On Mon, 31 Jan 2000, Stirling Westrup wrote:
> Is there anything stopping us from making sheets of rectennae (I gather
> that is a rectifying antenna, but I'm not sure) that were designed to
> pick up wavelengths in the visible spectrum? These would have to be mighty
> small antennae, so we would have to build them with chip fabrication
> techniques or even nanotech, but they should be highly efficient and
> very long lasting solar cells shouldn't they? What technological barriers
> are currently preventing us from trying this out?
One of our physics experts might comment in more detail, but I'll offer
my 2 cents. In radio, I believe that the energy is used to induce current
flow in the antenna (thus leading to the rectenna idea when you prevent
the current from oscillating). Now in the infrared range the energy
primarily contributes to atomic motion (heat) which I suppose in theory
could be harvested via standard "Carnot/steam cycle" methods. Once you
get up into the visible range the photons knock the electrons from
the valence band up to the conduction band where you can use them
for current circulation. So the trick in solar cell engineering is
one of matching the valence-to-current band-gap to the energy of the
incoming photons. Silicon does well for IR and red, GaAs or InP are
better for visible frequencies, CdS is even better still. Multi-layered
cells of different materials (and therefore different band-gaps) selectively
absorb photons of different energies and can push the efficiencies up
to 30% or so, though most current home systems are stuck between
10 and 15%.
To my mind, normal solar cells are "rectennas" for visible light.
There is a descriptive and methodological shift that occurs going
from more wave-like to more particle-like behavior in the radiation
when it moves from the very high radio-frequency range into the
far infrared range. So, "rectennas" are an energy collection
method that to my knowledge only work for radation in the radio,
esp. microwave range.
The lifetime of solar cells tends to be determined by temperature
stresses and radiation damage. If you can keep them relatively cool
and prevent most of the UV radiation from damaging the crystal structure
they should have relatively long lifetimes. These are engineering
problems that are addressed in current manufacturing methods presumably
on a cost dependent basis. The primary barriers are getting the volumes
high enough that the R&D and plant construction costs become a much
smaller fraction of the sales price. Why is steel or cement so cheap?
Because we use so much of the stuff that we can build really *big*
mines, trucks, transport barges, blast furnaces, etc. to move collect
the material and move it around. Once those investments have been
made their contribution to the final product costs gets very small.
Ultimately you get to the point where most of the cost is the energy
that goes into the product, you can see that clearly with aluminium.
The energy costs for solar cells are probably of the same
order of magnitude as Si for chips, Steel or Aluminium because
they all have to be melted at some point. Aluminium is more
expensive, not because its rarer (its actually more common)
but because the process of separating it from its bauxite
requires more energy. Silicon is even more common but is
typically required at much higher purity, so it too requires
more energy to get it to that stage. However, once the
manufacturing methods are perfected, solar cells should be very
very cheap because you need only a very thin layer (a few hundred
micro-meters) to harvest the electrons. Since glass is a major
component of solar cells (and it has to be melted as well), I
would guess that at some point the solar cell manufacturing costs
should fall to about the same level as the equivalent area of glass plate.
An interesting exercise (for some ambitious reader) would be to
determine the amount of glass plate manufactured, the amount
of solar cells manufactured and assuming the solar cell mass
is doubling every year (??) while glass plate volumes are
growing only a few % every year, when do the volumes become
equal? I would guess the average homeowner would probably
install about equal amounts of windows and solar cells
so this gives you a good idea of when the costs should be
roughly equivalent. It seems to make sense that if you could
design your solar cells as a roll-down window shade that people
who don't have anyone at home during the day would want to lower
these shades while they are away at work.
The only advantage "nano" brings to all of this is some type
of self-assembly, so maybe you "paint" the cells onto the
outside of your house. They might also offer 3-4 layer
nanostructured materials that might push the efficiencies
up a bit (40%+?) and then if they are "active" they might
provide self-healing or self-repair to some degree. And
of course if you have your Almost Anything machine (as
previously discussed), and the designs, you can program
it to make a square foot or so of nano-solar-tile a night,
so you can simply go outside and stick it to the side of
your house in the morning as you head off to work.
The real limit at this point is getting the volumes up,
nano makes that really easy but I suspect the volumes
will be pretty high before nano arrives.
The nice thing in my mind is what a lot of cheap solar power
will do for the balance-of-trade deficit. When electric
or fuel cell cars (driven by alcohol produced via solar
powered bacteria) are available there then becomes a huge
incentive for the U.S., Europe, Australia, etc. to move
away from oil. At that point it seems there would be a
huge ROI to the people in a country if the governments
pushed the solar cell development hard (through low
cost loans, tax incentives, etc.). It makes a lot more
sense to me for the developed countries to be spending
their "gas&heating" dollars on the installation of
the infrastructure required to give them a relatively
"free" energy future. It would give me a huge amount
of satisfaction to fill my car up from the ever replenishing
alcohol tank sitting on my roof.
Robert
This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:03:04 MDT