Today's PV market is a tremendous success story, with compound annual growth of more than 30% over the last 10 years, sales climbing to tens of billions of dollars, and the prospect of further growth fueled by a developing perfect storm of market forces - as long as the industry continues to achieve lower costs for solar-generated electricity.
PV now uses as much or more of the silicon feedstock supply as other electronic devices, contributing to the runup of prices for supplies of polysilicon and the scramble for long-term contracts. Today's best silicon PV modules deliver more than 20% conversion efficiency - a testament to engineering prowess and quality control and to the technology's maturity.
Silicon is one of the world's most studied materials, but the physics of the indirect band-gap absorber trumps any engineering. Roughly half of the cost of a photovoltaic module today is the cost of the silicon, and only incremental reductions at the shallow end of the learning curve are likely.
Researchers foresee a future where nearly ideal PV devices can be assembled molecule by molecule at the nanoscale. Copper indium gallium selenide (CIGS), an available and proven material, can spontaneously nanostructure itself through phase separation into the desired absorber and percolation network arrangement when manufactured properly.
Don't Hold The Mayo
CIGS is one example of the many potential advantages of "electronic mayonnaise" materials that we are now beginning to recognize and exploit. Somewhere between colloids and "hard" matter, they contain different phases of similar materials. Their electronic properties may depend as much on the spatial arrangement and electronic states of these phases as their proportions.
Working with these materials is as much metallurgy as chemistry. Compound semiconductors such as CIGS have several bonds and many ionic interactions that are very different from traditional silicon devices. Many of the old rules of thumb do not apply. Where variations in composition of one part per billion may render a silicon device inoperative, CIGS devices can vary by several percent and still function well, if the right nanostructures form within them. In fact, they may be the first widely deployed and profitable commercialization of nanotechnology.
Just as small-area integrated circuits opened up huge markets that couldn't be reached with discrete components and pc-board technology, large-area, high-efficiency, and low-cost PV integrated circuits will pave the way for a new energy mass market. Frost and Sullivan analysts recently estimated that thin-film PV could achieve a 25% market share by 2010.
This post-2010 PV market will make today's market pale by comparison. Despite all of its incredible growth, PV as an industry today is only where air conditioning was in the 1940s. Back then, if you were a well-heeled early adopter and wanted air conditioning for your home, you purchased a window unit designed for retrofit onto existing construction.
Up on the Roof
By 1960, few buildings and homes were designed without central air. Soon, mass-market central air conditioning dwarfed the sales of window units.
Today's PV system of discrete components retrofitted onto an existing roof is equivalent to the window unit of the 1940s, while tomorrow's building-integrated solutions will let buildings generate much of their own electric power through large-scale, high-efficiency electronic components that are part of the building envelope.
Making photovoltaic layers part of roofing materials or facades can eliminate the marginal cost of installation. Roofs could alternate areas of plantings and photovoltaics. Curtain wall systems could include opaque photovoltaic spandrels to use the large surface area of tall buildings to gather power. These components could be installed by roofers and glaziers and interconnected by electricians in the normal course of construction.
In the U.S., especially in the Sun Belt where much of today's construction is taking place, most homes and buildings could generate at least half of their electrical energy needs from PV components incorporated as roofing and cladding, avoiding spending billions of dollars every year to burn fossil fuels.
And, the potential growth for large-area thin-film electronics isn't limited to PV. Many other applications are emerging for displays, sensors, electrodes for batteries, fuel cells, capacitors, flexible electronics, and perhaps even high-temperature superconductors. There is a lot of electronic mayonnaise to spread around.