Thin-film solar cells hold the promise of low-cost, renewable energy source that could make fossil fuels obsolete, but thus far the cells' reliance on rare elements and expensive vacuum deposition manufacturing has impeded their progress. IBM Research has proposed solutions to both stumbling blocks by demonstrating a photovoltaic cell that uses common, abundant elements and is produced using an inexpensive nanoparticle- and spin-coat-based "printing" technique.

With 9.6 percent efficiency, the so-called kesterite solar cell beats the previous efficiency record of 6.8 percent for similar structures, bringing kesterite closer to the efficiency of established solar cell formulations, IBM said.

"Enough sunlight falls on the Earth in one hour to power the whole planet for a year, but solar cells currently contribute less than a tenth of a percent to our electricity supply, mostly because of their high cost and the short supply of key elements," said David Mitzi, IBM researcher and manager of photovoltaic science and technology. "We want to lower the manufacturing cost and increase the supply of the elements needed for thin-film photovoltaic devices."

Scanning-electron-microscope image of a thin-film photovoltaic struture consisting of a glass substrate onto which a kesterite compound is spin-coated.

Today, thin-film solar cells are based on chalcogenides such as copper-indium-gallium-selenium (CIGS) and cadmium telluride (CdTe). Indium and tellurium are rare elements, and the former is already in short supply because it is used to make transparent transistors.

By substituting more-abundant elements such as zinc and tin, IBM aims to lower the BOM cost for solar cells and enable mass production of kesterite photovoltaic devices, which are based on copper, tin, zinc, sulfur and selenium.

Indium is "in high demand by the flat-panel display industry," said Mitzi. "We replace the indium with zinc and tin, which not only lowers the cost but will enable the thin-film solar cell market to expand much more quickly, since these are more-abundant elements."

For CIGS, IBM had already developed a process that does not require expensive vacuum deposition. The company instead "prints" its films using solution-based mixing, spin coating and heat treatments.

To move from indium to tin and zinc, however, IBM had to invent a new deposition technique since zinc does not mix well into solutions. The new method reduces the zinc to nanoparticles that can be dispersed in a solution with the copper and tin, then spin-coated and heat-treated in the presence of sulfur or selenium vapor.

Last year, a Japanese research group at Nagaoka National College of Technology used a kesterite-based formulation to achieve thin-film cell that exhibited 6.8 percent efficiency. IBM says it was able to improve that by 40 percent.

The lab results demonstrate "that kesterites are a viable alternative to CIGS and CdTe, especially since our new deposition approach combines solutions and nanoparticle processing," said Mitzi.

IBM plans to optimize the architecture further, in hopes of raising its efficiency above 11 percent in order to beat both CIGS and CdTe formulations. The team will also experiment with alternative to spin coating, such as dip coating, spray coating and slit casting.