Currently, a perovskite/silicon tandem solar cell with 34.6% power conversion efficiency (PCE) is the most efficient dual‑junction solar cell under the AM1.5g spectrum [1]. Following this impressive efficiency improvement, perovskite/perovskite/silicon triple-junction solar cells have now gained significant attention and are rapidly developing [2–9] . In our work, we address some of the main aspects to realize monolithic two-terminal perovskite/perovskite/silicon triple‑junction solar cells.

First, in two-terminal multijunction solar cells, high open-circuit voltage (V_OC) is a key characteristic, as the voltages from different subcells add up. Therefore, minimizing voltage losses in individual subcells as well as the recombination layer between the cells is crucial.

To maximize V_OC, we identified the high-bandgap perovskite top cell as the main source of voltage loss in our triple-junction solar cell. We addressed this by optimizing both, the bulk quality of the perovskite and the interface between the perovskite and the electron transport layer (here C₆₀). Specifically, we replaced our reference triple-cation double-halide perovskite with a triple-cation triple-halide composition, which exhibits superior bulk quality. Additionally, a piperazinium iodide passivation layer [10] is introduced between the perovskite and C₆₀ to reduce the non-radiative recombination loss at this interface. These optimizations led to a high V_OC exceeding 3.0 V. Furthermore, we optimized the recombination layer between the two-perovskite sub cells leading to a V_OC of ∼3.1 V, which is among the highest values reported in literature for this structure. Finally, to investigate the remaining voltage losses in our structure, we conducted electrical simulations. The results indicate that by employing perfectly band aligned charge transport layers, the V_OC of our triple-junction cell has the potential of up to ~3.4 V [11].

Secondly, in a two-terminal multijunction solar cell, the sub-cell generating the lowest current limits the current of the whole device. Therefore, maximizing the current density (j_SC) requires current matching between the sub cells. This can be achieved through optimization of bandgaps and thicknesses of the perovskite middle cell and top cell.

To enhance the j_SC, we conducted optical simulations that revealed the middle cell as a significant limiting factor in our current triple-junction structure (~ 8.3 mA/cm² j_SC). Ideally, lowering the middle cell bandgap to 1.47 eV, would result in a current-matched triple-junction cell (at ~13 mA/cm²) [11]. However, this approach requires partially substituting lead with tin in the perovskite composition, which introduces significant stability challenges [12]. Therefore, we kept the bandgap of the perovskite middle cell at 1.55 eV and focused on increasing this absorber thickness with the potential of reaching > 11 mA/cm² j_SC for the triple-junction device [11]. In this regard, careful optimization of the crystallization process is necessary to ensure high bulk quality in the thick perovskite layer. For this purpose, we used a chloride-containing perovskite composition with a thickness of more than 750 nm. Our results highlight that optimizing the hole transport layer (HTL) and processing conditions, such as the choice of the antisolvent, is crucial for achieving a thick high-quality perovskite layer. Moving forward, we plan to implement this optimized middle cell in our triple-junction solar cell.