Gradient-Doped SnO2 for Perovskite Photovoltaics
Although single-junction perovskite solar cells (PSCs) are rapidly approaching the Shockley-Queisser limit, a persistent efficiency gap remains between the conventional (n-i-p) and inverted (p-i-n) architectures. While inverted devices have achieved a certified power conversion efficiency (PCE) of 28.0%,1 the certified PCE of conventional n-i-p PSCs has long stagnated near 26%. This discrepancy no longer originates from the perovskite absorber itself but increasingly from the electron-transport interface. However, tin oxide (SnO2) in conventional n-i-p PSCs still suffers from limited control over interfacial electronic properties and defect distribution.2 Against this backdrop, SnO2 deposited using the chemical-bath-deposited (CBD) method can achieve uniform coverage on textured substrates, and with its low-temperature processing and low cost, it is an ideal choice for realizing high-performance n-i-p devices.
Bottlenecks in CBD-SnO2
Despite its advantages in conformal deposition and industrial compatibility, CBD-SnO2 suffers from a fundamental limitation: the electronic structure of the film is difficult to precisely regulate during growth. During the CBD process, the oxidation kinetics of Sn2+ precursors strongly influence oxygen-vacancy (VO) formation and defect distribution. Notably, VOs in SnO2 can act as shallow-level donor defects, increasing electron density and electrical conductivity. However, when their concentration is too high, they introduce defect states that promote nonradiative recombination and accelerate material degradation. Therefore, for CBD-SnO2, the challenge lies not in maximizing or eliminating VOs but in balancing their beneficial and harmful effects. Furthermore, hydroxyl groups, residual ligands, and dangling bonds enriched on the SnO2 film surface introduce deep-level defects into the band gap, acting as Shockley-Read-Hall recombination centers that directly deplete photo-generated carriers. Consequently, conventional CBD-SnO2 still faces intrinsic bottlenecks in enabling high-performance and stable PSCs.
Toward graded CBD-SnO2
Previous studies have explored various optimization approaches. For example, Lin et al. proposed a pre-nucleation strategy to achieve SnO2 deposition driven by single-site heterogeneous nucleation, enabling carbon-based PSCs without a hole transport layer to achieve a record efficiency of 21.01%.3 Seo et al. obtained tin oxide films with complete surface coverage and low defect density through an excess ligand strategy, thereby reducing surface reflectance and improving charge extraction, ultimately increasing the certified efficiency of single-junction PSCs to 26.10%.4 Although CBD of SnO2 is the benchmark method for fabricating state-of-the-art n-i-p PSCs (including large-area devices and modules), it offers limited control over electronic properties. Previous research strategies have primarily focused on optimizing film formation, surface coverage, and defect passivation, viewing VOs as harmful defects while overlooking the fact that they are the main source of n-type doping in oxidized tin. Consequently, there is an urgent need to elucidate the underlying formation mechanisms and develop advanced strategies to drive the advancement of n-i-p PSCs.
Against this backdrop, Wang et al. found that the performance bottleneck in textured substrates stems from a synergistic effect between band misalignment and electron accumulation at the buried substrate interface.5 Further studies indicate that the incomplete oxidation of the Sn2+ precursor to Sn4+ during the CBD process is the key factor driving the incorporation of VO into the tin oxide lattice. This insight motivated the authors to fundamentally revisit the formation mechanism of CBD-SnO2 films and the microscopic origin of their electronic heterogeneity. Through combined characterization using cross-sectional transmission electron microscopy (TEM) and time-of-flight secondary ion mass spectrometry (TOF-SIMS), it was revealed that CBD-SnO2 is not simply a continuous dense film but rather a quasi-layer-by-layer (LBL) assembly of monodisperse colloidal nanoparticles (NPs) with diameters of 3–5 nm, formed through electrostatic adsorption and LBL stacking. This structural picture provides a mechanistic basis for spatially graded electronic structure design.
