Solar frontier / Tandem cells

Why two solar-cell layers can catch more light and still fail the commercial test

A silicon cell wastes some high-energy sunlight as heat and misses some lower-energy light. A perovskite layer tuned for higher-energy photons can sit above silicon, leaving another part of the spectrum for the lower cell.

2 minute readexplainerFrontierdeveloping evidenceReviewed 2026-08-03Next review 2026-11-03
Macro cutaway of a tandem solar cell splitting blue and red light between two layers while a crack and moisture bead show a durability question.
Macro cutaway of a tandem solar cell splitting blue and red light between two layers while a crack and moisture bead show a durability question.

Tandem cells promise more output from a given area because two absorbers handle different parts of sunlight. The top layer must also survive heat, moisture, light and mass production.

Two bandgaps divide the work

Every photovoltaic absorber has an energy threshold. Silicon captures a broad part of sunlight, but photons with much more energy than its bandgap lose the excess as heat. In a tandem, a wider-bandgap perovskite top cell converts more of the blue and higher-energy light. Lower-energy light passes to the silicon bottom cell. The pair can therefore use the spectrum more efficiently than either layer alone.

Mechanism diagram for Why two solar-cell layers can catch more light and still fail the commercial test.
How a perovskite top cell and silicon bottom cell divide sunlight between two absorbers. Source: ARENA, Cost-effective Si/Perovskite Tandem Modules on Passivating Contact Si Cells, checked 9 August 2026.

Integration creates new losses

A monolithic tandem deposits the perovskite device directly above the silicon cell and connects them as one package. The layers must be optically transparent where required, electrically compatible and current matched. Texturing that helps silicon capture light can make the upper film harder to coat uniformly. Interfaces, contacts and processing temperatures can improve one sub-cell while damaging the other.

Australian projects have moved beyond a tiny idea

ARENA-backed projects at ANU and partner institutions have demonstrated high-efficiency tandem cells on small and larger areas, developed textured silicon approaches and worked with industry partners. Current interim reporting says cells have passed important stability tests but further stability improvement is required for a commercial product. Project targets for modules describe the work programme, not guaranteed market performance.

Evidence boundary diagram for Why two solar-cell layers can catch more light and still fail the commercial test.
Commercial boundary: the current ANU project aims to develop stable scalable sub-cells, encapsulation and pilot-line evidence; cell efficiency is not itself a durable commercial module. Source: ARENA, Cost-effective Si/Perovskite Tandem Modules on Passivating Contact Si Cells, checked 9 August 2026.

Durability decides whether efficiency is valuable

Perovskite materials and their contacts can be sensitive to moisture, heat, oxygen, prolonged illumination and electrical stress. Encapsulation can slow degradation, but it adds materials and manufacturing steps. A product must deliver stable power across a large module, survive outdoor cycling, be manufactured consistently and manage material recovery. Extra laboratory efficiency has little value if it disappears too quickly or destroys production yield.

The tandem catches more of the spectrum; commercialisation must catch every failure mode around it.

The frontier is no longer a choice between excitement and scepticism. The optical mechanism is sound and high-efficiency cells exist. The honest question is whether researchers and manufacturers can preserve that advantage across area, time, cost and production. Two layers can catch more light. Only a durable module can turn that into cheaper electricity.

Sources and method

Primary source routes rechecked 2026-08-09; preserve project, demonstration and commercialisation boundaries. Recheck again immediately before authorised publication. Found a problem? See our correction process.