• Chinese Academy of Sciences researchers used Prussian Blue nanoparticles to push perovskite solar cells to 26.9% efficiency, in a study published in Science.
  • The same treatment dramatically improved durability: minimodules kept over 90% of output after 2,500 hours at 85°C and 85% humidity.
  • A 30×30 cm submodule hit a certified 22.9%, and a five-month outdoor test showed no decline versus a commercial silicon reference.

The advance pairs a record power conversion figure with durability evidence the technology has always missed. Structural templates control crystallization, and accelerated ageing projects about seven years to 80% output at 35°C.

Close-up of thin-film perovskite solar cells
National Renewable Energy Laboratory / Wikimedia Commons, public domain

pv magazine’s reporting details the advance: the Prussian Blue (PB) nanoparticles guided the crystallization of formamidinium lead iodide (FAPbI3) films, suppressing unwanted intermediate phases, favoring the photoactive α-phase, improving crystal alignment, and reducing lattice strain. Inverted (p-i-n) cells reached 26.9% efficiency (independently certified at 26.2%), up from 26.0% for untreated devices, while conventional (n-i-p) devices improved from 24.2% to 26.1%. pv magazine Australia confirmed the same results.

The 2,500-hour damp-heat result

Perovskite solar cells have been breaking efficiency records for a decade; the technology already rivals silicon in the lab. What has held it back is durability: perovskites degrade under heat, humidity and light, suffering from defect formation and ion migration that erode performance. A record efficiency without stability is a curiosity. Our deep dive on the perovskite durability problem explains the degradation mechanisms and what fixing them will take.

That’s what makes the Prussian blue solar cell results genuinely significant. The nanoparticles increased the activation energy for ion migration from 0.89 eV to 1.25 eV. In plain terms, they made it much harder for the ions inside the material to wander out of place, which is a key degradation pathway. Conventional minimodules retained over 90% of their initial efficiency after 2,500 hours at 85°C and 85% relative humidity (brutal accelerated-aging conditions) and over 80% after 2,200 hours of continuous operation. Accelerated aging projects a T80 lifetime (the point where output falls to 80% of initial) of about seven years at 35°C, though the researchers note that figure is extrapolated.

From lab cells to real modules

Another reason to take this result seriously: it scales. The team built a 30 cm × 30 cm inverted submodule with a 756 cm² aperture area that achieved a certified 22.9% efficiency, still excellent at that size, where perovskites typically lose performance.

The scaling penalty is worth quantifying. The record cell’s 26.9% drops to 22.9% on a 756 cm² submodule, a 4-point loss moving from postage-stamp to pan-sized. Compare that with the tandem race we track separately: LONGi’s 35.5% small-area cell loses 2.5 points scaling to 260.9 cm². The single-junction perovskite’s larger penalty suggests its crystallization tricks are harder to keep uniform over area, a problem that bites this technology hardest, because its entire commercial case rests on cheap, large-area processing.

And in a five-month outdoor test in Zhuhai, China, the devices showed no discernible performance decline versus a commercial silicon reference device.

Outdoor testing is the credibility check the perovskite field has needed. Lab aging protocols can be gamed, intentionally or not; five months under real sun, rain and heat in subtropical China is much harder to dismiss. Corresponding author Chong Liu said the structurally compatible PB framework “guides perovskite crystallization, mediates the conversion of abnormal lead and iodine species, and restricts A-site cation redistribution”: the materials science behind the performance.

Prussian Blue perovskite results from the CAS study in Science (figures via pv magazine)
Test Result
Inverted (p-i-n) cell efficiency 26.9% (certified 26.2%), up from 26.0% untreated
Conventional (n-i-p) cell efficiency 26.1%, up from 24.2% untreated
30 × 30 cm submodule (756 cm² aperture) Certified 22.9%
Damp heat: 2,500 h at 85°C / 85% RH Over 90% of initial efficiency retained
Continuous operation: 2,200 h Over 80% retained
Projected T80 lifetime at 35°C ~7 years (extrapolated)
Outdoor test, Zhuhai (5 months) No discernible decline vs commercial silicon reference
Ion migration activation energy 0.89 eV → 1.25 eV
LONGi tandem comparison (small area → module) 35.5% → loses 2.5 points at 260.9 cm²

Where perovskites go from here

The broader context is a technology approaching an inflection point. Perovskite-silicon tandems, which stack a perovskite layer on conventional silicon, have already pushed past 30% in the lab, and manufacturers are racing to commercialize. The remaining barrier has always been the same: proving a perovskite device can survive 25 years on a rooftop or in a solar farm. Nobody is claiming this paper solves that. Seven projected years is not twenty-five. But it attacks the problem at its root, in the crystal structure itself, rather than papering over it with better encapsulation.

There are still open questions. Lead content remains a concern for regulators and some markets, though encapsulation and recycling research is advancing in parallel. Manufacturing the nanoparticle treatment at scale, uniformly, and cheaply is unproven. And independent replication by other groups, the gold standard in materials science, will take months.

What perovskites offer that silicon cannot

Step back from the lab and the stakes become clear. Silicon solar is a mature, extraordinary technology, but it is approaching its practical efficiency limits, and the head-to-head comparison with silicon shows further cost reductions increasingly come from scale rather than science. Perovskites offer something silicon cannot: a fundamentally different material that can be processed at low temperatures, printed on flexible substrates, and stacked on top of silicon to harvest parts of the spectrum silicon misses.

That is why every major solar manufacturer and a flock of startups are invested in the technology’s success. The prize is a step-change in panel efficiency without a step-change in manufacturing cost: cheaper electricity from the same sunlight. Breakthroughs like the Prussian Blue treatment don’t deliver that prize by themselves, but each one that simultaneously lifts efficiency and durability shortens the road. The perovskite story has been “five years away” for a decade; results like this are how “five years away” eventually becomes “shipping.”

Replication and licensing: the two signals to watch

Watch for two things: replication studies from other labs, and whether any manufacturer licenses the approach for pilot production. A perovskite solar cell record that arrives with harsh-environment stability data, certified submodule results, and outdoor testing is the strongest signal yet that the technology’s defining weakness, durability, is yielding to materials engineering. If the trajectory holds, the solar panels of the 2030s may owe their performance to a pigment first synthesized in 18th-century Berlin.