Impact of the perovskite top layer on the thermal behavior of c-Si bottom cells in tandem photovoltaic applications

The crystalline-silicon (c-Si) bottom cell in a perovskite/c-Si tandem device operates under an optical environment that differs fundamentally from full-spectrum single-junction operation. It receives a spectrum filtered by the perovskite top layer, and that filter is itself temperature dependent. Building on established studies of tandem temperature behavior, this work focuses on the apparent change in c-Si bottom-cell temperature coefficients associated with perovskite optical filtering, considered separately from the electrical coupling of the sub-cells. We combine SCAPS-1D simulations (standalone c-Si, 4 T optical splitting, and perovskite-filtered illumination) with experiments on a single c-Si cell measured before and after perovskite coating. The coating serves as an optical filter, without an electrically active perovskite top cell. In simulation under AM1.5G illumination, the short-circuit-current (I SC ) temperature coefficient rises from 0.082 %/°C for standalone c-Si to 0.233 %/°C under fixed spectral splitting, 0.439 %/°C under temperature-dependent bandgap filtering, and 0.685 %/°C under full perovskite filtering. Under the distinct laboratory spectra used experimentally, the same c-Si cell shows an increase from 0.120 %/°C when uncoated to 0.476 %/°C after coating under visible-plus-infrared light; these values are compared for consistency of trend rather than absolute magnitude. Under visible-only illumination, the apparent coefficient becomes much larger, which is attributed to near-band-edge spectral filtering rather than to an intrinsic change in c-Si behavior. The amplified response is mainly photocurrent-driven, as the open-circuit-voltage temperature coefficient remains comparatively insensitive to the optical configuration. This filtered bottom-cell current response is directly relevant to temperature-dependent current matching in series-connected 2 T tandems; in independently operated 4 T devices, its relevance is to the silicon maximum-power contribution rather than current matching. The study characterizes the bottom-cell response and does not quantify complete-tandem power or outdoor energy yield.