio

Denis Flandre earned a bachelor’s degree in electrical engineering with a specialisation in microelectronics in 1986 at UCLouvain. He had two PhD options: photovoltaics or silicon on insulator (SOI) technology. ‘At the time,’ he recalls, ‘photovoltaics was seen as having no future in Belgium! To the point that UCLouvain researchers, who had a perfect command of this technology, left the country for California and were among the first to join SunPower, today a global giant in the sector, a subsidiary of the Total group.’ He earned his PhD in 1990, then undertook a postdoctorate in Barcelona. He was as FNRS researcher for 10 years until 2001, when he was appointed professor at UCLouvain. Prof. Flandre has always been concerned about applying his research to society. In the early 2000s, he participated in the creation of the start-up Cissoid, which specialised in electronic circuits that operate at very high temperatures, for example for the electric automobile, or withstand very great temperature differences, for example for the space sector. His research also led to the creation in 2014 of e-peas, a company that creates and sells autonomous electronic circuits with very low energy consumption, which recently won several innovation awards. He also participated, in the biomedical field, in the development of a pressure sensor system to be implanted in the eye for glaucoma treatment follow-up, and electronic sensors to detect the presence of very low numbers of bacteria. Since 2008, his laboratory has conducted research in the field of photovoltaics, particularly on improvements to be made to silicon, CIGS, CZTS and now CuO solar cells.