Korean Team's Perovskite Solar Cells Keep 95% of Efficiency After 1,000 Hours of Heat and Humidity

Perovskite solar cells are a promising next-generation technology, but they have long faced a trade-off between power conversion efficiency and durability. One cause is "spiro-OMeTAD," an organic semiconductor widely used in the hole transport layer. The material is important for achieving high power conversion efficiency (PCE), but the lithium salts and volatile additives mixed in to raise its conductivity have contributed to device degradation from heat and moisture. Using fewer additives could improve durability, but until now that meant insufficient conductivity and lower efficiency.

To address this, a joint research team led by Dr. SeJin Ahn of the Korea Institute of Energy Research (KIER) and Dr. Dong Suk Kim of the Ulsan National Institute of Science and Technology (UNIST) developed a new manufacturing process that combines a catalytic reaction with filtration. The findings were published online in Advanced Science on January 26, 2026. In the new process, fine nickel oxide (NiOₓ) particles speed up the oxidation of spiro-OMeTAD while also capturing lithium ions on their surfaces, and filtration then removes them from the solution. Perovskite solar cells made this way retained 95% of their initial conversion efficiency after 1,000 hours in a hot, humid environment of 85°C and 85% relative humidity. However, this result was obtained under specific laboratory conditions and does not directly prove the long-term service life of commercial solar panels.

{
  "type": "bar",
  "title": "Normalized efficiency retention after 1,000 hours of damp-heat testing (85°C, 85% RH)",
  "unit": "%",
  "categories": ["Developed device (NiOx catalysis and filtration extraction)"],
  "series": [
    {
      "name": "Retention relative to initial conversion efficiency",
      "values": [95]
    }
  ],
  "caption": "Accelerated aging test of encapsulated, lab-scale n-i-p perovskite solar cells. The 95% figure is retention relative to initial efficiency, not conversion efficiency itself. The absolute conversion efficiency values at the start and end of the test were not given in the published materials we could confirm.",
  "source": "Advanced Science (DOI: 10.1002/advs.202521825)"
}

What stands out about this result is not simply that the devices survived a 1,000-hour durability test. It is that the team removed components that cause degradation before the film is formed, without using the additives that have conventionally been considered necessary for high efficiency. Understanding the technical significance requires looking at both the test conditions and the chemical reactions in the hole transport layer.

What the 1,000-hour damp-heat test and continuous operation test showed

The team's finding of "95% of initial efficiency after 1,000 hours" does not mean the cells convert 95% of sunlight into electricity. It is the "normalized efficiency retention": taking the conversion efficiency at the start of the test as 100%, performance equivalent to 95% of that remained after 1,000 hours. And while 1,000 hours is roughly 42 days, that alone cannot establish that the cells would last 20 to 30 years on a home rooftop. The absolute conversion efficiency values at the start and end of the damp-heat test are also not stated in the published materials we could confirm.

The team fabricated small, lab-scale n-i-p perovskite solar cells. To evaluate durability, they used damp-heat conditions of 85°C and 85% relative humidity, which are also used in international reliability testing of solar cells. Encapsulated devices were placed in a dark, hot, humid environment to examine how heat and moisture affect performance. This test evaluates moisture ingress and heat-driven material degradation; it differs from conditions in which a cell keeps generating power under sunlight.

Separately, the team ran a maximum power point tracking (MPPT) test, in which simulated sunlight is applied continuously and power is actually drawn from the device while performance changes are monitored. The illumination condition was 1 sun, meaning continuous light at the standard intensity of 1,000 watts per square meter. In this continuous operation test, devices made with the new process also retained more than 90% of their initial conversion efficiency after 500 hours. By contrast, the comparison devices in the paper showed significant performance loss over the same period.

Good results in both the dark damp-heat storage test and the continuous operation test under illumination are meaningful. However, the two tests were run under different conditions, so the 95% retention figure cannot be interpreted as durability during continuous power generation. Outdoors, heat, moisture, light, and electrical load act simultaneously, so there are also limits to predicting real-world lifespan from either test result alone.

How nickel oxide captures lithium for removal by filtration

In perovskite solar cells, the hole transport layer (HTL), which carries holes (positively charged carriers) generated in the light-absorbing layer to the electrode, plays a key role in determining efficiency. The organic semiconductor widely used for this layer is spiro-OMeTAD. On its own, however, spiro-OMeTAD has low electrical conductivity, so it is generally p-doped through oxidation to improve hole transport.

In the typical conventional recipe, lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) is added as a dopant, along with 4-tert-butylpyridine (tBP). The tBP helps dissolve the LiTFSI and promotes uniform doping. But tBP is volatile, and in high-temperature environments it can change the composition and microstructure of the film, potentially leading to voids and defects. Meanwhile, lithium ions (Li⁺) from LiTFSI are small and mobile, so heat and electrical stress can drive them to diffuse through the device and accumulate at interfaces, triggering degradation reactions. Such ion migration and the formation of byproducts reduce the conductivity of the hole transport layer and impede charge transport.

To tackle this problem, the team devised a method that carries out the oxidation needed for doping in solution before film formation, removing degradation-causing components before the film is made. At its core is micrometer-sized nickel oxide (NiOₓ) powder. NiOₓ not only promotes the oxidation reaction but also captures lithium ions on its surface.

The new process proceeds as follows:

1. Adding NiOₓ to the precursor solution: Micrometer-sized NiOₓ powder is added to a solution containing spiro-OMeTAD and LiTFSI.
2. Accelerating oxidation: Redox reactions on the NiOₓ rapidly oxidize spiro-OMeTAD, generating the radical cations needed for hole transport. Spectroscopic measurements confirmed that oxidation proceeded within 5 seconds under optimized conditions.
3. Capturing lithium ions: The chemical state of the NiOₓ surface changes, promoting the dissolution and dissociation of LiTFSI without tBP. The released Li⁺ is captured by coordinating to oxygen sites on the NiOₓ surface.
4. Separation by filtration: When the solution is passed through a filter, the micrometer-sized NiOₓ particles are trapped and removed together with the lithium ions and related byproducts bound to their surfaces.
5. Forming the hole transport layer: The filtered solution is applied to a substrate by spin coating, forming a hole transport layer containing oxidized spiro-OMeTAD.

A distinctive feature of this method is that NiOₓ does not remain as a component of the finished solar cell. Some solar cells use NiOₓ as the hole transport layer itself, but in this study it serves as a redox catalyst and lithium adsorbent that treats the solution before film formation. Once the NiOₓ particles have completed the necessary reactions, they are filtered out along with the captured lithium ions. In other words, the catalyst is used midway through manufacturing and never carried into the final film.

The team evaluated the lithium removal using time-of-flight secondary ion mass spectrometry (ToF-SIMS). The results showed that lithium-derived signals dropped substantially in films made with the new process, indicating that residual lithium inside the device was greatly reduced compared with conventional processing. However, a large drop in the detected signal does not mean residual lithium has fallen to zero. The long-term effects of trace residues will need continued evaluation.

25.24% efficiency maintained, and another factor behind the durability

If improved durability came with a major loss of efficiency, the practical benefit would be limited. But the maximum power conversion efficiency (PCE) of perovskite solar cells made with the new process reached 25.24%. A comparison device made with the conventional additive recipe using LiTFSI, tBP, and other components reached a maximum of 25.21%, so the two performed at nearly the same level.

The difference is only 0.03 percentage points, so it would not be appropriate to conclude that power generation performance itself improved significantly. Solar cell efficiency measurements vary with factors such as the light source spectrum, measurement area, voltage scan conditions, and hysteresis (the phenomenon in which measured values differ depending on the direction of the voltage sweep). A slight difference in peak values therefore cannot establish which device performs better. The value of this result is that efficiency comparable to conventional devices was secured without volatile additives and with substantially less lithium.

Lithium removal is not the only reason durability improved. The paper also highlights changes in the glass transition temperature (Tg), which relates to the thermal stability of the hole transport layer. The glass transition temperature indicates the range at which molecular motion in an amorphous material becomes active and the material softens more easily. In the hole transport layer, this temperature characteristic affects whether the film maintains its shape and electrical properties at high temperatures.

The tBP used in the conventional recipe acts somewhat like a plasticizer in the spiro-OMeTAD film, lowering its glass transition temperature. When a film's thermal stability drops, molecules are more likely to rearrange and aggregate in hot environments, which can damage conductive pathways and film uniformity. By contrast, the team reports that the new method, which advances oxidation without tBP, keeps the film's glass transition temperature high. Suppressing lithium-ion migration while also improving the film's own thermal stability is thought to have contributed to the improved durability.

The new process also made the oxidation of spiro-OMeTAD more uniform and improved the electrical properties of the hole transport layer. The paper reports that microscope observations of the film and surface potential measurements showed that NiOₓ treatment made the material distribution more uniform and reduced local variations in conductivity. Such film uniformity is also important for extracting charge efficiently. The new process does more than reduce degradation-causing substances; it also improves the formation of the hole transport layer itself.

Why a 1,000-hour accelerated test does not guarantee long outdoor life

The durability test is an important result, but it does not mean the lifespan problem of perovskite solar cells has been completely solved. Accelerated aging tests are a way to evaluate in a short time how much materials and devices degrade under particular environmental stresses. Yet solar cells used outdoors face not only heat and moisture but also light, temperature swings, and mechanical loads acting in combination. Even if a device shows high durability under fixed conditions, the result cannot be applied directly to long-term outdoor use.

The 85°C, 85% relative humidity damp-heat test is widely used as a standard method for evaluating the moisture and heat resistance of solar modules. However, this damp-heat test was conducted in the dark with temperature and humidity held constant. Outdoor solar cells, by contrast, repeatedly warm up during the day and cool at night. Because they are stacks of multiple thin films with different coefficients of thermal expansion, such temperature changes can cause interface delamination and microscopic cracks.

Outdoors, cells are also affected by the breakdown of organic materials from ultraviolet light, mechanical loads from wind, rain, and snow, electrical stress from fluctuating sunlight, and condensation. This damp-heat test did not reproduce all of these. And to say how many years of outdoor use 1,000 hours of accelerated testing corresponds to, degradation models tailored to the material composition and encapsulation method must be checked against data from real outdoor environments. Perovskite solar cells come in many material systems and device structures, and no single lifetime conversion factor common to all of them has been established.

Caution is also needed in describing the lifespan of perovskite solar cells uniformly as "a few years." Durability varies greatly with the perovskite composition, charge transport layer materials, device structure, encapsulation technology, and operating environment. When comparing lifetime data across studies, one must check whether test conditions and the definition of efficiency retention match. This study succeeded in suppressing an important degradation pathway in the hole transport layer, but confirming the outdoor durability required of commercial products will take further long-term testing.

Challenges for large-area modules and tandem solar cells

Announcing the results, Dr. SeJin Ahn expressed hope that progress toward solving the long-standing lifespan problem could become an important milestone in promoting the commercialization of perovskite and tandem solar cells. Tandem cells, which layer perovskite on top of silicon solar cells, are attracting particular attention because each layer can efficiently absorb different wavelengths of light, offering the potential to exceed the efficiency of single-junction silicon cells. Efficiencies above 30% are anticipated, but because multiple material layers are combined, the long-term stability of each layer will determine practical viability.

Still, the conditions required of manufacturing differ greatly between cells a few millimeters across made in a lab and the large solar modules actually sold. Applying this result to industrial products will require not only materials chemistry but also establishing processes suited to mass production. In particular, how to incorporate solution treatment with catalyst powder and the filtration step into a production line is a key issue.

The first challenge is scaling up the catalytic reaction and filtration steps. In the lab, a small amount of solution can be treated by adding NiOₓ powder, stirring briefly, and filtering with a syringe filter. In a factory handling large volumes of precursor solution, however, catalyst particles and solution must be brought into uniform contact while stable reaction conditions are maintained. It is also necessary to reliably remove NiOₓ particles and lithium-containing residues after the reaction while preventing filter clogging. Whether these steps can run continuously will affect production efficiency and quality stability.

The second challenge is compatibility with large-area coating techniques. The spin coating used in this experiment forms a uniform thin film by spinning the substrate at high speed and using centrifugal force. But it makes poor use of material and is limited in how well it applies to large substrates, so mass production would likely consider methods such as slot-die coating or inkjet printing. Whether the tBP-free solution obtained from the new process can maintain appropriate viscosity, wettability, and drying characteristics with these coating methods remains to be verified. If scaling up to large areas causes unevenness in film thickness or doping, the efficiency and durability seen in the lab might not be reproduced.

The third challenge is manufacturing cost and waste handling. In this process, NiOₓ powder is used as a catalyst and then removed by filtration together with lithium ions. Material, equipment, and waste disposal costs will differ depending on whether the powder is discarded or the captured lithium is separated and reused. Even if the catalyst brings large performance gains, high costs for the added step would limit the commercial advantage. Building a system capable of low-cost mass production while maintaining stable quality is essential.

This study revisited the doping process long considered necessary for high-efficiency perovskite solar cells and showed that combining a redox reaction with physical filtration could deliver both performance and durability. In particular, the idea of using a material needed for the reaction midway through manufacturing and removing degradation-causing components before film formation offers a new option for designing hole transport layer materials. At the same time, this result is a demonstration with lab-scale single-junction cells, and applicability to commercial modules has not yet been confirmed. Going forward, the focus will be on verifying manufacturing quality and yield at larger areas, and on whether performance can be maintained over long periods in outdoor environments where multiple degradation factors overlap.