There is a gap between what a solar cell looks like and what it actually does, and that gap has been quietly undermining one of the most promising technologies in renewable energy for years. A solar cell can be chemically pristine, beautifully crystalline and optically efficient under laboratory conditions, and still collapse in ways that no standard measurement would have predicted. The problem is invisible to most of the tools researchers use to study these devices.

A new study published in Nature Energy has built a microscopy platform powerful enough to see inside operating perovskite solar cells at nanoscale resolution, tracking both the chemistry and the electrical behaviour of the same tiny patch of material before and after weeks of accelerated stress testing. What the researchers found rewrites several assumptions about what makes these cells fail and what engineers need to fix first.

What Makes Perovskites So Exciting

Perovskites are a family of crystalline materials with a particular atomic arrangement that turns out to be extraordinarily good at absorbing sunlight and converting it into electrical current. They can be manufactured at relatively low temperatures using solution based processes, meaning the technology could in principle be produced cheaply at scale compared to conventional silicon solar cells, which require energy intensive high temperature processing.

Over the past decade, perovskite solar cells have improved at a pace that has astonished the photovoltaic community. Their efficiency has climbed from a few percent to well above 25 percent in a remarkably short time, rivalling commercial silicon. When combined with silicon in tandem cells, stacking two different materials to capture different parts of the solar spectrum, the combined efficiencies can reach beyond 30 percent, higher than either material could achieve alone.

The problem is stability. Silicon solar cells routinely operate for 25 to 30 years in the field. Perovskite devices, for all their optical brilliance, tend to degrade far faster under the combined stresses of light, heat and electrical load. Understanding exactly why they fail and exactly where the damage begins has proven surprisingly difficult, because the standard tools of solar cell characterisation measure the whole device at once. They give you an average, and averages hide a great deal.

A Toolkit That Sees What Others Miss

The research team, led by scientists at the University of Cambridge and Helmholtz Zentrum Berlin, assembled a multimodal microscopy platform that combines several complementary imaging techniques applied to the exact same region of the same device.

The centrepiece is voltage dependent photoluminescence microscopy. When a perovskite is illuminated, it emits light, and the brightness and colour of that light carry information about the material's internal voltage and the quality of its crystal structure. By sweeping the voltage applied to the device while recording these light emissions at high spatial resolution, the team could extract a local current voltage curve from every point across the device, essentially mapping performance at the nanoscale without assuming any particular model of how the device should behave.

This optical measurement was combined with hyperspectral photoluminescence, which resolves the colour of the emitted light with great precision, and with synchrotron X-ray nanoprobe fluorescence, which uses a beam of hard X-rays focused to a spot just 50 nanometres wide to map the chemical composition of the perovskite at the same locations. Measurements from all three techniques were registered to the same spatial coordinates, allowing the team to directly compare what the chemistry looks like, what the internal voltage looks like, and how efficiently charges are extracted, all at the same spot on the same cell.

"We find that devices with lower PCE disorder correlate with higher initial performance and are also more stable under operational stress. By contrast, more disordered devices tend to be less stable and exhibit more severe phase segregation during stress."

The devices were then subjected to 100 hours of accelerated operational stress under the industry standard ISOS-L-2I protocol, held at open circuit voltage under continuous one sun illumination at 65 degrees Celsius. The same scan areas were then imaged again, allowing a direct before and after comparison at nanoscale resolution.

The Surprising Tolerance of Chemical Disorder

The first major finding concerns a feature called wrinkles. The perovskite layers studied in this work form a distinctive rippled morphology during crystallisation, and these wrinkled regions have previously been suggested to be harmful to device performance. The new measurements tell a more nuanced story.

Wrinkled areas do show altered chemistry. They contain slightly less bromine than their surroundings, which shifts their bandgap and changes the colour of their emission. They also emit more light, which partly compensates for this. The net effect on what really matters, the local power conversion efficiency, turns out to be almost negligible. The optical efficiency map of the best performing device was remarkably flat, varying by less than five percent relative across a region that showed dramatic chemical and morphological variation.

This is a striking result. The perovskite, in the best devices, tolerates chemical disorder surprisingly well. Charge carriers generated in regions with slightly different chemistry are funnelled into higher quality zones, maintaining a nearly uniform internal voltage across the device. The crystal can be chemically imperfect and still perform consistently.

What it cannot tolerate, the team found, is disorder in charge extraction.

The Critical Role of Charge Transport

When the team added a third element to the perovskite recipe, specifically a chloride source introduced through lead chloride in the precursor solution, the device initially appeared to improve. The addition widened the bandgap, which is useful for the top cell in a tandem device. But the nanoscale maps told a different story than the bulk measurements.

The chloride containing device showed dramatically greater spatial variation in its optical power conversion efficiency. Some regions were extracting charges well while others were not, and this patchiness persisted even in the fresh, unstressed device. After operational stress, these devices lost more than 40 percent of their initial efficiency on average, far worse than the chloride free devices. Regions of severe halide segregation appeared, where the bromine and iodine in the crystal separated under the combined stress of light and heat into distinct chemical phases, each with different optical properties.

The team then added yet another ingredient, methylammonium chloride, to produce the triple cation triple halide composition used in many of the highest performing perovskite tandem cells reported in the literature. Counterintuitively, this slightly increased compositional complexity delivered a large improvement in stability. Phase segregation was strongly suppressed. The photoluminescence signal remained stable under stress. The bulk of the perovskite, it appeared, had been stabilised by this compositional change even though the initial device performance was slightly lower.

The lesson was clear: for a given interface quality, the spatial uniformity of charge extraction, what the team defined as PCE disorder, predicted long term stability far better than any single bulk measurement.

KEY FACTS

What is a perovskite solar cell? A solar cell based on a class of crystalline materials with a specific atomic arrangement called the perovskite structure. These materials are exceptional light absorbers and can be processed at low temperatures, making them candidates for low cost, high efficiency photovoltaics. Their main challenge is long term operational stability.

What is PCE disorder? A measure of the spatial variation in local power conversion efficiency across a solar cell, extracted by mapping how well each nanoscale region of the device converts absorbed light into electrical current. Lower disorder means more uniform performance and, this study shows, better stability.

What is halide segregation? A degradation process in which the mixture of halide elements (such as iodine and bromine) in an alloyed perovskite separates into regions of different composition under illumination or heat. This changes the local bandgap and optical properties, generally reducing device efficiency.

What is the gain of the multimodal approach? By combining optical microscopy with synchrotron X-ray imaging on the same scan area before and after stress testing, the team could separately identify degradation in the bulk perovskite crystal, at the interfaces with charge transport layers, and in the spatial distribution of charge extraction. Standard macroscopic measurements cannot distinguish between these different loss mechanisms.

When Passivation Becomes a Double Edged Sword

The most cautionary finding of the study concerns surface passivation, a widely used strategy in which a thin chemical treatment is applied to the perovskite surface before depositing the electron transport layer on top. Passivation treatments can dramatically boost the open circuit voltage of a cell by suppressing the non-radiative recombination that normally occurs at this interface.

The team tested two passivation treatments on the same perovskite composition and compared them to a control device and to cells where the hole transport layer material was changed instead. Both passivation strategies reduced PCE disorder and raised initial device efficiency, with the best treated cells achieving open circuit voltages around 0.1 volts higher than the control.

But after operational stress, the results were sobering. The device with the adjusted hole transport layer degraded substantially and showed losses attributable to the interface rather than the perovskite bulk. The passivated devices showed open circuit voltage losses comparable to the unpassivated control, despite their higher starting values. Only one passivation treatment preserved the short circuit current well under stress, while the other showed heavier losses.

The disturbing conclusion is that a passivation treatment can reduce PCE disorder and make a device look excellent in initial characterisation while actually creating an interface that is less stable under real operating conditions. The two goals, high initial performance and robust long term stability, do not automatically go together, and screening only for initial performance can lead engineers in the wrong direction.

Reading the Device Before It Fails

What the study ultimately demonstrates is that operando nanoscale microscopy can function as a predictive tool. The spatial distribution of charge extraction efficiency in a freshly made device carries information about how that device will age. For devices where the interfaces are already stable, lower PCE disorder in the as-made cell predicts better retention of performance after stress.

This opens a practical path. Rather than running weeks of accelerated stability tests on every device variant, researchers could image the fresh device and use its PCE disorder as a fast proxy for likely stability. During manufacturing, a scaled version of the optical components of this toolkit could serve as a quality control screen, identifying devices likely to fail before they are deployed.

The study also points clearly toward the priorities that should guide the next phase of perovskite solar cell engineering. Stable interfaces must come first. Until the boundaries between the perovskite and the layers that carry away its charges are reliably engineered, neither compositional tuning nor passivation can guarantee a long lived device. Once those interfaces are stable, reducing the spatial disorder in charge extraction becomes the key lever for improving both performance and longevity.

The perovskite crystal itself, the study concludes, can be more chemically imperfect than researchers once feared. It is the invisible surfaces, the junctions between layers that no standard measurement sees directly, that ultimately decide whether a solar cell survives.

Publication Details: Year of publication: 2025 Journal: Nature Energy Publisher: Springer Nature Volume / Pages: Volume 10, January 2025, pp. 66–76 DOI: https://doi.org/10.1038/s41560-024-01660-1

Credit & Disclaimer: This article is based on the peer reviewed research paper. All scientific facts, findings, and conclusions presented here are drawn directly from the original study and remain unchanged. This popular science article is intended purely for general educational purposes. Readers are strongly encouraged to consult the full research article for complete experimental data, methodology, and detailed scientific analysis.