A new 2026 study connects microscopic defect populations to long-term photovoltaic degradation—and demonstrates a route toward more commercially viable perovskite solar cells
Perovskite solar cells have made extraordinary progress in efficiency, but one fundamental challenge continues to separate laboratory performance from large-scale commercialization: stability.
A solar cell can deliver an impressive efficiency when it is first fabricated, but that number means considerably less if the device rapidly loses performance during illumination, heating, electrical operation or outdoor exposure.
A new study published in Advanced Materials in June 2026 offers an important perspective on this problem. Qiu Xiong, Can Wang, Xiaofeng Huang and co-workers report that deep-level defects, despite being present at concentrations roughly three orders of magnitude lower than commonly discussed shallow-level defects, can dominate the degradation of perovskite solar cells. The researchers identify two particularly important defect species—(I_{FA}) and (I_{Pb})—and develop a molecular passivation strategy designed specifically to suppress their impact.
The result is not simply another incremental efficiency improvement. The study connects defect physics, energy losses, degradation, lifetime and levelized cost of electricity in a single device-engineering strategy.
The central question: Which defects actually control degradation?
Defects are unavoidable in semiconductor materials.
In a perovskite absorber, imperfections can arise from vacancies, antisite defects, under-coordinated atoms, grain boundaries and chemical reactions occurring during operation. Some defects create relatively shallow electronic states, while others generate deep electronic states inside the bandgap.
That distinction is important.
A defect does not need to be abundant to be technologically important. A small population of highly detrimental deep-level defects can introduce efficient non-radiative recombination pathways and progressively undermine the electrical quality of the device.
The new study addresses precisely this issue.
Instead of assuming that the most abundant defects must be the most important, the researchers quantitatively analyze the evolution of defect states during degradation using capacitance-frequency spectroscopy combined with detailed-balance analysis. Their conclusion is striking:
Deep-level (I_{FA}) and (I_{Pb}) defects are identified as the primary defects responsible for device degradation, even though their concentrations are approximately three orders of magnitude lower than those of commonly considered shallow defects.
In other words, defect concentration alone is not a sufficient indicator of degradation risk.
This is one of the most important messages of the work.
Why deep-level defects are so damaging
A photovoltaic device needs to separate and extract photogenerated electrons and holes before they recombine.
Deep electronic defect states can interfere with this process by acting as recombination centers. In simplified terms, instead of allowing photogenerated carriers to contribute to useful current, these defect states provide an energetically favorable pathway for carriers to recombine.
The consequences can include:
- increased non-radiative recombination;
- reduced quasi-Fermi-level splitting;
- increased voltage losses;
- poorer carrier extraction;
- deterioration of the photovoltaic parameters;
- accelerated performance degradation during operation.
The study therefore shifts attention from simply asking:
“How many defects are present?”
to the more important question:
“Which defects dominate the electronic losses and degradation pathway?”
That distinction could be highly significant for future perovskite manufacturing.
A quantitative way to track degradation
One of the notable aspects of the research is the use of capacitance-frequency spectroscopy to quantitatively examine defect populations.
Capacitance measurements can provide information about electrically active states within a semiconductor. By examining how capacitance changes as a function of frequency, researchers can extract information about defect states and their response to electrical perturbation.
The authors combine this information with detailed-balance theory to determine which defect states have the greatest influence on the photovoltaic performance.
This approach is valuable because conventional characterization can sometimes identify that defects exist without establishing whether a particular defect population is actually responsible for the dominant device loss.
Here, the researchers establish a stronger connection between:
defect evolution → electronic loss → photovoltaic degradation.
That is a much more useful framework for device engineering.
The surprising result: fewer defects can matter more
Perhaps the most counterintuitive result of the research is the difference between defect abundance and defect importance.
The deep-level (I_{FA}) and (I_{Pb}) defects occur at concentrations approximately 1,000 times lower than the shallow-level defects discussed in the study.
Yet they emerge as the dominant degradation-related defect states.
This has an important implication for perovskite solar-cell research.
A passivation strategy that simply reduces the overall defect density may not necessarily produce the maximum improvement.
Instead, researchers and manufacturers may need to identify and selectively suppress the small population of electronically dominant defects.
This could make defect engineering considerably more targeted.
The researchers’ solution: 3TU²⁺ molecular coordination
After identifying the degradation-related deep-level defects, the researchers developed a passivation strategy based on dual-end electropositive 3TU²⁺ ions.
The strategy is described as a non-intercalary ligand coordination approach.
The objective is to coordinate with and passivate the degradation-induced deep-level defect states without relying on an intercalation process that could disrupt the perovskite structure.
This is an important design principle.
A successful passivation molecule has to do more than simply bind to a defect. It must improve the electronic environment without introducing a new barrier to charge transport or compromising the structural integrity of the absorber.
The study reports that the 3TU²⁺ strategy effectively passivates the degradation-induced deep-level defects.
Rear-interface energy loss drops by more than half
One of the clearest quantitative demonstrations of the benefit is the reduction in energy loss at the rear interface.
Before the passivation strategy, the reported rear-interface energy loss was:
1.46%
After the 3TU²⁺ treatment, it was reduced to:
0.62%
That represents an absolute reduction of:
0.84 percentage points
and a reduction of approximately:
58%
relative to the original 1.46% value.
The paper describes this as an approximately order-of-magnitude improvement in the relevant energy-loss alignment, with the reported values changing from 1.46% to 0.62%.
This improvement is connected to better quasi-Fermi-level splitting alignment.
For a photovoltaic device, that matters because the quasi-Fermi-level splitting is closely related to the maximum voltage that the absorber can generate.
Reducing non-radiative losses therefore has the potential to translate directly into improved voltage and overall device efficiency.
Certified efficiency reaches 25.56%
The optimized device achieved a certified power conversion efficiency of 25.56%.
The significance of this number is not simply that it exceeds another efficiency threshold.
The more interesting point is the combination:
25.56% certified efficiency + defect-specific passivation + long projected lifetime.
For perovskite photovoltaics, the industry challenge is increasingly moving away from achieving high initial PCE alone.
A commercially relevant device must simultaneously provide:
- high efficiency;
- operational stability;
- reproducible manufacturing;
- low degradation;
- reasonable material and processing costs;
- predictable lifetime;
- competitive electricity-generation economics.
This research attempts to address several of those requirements simultaneously.
T₈₀ lifetime: more than 10 years
Efficiency is only half of the commercialization equation.
The study reports an extrapolated T₈₀ lifetime exceeding 10 years, according to the reported ISOS-LC-1 protocol.
T₈₀ refers to the time required for the device to decline to 80% of its initial performance.
If the initial efficiency is represented as:
100% → initial performance
then T₈₀ corresponds to:
80% → remaining performance
The reported extrapolated lifetime therefore suggests that the treated devices could maintain at least 80% of their initial performance for more than a decade under the conditions and extrapolation methodology used by the researchers.
It is important to emphasize the word extrapolated.
A projected lifetime is not equivalent to a 10-year field demonstration. Long-term commercial validation still requires standardized testing, extended outdoor operation and independent assessment across multiple device batches and module formats.
Nevertheless, the result is significant because it links a microscopic defect-passivation mechanism to a lifetime metric relevant to commercialization.
From defect physics to electricity cost
The researchers take the analysis one step further.
They calculate the impact of improved performance and lifetime on the levelized cost of energy (LCOE).
The reported LCOE is:
$0.148 kWh⁻¹
The authors state that this is comparable to silicon photovoltaics.
LCOE is particularly important because the value of a solar technology is ultimately determined not by efficiency alone, but by the cost of producing electricity over the operating lifetime of the system.
A simplified relationship is:
Lower LCOE = lower total lifetime electricity cost
Improving the initial efficiency increases energy production.
Improving lifetime increases the amount of electricity generated before replacement or significant performance loss.
Therefore:
Higher efficiency + longer lifetime → greater lifetime energy yield → potentially lower LCOE
This is why the defect work in this study has significance beyond materials science.
Why this matters for commercial perovskite modules
For commercial solar manufacturing, the most interesting aspect of the paper may not be the 25.56% efficiency figure.
It is the methodology.
The research suggests a pathway toward identifying which microscopic defects are economically important.
That creates a potential development workflow:
Step 1 — Identify defect populations
Use electrical and spectroscopic characterization to determine which defect states exist.
Step 2 — Quantify their electronic impact
Determine which states actually contribute to recombination and energy loss.
Step 3 — Track their evolution during degradation
Rather than examining only the fresh device, monitor how defect populations change during operation.
Step 4 — Develop selective passivation
Design ligands or interface treatments that target the dominant defect states.
Step 5 — Validate device-level improvement
Measure changes in voltage, current, fill factor, efficiency and stability.
Step 6 — Translate the improvement into economics
Use lifetime and efficiency data to estimate LCOE and commercial viability.
This is a more systematic approach than simply screening large numbers of passivation molecules and selecting the one that produces the highest initial PCE.
What the study changes about defect engineering
The paper challenges a common intuition in materials engineering:
The most abundant defect is not necessarily the most important defect.
A defect’s importance depends on its electronic activity.
A very low concentration of deep-level defects can potentially have a much greater impact on device performance than a much larger population of shallow defects.
This suggests that future research should increasingly combine:
materials characterization + semiconductor physics + degradation analysis + device modelling.
The objective should not simply be to minimize the total number of defects.
It should be to minimize the defects that dominate recombination and degradation.
What this means for perovskite-silicon tandems
The findings are also relevant to the broader development of high-efficiency tandem photovoltaics.
Perovskite top cells in tandem architectures must operate with high voltage, high efficiency and long-term stability.
Any deep-level defect that increases non-radiative recombination can directly undermine the voltage advantage needed for tandem architectures.
Interface engineering is therefore becoming increasingly important.
The new work reinforces the idea that future high-performance perovskite devices will require simultaneous control of:
- bulk defects;
- surface defects;
- buried interfaces;
- ionic defects;
- strain;
- carrier recombination;
- chemical degradation.
The challenge is no longer simply making a high-quality perovskite film.
It is maintaining the electronic quality of that film through thousands of hours of operation.
The bigger picture: stability is becoming a defect-physics problem
Perovskite solar-cell research has traditionally emphasized several major degradation pathways:
- moisture;
- oxygen;
- heat;
- light;
- ion migration;
- phase instability;
- electrode reactions;
- interface degradation.
The new research adds another important layer:
the evolution of electrically active deep-level defects during degradation.
This is important because degradation is not necessarily a single chemical event.
A device can gradually develop new defect states, which increase recombination, which reduces electrical performance, which can in turn accelerate additional degradation processes.
The resulting system can behave like a feedback loop:
operation → defect generation → recombination → energy loss → performance degradation
Breaking that loop through targeted defect passivation could therefore be a powerful route toward long-lived PSCs.
Key numbers from the study
ParameterReported resultPublicationAdvanced Materials, 2026DOI10.1002/adma.73679Identified dominant deep defects(I_{FA}), (I_{Pb})Deep-defect concentration vs. shallow defects~3 orders of magnitude lowerRear-interface energy loss1.46% → 0.62%Certified PCE25.56%Extrapolated T₈₀>10 yearsReported protocolISOS-LC-1Reported LCOE$0.148 kWh⁻¹Passivation strategyDual-end electropositive 3TU²⁺ ligand coordination
The values above are taken from the published article’s abstract and publisher metadata; additional numerical values from figures or supplementary datasets should be checked directly against the paper’s full text before being used for formal technical or commercial claims.
What the research means for the solar industry
The commercialization of perovskite photovoltaics will ultimately depend on more than record efficiencies.
Manufacturers need devices that can deliver predictable energy for years under real operating conditions.
That makes the central message of this research particularly relevant:
A tiny population of highly active defects can matter more than a large population of relatively benign defects.
The implication is powerful.
Instead of treating all defects equally, future perovskite manufacturing could increasingly use defect-specific diagnostics and targeted molecular engineering to identify the few defect populations that dominate energy loss and degradation.
The result reported by Xiong and colleagues demonstrates the potential of this approach: by targeting degradation-induced deep-level defects with a 3TU²⁺ coordination strategy, the researchers achieved a certified 25.56% efficiency, reduced the reported rear-interface energy loss from 1.46% to 0.62%, projected a T₈₀ lifetime beyond 10 years, and calculated an LCOE of $0.148 kWh⁻¹.
These numbers do not, by themselves, establish that perovskite photovoltaics are commercially ready. Module-scale validation, manufacturing reproducibility, field testing, encapsulation, environmental reliability and independent lifetime assessments remain essential.
But they demonstrate an increasingly important direction for the field:
The path to durable perovskite solar cells may depend less on eliminating every defect and more on finding—and neutralizing—the defects that matter most.
Reference
Xiong, Q., Wang, C., Huang, X., et al. “Quantifying Deep-Level Defects-Dominated Degradation for Commercially Viable Perovskite Solar Cells.” Advanced Materials (2026), e73679. DOI: 10.1002/adma.73679. First published online 11 June 2026.
