TOPCon is approaching SHJ performance, with Voc now exceeding 740 mV and the gap reduced to below 10 mV
Cost and scalability favor TOPCon, supported by lower CapEx, reduced silver use, and compatibility with existing PERC production lines
System-level performance remains complex, as tandem efficiency depends on sub-cell interaction, interconnect quality, and fill factor
Single-junction silicon cells are approaching their efficiency limits. Current cells are already close to 28%, with a practical ceiling near 29.4%, meaning further gains are limited. This is pushing the industry toward tandem architectures, where perovskite-silicon combinations have already exceeded 34% efficiency and show potential beyond 40%.
A recent study examines how different silicon bottom-cell technologies perform in these tandem structures, with a focus on the evolving role of TOPCon. It analyzes improvements in TOPCon cell design and tracks the narrowing performance gap with Silicon Heterojunction (SHJ). The review benchmarks TOPCon-based tandems against SHJ in terms of efficiency potential, cost, manufacturability, and scalability for large-scale deployment.
In a paper titled TOPCon-based bottom cells for perovskite/silicon tandem solar cells, published in Joule, the authors review recent progress in TOPCon-based tandems and benchmark their performance against SHJ-based designs.
In perovskite-silicon tandems, the top cell absorbs high-energy photons, while the silicon bottom cell captures the lower-energy part of the spectrum. This improves overall energy conversion. Among different configurations, 2-terminal (2T) designs are emerging as the most relevant for industrial adoption due to their simpler integration, despite the requirement for current matching.
Not all silicon technologies are suitable for this role. Al-BSF and PERC are limited by lower Voc and higher recombination losses. As a result, SHJ and TOPCon have become the main candidates. Given its high Voc above 740 mV and strong passivation, SHJ initially dominated the tandem landscape. It also offers both-side passivating contacts and integration flexibility, with TCO layers on the front side simplifying tandem integration.
TOPCon, however, has seen rapid progress. Improvements in poly-Si contacts, ultrathin oxide layers, and hydrogenation have reduced recombination losses. Laser-assisted processes such as LECO have further improved contact quality. This has pushed Voc values above 740 mV, reducing the gap with SHJ to below 10 mV.
A similar trend is seen in fill factor. While SHJ typically operates at 83-85%, TOPCon has improved from around 81-82% to above 84%. The difference remains, but its impact at the tandem level is less direct, as fill factor is also influenced by interconnect quality, localized defects, and current-limiting behavior in the device stack.
At the tandem level, performance depends on how both sub-cells interact. While Voc and current density are key drivers, fill factor is influenced by interconnect quality and defects. In some cases, a slight current mismatch can even improve performance due to FF compensation effects.
Simulation studies show that SHJ-based tandems can reach around 38% efficiency. TOPCon-based designs achieve about 36-37%, depending on passivation and surface texturing. Double-sided texturing improves the current generation and further reduces the gap.
Despite this difference, TOPCon offers advantages in manufacturing. The industry has already scaled TOPCon rapidly, with a market share north of 50%. It is also compatible with existing PERC production lines, reducing capital investment. In contrast, SHJ requires more expensive equipment and relies on indium-based TCO layers, which impose cost and material constraints.
From an LCOE perspective, both technologies are comparable. SHJ benefits from higher efficiency, but this is offset by higher manufacturing costs. This keeps TOPCon competitive for large-scale deployment.
TOPCon integration still faces challenges, including parasitic absorption, passivation losses on textured surfaces, sputtering damage, and hydrogen-related instability. Resolving these issues will aid further development of tandem architectures.