Crystal-Solvate Seeding Strategy Pushes Perovskite Solar Mini-Modules to 23.15% Efficiency
Perovskite solar cells have attracted intense research attention for a straightforward reason: they convert sunlight to electricity with high efficiency and can be fabricated using solution-based processes that are far cheaper and more scalable than the high-temperature vacuum deposition required for silicon. But translating laboratory-scale efficiency records into large modules that perform reliably over decades has proven stubbornly difficult. Every increase in cell area introduces new opportunities for defects, non-uniformities, and interface failures that reduce performance.
Inverted perovskite solar cells, in which the hole-transport layer sits beneath the perovskite absorber and the electron-transport layer sits on top, are particularly attractive for scaling because they are compatible with solution processing techniques and have demonstrated good stability. Their persistent weakness has been the quality of the buried interface - the point where the perovskite layer meets the hole-transport layer underneath. Controlling crystal growth and minimizing electronic defects at a surface that is literally buried beneath the rest of the cell is technically demanding, and failures there limit both efficiency and longevity.
Seeding Better Crystal Growth From Below
A team from the Qingdao Institute of Bioenergy and Bioprocess Technology (QIBEBT) at the Chinese Academy of Sciences, led by Professor Pang Shuping, addressed this problem by intervening before the perovskite film is deposited. Rather than trying to fix the buried interface after the fact, they created conditions that direct how crystals form in the first place.
The approach involves pre-depositing specially designed nanocrystals - called crystal-solvate (CSV) materials, with the chemical formula PDPbI4-DMSO - onto the self-assembled monolayer surface that typically serves as the hole-transport layer in inverted cells. These CSV nanocrystals have a rod-like shape. When deposited on the hydrophobic monolayer surface, they improve wettability - allowing the perovskite precursor solution to spread more uniformly when applied. Without this modification, the hydrophobic surface resists even spreading, producing non-uniform films that nucleate crystals erratically.
The CSV nanocrystals do more than improve wetting. During the subsequent crystallization process, they serve as nucleation sites. Rather than forming crystals randomly throughout the precursor film, the perovskite material preferentially nucleates at the pre-seeded locations, producing a more ordered, highly oriented crystal structure at the bottom of the film.
Lattice-Confined Solvent Annealing
A second function of the CSV design involves the DMSO solvent molecules incorporated into the crystal structure. Dimethyl sulfoxide is routinely used in perovskite processing because it forms intermediate phases with lead iodide that slow crystallization and improve film quality. In the CSV design, DMSO is locked within the nanocrystal lattice rather than free in solution. During thermal annealing, it is released in a controlled and gradual manner at the buried interface, creating what the team calls a lattice-confined solvent annealing microenvironment.
This local solvent atmosphere facilitates grain reorganization at the very bottom of the growing perovskite film without disturbing the rest of the film, producing a dense, well-oriented perovskite region at the buried interface that previously could not be achieved through standard annealing approaches. Dr. Sun Xiuhong, co-first author of the study, described the result as simultaneously addressing crystallization regulation and interface stabilization.
From Small Cells to Mini-Modules
The practical test of any perovskite advance is how well it performs at larger areas. The team fabricated a perovskite solar mini-module with an aperture area of 49.91 square centimeters using a slot-die coating process compatible with roll-to-roll manufacturing, and achieved a power conversion efficiency of 23.15%. The efficiency loss between small-area reference cells and this mini-module was less than 3% - a figure the authors describe as outperforming many previously reported results, where scale-up typically produces larger efficiency drops.
The study was published in Nature Synthesis on February 27, 2026.
Several challenges remain between these results and commercial viability. The long-term stability of cells incorporating CSV seeding under operational conditions - extended illumination, thermal cycling, humidity - requires validation over timescales of years rather than the accelerated testing protocols used in laboratory studies. The slot-die coating process demonstrated here is scalable in principle, but optimizing it for production-scale equipment introduces additional engineering variables. Professor Pang noted that the CSV material platform is modular by design - by varying the organic cation and solvent components, a library of different CSV materials could be developed for different perovskite compositions and device architectures, extending the concept beyond this particular cell configuration.