Light concentration increases solar-cell efficiency and reduces cost

Solar cells based on III-IV compound semiconductors and light concentration have potential applications in personal electronics and large-scale power generation.
17 May 2007
Claudio Pelosi and Matteo Bosi

The solar cell industry has grown quickly in recent years due to strong interest in renewable energy and the problem of global climate change. Currently, silicon solar cells rule the photovoltaic (PV) market. The best commercial Si PV modules have an efficiency of about 15% and cost between $4 and $5/Wp (Watts peak). PVs based on CdTe, CuInGaSe (CIGS), CuInSe (CIS), and organic materials are being developed with the aim of reducing the price per watt even if that means sacrificing conversion efficiency and reliability.1

Si solar cells were initially used in the exploration of space, which remains an important application today.2 Since their first development in the late 1950s, the efficiency of these cells has risen from 6% to more than 15%. This plus a decrease in cost and and increase in production throughput has permitted companies to sell them for terrestrial use. In the 1980s, a new generation of solar cells based on III-V compound semiconductors, namely GaAs and InGaP, came into use in space applications.3 Complex heterostructures based on arsenides and phosphide multijunction solar cells were developed and then realized on Ge substrates by means of metal-organic vapor phase epitaxy. Improvements in the 1990s permitted these solar cells to surpass the 20% efficiency mark, providing a significant boost to satellite power sources. Then, by the end of 2000, a triple junction InGaP/GaAs/Ge device had achieved 30% efficiency.4 Space remains the main application of these solar cells, as the cost of the device is a minor concern compared to efficiency, reliability and weight. But III-V semiconductors based on arsenides and phosphides also find use on Earth as the materials of choice for commercial optoelectronic devices as well as high speed and high frequency devices for mobile phones, satellite communication, LEDs and lasers.5,6

Due to cost and low production yield, III-V solar cells have not yet found general application as flat PV modules similar to the ones based on Si. But several research groups have investigated the possibility of employing III-V solar cells in conjunction with light concentration.7–9 Using optical devices such as mirrors, fresnel lenses, dichroic films, and light guides, it is possible to collect solar light and concentrate its energy on a single small area solar cell. This reduces the total cell area an amount equal to the concentration ratio and thus decreases the cost of the PV system, as a relatively inexpensive optical concentrator replaces the expensive semiconductor material. Moreover, using light concentration boosts the cell conversion efficiency.

The most advanced III-V solar cells presently in commercial use are monolithic devices composed of three P-N junctions working in series, each cell tuned to a particular region of the solar spectrum. Being designed for use in orbit, these cells are matched to the solar spectrum in space (air mass zero or AM0) to obtain the maximum conversion efficiency. Use of these cells here requires that they be redesigned due to the different solar spectrum on Earth (AM1.5), and because the light changes both during the day and over the course of the year.

Use of single junction GaAs or mechanically separated cells could overcome these problems (see Figure 1).10 A single high efficiency (∼30%) cell coupled to an optical concentrator simplifies the overall design of the module and, provided that concentration ratio is high enough (>1000×), reduces the dimensions of the device to those used in the optoelectronic industry. This makes it possible to use commercially-available processing techniques to fabricate the cells.

Figure 1. By linking several cells tuned to a particular spectrum region in (a) parallel or (b) series, more of the solar spectrum can be used.

On the other hand, a triple junction approach or a hybrid ‘mechanically separated multi-junction cell’ would permit use of the full solar spectrum, thus achieving a theoretical efficiency approaching 60% for a six junction device (see Figure 2).11 The United States Defense Advanced Research Projects Agency (DARPA) is currently pursuing this goal with a $50 million project aimed at developing low-dimension cells that would reduce the weight of soldiers' battery packs.

Figure 2. Efficiency of different types of solar cells under direct light (silicon) and concentration. The maxiumum efficiency currently obtained is reported for the Si, GaAs, and triple junction cells; the target efficiency is shown for the mechanically separated multi-junction cell.

Since the main obstacle to mass production of III-V solar cells is cost, large-area installations of such solar panels are some years off. The first application of these new PV cells could actually be to power or recharge portable equipment such as cell phones, MP3 players, or camcorders. Thanks to the higher efficiency of III-V solar cells as compared to Si, they could easily satisfy the demands of these low-power devices. Nevertheless, large area installations should not be neglected, as concentration could facilitate more rational use of land for installations as well as boosting the energy production of large solar power plants. Concentration could also challenge conventional rooftop Si panels, particularly since improved kWp/m2 generation would make a concentration system installation possible in large apartment buildings. Moreover, a better integration of PV elements with architectural design could be achieved.

In the near future, solar cells will probably be used in a wide variety of applications, from powering buildings to recharging personal electronics. At the same time solar cells will become increasingly evolved devices, exploiting emerging technologies and know-how, mainly from optics and materials science.

Claudio Pelosi
Institute of Materials for Electronics and Magnetism Italian National Research Council
Fontanini, Italy

Prog. Photovolt: Res. Appl. 12,
pp. 69–92, 2004. doi:10.1002/pip.541.
Mater. Sci. and Eng. R 40,
pp. 1–46, 2003.
Appl. Phys. Lett. 56,
p. 623, 1990.
28th IEEE Photovoltaic Specialists Conf.,
pp. 955–960, 2000. doi:10.1109/PVSC.2000.916044.
Appl. Phys. Lett. 69,
pp. 1532–1534, 1996.
IEEE Trans. Microwave Theory and Techniques 36,
p. 1966, 1988. doi:10.1063/1.117995.
Prog. Photovolt: Res. Appl. 8,
pp. 93–111, 2000.
Prog. Photovolt: Res. Appl. 15,
pp. 51–68, 2007. doi:10.1002/pip.715.
Proc. 20th Euro. Photovoltaic SolarEnergy Conf.,
pp. 82–85, 2005.
31st IEEE PhotovoltaicSpecialist Conf.,
p. 525, 2005.
Recent News