A new generation of perovskite-silicon tandem solar cells can break the theoretical 30% efficiency barrier, Science reports. Two research groups present different molecular methods to pull this off. The biggest challenge is the limited stability of these cells. ‘Extending the lifespan is now more important than a further increase in efficiency.’  

The more efficient you can convert the solar spectrum into energy, the higher the overall efficiency of a solar cell. Improving that efficiency can be achieved via a tandem solar cell, which consists of two layers. The top layer absorbs high-energy photons and the bottom layer absorbs low-energy photons. Two simultaneously published, but independent, papers in Science describe two different methods to make a perovskite-silicon tandem solar cell with an efficiency of more than 30%. In comparison, the theoretical maximum efficiency of a silicon solar cell is around 29%. Both groups used molecular methods to counteract the losses between perovskite and the cell’s upper conductive layer. 

In July 2022, Xin Yu Chin’s team (Ecole Polytechnique Fédérale de Lausanne) already reported a certified conversion efficiency of 31.25% with their monolithic perovskite-silicon tandem solar cell. They now describe the method they used: chemical vapour deposition of one of the two perovskite precursor materials in a controlled vacuum, followed by deposition of the second perovskite precursor from a solution. This is followed by conversion to the perovskite phase.  

Losses at the interface  

A silicon solar cell is not flat, which makes it difficult to apply an additional layer, as co-author Julian Steele of KU Leuven explain. ‘The silicon layer has a micropattern of pyramids, the size of a tenth of a human hair, to prevent reflection. You want to preserve that pattern when applying a conformal layer. But if you only deposit from a solution, the micropattern will be filled, resulting in another flat layer.’ They decided on chemical vapour deposition for the first part of the reaction. ‘This way, we create a uniform template layer on the silicon. In the solution step, the material bonds only with that template layer and the pattern is preserved.’ 

Unfortunately, that approach hampered the attachment of the conductive layer (fullerene, C60), which in turn reduced the cell’s overall performance. The team solved this by adding a phosphonic acid-based additive, which flows to the surface of the perovskite layer during crystallisation. Steele has a background in crystallography and he used X-ray spectroscopy to follow the growth of the perovskite crystal in situ. He watched the first and second layers merge into one. ‘The additive slowed down crystal growth. Crystal quality is increases with slow growth, as this creates larger domains with fewer defects. And that resulted in reducing the performance loss at the perovskite-C60 interface.’  

Meanwhile at the Helmholtz-Zentrum Berlin für Materialien und Energie, Silvia Mariotti and colleagues came up with another molecular method to counteract the losses at the perovskite-C60 interface. They used piperazinium iodide, an ionic liquid, for a tandem cell with a planar silicon substrate and a perovskite layer that they produced via solution. This substance contains electron acceptor groups and donor groups that interact with the surface defects of the perovskite. In 2002, they reported an efficiency of 32.5% using this approach.   

Chemically vulnerable  

The true Achilles’ heel of their solar cells is lifetime, says Steele. ‘For real-world applications, you need to stay as close to the industrial standard as possible. We started with the industrial standard for silicon solar cells and put an extra layer on top of that.’ But to become commercially attractive, solar cells have to score well on lifetime in addition to efficiency and cost. ‘We do well on the latter two, going from 25% to 31% efficiency with relatively cheap materials. Especially for small surfaces like roofs or even vehicles, that is a significant increase. But lifetime is always an issue with perovskites. Solar cells should have a 25-year warranty.’

Silicon is resistant to oxygen, as a thin oxide layer forms in the air that protects against further oxidation, but perovskites are chemically more vulnerable. ‘Chin’s cells stay functional for at least a year in a nitrogen cabinet, but right now, they will not make it outside in the wild.’ Steele says universities are working each other up in the race for efficiency records, but relatively few researchers are working on durability. ‘Actually, that’s more important now, but it just seems less sexy than going for high efficiency.’  

According to Sjoerd Veenstra, programme coordinator perovskite solars cells at TNO, the question remains whether these solar cells will make it to the market, but their much higher efficiency makes a strong case. ‘Stretching the efficiency of silicon cells further is very difficult. Tandem technology is the next step.’ Besides stability requirements, the cells also need to be scaled up in size. Veenstra: ‘They can produce cells of a few square centimetres, but you have to move to wafer scale: hundreds of square centimetres, so a few orders of magnitude higher.’  

Veenstra expects this kind of cell to conquer smaller markets first, but whether it will become a standard technology remains to be seen. ‘Silicon has had sixty years to evolve, so it is much more .’ But he finds Chin’s and Mariotti’s findings very exciting. ‘They can significantly increase efficiency and reduce costs, generating the same energy with a smaller surface area. That could make integration into the built environment feasible.’ 

Efficiency race

His colleague Valerio Zardetto says it is normal for these kinds of innovations not to concern themselves with longevity. ‘Universities are pushing the efficiency race. Lifetime experiments take a lot of time and are more expensive, which is not attractive to those institutions.’ You have to be able to reproduce a solar cell 1000 times to get statistically relevant data and the stress factors you evaluate have to be relevant to the real world situation. Zardetto: ‘It’s more complex and takes longer. I think there is a job here for institutes specialized in applied research.’  

Veenstra and Zardetto appreciate how both research groups devised their own way to mitigate the biggest loss - recombination at the perovskite electron transfer boundary. ‘It’s like cooking. You can use different materials and steps to achieve your desired result’, says Zardetto. Veenstra adds: ‘One group uses a textured silicon substrate to capture as much light as possible. To build a good perovskite cell on such a substrate, they have to modify the deposition process; the other group opts for a simple deposition process on a wafer with a flat surface and then has to put in additonal work to capture as much light as possible again. That gives the industry some freedom to choose a route.’ 

Xin Yu Chin, et al., Interface passivation for 31.25%-efficient perovskite/silicon tandem solar cells, Science (2023)

Silvia Mariotti, et al., Interface engineering for high-performace, triple-halide perovskite-silicon tandem solar cells, Science (2023)