Vacuum flash helps perovskite solar cells beat 20 per cent efficiency

A team at the École Polytechnqie Fédéral de Lausanne (EPFL) led by organic solar cell pioneer Michael Grätzel has grown perovskite crystals whose efficiency exceeds 20 per cent. Perovskites are crystals with a structure similar to a calcium titanium oxide salt found naturally in Russia; they can be grown under mild conditions without the need for vacuum and high temperatures associated with silicon film manufacture.

Grätzel’s team has developed a method for making a perovskite film that incorporates a low-pressure process they refer to a ‘vacuum flash’. The process starts with a solution of caesium-containing metal halide salts coated onto a glass cell treated so that it conducts electricity. Addition of caesium improves the thermal stability of the perovskite, making it more suitable for commercial application.

Normally, this solution is let to dry so that the salts crystallise out; but this tends to produce an irregular film of crystals with mall grain sizes; these have low photovoltaic efficiencies. The EPFL team has had success with spinning the glass cells while the solution is still wet; this removes some of the excess liquid, giving a flatter crystal film with larger grain sizes. This is accompanied by dripping a liquid known as an ‘anti-solvent’ onto the cell, which encourages crystallisation; unfortunately, the crystals produced tend to be impure, and the anti-solvents are toxic.

Attempting to further improve this effect, the team vacuum flashed the wet film, applying a mild vacuum to the cell for 20 seconds. This, they explain in a paper in Science, selectively removes volatile components of the solution, while also creating seed crystals that lead to the formation of a very smooth, regular crystalline film. It also does away with the need for anti-solvent.

The method makes perovskite cells larger than a square centimetre with an average PV conversion efficiency of 19.6 per cent and a maximum efficiency of 20.6 per cent, matching that of thin-film silicon solar cells. “How high can we go with efficiency? Theoretical efficiencies are around 30 per cent. While we might not get to theoretical efficiencies, we can still go much higher than we have currently,” said Grätzel. "We can very simply get remarkable results compared to anti-solvent methods. We are now excited about scaling up to make bigger modules."

One possibility for the technique is to make hybrid cells with the perovskite grown on top of a conventional silicon cell; high energy light would be absorbed by the top layer, lower energy by the silicon. Such cells would have a theoretical maximum efficiency of 44 per cent.