The Green MechanicsEnergy, engineering and sustainable mobility

Artificial photosynthesis: a promising technology for future energy production

In 2013 artificial photosynthesis meant two things: Grätzel's dye-sensitised solar cells and devices that split water into solar fuel. What each does, where the research stood, and the maths.

Line illustration for Green Technology

In October 2013, “artificial photosynthesis” was one of the most talked-about ideas in energy research. The phrase was being used for two quite different technologies. One was the dye-sensitised solar cell, which borrows a trick from leaves to turn light into electricity and had earned its inventor, Swiss-based chemist Michael Grätzel, one of the world’s largest technology awards. The other, closer to what plants really do, was a family of devices that use sunlight to split water or convert carbon dioxide into a storable fuel.

What plants do

Natural photosynthesis captures sunlight with pigments such as chlorophyll, uses that energy to split water into oxygen, protons and electrons, and then uses those electrons to turn carbon dioxide into sugars. The key point is storage: the sun’s energy ends up in chemical bonds that can be kept, transported and released later. The overall efficiency is low, typically a few per cent at best for crops, but it runs on cheap, abundant materials and repairs itself.

“Artificial photosynthesis” in the strict sense means doing the same thing in a man-made device: sunlight in, chemical fuel out. Such fuels, from hydrogen to methanol or other carbon-based fuels, are known as solar fuels.

Grätzel’s dye-sensitised solar cell

In 2010 Michael Grätzel of the École Polytechnique Fédérale de Lausanne (EPFL) received the Millennium Technology Prize, awarded in Finland, with a grand prize of €800,000, for the dye-sensitised solar cell (DSSC). He and Brian O’Regan had described the concept in Nature in 1991.

A DSSC works differently from a conventional silicon cell:

  1. A porous film of titanium dioxide nanoparticles, with an enormous internal surface, is coated with a single layer of light-absorbing dye, playing the role of chlorophyll.
  2. When the dye absorbs a photon, it injects an excited electron into the titanium dioxide, which carries it to the electrode and the external circuit.
  3. A redox electrolyte returns an electron to the dye, completing the circuit.

The separation of light absorption (dye) from charge transport (titanium dioxide) is what earned the comparison with leaves, where pigments and electron-transport chains are also separate. The attractions were cheap materials, simple manufacturing, good performance in diffuse light and the possibility of coloured or semi-transparent panels for buildings. The best laboratory DSSC efficiency reported in 2011 was 12.3%, below crystalline silicon but improving.

Strictly, though, a DSSC is bio-inspired photovoltaics: it makes electricity, not fuel.

Solar fuels: the research front in 2013

To make a fuel from water, a device must drive two reactions: oxidising water to oxygen at one electrode and reducing protons to hydrogen at the other. Thermodynamics sets a minimum of 1.23 V between them; in practice catalysts and losses push the requirement higher. Three results show where the field stood:

  • The “artificial leaf” (2011). A team led by Daniel Nocera reported a silicon solar cell coated with a cobalt-based oxygen-evolving catalyst and a nickel-alloy hydrogen catalyst that split water when dropped into a jar in sunlight, with no wires. Its solar-to-fuel efficiency was 2.5% in wireless form and 4.7% wired.
  • Rust as a photoanode (October 2013). Boston College chemists reported a hematite (iron oxide) photoanode, coated with a nickel-iron oxide catalyst, producing 0.6 V, about 50% more than the previous best for this material. Paired with their earlier photocathode, the system delivered more than 1 V, leaving a gap of about 0.2 V to unassisted water splitting using very cheap materials.
  • Perovskites (July 2013). In a separate development from Grätzel’s own group, perovskite-sensitised solar cells made by a new sequential deposition method reached about 15% efficiency. Their high voltage also made them interesting as a way to drive water-splitting cells.

How efficiency is measured

The standard figure of merit is solar-to-hydrogen (STH) efficiency:

STH = (j × 1.23 V) / P_sun

where j is the current density and P_sun is the incident sunlight, conventionally 100 mW/cm². To reach 10% STH:

j = 0.10 × 100 mW/cm² / 1.23 V ≈ 8.1 mA/cm²

What would that mean on the ground? Take a round figure of 5 kWh/m² of sunlight per day for a sunny tropical location (see our article on annual solar radiation in different locations). At 10% STH, one square metre would store about 0.5 kWh a day as hydrogen. With hydrogen’s higher heating value of about 39.4 kWh/kg, that is roughly 13 g of hydrogen per square metre per day, before any losses in collection, storage or compression.

Why it was described as promising

The promise of artificial photosynthesis lies in storage. Solar panels make electricity only when the sun shines; batteries in 2013 were expensive for long-duration storage. A device that turns sunlight directly into a fuel would sidestep that problem and could also supply hydrogen or carbon-based chemicals for industry. The challenges in 2013 were stability (many light-absorbing materials corrode in water), cost (the best catalysts used platinum, iridium or ruthenium), and efficiency at scale. Commercial systems were generally regarded as many years away.

For more context on how conventional solar cells were evolving at the same time, see our archive piece on PV module manufacturing consolidation. For a biological route from waste biomass to fuel, see Teck Guan’s cellulosic ethanol pilot.

What has changed since

  • Perovskites took off. Perovskite solar cells became one of the fastest-improving photovoltaic technologies, and perovskite-silicon tandem cells later surpassed the efficiency of the best single-junction silicon cells in the laboratory. Durability remained the main obstacle to large-scale use.
  • DSSCs found a niche. Dye-sensitised cells did not displace silicon for power generation, but found uses in indoor and low-light products where their performance under weak, diffuse light is an advantage.
  • Solar fuels remain mostly in the laboratory. Integrated water-splitting devices have improved in efficiency and lifetime, but in practice most “green hydrogen” today is made by conventional electrolysers powered by solar or wind electricity, a two-step route that has proved easier to scale than a single artificial-leaf device.

Sources and further reading

  1. Prof. Michael Grätzel awarded the Millennium Technology Prize 2010, EPFL
  2. Porphyrin-sensitized solar cells with cobalt (II/III)-based redox electrolyte exceed 12 percent efficiency (Science 334, 629), Yella et al., Science (2011)
  3. Sequential deposition as a route to high-performance perovskite-sensitized solar cells (Nature 499, 316), Burschka et al., Nature (2013)
  4. Wireless solar water splitting using silicon-based semiconductors and earth-abundant catalysts (the 'artificial leaf'), MIT News (2011), on Reece et al., Science 334, 645
  5. BC chemists are powering artificial photosynthesis, Boston College Chronicle (17 October 2013)