|
The
PV industry is expected to continue to grow at an annual rate of about
20%. The well-established technology base and ready availability,
proven performance, and salubrity of silicon, coupled with economies of
scale in larger factories, will likely allow Si to remain the dominant PV
material for the foreseeable future.
The demand for off-specification polycrystalline silicon feedstock
for PV use is likely to exceed the available supply by a factor of at
least 2 within the next 10 years, and this will probably be an impetus for
development of alternative feedstock material with adequate but not
excessive purity levels.
In
ingot growth, the trend for single crystals will be away from the smaller
100- and 125-mm-dia. sizes with more focus on 200-mm diameters.
Despite the potential advantages of FZ material, it is unlikely
that its role in PV will increase significantly because of higher costs
for the crack-free, long cylindrical feedstock it requires and the difficulty
in producing the larger FZ diameters.
In CZ growth, we are likely to see an increased effort to make hot
zones more energy efficient, to grow larger diameters, and to achieve
continuously melt-replenished long growth runs.
An effort will continue to evaluate tricrystalline growth or other
means of strengthening the ingots so as to improve breakage yields for
thin wafers. There will be a
continuing effort to achieve more wafers per length of ingot, and to take
advantage of potentially higher cell efficiencies afforded by thinner
wafers when back surface fields are used in the cell design.
Multicrystalline casting, directional solidification, and
electromagnetic casting are commanding an increasing share of the Si PV
market (53% of all ingot-based modules sold in 1998 were
multicrystalline). This trend is likely to continue because the processes
and equipment are simpler and the throughputs are higher (especially for
electromagnetic casting) by a factor of 5 to 20.
In
the ribbon- and sheet-growth technologies, a challenge for dendritic web
growth and edge-supported pulling will be to increase areal throughput via
wider ribbons, multiple ribbons, or other approaches.
Even though these methods have the advantage of minimal silicon
consumption and elimination of wafering, it is unlikely that they can
effectively compete with their current throughput of 1-2 m2/day.
This is because the effective areal throughputs of ingot growth
range between 30 and 600 m2/day, and other sheet technologies
produce 20 m2/day to >1,000 m2/day.
While capillary die growth of octagons produces about 20 m2/day,
experimentation is under way to grow large-diameter, thin-walled circular
tubes (as depicted in Silicon Ribbon and Sheet Growth) a
meter in diameter and much thinner than current octagonal tubes.
This would increase throughput to more than 75 m2/day.
So far, tubes with 0.5-m diameters have been made, effectively
doubling the current octagon areal throughput (Roy et
al., 1999). These tubes have
been grown less than 100 mm
thick.
We will probably see continued progress in horizontally pulled,
large-area solid/liquid interface sheets by some variant of the method
shown in Silicon Ribbon and Sheet Growth
because the throughput potential is enormous and one growth
furnace could easily generate material for 35 MWp/year or more of solar
cell production.
The
future is expected to bring continued exploration of thin-layer Si growth
approaches, in search of ones that have significant economic advantages
over the best ingot and sheet techniques.
Successful ones will have fast deposition rates, large grain sizes,
high efficiencies (at least 14% production efficiency), compatibility with
low-cost substrates, and amenability to low-cost cell-fabrication schemes.
It is not likely that production of thin-layer Si PV modules will
be a significant fraction of the mainstream PV market for at least 10
years, although they, like the ingot and sheet approaches, would have
substantial advantages over many other thin-film PV approaches.
These include the simple chemistry and relative abundance of the Si
starting material.
The
Earth's crust contains 27.7% Si, in contrast to 0.00002% Cd, 0.00001%
In, 0.000009% Se, and 0.0000002% Te (commonly used thin-film elements).
In addition, crystalline Si benefits from an extremely
well-established technology base, compatibility with SiO2 surface
passivation, relative salubrity with respect to toxicity, and stability
under light exposure.
__________________
Roy, A., Chen, Q.S., Zhang, H., Prasad, V., Mackintosh,
B., and Kalejs, J.P. (1999) Presentation at the 11th American
Conference on Crystal Growth & Epitaxy, Tucson August 1-6.
To be published in J. Crystal Growth. |
