A single chemical additive applied during perovskite film deposition helped lift a perovskite-silicon tandem solar cell to a certified stabilized efficiency of 32.76%, according to a peer-reviewed study published in a Nature journal. The result, achieved on an industrial-grade TOPCon silicon substrate rather than a lab-optimized wafer, targets a key barrier to translating high-efficiency tandem results onto commercially common silicon platforms.
Why Industrial Silicon Fights Back
Perovskite-silicon tandems stack a perovskite absorber on top of a silicon cell so each layer captures a different slice of the solar spectrum. The concept is straightforward, but execution on real-world silicon is not. Industrial TOPCon cells can draw heat away from a freshly deposited perovskite layer quickly compared with many research-grade substrates used in record-setting experiments. When a perovskite precursor solution is spin-coated onto these substrates, rapid heat transfer induces fast crystallization of the perovskite layer. That speed is the enemy of quality: the resulting films develop voids, uneven grain boundaries, and defects that drain electrical performance.
Most prior efficiency records have sidestepped this problem by using silicon heterojunction (SHJ) bottom cells with smoother, thicker wafers that slow heat dissipation naturally. TOPCon cells dominate the commercial silicon market, though, so any tandem technology that cannot work on TOPCon faces a steep path to mass production. The research team targeted this mismatch directly, designing their process around the thermal behavior and surface morphology of industrial cells rather than idealized substrates.
How One Molecule Slows the Crystal Race
The additive at the center of the result is 2-mercaptobenzothiazole, or MBT, a small sulfur-containing molecule already familiar to the rubber and corrosion-inhibitor industries. In this application, MBT serves a different purpose: the study reports that it binds to perovskite organic cations in two chemical modes, effectively acting as a molecular brake on crystallization. By latching onto the cations that drive crystal nucleation and growth, MBT gives the perovskite precursor more time to organize into a dense, uniform film before solidifying.
The reported experiments show that this dual-mode binding produces void-free perovskite layers on textured industrial TOPCon substrates, even under the fast heat-transfer conditions that usually degrade film quality. The practical payoff is a monolithic two-terminal tandem device with a certified stabilized power conversion efficiency of 32.76%. Measurement and reporting details, including current–voltage and maximum power point tracking procedures, are described in the paper’s methods and reporting sections.
Beyond morphology, the work indicates that MBT influences the energetics of the perovskite surface. By coordinating with organic cations, the molecule helps passivate undercoordinated sites that would otherwise act as traps. This combination of slowed crystallization and trap reduction allows the perovskite top cell to operate closer to its radiative limit, boosting open-circuit voltage without major changes to layer stack or processing temperature.
Cutting Losses at the Interface
Efficiency in a tandem cell depends not just on how well each absorber converts photons but on how cleanly charge carriers cross the boundary between layers. In perovskite-based devices, the dominant source of energy loss at these interfaces is non-radiative recombination, a process in which excited electrons release their energy as heat rather than electrical current. Research on interfacial loss channels in perovskite-silicon tandems has shown that this recombination stems primarily from defects at individual perovskite and transport layer junctions.
By producing a more uniform perovskite film with fewer grain-boundary defects, MBT appears to reduce the density of these recombination sites. The result is a cleaner electrical handoff between the perovskite top cell and the silicon bottom cell. This mechanism helps explain why a relatively simple additive, rather than an entirely new device architecture, can produce a meaningful efficiency gain. The improvement is not about absorbing more light; it is about wasting less of the energy that light generates.
The authors’ analysis aligns with broader efforts to tame interfacial losses in tandem structures. Detailed balance calculations and device modeling, such as those presented in recent tandem performance studies, highlight that once optical absorption is near optimal, incremental gains depend heavily on suppressing non-radiative pathways. MBT’s role as a crystallization modulator and defect passivator fits squarely into this strategy.
Where 32.76% Sits in the Bigger Picture
Perovskite-silicon tandems possess a theoretical efficiency ceiling that far exceeds what single-junction silicon cells can achieve. Recent analysis places the demonstrated performance of these tandems as nearing 35% power conversion efficiency. The 32.76% result reported here is significant not because it is the absolute highest number ever recorded for any tandem configuration, but because it was achieved on the type of silicon cell that factories already produce in volume.
That distinction matters for cost. A tandem process that requires exotic or non-standard silicon substrates adds expense and complexity that can erase the economic benefit of higher efficiency. By demonstrating that a single low-cost additive can bridge the quality gap on commodity TOPCon wafers, the work suggests a shorter path from lab to production line. For solar developers and utilities, higher module efficiency translates directly into less land, fewer mounting structures, and lower balance-of-system costs per watt installed.
The study also underscores how incremental, chemistry-focused interventions can unlock performance on par with more radical device redesigns. Rather than re-engineering the silicon bottom cell or introducing complex new interlayers, the researchers optimized a molecule that can be blended into existing perovskite precursor solutions. If compatible with roll-to-roll or slot-die coating, such an approach could dovetail with manufacturing lines already being developed for large-area perovskite modules.
Remaining Hurdles Before Factory Floors
Despite the promising efficiency, several challenges stand between this result and commercial deployment. The first is stability. Perovskite layers remain vulnerable to moisture, heat, and ultraviolet exposure, and any additive that alters crystallization could also affect long-term degradation pathways. The Nature Energy work focuses primarily on efficiency metrics; extended field-relevant stability testing under damp heat and light-soaking conditions will be needed to confirm that MBT does not introduce new failure modes.
Scaling is another open question. The reported devices are fabricated on industrial-type TOPCon cells but at laboratory scale. Translating MBT-assisted crystallization to full-size wafers and high-throughput coating tools will require tight control over solvent evaporation, temperature gradients, and additive concentration. The paper describes processing parameters and device fabrication in its methods sections, but industrial partners will still need to validate yield and uniformity across many cells on production tools.
There is also a broader ecosystem question. Tandem modules must integrate with existing balance-of-system components, inverters, and installation practices. Higher voltages and different current densities may require re-qualification of hardware and standards. While these issues are not unique to MBT-enabled tandems, they shape the timeline on which any new tandem architecture can move from record-setting cells to bankable products.
Even so, the demonstration that a single, readily available molecule can tame the thermal and morphological challenges of industrial TOPCon substrates marks an important inflection point. It suggests that the performance gap between champion cells and manufacturable tandems may be narrowed not only by new device stacks, but also by smarter chemistry at the nanoscale. If follow-up studies confirm durability and scalability, MBT or similar molecular brakes on crystallization could help push perovskite-silicon tandems from specialized prototypes toward mainstream deployment, bringing multi-junction efficiencies into the commercial solar landscape.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
