Prussian blue was born in early 18th-century Berlin, first as a cheap dye for Prussian army uniforms, later entering Van Gogh's *The Starry Night* and Hokusai's *The Great Wave off Kanagawa*. On September 25, it gained a new identity: in *Science* vol. 393, p. 1329, a team led by Academician Zheng Nanfeng and Professor Wu Binghui of Xiamen University used it to tackle the most stubborn problem in perovskite solar cells — runaway ion migration. The paper's title is refreshingly direct: *Prussian blue regulates ion dynamics in perovskite solar cells*.
Collaborators include Jinan University and Tsinghua University, Xi'an Jiaotong University and Sweden's Uppsala University, plus two Chinese Academy of Sciences institutes. The project's starting point was unexpectedly modest: a seminar class for first-year undergraduates in inorganic chemistry.
🏚️ Perovskite's Achilles' Heel
Perovskite is the fastest-improving photovoltaic technology of the past decade: power conversion efficiency has climbed from 3.8% to over 26%, and it can be processed from solution at low temperature like printing a newspaper. Its trouble is written into the material's nature: perovskites are soft ionic semiconductors whose lattice ions migrate and redistribute under light, heat, and electric fields. A *Strait Herald* report offered a fitting analogy: the lattice is an apartment building, ions are its residents, and light and fields are external vibrations — shake the building and residents start moving out, loosening the structure.
As residents relocate, local electronic environments change, defect evolution accelerates, and device performance decays. This is the biggest obstacle to commercialization.
Worse, the problem spans the material's entire life: structural defects left during thin-film crystallization and ion migration during device operation are causally linked stages of one continuous process, yet traditional research often treats them separately — crystallization over here, operation over there.
🎨 A Seminar's Three-in-One Toolkit
Prussian blue's structural chemistry offers exactly a three-in-one kit: iron ions bridged by cyanide ligands into an open framework, regularly arranged cavities within that framework, and iron centers in both divalent and trivalent states. Three properties taught as textbook exercises map onto three tasks for perovskites.
First, guiding crystallization. Prussian blue's lattice parameters closely match those of perovskite. Added nanoparticles act as scaffolding in the precursor solution, providing heterogeneous nucleation sites. In-situ photoluminescence and X-ray scattering experiments show more orderly nucleation, intermediate-phase transformation, and grain growth, yielding films with stronger preferred orientation and lower local strain.
Second, regulating defects. Perovskites produce abnormal-valence species during fabrication and operation — zero-valent lead and zero-valent iodine — which worsen non-radiative recombination losses and participate in decomposition. Prussian blue's Fe(II)/Fe(III) centers act as a redox buffer, pulling these bad actors back into line while its own framework stays intact.
Third, suppressing ion migration. The regular cavities in the open framework can house ions attempting to move. In-situ time-of-flight secondary ion mass spectrometry and Kelvin probe force microscopy show that long-range migration — especially of cesium ions — is significantly suppressed, and field-induced local potential changes are greatly reduced.
| Stage | What Prussian blue does | Evidence | |---|---|---| | Crystallization | Heterogeneous nucleation sites; matched lattice parameters | In-situ photoluminescence; X-ray scattering | | Defects | Fe(II)/Fe(III) buffers abnormal-valence species | Framework itself remains stable after buffering | | Operation | Cavities house migrating ions, chiefly cesium | TOF-SIMS; Kelvin probe force microscopy |
📏 Numbers and the Outdoors
Device level: small-area cells exceeded 26% efficiency in both conventional (n-i-p) and inverted (p-i-n) architectures. Scaled up, a 6 cm × 6 cm conventional module reached 23.4%, and a 30 cm × 30 cm inverted module earned a third-party certified efficiency of 22.9%. Thirty centimeters square — a bit larger than an A4 sheet — is a size now entering production-line discussions.
Four stability tests were run: damp-heat aging, continuous maximum power point tracking, accelerated aging at different temperatures, and the most rigorous one — a perovskite module and a commercial silicon cell monitored side by side outdoors for 151 days. The result: no obvious degradation trend of the perovskite module relative to the silicon reference. Five months is only a beginning for a cell promised 25 years of life, but stability relative to a reference speaks more loudly about suppressed ions than absolute duration does. [inference]
🔄 A Shift in Thinking
This work's significance lies less in another efficiency number than in the approach. Traditional perovskite defect research is patch-the-hole: find a defect, design a passivation molecule to fix it — static, passive, one problem one solution. The Prussian blue approach treats film formation and device operation as one continuous process, with a single material acting simultaneously on crystallization, defect evolution, and ion migration.
The design space this opens: instead of hunting for one passivation molecule, seek a class of open-framework materials combining structural guidance, redox buffering, and ion sequestration. Prussian blue is merely one representative — accidentally synthesized 300 years ago, its three structural properties reread 300 years later. [assessment]
The line from a freshman seminar to a *Science* paper is also worth noting: the flow between foundational teaching and frontier research is not always one-way. Sometimes returning to the most basic structural chemistry principles lets you see further.
🧭 Next Steps
Two test points are obvious. First, larger areas and longer outdoor durations — 151 days is too short for commercial promises. Second, whether the open-framework strategy transfers to other perovskite compositions and other soft ionic semiconductors. Jinan University's module validation path, backed by industrial support, suggests the distance from mechanism to production line is being taken seriously; what remains is time and scale.
---
Sources: *Science* 393, 1329–1334 (2026), DOI 10.1126/science.aef5487 (online 2026-09-25); Xiamen University domestic cooperation office press release; Jinan University New Energy Technology Institute press release; Strait Herald (2026-10-01).
Caveats: Efficiency and stability figures come from journal pages and institutional press releases and were not independently checked against the paper's full text; microscopic evidence for the three mechanisms was cross-confirmed via the Xiamen University release and *Strait Herald*; collaborator list follows the Xiamen University release; the climate conditions of the 151-day outdoor monitoring were not disclosed.