Australasian Science: Australia's authority on science since 1938

Turning Water into Fuel

Environmental footprint

A major challenge facing the world is to develop sustainable, non-carbon-based sources of energy. One of the most obvious, renewable and non-carbon-based sources of energy is sunlight.

By Zhiguo Yi and Ray Withers

A simple inorganic semiconductor could deliver an artificial photosynthesis process that will convert sunlight and water directly into hydrogen and oxygen, thus providing the renewable fuel of the future.

Modern society depends on a continuous, reliable supply of energy that must be available both day and night. At the moment, the burning of carbon-based fossil fuels such as coal and oil provide the overwhelming majority of the world’s current, and projected, energy needs.

The problem rushing headlong towards us is that fossil fuels are non-renewable, and the burning of them leads to ever-increasing levels of pollution as well as a systematic increase in the level of greenhouse gases such as carbon dioxide (CO2) in our atmosphere. This, in turn, leads to human-induced climate change.

A major challenge facing the world is to develop sustainable, non-carbon-based sources of energy. One of the most obvious, renewable and non-carbon-based sources of energy is sunlight. Given that sunlight is only available for part of the day, the task is how to capture and store that solar energy in a form that can be used at a later time and on a large-enough scale.

One approach is to efficiently convert solar energy into the chemical energy stored in chemical bonds, such as gaseous hydrogen (H2) and oxygen (O2), via the splitting of water (H2O). This water splitting reaction is accomplished via photosynthesis in nature.

Hydrogen and oxygen are the dream fuels of the environmentally conscious engineer. They can, for example, be combined in a fuel cell to produce electricity while the only by-product is pure clean water. Alternatively, the hydrogen gas obtained could be liquefied and used as a high energy density fuel to provide energy on tap.

The current difficulty is that the only practical (although not energy efficient) way to split water on a large scale is via electrolysis: the decomposition of liquid water (H2O) into gaseous hydrogen (H2) and oxygen (O2) by passing an electric current through the water in an electrolytic cell.

During electrolysis, electrons (e–) at the negatively charged cathode react with protons (H+ ions) in the water to form gaseous hydrogen, while hydroxyl ions (OH–) in the water give up electrons at the positively charged anode to form gaseous O2. This process, however, requires vast amounts of electrical energy. If you generate this electrical energy by burning coal then you are no better off in environmental terms than you were by burning coal directly in the first place.

The development of an alternative, energy-efficient and environmentally sustainable way of using sunlight to decompose water into hydrogen and oxygen gas is therefore an essential precursor to the development of the future so-called hydrogen economy.

As sunlight obviously does not directly transform water into gaseous hydrogen and oxygen, a suitable catalyst that can harvest sunlight is required to drive the reaction:

2H2O + light —> 2H2 + O2

Photocatalysts of this type use electrons and the absence of electrons, or holes (h+), to catalyse this water-splitting reaction. Electrons and holes are normally the domain of electronic engineers, enabling them to create devices from transistors to lasers.

Chemists get excited about free electrons for different reasons. Electron exchange is the basis of chemical reactions. But how can we generate the necessary electrons and holes from solids using only sunlight to drive the above reaction?

Consider the electrons in a single atom. These electrons occupy atomic orbitals around the atomic nucleus, much like planets moving around the Sun. But, unlike planets, an electron’s “orbit” as well as its energy is inherently determined by the laws of quantum mechanics. Some orbits are more attractive to electrons than others, so the ones requiring the least energy fill up first while the ones requiring the most energy fill up last. Each orbital can only take a set number of electrons and, when there isn’t any more space, electrons fill the next best orbital (requiring the next lowest amount of energy).

When several such atoms are brought together to form a molecule, the individual atomic orbitals combine together to form molecular orbitals, the number of which is proportional to the number of atoms. The original atomic orbitals and their associated energy levels thereby split. When a very large number of atoms are brought together to form a crystalline solid, the number of orbitals becomes so large that the allowed energy levels form a series of continuous energy bands that allow the electrons to move from one atom to the next.

In the case of a semiconducting solid, however, there remain intervals of energy for which no orbitals exist. What this means is that there are some energy values that any free electrons moving about the crystal are not allowed to have.

Like the atom, the electrons in the crystal fill up the allowed orbitals from the lowest energy orbital until the total number of electrons is exhausted. The highest energy band that is occupied by the electrons is known as the valence band. The lowest energy unoccupied band just above the valence band is known as the conduction band, and the forbidden energy interval separating them is known as the band gap.

If the photons from sunlight do not have enough energy to excite an electron from the valence band to the conduction band, nothing happens. The photons will not be absorbed and hence will pass right through the solid, making the material transparent to that wavelength of light.

However, once the energy of the photons becomes sufficient to catapult a valence electron right across the energy gap, the photons are strongly absorbed, thereby creating negatively-charged electrons in the conduction band and positively-charged holes left behind in the valence band of the material.

For these electrons and holes to be able to act as efficient oxidising and reducing agents they need to be spatially separated to minimise the rate at which they annihilate each other via recombination. They also need to be induced to migrate to the surface of the semi-conductor particle, where they can either reduce H+ ions adsorbed on the surface to H2 gas or oxidise OH– ions to water and O2 gas (Fig. 1).

One way in which the spatial separation of electrons and holes can be achieved is via a photoelectrochemical cell in which a thin film of the semi-conductor is deposited onto a conducting substrate to form a cathode. An applied potential then removes the photogenerated electrons from the cathode through a conducting wire to the anode. Meanwhile, the photogenerated holes, under the influence of the same potential, migrate to a surface site on the semi-conductor particle where they are available to oxidise OH– ions to form water and O2 gas.

The efficiency by which visible light can split water in such an electrochemical cell is determined by parameters such as:

• the band gap of the semiconductor (which determines its capacity to absorb visible light);

• the capacity to avoid recombination of electrons and holes (which determines the ability to utilise the electrons and holes created); and

• the rate of the photocatalytic splitting of water at the semiconductor’s crystal surface (which will be affected by things such as the size, surface area and morphology of the semiconducting crystalline particles).

The search for appropriate semi-conductor photocatalysts to drive the water-splitting reaction has been underway for the past 40 years, and is now on in earnest. In a recent publication in Nature Materials, we have demonstrated that a simple inorganic semiconductor, silver orthophosphate (Ag3PO4), has a remarkable ability to harvest sunlight to oxidise water to release oxygen. It can also be used to break down organic contaminants such as methlyene blue, Rhodamine B and other chemicals that may be undesirable in water supplies.

One of the things about silver orthophosphate that makes it so suitable for harnessing sunlight is that its band gap (2.36 eV) corresponds almost perfectly with the wavelengths of sunlight, making the process highly efficient. When powdered silver orthophosphate is placed on a conductive electrode underwater, interesting things begin to happen. Sunlight falling on the powder creates free electrons within the lattice and vacant holes where the electrons came from, and this drives a reaction that very efficiently oxidises the water as follows:

4Ag3PO4 + 6H2O + 12H+ +12e– —> 12Ag + 4H3PO4 +3O2

Indeed, the reaction is so vigorous that a plume of oxygen bubbles can be seen pouring from the anode (Fig. 2).

Once you’ve taken the oxygen out, however, the silver orthophosphate has become metallic silver and phosphoric acid, and the reaction stops. To keep the process going it’s necessary to electrolyse the solution to regenerate the silver orthophosphate, a process that also liberates hydrogen gas that can then be stored.

The driving voltage to electrolytically release this hydrogen from the solution, however, is only 0.6 V. In comparison, 1.3 V is required to do the same to water. Thus the energy consumption when using the silver orthophosphate process (0.6 V) is less than half of what would be required to electrolyse water (1.3 V).

As things stand, the new process would greatly reduce the costs of generating hydrogen and oxygen fuel via electrolysis. It may, however, be possible to take things a step further by adapting the chemistry so that it can generate both oxygen and hydrogen gas while at the same time regenerating the silver orthophosphate electrodes – and all with no net input of energy other than sunlight.

What we are working towards is a simple all-chemistry, artificial photosynthesis process that will convert sunlight and water directly into hydrogen and oxygen for use in fuel cells or as sources of hydrogen fuel. There are, of course, many problems to be solved before artificial photosynthesis or the splitting of water to produce hydrogen fuel can become a practical and economically viable alternative to petrol, but this work may represent a significant step on that road.

Zhiguo Yi is a Postdoctoral fellow and Ray Withers is Professor of Materials Chemistry at the Research School of Chemistry, Australian National University. The assistance of Tim Wetherell in the writing of this article is acknowledged.