Solar cell efficiency records: Spectrolab, NREL and the race to 50%

Your residential panel: 23%. The world's best cell: 47.6%. The gap is III-V semiconductors stacked in six layers with 200× light concentration. Almost nobody uses them, but they set the ceiling of what's possible.

The chart that matters: NREL Best Research

The National Renewable Energy Laboratory (USA) maintains since 1976 the famous 'NREL Best Research-Cell Efficiencies' chart. The progress bible: every verified record added. Latest edition graphs 28 different technologies.

Absolute record: 47.6% (NREL/Fraunhofer, 2022)

Six-junction cell (III-V), runs at 143 suns light concentration. Size: 0.025 cm². Not manufacturable in commercial panel, concept only. Light concentrated with Fresnel lens on tiny cell. Application: utility-scale in very sunny zones.

Space record: 39% (Spectrolab Boeing)

Spectrolab, Boeing subsidiary, makes satellite multijunction cells since the 90s. Record 39.5% (June 2024) in triple-junction GaInP/GaAs/Ge. These are the panels on commercial satellites (SpaceX, OneWeb) and ISS.

Pure silicon record: 27.3% (Longi 2024)

Back-contact heterojunction (HBC). Approaching the 29.4% Shockley-Queisser theoretical limit for silicon. Immediate commercial benefit: 24-25% mass-production panels by 2027.

Perovskite-silicon tandem record: 33.9% (Longi 2024)

The big leap. Stacking perovskite above silicon breaks the theoretical limit. Path to 35-40% commercial in 2030. Game changer: 30%+ panels on your roof within 5-7 years.

Thin film (CIGS) record: 23.6% (Solar Frontier)

Copper-Indium-Gallium-Selenide. Promising but stuck on manufacturing difficulty. Good for curved and semi-transparent surfaces (BIPV - Building Integrated Photovoltaics).

The dream: 50% commercial silicon

Impossible with pure silicon (29.4% limit). Possible with triple tandem (perovskite + silicon + something else). Estimate: 2032-2035. That means panels producing double today's in the same sqft. Total revolution.

More on evolution in 50 years of efficiency.