Science1 publisher3 min readPublished
Perovskite cell built for blue-green light runs at twice its sunlight efficiency in simulated seawater
Wen-Hua Zhang's team at Yunnan University reports a 1.96 eV perovskite cell reaching 34.71% efficiency under light filtered to mimic 10 m of water. The number comes from a 0.0895 cm2 cell under faint filtered light, so it mainly measures how well the cell matches that spectrum. It does not say how much power a subsea sensor could draw.
The Scientist · Science desk
What happened
- Water strongly absorbs light at 630 nm and beyond, so a few metres down only a faint 400 to 600 nm band remains for a solar cell to use.
- Under standard terrestrial sunlight, the same modified cell reached a certified efficiency of 16.79%.
- The group also built encapsulated modules with close to 29 square centimetres of active area.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- constraint The 34.71% is a ratio against a narrow, dim spectrum, so it cannot be ranked against terrestrial cell records, and anyone sizing a sensor power budget still needs output per unit area at depth.
- capability Heat and light degrade wide-bandgap perovskites quickly on land, leaving them too fragile for rooftops. Cool, dim deep water blunts that degradation, which makes them plausible candidates below the surface.
- decision Whether this can power a subsea sensor depends on the roughly 324-fold jump from test cell to module. Module efficiency at depth is the figure to ask for before designing around the result.
A bandgap of 1.96 eV puts the cell's absorption edge near 633 nm [1]. Water absorbs strongly from about 630 nm upward [2]. A standard 1.55 eV perovskite absorbs out to about 800 nm [2], and terrestrial cells rely heavily on red and infrared light between 800 and 1150 nm [1]. A few metres down, those longer wavelengths are gone [2]. According to Physics World, that mismatch has held back underwater solar harvesting, and with it autonomous deep-sea sensors and environmental monitors [4].
The wide bandgap comes from the perovskite's composition, a mixed-anion lead halide [7]. The polymer has a different job. On land, intense light and heat make wide-bandgap perovskites degrade quickly: halide ions migrate, phases separate and defects accumulate [8]. Below the surface, temperatures typically under 25 degrees C and weaker light remove much of that thermal stress [9]. The additive, polyhexamethylene guanidine hydrochloride (PHMG), was brought in for the defects and ion migration that remain [10].
PHMG pairs a water-repelling backbone with guanidinium cations. Those cations lodge in empty gaps in the crystal and hydrogen-bond to lead and halide ions [11]. The measured result is an activation energy for ion migration that rises from 0.07 to 0.21 eV [12], a threefold increase [3]. Ultraviolet photoelectron spectroscopy also showed the film switching from p-type to n-type conduction [13]. The band bending this creates at the surface speeds electron extraction and cuts charge losses to non-radiative recombination at the interface [13].
Physics World calls the result record-breaking [6], and the 34.71% has to be read against the narrow, dim spectrum it was measured under [9]. The same modified cell has a certified efficiency of 16.79% under standard sunlight [14], so the 34.71% underwater figure [6] is about 2.07 times higher [4]. Both numbers divide electrical output by incoming light. In sunlight, the red and infrared that terrestrial cells depend on [1] lies past this cell's roughly 633 nm edge [1]. In the simulated 10 m spectrum, the light sits in the 400 to 600 nm band the cell was designed to absorb [2][7].
I'd put more weight on the stability chemistry than on the percentage. It targets the failure that stops wide-bandgap perovskites lasting on land [8], in a setting that already removes most of the heat [9]. A higher barrier to ion migration helps a cell last longer, but it is not a lifetime measurement.
The record came from a 0.0895 cm2 cell in an underwater solar simulator fitted with custom multilayer interference filters [15]. The team also built encapsulated modules with nearly 29 cm2 of active area [16], about 324 times the test cell's area [5]. Module efficiency, output per square centimetre at depth and lifetime in water would settle the sensor question, and the Physics World text reviewed for this article does not include them.
What to watch
- Published efficiency for the roughly 29 cm2 modules under the 10 m spectrum, and whether it stays anywhere near the small-cell figure.
- Open-water trials that measure output and lifetime at real depths, where spectrum and intensity vary with water clarity.
- Long-duration stability data in water showing whether the higher 0.21 eV migration barrier translates into longer device life.