…

                                                                                              F E AT UR E
                                                                      B    Y        M     I    K   E       V    A    R    G     O




                                                   Solar for Everyone

                   Erik Ydstie has a vision. This past summer,            T he Attac k on Cos t                                     exposed to sunlight. This triggers a flow of
                   while on a mission through his church to teach         The main challenge in making solar energy                 current between the n-type and the positive p-
                   English, the Chemical Engineering professor            more widely used everywhere is to bring down              type layer, and the current is harnessed.)
                   visited Burkina Faso, one of Africa's poorest          the cost. Here in the U.S. at present, equipping                PV cells don't require silicon of the
                   countries. There he found isolated villagers           a modest-sized all-electric home with a full              extreme purity needed for microchips. For
                   generating electricity with solar units they           array of solar roof panels and related gear can           years, the fledgling solar industry could simply
                   had somehow obtained. "You'd see a single              cost around $50,000. Although one then begins             buy off-spec material rejected by the IC
                   light bulb, inside a home, running off sunlight        to save money by getting free electricity from            industry, but that ride is over. Demand has now
                   captured during the day," he says. "Or an              sunlight, the payback period can run to many              outstripped the supply of reject material and
                   entrepreneur would have a little business, with        years and the high initial price -- chiefly for the       driven up prices; hence the need for a low-cost
                   panels laid out on the ground to charge people's       panels -- is beyond the reach of many users.              process to produce solar-grade silicon directly.
                   radio batteries and cell phones."                            On the plus side, there are also many                     On this Ydstie is working with the firm
                         Ydstie is now seeking partners for a pilot       possible ways of cutting costs. Researchers at            SGS (for Solar Grade Silicon, naturally). At a
                   project to "get this sort of thing going on a          Carnegie Mellon are investigating three major             facility in Moses Lake, Washington, SGS has
                   larger scale." He sees the biggest future for          avenues of attack.                                        been testing a novel silane-gas reactor process
                   solar energy as being in the developing world,               The nearest-term savings may come from              that takes place inside a thin "tower" about one
                   where wired electric grids do not reach vast           Ydstie's ChemE team, which is focused on a new            foot in diameter and three stories high. Plans
                   areas. Leapfrogging straight to solar, instead         process for making solar-grade silicon. Most              are under way for a production plant filled with
                   of trying to build grid-based power systems,           photovoltaic (PV) cells today are close cousins           such towers. Ydstie and Ph.D. student Christy
                   could serve multitudes of people faster without        of microchips, in that both are semiconductor             White have had the crucial role of modeling the
                   the environmental hazards of burning more              devices built from layers of doped crystalline            process and writing software for process control
                   fossil fuels.                                          silicon. (In a PV cell, the n-type layer -- doped         -- a key to cost control because, as Ydstie says,
                         But meanwhile, back home, there is basic         with trace materials to have free electrons,              "it's all about the yield."
                   research to be done.                                   and thus a negative electrical potential -- is

8
ENGINEERING NEWS




                             In B ur k i n a Fa s o , o n e o f A f r i c a 's

                   p oor est count r i e s , r u r a l v i l l a g e r s g e n e r a te

                              elect ri c i t y w i t h s o l a r u n i t s .
Th e Powe r o f P l astics                           A Qua ntum L eap                                     be made with dots of differing compositions
On the other hand, why use silicon at all?           Next to Porter's office in Roberts Hall sits Elias   or sizes, to capture and convert energy from
Even cheap silicon isn't very cheap. About           Towe, professor of Electrical and Computer           different parts of the solar spectrum. (Today's
half of the material is later lost, as sawdust,      Engineering and Materials Science and                uniform cells are limited by the fact that any
when sawing it into ultra-thin wafers for the        Engineering. And across the hall, in a clean         single material interacts only with light within
solar cells. Other less cost-intensive materials     room, stands the machine. Built by Towe,             a limited bandgap). Quantum-dot cells could
have photovoltaic properties, notably certain        students and staff with U.S. Army funds, the         aim for efficiencies of 40%, 50% or more.
organic polymers, or plastics.                       machine could be a set piece for a sci-fi film:      They probably wouldn't be sold as cost-savers
      The problem with plastic cells has been        a bright metal hulk with tubes protruding, it        at first; more likely as specialty items for
efficiency. They convert about 5% of the             looks as if it might teleport you if you're not      users needing high power from a small unit.
energy from the incident sunlight to electrical      careful.                                             But if advances in nano-fabrication can keep
energy, whereas silicon cells deliver about                It is in fact a molecular beam epitaxy         lowering the initial price, the efficiency factor
18%. That alone makes polymers impractical           (MBE) machine, used for teleporting atoms            would loom ever larger, making cells like these
for many uses; they need far more panel area         through a vacuum chamber -- at controlled            the cost choice for a widening range of uses.
to get the equivalent output. "However, I don't      velocities -- and "reconstituting them, one
think you should count them out just yet," says      atomic layer at a time, on the other side," as       The F uture
Lisa Porter, professor of Materials Science and      Towe puts it, to make quantum dots. These            Which solar technology will prevail in the
Engineering.                                         tiny clumps of matter, deposited on a substrate,     long run? The real strength will lie in all of
      Porter is teaming with Chemistry               contain a hundred or so atoms each. The size of      them coming together, says Lisa Porter: "As
professors Tomasz Kowalewski and Richard             a quantum dot "is on the order of a wavelength       in other fields, it's often not just one material
McCullough, dean of the Mellon College               of an electron," Towe says, so "it does not          or application that comes to the fore, but
of Science, on new polymer-based research            follow the classical rules of physics, but           many options for many needs." Also, just one
that could lead to breakthroughs in a field          behaves according to quantum mechanics." He          material can itself be quite versatile. Since
called bulk heterojunction cells. The work           smiles. "You can get very different properties       polymer panels would be cheap but not so
is very early-stage and too complex for brief        than the usual."                                     efficient, Elias Towe envisions vast numbers of
explaining. What is simple, and may be                     And this could open the door to making         them spread out to make solar farms in areas
reachable, is the goal: not silicon efficiency,      very high-efficiency solar cells. Currently,         where sun and space are plentiful -- think of
but low cost with efficiency that's good enough      some high-end cells for critical uses are            Arizona, or Burkina Faso -- while Porter notes
for a lot of applications. Says Porter, "if we can   made with gallium arsenide, in a standard            that thin, flexible sheets of polymer would be
get to ten percent, we are real players."            crystalline form. Quantum dots built from            ideal for portable solar panels: you could roll
                                                     gallium, arsenic and other atoms could do even       them up and carry them around, to drape where
                                                     better. And better yet, says Towe, cells could       needed. You could wear solar clothing.
                                                                                                                Carnegie Mellon's role in solar energy is
                                                                                                          growing, especially through its university-wide
                                                                                                          Center for Nano-enabled Device and Energy           9
                                                                                                          Technologies. Starting from a few researchers,
                                                                                                          a true research community is emerging, with
                                                                                                          synergies and spinoffs. Porter and Towe have
                                                                                                                                                              FEATURE

                                                                                                          been working together; their research has
                                                                                                          common threads in nanotechnology; Erik
                                                                                                          Ydstie is mixing the worlds of computer
                                                                                                          modeling and fieldwork overseas. The future
                                                                                                          of solar depends on communities of this type,
                                                                                                          and Porter speaks for all when she says, "I
                                                                                                          think the future looks bright."




                                                                                                          Elias Towe, the director for the
                                                                                                          Center for Nano-enabled Device
                                                                                                          and Energy Technologies, uses a
                                                                                                          molecular beam epitaxy machine
                                                                                                          to make quantum dots.