Dual Absorber Electrodes

Abstract

Dual absorber electrodes are disclosed. In some embodiments, a dual absorber electrode includes a first absorber material, such as silicon, having a first bandgap, and a second absorber material, such as hematite, deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber. In some embodiments, the dual absorber electrodes of the present embodiment may be utilized in an electrolytic cell for water splitting.

Description

FIELD

The embodiments disclosed herein relate to photoelectric electrodes for use in water splitting.

BACKGROUND

Photoelectrochemical (PEC) water splitting offers the capability of harvesting the energy in solar radiation and transferring it directly to chemical bonds for easy storage, transport, and use in the form of hydrogen. Among the various considerations of a PEC system, the choice of photoelectrode materials is especially important because their properties, such as optical absorption characteristics and chemical stability, determine the system's performance. These materials should absorb light broadly, be inexpensive, and be resistant to photo corrosion.

SUMMARY

Dual absorber electrodes are disclosed. According to some aspects illustrated herein, there is provided a dual absorber electrode that includes a first absorber material having a first bandgap; and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber.

According to some aspects illustrated herein, there is provided a device for splitting water to generate hydrogen and oxygen that includes a first compartment having a first electrode, the electrode comprising a first absorber material having a first bandgap, and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber; a second compartment having a second electrode for catalyzing hydrogen generation; and a semi-permeable membrane separating the first compartment and the second compartment.

According to some aspects illustrated herein, there is provided a dual absorber electrode that includes a first electrode comprising first absorber material and a second absorber material having a larger band gap than the first absorber material; and a second electrode for catalyzing hydrogen generation in electrical contact with the first absorber material.

According to some aspects illustrated herein, there is provided a method of fabricating an electrode that includes obtaining a silicon nanostructure; forming by vapor deposition a layer of hematite on a surface of the silicon nanostructure by exposing the silicon nanostructure to gas precursors of hematite; and repeating the vapor-depositing to form additional layers of hematite on the surface of the silicon nanostructure.

BRIEF DESCRIPTION OF THE DRAWINGS

The presently disclosed embodiments will be further explained with reference to the attached drawings, wherein like structures are referred to by like numerals throughout the several views. The drawings shown are not necessarily to scale, with emphasis instead generally being placed upon illustrating the principles of the presently disclosed embodiments.

FIG. 1 illustrates an embodiment dual absorber electrode of the present disclosure.

FIG. 2 illustrates an embodiment electrolytic cell for water splitting suitable for use with a dual absorber electrode of the present disclosure.

FIG. 3 illustrates an embodiment dual absorber particle of the present disclosure.

FIG. 4 illustrates an embodiment method using the dual absorber particles presented in FIG. 3 for water splitting.

FIG. 5A and FIG. 5B present energy band schematics (quasi-static equilibrium under solar illumination) and photocurrent-voltage plots for water oxidation by Fe₂O₃ on Si nanowires or FTO (fluorine doped tin oxide).

FIG. 6A and FIG. 6B present microstructure of the Si/Fe₂O₃ nanowire photoelectrodes. FIG. 6A presents a scanning electron micrograph (SEM) showing the arrangement of chemically-etched Si nanowires. FIG. 6B presents a transmission electron micrograph (TEM) showing the crystalline quality of the ALD-grown Fe₂O₃ film on the Si nanowire surface. Insets: electron diffraction pattern (top right) and energy-dispersive X-ray spectroscopy (EDS) line scans revealing the compositional makeup across the Si/Fe₂O₃ interface (bottom).

FIG. 7 is a low-magnification transmission electron microscope (TEM) image showing the uniform and conformal nature of the ALD-deposited hematite film on a typical etched Si nanowire.

FIG. 8A and FIG. 8B present results of SiNWs/hematite photoanode testing. FIG. 8A presents a graph of photocurrent measured over 3 h under AM1.5 illumination and 1.0 V_RHE applied bias. Approximately every 15 min, the illumination was blocked for a few seconds to check the dark current, which remained around zero for the duration of the test. FIG. 8B presents J-V curves before and after this 3 h test, showing negligible change in performance.

FIG. 9A, FIG. 9B and FIG. 9C present J-V curves for hematite films on planar substrates under different illumination conditions.

FIG. 10A and FIG. 10B present band diagrams at pre-equilibrium and equilibrium in dark, respectively. At equilibrium, band bending at each junction results due to the equilibration of chemical potentials across the materials. Band edge positions and Fermi levels are approximate, determined theoretically and from literature reports, and the x-axis depth of each material is not to scale. The nature of the band behavior at the interface has been proposed in the literature.

FIG. 11A, FIG. 11B, FIG. 11C, and FIG. 11D demonstrate the dual-absorber nature of hematite/Si nanowires photoanodes for water oxidation. FIG. 11A presents J-V curves for a typical electrode under different and combined monochromatic illumination suggests a synergistic effect when combining low-energy and high-energy photons, and representative band energy diagrams show the two cases of independent excitation. FIG. 11B presents the photocurrent responses under stepwise monochromatic chopped illumination in the wavelength range 300-1000 nm, under applied bias of 1.0 V_RHE, reveal the characteristic responses of Fe₂O₃ and Si in the device. In reference to FIG. 11C, in the short wavelength range, photons are primarily absorbed by the Fe₂O₃, and stable anodic photocurrents are observed. As shown in FIG. 11D, at longer wavelengths, photons pass through Fe₂O₃ to excite only Si, producing zero net current.

FIG. 12A and FIG. 12B present the photocurrent responses of hematite on FTO under stepwise monochromatic chopped illumination in the wavelength range 300-780 nm under applied bias of 1.45 V_RHE.

FIG. 13A presents representative photocurrent-voltage plots for Fe₂O₃/Si nanowire and Fe₂O₃/FTO devices.

FIG. 13B presents representative plots for Fe₂O₃/Si nanowire devices of different Si doping characteristics. The details of the Si doping are tabulated below.

While the above-identified drawings set forth presently disclosed embodiments, other embodiments are also contemplated, as noted in the discussion. This disclosure presents illustrative embodiments by way of representation and not limitation. Numerous other modifications and embodiments can be devised by those skilled in the art which fall within the scope and spirit of the principles of the presently disclosed embodiments.

DETAILED DESCRIPTION

In order to completely split water, 1.6V of potential is needed. However, there are many photoanode materials with otherwise desirable properties that cannot produce 1.6V of potential. For example, research on using hematite (Fe2O3) to absorb solar light and split water is moving at a slow pace despite the positive prospect of hematite having the suitable bandgap and being low cost, and the limiting factor has been the intrinsic physical and chemical properties of this material. In particular, hematite suffers from a number of challenges, including, short hole diffusion distances, poor catalytic activities, and mismatch of hematite band edge positions with the reduction and oxidation potentials of water. A considerable voltage is needed to induce the water-splitting reaction (fundamentally 1.23 V but upwards of 2 V in practice), so a large band gap semiconductor would be desired. However, the band gap defines the minimum energy a solar photon must have to be absorbed, and thus larger bandgaps lead to decreased light absorption and therefore lower efficiencies. In actual photoelectrolysis cells, this limitation manifests as a dependence on an external applied bias to achieve the necessary voltage for water splitting. A primary goal is therefore to reduce or eliminate this dependence and achieve unassisted water splitting. To do this, the solar spectrum needs to be more efficiently utilized and a greater photovoltage needs to be generated.

The materials of the present disclosure address these challenges by direct coupling of a metal oxide, such as hematite, with smaller-bandgap materials, such as silicon, to create multiple-absorber systems. The materials of the present disclosure can utilize the portion of solar spectrum that cannot be absorbed by hematite (X>600 nm), which accounts for more than half of the total solar energy. Electrodes of the present disclosure can collect red and near-infrared photons (600 nm<λ<1100 nm) to provide extra photovoltages when hematite is interfaced with Si nanowires. The resulting photoelectrodes exhibit a photocurrent turn-on potential as low as 0.6 V_RHE, where V_RHE represents the applied potential relative to the standard potential of a reversible hydrogen electrode, or RHE. The combined materials of the present disclosure thus directly enhance the photovoltage obtainable on a single device. This results in a reduction in the magnitude of applied voltage which is necessary to achieve water splitting. Conventional approaches rely on external sources to provide the entire additional voltage, including electronic power supplies or photovoltaic modules. The improvement of the present disclosure comes from the more efficient utilization of the solar spectrum resulting from the use of two absorbers, a unique demonstration for metal oxide devices.

It should be noted however that the electrodes of the present can be utilized in other photovoltaic applications, not just water photoelectrolysis. Solid-state photovoltaic or photoelectrochemical cells in which the generated charge is collected in electronic form, rather than water splitting, could be designed by this same principle. The presently disclosed design can also be used to supply charge to other electrochemical reactions other than water splitting, including, but not limited to, photosynthesis of other useful molecules or fuels. The materials of the present disclosure may be used in photoelectrochemical synthesis applications in which photo-generated charge is used to drive chemical reactions, in photovoltaic cells to generate electricity, and in solar filters designed to selectively block light.

In reference to FIG. 1, electrodes 100 of the present disclosure comprise two absorber materials 102, 104 of different bandgap, wherere the absorbers are stacked together to allow the light to pass through the largest-bandgap absorber first. In some embodiments, the inner absorber 102 having a smaller band-gap may be in the form of a nanowire, nanoparticle, nanorod, nanonet or another nanostructure. In some embodiments, the higher bandgap absorber may form at least a partial shell around the smaller bandgap absorber. In some embodiments, the inner absorber material 102 of the presently disclosed electrodes comprises one or more smaller bandgap materials such as, for example, Si, Ge, InP, GaAs, and similar materials. The outer absorber 104 of the presently-disclosed electrodes may comprise one or more larger-bandgap materials which are designed to produce larger photovoltages as a result of having multiple complimentary light absorbers. Suitable larger-bandgap materials include, but are not limited to Fe₂O₃, WO₃, TiO₂, SnO₂, ZnO, and other similar metal oxides. In some embodiments, the electrode 100 may also include an electrical contact 106 electrically connected to the inner absorber maternal 102 to enable collection of electrons generated inside the inner absorber material 102.

In some embodiments, the dual absorber electrodes of the present disclosure include a Si nanowire core with hematite deposited over at least a part of the Si nanowire in a conformal fashion. When hematite is deposited on Si nanowires in a conformal fashion, hematite and Si can be independently excited by photons of different energies in the solar spectrum. Under simultaneous excitation, a dual-absorber mechanism develops. Charge flow is enabled only when both are excited in a synergistic manner, and the photopotentials developed within the two materials contribute to enhanced water splitting performance as evidenced by a cathodic shift in the photocurrent onset potential. The onset potential of 0.6 V_RHE represents one of the lowest reported for hematite photoanode devices, and is achieved without the use of catalysts, hematite doping, or surface treatments. Accordingly, the utility of hematite can be improved using direct coupling with small-bandgap materials to more efficiently utilize the solar spectrum and to enhance the photovoltages attainable by a single device. The device fabrication depends on the growth of high-quality thin films of hematite made possible by the ALD technique. Furthermore, the fact that the active materials are primarily composed of three of the four most abundant elements in Earth's crust (O, Si, Fe) offers promise that renewable energy harvesting by photoelectrochemical water splitting remains an achievable goal.

FIG. 2 shows an exemplary illustration of a device 1100 of the present disclosure for use in water splitting. The device 1100 includes two compartments, 1110 and 1120, each of which can be used for the half reactions of H₂ and O₂ generations. Solar energy is harnessed to separate charges, which then transfer to the redox pairs in the solutions to perform reactions. The appropriate energy alignment can be enabled by material choices (p-type for H₂ and n-type for O₂) and the adjustment of solution pH. Highly conductive components ensure efficient charge transport, thus completing the full reaction of H₂O splitting. In an embodiment, compartment 1110 is filled with an acidic solution, and compartment 1120 is filled with a basic solution. Compartments 1110 and 1120 are separated by a semi-permeable membrane 1140 that only allows ionic exchange to balance potential buildup. In an embodiment, the semi-permeable membrane 1140 is a charge-mosaic membrane (CMM). In the acidic compartment 1110, a p-type material acts to produce H₂ upon illumination. The device 1100 includes electrodes 1115 and 1125. In some embodiments, one electrode may comprises a dual absorber material of the present disclosure, such as silicon core embedded in a hematite shell. In some embodiments, the second electrode catalyzes hydrogen generation. The second electrode may be made of platinum or another metal capable of acting as a catalyst for hydrogen generation, supporting hydrogen generation or both.

The electrodes 1115 and 1125 can be connected together by external contacts 1150 to ensure charge balance. In the solution, opposite charges flow through the semi-permeable membrane 1140 to annihilate each other. Both the acidic and the basic solutions should be periodically refreshed by adding more acids or bases to maintain an appropriate chemical potential difference by maintaining a preset pH difference.

In reference to FIG. 3, in some embodiments, the materials of the present disclosure may be presented as particles 300 including a dual absorber material including an inner absorber 302 and an outer absorber 304, combined with a suitable catalyzing member 306 such as platinum or another catalyst metal. The catalyzing member 304 may essentially act as the second electrode for photochemical water splitting. The inner absorber 302 may be in electrical contact with the catalyzing member 306 to allow electrons to travel from the inner absorber 302 to the second electrode material.

In reference to FIG. 4, in some embodiments, a plurality of particles 300 may be added to a container 310 containing water 312 to effectuate splitting of water 312 into hydrogen an oxygen atoms upon absorption of light by the particles 300.

In some embodiments, an electrode of the present disclosure may include two (or more) semiconductors which absorb in different regions of the solar spectrum. In some embodiments, the outer absorber, such as hematite, operates as a typical photoelectrode for water photo-oxidation, while the inner absorber, such as the Si nanowire, uses the energy in long-wavelength photons to further increase the energy of electrons that will be ultimately utilized for water photo-reduction. The net effect is that hematite-based water splitting can be carried out at reduced external potentials.

In some embodiments, the outer absorber material of the present disclosure may include a photovoltaic junction, either a p-n junction or a p-i-n junction. In some embodiments, the outer absorber material of the present disclosure is formed by coupling an n-doped hematite and a p-doped hematite to form a p-n hematite junction. In some embodiments, the inner absorber material may include a photovoltaic junction, either a p-n junction or a p-i-n junction. In some embodiments, both the outer absorber material and the inner absorber material include or are a photovoltaic junction.

In some aspects, the present disclosure provides methods of growing photovoltaic hematite on lower bandgap material. Various gas phase deposition methods may be utilized to form photovoltaic hematite junctions of the present disclosure, including, but not limited to, atomic layer deposition, chemical vapor deposition, pulse laser deposition, evaporation and solution synthesis approach and similar methods. In some embodiments, a uniform interface is formed between the outer absorber material and the inner absorber material. In some embodiments, the outer absorber material is conformal to the surface of the inner material absorber. In some embodiments, the interface between absorbers has low defect densities and low impurity levels. Metal oxide deposition techniques such as hydrolysis, simple vapor phase deposition, may typically fail to produce an interface that meets this requirement. In some embodiments, the outer absorber material is deposited by atomic layer deposition to form an ultra-thin film of the outer absorber material one atomic layer at a time.

In some embodiments, the method of forming an electrode of the present disclosure includes depositing multiple layers of hematite (outer absorber material) on a surface of a silicon nanostructure (inner absorber material). First, a silicon nanostructure of a desired shape and size is obtained. Next, a layer of the hematite may be deposited on the silicon nanostructure by exposing the silicon nanostructure at a first temperature to a pulse of iron precursor, followed by a pulse of oxygen precursor. In some embodiments, the first temperature may be selected based on the reaction temperature of precursors, such as, for example between about 140 C and about 180 C, to enable formation of n-type hematite from precursors on the substrate surface. In some embodiments, the precursors may be maintained at a temperature from 120 C to 135 C to yield appreciable vapor pressure of precursor. This step may be repeated until the layer of the hematite is of a desired thickness. Subsequently, the electrode may be annealed at a temperature sufficient to crystallize the hematite.

The methods and materials of the present disclosure are described in the following Examples, which are set forth to aid in the understanding of the disclosure, and should not be construed to limit in any way the scope of the disclosure as defined in the claims which follow thereafter. The following examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the embodiments of the present disclosure, and are not intended to limit the scope of what the inventors regard as their invention nor are they intended to represent that the experiments below are all or the only experiments performed. Efforts have been made to ensure accuracy with respect to numbers used (e.g. amounts, temperature, etc.) but some experimental errors and deviations should be accounted for.

EXAMPLES

EXPERIMENTAL PROCEDURE: Fabrication of α-Fe₂O₃/Si NW devices: Silicon nanowire (NW) arrays were synthesized by an electroless chemical etching method previously reported. A cleaned n-type Si (100) substrate (P-doped, 5-15 Ωcm, University Wafer) was treated in a 3:1 (v/v) H₂SO₄/H₂O₂ at 90° C. for 15 min. and then rinsed thoroughly with DI water. The substrate was then immersed, with polished side facing upwards, into a solution of 4.4 M HF and 0.02 M AgNO₃ at 50° C. A typical etching time of 30 min. created NWs of approximately 8 μm in length. The substrate was then removed, rinsed in DI water, and immersed in concentrated HNO₃ for 15 min. to remove Ag from the Si surfaces. Finally, the substrate was gently but thoroughly rinsed in DI water and gently dried in an N₂ stream.

The general technique for Fe₂O₃ deposition has been previously reported. The as-prepared Si NW arrays (or other substrates, including cleaned planar Si and cleaned fluorine-doped tin oxide on glass) were transferred to an atomic layer deposition system (ALD; Cambridge Nanotech, Savannah 100) and the deposition chamber was evacuated to a base pressure of 0.36 Torr. Iron tert-butoxide (heated to 125° C.) and water (25° C.) served as the precursors for Fe₂O₃ and were pulsed alternatingly into the deposition chamber (heated to 180° C.) with a 10 cm³ min⁻¹ flow of N₂ as carrier gas. The specific pulse sequence was previously reported. Following deposition, a 15 min. heat treatment at 500° C. in O₂ was applied to all Fe₂O₃ samples to complete the synthesis of photoactive hematite.

Anode devices for photoelectrochemical testing were then prepared. For Fe₂O₃/Si NW devices, a small area of fresh Si was exposed on the NW array surface by scraping away the NWs with a blade and roughing the Si with a diamond pen. For ohmic contact to the n-type Si, a drop of Ga/In eutectic (Sigma Aldrich) was used for contact between the Si and a copper wire. Then, the contact area and the edges and backside of the substrate were passivated by application of insulating epoxy. The resulting exposed area of Fe₂O₃-coated Si NWs was defined as the active area to which photocurrents were normalized for each specimen.

Photoelectrochemical testing: The photoelectrochemical behavior of all samples was tested using a potentiostat (CH Instruments CHI604C) in a three-electrode configuration, with an alkaline/mercury oxide reference electrode (Hg/HgO/1M NaOH; CH Instruments) and a Pt wire counter electrode. An aqueous electrolyte solution of 1.0 M NaOH (pH 13.6), bubbled with N₂ gas, was used. In a typical J-V experiment, the voltage was swept linearly from 0.6 to 1.6 V_RHE at a rate of 10 mV s⁻¹. Illumination sources included an AM 1.5 solar simulator (100 mW cm⁻², Newport Oriel 96000), an ultraviolet lamp (365 nm wavelength, UVP UVGL-55), and an infrared laser (980 nm wavelength, 2000 mW cm⁻² power, AixiZ). For the monochromatic spectral illumination shown in FIG. 3b, a monochromator (Newport Oriel Cornerstone 260) was coupled to the solar simulator.

Structural characterization: A scanning electron microscope (JEOL JSM6340F) and transmission electron microscope (JEOL JEM2010F) were used for specimen imaging. The TEM specimen was prepared by dispersing scraped Fe₂O₃/Si nanowires in isopropyl alcohol and dropping them onto Cu grids with lacey C coatings. Assuming a cylindrical cross-section, a Si nanowire diameter of 50 nm, and a Fe₂O₃ thickness of 20 nm (typical values), the specimen thickness at the Fe₂O₃/Si interface (along the electron beam path) was approximately 75 nm. An energy-dispersive X-ray spectrometer (Oxford Inca) was used in conjunction with the TEM to determine material compositions. All TEM and EDS data were obtained using a 200 keV accelerating voltage.

RESULTS: FIG. 5A shows a Fe₂O₃/Si system that relies on dual light absorbers for charge generation and transfer, with Fe₂O₃ absorbing higher-energy photons (>2.1 eV) and Si absorbing lower-energy photons. Holes generated in Fe₂O₃ transfer to the electrolyte to oxidize water and evolve O₂ gas. Photoexcited electrons in the Fe₂O₃ conduction band undergo favorable recombination with holes from Si at the junction. Lastly, photoexcited electrons in Si, which achieve more negative energies than in Fe₂O₃ itself, travel via the external circuit to perform the water reduction and generate H₂ gas.

FIG. 5B provides photocurrent-voltage plots under simulated solar illumination (AM 1.5, 100 mW cm⁻²) in 1.0 M NaOH aqueous electrolyte (pH: 13.6; scan rate: 10 mV s⁻¹) show that the presence of the Fe₂O₃/Si junction leads to a significant cathodic shift of the photocurrent onset potential and the nanowire array morphology results in enhanced photocurrent (red), as compared to Fe₂O₃ on FTO (black).

FIG. 5B compares the photocurrent density with applied potential (J-V) plots of hematite photoelectrodes with and without the second absorber of Si nanowires. Most obvious is the significant shift of the curve toward the cathodic direction. Because the two photoelectrodes were prepared by the same batch of hematite growth, the possibilities of unintentional doping or other phenomenological surface effects such as passivation or catalyst decoration can be ruled out. A second artifact considered was whether the current came from photo corrosion of Si, which would be a result of unsuccessful coverage of Si nanowires by hematite. Electron microscopy studies revealed that the hematite layer was conformal, as shown in FIG. 6A, FIG. 6B and FIG. 7.

The crystalline nature of hematite was also confirmed by the inset electron diffraction pattern in FIG. 6B. While n-type doping by Si has been shown to enhance electron conductivity in hematite, no appreciable Si concentration was detected in the hematite layer of our Fe₂O₃/Si nanowire devices, suggesting negligible outward diffusion of Si into hematite took place during the mild post-growth annealing treatment (500° C.).

In reference to FIG. 8A and FIG. 8B, sustained photocurrent without decay was measured for up to 3 h of continued photoelectrochemical reactions. The data presented in FIG. 8A and FIG. 8B support the conclusion that photocurrent is not due to Si oxidation and that the atomic layer deposition (ALD) coating of hematite protects the device from degradation. FIG. 8B shows that the photocurrent did not degrade over 3 h, but instead increased slightly. Before and after J-V curves in FIG. 8B show that the device performance was not significantly affected by the long test.

If a significant portion of the measured current did come from photo corrosion of Si, a drastic decay of photocurrent would be expected. Taken as a whole, it was concluded that the cathodic shift of the photocurrent as shown in FIG. 6B is indeed a result of additional photovoltage produced at the Fe₂O₃/Si junction by Si nanowires.

In the systems of the present disclosure, the overall photocurrent may be limited by the lowest performing component which, in the present case, is hematite. As such, unless light absorption by hematite is significantly improved, a dramatic increase of photocurrent is not expected.

FIG. 9A, FIG. 9B, and FIG. 9C present J-V curves for hematite films on planar substrates under different illumination conditions. FIG. 9A shows that a planar Si substrates yield hematite devices with similarly low onset potentials (<0.8 V_RHE) but smaller photocurrent magnitudes, when compared with the Si nanowire based devices. The lower current is largely due to the smaller surface area and poorer light absorption by planar devices. It should be noted that the photocurrents were even lower than those measured on planar FTO substrate by a factor of 2. The difference may be attributed to how the substrate influences the hematite growth. The variety of crystal surfaces on etched Si nanowires may be preferred over planar Si (100) which may not be ideal for ALD hematite growth.

FIG. 9B shows planar device behavior under monochromatic illumination exhibits some differences from nanowire devices. Primarily, the UV lamp illumination yields performance close to the combined UV and IR illumination condition. This may be due to the planar substrate leading to shorter photon path lengths in Fe₂O₃ (i.e. the Fe₂O₃ film appears thinner to photons passing through along paths normal to the film surface) and therefore more UV photons pass through Fe₂O₃ to excite the Si.

FIG. 9C presents data for hematite devices on FTO substrates, which suggest that such devices respond only to the UV illumination, with no difference caused by adding the high-power IR laser illumination. This is expected for a single-absorber (Fe₂O₃) system.

In reference to FIG. 9A, when hematite was grown on planar Si substrates, low photocurrents were measured. The higher photocurrent measured on the hematite/Si nanowires combination may be a result of increased light absorption by hematite on a textured substrate (Si nanowires), and may be due to the enhanced path length for photon absorption and the increased surface area. Dynamic electrolyte diffusion into and from the regions between adjacent Si nanowires is poor, since an electrochemically active effective surface-area-to-projected-area ratio of <2, which is two orders of magnitude lower than the true surface area, has been measured on Si nanowires prepared by the same method in the Me₂Fc/Me₂Fc⁺ electrochemical system. The cathodic shift of the J-V plots observed on the planar hematite/Si system, as shown in FIG. 9A, was comparable to that shown in FIG. 5B.

In reference to FIG. 10A and FIG. 10B, in combination with FIG. 5A, the band diagrams predict that the magnitude of cathodic shift should depend on how much photovoltage can be produced by Si nanowires, which in turn is determined by the extent of band bending within Si. One way to test this prediction is to change the position of the Fermi level within Si by, for instance, using Si with different doping levels or doping type or both. When the Fermi level is closer to the valence band edge (such as in p-type Si), the photovoltage obtained on Si would be smaller, resulting in a lower value in the cathodic shift. Combined with the fact that an appreciable band-bending depth is necessary for charge separation, a lightly n-doped Si is desired for the dual absorber system to exhibit the most significant cathodic shift. This indeed was the case. As shown in FIG. 13B, Si n-doping levels of 10¹⁴, 10¹⁶, and 10¹⁸ cm⁻³ were examined which led to typical onset potentials of approximately 0.6, 0.8, and 0.9 V_RHE, respectively, whereas using p-type Si (10¹⁵ cm³) resulted in no cathodic shift as compared to hematite/FTO (fluorine doped tin oxide) devices.

In some embodiments, the dual absorber is designed such that both photon-to-charge conversion processes take place in a concerted fashion. To this end, a set of experiments was carried out to prove the dual absorber nature of these devices using monochromatic light for excitation. First, an ultraviolet (UV) lamp was used to illuminate the device with a wavelength of 365 nm at a relatively low power (3 mW cm⁻²). These UV photons have sufficient energy to excite hematite and therefore most of them are absorbed within hematite itself, rather than penetrating through to the Si. The resulting J-V curve and representative band diagram are shown in FIG. 11A (purple). When hematite alone is excited, photogenerated electrons flow towards Si where they encounter a barrier which prevents their flow. This results in charge accumulation but no net current flow through the device until a sufficient anodic bias is applied to allow electrons to tunnel through to the Si conduction band. The resulting photocurrent onset potential of ˜1.0 V_RHE reflects the expected performance for a hematite-only device without significant contribution from Si.

Conversely, when an infrared (IR) laser of 980 nm wavelength was used as the lone light source, the photons had insufficient energy to excite hematite and instead passed through to be absorbed by Si. Despite the high power of the IR illumination (laser power ˜2000 mW cm⁻²), no appreciable photocurrent was observed until a considerable anodic potential was applied, with only small photocurrents emerging when biased above 1.2 V_RHE (FIG. 11A, red). This may be due to the fact that photoexcited holes in Si cannot be annihilated by electrons from the hematite conduction band, and furthermore they do not have sufficient energy to inject into hematite and perform water oxidation without large anodic bias.

However, combining the two light sources for simultaneous illumination of the hematite/Si nanowires device created a synergistic effect that resulted in full development of the photopotential and cathodic shift of the onset potential. The J-V curve (FIG. 11A, blue) shows that, in fact, two combined single-wavelength light sources are capable of producing a curve shape similar to that obtained under full solar spectrum illumination as presented in FIG. 5B, albeit with a lower photocurrent magnitude due to the bottleneck in hematite caused by using low-power UV illumination. For the case of dual excitation, the band diagram in FIG. 5A portrays the cathodic shift of the bands of each n-type material under illumination and the electronic current flow that is produced. The synergy produced by combining the UV and IR illumination sources illustrates the dual-absorber nature of the hematite/Si nanowires photoelectrodes. This clearly contrasts the single-absorber mechanism of typical hematite/FTO devices, wherein only the UV illumination elicited a photocurrent response (FIG. 9C). It is also different from when Si is the lone active photoanode material, with hematite acting in a passivating or catalytic role.

As further evidence of the device response to photon energy, the wavelength-dependent photocurrent was measured using monochromatic light in the wavelength range 300-1000 nm under a 1.0 V_RHE applied bias. The illumination was achieved by passing simulated solar light (AM 1.5 spectrum; intensity adjusted to 100 mW cm⁻²) through a monochromator. Depicted in FIG. 11B, the photocurrent response reflected the behavior predicted by the J-V curves and band diagrams in FIG. 11A. In the short-wavelength range (300-580 nm) where the primary absorber is hematite, appreciable anodic photocurrent was observed. Since the photocurrent was nonzero, the photoexcited electrons in hematite were reaching the external circuit, meaning that under this applied anodic bias photoexcited electrons from the hematite conduction band can inject into Si to produce a stable photocurrent. In stark contrast, no net current flow was observed in the long-wavelength region (580-1000 nm). Also different from the photocurrents measured under short-wavelength photon illumination was the apparent noise level, that in the long-wavelength region being significantly higher. It is possible that this “noisy” current is due to random collection of photogenerated electrons and holes, both from Si. This may occur because the photogenerated hole transfer through hematite into water is forbidden due to the lack of short-wavelength photons and the large energy barrier to the hematite valence band.

FIG. 11B shows how the photocurrent levels change and the unique transient behaviors. To better present these features, magnified views of currents in two spectral regions, 320-400 nm and 700-780 nm are re-plotted as FIG. 11C and FIG. 11D, respectively. In FIG. 11C, it can be seen that the photocurrent first went up with the increasing wavelength. This trend tracks the abundance of photons in the solar spectrum within this region. Further increasing the photon wavelength beyond 440 nm, however, resulted in a lower photocurrent. This may be due to poorer light absorption by hematite in the longer wavelength regions. The net photocurrent was diminished at 580 nm and beyond, as has been previously discussed.

The second feature concerns the obvious transient phenomenon, which is manifested in the form of current spikes when light was switched on and off. To understand the nature of these transient spikes, it is necessary to clarify that the anodic photocurrent of an electrode is a measure of how fast electrons are collected. An anodic spike in the chronoamperometry plot would indicate a surge of electrons and is often explained by charging and discharging effect of trap states. For the hematite/Si nanowire system, these transient spikes may be originated from two sources, trap states at the hematite/electrolyte interface or those at the hematite/Si interface.

FIG. 12A and FIG. 12B present the photocurrent responses of hematite on FTO under stepwise monochromatic chopped illumination in the wavelength range 300-780 nm under applied bias of 1.45 V_RHE. In reference to FIG. 12A, the hematite film is the only contributor to photocurrent, and thus photocurrent is produced only by wavelengths up to 580 nm. The dotted box denotes the area detailed in FIG. 12B, where it can be seen that the transient photocurrent behavior is similar to that of the Si nanowires/hematite devices shown in FIGS. 11A-11D. Photocurrents in FIG. 12B are higher than those of FIGS. 11A-11D because FTO/hematite devices do not have the heterojunction barrier that Si nanowires/hematite devices exhibit.

By performing control experiments on hematite/FTO substrates under similar conditions (FIGS. 11A-11D versus FIGS. 12A-12B) it was concluded that while the transient behaviors under UV and blue illumination may be explained by the nature of the hematite/electrolyte interface, those under red and near-infrared illumination can only be explained by charging and discharging of the hematite/Si interface. That is, when light is switched on, rapid charge separation takes place within Si, electrons being collected to produce an anodic photocurrent and holes moving to the Si/hematite interface to be trapped there. If these trapped holes recombine with photogenerated electrons from hematite, which would take place under dual-illumination conditions as that shown in FIGS. 1A-1B, a steady-state photocurrent will be measured; in the absence of effective photocharge generation within hematite, however, the initial photocurrent would quickly decay to the base level (zero net current), resulting in a transient spike. When light is switched off, annihilation of the initially trapped holes requires back-electron transfer into Si, leading to a cathodic photocurrent spike. This feature can be clearly observed in FIG. 11D.

Accordingly, the interfaces between Si and hematite are preferable sites for photogenerated holes (from Si) and electrons (from hematite) to recombine. Such a recombination enables forward current flow and is critically important for the realization of the dual-absorber-based “Z-scheme” as shown in FIG. 5A.

FIG. 13A presents representative photocurrent-voltage plots for Fe₂O₃/Si nanowire (NW) and Fe₂O₃/FTO devices. The top-performing Fe₂O₃/Si NW device (red) and three other typical preparations (pink) show similar photocurrent onset potentials as well as peak photocurrent densities. Three Fe₂O₃/FTO devices (black and grey) show the range of behaviors with sample variation. As can be seen from the plots, the Fe₂O₃/Si NW samples showed less sample variation than the Fe₂O₃/FTO samples. FIG. 13B presents representative plots for Fe₂O₃/Si NW devices of different Si doping characteristics. The details of the Si doping are tabulated in Table I below.

TABLE 1
Silicon doping characteristics
	FIG. S6b		Resistivity	Approx. doping
	color	Dopant	(Ω cm)	level (cm−3)
	Red	P	5-15	1014
	Green	P	0.04-0.12	1016
	Blue	Sb	0.007-0.020	1018
	Pink	B	8-20	1015

In an embodiment, a dual absorber electrode includes two absorber materials of different bandgap coupled together and stacked together so to allow the light to pass through the largest-bandgap absorber first. In some embodiments, the dual absorber electrodes of the present embodiment may be utilized in an electrolytic cell for water splitting.

In some embodiments, a dual absorber particle includes a first absorber material and a second absorber material having a larger band gap than the first absorber material, and a suitable second electrode material, such as platinum or another catalyst metal, in electrical contact with the first absorber material.

In some embodiments, a dual absorber electrode includes a first absorber material having a first bandgap; and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber.

In some embodiments, a device for splitting water to generate hydrogen and oxygen includes a first compartment having a first electrode, the electrode comprising a first absorber material having a first bandgap, and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber; a second compartment having a second electrode for catalyzing hydrogen generation; and a semi-permeable membrane separating the first compartment and the second compartment.

In some embodiments, a dual absorber electrode that includes a first electrode comprising first absorber material and a second absorber material having a larger band gap than the first absorber material; and a second electrode for catalyzing hydrogen generation in electrical contact with the first absorber material.

In some embodiments, a method of fabricating an electrode that includes obtaining a silicon nanostructure; forming by vapor deposition a layer of hematite on a surface of the silicon nanostructure by exposing the silicon nanostructure to gas precursors of hematite; and repeating the vapor-depositing to form additional layers of hematite on the surface of the silicon nanostructure

All patents, patent applications, and published references cited herein are hereby incorporated by reference in their entirety. It will be appreciated that several of the above-disclosed and other features and functions, or alternatives thereof, may be desirably combined into many other different systems or applications. Various presently unforeseen or unanticipated alternatives, modifications, variations, or improvements therein may be subsequently made by those skilled in the art which are also intended to be encompassed by the following claims.

Claims

1. A dual absorber electrode comprising a first absorber material having a first bandgap; and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber.

2. The electrode of claim 1 wherein the first absorber material is a silicon nanostructure and the second absorber material is a hematite thin film deposited on the surface of the silicon nanostructure.

3. The electrode of claim 1 further comprising a uniform interface between the first absorber material and the second absorber material with low defect densities and low impurity levels.

4. The electrode of claim 1 wherein the second absorber material is deposited over the first absorber material in a conformal fashion.

5. The electrode of claim 1 wherein the second absorber material is a thin film conformally deposited on the surface of the first absorber material.

6. The electrode of claim 1 wherein the first absorber material is a photovoltaic junction.

7. The electrode of claim 1 wherein the second absorber material is a photovoltaic junction.

8. The electrode of claim 1 further comprising a catalyzing member in electrical contact with the first absorber material for catalyzing hydrogen generation.

9. The electrode of claim 8 wherein the first absorber material is a silicon nanostructure and the second absorber material is a hematite thin film deposited on the surface of the silicon nanostructure.

10. The electrode of claim 8 wherein a uniform interface is formed between the first absorber material and the second absorber material with low defect densities and low impurity levels.

11. The electrode of claim 8 wherein the second absorber material is a photovoltaic junction.

12. A device for splitting water to generate hydrogen and oxygen comprising: a first compartment having a first electrode, the electrode comprising a first absorber material having a first bandgap, and a second absorber material deposited on a surface of the first absorber material, the second absorber material having a second bandgap larger than the first bandgap of the first absorber;a second compartment having a second electrode for catalyzing hydrogen generation; anda semi-permeable membrane separating the first compartment and the second compartment.

13. The device of claim 12 wherein the first absorber material is a silicon nanostructure and the second absorber material is a hematite thin film deposited on the surface of the silicon nanostructure.

14. The device of claim 12 wherein a uniform interface is formed between the first absorber material and the second absorber material with low defect densities and low impurity levels.

15. The device of claim 12 wherein the second absorber material is deposited over the first absorber material in a conformal fashion.

16. The device of claim 12 wherein the second absorber material is a thin film conformally deposited on the surface of the first absorber material.

17. The device of claim 12 wherein the first absorber material is a photovoltaic junction.

18. The device of claim 12 wherein the second absorber material is a photovoltaic junction.

19. A method of fabricating an electrode comprising: obtaining a silicon nanostructure;forming by vapor deposition a layer of hematite on a surface of the silicon nanostructure by exposing the silicon nanostructure to gas precursors of hematite; andrepeating the vapor-depositing to form additional layers of hematite on the surface of the silicon nanostructure.

20. The method of claim 19 wherein the silicon nanostructure is a photovoltaic junction.