Researchers looking for ways to reduce the cost of solar photovoltaics have recently begun studying perovskites. These are inexpensive materials with the same crystal structure as calcium titanium oxide, from which they draw their name. To date, researchers have demonstrated a series of photovoltaic cells that incorporate perovskites in relatively complex device architectures that generally require complex processing steps and high temperatures. Now, Mingzhen Liu and colleagues at the University of Oxford have demonstrated a simple planar perovskite-based solar cell that avoids these issues and achieves over 15% efficiency. They reported their results in the September 19 issue of Nature (DOI: 10.1038/nature12509; p. 395).

The active layer in photovoltaic cells has three essential functions: absorption of light, generation of free carriers, and transport of those carriers to their respective contacts. Researchers have recently investigated the potential of organometallic trihalide perovskites to play one or more of these roles in nanostructured solar cells, although in all of these cases, other materials were also present to serve as a structured scaffold. The Oxford researchers hypothesized that it might be possible to build a high-efficiency device using perovskites with a simple planar heterojunction architecture, eliminating the scaffolding and using the perovskite material to serve all three essential functions of the active layer. This would put the device in the same structural category as industrially relevant silicon and thin-film solar cells, and therefore be an important step forward.

To test this hypothesis, the researchers fabricated devices based on a fluorine-doped tin oxide (FTO)-coated glass (the device anode), followed by a spin-coated layer of compact TiO₂ as an electron-sensitive contact. Next, they deposited a layer of the mixed halide perovskite CH₃NH₃PbI_3–x Cl _x using two different methods: two-phase vapor deposition, and spin-coating. Both approaches used methylammonium iodide and lead chloride as the organic and inorganic precursor salts, respectively. Finally, the researchers spin-coated a p-type hole conductor—2,2´,7,7´-tetrakis-(N,N-di-p-methoxyphenylamine)9,9´-spirobifluorene(spiro-OMeTAD)—on the perovskite layer, followed by a silver cathode to complete the device.

Using x-ray diffraction, the researchers found good phase purity in the perovskite layer for both vapor-deposited and spin-coated devices. However, the morphologies were quite different; the vapor-deposited film had a uniform thickness of 330 nm, while the spin-coated film was highly structured, varying between 0 nm and 465 nm. When tested under simulated sunlight, the most efficient vapor-deposited device displayed 15.4% efficiency, while the best solution-processed device showed a 8.6% efficiency. An ensemble of 12 vapor-deposited devices displayed an average efficiency of 12.3%, demonstrating some variability.

These results are highly encouraging for the potential use of perovskite materials in solar cells, particularly because these devices achieve high efficiencies with a simple planar heterojunction architecture and because vapor deposition is compatible with existing silicon and thin-film solar-cell processing infrastructure. The researchers also said that vapor-deposited perovskites might be able to play an important role as a “top cell” in multijunction silicon or copper indium gallium diselenide devices, improving their overall efficiencies. It is clear that perovskites are worth watching for high-efficiency, low-cost solar power.