Researchers at the Fraunhofer Institute for Solar Energy Systems ISE in Freiburg have demonstrated a peak solar-to-hydrogen efficiency of 31.3 percent with HyCon, a directly coupled concentrator photovoltaic and PEM electrolysis module. The peer-reviewed study reports the result from outdoor testing over 13 summer days and describes it, to the authors’ knowledge, as the highest efficiency yet reported for a solar hydrogen system measured outdoors.
The architecture needs to be described precisely. HyCon still converts sunlight into electricity before that electricity drives electrolysis, but the two subsystems are electrically matched and connected without an inverter or DC-to-DC power converter. The prototype is also smaller than a briefcase: its Fresnel lens aperture covers 64 square centimetres, while its copper baseplate measures 15 by 15 centimetres, according to technical reporting on the system.
What 31.3 percent actually means
Solar-to-hydrogen efficiency, or STH, measures how much incoming solar energy ends up stored as chemical energy in hydrogen. Fraunhofer calculated the 31.3 percent result using hydrogen’s higher heating value, which is the basis used throughout the research paper.
The figure was a peak, not a full-day average. Across more than 13,000 measurements, efficiency moved between 25 and 31.3 percent as irradiance, atmospheric conditions and water temperature changed. An 11-hour cycle without assisted heating produced a daily STH efficiency of 28.8 percent, while a test with assisted heating reached 29.5 percent.
The result should not be presented as more than twice the performance of every separate photovoltaic-electrolyser system. The researchers place published PV-electrolysis systems broadly between 20 and 30 percent and say HyCon performed about 5 percent better than the strongest comparison in that literature. At the record operating point, the concentrator photovoltaic array reached 34.7 percent efficiency and the two-cell PEM stack reached 91.1 percent on the paper’s thermoneutral-voltage basis.
The cell that came from spacecraft technology
HyCon uses four parallel-connected, four-junction III-V solar cells, each with an active area of seven square millimetres. The cells are produced by wafer-bonding two dual-junction structures, combining gallium indium phosphide, gallium arsenide, gallium indium arsenide phosphide and gallium indium arsenide.
Each junction absorbs a different part of the solar spectrum, allowing the device to convert more light than a conventional single-junction silicon cell. Fraunhofer describes III-V cells as the world’s most efficient solar-cell class and notes that they are currently used primarily in space, where high performance and long-term stability justify their cost. The institute’s official announcement identifies Juan Francisco Martínez Sánchez as project manager, Tom Smolinka as head of membrane electrolysis and Frank Dimroth as head of III-V photovoltaics and concentrator technology.
How direct coupling works
A two-by-two Fresnel lens array concentrates direct sunlight onto the four photovoltaic cells. Those cells are connected in parallel so that their combined current can drive two PEM electrolysis cells connected in series.
The crucial step is matching the voltage-current characteristics of the photovoltaic array and the electrolyser stack. Their operating point falls close to the photovoltaic cells’ maximum power point, allowing the system to use most of the available current without an electronic maximum-power-point tracker or separate power-conditioning stage.
There are still electrical connections inside the module. Current passes through the copper mounting structure and titanium components between the photovoltaic cells and the electrolyser. What HyCon removes is the intermediate conversion hardware that a conventional installation would use to condition and control electricity flowing from a separate solar array.
How the water gets split
The electrolysis section contains two PEM cells connected in series. Each uses a 175-micrometre perfluorosulfonic acid membrane with an active area of 1.13 square centimetres, with iridium on the anode side and platinum on the cathode side.
Machined chlorinated PVC plates guide deionised water into the reaction chamber. Titanium mesh distributes the water, carries current and helps remove the resulting hydrogen and oxygen from the membrane assembly.
The researchers reported no measurable performance degradation during 107 hours of operation and 13 dynamic outdoor cycles. That is encouraging for a proof of concept, but it is not a long-duration reliability demonstration. Commercial PEM systems are expected to operate for tens of thousands of hours, so substantially longer testing will be needed.
Why thermal coupling matters
PEM electrolysis becomes more efficient as its operating temperature rises within the appropriate range. In Fraunhofer’s measurements, lowering the water temperature by 40 K reduced solar-to-hydrogen efficiency by as much as 2.8 percentage points because the electrolyser required a higher voltage.
HyCon was designed to transfer heat from the concentrator photovoltaic cells into the electrolysis water. The first prototype achieved only a 2 K temperature rise through passive coupling, however, so the researchers actively preheated the inlet water during the highest-efficiency tests. A future design would place the water flow closer to the hot solar-cell substrate so that waste heat can raise the electrolyser temperature without a separate heater.
Where the result sits against the rest of the field
Previous Fraunhofer work reached 19.8 percent solar-to-hydrogen efficiency with an outdoor concentrator module. Other researchers achieved about 30 percent under controlled indoor illumination and 24.4 percent outdoors with a larger concentrated solar reactor.
The HyCon paper says its 31.3 percent result is, to the authors’ knowledge, the highest reported outdoor value once real optical, temperature and atmospheric effects are included. Specialist coverage of HyCon has therefore focused on the way the photovoltaic and PEM components are matched, rather than treating it as a new chemical method of splitting water.
Why green hydrogen remains difficult
The wider green hydrogen sector continues to struggle to move from announcements to operating projects. Coverage of the implementation gap highlighted a 2025 Nature Energy study that tracked 190 projects and found that only 7 percent of the global production capacity scheduled for 2023 was completed on time.
Conversion efficiency is one part of that problem, but it is not the only one. Project economics also depend on equipment cost, operating hours, renewable-resource quality, hydrogen buyers, storage, transport infrastructure and financing. HyCon addresses conversion losses and some balance-of-system hardware, but it does not remove those wider commercial obstacles.
The other paths being tried
Concentrator photovoltaics are only one route to solar-powered hydrogen. A recent UK floating-solar study modelled conventional silicon arrays on two reservoirs and directed surplus electricity into PEM electrolysis. The study also found that covering part of the reservoir surface could reduce evaporation, although it did not demonstrate a directly integrated reservoir-water feed or cooling system.
Perovskite photovoltaics are another fast-moving area. An IDTechEx overview highlights their low weight, flexibility and potential for lower production costs alongside continuing development in green hydrogen and electrolyser technology. Those qualities could eventually make high-efficiency solar hydrogen systems cheaper, but they do not yet make perovskites a direct substitute for HyCon’s established III-V cells.
The scale question
HyCon is a proof-of-concept device, not an industrial hydrogen plant. Dimroth assessed it at technology readiness level 3, meaning that its basic concept has been demonstrated experimentally but has not yet been validated as a full pilot system.
Scaling it would require larger lens and cell assemblies, accurate dual-axis tracking, reliable optical alignment, water circulation, safe gas collection and much longer operating tests. The cost and availability of III-V semiconductor material, iridium and platinum would also influence whether the efficiency advantage survives at commercial scale.
The research paper estimates that a system operating in a region with strong direct sunlight could approach hydrogen costs of about $3 per kilogram. That calculation assumes concentrator electricity costing $15 to $25 per megawatt-hour and a capacity factor of roughly 31.6 percent. It is a modelled projection, not a cost demonstrated by the prototype.
What comes next in Freiburg
The next technical step would be a larger pilot module with better thermal coupling, allowing heat from the photovoltaic cells to warm the electrolysis water without active inlet heating. Dimroth has said that the team does not yet have funding for that pilot and needs partners to develop the system further.
Fraunhofer is also seeking investors for a planned spin-off called Clearsun Energy, which would work on commercialising concentrator photovoltaics and could eventually develop the solar hydrogen module as a product. For now, the 31.3 percent result is a verified outdoor efficiency milestone, but the decisive test will be whether a larger system can combine similar performance with long operating life, manufacturability and competitive hydrogen costs.
