Crystallization Control Unlocks 30.1% Efficient All-Perovskite Triple-Junction Solar Cells

Perovskite solar cells have spent the past decade shattering one efficiency record after another, but the technology’s most ambitious configuration—stacking three perovskite absorbers on top of one another in a single monolithic device—has remained stubbornly out of reach. Now, a team at Nanjing University led by Hairen Tan reports in Nature Energy a decisive breakthrough: by taking control of how ultrawide-bandgap perovskite films crystallize, they have built an all-perovskite triple-junction solar cell with a power conversion efficiency of 30.1 percent, independently certified at 29.3 percent, together with robust operational stability.

The crux of the challenge lies in the top cell of the stack. In a triple-junction architecture, three subcells with different bandgaps are stacked so that each absorbs a different slice of the solar spectrum: a wide-bandgap top cell captures the most energetic photons, while middle and bottom cells harvest the redder light that passes through. For an all-perovskite design, the top absorber needs a bandgap of roughly 2.0 electronvolts, which requires a perovskite composition dominated by bromide rather than iodide. The problem is that pushing the bromide-to-iodide ratio that high triggers a cascade of material pathologies—surface wrinkling, chemical inhomogeneity, and large open-circuit voltage deficits—that have historically throttled the performance of these ultrawide-bandgap films.

The Nanjing researchers traced these problems back to the physics and chemistry of film formation. When a mixed bromide-iodide perovskite precursor solution is spun into a thin film and crystallized, the two halides do not necessarily solidify in lockstep. Iodide-rich and bromide-rich regions can nucleate at different moments, leaving the finished film with local variations in halide composition. Those variations create tiny bandgap fluctuations across the film, which in turn generate internal stress as different regions of the crystal lattice strain to accommodate one another. The visible symptom is wrinkling of the film surface—morphological disorder that degrades the uniformity of the interface where charge extraction begins.

To combat this, the team developed a strategy they describe as surface reconstruction combined with halide homogenization, executed through synergistic solvent and additive engineering. The first element involves a solvent treatment that induces a controlled reconstruction of the perovskite surface, suppressing the wrinkle formation that normally accompanies crystallization of these high-bromide compositions. The second element is a transient chlorine additive delivered via oleylammonium chloride. During film formation, the chloride ions are temporarily incorporated into the crystal lattice, where they act as a chemical mediator: their presence synchronizes the crystallization of bromide and iodide, ensuring that both halides lock into the lattice at the same rate rather than segregating into iodine-rich and bromine-rich domains.

The payoff of this coordinated approach is visible at every scale the researchers examined. Kelvin probe measurements showed that the treated films exhibit a uniform surface potential, in contrast to the patchy electrostatic landscapes of conventional ultrawide-bandgap perovskites. Time-resolved spectroscopic characterization revealed suppressed non-radiative recombination—the parasitic process in which photoexcited carriers annihilate each other as heat rather than contributing to current—and improved carrier mobility. Most strikingly, the open-circuit voltage of a single-junction device built on the 2.0-eV absorber reached 1.46 volts, a figure that dramatically narrows the notorious voltage deficit that has plagued wide-bandgap perovskites and that represents a critical step toward the theoretical performance limits of multijunction stacks.

Open-circuit voltage is the parameter that matters most in a triple-junction cell. Unlike single-junction devices, where current is the chief battleground, a series-connected multijunction stack forces all three subcells to operate at the same current, and the total voltage is the sum of the individual subcell voltages. Every millivolt lost to recombination or halide disorder in the top cell subtracts directly from the module’s total output. By achieving a 1.46-volt open-circuit voltage from a 2.0-eV absorber, the Nanjing team recovered a substantial fraction of the voltage that previous designs forfeited, converting the material-science fix into a direct efficiency gain at the device level.

To build the full triple-junction device, the researchers integrated their improved ultrawide-bandgap top cell with two carefully optimized lower absorbers, with bandgaps of 1.6 and 1.22 electronvolts respectively. The resulting monolithic device, in which all three junctions are grown sequentially on a single substrate, delivered a power conversion efficiency of 30.1 percent under standard test conditions, with a certified value of 29.3 percent—an efficiency that places all-perovskite triple-junction technology in the same league as the best perovskite-silicon tandems while using only low-cost, solution-processable absorbers. Importantly, the devices also demonstrated robust operational stability, addressing one of the most persistent doubts about perovskite photovoltaics.

The significance of the work extends beyond a single record number. The field’s broader goal, as the authors note, is all-perovskite multijunction photovoltaics exceeding 35 percent efficiency—a threshold that would rival the performance of the III-V compound semiconductors used in space-grade solar panels at a small fraction of the manufacturing cost. Triple-junction architectures are the most credible route to that target, but they live or die by the quality of the ultrawide-bandgap top cell, because the 2.0-eV absorber must absorb the harsh blue end of the spectrum without squandering voltage. The demonstration that solvent-induced surface reconstruction and transient chloride incorporation can tame the crystallization of these difficult compositions offers the field a reproducible recipe, rather than a one-off material trick.

There are also practical signals in how the result was achieved. Both components of the strategy—solvent engineering and small-molecule halide additives—are compatible with existing deposition workflows for perovskite films, including the antisolvent and thermal annealing steps used across the industry. The approach required no exotic processing equipment or entirely new material system, which improves the odds that the technique can be scaled to larger areas and integrated into commercial production lines. The work, which involved collaborators at the Australian National University and ULVAC-PHI Instruments in addition to Nanjing University, has already resulted in a granted patent held with Renshine Solar, a perovskite commercialization company founded by Tan, underscoring the team’s intent to translate the laboratory result toward manufacturing.

For now, the result stands as the most convincing demonstration yet that all-perovskite triple-junction solar cells are a viable technology rather than a theoretical curiosity. By showing that the stubborn problems of surface wrinkling, halide heterogeneity, and voltage deficit in 2.0-eV perovskites can be resolved through a rational, chemistry-level understanding of crystallization, the Nanjing team has removed one of the central bottlenecks on the road to ultrahigh-efficiency, low-cost solar power. If subsequent studies can preserve these gains while scaling the films and extending device lifetimes, the 35-percent-efficiency milestone that once seemed distant may arrive sooner than expected.

The voltage figure reported for the ultrawide-bandgap subcell deserves particular attention when set against the radiative limit. A 2.0-electronvolt absorber in an ideal diode would deliver an open-circuit voltage approaching 1.7 volts under one-sun illumination, so the 1.46 volts achieved here still leaves a deficit, but one that is markedly smaller than the losses of roughly 0.7 volts or more that have typified high-bromide perovskites in earlier studies. Because voltage losses in mixed-halide films are largely attributable to non-radiative recombination at halide-segregation-induced defects, the observed improvement is consistent with the idea that homogenizing the bromide and iodide distribution removes the very sites where carriers were previously lost.

The transient role of the chloride additive fits a broader pattern in halide perovskite chemistry. Chloride has long been known to influence grain growth, crystallite orientation, and defect density in lead-halide films, but it typically remains in the lattice only in small quantities or is expelled during annealing. Using oleylammonium chloride as a delivery vehicle adds a further dimension: the long organic cation can interact with the precursor solution and the developing film surface, slowing nucleation and giving the bromide and iodide species time to intermix before the lattice locks in. The fact that chlorine is largely absent from the final absorber means the strategy improves the film without introducing a foreign species whose long-term stability might be questioned.

The characterization approach used by the team also reflects how perovskite research has matured. Kelvin probe force microscopy maps surface potential with nanoscale resolution, allowing researchers to see whether electrostatic inhomogeneity persists after treatment, while time-resolved photoluminescence distinguishes radiative from non-radiative decay pathways. Combining these probes with device-level measurements makes it possible to connect a processing intervention directly to the microscopic defect physics and then to the macroscopic efficiency gain, a chain of evidence that strengthens confidence that the improvement is mechanistic rather than coincidental.

It is also worth situating the result within the trajectory of the field. All-perovskite tandems crossed 29 percent efficiency only recently, and triple-junction perovskite devices have lagged behind their perovskite-silicon counterparts because the ultrawide-bandgap top absorber was the weakest link. The certified 29.3 percent reported here, achieved entirely with solution-processed perovskite absorbers, suggests that the remaining headroom toward 35 percent depends less on inventing new architectures than on continuing to reduce voltage losses and optical losses in each subcell. The operational stability data accompanying the efficiency figures will be scrutinized closely, since encapsulated multijunction devices must endure prolonged illumination and thermal cycling, but the demonstration that a heavily brominated top cell can be made both efficient and durable removes a long-standing objection to the all-perovskite approach.

Subject of Research: Crystallization control of ultrawide-bandgap 2.0-eV perovskite absorbers for high-efficiency all-perovskite triple-junction solar cells

Article Title: Crystallization control of 2.0-eV bandgap perovskites for all-perovskite triple-junction solar cells

Article References: Zhang, Y., Wang, E., Liu, H., Zhou, D., Lin, R., Li, H., Xu, D., Lou, J., Zhu, H., Li, M., Wang, Y., Duan, C., Zhu, Y., Bui, A. D., Nguyen, K., MacDonald, D., Yang, O., Ju, H., Li, L., … Tan, H. (2026). Crystallization control of 2.0-eV bandgap perovskites for all-perovskite triple-junction solar cells. Nature Energy. https://doi.org/10.1038/s41560-026-02135-1

Image Credits: AI Generated

DOI: 10.1038/s41560-026-02135-1

Keywords: perovskite solar cells, triple-junction, ultrawide-bandgap, crystallization control, halide homogenization, surface reconstruction, open-circuit voltage, power conversion efficiency, multijunction photovoltaics, oleylammonium chloride, non-radiative recombination, Nature Energy