Why are solar cells made of silicon?

By Marie Mayer

The solar industry — a hot topic as of late due to subsidies — was built on silicon.  But we rarely discuss what is so special about the gold standard. We hear more frequently about business opportunities for “thin film” photovoltaics made of compounds like cadmium telluride (First Solar) or copper-indium-gallium-diselenide (Nanosolar, Miasole and dare I mention them…Solyndra) .  As we look to new materials systems, both in research and in industry, to solve all of our solar problems, it’s quite relevant to recall what was so special about the first one—the material that made up roughly 75% of installations in 2011.

In short, we use silicon for solar cells because we’re good at making silicon devices, not necessarily because silicon is the best material for solar cells. It’s no coincidence that we live in Silicon Valley.  Integrated circuits built from silicon-based components are the foundation of the multi-billion dollar microelectronics industry.  First patented in the south bay by Robert Noyce of Fairchild Semiconductor (who went on to found Intel), the commercial success of the integrated circuit has led to the production of more than 1 billion transistors annually per person worldwide!  One of the main reasons that silicon is the semiconductor material of choice in microelectronics is that it forms a unique oxide on the surface when heated to high temperatures.  This facilitates device fabrication for two reasons: (1) it neutralizes defects on the silicon surface and (2) it allows for straightforward planar processing.

To turn the clock back a bit further, the first photovoltaic effect (electricity generated by light) was observed in 1839 by Edmond Becquerel using wet chemistry.  Independently, silicon was one of many materials used in pre-WWII radio equipment—a sharp metal point jabbed into unrefined silicon formed a crude version of a diode, an important element in radio receivers.  These two developments led to the accidental discovery of the silicon solar cell in 1940, when, at Bell Telephone Labs, Russell Ohl observed the photovoltaic effect in a half-purified bar of silicon.  From serendipity, research on the processing and physics of silicon grew until the integrated circuit crystallized its role in history.

Today, we can grow near-perfect crystals of silicon at a commercial scale.  Solar cells made from these crystals allow solar-generated power to reach the external electrical circuit.  The only problem is that silicon does not absorb sunlight as efficiently as other materials, so silicon solar cells need to be 10-100(s) of times thicker than thin film solar cells.  Microelectronics grade silicon is too expensive for economical large area coverage, but totally unrefined silicon solar cells don’t work.

To lower the costs, research is underway to design silicon solar cells that operate using silicon of lower purity.  Here at Berkeley, Professor Peidong Yang’s group has investigated nanostructured Si solar cells, which work well even when low purity silicon is used.  Berkeley graduate Prof. Tonio Buonassisi at MIT continues his work on “dirty” silicon, or unpurified material, and “black” silicon, which gets its name from enhanced light absorption that changes the color of cells from blue to black.  Furthermore, researchers in Dr. Wladek Walukiewicz’s group at LBNL have designed devices that combine traditional silicon with other thin film materials to optimize performance.  (This is, of course, not the end of the story, but space constraints apply!)

The takeaway: silicon is the preeminent solar cell of the day because it is the first semiconductor we learned to commercialize.

Do you have a pressing solar or cleantech technology question?  E-mail Marie at mamayer@berkeley.edu and your question can be the subject of future columns.