Perovskites are a family of crystalline materials whose electronic properties can be tuned through their chemical composition. In solar cells, they absorb sunlight efficiently and generate mobile charge carriers, making them attractive for lightweight, flexible and potentially low-cost photovoltaic technologies. Screen printing could make these devices especially scalable because it deposits functional materials through patterned meshes, much like industrial printing processes. In a fully screen-printed architecture, multiple layers can be deposited sequentially, including the charge-transporting components, the perovskite absorber and the carbon electrode. The approach reduces reliance on vacuum equipment and could simplify manufacturing. But thick printed layers create a difficult physical environment for crystallization. A precursor solution must penetrate downward through the porous structure, react and solidify in the correct sequence, and form a continuous semiconductor without leaving voids, defects or mechanical stress behind.

The researchers identify inefficient vertical phase transformation as a central obstacle. During fabrication, the liquid perovskite precursor must undergo a transition from a solution containing dissolved and dispersed chemical components into an ordered crystalline solid. If crystallization begins too early at the surface, a crust can form before the underlying material has been fully infiltrated. That premature surface nucleation blocks further penetration and leaves the lower portions of the film poorly converted. Incomplete infiltration can create disconnected regions, while uneven crystallization generates defects that trap charge carriers. Residual stress may also accumulate as the material shrinks or rearranges during solidification. Each of these problems can reduce the current extracted from the cell and increase recombination, a process in which electrons and holes meet before they can do useful work in an external circuit. In thick printed films, controlling the path and timing of the liquid is therefore as important as controlling the final crystal chemistry.

To address the problem, the team developed what it calls a fluid motion crystallization strategy. The method uses a co-solvent system composed of the ionic liquid methylammonium propionate and butyronitrile. Ionic liquids are salts that remain liquid under relatively mild conditions and can strongly interact with precursor species, while butyronitrile is used here to alter how the precursor solution flows and solvates its ingredients. According to the study, butyronitrile reduces the resistance to fluid motion and helps disperse aggregates of lead iodide, a key precursor component. The result is an optimized solvation structure: rather than allowing large or poorly dispersed precursor clusters to impede movement, the liquid remains sufficiently mobile to travel rapidly and deeply into the printed film. This fluid-control step is crucial because it shifts crystallization from a surface-dominated event to a more coordinated transformation throughout the film’s depth.

The proposed mechanism begins with rapid, deep infiltration of the precursor into the multilayered structure. Once the liquid has reached the lower regions, the material undergoes a bottom-up, ordered phase transition. In practical terms, crystallization starts in the interior or lower portion of the film and progresses upward before a competing crystalline layer can form at the exposed surface. This sequence helps the entire precursor volume participate in the conversion, reducing the likelihood of unfilled pockets and poorly connected grains. It also suppresses the formation of defects associated with abrupt or incomplete solidification. The importance of this ordering lies in the fact that a solar-cell absorber is not simply a layer of light-absorbing material; it must also provide a continuous route for photogenerated charges to reach the electrodes. A film that appears visually complete can still contain microscopic barriers that cause electrical losses if its crystals are poorly connected or riddled with defect sites.

After the controlled phase transition, the researchers observed the growth of a dense, interconnected network of perovskite nanocrystals. Nanocrystals are crystalline domains measured on the nanometre scale, and their connectivity determines how efficiently charges can move through the absorber. The reported network is accompanied by an island-like surface morphology rather than an entirely flat interface. That morphology strengthens contact between the perovskite and the carbon electrode deposited above it. The interface is a critical region in a printed solar cell: photogenerated electrons and holes must be transferred across it without becoming trapped or recombining. Intimate physical contact can lower interfacial resistance and provide more direct pathways for charge extraction. By combining a compact internal crystal network with a better-connected surface, the strategy appears to address two linked sources of loss—poor transport through the absorber and recombination near the electrode boundary.

The performance figures suggest that the processing method does more than improve an isolated laboratory measurement. The air-processed, fully screen-printed cells reached 22.41% power conversion efficiency, meaning that fraction of incident solar power was converted into electrical power under the reported testing conditions. The certified value of 21.86% is particularly significant because certification provides an independent assessment of the device’s measured output. Efficiency alone, however, is an incomplete measure of photovoltaic progress. Perovskite materials have historically faced concerns over long-term stability, with heat, light, moisture and electrical operation all capable of accelerating degradation. In this study, the devices preserved over 90.5% of their starting efficiency after 2,000 hours under ISOS-L-1 accelerated light-soaking conditions. They also showed no degradation after 900 hours at 85 °C and 50% ± 10% relative humidity under the ISOS-L-3 operational protocol, combining elevated temperature, moisture and continuous operation.

Those durability results are closely connected to the film’s microscopic structure. Defects and voids can act as chemical and electrical weak points, allowing moisture or heat to trigger local deterioration and creating pathways for further damage. Residual stress can have a similar effect by making the film more vulnerable to cracking or interfacial failure during thermal cycling. A dense, interconnected nanocrystal network may limit these vulnerabilities, while stronger contact with the carbon electrode can help preserve the electrical connection as the device operates. The study does not present stability as a separate coating or after-treatment solution; instead, it links durability to how the precursor flows and crystallizes during fabrication. That is an important shift in emphasis. Controlling the earliest stages of film formation may prevent the structural imperfections that later become visible as efficiency losses, rather than attempting to repair them after the solar cell has already been built.

The work also illustrates why manufacturing physics can determine whether an emerging photovoltaic technology remains a laboratory curiosity or becomes an industrial contender. Perovskite absorbers can be deposited at relatively low temperatures and are compatible with solution-based processing, but those advantages are realized only if the liquid precursor can be controlled across large areas and through multiple layers. Screen printing offers a potentially high-throughput route, yet its patterned deposition process naturally creates films whose thickness and porosity must be managed during conversion. A co-solvent that tunes fluid resistance, precursor aggregation and the order of crystallization could therefore be valuable beyond the specific devices tested in this study. The reported approach is not merely an adjustment to the final electrode or a small optimization of a laboratory coating; it targets the coupled relationship between fluid dynamics, chemical solvation, nucleation and charge transport. That integrated control is what makes the result potentially relevant to scalable production.

The researchers’ results do not eliminate every challenge facing perovskite solar technology, and the reported tests do not by themselves establish how the cells would perform over many years outdoors or across industrial-scale modules. Nevertheless, the combination of air processing, full screen printing, certified efficiency above 21%, and strong resistance to demanding light, heat and humidity tests gives the strategy unusual news value. The core insight is both technically specific and broadly understandable: in a thick printed solar cell, the route taken by a liquid before it becomes a crystal can determine how well the finished device works. By making the precursor more mobile, dispersing lead iodide aggregates and forcing crystallization to proceed from the bottom upward, the team created a more continuous absorber and a more effective carbon interface. The result is a perovskite solar cell designed not only to capture sunlight efficiently, but also to survive the physical stresses imposed by practical operation.