Publication date: 15th December 2025
Self-assembling molecules (SAMs) are currently considered to replace PEDOT:PSS as a standard hole-transport layer (HTL) for tin-lead (Sn-Pb) narrow-bandgap (NBG) perovskite solar cells. Such NBG perovskites require precursor additives for suppression of Sn2+ oxidation and to balance crystallization, in order to achieve high solar cell efficiencies. Pseudo-halide precursor additives such as lead(II)-thiocyanate (Pb(SCN)2) are commonly employed for such purposes [1-3]. We report on an unintended effect that involves Pb(SCN)2 in SAM based NBG perovskite solar cells. Upon increasing the amount of Pb(SCN)2 up to 6 %, SAM-based layer stacks exhibit progressively reduced non-radiative carrier recombination and very good film morphology. Photoconversion efficiency reaches a maximum of 18.5 % for 1 % Pb(SCN)2 in the precursor and drops drastically for higher concentrations. Time resolved photoluminescence (trPL) and surface photovoltage (trSPV) measurements reveal impeded carrier extraction in Pb(SCN)2-containing layer stacks with Pb(SCN)2 concentrations greater than 1 %. Surprisingly, from UV- and X-ray photoelectron spectroscopy it becomes clear that the amount of SAMs is reduced at the buried interface and that SAMs appear on the top surface of the perovskite layer. We conclude that the hole-selective SAMs relocate from the ITO to the perovskite top surface in presence of Pb(SCN)2 and subsequently form an electron blocking layer between perovskite and C60. The relocated SAMs can be addressed by washing the perovskite surface, which recovers electron extraction into the electron-transport layer (ETL), however overall device performance remains impeded. This study highlights that commonly used additives that were explored for PEDOT:PSS based layer stacks have to be adapted accordingly, when introducing SAMs in NBG solar cells.
The authors thank the laboratory technicians for the technical support in HySPRINT Helmholtz Innovation Lab. The authors thank T. Dittrich for the technical assistance and support during data evaluation for trSPV measurements. The authors thank for funding by the Helmholtz Association for funding the HySPRINT Helmholtz Innovation Lab as well as the “Zeitenwende” project. In addition, we acknowledge the research school “HyPerCells”. The authors specifically thank for funding through the Deutsche Forschungsgemeinschaft (DFG) project “HIPSTER PRO” (grant no. AL 2241/1-1) and EUROPEAN CLIMATE, INFRASTRUCTURE AND ENVIRONMENT EXECUTIVE AGENCY (CINEA), Project 101075605 – SuPerTandem.
