Mixed-halide perovskite materials are promising candidates for multijunction solar cells aiming to surpass the detailed-balance limit. However, perovskite solar cells (PSCs) still suffer performance losses due to nonradiative charge recombination, primarily caused by electronic defects (traps). Efficient extraction of charge carriers from the absorber layer to the electrodes is also crucial for achieving high device performance. Under operational conditions, recombination and extraction processes compete, and understanding their relative contributions is essential for improving PSC efficiency.

Photoluminescence (PL) spectroscopy is a widely used technique to probe nonradiative losses in semiconductors. In particular, transient PL (tr-PL) measurements enable time-resolved analysis of recombination dynamics. In metal-halide perovskites, tr-PL decays are often governed by trapping and de-trapping at shallow defects, leading to long decay times.^[1] Differentiating between charge recombination and extraction mechanisms typically requires steady-state or voltage-dependent PL measurements.^[2] Notably, under short-circuit conditions, PSCs often exhibit strong PL emission, suggesting impeded charge extraction due to poorly optimized transport layers or ion-induced screening effects.^[3]

In this work, we introduce voltage-dependent transient PL spectroscopy (tr-PL(V)) to disentangle charge recombination and extraction processes in PSCs. By comparing the tr-PL decay under open-circuit and short-circuit conditions, we show that charge extraction dominates after a specific laser delay time, resulting in fast quenching of the PL signal. We demonstrate the applicability of the tr-PL(V) method by intentionally modifying the charge extraction rate by tuning the C₆₀ layer thickness and inserting passivation layers at the perovskite/C₆₀ interface. We identify that ion-induced screening affects the PL decay, changing the recombination dynamics. Hence, we apply a pre-bias before switching to short-circuit during tr-PL(V) measurements to enforce a homogeneous mobile ion distribution across the perovskite layer. This pre-biasing leads to enhanced charge extraction and confirms that mobile ions affect (transient) PL(V) measurements under short-circuit operation, hindering extraction.

Finally, we reveal that mobile ions also affect regular tr-PL measurements under open-circuit conditions in complete devices, whereas in half-stacks (without electrodes), their influence is significantly reduced.

Our results demonstrate that tr-PL(V) spectroscopy is a powerful tool to distinguish between charge recombination and extraction in PSCs. Moreover, applying a pre-bias enables the characterization of ion-free charge extraction dynamics, revealing the true extraction velocity.