WO2006010926A1

WO2006010926A1 - Photoactive layer, photoactive electrode and process for preparation

Info

Publication number: WO2006010926A1
Application number: PCT/GB2005/002946
Authority: WO
Inventors: Andrew Stevenson
Original assignee: Hydrogen Solar Ltd
Priority date: 2004-07-26
Filing date: 2005-07-26
Publication date: 2006-02-02
Also published as: GB0416616D0; GB2416778A

Abstract

A preferred embodiment of the present invention relates to a photoactive layer comprised of a primary structure of particles of photoactive material, and a secondary structure of particles of a material that is more electrically conductive than said photoactive material, wherein one of said first and second structures forms a relatively low resistance passageway for charge carrier mobility through the layer to thereby improve charge carrier mobility in the layer as a whole.

Description

PHOTOACTIVE LAYER, PHOTOACTIVE ELECTRODE AND PROCESS FOR PREPARATION

A first embodiment of this invention relates to a photoactive layer. Another embodiment of the invention pertains to an electrode including one or more such layers, and another embodiment pertains to a photoelectrochemical cell (operable on illumination to split water or other hydrogen containing liquid, to yield hydrogen) that includes such an electrode.

As is well known in the art, certain semiconductors have a bandgap that is sufficiently small to allow electrons in the valance band of the semiconductor to' be excited to the conduction band (to generate positive charge carriers (i.e. "holes") in the valance band) by incident photons that have an energy which exceeds the bandgap energy of the semiconductor.

These semiconductors are generally referred to as being "photoactive" semiconductors, and on immersion of a photoactive semiconductor in an appropriate electrolyte (typically an aqueous electrolyte) and illumination with light of an appropriate energy (i.e. wavelength), it is possible to electrolyse the electrolyte to yield a gas.

One such semiconductor is Iron (III) Oxide, Fe₂O₃. However charge mobility in iron oxide is poor and additional energy is required for example by means of applying a bias voltage to an electrode including an iron oxide layer to bring the charge carriers to the surface in contact with the electrolyte and thus to enable them to split water into gaseous components. Limited charge carrier mobility is a serious problem that not only limits the efficiency of small laboratory samples (e.g. lcm) but also the scale-up of electrodes to sizes needed for commercial production of hydrogen (e.g. 10- 100cm) in a cell.

International PCT Patent Publication No. WO 01/02624 discloses one such cell that is colloquially known as the "Tandem Cell". The Tandem Cell includes a first cell with a metal oxide photoactive layer (typically an iron oxide or a tungsten trioxide photoactive layer), and a second cell in electrical communication with the first cell that functions as a light driven electric bias. In operation of the Tandem Cell, valance band holes generated on illumination of the photoactive iron oxide layer in the first cell cleave water in the cell electrolyte to yield oxygen (which can be collected) and protons.

Electrons promoted to the conduction band (by virtue of the band-gap excitation of the photoactive layer) are collected, and the light-driven electric bias of the second cell is operable to increase the electrochemical potential of these electrons to a level where, in the second cell, they can reduce the protons produced in the first cell to yield hydrogen for collection. The collected hydrogen can then be used as a valuable fuel source.

The Tandem Cell disclosed in the aforementioned PCT publication provided an overall solar to chemical conversion efficiency of something in the order of 2%, and it would be beneficial for a practical working system of the type disclosed in our co-pending UK Patent Application No. 0411066.4 (the contents of which are incorporated herein by reference) to increase this efficiency yet further. The efficiency of the Tandem cell tends to be limited by the charge carrier mobility in the photo-active layer and this in turn is limited by the effect of defects in structure (for example dislocations, kinks or grain boundary defects) which cause electron-hole recombination, stopping charge carrier transport.

One simple way to improve the efficiency of the photo-oxidation process would be to use a photoactive electrode fabricated from a single- crystal semiconductor. However, this approach is costly and impractical for hydrogen production purposes. As another means to increase the efficiency of the photo-oxidation process, it has been proposed to increase the photo-activity of the photoactive layer by doping the photoactive semiconductor with one of a range of different dopants. For example, we have previously proposed (see European Patent Publication No. 1175938 for example) doping Iron (III) Oxide to form an n-type semiconductor (wherein electrons are the majority carriers and holes the minority charge carriers) or a p-type semiconductor (wherein holes are the majority carriers and electrons the minority charge carriers). N-doping, for example, of the Iron (III) Oxide semiconductor is achieved by introducing a suitable substitutional dopant into the Iron (III) oxide lattice (or in other words by replacing one or more Iron atoms in the Iron (III) Oxide molecule with one or more dopant atoms).

When this doped semiconducting layer is placed in contact with the electrolyte a depletion layer is formed at the semiconductor surface. The electric field present in the space charge layer assists in the separation of electron-hole pairs produced by light excitation of the oxide, and boosts the quantum yields for photocatalytic oxygen evolution on the semiconductor.

Another set of investigations described by V M Aroutiounian et al in a paper entitled "Investigations of the Fei.ggTio.oiOrelectrolyte interface" (Electrochimica Acta, 45, (2000), 1999-2005) measured photocurrents achievable with a polycrystalline Iron (III) Oxide layer doped with 0.5 at.% Ti. Messrs Aroutiounian et al also found that doping could lead to an improvement in achievable photocurrent. It is the case, therefore, that doping does provide a useful improvement to the photocurrent achievable from any given photoactive layer. However, doping of a photoactive material only addresses the issue of the rate of electron/hole pair generation inside the particles of photoactive material, and does not - for example - address the issue of charge carrier mobility between particles of the photoactive layer.

This problem of poor charge carrier mobility was reported by Messrs Arountiounian et al in their paper as a limitation of the achievable photocurrent, and as a solution they have suggested reducing the level of dopant in the photoactive layer to provide a more complete crystalline structure (i.e. something closer to a single crystal structure).

However, if doping levels were reduced substantially - as proposed by Messrs Arountiounian et al - then the issue of contamination (or more exactly the avoidance of contamination) would come to the fore. In the semiconductor industry, for example, contamination is avoided by using very expensive clean rooms, and whilst something similar could be utilised to provide the necessary degree of cleanliness, the expense would be prohibitive as regards producing a layer for a commercial hydrogen production system. It is also the case, that doping does not address the problems associated with charge carrier mobility, the mobility of those charge carriers still being vulnerable to electron/hole recombination - particularly at boundaries between particles of photoactive materials. The principle reason for this is that whilst photoactive materials may be good at generating electron/hole pairs, they are not good conductors and hence generated electron/hole pairs often recombine before the electrons can be carried away from the layer by an applied conductor.

The present invention has been conceived with the aim of improving interparticle charge carrier mobility and charge carrier mobility through the layer as a whole - both of which are factors in the achievable photocurrent that have, to date, not previously been addressed. The teachings of the present invention may be used in place or in addition to previously proposed doping techniques.

To this end, a preferred embodiment of the present invention provides a photoactive layer comprised of a primary structure of particles of photoactive material, and a secondary structure of particles of a material that is more electrically conductive than said photoactive material, wherein one of said first and second structures forms a relatively low resistance passageway for charge carrier mobility through the layer to thereby improve charge carrier mobility in the layer as a whole. In one preferred embodiment, the secondary material comprises a network of conducting particles. The network may, for example, provide a relatively highly conductive charge carrier path to a substrate on which the layer is provided. In another preferred embodiment, the primary structure of photoactive material consists of photoactive crystals having a microstructure that includes a relatively low number of grain boundaries

In a particularly preferred embodiment, the substrate is coated with an at least substantially continuous conductive layer, on which said photoactive layer is provided. By providing a composite structure with an increased concentration of conductive impurities (as compared for example to the arrangement proposed by Messrs Arountiounian et al) it is possible - inter alia - to fill interstices between at least some of the neighbouring semiconductor particles with conductive material that improves charge carrier mobility between these particles, and hence improves the photocurrent through the layer as a whole. This is in stark contrast to the teaching propounded by Messrs Arountiounian et al. that a reduction in doping (i.e. impurities) - to provide a more compact structure - would be most beneficial. A particular advantage of our proposal, in contrast to the reduction of dopant concentration proposed by Messrs Arountiounian et al, is that by adopting the structure disclosed it is possible to enhance charge mobility (i.e. conductivity) over larger areas, and hence fabricate a larger electrode. A particularly preferred embodiment of the present invention provides a layer which directly addresses the problem of charge mobility. In one embodiment, a conducting network is formed, either at the scale of individual crystal particles, by incorporating suitable amounts of secondary material in the recipe of spray pyrolysis precursors or it may be at macroscopic scale by providing a network of conducting tracks between or beneath layers.

A further anticipated benefit of the preferred embodiment is that the particles of secondary material (or a tertiary material in addition to the primary and secondary materials) may be intrinsically sufficiently advanced in the galvanic series to provide a local electrical bias when in contact with electrolyte. By this we mean that the particles of secondary (or tertiary) material, if sufficiently different in the galvanic series from the particles of first material, may be capable of inputting sufficient energy to the layer to enable the layer to directly split water without needing a discrete electrical

bias to be provided.

Other advantages of the present invention and preferred features thereof will become apparent from the following detailed description in which various preferred embodiments of the present invention are described, by way of illustrative example only, with reference to the accompanying drawings, in

which:

Fig. 1 is a schematic representation of a cell that includes a photoactive layer according to an embodiment of the invention; Fig. 2 is a plan view of the cell depicted in Fig. 1; Fig. 3 a is a schematic plan view of a layer in accordance with an embodiment of the present invention;

Fig. 3b is a side elevation of the layer depicted in Fig. 3 a; Figs 4a and 4b are, respectively, scanning electron micrographs for a layer produced in accordance with the teachings provided herein, and a previously proposed layer;

Figs. 4c to 4e are images acquired by means of Secondary Ion Mass Spectrometry (SIMS) investigations of a layer in accordance with a preferred embodiment of the present invention; Fig. 5 a is a scanning transmission electron microscopy (STEM) image

(Z-contrast) of a photoactive layer;

Fig. 5b is an enlargement of part of the image of Fig. 5 a; Figs. 6a and 6b are further STEM images of the photoactive layer, and Figs. 6c and 6d are STEM images for comparison of another photoactive layer;

Fig. 7 is a graph of photocurrent versus Titanium concentration in the precursor mix;

Fig. 8 is a graph of photocurrent versus Tin and Antimony concentration in the precursor mix; Fig. 9 is a graph depicting the effect of substrate temperature on photocurrent;

Fig. 10 is a schematic representation of a photoelectrochemical cell incorporating a photoactive layer, the cell being operable on illumination to electrolyse an electrolyte; and

Fig. 11 is a schematic representation of another photoelectrochemical cell.

As mentioned above, Fig. 1 illustrates schematically a cell 1 that includes a photoactive layer 7 that is in accordance with a preferred embodiment of the present invention.

As depicted, the photoactive cell 1 consists of a substrate 3, such as glass or some other optically transparent material (or at least substantially optically transparent at least to the wavelengths to which the cell is reactive), on which a conducting layer 5 has been deposited (although as will later be described the provision of a conducting layer may not in fact be necessary).

A photoactive layer 7 is deposited on the conducting layer 5, and is typically manufactured by depositing a number of layers (for example, 3 to 7 - preferably 3 to 5 - layers) one on top of the other on the conducting layer 5. The substrate 3, conducting layer 5 and photoactive layer 7 together form a photoactive electrode 8.

As depicted, the electrode 8 is immersed in an electrolyte 9 (such as seawater, or water with a suitable electrolyte added) in use. To measure the efficiency of the cell a counter electrode 11 is provided, and an electrical connection is made - in this instance - between the conducting layer and, in this example, the counter electrode. The electrolyte completes the electrical circuit and photocurrent on illumination of the photoactive layer 7 is measured through this circuit, in units of milliamps per unit area of photoactive layer.

In a practical system - for example for the photolysis of water (the like of which is later described) - the counter electrode 11 would most likely be coupled to the photoactive electrode 8 via a second cell which operates as a voltage bias to increase the electrochemical potential of electrons emitted by the photoactive layer 7 and collected by the conducting layer 5, and those electrons could then be used to reduce protons (i.e. H⁺ ions) generated on the electrolysis of the electrolyte by the photoactive layer 7 into Hydrogen which can be collected, for example for fuel. As mentioned above, on illumination of the cell, incident light travels through the substrate (as indicated) to the photoactive layer 7 of the electrode 8, and electron/hole pairs are generated in particles of the photoactive material (as described above). Some of the electron/hole pairs recombine at boundaries, and some of the electrons travel through the thickness of the photoactive layer to reach the conducting layer, which collects the electron flow. It is this recombination of electrons and holes that adversely affects the photocurrent achievable.

Fig. 2 is a plan view of the cell of Fig. 1 illustrating, in particular, a preferred form of conducting layer 5. As will be appreciated by those persons skilled in the art, for reasons of efficiency it is important that the conducting layer does not occlude too much of the photoactive layer, as any occluded area of photoactive layer would not be illuminated and hence would not generate electron/hole pairs in the manner aforementioned. One way to avoid occluding the photoactive layer is to provide a conducting layer 5 that, whilst being a continuous surface, is transparent - at least to radiation of a range of wavelengths to which the photoactive layer is responsive. For example, the conducting layer could comprise fluorine doped tin oxide, or indeed any of a number of other suitable materials that will be apparent to persons skilled in the art.

Whilst this arrangement is advantageous in that it avoids occluding the photoactive layer, a significant drawback is that whilst fluorine doped tin oxide is electrically conductive, it is not possessed of a particularly high conductance.

Another way to avoid occluding too much of the photoactive layer, which additionally provides for an improvement in the conductance of the conducting layer, is to provide a patterned conducting layer (as depicted) which consists of regions of less dense, more optically transparent and less electrically conducting material 13, and regions of more dense, more optically opaque and more conducting material 15.

As depicted in Fig. 2, the more dense material 15 is preferably arranged in the form of a lattice or grid. For reasons of manufacturing efficacy it is preferred for the region of less dense material to be formed as a continuous surface, and for the grid of more denser material to be formed on the less dense surface - for example by screen printing, ink-jet printing or any other direct write technique. It will be apparent, however, that the less dense material could instead be provided as a plurality of discrete shapes - in this instance "squares" - fitted into the gaps in the grid.

By way of illustration, the less dense material could comprise fluorine doped tin oxide, and the more dense material (formed as a grid or lattice) could comprise a highly conductive material such as silver, gold, aluminium or any other conducting material. If the more dense material is formed on the less dense material by means of a direct write technique, then the highly conductive material could comprise particles suspended or otherwise incorporated in a suitable printable fluid.

As an alternative to providing a discrete conducting layer, it is eminently possible that the conducting secondary material may itself form the conductor that is operable to collect the electrons emitted by the photoactive material.

As is mentioned above, a key problem with photoactive layers is that generated electron/hole pairs can tend to recombine at boundaries to reduce the achievable photocurrent. As will now be explained the layer of the preferred embodiment addresses such problems by providing for good charge carrier mobility.

Although the reasons why the layer described provides for a better photocurrent are not fully understood, investigations have suggested that charge carrier mobility is increased in the layer because the photoactive layer provided by the method to be described is formed with a morphology that is particularly well suited to the passage of charge carriers through the layer to the substrate. One investigation suggests that this occurs because the conductive material provides an improved electrical connection between crystals of photoactive materials, and another investigation suggests that the presence of the conductive material causes the photoactive crystals to form in such as way that they provide a relatively highly conductive pathway for charge carriers through the layer to the substrate. What is clear, however, is that the presence of conductive material in the precursor solution results in the formation of a layer that consists of a primary structure of particles of photoactive material and a second structure of conductive material, wherein one of said first and second structures provides a preferential pathway for charge carriers through the layer. What is also clear is that whilst the physical reasons as to why these structures have been achieved are not fully known, the structures achieved are repeatable and have positive implications for the photocurrent achievable from the layer.

Figs. 3a and 3b are schematic plan and side views of a layer structure comprising two discrete networks of material, a first photoactive material 17 and a second conductive material 19. These figures are provided merely for illustration and are not intended to be limiting in so far as the shape of the particles, or the relative sizes of photoactive and conducting particles, are concerned. It is eminently possible, for example, for the conducting particles to be larger than the photoactive particles, or indeed for the conducting and photoactive particles to be of a similar size.

As is depicted in Figs. 3a and 3b, our investigations have suggested that particles of conductive material - by virtue of the production method used (which will later be described) - tend to intermix with particles of photoactive material 21. This intermixing can comprise one or both of a location of the particles of conductive material in interstices between at least some of the particles of photoactive material, and an at least partial coating of at least some of the particles of photoactive material 21 by the particles of conductive material.

For simplicity Figs. 3a and 3b only show particles of conducting material located in interstices between particles of photoactive material - but this representation should not be construed as a limitation of the scope of the invention. As mentioned above, particles of conducting material may coat, and/or locate in interstices between, particles of photoactive material

In either event, our investigations suggest that the location of particles of conductive material in interstices and the coating of particles of photoactive material by particles of conductive material tend to provide a means to avoid features of the layer that would otherwise present a relatively formidable barrier to good charge carrier mobility. As will later be described, our investigations would also suggest that the secondary material tends to form a conductive network, or in other words to form relatively highly conductive pathways, through the layer. By way of contrast, a doped layer would necessarily include much larger interstices (which hinder charge carrier mobility). Operation of the doped layer would be wholly reliant on charge carriers moving between abutted photoactive particles that are relatively poor conductors (as compared for example to the conductive particles aforementioned), and hence the likelihood of electron/hole recombination occurring would be significantly increased.

As will be appreciated by those persons skilled in the art, the particles of conductive material 19 function to provide a relatively low resistance route for electron flow through the layer. As mentioned above, the particles of conductive material also appear to form a relatively highly conductive network through by means of which electrons can quickly pass to a conductive layer, for example, on which the photoactive layer can be provided. As the electrons can more easily leave any given particle of photoactive material the chances of those electrons recombining with holes at the boundary is reduced, and hence the electron mobility between the particles of said photoactive material and through the layer as a whole is improved thereby providing a gain in achievable photocurrent. hi very general terms, in this suggested layer structure the conductive particles function better to "connect", in an electrical sense, the photoactive particles of the layer both to themselves and to any conductor on which the layer is provided.

Another advantage of this proposed layer structure is that some interstices 21 still exist between particles of conductive material 19 and particles of photoactive material 17, and these persistent interstices allow the photoactive layer to retain its porous or partly porous nature. As the photoactive layer is at least partly porous the electrolyte can permeate through the layer thereby providing intimate contact between the electrolyte and the surface of the photoactive particles.

It is also the case that when the photoactive layer is provided on a conducting layer (as it is for example in the aforementioned preferred photoactive electrode 8), metal ions or metal oxide may diffuse at elevated temperature (as may well occur when the layer is illuminated) from the conducting layer into the grain boundary region of the photoactive layer and this phenomenon may further enhance charge carrier transport and help to reduce electron-hole recombination. A key structural difference between the layer depicted and doped layers of the type previously proposed which is immediately evident from Figs. 3a and 3b but nevertheless worth noting, is that irrespective of whether the conductive particles coat the photoactive particles or locate in interstices between the photoactive particles, the conductive particles interact with and contact only the boundary regions of the photoactive particles and do not penetrate them to form a new mixed or doped crystal type.

In the preferred embodiment of the present invention, the conductive particles are significantly more electrically conducting that the particles of photoactive material, and in a highly preferred embodiment the conductive particles are of one or more conducting metal oxides (such as Titanium Dioxide, Tin Oxide, Antimony Oxide, Zinc Oxide, Zirconium Oxide or equivalent), and the photoactive particles are of a photoactive metal oxide (such as Tungsten Trioxide, Iron (III) Oxide, or equivalent) which (as mentioned above) is in itself conductive, but is very much less conductive than the particles of conducting metal oxide. It is even conceivable for the photoactive particles to comprise a mix of one or more of the aforementioned photoactive metal oxides. In a modification of this preferred arrangement, it is of course conceivable that the conductive particles could also be photoactive. It remains the case, however, that the conductive particles - if also photoactive - should be more conductive that the photoactive particles.

As mentioned above, the conductive particles of the preferred embodiment better serve to electrically connect the photoactive particles, and this feature of the layer can be inferred from Figs. 4a and 4b of the accompanying drawings.

Fig 4b is a scanning electron micrograph for a previously proposed layer, and as immediately evident the particles of photoactive material are relatively rounded in shape, and as a result the layer is likely to have relatively large interstices.

Fig. 4a, in contrast, is a scanning electron micrograph for a layer produced in accordance with the teachings of the present invention, and it is immediately evident that the particles of the layer are much flatter (i.e. more oval in two dimensions) than those of the layer depicted in Fig. 4b.

A first advantage associated with this particle shape is that the surface area of photoactive material that is directly illuminated by light is increased. A second advantage is that the volume of any interstices is correspondingly reduced. As a consequence, the photocurrent achievable is higher, and in the particular case illustrated the photocurrent achieved by the layer of Fig. 4a was found to be roughly three times larger than the photocurrent achieved with the layer of Fig. 4b. Our studies have established that a desirable grain morphology for a good photoactive layer is that the particles are generally oval in two dimensions (as depicted in Fig. 4a) which are roughly 10 to 50 nm thick at their maximum extent, and 200 to 800 nm in diameter.

Considering now the cross-sectional structure of the layer, Figs. 4c to 4e are images acquired by means of Secondary Ion Mass Spectrometry (SIMS) investigations of a layer produced by the method described herein.

In general terms, a SIMS investigation involves bombarding the surface of a sample with a continuous, focused beam of primary ions (such as Ga⁺ or In⁺). The impact of the ions with the surface sputters (ejects) atoms from the surface of the sample, producing secondary ions. These secondary ions are extracted into a mass spectrometer that detects ions of particular mass-to-charge ratios, and by passing the beam over the sample in a raster scan it is possible to generate an image which provides a visual indication of the make-up of the sample, in particular as regards the position, in the sample, of the particular element that the spectrometer is set-up to detect. By repeating the scan for different elements it is possible to construct a series of images that together depict the elemental make-up of the sample. In this instance a sample layer was prepared, in a manner described below, using a conventional spray pyrolysis technique wherein the precursor solution had been modified by including - in this particular case - two conducting metal oxide salts. The images shown were obtained by investigation of a sample produced, by spray pyrolysis, from a precursor solution comprising iron chloride (which will ultimately be present in the layer as iron oxide (the photoactive component)), and a mixture of 20% titanium chloride and 10% aluminium chloride.

The image depicted in Fig. 4c was obtained by setting the spectrometer to detect secondary ions having a mass to charge ratio that matches that of iron. The lighter part of the image depicted indicates instances where iron was detected, and as is clearly apparent iron is pervasive throughout the layer under investigation. This is entirely consistent with what one would expect to see given that the photoactive constituent of the layer is iron oxide. Referring now to Figs. 4d and 4e, respectively, it is clear that both

Aluminium and Titanium are also present in the layer, and that these elements are only present at discrete locations in the layer (as opposed to being all pervasive as iron is). If Aluminium or Titanium were acting to dope iron particles in the layer, then one would naturally expect Aluminium and Titanium to be all pervasive throughout the layer, and the resultant images to be similar in distribution to that generated for Iron but fainter. The fact that the images produced for Aluminium and Titanium clearly show that these elements are not pervasive throughout the layer can only mean that these elements are not acting to dope the iron, but are instead clustering to form a secondary structure which is discrete from that formed by the iron. This is entirely consistent with the structure model described above with reference to Figs. 4a and 4b of the accompanying drawings. As mentioned above, our investigations have also suggested that the enhanced photoactive layer may have a different structure to that described immediately above, and the details of that structure are described below.

Our proposal is that this layer structure forms because the particles of conductive metal salt in the spray pyrolysis precursor act as seed crystals which cause the photoactive crystals to form in such as way that they are elongated and extend through the layer (substantially from top to bottom) to provide a relatively highly conductive pathway for charge carriers through the layer to the substrate which is relatively free of grain boundaries that would otherwise inhibit charge carrier movement. Evidence for this proposal is provided in Figs. 5a, 5b, 6a and 6b. Fig.

5 a depicts a scanning transmission electron microscopy (STEM) image (Z- contrast) of a sample photoactive layer prepared by the spray pyrolysis method described hereafter wherein the precursor solution included an iron salt as the photoactive component and a tin salt as the conducting component. Referring to Fig. 5b, an enlargement of part of the image of Fig. 5a, it is clearly apparent that the presence of a tin salt in the precursor solution has caused the photoactive Fe₂O₃ to form with a "starburst" or columnar grain structure that extends substantially from the top of the photoactive film to the bottom. The columnar structure, being relatively tree of grain defects such as grain boundaries, provides a relatively low resistance route for charge carrier transport through the photoactive layer.

A similar structure is clearly visible in the STEM images of Figs. 6a and 6b, and in this instance the photoactive crystals have a particularly pronounced columnar structure that is relatively free of grain boundaries which would otherwise hamper charge carrier transport through the layer.

Figs. 6c and 6d, in contrast, are STEM images of a photoactive layer prepared by spray pyrolysis from a precursor solution that does not contain a conductive metal salt. As is clearly visible the Fe₂O₃ photoactive layer formed in this instance does not have a columnar microstructure, and as such is much more polycrystalline in nature with a large number of grain boundaries that would together provide a significant barrier to enhanced charge carrier transport through the layer. As before, it is also the case that when the photoactive layer is provided on a conducting layer (as it is for example in the aforementioned preferred photoactive electrode 8), metal ions or metal oxide may diffuse at elevated temperature (as may well occur when the layer is illuminated) from the conducting layer into the grain boundary region of the photoactive layer and this phenomenon may further enhance charge carrier transport and help to reduce electron-hole recombination.

In general terms our investigations have suggested that the presence of a conducting metal salt, in this case a tin salt, in the precursor solution has altered the photoactive film growth kinetics (as compared to methods without such salts) such that the photoactive material assumes a columnar, low grain boundary microstructure which provides an enhanced pathway for charge carrier transport. In contrast, the film produced from a precursor solution that did not include a conducting metal oxide, as depicted in Figs. 6c and 6d, exhibits less attractive growth kinetics and has a microstructure with a relatively large number of grain boundaries that together provide a substantial resistance to charge carrier mobility through the layer.

It is clear from the above, that by means of the method hereafter described it is possible to produce a photoactive layer having improved charge carrier mobility, and hence achievable photocurrent, than has previously been proposed. It is of particular note that the method hereafter described provides a much improved photocurrent than conventional methods which seek to substitutionally dope photoactive crystals by including relatively small quantities of dopant in the precursor solution.

As has previously been mentioned above, the layers disclosed may be produced by a spray pyrolysis method, the details of which are well known to those persons skilled in the art and as such will not be described in detail herein. A few specific examples will now be provided, and our investigations have led us to the conclusion that in general terms, the layer of the present invention can be made from any photoactive metal oxide or oxides and any conducting metal oxide or oxides by combining appropriate metal salts in a precursor solution. In one example an iron oxide coating was produced by spray pyrolysis of 0.05M FeCl₃ in Ethanol, and this coating gave, on illumination, a photocurrent of 0.9 mA/cm2 when measured at a bias voltage of 70OmV. When a secondary grain structure of Titanium dioxide was added by adding TiCl₃ to the precursor mix, the measured photocurrent was found to increase significantly.

Various layers with differing concentrations of Titanium in the precursor mix were produced, and Fig. 7 is a graph of measured photocurrent for the resultant layers versus Titanium concentration in the precursor mix. As is clearly shown, the measured photocurrent increased from 0.9 mA/cm² to

2.5 mA/cm² as the Titanium concentration is increased from 5% to 20%.

Similar experiments were conducted using layers produced from precursor mixes that included Iron Chloride and increasing amounts of Tin salt, and Antimony salt (in this instance Tin Chloride and Antimony Chloride), and Fig. 8 is a graphical representation of the resultant photocurrent versus Tin or Antimony salt concentration in the precursor mix.

As is clearly visible from Fig. 8, in all cases the addition of Tin and Antimony salt to the precursor mix resulted in an increased photocurrent. An increase in photocurrent of nearly 4x was measured by adding 1.25% tin chloride to the iron chloride precursor solution.

Similar results were obtained for the layer produced, by spray pyrolysis, from a precursor solution comprising iron chloride (which will ultimately be present in the layer as iron oxide (the photoactive component)), and a mixture of 20% titanium chloride and 10% aluminium chloride (i.e. the layer depicted in Figs. 4c to 4e of the accompanying drawings).

Of ancillary interest insofar as the Tin / Antimony layer is concerned is the fact that measuring the achievable photocurrent for a batch of annealed Iron/antimony layers and a batch of annealed Iron/tin layers revealed that annealing actually acts to decrease the achievable photocurrent. This observation reinforces the view that the initial formation of the grain structures in the layer is a more important factor for achieving a good photocurrent than certain subsequent treatments such as annealing. One process factor that was found to positively affect the achievable photocurrent for a photo-active layer comprising of 1.25% Sn and 98.75% Fe was the substrate temperature during the spray pyrolysis.

Specifically, as depicted graphically in Fig. 9, increasing the substrate temperature from 350 degrees centigrade to 525 degrees centigrade increased the photocurrent by about a factor of three times. We believe that this increase results from the morphology of the layer structure becoming more interconnected and interlocked at the temperature increases. Other factors such as gas flow rate -which can cool the substrate surface during spray pyrolysis, liquid volume and liquid concentration, have also been found to affect the achievable photocurrent.

It is apparent from the foregoing that it is possible to fabricate a layer that provides an increased photocurrent by combining a photoactive metal salt or salts and any conducting metal salt or salts in the precursor solution, prior to depositing the layer on a substrate by a conventional spray pyrolysis process. The photocurrent achievable may be further enhanced by controlling the temperature of the substrate that the layer is applied to, as well as by varying other process factors. It is also the case that subsequent treatments, such as annealing, tend to adversely affect the layer structure initially obtained and drive down the achievable photocurrent.

As has been mentioned previously, layers of the type disclosed find utility in a photoelectrochemical cell for the conversion of water, for example, to hydrogen and oxygen, and Fig. 10 is a schematic representation of one such cell.

Fig. 10 is a schematic representation of a so-called Tandem Cell (similar in structure to that described in PCT application no. WO 01/02624 - the contents of which are incorporated herein by reference) for the cleavage of water into hydrogen and oxygen by visible light. The device consists of two photo systems electrically connected in parallel. The cell on the left (as depicted) comprises a compartment 30 which contains an aqueous electrolyte that is subjected to water photolysis. In the preferred arrangement the electrolyte comprises water to which an electrolyte has been added for ionic conduction, or seawater. Light enters from the left side of the cell through a glass window 32.

The light then crosses the electrolyte and impinges upon the front face of a photoactive electrode (34, 36, 38) which has been produced in accordance with the teachings provided above. Specifically, the electrode comprises a 46

26 layer fabricated by spray pyrolysis from a precursor which includes a photoactive oxide (such as tungsten chloride) and one or more conducting oxides (such as titanium chloride). The electrode comprises a glass plate 34, a conducting layer 36 provided on the plate 34, and a photoactive layer 38 provided on the conducting layer. The photoactive layer 68 is operable to absorb the blue and green part of the solar spectrum, and to transmit the red and yellow part to a second cell which in this instance is provided behind the back face of the photoactive electrode.

The second cell, in this example, contains a dye-sensitised mesoporous TiO₂ film 40 which functions as a light driven electric bias that is operable to increase the electrochemical potential of the electrons that emerge from the photoactive layer. The TiO₂ film is formed on a transparent conductor 42 which has been formed on the back face of the glass plate 34 of the photoactive electrode. The film is in contact with an organic redox electrolyte 44 that is provided between the film 40 and a transparent counter electrode 46 which is rendered conductive on the side facing the electrolyte by means of an applied conductive layer. Behind the counter electrode 46 there is provided a chamber bounded by a glass plate 48 in which an electrolyte 50 (of the same composition as that provided in the first cell) is provided, the two electrolytes 30, 50 being in fluid communication with one another by means of a glass frit 52 or ion conducting membrane. 46

27

As depicted in Fig. 10, incident light is used to cleave water so that Oxygen is evolved from the compartment 60 in the first cell, and hydrogen is evolved at a cathode 84 immersed in the chamber provided in the second cell. The principal advantage associated with using the principles taught herein to manufacture an electrode for use in the aforementioned cell is that the increase in achievable photocurrent increases the yield of hydrogen and oxygen when the cell is illuminated.

In a further embodiment of the invention, the aforementioned dye sensitised mesoporous TiO₂ film may be replaced with a photovoltaic cell, such as a conventional silicon photovoltaic cell (or some other photovoltaic cell which is chosen for its response to the particular wavelengths of light that are transmitted by the photoactive layer). Such an arrangement provides an increase in biasing voltage, improvements in durability, and is less expensive to manufacture. A simple example of one photovoltaic cell, as is well known in the art, comprises an n-type silicon layer and a p-type silicon layer which have been abutted to form a P-N junction therebetween. Current is extracted from the silicon cell on illumination by means of a contact grid abutted against the n- type layer, and a conductive back plate abutted against the p-type layer. Fig. 11 illustrates a photoelectrochemical cell for water cleavage which employs a photovoltaic cell 56 (such as one of the aforementioned silicon photovoltaic cells) in place of the dye sensitised mesoporous titanium dioxide film used in the cell of Fig. 8. As shown, the cell on the left (as depicted) comprises a compartment which contains an aqueous electrolyte that is subjected to water photolysis. In the preferred arrangement the electrolyte comprises water to which an electrolyte has been added for ionic conduction, or seawater. Light enters from the left side of the cell through a glass window 32. The light then crosses the electrolyte and impinges upon the front face of a photoactive electrode (34, 36, 38) which has been produced in accordance with the process described above (the electrode comprising a glass plate 34, a conducting layer 36 and a photoactive layer 38). The photoactive layer 38 absorbs the blue and green part of the solar spectrum, and transmits the red and yellow part to a photovoltaic cell 56 which in this instance is provided behind the back face of the photoactive electrode.

As before, the second (photovoltaic) cell 56 functions as a light driven electric bias which is operable to increase the electrochemical potential of the electrons that emerge from the photoactive film. Behind the second cell there is provided a chamber bounded by a glass plate 48 in which an electrolyte (of the same composition as that provided in the first cell) is provided, the two electrolytes being in fluid communication with one another by means of a glass frit 52 or ion conducting membrane.

As depicted in Fig. 11, incident light is used to cleave water so that Oxygen is evolved from the compartment in the first cell, and hydrogen is evolved at a cathode 54 immersed in the chamber provided in the second cell.

It can be seen from the above that the teachings of the present invention provide significant improvements over previously proposed arrangements. It will also be understood that whilst preferred embodiments of the invention have been described above in detail, modifications and alterations may be made to these embodiments without departing from the scope of the invention as set out in the claims. For example, in the embodiments described, the charge carriers have been characterised as being electrons. It is conceivable, however, that the charge carriers collected by the conducting layer could be holes rather than electrons.

A further modification, mentioned briefly above, is that it is conceivable that a second cell for providing an electrical bias may not be required if the particles of secondary material (or a third material provided in addition to the primary and secondary materials) are intrinsically sufficiently advanced in the galvanic series to provide a local electrical bias when in contact with electrolyte. By this we mean that the particles of secondary (or tertiary) material, if sufficiently galvanic, may be capable of inputting sufficient energy to the layer to enable the layer to directly split water without needing a discrete electrical bias to be provided.

As a yet further modification, doctor blading may be used as an alternative to spray pyrolysis. This is particularly the case if the layer includes tungsten trioxide.

Lastly, it will be apparent to those persons skilled in the art that a variety of different semiconducting photovoltaic devices have been proposed, and as such the cell descriptions provided herein should not be read as being limited solely to the use of a silicon photovoltaic device as a means for applying an electric bias to the photoactive layer. Any of the many previously proposed photovoltaic devices may utilised in place of the silicon

photovoltaic device mentioned above.

Claims

1. A photoactive layer comprised of a primary structure of particles of photoactive material, and a secondary structure of particles of a material that is more electrically conductive than said photoactive material, wherein one of said first and second structures forms a relatively low resistance pathway for charge carrier mobility through the layer to thereby improve charge carrier mobility in the layer as a whole.

2. A layer according to Claim 1, wherein the secondary structure is discrete from the primary structure and particles of said secondary material are intermixed with particles of said primary material.

3. A layer according to Claim 1 or Claim 2, wherein said particles of said secondary material are located in interstices between at least some of the particles of said photoactive material.

4. A layer according to any of Claims 1 to 3, wherein particles of said secondary material at least partly coat at least some of the particles of said photoactive material.

5. A layer according to any preceding claim, wherein said secondary structure forms a relatively highly conductive network for charge carrier transport.

6. A layer according to Claim 5, wherein said network functions better to connect, in an electrical sense, the photoactive primary structure to a conductive layer on which said photoactive layer is formed.

7. A layer according to Claim 1, wherein said primary structure of photoactive material consists of photoactive crystals having a microstructure that includes a relatively low number of grain boundaries to thereby provide said relatively low resistance passageway for charge carrier mobility.

8. A layer according to Claim 7, wherein said microstructure includes photoactive crystals that extend transversely, in the direction of incident illumination, at least substantially through said layer.

9. A layer according to Claim 7 or 8, wherein said photoactive crystals form a columnar microstructure.

10. A layer according to Claim 9, wherein each column of said columnar microstructure includes relatively few grain defects and extends substantially from the top of said layer to the bottom of said layer.

11. A layer according to Claim 7, wherein said photoactive crystals form a starburst microstructure, elements of said starburst extending substantially through said layer.

12. A layer according to any preceding claim, wherein the material of said photoactive primary structure comprises one or more photoactive metal oxides.

13. A layer according to Claim 12, wherein said one or more photoactive metal oxides are selected from Iron and Tungsten.

14. A layer according to any preceding claim, wherein said particles of secondary material are significantly more conductive (for example at least twice as conductive) as the photoactive particles.

15. A layer according to any preceding claim, wherein the material of said secondary grain structure comprises one or more metal oxides.

16. A layer according to Claim 15, wherein said one or more metal oxides are selected from titanium, aluminium, antimony, tin, zinc or zirconium.

17. A photoactive electrode comprising a substrate, a conducting layer provided on the substrate, and one or more layers according to any of Claims 1 to 16 provided on the conducting layer.

18. A photoactive electrode comprising a substrate and one or more layers according to any of Claims 1 to 6 provided on the substrate, wherein the particles of secondary material together define a conductive network that functions to improve conductivity, and hence charge carrier mobility, through the layer.

19. A photoactive electrode comprising a substrate and one or more layers according to any of claims 7 to 11 provided on the substrate, wherein said primary structure of photoactive material consists of photoactive crystals arranged with a microstructure that includes a relatively low number of grain boundaries to thereby provide said relatively low resistance passageway for charge carrier mobility through the layer.

20. A photoactive electrode according to Claim 17, wherein the conducting layer is substantially continuous, and optically transparent at least to light to which the photoactive structure is responsive.

21. A photoactive electrode according to Claim 20, wherein the conducting layer is of fluorine doped tin oxide.

22. A photoactive electrode according to Claim 20, wherein the conducting layer is provided in a pattern in order to avoid occluding a significant proportion of the photoactive layer.

23. A photoactive electrode according to Claim 22, wherein the pattern comprises regions of less dense, more optically transparent and less electrically conducting material; and regions of more dense, more optically opaque and more conducting material.

24. A photoactive electrode according to Claim 23, wherein the regions of more dense material are arranged in the form of a lattice or grid.

25. A photoactive electrode according to Claim 23 and 24, wherein the region of less dense material comprises a continuous layer on which said lattice or grid of more dense material is provided.

26. A photoactive electrode according to Claim 23 and 24, wherein the less dense material is provided as a plurality of discrete shapes, for example "squares", fitted into gaps in the lattice or grid of more dense material.

27. A photoactive electrode according to Claim 25 or 26, wherein the less dense material comprises fluorine doped tin oxide, and the more dense material (formed as a grid or lattice) comprises a highly conductive material such as silver or gold. 5 002946

36

28. A photoactive electrode according to any of Claims 17 to 27 wherein the substrate is optically transparent, at least to light to which the photoactive material is responsive.

29. A photoelectrochemical system for the cleavage of water into hydrogen and oxygen by light, the system including an electrode according to any of Claims 17 to 28, wherein the photoactive particles of said electrode are operable when in contact with an aqueous solution of an electrolyte in use to absorb a first range of wavelengths of light to evolve oxygen and generate protons, and particles of the secondary material and/or a tertiary material are intrinsically sufficiently advanced in the galvanic series to provide a local electrical bias when in contact with electrolyte in use to drive the reduction of said protons to hydrogen.

30. A photoelectrochemical system for the cleavage of water into hydrogen and oxygen by light, the system including first and second electrically connected cells, the first cell including an electrode according to any of Claims 17 to 28, said electrode being operable when in contact with an aqueous solution of an electrolyte in use to absorb a first range of wavelengths of light to evolve oxygen and generate protons, the second cell comprising a photovoltaic cell which is operable on illumination in use by a second range of wavelengths to drive the reduction of said protons to hydrogen.

31. A system according to Claim 30, wherein said second cell comprises a dye-sensitised mesoporous TiO₂ film in contact with an organic redox electrolyte.

32. A system according to Claim 30, wherein said second cell comprises a silicon photovoltaic cell.

33. A system according to Claim 31 or 32, wherein said second cell is provided behind said first cell.

34. A method of depositing a photoactive layer on a substrate by spray pyrolysis, the method comprising the steps of preparing a precursor solution consisting of a combination of one or more photoactive metal salts and one or more conducting metal salts, and undertaking spray pyrolysis of said precursor solution to form said photoactive layer on said substrate.

35. A method according to Claim 34, wherein said precursor solution comprises 0.5% to 30% conducting metal salt.

36. A method according to Claim 34 or 35, wherein said precursor solution comprises Iron salt (e.g. iron chloride) as said photoactive metal salt, and a mixture of 15 to 30% Titanium salt (e.g. Titanium Chloride) and 5 to 15% Aluminium salt (e.g. Aluminium Chloride) as said one or more conducting metal salts.

37. A method according to Claim 36, where the precursor includes 20% Titanium salt and 10% Aluminium salt.

38. A method according to Claim 34 or 35, wherein said precursor solution comprises Iron salt as said photoactive metal salt, and Tin salt as said one or more conducting metal salts.

39. A method according to Claim 38, wherein said precursor solution comprises 0.5 to 5%, preferably 1.25% Tin salt.

40. A method according to Claim 34 or 35, wherein said precursor solution comprises Iron salt as said photoactive metal salt, and Antimony salt as said one or more conducting metal salts

41. A method according to Claim 34 or 35, wherein said precursor solution comprises Iron salt as said photoactive metal salt, and Titanium salt as said conducting salt.

42. A method according to Claim 41, wherein said precursor comprises from 5% to 20% Titanium salt.

43. A method according to any of Claims 34 to 42, wherein the temperature of said substrate is controlled to be at least 350 degrees centigrade, and preferably between 350 and 525 degrees centigrade.