CyphrQ
Menu

Perovskite Solar Cells: Efficiency Records, Durability and Storage

Perovskite cells pass 26% in the lab and a perovskite-silicon tandem reached 34.85%. Durability is the open question. What the records mean.

By CyphrQ.ai

Written with AI assistance and human editorial review. Every figure linked to its source.

9 min read

Perovskite solar cell efficiency has climbed from about 3% in 2009 to over 26% on small lab cells, according to the US Department of Energy's page as read on 26 October 2026. Stacked on silicon, a certified tandem cell reached 34.85% in April 2025, its manufacturer LONGi says. What nobody has shown yet is that these cells last as long as silicon panels outdoors. That is the question to watch.

This post covers the records, why durability is still open, and one grid storage result from a US national laboratory that is worth knowing about.

What is a perovskite solar cell?

A perovskite is a crystal structure, not a single material. The US Department of Energy (DOE) describes perovskite solar cells this way: "Metal-halide perovskites are the main absorbing material," and "a thin layer of perovskite absorbs light, which excites charged particles called electrons." Those electrons are collected as electric current.

Much of the interest comes from how fast efficiency has risen and from the hope of a cheaper way to make cells. That hope is still unproven: DOE lists manufacturability among the main challenges.

How efficient are perovskite solar cells now?

DOE's page gives the headline: "perovskite solar cell efficiency increased from about 3% in 2009 to over 26% today on small area devices (about 0.1 cm2)." DOE's page does not date the word "today". It credits the figure to the Best Research-Cell Efficiency Chart kept by the National Laboratory of the Rockies (NLR), the US national laboratory whose chart page says "NLR maintains" it. Older documents, including LONGi's 2025 announcement below, name the certifying lab as the National Renewable Energy Laboratory (NREL).

Note the size. A tenth of a square centimetre is smaller than a fingernail. A record on a cell that small does not automatically carry over to a full-size panel.

What is a perovskite-silicon tandem cell?

A tandem cell stacks two light absorbers. The top layer (here, perovskite) captures more of the blue end of sunlight, and the bottom layer (silicon) captures more of the red. Together they convert more of the sunlight than either could alone.

In practice: Chinese manufacturer LONGi says that on 21 April 2025 it reached "a conversion efficiency of 34.85%, certified by the U.S. National Renewable Energy Laboratory (NREL)" for a crystalline silicon-perovskite tandem cell. The company lists its earlier steps as 33.9% in November 2023 and 34.6% in June 2024. The page does not give the cell's area or any stability data. DOE's own page says tandems "have reached efficiencies of almost 34%," which looks like it was written before LONGi's April 2025 result.

Is there a newer record? The chart page gives its figures only in the chart image and data files, not in its text, and I could not check those for this post. So, as of 26 October 2026, treat 34.85% as the latest certified tandem figure I could confirm from a primary source, not as the current record. Check the chart itself before quoting it.

Not investment advice. A company record is a research result, not a product you can buy.

How are efficiency records checked?

The Best Research-Cell Efficiency Chart is the reference most labs and companies point to. Its rules matter:

  • Efficiencies must be "confirmed by independent, recognized test labs," such as NLR, AIST in Japan, the EU's JRC-ESTI and Fraunhofer ISE in Germany.
  • Measurements follow standard test conditions set out in IEC 60904-3 Edition 2 or ASTM G173, at 25°C.
  • The area is either the full cell area or an area defined by an aperture (a mask over the cell).

What the chart does not measure: lifetime, cost, or performance outdoors. A chart entry tells you what a cell did on a test bench, once, in controlled light.

Why is durability the open problem?

DOE lists "Cell stability and durability" as one of four main challenges for perovskites, alongside manufacturability. It also says plainly that "perovskite PV is not yet manufactured at scale."

Solar panels are bought to run outdoors for decades. Perovskites have not built up that kind of track record, and a lab bench cannot supply one.

That is why DOE funded an independent test centre. In July 2021, Sandia National Laboratories announced the Perovskite PV Accelerator for Commercializing Technologies (PACT) Center, backed by a $14 million DOE award, with NREL, Los Alamos National Laboratory, the Electric Power Research Institute, Black and Veatch and CFV Labs as partners. Its stated job is to act as a neutral evaluator and to build common test protocols so devices can be fairly compared.

Sandia's Joshua Stein said at the launch that "there is still much that is unknown about this technology, especially in terms of outdoor performance and reliability." The aim of accelerated testing, the release says, is to "predict the long-term reliability of these panels in the real world without having to wait 30 years."

What is PNNL's iron flow battery, and how does it compare with vanadium?

Solar output stops at night. Grid storage fills that gap, and researchers are looking for cheaper, safer ways to store energy for many hours at a time.

In practice: in March 2024, Pacific Northwest National Laboratory (PNNL) reported a lab-scale, all-liquid iron flow battery in Nature Communications. A flow battery stores energy in liquids held in tanks and pumped past electrodes, so you can add storage time by adding bigger tanks. PNNL's key points:

  • It ran for over 1,000 consecutive charging cycles while keeping 98.7% of its maximum capacity.
  • The electrolyte uses iron and a phosphonate chemical (NTMPA) that is already sold in industrial quantities for water treatment.
  • Energy density reached up to 9 Wh/L, against 25 Wh/L for commercial vanadium flow batteries.
  • The work was funded by the DOE Office of Electricity.

PNNL is candid that this is an initial, lab-scale design with lower voltage than vanadium systems. Next steps named in the release are raising voltage and electrolyte concentration, and scaling up at PNNL's Grid Storage Launchpad.

What do these results not show yet?

  1. Perovskite lifetime outdoors. No source I read gives a field lifetime that matches silicon.
  2. Full-size panels at record efficiency. The 26% figure is for about 0.1 cm². LONGi did not publish its cell area.
  3. Manufacturing at scale. DOE says perovskites are not yet in high-volume production.
  4. Cost per kilowatt-hour for either technology. None of the sources give one.
  5. A grid-scale iron flow battery. PNNL's result is lab-scale, with a third of vanadium's energy density.

What perovskite and storage results should you watch for?

  • Independent outdoor data from PACT and from national labs, not just company announcements.
  • Module results, not cell results. Look for "module" or "mini-module" and an area in square centimetres.
  • New entries on the efficiency chart, checked against the chart itself.
  • Scale-up of iron flow batteries at PNNL's Grid Storage Launchpad, with energy density and cycle figures.

Going deeper: for security and IT teams

This section is for technical readers and decision-makers: facilities and IT leads planning on-site solar or storage, sustainability officers, and anyone signing off a procurement that mentions "next-generation" panels.

How to read an efficiency claim

Question Why it matters
Cell, mini-module or module? Small-cell records don't automatically carry over to larger areas. The 26% DOE figure is for about 0.1 cm².
Total area or aperture area? The chart allows either. An aperture can exclude edges and inactive parts.
Which test lab certified it? The chart requires an independent lab such as NLR, AIST, JRC-ESTI or Fraunhofer ISE.
Under what conditions? Chart entries follow IEC 60904-3 Ed. 2 or ASTM G173 at 25°C. Real rooftops run hotter and see variable light.
Is there stability data with it? LONGi's record page gives none. A record without stability data is a physics result, not a product.

Why lab stability and field stability differ

Accelerated tests stress a device with heat, light or humidity to predict years of use in weeks. PACT exists because nobody yet has protocols that reliably translate those tests into field lifetime for perovskites. Sandia's 2021 release frames the goal as predicting decades of reliability "without having to wait 30 years". In other words, that prediction is still being built.

Bankability

Bankability means lenders and insurers will finance a project using the technology, which needs independent evidence of performance and lifetime. PACT's plan at launch was to deliver bankability roadmaps for at least two companies within two years and full assessments, each taking about six months, by year four. If a supplier says its perovskite product is bankable, ask who carried out the assessment and when.

Reading the PNNL storage figures

98.7% capacity retained over 1,000 cycles means about 1.3% lost. Spread evenly, that is roughly 0.0013% per cycle (my arithmetic). At one cycle a day, 1,000 cycles is under three years, so this is an early lab indicator, not a lifetime figure. Energy density of 9 Wh/L against vanadium's 25 Wh/L means roughly three times the tank volume for the same stored energy, which matters for site footprint.

Questions for a supplier or project developer

  1. What is the module efficiency, measured by which independent lab, on what area?
  2. What outdoor data do you have, for how long, at which site?
  3. Which qualification and reliability tests has the product passed? Ask for standard numbers, editions and test reports.
  4. What warranty applies, and who backs it if the company fails?
  5. For storage: capacity retention after how many cycles, at what depth of discharge, and what energy density per litre?

What I did not verify. I did not open module qualification standards or the chart's image and data files. I quote the chart's test conditions only as the chart page states them.

What should you do before buying "next-generation" solar or storage?

  1. Separate records from products. Treat cell records as signals of where research is heading.
  2. Ask for module and outdoor data before any perovskite or tandem purchase.
  3. Check the efficiency chart yourself rather than relying on a press release.
  4. For storage plans, compare energy density and cycle data, not just chemistry names.
  5. Review once a year. Durability data is where the next real news will come from.

Key takeaways

  • DOE says perovskite cell efficiency rose from about 3% in 2009 to over 26% on cells of about 0.1 cm².
  • LONGi says its perovskite-silicon tandem cell reached 34.85% in April 2025, certified by NREL.
  • Chart records need independent certification under standard test conditions and say nothing about lifetime.
  • DOE lists stability and durability among the main challenges and says perovskites are not yet made at scale.
  • PNNL's lab-scale iron flow battery kept 98.7% capacity over 1,000 cycles at up to 9 Wh/L.

Sources