PHOTOELECTROCHEMICAL SYSTEM

This invention relates to so-called photoelectrochemical systems.

In such systems light, for example solar illumination, is arranged to illuminate one or more photoactive semiconductors (i.e. semiconductors that generate electrons or holes on illumination by light) and thereby generate a voltage of such a magnitude that cleavage of an electrolytic medium (for example an aqueous electrolytic medium) becomes possible. The liberated constituents of that medium, in particular hydrogen, can then be collected for use as a more environmentally friendly fuel.

One previously proposed system, described in European Patent No. 1198621 , consists of two photosystems connected in series. The first of these systems comprises a tungsten trioxide or iron (iii) oxide semiconducting film that absorbs the blue and green parts of the electromagnetic spectrum and transmits the red and yellow parts to a second system arranged behind the first. The second system acts as an electric bias to increase the electrochemical potential of electrons generated by the first to beyond the point where cleavage of the electrolyte to liberate hydrogen becomes possible.

This previously proposed system provides an overall solar light to chemical conversion efficiency of something in the order of 5 to 10 percent, and whilst this represents a significant improvement over previously proposed systems the commercial viability of a large scale system could be improved if the efficiency of the system were to be increased.

Alternative systems with more commercially attractive, i.e. higher, conversion efficiencies in the region of 10 to 12 percent have previously been proposed (for example in the paper by O. Khaselev and J. Turner, Science 280, 1998, pg. 455), but these systems achieve such efficiencies by employing very expensive single crystal semiconductors and hence the cost of these systems currently prohibit their use on a commercial scale.

Clearly, for hydrogen fuel to provide a commercially viable alternative to fossil fuels it is desirable for hydrogen production systems to be developed which are capable of generating hydrogen at a reasonable unit cost. For this to be achievable the hydrogen production systems employed must both be relatively efficient and reasonably inexpensive. The present invention has been conceived with the aim of providing a system which provides comparable efficiency to that provided by expensive single-crystal devices, but at a lower unit cost.

With this aim in mind, one presently preferred embodiment of the present invention provides a photoelectrochemical system for the generation of hydrogen by cleavage of an aqueous electrolyte, the system comprising: an electrolyser for housing an electrolyte, said electrolyser comprising first and second electrodes and being configured for collection of hydrogen liberated at one of said electrodes on operation of the system; a photoactive element separated from said electrolyte and consisting of first and second photoactive

systems electrically connected to one another in series, said first and second photoactive systems being electrically coupled to respective ones of said first and second electrodes; wherein said first photoactive system is configured to absorb and be responsive to illumination of a first range of wavelengths and at least substantially transparent to the remainder of said illumination, and said second photoactive system is configured to be responsive to illumination of a second range of wavelengths comprising at least part of the remainder transmitted by said first system.

This arrangement is advantageous, for example as. compared with that described in the abovementioned European Patent, for a number of reasons. Firstly, by providing a photoactive element that is not immersed in the electrolyte of the electrolyser (as previously proposed), it is then possible to provide an arrangement whereby hydrogen is supplied from the electrolyser at a desired positive pressure. This is advantageous as it allows the electrolyser to be directly coupled to a hydrogen storage device (such as a metal halide canister) without first having to pressurise gas output by the electrolyser.

Secondly, by separating the electrolyser from the photoactive element it is possible to utilise any one of a number of widely available electrolysers, and thus avoid having to construct a specialised arrangement such as that taught in the abovementioned European patent. This means that the device of the present invention can more easily, quickly and cheaply be constructed than the device disclosed in the aforementioned European patent.

It is also the case, that as the incident illumination need not traverse the electrolyte (as is required in the system of the aforementioned European patent) it is possible to reduce the amount of glass, particularly conductive glass, used in the system. This helps reduce both the overall cost of the system, as well as the weight of the system (weight being an important consideration, in particular for roof-borne systems). Furthermore, as incident light does not traverse the electrolyte (in the electrolyser) in the presently proposed configuration it is thereby possible to avoid potential problems associated with contamination of the electrolyte, for example by algal growth, that may otherwise reduce the extent to which the semiconducting layers of the device are illuminated, and hence reduce the overall efficiency of the system.

Thirdly, the use of a discrete electrolyser means that only those semiconductors which absolutely must be in contact with electrolyte need be so. Indeed, in the preferred arrangement, none of the photoactive semiconductors are in contact with the electrolyte in the electrolyser. This helps avoid problems associated with the electrolyte acting to degrade the semiconductor, as might happen with the previously proposed arrangement (for example seawater electrolyte may, over time, act to degrade the photoactive film in the front cell of the system). It is even possible, in a highly preferred embodiment of the present

invention, to avoid using electrolytic solutions in the electrolyser by utilising an electrolyser with a solid electrolytic membrane (such as a so-called PEM electrolyser for example). This arrangement is particularly advantageous as it then allows water, for example pure water, to be used in place of an electrolyte. In a highly preferred embodiment of the invention the photoactive element includes, in said first photoactive system, a dye-sensitised solar cell (DSC) that is traversed by incident illumination. This arrangement, surprisingly, provides a greater overall efficiency than previously proposed arrangements wherein the DSC was employed merely as an electric bias to increase the electrochemical potential of electrons generated on illumination of the front cell.

Another presently preferred embodiment of the present invention provides, in or for a photoelectrochemical system for the generation of hydrogen, a photoactive element substantially as herein described. Investigations have shown that this element is significantly more efficient than the photoactive element described in the aforementioned European patent. In general, the efficiencies achieved are comparable to those achieved with the aforementioned single crystal devices, and as DSCs are relatively inexpensive the cost of the presently proposed system is significantly less than the cost of a corresponding single crystal system.

Other preferred aspects and advantages of each of these embodiments are set out in the accompanying claims, and elsewhere in the following description.

With the above in mind, one particularly preferred embodiment of the present invention will now be described, by way of illustrative example only, with reference to the accompanying drawings, in which:

Fig. 1 is a schematic representation of a previously proposed so-called tandem cell; Fig. 2 is a schematic representation of a photoelectrochemical system in accordance with a preferred embodiment of the invention, that is to say a system comprising a photoactive element and an electrolyser;

Fig. 3 is a schematic representation of one embodiment of the photoactive element depicted in Fig. 2; Fig. 3a is a schematic representation of another embodiment of the photoactive element depicted in Fig. 2;

Fig. 4 is a schematic representation of the electrolyser depicted in Fig. 2;

Fig. 5 is a graph illustrating the light transmittance of an illustrative photoactive system incorporated within the element of Fig. 2; Fig. 6 is a graph illustrating the spectral response of a second illustrative photoactive system incorporated within the element of Fig. 2; and

Fig. 7 is a graph illustrating the overall performance of the photoactive element as a

whole.

Referring now to Fig. 1 , for illustrative background purposes, there is shown a photoelectrochemical system of the type described in the aforementioned European Patent.

As shown, the system 20 consists of two photoactive systems connected in series. The cell on the left (as shown) contains the aqueous electrolyte that is subjected to water photolysis. The electrolyte is composed of water as a solvent to which an electrolyte has been added for ionic conduction. Saline seawater can also be used as a water source, in which case the addition of electrolyte becomes superfluous. Light enters from the left side of the cell through a glass window 1. After traversing the electrolyte 2 the light impinges on a back wall of the cell that comprises a mesoporous semiconductor film 3 composed of an oxide such as tungsten trioxide (WO ₃ ) or iron (iii) oxide (Fe ₂ O ₃ ). This film is deposited onto a transparent conducting oxide film 4 made from a material such as fluorine-doped tin dioxide that serves as current collector and which is deposited on the glass sheet 1. The oxide absorbs the blue and green part of the solar spectrum whilst yellow and red light is transmitted through it.

The transmitted yellow and red part of the solar spectrum is captured by a second cell mounted behind the back wall of the first cell. This second cell contains a dye- sensitized mesoporous TiO ₂ film, and functions as a light-driven electric bias increasing the electrochemical potential of the electrons that emerge from the WO3 film under illumination to render the reduction of water to hydrogen possible.

The second cell further comprises a transparent conducting oxide film 4 deposited on the back side of the glass sheet 1 that constitutes the back wall of the first cell. The conducting oxide film is covered by the dye-derivative nanocrystalline titania film 6, and this film is in contact with the organic redox electrolyte 7 and a counter electrode 8 that comprises glass that is rendered conductive on the side of the organic electrolyte by the deposition of a transparent conductive oxide layer.

Behind the counter electrode there is a second compartment 9 which contains an aqueous electrolyte of the same composition as in the front compartment 2. Hydrogen is evolved at the cathode 10 that is immersed in this second electrolyte compartment. The two electrolyte compartments 2 and 10 have the same composition and are separated by an ion-conducting membrane or a glass frit 11.

The thin film of nanocrystalline tungsten trioxide absorbs the blue part of the solar spectrum.

\NO ₃ + hv → \NO ₃ (e , h ⁺ ) The valence band holes h ⁺ created by band gap excitation of the oxide serve to oxidize water forming oxygen and protons:

4h ⁺ + H ₂ O → O ₂ + 4H ⁺

while the conduction band electrons are collected on the conducting glass support forming the back wall of the first photocell. From there the conduction band electrons are fed into the second photocell that is mounted directly behind the WO3 film to capture the part of the solar spectrum that is transmitted through the top cell. The second photocell acts as a photo-driven bias to increase the electrochemical potential of the electrons so that they can reduce the protons produced during water oxidation to hydrogen.

4H ⁺ + 4e → 2H ₂

The overall reaction corresponds to the splitting of water by visible light.

H ₂ O → H ₂ + 0.5O ₂ Referring now to Fig. 2 of the accompanying drawings, the photoelectrochemical system of a preferred embodiment of the present invention comprises a photoactive element 22 electrically coupled to a discrete electrolyser 24, such as a so-called PEM electrolyser or a liquid electrolyte electrolyser. The electrolyser 24 is depicted schematically as comprising two regions 26, 28 separated by a barrier 30 that may comprise (in the case of a liquid electrolyte electrolyser) a simple gas impermeable membrane or (in the case of a PEM electrolyser) a proton exchange membrane. Separation of the electrolyser from the photoactive element may prove to be advantageous as it allows the electrolyser to be remotely located from the photoactive element, and this could prove advantageous in circumstances where the electrolyte of the electrolyser might react adversely to cold ambient temperatures (as might be experienced if the device were to be installed on a rooftop).

The photoactive element comprises, as aforementioned, a first photoactive system 32 coupled electrically in series to a second photoactive system 34. The first photoactive system 32 is configured to absorb and be responsive to a first wavelength range, and be at least substantially transparent to the remaining wavelength range of the incident illumination. The second photoactive system 34 is configured to absorb and be responsive to at least a second wavelength range falling within the remaining wavelength range transmitted by the first system 32.

In the preferred arrangement the first photoactive system 32 forms the negative electrode of the photoactive element and is coupled to the cathode of the electrolyser 24. The second photoactive system 34 forms the positive electrode of the photoactive element and is coupled to the anode of the electrolyser 24.

In use the photoactive element is arranged such that incident illumination (e.g. solar illumination) falls upon the first photoactive system so that a predetermined first wavelength range of the illumination is absorbed by the first system. The remainder of that incident illumination traverses the first system and falls upon the second system, and the second system absorbs and is responsive to a second range of wavelengths comprising at least

part of the range of wavelengths transmitted by the first system.

In the preferred embodiment, the first photoactive system comprises a dye sensitised solar cell (DSC). Such cells comprises a porous photoactive semiconductor structure, preferably a mesoporous structure (mesoporous structures being those structures which are porous and include pores of in the region of approximately 2 to 50 nm), that has been sensitised by the incorporation of dye particles.

The mesoporous dye-sensitised photoactive semiconductor is immersed in an organic redox electrolyte, preferably a non-volatile organic redox electrolyte or an ionic liquid. In the preferred arrangement, the photoactive semiconductor is of titanium dioxide, TiO ₂ , and consists of roughly 20nm sized anatase particles. Ruthenium dye, preferably Ru(4,4'-dicarboxylic acid-2,2'-bipyridine) (4,4'-bisnonly )- 2,2'-bipyridine)(NCS) ₂ ] coded as Z907 is preferred as a sensitiser, although it will be apparent to persons skilled in the art that alternative sensitisers such as Ru(4,4'-dicarboxylic acid-2,2'-bipyridine) (4,4'-bis(p- hexyloxystyryl)- 2,2'-bipyridine)(NCS) ₂ , coded as K-19 or complexes with a related structure could instead be employed. The mesoporous titanium dioxide structure is immersed, in the preferred embodiment, in a non-volatile organic electrolyte based on the ionic liquid, N- propyl-N-methyl imidazolium iodide or in another embodiment the mesopores of the titanium dioxide may be impregnated with a solid organic hole conductor. The DSC of the first photoactive system is constructed as a self-contained solar cell, that is to say that the photoactive material, immersed in electrolyte, is produced as a sealed unit. Previously DSCs have suffered from degradation of the electrolyte over time, but tests with the DSC of the presently preferred embodiment have indicated that with this proposed DSC design, degradation over time is less of a concern. Convincing evidence of long term stability of DSCs of this type is provided in the following publications: (1) P. Wang, S. M. Zakeeruddin, J. -E. Moser, Md. K. Nazeeruddin, T. Sekiguchi and M. Gratzel "A stable quasi-solid-state dye-sensitized solar cell with an amphiphilic ruthenium sensitizer and polymer gel electrolyte" Nature Materials, Vol. 2, 402-407 (2003); (2) R.F. Service "Solar Cells, Beating the Heat Help Panels See the Light" Science VoI 300, Issue 5623, 1219 , 23 May 2003; and (3) P.Wang, C. Klein, C. R.Humphry-Baker, S.M. Zakeeruddin, S. and M. Gratzel, "New Amphiphilic Dye for Stable >8% Efficient Dye-Sensitized Solar Cell". Appl. Phys. Lett., 86. 123508 (2005).

As is known in the art, DSCs absorb photons in particular spectral regions, and are at least substantially transparent in others. This makes DSCs a particularly suitable choice for multi-component photoactive elements, as it allows the individual components of the element to be "tuned" to particular spectral regions, to thereby utilise an increased proportion of the incident illumination. It is also the case that the short-circuit photocurrent of DSCs can readily be varied by changing the thickness of the film, the effective pore size

or the density, and furthermore that DSCs can readily (and relatively inexpensively) be formed by simple techniques such as screen printing or doctor blading.

The second photoactive system 34 comprises a thin-film photovoltaic semiconductor, or semiconductors. In a particularly preferred embodiment the second photoactive system includes a so-called CIGS cell (Cu(In ₁ Ga)Se ₂ ) responsive to at least part of the range of wavelengths of illumination that the DSC is transparent to. The CIGS cell is electrically coupled in series with the DSC of the first photoactive system by means of an appropriate conductor that is transparent at least to those wavelengths to which the CIGS cell is intended to be responsive. Whilst a CIGS cell is preferred, other suitable types of cell such as a thin film silicon cell or a TiO ₂ film sensitized by a quantum dot or by a near IR wavelength absorbing dye, may instead be used.

On illumination, light of the first range of wavelengths is absorbed by the dye of the DSC thereby exciting the dye molecules and freeing electrons which are injected (in this case) into the mesoporous titanium dioxide that supports the sensitiser. The holes generated on excitation of these electrons are transported by the electrolyte out of the first photoactive system to a conducting transparent interconnect, formed in this embodiment by the transparent conductor 39, that separates the top cell 32 from the bottom cell 34 of the photoactive element. The holes flow via the transparent conductor 39 to the series- connected second photoactive system, and the second photoactive system acts to increase the electrochemical potential of the holes to the point where those holes can oxidize water at the anode of the electrolyser to oxygen and protons. The electrons injected into the semiconductor (which in this case is mesoporous TiO ₂ ) are used to reduce the protons to hydrogen at the cathode of the electrolyser.

Referring now to Fig. 3 of the accompanying drawings, there is depicted a more detailed view of a photoactive element 36 in accordance with a preferred embodiment of the present invention.

As shown the element 36 comprises a first photoactive system 38 (typically this will be the front system, i.e. the system that is first illuminated), and a second photoactive system 40 closely coupled behind the first in a sandwiched arrangement. The second system is coupled to the first in series by means of a transparent interconnect 39 that comprises a transparent conductor such a conducting glass or transparent conducting organic material. Behind the second cell there is provided a reflective layer 41 so that illumination traversing through both systems is reflected back through the first and second systems to thereby increase the proportion of illumination absorbed. The arrangement depicted in Fig. 3a is similar to that of Fig. 3, except that in this embodiment the second system 40 comprises two photovoltaic cells 42, 44 coupled together in series. In this instance the second of these photovoltaic cells is separated from

the first photoactive system 38 by means of a transparent insulator 46 such as a relatively thin glass or plastic sheet so as to maintain the in-series electrical connection between the first and second photoactive systems 38, 40.

As will be apparent to persons skilled in the art, whilst a second photoactive system configuration of two photovoltaic cells is depicted, the scope of the present invention is not limited to this particular arrangement and instead includes any number of photovoltaic cells connected to one another and coupled in series with the first photoactive system.

Although not shown in Figs. 3 and 3a, the photoactive element may be encapsulated in a suitably transparent and impervious material so that the first and second photoactive systems are protected from damage.

Fig. 4 is a schematic representation of a previously proposed PEM electrolyser, the like of which is commonly available, e.g. from H-TEC AG (Germany), and to which the photoactive element of the preferred embodiment can conveniently be connected.

The electrolyser 24 comprises a housing 48 in which respective electrolyte inlets 50 are formed. In the example depicted the electrolyte is shown as being an aqueous electrolyte - specifically water - but it will be appreciated by those persons skilled in the art that substances may be added to water to improve its conductivity (although it will equally be appreciated that in a PEM electrolyser this is not necessary). It will also be appreciated that the electrolyte need not necessarily be aqueous, in which case whilst hydrogen would be liberated at one electrode, another gas would be liberated at the other. For simplicity hereafter it will be assumed that the electrolyte is aqueous (and thus that hydrogen and oxygen will be evolved), but it should be noted and remembered that this need not be the case. To enjoy the fruits of the present invention it is only essential that hydrogen be evolved at one electrode, and some other gas be evolved at the other. In the specific embodiment illustrated, the housing 48 is divided into an oxygen collection region 52 and a hydrogen collection region 54 by means of a so-called proton exchange membrane or PEM 56. Such membranes are well known in the art, and typically comprise complex polymer bodies which are porous to liquid, but gas impermeable (Nafion ^® , manufactured by Dupont, is one example of a suitable PEM material). Each side of the PEM 56 is coated with a catalyst 58 (typically of platinum or ruthenium/platinum), and it is these catalysed regions of the PEM that form the electrodes, specifically an anode 60 and a cathode 62, of the electrolyser. The anode 60 is coupled, as mentioned above, to the first photoactive system of the photoactive element, and the cathode is coupled to the second photoactive system of the photoactive element. One or both of the aforementioned collection regions of the electrolyser can be configured so that generated gasses, in particular hydrogen, are output from the electrolyser at a desired positive pressure. This could be implemented by providing the electrolyser with

walls that can withstand elevated pressures, and equipping the electrolyser with one or more pressure responsive valves that are configured to open (and release electrolytically formed hydrogen and/or other gasses) only once the internal pressure of the electrolyser reaches a predetermined threshold pressure. Such an arrangement can be advantageous, for example if it should be desired to store the hydrogen in a hydrogen storage device such as a metal halide canister as it obviates the need for a separate device for pressurising generated gasses.

In operation of the system, a DC voltage between the anode and cathode and generated on illumination of the photoactive element causes a DC electric current to flow through the PEM. Protons are drawn to the cathode and are discharged as H atoms by combination with electrons (e ^" ) at the metal cathode surface (M). Pairs of adsorbed H atoms then combine to make molecules of H ₂ gas which escape, freeing the electrode surface for more proton discharge:

4H + 4e →4M-H 4M-H →4M + 2H ₂

At the positive electrode or anode, electrons are lost by incoming water molecules creating O ad-atoms, and protons. The electrons are shunted to the cathode, protons enter the membrane, and two O atoms combine to release O ₂ gas:

2H ₂ O → 2M-0 + 4 H + 4 e ^" 2M-0 → 2M + O ₂

Although the overall process or mechanism is complex, its sum or balance is simply equivalent to producing two molecules of hydrogen and one molecule of oxygen from two molecules of water:

2H ₂ O -^-2H ₂ + O ₂ As mentioned above, a particularly preferred DSC comprises a mesoporous ^" IϊO2 film of 5 to 10 micron thickness that consists of roughly 20 nm sized anatase particles, ruthenium dye Z907 as a sensitizer and a nonvolatile electrolyte based on the ionic liquid N- propyl-N-methyl imidazolium iodide. The thickness of the top cell is controlled to provide a short circuit photocurrent (J _sc ) output of roughly 10-11 mA/cm ² and an open circuit voltage (V _oc ) between roughly 800 and 900 mV under AM 1.5 solar radiation. As is evident from

Fig. 5 of the accompanying drawings, the aforementioned DSC has good transparency in the wavelength region above about 700 nm.

For the second photoactive system it is particularly preferred to provide a pair of (for example) CIGS or silicon cells coupled together, and to the first photoactive system, as depicted schematically in Fig. 3a. When illuminated in isolation (i.e. when decoupled from the first photoactive system) each of these cells provide a photocurrent of roughly 35- 40

mA/cm ² and an open circuit photovoltage of roughly 650 mV in full sunshine. A typical spectral response of the photocurrent of a CIGS cell is shown in Figure 6, and it is clearly apparent that these cells have a good response to light of those wavelengths transmitted by the first photoactive system (namely, light of a wavelength above about 700 nm). Silicon cells exhibit a similar action spectrum of their photocurrent.

When the CIGS or silicon cells are used in a tandem configuration (i.e. behind the first photoactive system) as depicted in Fig. 3a, the photocurrent reduces to about 23 mA/cm2 and the V ₀₀ to 630 mV as a consequence of the absorption of a proportion of the incident illumination by the first photoactive system. Despite this reduction, the two in series connected CIGS or silicon cells are capable of supporting the photocurrent of roughly 10-11 mA/cm ² generated by the top cell.

Overall, the total open circuit photovoltage of the photoactive element, as depicted schematically in Fig. 7, is over 2 V. The photocurrent of the photoactive element, at the maximum power point is close to 9-10 mA/cm, and the voltage is roughly 1.6 V, which is enough to sustain electrolysis of an electrolyte in a PEM electrolyser. The overall efficiency of hydrogen generation by sunlight is roughly 14- 15 %, significantly more than previously achievable with the so-called tandem cell depicted schematically in Fig. 1 , and more than that achievable with more expensive single-crystal type photoelectrochemical systems of the like mentioned above. It will be apparent from the foregoing that the arrangement described herein provides a significant improvement in efficiency, without having to resort to using expensive single crystal type semiconductors.

It will also be apparent that whilst certain preferred embodiments of the present invention have been described herein by way of illustrative example, the scope of the present invention is not limited to those embodiments and includes modifications and alterations to the embodiments described. For example whilst the electrolyser is shown in Fig. 2 as being separate from the photoactive element, in an advantageous embodiment the electrolyser may be integrated into the rear reflective wall of the photoactive element. This could be achieved by forming the various layers of the photoactive element directly onto a wall of the electrolyser, and would provide a particularly compact arrangement.