A new preprint reports an all-perovskite tandem solar cell whose two light-absorbing layers were both made by sequential thermal evaporation. The researchers mapped how deposited materials turn into working crystals and found that the two layers need different processing conditions. That finding addresses a manufacturing problem: a recipe that works for one absorber may leave the other unevenly mixed.
Sequential thermal evaporation deposits starting materials as separate solid layers under vacuum, then heats them to promote mixing and crystallization. It provides control over layer thickness without the solvents used in many high-efficiency perovskite cells. To see where conversion stalled, the researchers stopped samples at different times and examined their structure and chemical composition.
A tandem stacks absorbers with different bandgaps, the minimum light energy each can absorb, to collect different parts of the spectrum. In the wider-bandgap layer, the team found that bromide and iodide mixed before the main perovskite crystal structure formed. The amount of deposited lead bromide determined the final bandgap; adding more lead iodide did not reliably change it because some could remain unreacted.
The narrower-bandgap tin-lead layer had the reverse problem: a lead-rich crystal structure formed before the materials had fully mixed. The authors attribute the remaining uneven tin distribution to that early structure restricting further movement through the film. Even after an hour of heating, the surface and buried interface had different tin fractions. Making crystals faster would therefore not solve both layers' processing problems.
Using those findings, the team stacked a wider-bandgap top cell over a narrower-bandgap bottom cell. The best device reached 19.2% efficiency in a current-voltage scan over a 0.0616-square-centimeter test area; a 60-second check at its maximum-power setting stabilized at 18.5%. The result was not independently certified, and the authors did not correct for differences between the test lamp's spectrum and the reference spectrum. It establishes a laboratory fabrication route, while solution-processed all-perovskite tandems cited by the authors exceed 30%.
The durability evidence is also limited. One sealed device retained at least 80% of its initial output for more than 130 hours under continuous light in nitrogen at 54 to 57 degrees Celsius. A separate heated-storage test, conducted in the dark, reached an average of 80% after 1,200 hours at 65 degrees Celsius for six devices; two others were excluded after losing external silver contacts. The longer result measures storage stability, so it cannot establish 1,200 hours of electricity production.
Sources
- Paper abstract and submission record
- Full paper and supplementary information
- Illustrative perovskite module; it is not the tandem device used in the study. Image source: Nejaby / Wikimedia Commons, licensed under CC BY-SA 4.0.
