New concepts

Thin film solar cells

Using the ELO process thin layers of III-V materials can be integrated onto an inexpensive glass substrate. Initially, however, subsequent production into a well-functioning solar cell seems to be far harder than with III-V layers onto their original crystal wafer. This is related to the fact that such material properties as the thermal expansion and chemical durability of the glass (or plastic) medium differ greatly from those of the III-V film. Because of this, the first III-V thin film solar cells made at Radboud University had a far lower output than comparable cells on a wafer.

Intensive scientific and technological research has now resulted in great improvements in the ELO technique and the subsequent processing of the layer structures onto their new medium. The fact that this allows the thin films to be worked on each side opens up possibilities for the development of completely new cell structures. For instance, cells have been made, which on the reverse as well as on the front have a lattice contact pattern instead of the conventional full metal contact. This results in semi-transparent solar cells capable of capturing light on both sides. Furthermore, it has been demonstrated that, if it is applied in the right way, a full contact on the reverse serves simultaneously as a reflector. This means that the thickness of the already thin film can be halved once again, while this also creates a small increase in output. In the meantime in Nijmegen the output of this ‘mirror-solar cell’ has been boosted to 24.5%, a world record for single thin films. The current research in this direction is targeted at boosting the output to far above the 30% mark by using multi-junction solar cells in concentrator systems.

: Semi-transparent InGaP solar cell with a lattice contact pattern on front and back

: Single-junction thin film GaAs solar cell with a world record output of 24.5%

Multi-junction solar cells

: Quadruple solar cell structure with which in theory an output of 46% is achievable

For an output of over 30%, a so-called multi-junction solar cell is needed. In fact this is several solar cells stacked together, each sensitive to a different part of the solar spectrum. In this way a greater portion of the incident sunlight can be converted into electrical energy. Researchers in the United States have already made wafer based InGaP/GaAs/Ge triple-junction cells with an output of 32%. By using germanium (Ge) as the third cell this triple-cell cannot be separated from its crystal wafer by using the ELO process. Therefore, in Nijmegen alternative multi-junction cells are being developed such as a triple-cell in which the third Ge layer is replaced by the material InGaAs. This material does not completely suit the structure of the crystal wafer on which the cell is produced in the first place. This creates mechanical stresses in the InGaAs layer of the cell, which however are partly cancelled out following removal of the wafer. Moreover, a quadruple cell is being developed. This cell will consist of two separate tandem cells, each produced on the most suitable crystal wafer. Following removal of the wafers, the cells are stuck on top of one another. In this way there is a subcell suited to every part of the solar spectrum in order to convert light into electricity. Calculations show that under standard conditions an output of 46% is achievable with this. For this purpose besides optimum subcells this double tandem also requires optimum anti-reflection coatings (ARC) and tunnel junctions (TJ). Experimental study should reveal the extent to which this is also feasible in practice.

Concentrator systems

: Schematic sketch of a concentrator system

A very interesting method for greatly reducing the costs of generating electricity using solar cells is the application of a concentrator system. In this, light is captured over a great surface, focussed by lenses and mirrors onto a small solar cell. In this way costly semiconductor material is replaced by inexpensive components in glass or plastic. The current yield of the solar cell increases in proportion with the concentration factor of the system. Moreover, the voltage of the cell becomes a touch higher so that the delivered capacity (equal to current times voltage) of a solar cell rises more than proportionally with the concentration factor. This means that the output of the cell also increases with the concentration factor. This effect is the greatest with multi-junction cells. For instance, under standard conditions a maximum output of 32% is attained with a triple-junction cell, while under high concentration the limit of 40% has now been passed. With a quadruple solar cell as described above it even ought to be possible to achieve an output of over 55% with a concentration factor of 500 times or higher. Because such high concentration factors result in very high electrical currents and a high thermal loading, it imposes special requirements on the design of the solar cells.

Flat luminescent panels

: Schematic sketch of a luminescent panel concentrator

In contrast with a concentrator system based on lenses and mirrors, a flat luminescent panel concentrator captures sunlight from all directions. The light is absorbed in the panel in optically active particles and then radiated out again isotropically. In this way the light is led to solar cells at the edges of the panel via internal reflection. The spectrum at the edge of the panel is highly dependent on the type of optically active particles that are used. Radboud University is conducting research in collaboration with ECN in order to harmonise the properties of the panel and the solar cells as well as possible. However, absorption losses in the panel still make the concentration factors attained very low (typically a factor of two).

III-V solar cells as model system

: Schematic sketch of an intermediate band solar cell. Besides the blue part, the intermediate band also uses the green and the red parts of the solar spectrum

III-V materials possess a high degree of perfection and their properties can be adapted to a significant extent to suit the user’s wishes. This turns these materials into an ideal model system for studying the fundamental limits on the performance of solar cells experimentally. Therefore, from the point of view of device efficiency, striving for maximum functionality with minimum material usage is central. For the optimum solar cell of the future, this means generating much energy in a thin material layer. Taking off this energy leads to big current densities and saturation effects in the thin films. With thin film III-V solar cells these effects can be studied in practice, because it is now the only system combining a thin film solar cell with an output above that of the present generation of wafer based crystalline Si cells. Furthermore, scientific literature describes a number of promising new concepts such as using photonic crystals and intermediate band gaps with which in theory solar cells could attain an output of 63%. III-V materials are ideally suited for examining whether these concepts also genuinely work in practice.