Perovskite–silicon tandem solar cells: manufacturing and commercialisation

Perovskite–silicon tandem manufacturing adds a light-absorbing perovskite device to a silicon solar cell so that the combined structure can use more of the solar spectrum. Turning a high-efficiency cell into a saleable module requires scalable deposition, repeatable interfaces, high production yield, reliable interconnection and durable encapsulation. Commercial shipments have been announced, but a laboratory record, pilot-line output and high-volume qualified production remain different achievements. Buyers should assess the actual module and manufacturing configuration rather than transfer a headline cell result to an entire factory.

Why does the tandem architecture change manufacturing?

A tandem combines absorbers with different responses to light. The top perovskite cell captures higher-energy light while silicon uses more of the lower-energy part. Fraunhofer ISE's manufacturing research explains why this architecture is attractive and why applying the additional layers at scale remains difficult. The industrial challenge is to obtain the combined electrical performance across useful cell area without damaging the underlying device.

The added process must fit the silicon cell's surface, handling requirements and temperature tolerance. A manufacturer needs a controlled sequence for deposition, conversion of the absorber, interfaces and contacts, followed by measurement and sorting. Adding a high-performing top cell is valuable only if the resulting tandem survives subsequent processing. The practical unit of analysis is therefore the whole process route, including upstream silicon quality and downstream module assembly.

Which deposition results are relevant to scale-up?

In June 2025, Fraunhofer ISE and KAUST reported a hybrid route combining evaporation with blade coating, replacing a spin-coating step and achieving 27.8% efficiency in fully textured tandem cells. This is specific research evidence for a more scalable coating approach. It does not report that a commercial production line achieves that efficiency at full throughput or yield. The researchers also identify coating dynamics as an important part of process optimisation.

For a manufacturing decision, request coated width or cell format, deposition time, material utilisation, uniformity, defect frequency and the number of repeated runs. A fast coating step can still create a bottleneck if conversion, cooling or inspection takes longer. Establish how material changes affect the process window and how quickly the line returns to specification after a stoppage. These are proposed diligence questions, not claims that one deposition route has already solved every issue.

The relevant experiment should use production-representative substrates and handling. A small, carefully selected sample can exclude edge effects and rare defects that become consequential over large areas. Track the distribution of electrical output and reject causes across batches. A record cell demonstrates a capability; the saleable distribution shows whether a manufacturer can repeatedly produce what a customer is being offered.

How do cells become modules?

Interconnection, cutting, encapsulation and electrical layout can change the value of the cell. Oxford PV and Fraunhofer ISE's June 2026 module announcement describes tandem cells cut into shingles, connected using conductive adhesive and encapsulated in glass-glass modules with edge sealing. The research modules reached a reported 25.6% efficiency over total module area. That result concerns the stated module design, not all tandem products.

This example makes the system boundary visible. Ask whether quoted efficiency uses active cell area, total cell area, aperture area or total module area, and who measured it. Check rated output, dimensions, electrical characteristics and test conditions for the offered product. A module buyer ultimately installs a physical package with connectors, mounting and inverter requirements; an isolated laboratory-cell percentage leaves those questions unanswered.

Manufacturing teams should also examine repairability and traceability. If an inspection step detects a defective cell or interconnection, can the module be reworked without undermining durability? Which incoming-material and process records are retained for each finished unit? A small loss in one process step may compound across a long sequence, making overall accepted output more informative than the nominal speed of any individual tool.

What does qualification establish about lifetime?

The public description of IEC 61215-1:2021 covers design qualification of terrestrial photovoltaic modules and explicitly warns that test results are not a quantitative lifetime prediction. Buyers should request the applicable qualification reports for the exact product and materials configuration. Passing a named test should not be converted into an unconditional claim about decades of field output in every climate.

For tandem products, the useful evidence package links laboratory stress testing with field observations and a clear failure analysis. Ask how moisture protection, thermal exposure, light exposure and electrical operating conditions were evaluated, and whether the test sequence reflects the offered construction. Record any differences between demonstration modules and the saleable product. A warranty also needs a financially and operationally credible party able to honour it.

Qualification changes when the product changes. A new encapsulant, contact layer or production supplier may alter behaviour even if initial efficiency remains unchanged. Require a documented change-control process explaining when additional testing is necessary. The customer should know which evidence covers the current bill of materials and which evidence belongs to an earlier development version.

What commercial evidence is available?

Oxford PV announced its first commercial tandem-panel shipment in September 2024. Its June 2026 module announcement still describes production at a pilot facility in Brandenburg. These statements can coexist: commercial sale does not require that a technology has already reached commodity-scale output. The Oxford PV company resource examines that business separately from this technology-level manufacturing assessment.

For a new project, establish available module configuration, delivery quantity, lead time, documentation and warranty terms directly. For an investment in manufacturing equipment, ask for repeated production-run data, accepted yield, uptime and the cost of consumables and rejects. The commercially useful question is whether the additional output per installed area justifies the complete delivered system and its risks. A higher cell efficiency alone cannot settle that calculation.

Which milestone should a buyer or manufacturer pursue next?

Choose the next commitment according to the unresolved constraint. A deposition programme may need repeated large-area runs; a module programme may need environmental testing; an installation buyer may need monitored field operation and a support contract. The technology-readiness guide provides a way to state that distinction. Advancing one milestone should trigger reassessment of the remaining manufacturing and operating evidence, not a blanket conclusion that the entire technology category is ready for every use.