Layer Order Surprise: Atomic-Deposited Tin Oxide Boosts Perovskite Solar Cells Only When Hidden

Perovskite solar cells have dazzled the photovoltaics world with efficiencies that now rival and in some cases exceed conventional silicon, yet the race to make them commercially viable hinges on details as small as a few nanometers. One of the most consequential details is the electron transport layer, the thin film that sits between the transparent electrode and the light-absorbing perovskite and ushers electrons out of the device. A new open-access study from researchers at Hankyong National University in South Korea, published in Advances in Industrial and Engineering Chemistry, takes a hard look at one of the most celebrated techniques for making that layer, atomic layer deposition of tin dioxide, and arrives at a conclusion that upends a common assumption: the pristine, atomically precise film that ALD produces can actually sabotage a solar cell when it touches the perovskite directly.

Tin dioxide has largely displaced titanium dioxide as the electron transport material of choice in the standard perovskite device architecture. Titanium dioxide was the original workhorse, offering a well-matched conduction band and robust chemistry, but it carries three burdens that have become increasingly costly as the field matures. Its electron mobility is modest, which throttles charge transport; its deposition typically demands temperatures above 450 degrees Celsius, ruling out flexible plastic substrates and inflating manufacturing costs; and under ultraviolet illumination it acts as a photocatalyst that slowly chews apart the perovskite layer it is supposed to protect. Tin dioxide, by contrast, offers higher electron mobility, a wider bandgap that keeps it optically transparent, and processing temperatures low enough for industrial roll-to-roll production, making it the backbone of many record-setting devices.

How that tin dioxide film is laid down matters enormously. The dominant laboratory method is spin coating, in which a colloidal suspension of tin dioxide nanoparticles is pipetted onto the spinning substrate and annealed. It is cheap, fast, and works in ambient air, but it is also notoriously uneven: incomplete coverage and pinholes expose the underlying electrode, creating defect-rich hotspots where electrons recombine with holes and their energy is wasted as heat rather than harvested as current. Atomic layer deposition promises to fix all of that. By alternating exposures to a tin precursor and an oxidant, ALD builds films one atomic layer at a time, yielding thickness control, uniformity, and conformality that no solution method can match, all at a modest 180 degrees Celsius in the Korean team’s process. On paper, it should be the perfect electron transport layer.

The new study shows that on paper is exactly where that story ends. The researchers fabricated five device configurations on indium tin oxide glass: no transport layer at all, spin-coated tin dioxide alone, ALD tin dioxide alone, and two bilayer stacks in which the two deposition methods were combined in opposite orders. They then probed each interface with photoluminescence spectroscopy, a technique in which the perovskite is optically excited and the light it re-emits is measured. Efficient charge extraction quenches that emission, so dimmer fluorescence generally signals better electron removal. But the measurements revealed a subtle trap. Devices in which ALD-grown tin dioxide touched the perovskite directly, as well as devices with no transport layer at all, showed complete suppression of the perovskite’s characteristic 810-nanometer emission peak, which superficially looks like superb charge extraction.

Time-resolved photoluminescence exposed what was really happening. In those configurations the emission did not merely dim; it collapsed almost instantaneously, a signature not of electrons being efficiently swept away but of severe energy loss. The culprit is the very perfection of the ALD film. Solution-processed tin dioxide is riddled with oxygen vacancies, defect states that donate free carriers and enhance conductivity. ALD-grown tin dioxide, being dense and nearly stoichiometric, lacks those vacancies, so its intrinsic conductivity is low and its conduction band sits misaligned with the lowest unoccupied molecular orbital of the perovskite. Electrons arriving at that interface encounter an energetic wall, become trapped, and recombine non-radiatively. The same fate befalls electrons dumped directly into bare indium tin oxide, which, as a degenerate conductor rather than a true semiconductor, offers no well-defined conduction band to receive them gently. Suppressed luminescence, in other words, can masquerade as good charge extraction while actually broadcasting catastrophic energy dissipation.

Atomic force microscopy added a morphological dimension to the story. Bare indium tin oxide has a root-mean-square roughness of about 2.68 nanometers, which spin-coated nanoparticles smooth to 2.02 nanometers by partially filling the surface’s valleys. The ALD film, however, grew rougher still at 4.00 nanometers, because the process replicates the underlying topography layer by layer rather than leveling it as a liquid does. When ALD was deposited on top of spin-coated nanoparticles, roughness rose again to 2.46 nanometers. But reversing the order produced a striking result: spin-coating nanoparticles onto the ALD film yielded the smoothest surface of all, at just 1.03 nanometers. The researchers attribute this to chemical affinity, since both layers are tin oxide, allowing the nanoparticle suspension to wet the ALD surface uniformly and pack into a compact coating with minimal voids.

The device physics followed directly from the interface chemistry. Current density-voltage measurements, taken under standard one-sun illumination with a Class A+A+A+ LED solar simulator, showed that ALD-only devices and those with no transport layer performed worst, with pronounced hysteresis between forward and reverse voltage scans, a symptom of trapped interfacial charge. Dark-current measurements told the same story from the opposite direction: devices with ALD tin dioxide facing the perovskite leaked significantly more current between 0.2 and 1.0 volts, indicating degraded diode behavior and rampant recombination, while the favorable configurations behaved like near-ideal diodes. Notably, the short-circuit current density exceeded 20 milliamperes per square centimeter in every configuration, confirming that the transport layer’s order influenced voltage losses and fill factor rather than how much light was absorbed.

The winning architecture, ALD beneath and spin-coating above, delivered a power conversion efficiency exceeding 22 percent, comfortably outperforming both single-layer alternatives and the reversed bilayer, and box plots across multiple devices showed the improvement was reproducible rather than a lucky champion cell. Durability improved as well. In a ten-day ambient storage test, the bilayer device held its initial efficiency throughout, whereas the spin-coated-only control began a sharp decline after day eight, suggesting the dense ALD film acts as a robust barrier protecting the stack from below. The hysteresis index, a quantitative gauge of charge trapping, was lowest for the winning configuration, tying together every strand of the optical, morphological, and electrical evidence.

The broader lesson reaches beyond a single lab result. Atomic layer deposition is already central to perovskite-silicon tandem cells, where conformal transport layers must coat textured silicon surfaces, and this study supplies a design rule those efforts cannot ignore: ALD-grown tin dioxide is best deployed as an embedded passivation and barrier layer, never as the direct electronic handshake with the absorber. The work also carries a methodological warning for the field at large, since steady-state photoluminescence quenching alone would have led researchers to celebrate the very interfaces that were destroying their devices. As perovskite photovoltaics edge toward factories, the difference between a record and a dud may come down to which side of a two-nanometer film the nanoparticles land on.

Subject of Research: The role of atomic layer deposited tin dioxide electron transport layers in perovskite solar cell efficiency and stability

Article Title: Unveiling and optimizing the role of ALD-SnO2 in Perovskite solar cells

Article References: Gong, J., Song, M., Kim, Y.-K., & Kim, G. M. (2025). Unveiling and optimizing the role of ALD-SnO2 in Perovskite solar cells. Advances in Industrial and Engineering Chemistry, 1(1), Article 10. https://doi.org/10.1007/s44405-025-00010-3

Image Credits: AI Generated

DOI: 10.1007/s44405-025-00010-3

Keywords: perovskite solar cells, tin dioxide, atomic layer deposition, electron transport layer, spin coating, charge recombination, interfacial passivation, photoluminescence, power conversion efficiency, tandem solar cells, oxygen vacancies, thin films