“MoOx has traditionally been exploited as a hole contact layer in organic electronics and photovoltaics but, so far, no-one had tried applying it to the most common semiconductor silicon,” explained team member Corsin Battaglia. “We have, and have found that it works even better than we expected.”

A solar cell generally consists of a light-absorbing semiconductor layer bound on one side by a selective contact that transmits holes and blocks electrons. The other side of the semiconductor layer is bound by a complementary selective contact, which, for its part, does the opposite – that is, blocks holes and transmits electrons. When sunlight falls on the device, these selective contacts act like sinks for either electrons or holes and each contact establishes a chemical potential gradient, which, in turn, generates a diffusion current.

Passivating surface defects

To reach high efficiencies in solar cells, the surface defects on the semiconductor surface must be passivated so that they do not hinder the flow of photogenerated electrons and holes. This allows the charge carriers to be transmitted through the contacts and so produce useful current before they recombine.

So-called silicon heterojunction solar cells, which have efficiencies of as high as 24.7%, are often touted as a model for this class of cell. These devices contain a thin hydrogenated amorphous silicon (a-Si:H) surface-passivating layer, thanks to which record open-circuit voltages of above 750 mV can be achieved.

However, the problem is that a-Si:H has a bandgap of only 1.7–1.8 eV and is highly defective itself. So, even if this passivating layer is only a few nanometres thin, it still parasitically absorbs light in the UV and visible parts of the solar spectrum. This problem is set to get even worse as the light-absorbing semiconducting layers in solar cells become ever thinner in an effort to improve open-circuit voltages. Although researchers have tried overcoming this drawback by replacing a-Si:H with wider-bandgap a-SiOx:H or a-SiCx:H, such strategies have met with little success.

Efficient hole injection and extraction

Now, a team led by Ali Javey at the University of California at Berkeley and the Lawrence Berkeley National Lab, has come up with a radically different hole contact scheme for n-type silicon heterojunction solar cells based on thermally evaporated molybdenum trioxide (MoOx) thin films. MoOx has a higher work function than any elemental metal, which places its Fermi energy level closer to the low-lying valence band states of semiconductors, which means that holes can be injected or extracted into these materials very efficiently.

The team began by fabricating an unpassivated MoOx/crystalline silicon (c-Si) solar cell that had an efficiency of 14.3%. “By then inserting a a-Si:H passivating layer between the oxide contact and the silicon absorber, MoOx/a-Si:H/c-Si, we obtained an efficiency of 18.8%,” said Battaglia. “With a bandgap of 3.3 eV, MoOx contacts allow for an impressive current gain of 1.9 mA/cm2 while maintaining the same open-circuit voltage as a standard silicon heterojunction cell.

“Our technology is a general one and could readily be implemented in existing photovoltaic production lines based on, for example, the high-efficiency HIT cell concept pioneered by Sanyo (now Panasonic),” he told nanotechweb.org. “Many other companies around the world are working on this concept and photovoltaic modules with our new MoOx hole contacts could hit the market within less than a year.”

The work, carried out in collaboration with the PV-Lab at the Ecole Polytechnique Fédérale de Lausanne (EPFL) in Switzerland, is reported in Applied Physics Letters 104 113902.