Metal halide perovskite solar cells (PSCs) are rapidly progressing toward commercialization. Among the various device configurations, the p–i–n (inverted) architecture has emerged as particularly promising due to the very high record efficiencies achieved with this design[1], its compatibility with conventional silicon solar cells for tandem applications, and the elimination of expensive and relatively unstable doped hole transport layers such as Spiro-OMeTAD[2]. The charge transport layers, which represent the primary distinction between the normal and inverted architectures, offer significant potential for further improvement and optimization owing to the wide range of semiconducting materials available. In this work, we investigate both charge transport layers.

The benchmark electron transport material (ETM) for inverted PSCs is C₆₀/PCBM, a fullerenic carbon allotrope or its soluble derivative. While these materials are reliable and enable highly efficient devices, non-fullerene alternatives may offer enhanced stability due to closer molecular packing, which can provide improved barrier properties. Naphthalene diimide (NDI) has been widely used as a building block in organic semiconductors and represents a promising ETM for PSCs owing to its strong electron-accepting characteristics, high tunability—with or without bandgap modification—and relatively low synthesis cost[3]. We synthesized several NDI derivatives with different solubilities via one-step condensation reactions, characterized them electrochemically and thermally, and fabricated PSCs employing these non-fullerene ETMs.

In addition, the selection and deposition of the hole transport material (HTM) in inverted PSCs are strongly influenced by the choice of transparent conductive substrate. For large-scale commercialization, fluorine-doped tin oxide (FTO) offers a key advantage over indium tin oxide by avoiding reliance on the critical and costly resource indium. However, the deposition of self-assembled monolayers (SAMs) as HTMs on FTO is challenging due to the high surface roughness[4] and the absence of indium required for the formation of In–O–P bonds with common SAM anchoring groups[5]. To address these challenges, we optimized both the SAM deposition process and molecular selection for FTO substrates, and compared the device performance of SAM-only HTMs with that of combined SAM and nickel oxide layers.