Perovskite Solar Module Reaches 22.7% Efficiency With Slot-Die Coating in Open Air

A research team led by Professor Yang Dong at the Dalian Institute of Chemical Physics, Chinese Academy of Sciences (CAS), together with Dalian Jiaotong University, has fabricated a methylammonium-free (MA-free) perovskite solar module using slot-die coating in ambient air. According to a paper published online in Advanced Materials on October 27, 2026 (first author: Lianjie Duan et al.), a glass-substrate module with 14 cells connected in series (active area 54.80 ${\text{cm}}^2$) achieved a power conversion efficiency (PCE) of 22.70%. In scaling up perovskite solar cell production, one major challenge in moving from lab-standard spin coating to continuous coating processes is controlling solvent evaporation and crystallization in air. In this study, combining two additives widened the processing window between coating and annealing, allowing the team to form uniform films even over large areas.

Two thiocyanate additives extend the crystallization window to over 100 seconds

Methylammonium (MA), widely used in perovskite solar cells, is highly volatile and prone to thermal decomposition. For this reason, ${\mathrm{FA}}_x{\mathrm{Cs}}_{1-x}{\mathrm{PbI}}_3$ materials, which combine thermally stable formamidinium (FA) and cesium (Cs), have attracted attention. However, these materials crystallize slowly in air and are prone to phase separation. In liquid-film processes such as slot-die coating, film quality varies greatly with the time between applying the ink and starting heat treatment (annealing). Note that slot-die coating was used here only to form the perovskite light-absorbing layer; the whole module was not printed.

The paper defines the time from ink coating to vacuum flash treatment as $\Delta t_1$ and the subsequent wait before annealing as $\Delta t_2$, with their sum defined as the processing window $\Delta t$. To widen this window, the team developed a method of adding formamidinium thiocyanate (FASCN, 2 mg/mL) and rubidium thiocyanate (RbSCN, 5 mg/mL) to the precursor ink. Using the two additives together is intended to control crystal formation and stabilization at the same time.

Each additive plays a different role. In comparison experiments with single additives, FASCN promoted nucleation and crystal growth during vacuum flash treatment but could not maintain a uniform crystal phase on its own. RbSCN, by contrast, had a relatively modest effect on early nucleation but suppressed the formation of the yellow phase ($\delta$ phase), which does not contribute to photovoltaic conversion, and helped stabilize the light-absorbing black phase ($\alpha$ phase).

This difference was also confirmed in experiments observing the crystallization process. In in situ photoluminescence (PL) measurements with the waiting time $\Delta t_2$ extended to 5 minutes (300 seconds), the luminescence signal from $\alpha$-FAPbI₃ disappeared partway in the additive-free control film, while the dual-additive film maintained a stable signal. In X-ray diffraction (XRD) measurements at the 70°C heat-treatment stage, the additive-free film and the FASCN-only film remained in the $\delta$ phase, which is unsuitable for photovoltaics, while the RbSCN-containing film and the dual-additive film quickly converted to the $\alpha$ phase.

The team explains that shifts in the $\mathrm{C}\equiv \mathrm{N}$ stretching vibration of the thiocyanate ion (${\mathrm{SCN}}^{-}$) seen in infrared spectroscopy (FTIR), and changes in the bonding state around lead (Pb) seen in X-ray photoelectron spectroscopy (XPS), indicate that the additives form coordination bonds with lead complexes in the precursor solution. In addition, density functional theory (DFT) calculations estimated that rubidium ions (${\mathrm{Rb}}^{+}$) entering the A-site of the crystal lattice relieve lattice strain, lowering the activation energy required for the $\delta$-to-$\alpha$ phase transition from 0.68 eV to 0.64 eV. However, this energy reduction is based on theoretical calculation and was not measured directly in experiments.

In comments published on EurekAlert!, co-author Zhihua Zhang of Dalian Jiaotong University explained that, "rather than tightening manufacturing process control further, we increased the tolerance of the crystallization process itself to manufacturing variation." Thanks to the synergy of the two additives, the processing window allowing uniform crystallization exceeded 100 seconds, and the team formed dense films with large grains and low residual stress.

26.16% for a small cell and 22.70% for a 54.80 cm² module

When evaluating solar cell performance, it is necessary to distinguish between small-area single cells made in the lab and large-area modules of multiple series-connected cells. This study reports measurements for both. Measurements used the standard simulated sunlight condition AM 1.5G (irradiance 100 ${\text{mW/cm}}^2$).

The best small-area cell made under optimal conditions had an open-circuit voltage ($V_{\mathrm{oc}}$) of 1.19 V, a short-circuit current density ($J_{\mathrm{sc}}$) of 26.15 ${\text{mA/cm}}^2$, and a fill factor (FF) of 84.19%, reaching a PCE of 26.16%. The integrated current density calculated from the external quantum efficiency (EQE) spectrum was 25.80 ${\text{mA/cm}}^2$, consistent with the current-voltage measurements. In a 300-second operation at the maximum power point (MPP), the cell also maintained a stabilized efficiency of 26.14% at an operating voltage of 1.06 V.

The average efficiency of 10 small-area cells made to check reproducibility was about 26%, compared with about 20% for additive-free control cells. These are approximate figures, however, and not results of a rigorous statistical significance test. To evaluate the uniformity of the large-area coated film, eight planar p-i-n cells were cut from a 10 cm × 10 cm substrate and measured, and each showed an efficiency of about 24%. The efficiency deviation was ±1.16 or more for additive-free films but stayed below ±0.48 for dual-additive films, confirming reduced variation in film quality.

• 複合添加剤（本研究）
• 無添加（対照群）

データを表で見る

	複合添加剤（本研究） (%)	無添加（対照群） (%)
小面積セル（単一素子）	26.16	20
大面積モジュール（54.80 cm²）	22.7	15.8

The large-area module, with 14 series-connected cells on a glass substrate, reached a maximum PCE of 22.70% over an active area of 54.80 ${\text{cm}}^2$. The additive-free control module reached only 15.80%, so the dual additives improved efficiency by 6.90 percentage points. However, the paper does not report an official certified efficiency from a third-party certification body; the 22.70% is a value the research team measured in its own laboratory.

1,300-hour continuous operation test at 65°C in a nitrogen atmosphere

For practical use of perovskite solar cells, durability under prolonged exposure to heat and light matters as much as high conversion efficiency. The team ran accelerated degradation tests under several conditions using unencapsulated devices.

In a test under continuous AM 1.5G simulated sunlight (100 ${\text{mW/cm}}^2$) with the device held at 65°C and operating at the maximum power point, the device with the dual additives retained 97.76% of its initial efficiency after 1,300 hours. The additive-free device tested under the same conditions fell to 65.86% of its initial efficiency. The 1,300 hours correspond to about 54.2 days of continuous operation, indicating that the dual additives substantially suppressed degradation during long-term operation.

This result cannot, however, be equated with outdoor durability. The test was conducted in a nitrogen atmosphere that excludes moisture and oxygen, which differs from real outdoor conditions. The paper labels the test "ISOS-L-3," but under the strict classification of the international consensus statement on solar cell stability testing published in 2020 (the ISOS consensus), a variant test conducted in an inert atmosphere would correspond to "ISOS-L-3I," with an "I" appended. ISOS is a set of procedures for standardizing degradation test conditions and evaluation methods, and it does not replace International Electrotechnical Commission (IEC) certification for assessing long-term outdoor reliability.

The degradation-suppressing effect of the dual additives was also confirmed in tests without light exposure. In a thermal stability test with dark heating at 85°C in a nitrogen atmosphere, the dual-additive device retained over 96% of its initial efficiency after 1,000 hours, while the additive-free device dropped below 90% within 600 hours. In a test where devices were stored in the dark, without generating power, in ordinary air at 30–50% relative humidity, the efficiency of the additive-free device fell to 41% of its initial value after 1,600 hours, while the dual-additive device maintained over 97%. These results support improved resistance to heat and humidity, but dark storage tests need to be distinguished from lifetime evaluation in outdoor power-generating environments, where light, moisture, and oxygen act simultaneously.

Comparison with prior work and remaining hurdles to mass production

When comparing the performance of large-area perovskite solar cells, attention must be paid not only to conversion efficiency but also to the coating method and the measured area. A 2025 paper in Joule involving the same research group reported a module in which the electron transport layer, perovskite absorber layer, and hole transport layer were all formed by blade coating. The main differences from the present study are as follows.

Item	This study (Advanced Materials 2026)	Prior study (Joule 2025)
Coating method and target layers	Slot-die coating (perovskite layer only), in air	All-blade coating (electron transport, perovskite, and hole transport layers), in air
Module area and definition	Active area: 54.80 ${\text{cm}}^2$	Aperture area: 57.20 ${\text{cm}}^2$
Highest reported efficiency (PCE)	22.70% (author-reported)	22.76% (author-reported)
Certified efficiency / stabilized output	No certified value / 26.14% in a 300-second MPP measurement of a small cell	Certified PCE: 21.60% / Stabilized output: 22.20%
Efficiency of control group (e.g., additive-free)	15.80% (additive-free module)	20.72% (control module)

The numbers from the two studies do not allow a simple judgment of which performs better. First, the 54.80 ${\text{cm}}^2$ in this study is the "active area," which excludes the non-generating regions (dead area) between series-connected cells. The 57.20 ${\text{cm}}^2$ in the 2025 study, by contrast, is the "aperture area," based on the entire opening of the module, so the area definitions used to calculate efficiency differ. In addition, the prior study reported a certified efficiency of 21.60% from an independent certification body, whereas the 22.70% here is a measurement by the authors themselves. The manufacturing processes also differ: in this study only the perovskite absorber layer was formed by slot-die coating, not all layers including the electron and hole transport layers.

What matters in this study is that it showed that widening the processing window from coating to crystallization makes it possible to form uniform perovskite films even over large areas. Beyond achieving 22.70% efficiency in a 54.80 ${\text{cm}}^2$ module, suppressing device-to-device efficiency variation on a 10 cm × 10 cm substrate is meaningful for considering scale-up to large-area manufacturing. On the other hand, application to continuous roll-to-roll processes remains a future goal, and the paper does not present validation on an actual continuous coating line, product yield, coating speed, or manufacturing cost estimates.

Going forward, in addition to verification of module efficiency by a third-party certification body, modules with proper encapsulation will need to be operated outdoors over long periods to evaluate durability. Whether the crystallization-control effect confirmed here can be reproduced in continuous manufacturing processes will be the key to turning laboratory results into commercial production.