Perovskite-silicon tandem solar cells
Also known as Perovskite-on-silicon tandem, Perovskite/silicon tandem solar cells, Monolithic perovskite-silicon tandems
Perovskite-silicon tandem solar cells stack a thin perovskite layer on top of a conventional silicon cell so that each layer captures a different part of sunlight. Certified laboratory cells have passed the efficiency limit of single-junction silicon, a certified 60 cm² device reached 28.9% in 2025, and first commercial modules shipped in 2024, but long-term outdoor durability is still being proven.
In plain language
Sunlight contains a mix of colours, and a single solar cell material can only use part of that mix efficiently. A tandem cell stacks two materials: a top layer that turns blue and green light into electricity, and a bottom silicon layer that uses the red and near-infrared light that passes through. Together they waste less of the sunlight that hits them.
The top layer is made from a perovskite, a family of crystals that can be deposited as very thin films and tuned to absorb particular colours. Laboratory tandem cells now turn more than a third of incoming sunlight into electricity, beyond what silicon alone can achieve in theory. Results are now also reported on larger devices: a 60 cm² laboratory cell was certified at 28.9% in 2025. That is still far smaller than a full-size panel. The open question is whether perovskite layers can last for the 25 years or so that people expect from rooftop and solar farm panels.
Technical explanation
In a monolithic (two-terminal) tandem, a wide-bandgap perovskite top cell, typically around 1.68 eV, is grown directly on a crystalline silicon bottom cell and connected through a recombination or tunnel junction. Because the two sub-cells are in series, their currents must be matched, and the device voltage is roughly the sum of both. The first two-terminal perovskite/silicon cell, reported in 2015, reached a stable 13.7% on 1 cm² using a silicon tunnel junction.
Recent gains have come mainly from better interfaces. Self-assembled monolayer hole contacts, reported in a 2020 Science paper, cut non-radiative recombination and stabilised the perovskite phase, giving a certified 29.15%. By 2024 tandems had passed the 33.7% Shockley-Queisser limit for a single junction, and LONGi reports an NREL-certified 34.85% for a two-terminal cell in April 2025.
Scaling up is the next test. Artuk and colleagues (2025) used phosphonic-acid treatments and piperazinium chloride to passivate defects in the perovskite top cell and reduce recombination at its interfaces. Two similar two-terminal 1 cm² devices were certified by JRC-ESTI at steady-state efficiencies of 31.6% and 31.14%, and a certified open-circuit voltage of 2.0006 V was reported in the latter certification. A 60 cm² active-area device was certified at 28.9% by Fraunhofer ISE CalLab (29.0% on a reverse scan). The authors also report an average of 29.5% across eight large-area devices in two batches, but the main text does not say whether shunted devices are excluded from that average. The paper's 32.9% current-voltage scan and about 32.45% maximum-power-point value are presented separately from its certified results and should not be compared directly with them. Efficiency values should be compared only when area definitions, certification, scan protocol and measurement conditions match.
The move to industry raises different problems: scalable deposition methods, manufacturing throughput, module integration and reliable predictions of field energy yield. Perovskites are sensitive to moisture, heat and light, so stability testing to IEC standards and real outdoor data are now the main evidence gaps. Oxford PV began commercial shipments of 24.5%-efficient tandem modules in September 2024.
What the evidence says
Each statement is typed by the strength of its evidence and linked to the source that supports it.
Demonstrated Shown in a published experiment, test or shipping product.
The first two-terminal perovskite/silicon tandem, joined by a silicon tunnel junction, reached a stable 13.7% power conversion efficiency on a 1 cm² device with an open-circuit voltage up to 1.65 V.
Sources: [1] · AbstractA monolithic perovskite/silicon tandem with a 1.68 eV perovskite and a self-assembled carbazole monolayer hole contact achieved a certified 29.15% efficiency and an open-circuit voltage up to 1.92 V; unencapsulated in air, it kept 95% of its initial efficiency after 300 hours of operation.
Sources: [2] · AbstractPerovskite/silicon tandem cells exceeded the 33.7% Shockley-Queisser limit for single-junction cells in 2024.
Sources: [6] · AbstractTwo similar two-terminal perovskite-on-silicon tandems with a 1 cm² device area were certified by JRC-ESTI at steady-state efficiencies of 31.6% and 31.14%; a certified open-circuit voltage of 2.0006 V was reported in the latter certification.
Sources: [7] · Abstract; ResultsA two-terminal perovskite-on-silicon tandem with a 60 cm² active area was certified by Fraunhofer ISE CalLab at 28.9% steady-state efficiency (29.0% on a reverse scan), using phosphonic-acid and piperazinium chloride passivation.
Sources: [7] · Abstract; ResultsThe paper reports an average efficiency of 29.5% across eight large-area devices in two batches; its main text does not state whether shunted devices are excluded from this average.
Sources: [7] · Results · medium confidenceIn single-junction test devices tracked at maximum power point for 115 hours in nitrogen, without encapsulation at about 35 °C, devices with the paper's passivation treatment (labelled SIP) fell from 21% to 16%, while control devices rose from 17.5% to 20%. The authors hypothesise that the faster loss stems from ionic losses due to extra mobile ions and a weakened perovskite/C60 interface.
Sources: [7] · Results (single-junction stability) · medium confidenceReported, unverified Stated by a company or team but not independently confirmed.
LONGi reports that NREL certified a two-terminal crystalline silicon-perovskite tandem cell at 34.85% efficiency in April 2025, following its own 33.9% (November 2023) and 34.6% (June 2024) results.
Sources: [4] · medium confidenceOxford PV reported the first commercial shipment of perovskite-on-silicon tandem modules in September 2024: 72-cell panels with 24.5% module efficiency, made on its pilot line in Brandenburg an der Havel, Germany, which it claims produce up to 20% more energy than a standard silicon panel.
Sources: [3] · medium confidenceSome companies claim their perovskite modules will keep 80% of initial efficiency after 25 years, and many have passed IEC qualification standards, but reviewers say the next two to three years of field experience will be decisive in testing these claims.
Sources: [6] · Abstract · medium confidenceThe paper reports a top-performing device reaching 32.9% in a current-voltage scan, and about 32.45% under maximum-power-point tracking; these values are not presented as certified results.
Sources: [7] · Results · medium confidenceProposed A design or plan that has not yet been built or tested.
Moving tandems from the laboratory to industry requires scalable materials and manufacturing processes, more focus on long-term stability, faster production and characterisation, successful module integration, and accurate prediction of field performance.
Sources: [5] · AbstractLimitations and open problems
- Long-term outdoor stability is not yet proven at scale; manufacturer claims of 80% efficiency retention after 25 years remain to be tested in the field.
- Perovskite layers are sensitive to heat, light and moisture and need careful encapsulation.
- In a 2025 study, single-junction test devices with a passivation treatment fell from 21% to 16% during 115 hours of maximum-power-point tracking in nitrogen, without encapsulation at about 35 °C, while control devices rose from 17.5% to 20%. The authors hypothesise that the faster loss stems from ionic losses and a weakened perovskite/C60 interface, and report that full recovery is not possible even when those layers are redeposited.
- Record efficiencies are on small laboratory cells; the largest-area certified result cited in this profile is 28.9% on 60 cm², a laboratory device rather than a full-size module. The first commercial modules are much less efficient (24.5% for Oxford PV's first shipped panels).
- The 2025 large-area study reports no outdoor testing. Its perovskite contains lead, and the paper does not discuss leakage, containment or recycling.
- Industrial production needs scalable deposition methods, higher throughput and reliable module integration that are still being developed.
- Current matching between the two sub-cells makes output sensitive to changes in the solar spectrum, so field energy yield must be predicted and checked carefully.
Possible future uses
Ideas, not predictions. None of these has been demonstrated.
- Upgrading existing silicon production lines with a perovskite top-cell step to raise panel output without new silicon capacity.
- High-efficiency panels for space-limited rooftops and vehicles, where more power per square metre is especially valuable.
- Triple-junction stacks that add a second perovskite layer to push laboratory efficiencies well beyond today's two-junction records.
Timeline
- Mar 2015Researchers led from MIT report the first two-terminal perovskite/silicon tandem, reaching a stable 13.7% on 1 cm². Sources: [1]
- 11 Dec 2020A Science paper reports a certified 29.15% monolithic perovskite/silicon tandem using a self-assembled monolayer hole contact. Sources: [2]
- Nov 2023LONGi reports a 33.9% crystalline silicon-perovskite tandem cell. Sources: [4]
- 5 Sept 2024Oxford PV announces the first commercial shipment of perovskite-on-silicon tandem modules, to a US utility-scale customer. Sources: [3]
- 21 Apr 2025LONGi announces an NREL-certified 34.85% two-terminal tandem cell. Sources: [4]
- 30 Sept 2025Artuk and colleagues report a certified 28.9% two-terminal tandem with a 60 cm² active area, measured by Fraunhofer ISE CalLab, in Nature Communications. Sources: [7]
Futures this could help bring about
Related breakthroughs
Glossary
- Perovskite
- A family of crystal materials with a particular structure; the ones used in solar cells are usually lead halide compounds that absorb light strongly in very thin layers.
- Tandem cell
- A solar cell made of two stacked absorbers, each tuned to a different part of the solar spectrum.
- Shockley-Queisser limit
- The theoretical maximum efficiency of a single-junction solar cell, about 33.7% under standard sunlight.
- Bandgap
- The minimum energy a photon needs to free an electron in a material; it sets which colours of light a solar cell can use.
- Certified efficiency
- An efficiency measured by an independent accredited test laboratory rather than by the team that made the cell.
- Passivation
- Treatment that reduces electrically harmful defects at a material surface or interface.
- Active area
- The defined light-harvesting area used for a device's performance measurement.
- Shunt
- An unintended conductive path that allows current to bypass the intended solar-cell junction.
- Maximum power point tracking
- Measurement or control that follows the voltage-current point where a solar device delivers its maximum output.
Sources
- peer reviewed paper
A 2-terminal perovskite/silicon multijunction solar cell enabled by a silicon tunnel junction
Jonathan P. Mailoa, Colin D. Bailie, Eric C. Johlin et al.. Applied Physics Letters (AIP Publishing), 23 Mar 2015. DOI: 10.1063/1.4914179. Accessed 20 Oct 2026. - peer reviewed paper
Monolithic perovskite/silicon tandem solar cell with >29% efficiency by enhanced hole extraction
Amran Al-Ashouri, Eike Köhnen, Bor Li et al.. Science (AAAS), 11 Dec 2020. DOI: 10.1126/science.abd4016. Accessed 20 Oct 2026. - company announcement
20% more powerful tandem solar panels enter commercial use for the first time in the US
Oxford PV, 5 Sept 2024. Accessed 20 Oct 2026. - company announcement
34.85%! LONGi Breaks World Record for Crystalline Silicon-Perovskite Tandem Solar Cell Efficiency Again
LONGi Green Energy Technology, 21 Apr 2025. Accessed 20 Oct 2026. - review article
Pathways toward commercial perovskite/silicon tandem photovoltaics
E. Aydin, T. G. Allen, M. De Bastiani et al.. Science (AAAS), 12 Jan 2024. DOI: 10.1126/science.adh3849. Accessed 20 Oct 2026. - review article
Stability and reliability of perovskite photovoltaics: Are we there yet?
K. T. Tanko, Z. Tian, S. Raga et al.. MRS Bulletin, 18 Mar 2025. DOI: 10.1557/s43577-025-00863-5. Accessed 20 Oct 2026. - peer reviewed paper
60 cm² perovskite-silicon tandem solar cells with an efficiency of 28.9% by homogeneous passivation
Kerem Artuk, Aleksandra Oranskaia, Deniz Turkay et al.. Nature Communications (Springer Nature), 30 Sept 2025. DOI: 10.1038/s41467-025-63673-y. Accessed 21 Oct 2026.
Revision history
6 changes since this profile was created.