For more than a decade, the standard approach to protecting perovskite solar cells has been planar interface engineering. Researchers deposit thin, flat interlayers of organic molecules, polymers, or inorganic salts between the perovskite absorber and the adjacent charge-transport layer. These interlayers serve two purposes: they chemically passivate dangling bonds and defects at the interface, reducing the recombination of charge carriers, and they act as a kind of glue, interlocking the two layers mechanically. The trouble is that these goals pull in opposite directions. A thicker, denser interlayer bonds the layers together more robustly, but it also adds an insulating barrier that impedes the flow of electrons or holes across the junction. A thinner interlayer lets charges pass freely but offers little mechanical reinforcement. The result has been a persistent trade-off, in which gains in stability come at the cost of efficiency, and vice versa.

The Hong Kong team, led by Yuanyuan Zhou with Pengfei Guo and Wenjian Yu as co-first authors, sidestepped this trade-off by abandoning the flat interlayer altogether. Instead of building a wall between the perovskite and the charge-transport layer, they built bridges through it. Their trans-interface engineering creates vertically aligned, mechanically tough nanolinks that penetrate into both layers simultaneously, anchoring the interface from both sides at once. Because these molecular links are discrete and vertically oriented rather than continuous and planar, they provide mechanical reinforcement without blanketing the interface in insulating material. Charge carriers can still travel across the junction through the unoccupied regions between the nanolinks, while the links themselves hold the two layers together under the thermal and mechanical stresses of operation.

The chemistry behind the nanolinks is a two-step molecular handshake. First, the researchers pre-incorporated two different organic linkers into the device stack: a triple organic linker embedded within the perovskite layer, and a single organic linker distributed in the charge-transport layer. Each linker is designed to integrate into its host layer without disrupting its function. Then came the trigger: the team applied radicals derived from azobisisobutyronitrile, a common free-radical initiator better known in polymer chemistry as AIBN. These cyanoisopropyl radicals initiated polymerization of the two linkers within their respective layers, and, crucially, also bonded the two linker populations to each other vertically at the heterointerfacial contacts. The result is a covalently stitched interface, with polymerized molecular tethers running from the body of the perovskite layer, across the junction, and into the body of the charge-transport layer.

Mechanical testing showed that this stitched interface outperformed other reported interfacial approaches by a wide margin. This matters because mechanical failure, not chemistry alone, has emerged as a leading cause of perovskite device degradation. Metal halide perovskites are soft, brittle ionic crystals, and the repeated thermal cycling, humidity exposure, and mechanical stress of real-world operation cause delamination, cracking, and grain-boundary failure that begin precisely at the heterointerfaces. Previous work from other groups has explored interfacial toughening with self-assembled monolayers, chiral-structured heterointerfaces, and interpenetrating networks, but the trans-interface approach is distinctive in reinforcing the interface from within both layers rather than merely coating it from one side. By suppressing the mechanical failure modes that initiate degradation, the nanolinks address degradation at its root rather than treating its symptoms.

The photovoltaic results are striking. Devices built with the TIE-treated interface achieved power conversion efficiencies of up to 27.04 percent, with an independently certified value of 26.67 percent, placing them among the most efficient perovskite solar cells ever reported. But efficiency records in this field are common; what has been rare is stability delivered reproducibly. Under maximum-power-point tracking at one-sun-intensity illumination, the standard operating condition used to assess how a cell performs while actually generating electricity, 92 percent of the tested devices reached T80 lifetimes of 1,400 hours, meaning they retained at least 80 percent of their initial performance for nearly two months of continuous simulated sunlight. That statistical consistency across a large population of devices is arguably the more important headline, because industrialization depends not on a single champion cell but on thousands of cells that all behave the same way.

The reproducibility stems largely from what the nanolinks prevent. In conventional devices, tiny variations in interface quality from sample to sample translate into wildly different degradation rates, so one cell might last ten times longer than its nominally identical twin. By mechanically reinforcing every interface in the same way, trans-interface engineering narrows that distribution. The team attributes the high reliability chiefly to the suppression of mechanical failure during operation: when the interface cannot delaminate or crack, the chemical degradation pathways that follow mechanical damage never get started. This reframing of stability as a mechanical problem as much as a chemical one echoes a growing consensus in the field, articulated in recent reviews of the mechanical reliability of metal halide perovskites, that fracture and delamination deserve the same attention traditionally reserved for ion migration and moisture attack.

The implications extend well beyond a single efficiency chart. Perovskite solar cells are the leading candidates for next-generation photovoltaics because they can be made cheaply, at low temperatures, on flexible substrates, and in tandem configurations stacked on silicon to break through the efficiency limits of single-junction cells. But investors and module manufacturers have been wary of a technology whose lab champions fade within weeks and whose production batches vary unpredictably. A strategy that simultaneously boosts efficiency, extends lifetime, and tightens the statistical spread of device performance directly addresses the three objections that have kept perovskites out of mass production. The fact that the TIE process relies on well-understood radical polymerization chemistry, compatible with existing solution-processing fabrication, suggests it could be integrated into manufacturing without exotic equipment.

Perhaps the most far-reaching claim in the paper is that trans-interface engineering is not limited to photovoltaics. Multilayered optoelectronic devices of every kind, including organic light-emitting diodes, photodetectors, and perovskite-based transistors, suffer from the same fundamental weakness: dissimilar layers stacked on top of one another form fragile heterointerfaces that fail under stress. A generalizable method for growing vertically aligned molecular links across any such junction, decoupling mechanical reinforcement from charge transport, offers a design principle for the entire field of layered electronics. A US patent has been filed based on the work by Zhou, Guo, and Yu, signaling commercial interest in translating the concept. If the approach survives the transition from laboratory cells to full-size modules and outdoor field testing, the molecular stitching demonstrated here may be remembered as the moment perovskite solar cells stopped being a laboratory marvel and started becoming a product.

Subject of Research: Trans-interface engineering of perovskite solar cell heterointerfaces for efficiency and stability

Article Title: Trans-interface engineering for reliable perovskite solar cells with reproducible stability

Article References: Guo, P., Yu, W., Hao, M., Wang, K., Duan, T., Li, S., & Zhou, Y. (2026). Trans-interface engineering for reliable perovskite solar cells with reproducible stability. Nature Photonics. https://doi.org/10.1038/s41566-026-02007-w

Image Credits: AI Generated

DOI: 10.1038/s41566-026-02007-w

Keywords: perovskite solar cells, trans-interface engineering, nanolinks, power conversion efficiency, operational stability, interface engineering, radical polymerization, mechanical reliability, charge transport, photovoltaics, heterointerfaces, reproducibility

The elegance of the design lies in pairing two materials with complementary chemistries. Manganese dioxide is a veteran of pseudocapacitor research: it is cheap, abundant, environmentally benign, and boasts a theoretical capacitance of roughly 1,370 farads per gram thanks to reversible Mn3+/Mn4+ redox reactions at or near its surface. But it has well-known Achilles heels. Its bulk is electrochemically nearly inaccessible, its electrical conductivity is poor, and it gradually dissolves and aggregates during repeated cycling. Prussian blue, the deep-blue iron hexacyanoferrate pigment first synthesized in the eighteenth century, brings exactly what manganese dioxide lacks. Its three-dimensional open framework of iron-nitrogen and iron-carbon coordination units, stitched together by cyanide bridges, creates interconnected channels through which electrolyte ions can migrate freely. More importantly, it hosts its own family of Fe2+/Fe3+ redox centers, adding a second, independent source of faradaic charge storage.

Fabrication followed a two-step sequence chosen for its industrial simplicity. First, strips of carbon cloth were activated in a hot mixture of nitric and sulfuric acid, which cleans the fibers and grafts oxygen-rich functional groups onto their surfaces, improving wettability and creating nucleation sites. The cloth was then sealed in an autoclave with acidified potassium permanganate solution and heated to 180 degrees Celsius for 24 hours, growing dense forests of manganese dioxide nanoneedles and nanorods directly on every fiber. In the second step, Prussian blue was electrodeposited onto the coated cloth by cyclic voltammetry in a solution of ferric chloride, potassium ferricyanide, hydrochloric acid, and supporting potassium chloride. The number of deposition cycles, varied from 15 to 100, became the tuning knob for how much Prussian blue loaded onto each electrode.

Microscopy revealed why the choice of 50 cycles proved decisive. With only 15 cycles, sparse Prussian blue particles dotted the carbon fibers, leaving most of the conductive network uncovered and the active material content too low. At 100 cycles, the opposite failure appeared: thick, cracked, agglomerated crusts of Prussian blue choked the porous architecture, blocking electrolyte diffusion and raising charge-transfer resistance. The PBMC-50 electrode, named for its 50 deposition cycles, hit the sweet spot, with a uniform, well-dispersed coating that preserved the interconnected fibrous framework of the cloth while maximizing the electroactive surface area. X-ray diffraction, performed on films grown on transparent conductive substrates to avoid interference from the carbon background, confirmed the coexistence of tetragonal alpha-MnO2 and face-centered cubic Prussian blue, with no impurity phases detected.

Electrochemical testing told a consistent story across every measurement technique. In a three-electrode configuration, the optimized PBMC-50 electrode delivered an areal capacitance of 636 millifarads per square centimeter at a modest current density of 0.7 milliamperes per square centimeter, and still retained 426 millifarads per square centimeter at 2 milliamperes per square centimeter. Cyclic voltammetry and galvanostatic charge-discharge measurements, which probe the electrode on different timescales, both independently identified PBMC-50 as the best performer, ruling out the possibility that the result was an artifact of a single method. Impedance spectroscopy added a mechanistic explanation: adding Prussian blue left the overall series resistance essentially unchanged, around 172.5 ohms, but sharply reduced interfacial charge-transfer resistance and produced a near-vertical low-frequency response with a phase angle approaching 80 degrees, a signature of efficient ion transport.

Kinetic analysis added a layer of nuance rarely reported for such electrodes. By separating the current response into surface-controlled capacitive and diffusion-controlled contributions, the researchers discovered markedly asymmetric behavior between the oxidation and reduction branches. The anodic b-value of 0.36 indicated a strongly diffusion-limited oxidation process, while the cathodic b-value of 0.66 pointed to a mixed regime. At the slowest scan rate of 2 millivolts per second, diffusion-controlled processes accounted for about 88 percent of the anodic charge and 77 percent of the cathodic charge, with surface-controlled contributions growing as the scan rate increased. This mixed, diffusion-influenced mechanism reflects the genuine bulk participation of the redox-active phases, a double-edged property that grants high capacity at moderate rates but limits the deepest active sites at very fast charging.

The practical payoff came when the team assembled complete devices. Using a polyvinyl alcohol-sodium sulfate gel electrolyte, they constructed six different symmetric and asymmetric all-solid-state supercapacitors and compared them head to head. The symmetric PBMC//PBMC configuration emerged as the clear winner, achieving an areal capacitance of 117 millifarads per square centimeter, more than four times that of the weakest configuration tested. After systematically mapping the voltage window with cyclic voltammetry and charge-discharge curves, the researchers settled on a wide operating range of 2.2 volts, extending from minus 1.4 to plus 0.8 volts. They deliberately avoided pushing to the highest voltage the device could technically tolerate, noting that beyond this range, polarization and incipient electrolyte oxidation begin to contaminate the response with parasitic, weakly reversible reactions.

The full-cell figures place the device among the competitive entries in the flexible supercapacitor field. The PBMC//PBMC device delivered a maximum areal energy density of 0.084 milliwatt-hours per square centimeter at a power density of 0.48 milliwatts per square centimeter, and held a capacitance of 31 millifarads per square centimeter even at a demanding 5 milliamperes per square centimeter. Endurance testing over 10,000 consecutive cycles at 7 milliamperes per square centimeter left 89.83 percent of the initial capacitance intact. Mechanical robustness proved equally impressive: capacitance retention measured at bending angles of 0, 90, and 180 degrees came in at 100, 99.14, and 98.67 percent respectively, meaning the device essentially did not notice being folded in half. Three devices wired in series successfully lit an array of ten red and yellow light-emitting diodes, and a single unit, light enough to rest on a plant leaf without bending it, demonstrated the ultralight character the design was after.

What distinguishes this work is not a single record-breaking number but the coherent integration of materials chemistry, electrode engineering, and device demonstration. By coupling manganese-based and iron-based redox systems on a conductive, mechanically resilient carbon cloth scaffold, the researchers eliminated the binders and insulating additives that typically degrade flexible electrode performance, while the Prussian blue framework simultaneously added redox capacity, improved ion pathways, and helped suppress the structural degradation that usually shortens manganese dioxide lifetimes. The hydrothermal-then-electrodeposition route uses inexpensive reagents, water-based processing, and standard laboratory equipment, making it a credible candidate for scale-up. As the market for wearables, electronic skin, medical sensors, and bendable displays continues to expand, electrode architectures of this kind, which balance energy density, durability, and manufacturability, may prove to be the quiet enabling technology behind the next generation of devices that flex with the human body.

Subject of Research: Binder-free Prussian blue-MnO2 heterostructure electrodes on carbon cloth for flexible solid-state supercapacitors

Article Title: Engineering Prussian blue-MnO 2 heterointerfaces on carbon cloth as binder-free electrodes for high-performance flexible solid-state supercapacitors

Article References: Moosavi, N., & Momeni, M. M. (2026). Engineering Prussian blue-MnO2 heterointerfaces on carbon cloth as binder-free electrodes for high-performance flexible solid-state supercapacitors. Results in Chemistry, 30, Article 103847. https://doi.org/10.1016/j.rechem.2026.103847

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103847

Keywords: flexible supercapacitors, Prussian blue, manganese dioxide, carbon cloth, binder-free electrodes, solid-state energy storage, pseudocapacitance, heterointerfaces, electrodeposition, wearable electronics, energy density, cycling stability