Cheaper Silicon Found Effective for Solar Cells
A research team from the University of California at Berkeley, Lawrence Berkeley National Laboratory, Argonne National Laboratory, and Pacific Northwest National Laboratory, using U.S. Department of Energy (DOE) synchrotron light sources, has successfully shown that inexpensive silicon has the potential to be used for photovoltaic (PV) devices, commonly known as solar cells. In a new approach—whose findings were published online in Nature Materials (August 14, 2005)—the researchers used nanodefect engineering to control transition metal contamination in order to produce impurity-rich, performance-enhanced multicrystalline silicon (mc-Si) material.
"Solar energy is often touted as the most promising and secure energy source, capable of reducing our dependence on foreign fuels while reducing the emission of dangerous gases that change world climate. Even though the current, worldwide growth rate of PV of over 25 percent per year is nothing short of amazing, photovoltaics could grow much faster if researchers and manufacturers could further reduce the cost of solar cells and overcome the shortage in the high-quality, semiconductor-grade silicon used presently to make commercial solar cells." said Eicke Weber (UCB), the project's principal investigator.
To that end, Tonio Buonassisi, Andrei Istratov, and Eicke Weber, of the University of California, Berkeley, teamed up with scientists from three DOE national laboratories: Barry Lai and Zhonghou Cai, of the Advanced Photon Source at Argonne National Laboratory; Steven Heald, of Pacific Northwest National Laboratory; and Matthew Marcus, of the Advanced Light Source at Lawrence Berkeley National Laboratory. The team studied transition metals, native contaminants to less purified silicon, and their effect on solar cell material performance using highly sensitive, state-of-the-art synchrotron-based analytical techniques.
The researchers’ work at finding a way to supplement the currently used high-quality silicon feedstock with what they describe as "cheaper but dirtier alternative feedstock materials" was spurred on by the high cost of polysilicon (polycrystalline silicon). Currently, polysilicon is purified in a complex refining procedure and, thus, only a limited amount (about 30,000 tons) is produced each year. At the same time, the photovoltaic industry, a major user of polysilicon, has steadily increased its utilization of the material over the past few years. In fact, 2004 became the first year that the available supply for silicon feedstock did not meet its demand.
Many experimenters in the past have attempted to use lower quality, abundantly available solar-grade silicon (SoG-Si) feedstock to produce cost-effective solar cells. However, the traditional solar cell processing steps designed to reduce the detrimental impact of metal contamination on material performance, phosphorus diffusion gettering and hydrogen passivation, were limited in their capacity to improve material containing very high amounts of metal impurities. A new approach was necessary.
In this work, it was discovered that the size and spacing of metal nanodefect clusters controls the diffusion lengths of minority carriers, a key component in establishing performance quality in solar-cell devices. To understand this relationship between minority-carrier diffusion length and metal impurities, the researchers turned to the 2-ID-D and 20-ID-B beamlines at the Advanced Photon Source and the 10.3.2 beamline at the Advanced Light Source.
Using recent advances in synchrotron-based X-ray fluorescence microscopy, X-ray absorption microspectroscopy, and spectrally resolved X-ray-beam-induced current, the researchers were able to (1) map the distribution of metal impurity nanoprecipitates with sub-micron spatial resolution, (2) analyze the chemical state of metal impurity clusters, and (3) map the minority-carrier diffusion length in situ. Tonio Buonassisi and Andrei Istratov (UCB/LBNL) explain: "It was only with this suite of highly-sensitive synchrotron X-ray microprobe techniques capable of detecting metal clusters as small as 30 nanometers that we could determine the chemical state of metal impurities, their spatial distributions, and their impact on solar cell performance. We have, in essence, directly observed the impact of nanometer-sized defects on centimeter-sized devices." Coauthor Barry Lai ( ANL/APS) adds: "The identification of the types of defect present in mc-Si is crucial in determining their impacts on device performance and hence the development of proper remediation strategies."
Within the paper, the researchers suggested that rather than removing all metal defects from silicon feedstock, which is expensive and time-consuming, large amounts of metals inside the feedstock could remain as long as their individual sizes and distances apart from one another are restricted by the application of nanodefect engineering. They found that "to maximize solar-cell efficiency without changing the total metal concentration, all metals must be completely contained in large, micrometer-sized clusters separated by several hundreds of micrometers, thus minimizing the interaction between metal atoms and charge-carrying electrons."
The researchers demonstrated that one possible means to achieve a more beneficial distribution of metals is by tailoring the cooling rate after high-temperature processing. It was demonstrated that even in heavily contaminated mc-Si the minority-carrier diffusion length could be raised by a factor of four. Such an improvement indicates that the use of lower-quality silicon feedstocks with strict contamination engineering may be an economical alternative in manufacturing low-cost commercial solar cells.
William Arthur Atkins
The work is supported by the U.S. Department of Energy's (DOE) National Renewable Energy Laboratory (NREL) and its University Crystalline Silicon Research Project. Richard Matson, the project's manager, stated that, "The group's contribution to the research and development of silicon solar cells is seminal and their synchrotron based X-ray characterization technique is both quite powerful and unique in the field."
See: Tonio Buonassisi, Andrei A. Istratov, Matthew A. Marcus, Barry Lai, Zhonghou Cai, Steven M. Heald, and Eicke R. Weber. "Engineering metal-impurity nanodefects for low-cost solar cells," Nature Materials, online August 14, 2005.