FIELD OF INVENTION The present invention relates to a composite electrode in which a transparent conductive oxide current collector, photoactive and electroactive elements are all integrated into a single module designed to enhance light harvesting and/or charge extraction efficiency in photovoltaic, photolytic water splitting devices and other types of electro-chemical electrode.

BACKGROUND OF INVENTION AND PROBLEMS TO BE SOLVED

Twenty days of sunlight impinging the earth's surface is equivalent to the energy stored in all known fossil fuel reserves! Almost 60% of America's electricity needs could be met by deploying existing solar cell technology on existing building roofs and facades! These are amazing statistics. Clearly the potential of photovoltaics, the direct conversion of solar energy into electricity is enormous and as a result photovoltaics are becoming increasingly important as an essential component for the production of sustainable energy (Ginley, D., Green, M.A., Collins, R., MRS Bulletin, 2008, 33, 368, Photovoltaics, Nature Photonics, 2008, 2, 277).

An alternative to silicon-based photovoltaic devices is the dye sensitized solar cell DSSC (B. O'Regan, M Gratzel, A low-cost, high-efficiency solcar-cell based on dye-sensitized colloidal TiO2 films. Nature (1991 ) vol. 353 (6346) pp. 737-740, Graetzel, M., Phil. Trans., 2007, 365, 993). Its low cost of production and a relatively high photon to electron conversion efficiency (PECE) of 1 1 % puts it within reach of silicon alternatives (Qing Wang, Seigo Ito, Michael Graetzel, Francisco Fabregat- Santiago, Ivan Mora-Sero, Juan Bisquert, Takeru Bessho, Hachiro Imai. Characteristics of high efficiency dye-sensitized solar cells. J Phys Chem B (2006) vol. 1 10 (50) pp. 25210-25221 , Graetzel. The advent of mesoscopic injection solar cells. Prog Photovoltaics (2006) vol. 14 (5) pp. 429-442). In brief, DSSCs are constructed from a random mesoporous network of anatase TiO2 nanocrystals with a surface anchored ruthenium-based dye to create an optically transparent electrode which is placed in an iodine-triiodide redox electrolyte with a platinum counter electrode. Absorption of visible light by the dye causes injection of photoelectrons into the conduction band of nanocrystalline anatase with electron percolation to the current collecting electrode. The redox cycle is completed at the counter electrode where iodine is reduced to iodide, which is subsequently oxidized to iodine by electron transfer to the tethered dye.

As things currently stand considerable molecules and materials engineering remains to bring the PECE of DSSC to silicon pin-junction solar cell standards. It is believed by most practitioners in the field of DSSC that dramatic improvements in device efficiency to bring it on par with the performance of silicon photovoltaics will only be realized through the design and implementation of cells that incorporate innovative schemes for capturing more photons.

Reports in the open and patent literature to enhance the efficiency of solar cells in general fall into two main categories (i) concentrator solar cells using optics like Fresnel lenses to intensify the light incident on the solar cell to enhance the production of charge carriers (ii) geometric optics based on random light scattering and texturing of active layers and substrates, photonic crystals founded on stop band reflection, surface resonant modes and slow light, orchestrated to increase the effective path length of light and boost charge carrier production.

In this invention we move beyond these well practiced approaches and herein present an innovative strategy that simultaneously amplifies the harvesting of both photons and improves charge collection in a new kind of DSSC. Enhanced light harvesting is actively being explored through novel dyes with improved absorption characteristics or using innovative materials like photonic crystals (A Mihi, M. E Calvo, J. A Anta, H Miguez. Spectral response of opal-based dye-sensitized solar cells. J Phys Chem C (2008) vol. 1 12 (1 ) pp. 13-17). In terms of improving charge collection, it is worth noting that percolation of charge back to the TCO electrodes takes milliseconds (Juan Bisquert, Michael Graetzel, Qing Wang, Francisco Fabregat-Santiag. Three-channel transmission line impedance model for mesoscopic oxide electrodes functionalized with a conductive coating. J Phys Chem B (2006) vol. 1 10 (23) pp. 1 1284-1 1290). This is rendered possible due to well chosen electrolytes which offer low recombination rate with electrons in the TiO2. This limitation has proven long lasting with only modest efficiency improvements over the last 15 years with only a handful of developed electrolytes. Given that the electron transport in DSSCs is controlled by trap-limited hopping through a relatively long and tortuous path to the transparent electrode (L Dloczik, O lleperuma, I Lauermann, LM Peter, EA Ponomarev, G Redmond, NJ Shaw, I Uhlendorf. Dynamic response of dye- sensitized nanocrystalline solar cells: Characterization by intensity-modulated photocurrent spectroscopy. J Phys Chem B (1997) vol. 101 (49) pp. 10281 -10289), Martinson et al. proposed radial ITO collectors covered by a nanometer thick dye sensitized ΤΊΟ2 film as a way to improve charge collection (Alex B. F Martinson, Jeffrey W Elam, Jun Liu, Michael J Pellin, Tobin J Marks, Joseph T Hupp. Radial electron collection in dye-sensitized solar cells. Nano Lett (2008) vol. 8 (9) pp. 2862- 2866). Although promising, Martinson's design provided for very low light harvesting due to reduced surface area.

In terms of water splitting, it remains an ultimate goal of the photoelectrochemist to efficiently and in a sustainable fashion split water using solar illumination as the only energy input. Surmounting this challenge would provide a means to convert energy from our most abundant renewable source, the Sun, into dihydrogen, which would then be employed as an energy vector in a carbon-neutral market( Turner, J. A. Science 2004, 305, 972). Indeed, interest in photoelectrochemical (PEC) water splitting into molecular hydrogen and oxygen intensified in the advent of the petroleum crisis in the 1970s, and was first demonstrated with a semiconductor-liquid junction (SCLJ) using TiO2 as a photoanode in 1972 ( Fujishima, A.; Honda, K. Nature 1972, 238, 37).

Hematite (a-Fe ₂ O ₃ ) is a promising material for water splitting. With a favorable band gap of 2.0 - 2.2 eV, chemical stability in aqueous environments, and matchless abundance, its use as a photocatalyst to produce dihydrogen at a scale corresponding to the world energy demand is realistic. Many research groups have examined this material as a photoelectrode for water-splitting in the past.( Dareedwards, M. P.; Goodenough, J. B.; Hamnett, A.; Trevellick, P. R. J. Chem. Soc. Faraday Trans. 1983, 79, 2027. Quinn, R. K.; Nasby, R. D.; Baughman, R. J. Mater. Res. Bull. 1976, 1 1 , 101 1 . Sanchez, C; Sieber, K. D.; Somorjai, G. A. J. Electroanal. Chem. 1988, 252, 269. Shinar, R.; Kennedy, J. H. Sol. Energy Mater. 1982, 6, 323.)

As a drawback, hematite has a very small hole diffusion length (2 - 4 nm) (Kennedy, J. H.; Frese, K. W. J. Electrochem. Soc. 1978, 125, 709.) as compared to the light penetration depth (a-1 = 1 18 nm at λ = 550 nm) (Balberg, I.; Pinch, H. L. J of Magn. Magn. Mater. 1978, 7, 12.). This causes most photons to be absorbed in the bulk far from the SCLJ, creating photogenerated holes with low probability of participating in water oxidation. Recently, nanostructuring techniques have proven useful in increasing the performance of hematite for water-splitting (van de Krol, R.; Liang, Y. Q.; Schoonman, J. J. Mater. Chem. 2008, 18, 231 1 .), and several groups have reported various approaches (Glasscock, J. A.; Barnes, P. R. F.; Plumb, I. C; Savvides, N. J. Phys. Chem. C 2007, 1 1 1 , 16477. Hu, Y. S.; Kleiman-Shwarsctein, A.; Forman, A. J.; Hazen, D.; Park, J. N.; McFarland, E. W. Chem. Mater. 2008, 20, 3803-3805. Liang, Y. Q.; Enache, C. S.; van de Krol, R. Int. J. Photoenergy 2008, 739864.). Graetzel and coworkers have reported benchmark performance with nanostructured and silicon doped hematite thin films prepared by atmospheric pressure chemical vapor deposition (APCVD) (Kay, A.; Cesar, I.; Gratzel, M. J. Amer. Chem. Soc. 2006, 128, 15714. Murphy, A. B.; Barnes, P. R. F.; Randeniya, L. K.; Plumb, I. C; Grey, I. E.; Home, M. D.; Glasscock, J. A. Int. J. Hydrogen Energy 2006, 31 , 1999.). Even in this nanostructured system the quantum conversion efficiencies are relatively low and are especially poor (< 20%) in the region where hematite has an indirect band-gap transition (λ = 610 - 450 nm). Importantly, there is over 10 mA cm-2 of solar photocurrent available in this wavelength range.

An answer to the persisting problem of small hole diffusion length and large photon penetration depth proposed by Itoh and Bockris is to use ultra-thin films in a stacked formation to increase the proximity of the photogenerated holes to the SCLJ (Itoh, K.; Bockris, J. O. J. Electrochem. Soc. 1984, 131 , 1266.). While this solution could fundamentally resolve the issue, it is cumbersome and expensive to implement, and in practice the thin iron oxide films were found to have poor performance due to the increased recombination of the photogenerated holes (Itoh, K.; Bockris, J. O. J. Appl. Phys. 1984, 56, 874.). An increased photoresponse should also be obtainable by coating these thin films conformally on a suitable nanostructured collector with a large specific surface area, in analogy to the DSSC (Gratzel, M. Chemistry Letters 2005, 34, 8.). In the case of the DSSC, the photoanode is constructed from a light absorbing dye molecule anchored to a high surface area mesoporous ΤΊΟ2 scaffold, which transports and collects the photo- excited electrons. The DSSC's host-scaffold/guest-absorber approach effectively decouples light harvesting and charge transport while maximizing the incident photon-to-current efficiency (IPCE) by means of the thin (mono) layer of absorber on the high surface area, transparent collector.

For water splitting, no single material is likely to meet the many criteria needed to achieve high efficiency; therefore, the approach described in this invention is to focus on advanced nanocomposite architectures. Distinct phases will provide complementary functions, thereby enhancing performance and decreasing the number of criteria that any single phase must meet.

In this class of structures, nanoscopic semiconductor particles (the guest) are supported on a host semiconductor with modest surface area. The guest semiconductor absorbs photons and enables efficient charge separation because of its small size. The host efficiently transports charge to an external circuit. This design is particularly well suited for guest semiconductors that have small excited-state lifetimes: the proximity of the electrolyte and host semiconductor will allow efficient charge collection. Based on the optoelectronic properties already mentioned for one promising material (a-Fe2Os), the guest-host approach has the potential to increase the state-of-the-art photocurrent of 2.2 mA cm ^"2 (at AM 1 .5G 1000 W m ^"2 and 1 .23 V vs. RHE) to well over 10 mA cm-2 assuming a modest 67% average quantum conversion efficiency of absorbed photons. SUMMARY OF THE INVENTION AND DESCRIPTION OF EMBODIMENTS

The present invention provides a generic electrode for photovoltaics, water splitting cells or any electro-chemical device where a three-dimensional transparent conductive backbone provides efficient charge extraction and can support an active guest material for light harvesting or electro-chemical reaction. Enhanced light harvesting is provided through scattering or through photonic crystal effects from the disordered or ordered three-dimensional structure, respectively. Enhanced charge collection is provided by the conductive/transparent conductive tri-dimensional backbone forming a matrix in which the active material can be intercalated. Enhanced light harvesting, charge extraction and thus improved photocurrent and efficiency are possible with this method.

The essence of the invention described herein is predicated upon a composite anode in which current collector, photoactive element, sensitizer and electrolyte are together integrated into a single unit without sacrificing light harvesting capabilities. Dramatic efficiency gains are anticipated using the proposed new design of DSSC that are expected to go well beyond the incremental improvements in performance currently being realized by continued refinement of molecules and materials exploited by DSSC practitioners around the world.

The perceived advantages for DSSCs with this construction over all previous designs is that separate TCO current collector and photoactive anode are combined into a single composite anode in which multiple light scattering from a pore network with predetermined disorder will enhance the effective optical path length of the incident light in the dye sensitized anatase nanocrystals, while at the same time intimate contact between the high surface area TCO, anatase, anchored dye and electrolyte facilitates rapid transport and collection of charge carriers. A faster charge collection will reduce the recombination at the anatase-dye interface and increase the efficiency significantly. An indirect but important consequence of a lower recombination rate is that the choice of liquid and solid electrolytes will be broader and thereby allow further research on hole transport materials. Working synergistically these effects are expected to manifestly amplify photon and charge carrier harvesting and boost the overall PECE of the DSSC compared to all other reported DSSC designs.

In one aspect of the present invention there is provided a disordered or ordered three-dimensional transparent conductive backbone for enhanced charge extraction for photovoltaics or hydrogen production through water splitting or any kind of electrochemical device in which:

a) a porous, disordered or ordered three-dimensional backbone made of a transparent conductive material is formed on a transparent substrate;

b) the pores are partially or completely filled with an active material like semiconductors, semiconductor oxides, quantum dots or others to form the photoanode/photocathode/electrode; and, optionally,

c) the photoanode/photocathode/electrode is assembled into the resulting dye sensitized solar cell, organic-based photovoltaic cell, quantum dots-based cell, water splitting cell, tandem cell, etc;

In one aspect of the present invention there is provided optical effects using an ordered or disordered three-dimensional transparent conductive backbone and/or enhanced light harvesting through photonic crystal and/or light scattering effects for photovoltaics or hydrogen production through water splitting or any kind of photoactive device in which:

a) a porous, ordered or disordered three-dimensional backbone made of a transparent or partially transparent conductive material is formed on a transparent conductive or non conductive substrate;

b) the pores are partially or completely filled with optically active material like semiconductors, semiconductors oxides, quantum dots or others to form the photoanode/photocathode/electrode; and, optionally,

c) the photoanodephotocathode/electrode is assembled into the resulting dye sensitized solar cell, organic-based photovoltaic cell, quantum dots-based cell, water splitting cell, tandem cell, etc;

In an aspect, the present invention provides an electrode comprising: (a) a three-dimensional, porous conductive matrix, which may be provided on a conductive substrate; and (b) a material selected from a photoactive material, an electrochemically active material, a water splitting material, a catalyst, and combinations thereof, said material covering partially or totally the surface of said porous conductive matrix.

In an aspect, the present invention provides a method of preparing an electrode, the method comprising the steps of: providing an assembly of template particles, said particles having a mean diameter of 20 nm to 100 μιτι, in accordance with the pore size of the electrode to be obtained;applying, to said particles, a conductive material; and

removing said template particles.

In further aspects, the invention provides an electrochemical device comprising the electrode of the invention and the use of the electrode in electrochemical devices.

In an aspect, the invention provides an electrode comprising a transparent conductive material (TCM), said transparent conductive material being characterised by one or more of the following features:

- may or may not be provided on a surface/substrate;

- may have two opposed main sides;

- may form a surface, which may be continuous;

- it is three-dimensional, preferably on a micro- and/or mesoscopic scale; - it is porous, in particular nano-, meso- and/or microporous;

- it exhibits an increased surface;

- it is order and/or disordered.

In an aspect, the invention provides a composite electrode in which a transparent conductive current collector, and an active element, selected, for example, from a light harvesting polymer, sensitizers, water splitting material and catalysts are integrated into a single unit, which preferably enhances light harvesting and charge extraction, in particular in photovoltaic cells and photolytic water splitting cells, whereby the composite photoanode comprises: (a) a three-dimensional porous conductive matrix on a conductive substrate; (b) photoactive or electrochemically active materials fully or partially filling the porous matrix and/or fully or partially coating or covering the surface of the matrix.

Further aspects and preferred embodiments are provided in the appended claims and in the detailed discussion herein below.

On a macroscopic scale, and in particular on the scale of 0.5 or 1 cm and more, the electrode of the invention may appear as a surface or layer, in particular as a layer that may have substantially two opposed main sides (or surfaces) that are substantially opposed and/or parallel to each other. The electrode preferably forms a continuous layer or surface. The electrode may be applied, at least during manufacturing, on a substrate layer and have thus the macroscopic form and/or appearance imposed by the substrate. Since the electrode may be flexible, its form may, however, be modified following manufacturing. For example, the electrode may be flat, two-dimensional or it may be parabolic on a macroscopic scale, for example of >0.5 cm. The substrate, if present also following manufacturing, for example in the device comprising the electrode, may be conductive. The substrate may also be transparent. Glass and transparent plastics may be used as substrates, for example.

"Transparent" preferably means that 30%, 40%, 50%, 60%, 70%, 75%, 80%, 85%, 90%, 91 %, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% transparency of incident light. Incident light is preferably light corresponding to AM 1 .5 solar conditions at l OO mWcm ^"2 . Preferably, only visible light is considered in the above percentages.

"Conductive", in the expression transparent conductive material, preferably means conductivities of at least ( >) 1 (Ω cm) ^"1 , preferably >5(Ω cm) ^"1 , more preferably >10(Ω cm) ^"1 , >10 ² (Ω cm) ^"1 , >5χ10 ² (Ω cm) ^"1 and most preferably at least >10 ³ (Ω cm) ^" >1 .1 x10 ³ (Ω cm) ^"1 , >1 .2 x10 ³ (Ω cm) ^"1 , >1 .3 x10 ³ (Ω cm) ^"1 , >1 .5 x10 ³ (Ω cm) ^"1 , >2 x10 ³ (Ω cm) ^"1 , >3x10 ³ (Ω cm) ^"1 or >5χ10 ³ (Ω cm) ^"1 , for example >10 ⁴ (Ω cm) ^"1 , even more preferably >2x10 ⁴ (Ω cm) ^"1 . The unit (Ω cm) represents the resistivity Ohm ^* cm, and (Ω cm) ^"1 is the reciprocal of resistivity, thus conductivity in S/cm (or mho/cm).

The transparent conductive material (TCM) may be any material fulfilling the requirements of transparency and conduciveness and which may be processed to a three-dimensional structure, for example a meso- and/or micro-porous structure. Said three-dimensional, porous conductive matrix preferably comprises and/or consists of said TCM.

For example, the TCM may be provided as a porous material, with pore sizes of 50nm to 7μηη, 70nm to 6μηη, 100nm to 5.5μηη, 150nm to 5μηη, 200nm to 4μηη, 300nm to 3μηη, 350nm to 2μηη, for example. According to an embodiment, the TCM is characterized in that it comprises some pores of the indicated pore sizes. According to an embodiment, these pore sizes are mean pore sizes, corresponding to the mean diameter of the pores.

Pore sizes may be determined by scanning electron microscopy on the spheres or the 3 D backbone, for example.

The TCM may comprise 1 , 2, 3, 4, 5, 6 or possibly more, for example 1 -20, preferably 2-10 layers, most preferably 3-7 of particles and/or cavities. Cavities may be formed by a TCM filling the interstitial volume of template pores or particles, which template particles (in particular the according number of layers of template particles as mentioned above with respect to the cavities) being removed following the application of the TCM. In Figure 1 D, three layers of round cavities are shown, formed following the removal of three layers of template particles made from polystyrene. The number of layers of cavities/pores/particles/template particles may be an average number, since they may not form uniform and/or ordered layers throughout the surface of the electrode.

Preferred materials for TCMs are transparent conductive oxides (TCO). These incude materials comprising doped or non-doped oxides, preferably doped or non- doped zinc oxides, tin oxides and indium oxide materials. For example, the TCM may be antimony doped tin oxide. More generally, Si, TiO ₂ , SnO ₂ , Fe ₂ O ₃ , WO ₃ , ZnO, Nb ₂ O5, CdS, ZnS, PbS, Bi ₂ S3, CdSe, GaP, InP, GaAs, CdTe, CulnS ₂ , and/or CulnSe ₂ , be it doped or non- doped, and combinations thereof, may be used as TCM, for example. The oxide semiconductor materials listed further below with respect to the optically and/or electrochemically active material (O/EAM) may also be used as TCM/TCO, for example in the form of doped oxide materials. It is noted that said O/EAM may be used as TCM in particular if the former is used in a dense, non-nanoporous or non- particulate form. Other materials include materials comprising carbon nanotubes and/or organic transparent conductive materials, for example transparent conductive polymers. Examples of materials further include carbon nanotube (CNT) films, graphene films, processed graphene, and metal nanowire gratings. Also materials comprising metals and/or alloys may be used as transparent conductive materials, for example if the metal and/or alloy is provided in a very thin form so as to be substantially transparent.

According to an embodiment, the TCM may itself not be optically and/or photoactive. For example, the TCM may not be a semiconductor.

The TCM of the electrode is preferably in contact with an optically and/or electrochemically active material (O/EAM), or, more generally, with a "guest" material, which determines the use of the device/electrode. For example, the guest material may be selected from a photoactive material, an electrochemically active material, a water splitting material, a catalyst, and combinations thereof. Said "guest" material may cover partially or totally the surface of said porous conductive matrix. Preferably, the O/EAM is photoactive. The O/EAM may be provided as a layer on one side and/or surface of the porous TCM, and/or it may be partially or totally filling the pores of the porous TCM, preferably only on one side of the TCM and/or electrode. The O/EAM may be a semiconductor material or other materials as mentioned above.

Preferably, the O/EAM covers completely and/or forms at least a complete, continuous layer on the TCM, in particular on one side of the TCM, which is opposed to a two-dimensional and/or flat surface on which the TCM may be provided and/or applied. According to an embodiment, the O/EAM is itself porous. Preferably, the O/EAM comprises a nano- or mesoporous surface and/or layer. For example, the O/EAM is based on and/or comprises particles, for example semiconductor particles having a mean particle size (diameter) of 1 nm to 1 μηη, preferably 4nm to 700nm, more preferably 5nm to 500nm, 6nm to 400nm, 7nm to 300, 8nm to 200nm an most preferably 10nm to 100nm, 5nm to 50nm, 10nm to 40nm, 5nm to 30nm, and 5nm to 20nm for example. State of the art dye-sensitized solar cells generally use a porous structure made of TiO ₂ nanoparticles having a mean particle size of generally not more than about 50m, for example up to about 20 nm. Such nanoparticulate anastase can be used as photoactive semiconductor material.

For example, the O/EAM has a roughness factor of greater than (>) 5, preferably >10, >20, preferably > 150, more preferably > 200, >500, >700, >1000, >1500, >1700, and >2000. A sol-gel process for preparing one or more nanoporous O/EAM layer(s) is disclosed in EP 0333641 B1 , for example, and/or in Wang, P.; Zakeeruddin, S. M.; Comte, P.; Charvet, R.; Humphry-Baker, R.; Gratzel, Enhance the Perfomance of Dye-Sensitized Solar Cells B Co-Grafting Amphiphilic Sensitizer and hexadecylmalonic Acid on Ti02 Nanocrystals M. J. Phys. Chem. B 2003, 107, 14336, in particular page 14337, in order to obtain a photoanode consisting of a double layer structure, with a transparent layer (20 nm particle) and scattering layer (400 nm particle) of 6.8 μηη and 4 μηη thickness, respectively.

Preferably, the O/EAM has a lower-sized porosity structure than the TCM. For example, the mean pore size of the O/EAM is at least 2x (2 times), 3x, 5x, 10x, 20x, 50x, 80x, 100x, 150x, 200x, 250x, 300x, 400x, 500, 600, 700x, 800x, 900x smaller than the mean pore size of the TCM. In this way, the O/EAM is provided in the pores of the TCM.

The O/EAM may be a material comprising or consisting any one or a combination of more of Si, TiO ₂ , SnO ₂ , Fe ₂ O ₃ , WO ₃ , ZnO, Nb ₂ O5, CdS, ZnS, PbS, Bi ₂ S3, CdSe, GaP, InP, GaAs, CdTe, CulnS ₂ , and/or CulnSe ₂ , be it doped or non- doped, and combinations thereof.

More generally, preferable examples of the semiconductor materials include oxides of Cd, Zn, In, Pb, Mo, W, Sb, Bi, Cu, Hg, Ti, Ag, Mn, Fe, V, Sn, Zr, Sr, Ga, Si, and Cr; perovskite such as SrTiO3 and CaTiOs; or suffides such as CdS, ZnS, ln ₂ S3, PbS, Mo ₂ S, WS ₂ , Sb ₂ S ₃ , Bi ₂ S ₃ , ZnCdS2, and Cu ₂ S; metal chalcogenide such as CdSe, ln ₂ Se ₃ , WSe ₂ , HgSe, PbSe, and CdTe; GaAs; Si; Se; Cd ₃ P ₂ ; Zn ₃ P ₂ ; InP; AgBr; Pbl ₂ ; Hgl ₂ ; and Bil ₃ . Alternatively, complexes containing at least one selected from the above semiconductors are preferable, such as CdS/TiO ₂ , CdS/Agl, Ag ₂ S/Agl, CdS/ZnO, CdS/HgS, CdS/PbS, ZnO/ZnS, ZnO/ZnSe, CdS/HgS, CdS _x /CdSei _-x , CdS _x /Tei _-x , CdSe _x /Tei _-x , ZnS/CdSe, ZnSe/CdSe, CdS/ZnS, TiO ₂ /Cd ₃ P ₂ , CdS/CdSeCdyZni _-y S, and CdS/HgS/CdS. TiO ₂ is the preferred O/EAM, in particular semiconductor material for dye sensitized solar cells.

According to an embodiment, the O/EAM is selected from semiconductor materials typically used and/or usable for solar cells, in particular dye-sensitized solar cells and/or water splitting devices.

The O/EAM may also form a layer on one side/surface of the TCM.

In particular in solar cells, the surface of the O/EAM not in contact with the TCM and/or the side of the O/EAM opposed to the face in contact with the TCM may be coated and/or in contact with a further optically and/or electrochemically active material (O/EAM), in particular a sensitizing material. Such materials include but are not limited to dyes, polymers, quantum dots, for example as used in photovoltaics. Dyes may be metal (for example ruthenium)-based dyes (organometallic compounds) or organic dyes. Exampels of dyes are disclosed in EP0613466, EP0758337, EP0983282, EP1622178, WO2006/038823, WO 2009107100, WO2009098643, for example. The organic or organometallic photoactive dye compounds may be attached and/or adsorbed by anchoring groups, such as carboxylic acid groups, phosphonic, phosphinic acid groups, and the like, to the (preferably mesoporous) surface of the O/EAM, as disclosed, for example, in the above listed patent documents. Along with sensitizing compounds, also co-adsorbing compounds may be attached to the surface of the O/EAM, preferably on the side which is not in contact with the TCM. The sensitizers and possibly co-adsorbed compounds may also form a layer, for a example a substantially continuous layer.

The electrode of the invention may be a photoanode or a photocathode, for example.

The invention encompasses a photoelectrode comprising the electrode of the invention. The invention encompasses electrochemical cells, waiter splitting devices, solar cells, photoelectroc conversion devices, for example dye-sensitized solar cells, optoelectronic devices comprising the electrode of the invention.

The electrode including the TCM and the O/EAM (for example a semiconductor material), and possibly the further O/EAM (for example, the sensitizer material) may be considered as a composite electrode.

In an aspect, the invention encompasses a photoelectrochemical device comprising a photoelectrode, which is the electrode of the invention, and a counter electrode. Preferably, a charge transport material (CTM) is provided between the photoelectrode and the counter electrode. The CTM may be in contact with the O/EAM or, if present, with the sensitizer (more generally, the further O/EAM) adsorbed on the O/EAM.

The CTM may be provided in the form of an electrolyte, in which charge transport is generally obtained by material diffusion. Electrolytes include solvent- based electrolytes and ionic liquid-based electrolytes. Alternatively, the charge transport may be provided by an electrically conductive material, such as electron and/or hole transporting materials, for example organic materials, in which charges are transported by electronic motion. Examples of electrically conductive materials suitable as CTMs are disclosed in WO2007/107961 .

Electrolytes may contain solvent or may be solvent free. They may comprise one or more ionic liquids. Electrolytes useful for photelectrochemical devices are disclosed in EP0737358, EP1507307, WO2007/093961 , WO2009083901 , for example.

According to an embodiment of the electrochemical device and/or electrode of the invention, the photoelectrode is a photoanode, and the counter electrode is a cathode.

The present invention provides a generic photoanode for photovoltaics and/or water splitting cells where a three-dimensional transparent conductive backbone provides efficient charge extraction and supports guest photoactive materials for light harvesting. Enhanced light harvesting is provided through scattering or through photonic crystal effects from the disordered or ordered three-dimensional structure respectively. Enhanced light harvesting, charge extraction and thus improvements in photocurrent and efficiency are possible with this method.

To expand upon, the composite photoanode preferqably comprises a macroporous transparent conducting oxide with built-in controlled order or disorder for enhanced light harvesting using photonic crystal effects or scattering respectively (see Figure 1 d). The pores will be fully or partially filled using light harvesting material(s) including anatase nanocrystal and dye for dye sensitized solar cells, organic polymers for polymer-based solar cells, quantum dots or semiconductors for hydrogen production by water splitting (see Figure 1 e & f).

The optically transparent electrically conducting macroporous framework of the composite anode can be made of either dense or mesoporous antimony tin oxide, fluorine tin oxide or any other transparent conductive materials. It can, for example, be made by inversion of an opal film template comprised of disordered or fee microspheres.

The perceived advantages of this construction over all previous designs is that separate transparent conductive current collector and photoactive materials are preferably combined into a single composite anode in which multiple light scattering from a pore network with predetermined disorder or photonic crystal-based optical effects will enhance the effective optical path length of the incident light, while at the same time intimate contact between high surface area transparent conducting material and light harvesting materials for photovoltaics or hydrogen production by water splitting will facilitate rapid transport and collection of charge carriers.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 : Schematic representation of one of the possible fabrication methods for a three-dimensional transparent conductive charge extraction backbone where (a) the three-dimensional template is grown onto a transparent conductive substrate; (b) the interstitial template pores are filled by vapor or solution phase methods with a transparent conductive material; (c) the template is removed chemically or by annealing; (d) the three-dimensional transparent conductive backbone is formed; (e) the porous three-dimensional TCO backbone is infiltrated with an active material, e.g. porous TiO2 for DSCs or Fe2O3 for water splitting; (f) For DSC, the dyed TiO2 infiltrated 3D TCO backbone is filled with electrolyte and the cell is closed with a cathode to complete the circuit. For water splitting, the Fe2O3 infiltrated 3D TCO is immersed in water which contains the cathode to complete the circuit.

Reference numbers used in Figure 1 :

1 . Substrate, for example glass.

2. Conductive coating, for example a transparent conductive coating, such as a TCO coating (ITO, FTO, etc.).

3. Template spheres.

4. Three-dimensional, porous conductive matrix, also referred to as transparent conductive host backbone.

5. Void; lumen of spheres, included in the communicating space comprising the interstitial space between said spheres (or balls) and the inner lumen of said spheres (or balls).

6. Guest material, for example nanoporous semiconductor layer

(for example T1O2) in case of a DSSCs, for example.

7. Electrolyte or charge transport material, such as hole or electron conductor, moving charges by electronic motion.

Figure 2: Cross-sectional view of an AI:ZnO | T1O2 3D host-guest structure, b, Top-view of an inverse AI:ZnO opal after top surface removal by RIE and template removal by annealing, c, High-magnification micrograph showing a T1O2 covered AI:ZnO host in direct contact with the front FTO electrode, d, High-magnification micrograph showing a 3D AI:ZnO conformally coated with a 25 nm thin film of dense T1O2 inside and out. e, Complete photoanode after infiltration with and calcination of the mesoporous T1O2 nanoparticle paste.

The following examples are illustrative of some of the devices and methods of making the same falling within the scope of the present invention. They are not to be considered in any way limitative of the invention. Changes and modifications can be made with respect to the invention. The skilled person will recognize many variations in these examples to cover a wide area of devices and processes. EXAMPLE 1

As an example of the invention disclosed herein, a method is described for fabricating a three-dimensional transparent conductive oxide charge extraction backbone to be used in photovoltaics dye sensitized solar cell. First, polystyrene particles are crystallized by convection induced self-assembly using a glass slide or a silicon wafer as substrates (Figure 1 a & b), as described in the literature, see J. F. Bertone, P. Jiang, K. S. Hwang, D. M. Mittleman, V.L. Colvin. Phys.Rev. Lett. 1999, 83, 300. In the present example, the transparent conductive material will be antimony doped tin oxide deposited from a tin precursor containing between 1 -10% antimony pentachloride. The uniform coating is deposited by chemical vapor deposition in a layer-by-layer fashion within the colloidal photonic crystal template at room temperature and atmospheric pressure (see Figure 1 b) as described by Miguez et al. Mechanical stability enhancement by pore size and connectivity control in colloidal crystals by layer-by-layer growth of oxide. Chem Commun (2002) (22) pp. 2736- 2737. The template is later remove and the transparent conductive oxide backbone is annealed at high temperature to induced crystallisation (Figure 1 d). It will then be coated by a few nanometers of dense TiO2 from titanium tetrachloride precursor to reduce charge recombination and improve electron injection. Next an internal coating of anatase nanocrystals uniformly throughout the network of pores in the inverse TCO will be achieved, for example, through nucleation and growth from a hexafluorotitanate solution (Figure 1 e). This is followed by carboxyl group anchoring of a standard ruthenium bipyridyl-based dye sensitizer onto the surface of anatase nanocrystals. Electrolyte will be diffused into the residual void spaces that exist in the composite anode thereby enabling construction and testing of the efficacy of the new genre of DSSC (Figure 1 f).

EXAMPLE 2

As an example of the invention disclosed herein, a method is described for fabricating a three-dimensional transparent conductive oxide charge extraction backbone to be used in hydrogen production through water splitting. First, polystyrene or silica particles are crystallized by convection induced self-assembly using a glass slide or a silicon wafer as substrates (Figure 1 a & b), as described in the literature, see J. F. Bertone, P. Jiang, K. S. Hwang, D. M. Mittleman, V.L. Colvin. Phys.Rev. Lett. 1999, 83, 300. In the present example, the transparent conductive material will be antimony doped tin oxide deposited from a tin tetrachloride precursor containing between 1 -10% antimony pentachloride. The uniform coating is deposited by chemical vapor deposition in a layer-by-layer fashion within the colloidal photonic crystal template at room temperature and atmospheric pressure (Figure 1 c) as described by Miguez et al. Mechanical stability enhancement by pore size and connectivity control in colloidal crystals by layer-by-layer growth of oxide. Chem Commun (2002) (22) pp. 2736-2737. The template is later removed and the transparent conductive oxide backbone is annealed at high temperature to induced crystallisation (Figure 1 d). It will be coated by a few nanometers of dense hematite Fe ₂ O ₃ using chemical vapor deposition techniques like atomic layer deposition or atmospheric pressure chemical vapor deposition from Fe(CO) ₅ (Figure 1 e & f).

EXAMPLE 3

As an example of the invention disclosed herein, a method is described for fabricating a three-dimensional transparent conductive oxide charge extraction backbone to be used in photovoltaics dye sensitized solar cell. The overall fabrication method is inspired by micromolding in inverse opals (MISO) first developed for the synthesis of large-area oxide inverse opals from polymer templates. In particular, by speeding up the evaporation-induced self-assembly of polystyrene macrospheres (Φ = 2.2 m) the capillary forces are manipulated to obtain a highly disordered opal template of uniform thickness. Using a quick self-assembly of large monodispersed spheres enables the formation of a uniformly thick template with predictably large interconnecting pores extending in all 3D. The thickness of the template was adjusted through the volume concentration of spheres in the isopropanol dispersion (~ 22 vol% for 10-12 μιτι templates). The dispersion is then doctor bladed. Adjustment by doctor blading represents a scalable fabrication technique, on a transparent conductive FTO-covered glass substrate to reveal a large-area disordered opal template (Fig. 1 B). The template is then heated and infiltrated with 90 nm of AI:ZnO (1 :9) or other TCMs in a highly conformal fashion by atomic layer deposition (ALD) by alternating pulses of organometallic precursors and an oxidizing agent (e.g. H ₂ O) in exposure mode (Fig. 1 C). The deposition temperature is optimized to increase the template interconnecting pore size by softening of the polymer beads ensuring proper sphere necking. Large interconnecting pores will enable proper filling with the guest ΤΊΟ2 later on. A quick dry etching of the infiltrated opal's top surface oxide is performed before removing the polymer template by annealing (300°C, 15 min.). We thus obtain a 3D TCM host backbone that is well connected to the underlying FTO-glass front electrode. The direct electronic connection will ensure efficient charge extraction throughout the interconnected 3D AI:ZnO and ΤΊΟ2 electrode. The TCO backbone is then conformally coated with 25nm of dense TiO ₂ by ALD in order to reduce interfacial recombination between electrons in the highly conducting TCM backbone and the oxidized electrolyte. The photoanode is then calcined (500°C,15 min.) to remove all traces of organics from the T1O2 paste. In order to provide enough surface area for dye sensitization and efficient light harvesting, the macroporous inverse TCM backbone is infiltrated by sequential doctor-blading of a low viscosity 17nm anatase nanoparticle paste (Fig 1 E).

Scanning electron micrographs of AI:ZnO | T1O2 host-guest photoanodes are presented in Figure 2. The cross-sectional and top view (AI:ZnO host) micrographs (Fig. 2 a,b) clearly show the high pore connectivity throughout the inverse 3D structure as well as the significant disorder inherited from the polystyrene template. From these micrographs, we can conclude that the backbone only uses a small volume fraction of the 3D film. This is an additional advantage of the inverse opal structure where the high degree of necking between the templating spheres and the very thin (90 nm) film of metal oxide host deposited leave the vast majority of the volume for filling with the photoactive guest material. The extremely porous structure further enables the conformal coating of the inverse opal host on both sides with dense T1O2 by ALD (Fig. 2 c,d). The role of the TiO2 overlayer is to passivate the AI:ZnO surface to improve its stability and further reduce recombination rate at the conductive transparent oxide electrolyte interface. Visible in Figure 2 c is the direct contact between the 3D host and the FTO-glass that will ensure unhindered electron transfert from the 3D host-guest structure and the front DSC electrode. Additionally, we find that the subsequent T1O2 deposition covers the FTO in addition to the AI:ZnO to effectively block recombination at the FTO - electrolyte interface. Figure 2e shows the 3D AI:ZnO | T1O2 host-guest electrode after partial filling with mesoporous anatase to constitute the finished 3D TCM photoanode. We estimate from the latter picture that about 50% on the internal volume is filled with the anatase nanoparticles. The disordered structure and the partial filling at the wavelength scale combine to induce high scattering in the 3D Host-guest structure.

Complete DSC devices are prepared using 10-12 μm thick 3D TiO ₂ | TiO ₂ and AI:ZnO | TiO ₂ host-guest photoanodes on a FTO-glass substrate. Then, the electrodes are dipped for 10min in a 0.3 mM Z907 (90 vol% 1 :1 acetonitrile:tert- butanol, 10 vol% dimethylformamide) for 10 min and washed off in pure acetonitrile. Z907 has been supplanted recently by ruthenium and organic dyes with higher molar coefficient but its well-characterized photovoltaic characteristics enabling a direct comparison with published data (Z907: J _sc = 17.13 mA cm ² , V _oc = 730 mV, FF = 0.724, η = 9.05). Following the immersion procedure, the dye-sensitized electrodes were rinsed with acetonitrile and dried in air. The working photoanodes are then assembled using a thermally platinized FTO/glass (Tek15) counter electrode using a 25 Mm thick hot melt ring (Surlyn, DuPont) and sealed by heating. The cell internal space was filled with a volatile electrolyte (Z960: 1 .0 M 1 ,3-dimethylimidazolium iodide, 0.03 M iodine, 0.5 M tert-butylpyridine, 0.05 m Lil, 0.1 M guanidinium thiocyanate), in an 85:15 acetonitrile/valeronitrile mixture through a pre-drilled hole using a vacuum pump.

We find an increase in short-circuit current going from a TiO ₂ (J _sc = 6.9 mA cm ² ) to AI:ZnO (J _sc = 7.9 mA cm ² ) 3D host. It is possible that the increase in photocurrent observed is a direct consequence of the high electron mobility in the AI:ZnO host (241 cm ² /Vs) which is 7 orders of magnitude higher than in anatase nanoparticles (2.30 x 10 ^"4 cm ² /Vs). In addition, we find that the fill factor is greater for the AI:ZnO {FF = 0.77) host than for TiO ₂ {FF = 0.73) host. Regardless of the host material, we find that the 3D host-guest photoanodes all present photovoltages well above the V _oc = 730 mV obtained with the best performing Z907 DSCs. Indeed, we find the an improvement of 60 mV and 1 10mV for TiO ₂ {V _oc = 791 mV) and AI:ZnO ( Voc = 842 mV) hosts respectively.