Mimicking nature: Producing hydrogen from sunlight
When Daniel Nocera of the Massachusetts Institute of Technology (now at Harvard University) demonstrated his ‘artificial leaf’ in 2011,1 he made headlines with promises of a breakthrough technology that could mimic photosynthesis, using sunlight to split water into hydrogen and oxygen at room temperature. Just sixteen years earlier, this kind of artificial photosynthesis had been called the ‘Holy Grail of science’.2 However, while the postage stamp-sized ‘artificial leaf’ works in a small beaker of water in the lab, major challenges need to be overcome before it can be scaled up for commercial use.
As an excellent energy carrier, hydrogen has been a prime candidate for renewable power and energy storage for many years. But hydrogen is rarely found in its pure form in nature, as it is lighter than air and quickly rises into the atmosphere. Rather, hydrogen is a constituent of a multitude of naturally occurring molecules, like water, from which it has to be separated.
The principle of mimicking photosynthesis to produce hydrogen using solar energy was first described over 100 years ago. Then a team of Japanese researchers published a description of a functioning prototype in the journal Nature in 1975. In this process, a photovoltaic wafer comprising two electrodes separated by a membrane is immersed in an electrolyte solution - in this case, water. Each electrode is made of a photosensitive semiconductor material, coated in a catalyst that helps to generate oxygen (at the anode) and hydrogen (at the cathode). Once the water has been split, the hydrogen can be recovered.
Using these principles, Nocera’s ‘artificial leaf’ consists of a thin sheet of semiconducting silicon, onto which is bonded a layer of a cobalt-based catalyst, which releases oxygen in the presence of sunlight. The other side of the silicon sheet is coated with a layer of a nickel-molybdenum-zinc alloy, which releases hydrogen from the water molecules. What makes his ‘artificial leaf’ different from other attempts to produce hydrogen using sunlight is that it uses abundant and generally inexpensive materials instead of corrosive solutions or relatively rare and expensive materials, such as platinum. The main challenge for the technology to be really useful is scale and further developing the catalysts.
Large-scale solar production of hydrogen
While there are several potential technologies for producing hydrogen using solar energy, only a few are feasible on a large scale. One currently under development is looked into by the FCH-JU co-funded CoMETHy project that uses molten salts to store heat derived from a number of renewable energy sources, including solar energy from concentrating solar plants. This heat is then used to convert a range of possible fuels to produce hydrogen. This is a ‘greener’ variant of a more mature technology that involves reforming methane or natural gas at high temperatures. The gases react with steam in the presence of a catalyst to produce hydrogen, along with carbon monoxide and some carbon dioxide - so they are not completely ‘carbon free’.
One of the most promising technologies to produce hydrogen on a large scale is the so-called hybrid-sulphur cycle (or Westinghouse cycle). In this process, both thermochemical and electrochemical cycles are used to split water by the reduction and oxydation (‘redox’) of sulphur compounds, often recycled from metallurgical industries. The net result is hydrogen and sulphuric acid, which is also marketable.
Whereas nuclear energy is often used to provide the high temperature heat for the thermochemical cycles, the SOL2HY2 project, which is part of the European Fuel Cells and Hydrogen Joint Undertaking (FCH-JU) public-private partnership, is exploring the use of solar energy for both the electrolysis cycles and, using molten salts, also for the thermochemical cycles. Another innovation being explored in SOL2HY2 is to integrate the hydrogen production process with existing sulphuric acid production plants. According to a SOL2HY2 report, the “…highly efficient co-use of existing plants drops H2 costs by about 50-70 % compared to traditional HyS process designs.”
Another FCH co-funded project, HYDROSOL-3D, which follows on from the successful HYDROSOL and HYDROSOL2 projects, aims to exploit solar energy for the catalytic dissociation of water and the production of hydrogen. One of the main technical obstacles to producing hydrogen by splitting water is the high temperature needed. To get around this, catalysts are used to lower the reaction temperature. The project is preparing to build a 1 MW demonstration plant at a site in Spain for the thermochemical production of hydrogen using a solar monolithic reactor, with on-site storage.
In an earlier phase, the team developed an innovative solar reactor to produce hydrogen by splitting steam at moderate temperatures (800-1200°C) using solar energy. This used special refractory ceramic, thin-wall, honeycomb monoliths, optimised to absorb solar radiation. The monoliths are coated with special, highly-active oxides with redox properties that trap oxygen and split water. HYDROSOL-3D now intends to carry out all the remaining steps to build the 1MW solar demonstration plant using this technology in order to ensure long-term, reliable solar-aided hydrogen production at industrially attractive yields.
The SOPHIA project, also part of FCH-JU, is exploring another high temperature path to the dissociation of water using solar energy, namely high temperature steam electrolysis (so-called HTE or SOE for Solid Oxide Electrolysis). This process involves the joint electrolysis of CO2 and H2O to produce syngas (H2+CO), which is the standard intermediate for the subsequent production of methane or other gaseous or liquid fuels after an additional processing step. The main goal of SOPHIA is to develop and operate a 3 kWe pressurized HTE system, coupled to a concentrated solar energy source for proof of principle. A secondary aim is to prove the concept of co-electrolysis at the stack level.
Mimicking nature: scaling up the artificial leaf
There is now considerable R&D activity to try to overcome the limitations of scale and catalyst materials encountered when mimicking nature to produce hydrogen via artificial photosynthesis. In Europe, under the FCH-JU initiative, the ArtipHyction project is using artificial versions of enzymes involved in photosynthesis in the leaf, in particular hydrogenase and so-called Photosystem II (PSII). These serve an equivalent function to the catalysts used in other processes. The aim is to develop an artificial device to convert solar energy into hydrogen with close to 10% efficiency, by splitting water at ambient temperature.
This version of the ‘artificial leaf’ uses an electrode (anode) exposed to sunlight, carrying a PSII-like chemical mimic deposited on a suitable transparent electron-conductive porous electrode material. A membrane enables the transport of protons through a pulsated thin water gap. On the other side of the membrane is a cathode carrying a mimic of hydrogenase coated onto a porous electron-conducting support. Meanwhile, an external wire conducts electrons between the electrodes. At the cathode, protons and electrons are combined into pure molecular hydrogen. The goal is to assemble and test a proof-of-concept prototype of about 100 W (3 gr H2/h) by the end of the project, for a projected lifetime of over 10,000 hours.
Meanwhile, PHOCS (Photogenerated Hydrogen by Organic Catalytic Systems), a three-year EC FP7 project due to end in November 2015, has successfully developed a hybrid photoactive device that combines organic semiconductors and inorganic materials to convert water into hydrogen using sunlight. Unlike other attempts to produce hydrogen using organic semiconductors, PHOCS places layers of nanometric titanium oxide as a physical barrier between the photovoltaic part of the cell and the catalyst that stimulates the hydrogen generation reaction. This technique overcomes problems of corrosion and stability usually associated with organic semiconductors.
In another FCH-JU initiative, PECDEMO will make hydrogen using the theoretically simple process of photo-electrochemical (PEC) water splitting. In this process, light is absorbed by a semiconductor photo-anode and/or photocathode, and converted into energetic electron-hole pairs. The electrons reduce water to form hydrogen gas at the cathode, while the holes oxidise water to form oxygen at the anode. The main focus for innovation in PECDEMO is to design earth-abundant photo-electrode and catalyst materials that are both highly efficient and chemically stable.
Building on impressive breakthroughs in the EU-funded NanoPEC project, which ran from 2009 to 2011, PECDEMO will use a chemically stable metal oxide-based photo-electrode, combined with an efficient photovoltaic (PV) solar cell in a hybrid device. The next step will be to scale up the electrode areas from a few square centimetres to 50 cm2 within three years, with the ambition to demonstrate an 8% efficient device that is stable for more than 1000 hours.
The last word?
All of these technologies seem elaborate, though, compared to the natural simplicity of photosynthesis. But now a group of researchers at the Max Planck Institutes for Chemical Energy Conversion and Coal Research in Germany has found a way to exploit the capacity of hydrogenase enzymes in microalgae to produce hydrogen naturally, as part of the complex processes of photosynthesis. The problem with using microalgae has always been the vast surface areas required to produce hydrogen in useful volumes. But by modifying key amino acids and an enzyme in the microalgae cells, the researchers have been able to multiply the volume of hydrogen produced by hydrogenases by a factor of five. A multiplication of 10 to 100 will be needed for commercial exploitation and, even so, tanks of microalgae cultures covering huge areas will be required. But it may be as close to nature as the technology will get.
For more information:
1 Science 4 November 2011. Vol. 334 no. 6056 pp. 645-648
2 Allen Bard and Marye Fox Accounts of Chemical Research 1995, 28 (3), pp 141–145