Device and method for manufacturing same

Abstract

A method for manufacturing a device is provided that can stabilize the device surface during operation. A method of manufacturing a device includes providing a substrate for the device and forming a structure for the device, the structure being supported by the substrate and having a semiconductor composition and including a surface having nitrogen present thereon, the method further including introducing oxygen to the surface to form a stabilization layer on the surface.

Description

[Technical field]

[0001] [CROSS REFERENCE TO RELATED APPLICATIONS] This application claims the benefit of U.S. Provisional Application No. 63 / 154,428, entitled "Oxynitride-Based Surface Passivation," filed February 26, 2021, the entire disclosure of which is expressly incorporated herein by reference. [STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT] This invention was made with Government support under Contract No. DE-EE0008086 awarded by the Department of Energy and under Contract No. W911NF2110337 awarded by the U.S. Army Research Laboratory. The Government has certain rights in this invention. [Technical field to which the invention pertains] The present disclosure relates generally to the stabilization and activation of device surfaces during water splitting and other reactions, as well as during other types of device operation. [Background technology]

[0002] Photoelectrochemical (PEC) water splitting mimics plants to generate sustainable clean fuels using two of the most abundant resources on Earth: sunlight and water. One common scheme to artificially achieve this solar-to-chemical energy conversion is to monolithically integrate a semiconductor material with an electrocatalyst in a photoelectrode. In this approach, the semiconductor efficiently harvests solar energy, while the electrocatalyst, sometimes also acting as a protective layer, reduces overpotentials, mediates charge carrier transfer, and provides active sites for chemical reactions at the solid / liquid interface.

[0003] Unfortunately, efficient light absorbers are not durable (e.g., Si, III-V), while durable materials show poor efficiency (e.g., TiO2, SrTiO3). For example, both Si and III-V compound semiconductors suffer from poor stability due to chemical and photochemical corrosion. Compared with PV-EL (photovoltaic electrolyser) devices, the light absorber in PEC devices is often in direct contact with the electrolyte, resulting in more rapid degradation. Semiconductor corrosion is influenced by many factors, including the intensity of light irradiation, bias conditions, catalyst, surface passivation, semiconductor electronic band structure, electrolyte composition, and semiconductor / electrolyte and catalyst / electrolyte interfaces. These factors can potentially be addressed by exploring thermodynamic and kinetic protection schemes.

[0004] Recent efforts have been directed towards providing stable coatings to obtain stable photoelectrodes. These coatings are usually a combination of a charge transport layer combined with a catalyst, thus promoting efficient charge collection and separation and preventing corrosion. Various oxide, nitride, selenide, and sulfide materials have been investigated for the protection of Si-based photocathodes in the presence of catalysts for the hydrogen evolution reaction (HER). Among these, gallium nitride (GaN) is an interesting option due to the chemically stable nature of its N-rich surface. GaN is suitable as an efficient protective layer for silicon photocathodes due to the nearly ideal band alignment of the two materials to provide optimal electron transport.

[0005] Other protection schemes have been developed to enhance the stability of photoelectrodes. Kinetic protection for a given photoelectrode is possible by using a synergistic combination of a stable surface protection layer and a highly active cocatalyst. To further improve the stability of the device, other photoelectrodes have used relatively thick metal oxides, such as TiO2, Al2O3, and IrOx, as passivation layers in addition to the use of a suitable cocatalyst. Although the stability of these devices has improved, one major problem is the loss of photocurrent due to poor charge transfer and, in some cases, unwanted light absorption by the protection layer. Summary of the Invention

[0006] According to one aspect of the disclosure, a method of manufacturing a device comprises providing a substrate for the device and forming a structure for the device. The structure includes a surface supported by the substrate, having a semiconductor composition and having nitrogen present therein. The method further includes introducing oxygen to the surface to form a stabilization layer on the surface. The introducing of oxygen is performed such that the stabilization layer includes a uniform distribution of oxynitride material.

[0007] According to another aspect of the present disclosure, an apparatus includes a substrate and a structure having a semiconductor composition. The structure includes a surface supported by the substrate and having nitrogen present on the surface. The apparatus further includes a stabilization layer disposed on the surface of the structure. The stabilization layer includes a uniform distribution of an oxynitride material.

[0008] According to another aspect of the present disclosure, an apparatus includes a substrate and a structure having a semiconductor composition. The structure is supported by the substrate and includes a surface. The apparatus further includes a stabilization layer disposed on the surface of the structure. The stabilization layer includes an oxynitride material. The oxynitride material of the stabilization layer is continuously distributed across the surface of the structure.

[0009] According to another aspect of the present disclosure, an apparatus includes a substrate and a structure having a semiconductor composition. The structure is supported by the substrate and includes a non-bipolar sidewall surface. The apparatus further includes a passivation structure disposed along the non-bipolar sidewall surface of the structure, and a stabilization layer disposed between the structure and the passivation structure. The stabilization layer includes an oxynitride material.

[0010] According to another aspect of the disclosure, a method of manufacturing a device comprises providing a substrate for the device and forming a structure for the device. The structure is supported by the substrate, has a semiconductor composition, and includes a surface. The method further comprises forming a passivation layer on the surface. The passivation layer includes an oxynitride material. The method further comprises depositing an insulating structure adjacent to the structure such that the passivation layer is disposed between the surface of the structure and the insulating structure.

[0011] In connection with any one of the aforementioned aspects, the methods, apparatus, and / or systems described herein may alternatively or additionally include any combination of one or more of the following aspects or features.

[0012] The stabilization layer is configured as an activation layer The stabilization layer is configured as a passivation layer

[0013] Introducing oxygen includes performing an oxidation reaction to form a stabilization layer. Introducing oxygen includes performing an electrochemical procedure. The electrochemical procedure is performed for a period of time on the order of minutes. The electrochemical procedure includes performing a water splitting reaction in which the device is immersed in water. Introducing oxygen includes annealing the surface. Introducing oxygen includes depositing an oxygen-containing material on the surface. The oxygen-containing material includes an oxide. The oxide is aluminum oxide. The oxide is hafnium oxide.

[0014] Introducing oxygen can include treating the surface prior to depositing the oxygen-containing material. Introducing oxygen can further include annealing the surface after depositing the oxygen-containing material. The oxygen-containing material can include a ferroelectric material.

[0015] Forming the structure includes forming an array of conductive protrusions supported by and extending outwardly from the substrate, the array of conductive protrusions comprising the structure. Forming the array of conductive protrusions includes performing a molecular beam epitaxy (MBE) growth procedure such that each conductive protrusion in the array of conductive protrusions comprises a respective nanowire. The MBE growth procedure is performed under nitrogen-rich conditions such that a sidewall of each conductive protrusion in the array of conductive protrusions is nitrogen-terminated.

[0016] The substrate includes silicon. The semiconductor composition of the structure includes an oxynitride material of GaO x N 1-x The gallium nitride includes gallium nitride such that

[0017] The stabilization layer has a thickness ranging from about one monolayer to several monolayers. The surface is oriented along a non-polar face of the semiconductor composition. The surface is oriented along a polar face of the semiconductor composition. The semiconductor composition is configured to effect partial oxygen replacement of nitrogen by performing an oxidation reaction. The semiconductor composition has a wurtzite crystal structure. The nitrogen at the surface is arranged in a compound semiconductor array of the semiconductor composition.

[0018] The stabilization layer is configured as a passivation layer. The structure is configured as a transistor channel. The device further includes a gate oxide layer. The stabilization layer is disposed between the transistor channel and the gate oxide layer and stabilizes an interface between the transistor channel and the gate oxide layer. The transistor channel includes gallium nitride. The gate oxide layer includes aluminum oxide. The gate layer includes hafnium oxide.

[0019] The structure includes a stack of semiconductor layers configured for light emission. The device further includes a metal contact layer adjacent to the stack of semiconductor layers, with a stabilization layer disposed along an interface between the metal contact layer and the stack of semiconductor layers. The stack of semiconductor layers includes an active layer. The stabilization layer is disposed along a sidewall of the active layer. The stabilization layer is configured as an activation layer. The stabilization layer is configured to define catalytic sites along a surface of the structure.

[0020] The surface is oriented along a non-polar face of the semiconductor composition. The surface is oriented along a polar face of the semiconductor composition. The stabilization layer is comprised of partial oxygen substitutions of nitrogen. The semiconductor composition has a wurtzite crystal structure. The nitrogen at the surface is disposed in a compound semiconductor array of the semiconductor composition. The surface includes a sidewall. The stabilization layer is disposed along the sidewall.

[0021] The stabilization layer has a thickness ranging from sub-nanometer thickness to several nanometers thick. The substrate comprises silicon. The semiconductor composition of the structure is such that the oxynitride material is GaO x N 1-x The surface does not include catalytic nanoparticles.

[0022] The device further includes an array of conductive protrusions supported by and extending outwardly from the substrate, the array of conductive protrusions including a structure, each conductive protrusion in the array of conductive protrusions having a surface protected by an oxynitride layer.

[0023] The electrochemical system includes a working electrode configured according to the device described herein. The surface is a non-polar surface of the structure. The non-polar surface is oriented along an m-plane of the lattice of the structure. The non-polar surface is a sidewall of the structure.

[0024] The device further includes a passivation cap disposed on the polar surface of the structure. The polar surface is oriented along a c-plane of the lattice of the structure. The passivation cap includes a nitride material or an oxide material. The device further includes a passivation structure partially covering the non-polar surface. The passivation structure includes parylene. The passivation structure includes a nitride material or an oxide material.

[0025] The structure is configured as a nanowire. The device further includes an electrode supported by the structure. The stabilization layer is disposed between the structure and the electrode. The semiconductor composition includes nitrogen such that nitrogen is present at the surface. The semiconductor composition includes oxygen such that oxygen is present at the surface. The surface is a polar surface of the structure. The structure is configured as a transistor channel.

[0026] The device further includes a gate oxide layer. The stabilization layer is disposed between the transistor channel and the gate oxide layer to stabilize an interface between the transistor channel and the gate oxide layer. The surface is oriented along a c-plane of the lattice of the structure. The semiconductor composition includes nitrogen such that nitrogen is present at the surface. The semiconductor composition includes oxygen such that oxygen is present at the surface.

[0027] The device further includes a passivation structure disposed along a sidewall of the structure and oriented along an m-plane of the lattice of the structure, and a further stabilization layer between the passivation structure and the sidewall of the structure. The non-bipolar sidewall surface is an m-plane of the lattice of the structure. The passivation structure partially covers the non-bipolar sidewall surface. The stabilization layer covers the entirety of the non-bipolar sidewall surface, and the passivation layer passivates the entirety of the non-bipolar sidewall surface. The surface is a non-polar surface of the structure. The surface is a polar surface of the structure.

[0028] Depositing the insulating structure includes covering a c-plane surface of the structure. Depositing the insulating structure includes partially covering an m-plane surface of the structure. Depositing the insulating structure is performed after forming a stabilization layer. Forming the stabilization layer includes depositing an oxygen-containing material on the surface and treating the surface prior to depositing the oxygen-containing material. Forming the stabilization layer further includes performing an anneal to form an oxynitride material from oxygen of the oxygen-containing material. Forming the stabilization layer includes depositing a nitrogen-containing material on the surface and performing an anneal to form an oxynitride material from nitrogen of the nitrogen-containing material.

[0029] For a more complete understanding of the present disclosure, reference should be made to the following detailed description and the accompanying drawings, in which like elements are identified with like reference numerals, and in which: [Brief description of the drawings]

[0030] [Figure 1] FIG. 1 is a schematic and block diagram of an electrochemical system having a photocathode with an array of nanostructures having a stabilizing layer for stable activation of hydrogen generation by water splitting, according to one embodiment. [Diagram 2] 1 is a flow diagram of a method for fabricating a photocathode having a stabilization layer, according to an example. [Diagram 3] 1 is a schematic cross-sectional view of an electronic device having a stabilization layer according to one embodiment. [Figure 4] 1 shows a graphical plot of photocurrent density for several photocathodes, including photocathodes having an oxynitride layer for stabilizing surface activation, according to some embodiments. [Diagram 5] 1 shows topographical maps of a photocathode without and with an oxynitride layer for stabilizing surface activation, along with graphical plots of the photocathode surface profile and photocurrent, according to an example. [Figure 6]Illustrated are scanning transmission electron microscopy (STEM) images and electron energy loss spectroscopy (EELS) mapping of a photocathode without an oxynitride layer and a photocathode with an oxynitride layer for stabilizing surface activation, according to one embodiment, along with a graphical plot of the chemical analysis of the photocathode. [Figure 7] 1 shows a schematic diagram of the chemical makeup of a gallium nitride surface along with a graphical plot of the energy of formation of a gallium nitride surface with varying degrees of substitution of nitrogen by oxygen. [Figure 8] A schematic diagram of the wurtzite structure of gallium nitride and a schematic diagram of a plan view of the wurtzite structure of gallium nitride are shown together with atomic force microscope (AFM) and high resolution transmission electron microscope (HRTEM) images of the gallium nitride surface. [Figure 9] FIG. 1 shows a schematic of the chemical makeup of gallium nitride and gallium oxynitride surfaces for density functional theory (DFT) simulations. [Figure 10] 1 is a schematic cross-sectional view of a light-emitting device having a stabilization layer according to an example. [Figure 11] 4 is a schematic cross-sectional view of a light emitting device having a stabilization layer according to another embodiment. [Figure 12] FIG. 1 shows a schematic perspective view of an exemplary device having nanostructures with either bipolar c-plane or non-bipolar m-plane surfaces exposed to an electrolyte, scanning electron microscope images of the nanostructures, and a schematic of the wurtzite structure of the GaN structure to show in more detail the bipolar c-plane top surface and the non-bipolar m-plane sidewalls. [Figure 13] (a) Linear sweep voltammetry curves for devices with bipolar c-face or non-bipolar m-face in 0.5 M H2SO4 at a daily AM1.5G irradiance and (b) graphical plot of photocurrent and hydrogen evolution for a device with an exposed non-bipolar m-face, showing a faradaic efficiency of nearly 100%. [Figure 14] Nyquist plots of an exemplary device having a bipolar c-plane or a non-dipole m-plane are shown along with graphical plots of Mott-Schottky measurements of the exemplary device. [Figure 15]1 shows a graphical plot of the linear scan voltammograms of exemplary devices having example parylene coated GaN nanowires under one day of light irradiation, along with schematic diagrams of two of the exemplary devices. [Figure 16] A schematic of the density functional theory (DFT) construction of gallium oxynitride formation by oxygen substitution on the c- and m-planes and a schematic of the band bending induced by oxygen substitution are shown. [Figure 17A] FIG. 2 is a schematic diagram of a nanostructure having a passivated c-face, according to an example. [Figure 17B] FIG. 2 is a schematic diagram of a nanostructure having a passivated c-surface and a partially passivated m-surface, according to an example. [Figure 17C] FIG. 1 is a schematic diagram of a device having planar nanostructures with a stabilized surface according to an example. [Figure 18] 1 is a flow diagram of a method for manufacturing a device having a stabilization layer, according to various embodiments. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] The disclosed device, system, and method embodiments may take a variety of forms, and while specific embodiments are illustrated in the drawings and described below, it is understood that the disclosure is intended to be illustrative, i.e., the disclosure is not intended to limit the invention to the specific embodiments described and illustrated herein.

[0032] Photoelectrochemical, photocatalytic, electronic, and other systems and devices are described. In some cases, the disclosed systems and devices (e.g., photoelectrodes) can be configured for water splitting and other chemical reactions. Methods for manufacturing devices (e.g., photoelectrodes or photocatalytic devices) are also described. The photoelectrode or other device includes a stabilizing layer (e.g., a stabilizing activation layer) disposed on one or more surfaces thereof.

[0033] The stabilization layer includes an oxynitride material. As described herein, the disclosed methods include incorporating oxygen into the structure surface (e.g., oxidizing the structure surface) to form an oxynitride material (e.g., a continuous or other uniform distribution of the oxynitride material). For example, an oxidation reaction can be performed that is configured to form a continuous or other uniform distribution of the oxynitride material. In some cases, the uniformity of the oxynitride distribution can relate to the comprehensive formation of the oxynitride material at each nitrogen site of the surface. Alternative or additional methods in which the distribution is uniform are provided below.

[0034] The composition of the oxynitride material may vary. For example, the ratio of nitrogen to oxygen may vary. Thus, in some cases, the oxygen composition may change (e.g., decrease) with increasing distance from the surface. Oxynitride materials may be O x N 1-x For example, in a device having a GaN device structure, the oxynitride material may be GaO x N 1-x Thus, the stabilization layer in such a case may comprise gallium oxynitride (gallium oxynitride) in combination with a very thin layer (e.g., 1 nm or less) of gallium and oxide (e.g., on top of the gallium oxynitride), GaO X N 1-X The regions of decrease in oxygen concentration with increasing distance from the surface. Thus, reference to an oxynitride material of the stabilization layer can include materials having various levels of oxygen content. The oxygen content level can vary depending on several factors, including, for example, surface orientation, surrounding environment (e.g., electrolyte, metal layer, passivation layer, etc.), and / or processing conditions. In either case, the oxynitride material of the stabilization layer provides surface passivation and device performance enhancement, as described below. The oxynitride material may be altered in alternative or additional ways. For example, the nitrogen and oxygen may be disordered. In some cases, the arrangement of the nitrogen and oxygen may be interchanged or otherwise altered.

[0035] In some cases, the oxynitride layer is configured as an activation layer. For example, the oxynitride layer can activate the device surface by establishing or otherwise defining catalytic sites along the device surface. Alternatively or additionally, the oxynitride layer can activate the device surface by establishing or otherwise defining sites along the device surface for efficient charge transfer. Thus, the resistance present at an interface, such as a metal-semiconductor interface in a light-emitting device, can be reduced. Thus, the activation aspect of the oxynitride layer can improve the efficiency or performance of the disclosed devices, as described in connection with some embodiments herein.

[0036] In addition to providing such an activation function, the stabilization layer can provide protection or other stabilization of the device surface. Thus, the structure in which the stabilization layer is disposed can be more durable. For example, in the electrochemical case, the stabilization layer can provide protection against corrosion or other degradation. Such protection or other stabilization is achieved without adversely affecting device performance. In fact, on the contrary, the oxynitride layer provides such protection or other stabilization while unexpectedly improving device performance.

[0037] In some cases, the stabilization layer is configured as a passivation layer rather than an activation layer. The stabilization layer may be used to stabilize an interface between a device structure, such as a transistor channel, and an adjacent structure, such as a transistor gate oxide. The stabilization layer may be disposed between such structures to stabilize the interface therebetween. In some cases, the stabilization layer may reduce interface traps at the interface of an electronic device. Such reduction may be useful during the operation of transistors and a wide variety of other electronic devices. The stabilization layer may provide an alternative or additional stabilization function, including, for example, making a conventionally unstable interface (e.g., a high dielectric constant interface, such as a III-nitride interface) more chemically stable. Such increased stability may be useful in the context of a wide range of electronic devices, such as high electron mobility transistors (HEMTs), metal oxide semiconductor field effect transistors (MOSFETs), fin field effect transistors (FinFETs), and vertical transistors. Additionally, further stabilization may be provided in the context of light emitting devices via a passivation layer configured to reduce current leakage and / or surface recombination.

[0038] Water splitting and other chemical reactions may be solar-powered (e.g., solar water splitting). Thus, the disclosed devices, systems, and methods may in some cases be considered to perform artificial photosynthesis. In some cases, the disclosed devices and systems include an array of conductive protrusions, such as nanowires or other nanostructures. Each nanostructure or other conductive protrusion may establish a structure to support or provide a catalyst for water splitting or other chemical reactions. In other cases, the disclosed devices may be comprised of other types of structures, including planar structures. For example, the functionality of the device may instead be provided by several planar structures supported by a substrate.

[0039] The disclosed methods, apparatus, and systems include structures having a surface that includes nitrogen. For example, in the case of a transistor, the structure may be a transistor channel that is composed of or otherwise includes a compound semiconductor that establishes a nitrogen-based surface. In the case of a photoelectrode that includes an array of nanostructures, each nanostructure may be composed of or otherwise include a compound semiconductor that establishes a nitrogen-based surface. Alternative or additional surface protection may be provided by nitrogen-based aspects of the surface. For example, the nanostructure may be composed of GaN in a configuration in which the surface is nitrogen-terminated. Such nitrogen termination may provide protection or other stabilization of the structure. Alternatively or additionally, the disclosed photoelectrode or other device may have one or more surfaces on which nitrogen or other nitrogen-based layers are disposed.

[0040] Although described in the context of nanowire arrays, the surface passivation and activation schemes described herein may be applied to a variety of different device structures. A variety of planar catalyst surfaces may be modified to include nitrogen. For example, a nitrogen layer (or other nitrogen-based layer) may be deposited or otherwise applied to a variety of different electrode surfaces.

[0041] Photoelectrochemical, photocatalytic, and other water splitting provided by the disclosed devices and systems can include solar-hydrogen conversion. The disclosed devices and systems provide improved efficiency of photoelectrochemical water splitting and / or other water splitting (e.g., photocatalytic water splitting). The disclosed devices and systems can include one or more other aspects directed to improved efficiency. For example, the structure can include a dual junction configuration, such as a multi-band InGaN nanowire array configured to provide a second junction.

[0042] Although described in the context of photoelectrochemical water splitting, the disclosed devices and systems may be used in other chemical reactions and applications. For example, the disclosed devices and systems may be useful in conjunction with various types of photocatalytic and / or other systems and / or in conjunction with other reactions including, for example, nitrogen reduction to ammonia, CO2 reduction to various fuels and other chemicals, and C-H bond activation for the production of various chemicals.

[0043] Although described herein in the context of electrodes having GaN-based nanowire arrays for water splitting, the disclosed devices and systems are not limited to GaN-based nanowire arrays. A wide variety of other types of nanostructures and other conductive protrusions can be used. In some cases, the electrodes of the disclosed systems do not include an array of protrusions, but are instead planar devices. Thus, the nature, structure, composition, properties, shape, and other aspects of the electrodes on which water splitting is performed may vary.

[0044] The surface passivation and activation of the disclosed methods and devices are also not limited to electrochemical or other chemical contexts. Although described herein in the context of field effect transistors, surface passivation can be applied in a variety of electronic contexts. For example, the disclosed devices can be any type of optoelectronic, photoelectronic, or other electronic device. In such cases, the structures formed by the disclosed methods can be or include a variety of other types of device structures, including, for example, structures other than transistor structures. The manner in which the surface of the structure is protected or stabilized by the oxynitride material of the stabilization layer of the disclosed methods and devices can be modified accordingly.

[0045] FIG. 1 shows a system 100 for hydrogen generation by water splitting. System 100 may also be configured for other reactions. System 100 may be configured as an electrochemical system. In this example, electrochemical system 100 is a photoelectrochemical (PEC) system in which sunlight and / or other radiation is used to promote hydrogen generation and water splitting. Water splitting may be assisted or unassisted as described herein. The manner in which PEC system 100 is illuminated may vary. The wavelength and other characteristics of the radiation may vary accordingly.

[0046] The electrochemical system 100 includes one or more electrochemical cells 102. A single electrochemical cell 102 is shown for ease of illustration and description. The electrochemical cells 102 and other components of the electrochemical system 100 are shown generally in FIG. 1 for ease of illustration. The cells 102 include an electrolyte 104. In some cases, CO2 and / or other sources are applied. In some cases, the electrolyte is saturated with CO2. Potassium bicarbonate, KHCO3, can be used as the electrolyte. Additional or alternative electrolytes can be used, as described below. Further details regarding examples of the electrochemical system 100 are provided below.

[0047] In the example of FIG. 1, the electrochemical cell 102 has a three-electrode configuration. The electrochemical cell 102 includes a working electrode 108, a counter electrode 110, and a reference electrode 112, each of which is immersed in the electrolyte 104. The counter electrode 110 may be or may include a metal wire, such as a platinum wire. The reference electrode 112 may be configured as a reversible hydrogen electrode (RHE). The positioning of the reference electrode 112 may vary from the example shown. For example, the reference electrode 112 may be adjacent to the counter electrode 110 in other cases. The configurations of the counter electrode 110 and the reference electrode 112 may vary. For example, the counter electrode 110 may be a hydroxide (2H2O O2+4e - +4H +) occurs. In some cases, the counter electrode 110 is configured as or includes an IrOx electrode. The configuration of the electrochemical cell may vary. For example, in other cases, two-electrode or other configurations may be used.

[0048] Hydrogen evolution occurs at the working electrode 108 as follows: Hydrogen generation: 2H + + 2e - ⇔ H2 To do so, electrons can flow from the counter electrode 110 through a circuit path external to the electrochemical cell 102 to reach the working electrode 108. The working electrode 108 and the counter electrode 110 can therefore be considered as a cathode and an anode, respectively.

[0049] 1, the working and counter electrodes are separated from each other by a membrane 114, such as a proton exchange membrane. The structure, composition, morphology, and other properties of the membrane 114 can vary.

[0050] In some cases, no bias voltage is applied (e.g., in a non-auxiliary system). In the example of FIG. 1, a circuit path includes a voltage source 116 of the electrochemical system 100. The voltage source 116 is configured to apply a bias voltage between the working electrode 108 and the counter electrode 110. The bias voltage can be used to establish a ratio of CO2 reduction to hydrogen (H2) generation, and / or another reaction ratio. The circuit path can include additional or alternative components. For example, the circuit path can include a potentiometer in some cases.

[0051] In this example, the working electrode 108 is configured as a photocathode. Light 118, such as solar radiation, can be incident on the working electrode 108 as shown. Thus, the electrochemical cell 102 can be considered a photoelectrochemical cell and configured as a photoelectrochemical cell. In such a case, illumination of the working electrode 108 can cause charge carriers to be generated in the working electrode 108. Electrons arriving at the surface of the working electrode 108 can then be used for hydrogen generation. The photogenerated electrons can augment the electrons provided via the current path. Alternatively or additionally, the electrons provided via the current path can recombine with photogenerated holes at the backside or other contact. Further details regarding examples of photocathode are provided below.

[0052] The working electrode 108 includes a substrate 120. The substrate 120 of the working electrode 108 can form part of the architecture, scaffolding, or other support structure of the working electrode 108. The substrate 120 can be homogeneous or composite. For example, the substrate 120 can include any number of layers or other components. Thus, the substrate 120 may or may not be monolithic. The substrate 120 may also have a variety of shapes. For example, the substrate 120 may or may not be planar or flat.

[0053] In the embodiment of FIG. 1, the substrate 120 is doped or otherwise configured to provide a junction. Thus, the substrate 120 of the working electrode 108 may be active (functional) with respect to photogeneration of charge carriers. Alternatively or additionally, the substrate 120 is passive (e.g., structural). The substrate 120 may be configured and act as a support structure for the catalytic arrangement of the working electrode 108, as described below. Alternatively or additionally, the substrate 120 may be constructed of or otherwise include a material suitable for growth or other deposition of the catalytic arrangement of the working electrode 108.

[0054] If active or functional, the substrate 120 may include a light absorbing material. The light absorbing material is configured to generate charge carriers upon sunlight or other illumination. The light absorbing material has a band gap such that incident light generates charge carriers (electron-hole pairs) within the substrate. Some or all of the substrate 120 may be configured for photo-generation of electron-hole pairs. To that end, the substrate 120 may include a semiconductor material. In some cases, the substrate 120 is composed of or includes silicon. For example, the substrate 120 may be provided as a silicon wafer.

[0055] Silicon may be doped. In the embodiment of FIG. 1, substrate 120 includes a heavily n-doped layer 122, a moderately or lightly p-doped layer 123, and a heavily p-doped layer 124. The arrangement of layers 122-124 establishes a junction within substrate 120. The doping arrangement may vary. For example, one or more components of substrate 120 may be undoped (intrinsic) or effectively undoped. Substrate 120 may include alternative or additional layers, including, for example, supports or other structural layers. In other cases, substrate 120 does not absorb light.

[0056] The substrate 120 of the working electrode 108 establishes a surface on which several structures are provided. Each structure may be disposed on or supported by the substrate 120. In this example, the structure is configured to provide a catalyst arrangement. In this case, the structure is configured to provide catalytic functionality. In such a case, the structure does not include catalyst nanoparticles. In other cases, the structure is configured as a support structure or scaffold for the catalyst (e.g., catalyst nanoparticles) of the electrode 108. As described below, the structure may include an array of conductive protrusions extending outward from the surface of the substrate 120. In other cases, the structure is not configured as conductive protrusions or does not include conductive protrusions. For example, the structure may include one or more planar structures, such as one or more layers supported by the substrate 120.

[0057] 1, working electrode 100 includes an array of nanostructures 126 (or other conductive protrusions) supported by substrate 120. Each nanostructure 126 is configured to extract charge carriers (e.g., electrons) from substrate 120. Extraction transports the electrons along nanostructure 126 to an external site for use in hydrogen generation. In some cases, each nanostructure 126 is configured as a nanowire.

[0058] Each nanostructure 126 and / or other structures supported by substrate 120 has a semiconductor composition. In some cases, the semiconductor composition is or otherwise includes a semiconductor core 128. For example, core 128 may be composed of or include a III-V nitride semiconductor material, such as gallium nitride (GaN). Additional or alternative semiconductor materials may be used, including, for example, indium gallium nitride (InGaN) and / or other III-V nitride semiconductor materials.

[0059] The core 128 of each nanowire or other nanostructure 126 may be or include a pillar, post, or other elongated structure that extends outward (e.g., upward) from the plane of the substrate 120. The semiconductor nanowires or other nanostructures 126 may be grown or formed as described in U.S. Patent No. 8,563,395, the entire disclosure of which is incorporated herein by reference. The nanostructures 126 may be referred to herein as nanowires, with the understanding that the dimensions, size, shape, composition, and other properties of the nanostructures 126 or other conductive protrusions may vary.

[0060] The semiconductor composition of each nanostructure 126 may or may not be configured to facilitate the reaction(s) supported by the electrochemical system 100. In some cases, the semiconductor composition may be configured for photogeneration of charge carriers. Alternatively or additionally, the semiconductor composition may be configured to act as a catalyst for the reaction. The semiconductor composition may provide other functions, including, for example, protection of the substrate 120. As mentioned above, the semiconductor composition may include GaN and / or InGaN. Further details regarding some examples including GaN are provided below. Additional or alternative semiconductor materials may be used, including, for example, indium nitride, aluminum nitride, boron nitride, aluminum oxide, silicon, and / or alloys thereof.

[0061] The semiconductor composition of each nanostructure 126 may be configured to provide or support surface passivation and / or other protection of the photoelectrode 108. As described herein, nitrogen is present on one or more surfaces 130 of each nanostructure 126. Thus, the surfaces 130 may be considered to include nitrogen. For example, in some cases, the semiconductor composition is terminated with nitrogen. The nitrogen termination or other nitrogen-based aspects of the nanostructures 126 may protect the nanostructures 126 and / or other components of the electrode 108 (e.g., substrate 120) during operation, for example, from corrosion. In other cases, alternative or additional nitrogen-based protection schemes may be used. For example, a layer including nitrogen may be deposited or disposed along the surface of each nanostructure 126 and / or other elements of the electrode 108. As described herein, nitrogen may be present at the surface in a variety of ways. In some cases, the nitrogen may be disposed below the surface (e.g., about 1 nm below), but may be disposed close enough to the surface to be functionally active (e.g., chemically active for the formation of an oxynitride layer). Thus, the terms "at," "present," and "presence" are used herein to include various situations in which nitrogen is functionally present on a surface, thereby not being limited to situations in which nitrogen is located solely on the surface, but including situations in which nitrogen atoms are located along the surface or sufficiently close to the surface to be functional for the activity of the surface (e.g., chemically functional in connection with an oxidation reaction at the surface).

[0062] The nanostructures 126 can facilitate hydrogen evolution and / or another chemical reaction in one or more ways. For example, each nanostructure 126 can be configured to extract charge carriers (e.g., electrons) generated within the substrate 120. The extraction transports the electrons along the surface 130 of the nanostructure 126 to an external site for use in hydrogen evolution and / or another chemical reaction. The composition of the nanostructures 122 may also form an interface well suited for hydrogen evolution and / or another chemical reaction, as described below.

[0063] Each nanostructure 126 may be or may include a pillar, post, or other elongated structure that extends outward (e.g., upward) from the plane of the substrate 120. The dimensions, size, shape, composition, and other properties of the nanostructures 126 may vary. For example, each nanostructure 126 may or may not be elongated, such as a nanowire. Thus, other types of nanostructures from the substrate 120 may be used, such as nanocrystals of various shapes.

[0064] In some cases, the nanostructures 126 are configured to generate electron-hole pairs upon illumination. For example, the nanostructures 122 may be configured to absorb light at a different frequency than other light-absorbing components of the electrode 108. For example, one light-absorbing component, such as the substrate 120, may be configured for absorption in the visible or infrared wavelength range, while another component may be configured to absorb ultraviolet wavelengths of light. In other cases, the nanostructures 126 are the only light-absorbing component of the electrode 108. In still other cases, the substrate 120 is the only light-absorbing component of the electrode 108.

[0065] Each nanostructure 126 may include a layered or segmented array of semiconductor materials. For example, in the example of III-nitrides, layers or segments of the array may have different Group III (e.g., indium and gallium) compositions. One or more layers or segments in the arrangement may be configured for absorption of a respective wavelength range. Other layers or segments may be directed to establishing tunnel junctions. Each nanostructure 126 may also include additional or alternative segments.

[0066] In other cases, layered arrangements of semiconductor materials are also used to establish multi-band structures, such as quadruple band structures. Each layer or segment of the arrangement may have a different semiconductor composition to establish a different band gap. Different band gaps may be useful in relation to absorbing different wavelengths of light. Further details regarding the formation and morphology of multi-band structures, including, for example, triple band structures, are provided in U.S. Pat. No. 9,112,085 ("Highly Efficient Broadband Semiconductor Nanowire Devices") and U.S. Pat. No. 9,240,516 ("Highly Efficient Broadband Semiconductor Nanowire Devices"), the disclosures of which are incorporated by reference in their entirety. Other layered or non-layered arrangements may be used for the semiconductor composition of the disclosed device structures.

[0067] The semiconductor composition of each nanowire 126 may be configured to improve the efficiency of water splitting in additional ways. For example, in some cases, the semiconductor composition of each nanowire 126 may include doping to promote separation and extraction of charge carriers, as well as doping to promote the establishment of a photochemical diode. For example, the dopant concentration of the semiconductor composition may vary laterally.

[0068] In embodiments involving III-nitride compositions, the dopant may be or include magnesium. Further details regarding the manner in which magnesium doping promotes separation and extraction of charge carriers are described in U.S. Patent No. 10,576,447 ("Method and System for Photochemical Water Splitting"), the entire disclosure of which is incorporated by reference. Depending on the semiconductor light absorber selected, additional or alternative dopant materials may be used, including, for example, silicon, carbon, and beryllium.

[0069] In some cases, the working electrode 108 can further include catalytic nanoparticles disposed on the array of nanostructures 126. For example, the nanoparticles can be distributed across or along the exterior surface 130 (e.g., sidewall) of each nanowire 126. In some cases, the nanoparticles are configured to promote or facilitate the proton reduction reaction. For example, each nanoparticle 136 includes a metal such as platinum. Other metals or materials can be used, including alloys, oxides, and / or other metals or combinations of metals. Further details regarding the formation, morphology, functionality, and other properties of the nanoparticles in combination with the nanowire arrays are described in one or more of the U.S. patents referenced above. Further details regarding the distribution of the nanoparticles are provided below in connection with several examples.

[0070] The electrode 108 also includes a stabilization layer 132 disposed on the surface 130 of each nanostructure 126. In this case, the stabilization layer 132 is configured to provide both a stabilization function and an activation function for the electrode 108. Thus, the stabilization layer 132 is configured as an activation layer or a stabilized activation layer. The term "activation" is used to distinguish an activation layer from layers configured to act only as passivation or other protective layers, as well as other non-activating layers. In this electrode example, the stabilization function may include protection against corrosion and / or other passivation or protection. The activation function may include defining or otherwise establishing catalytic sites along the surface 130 for the electrode 108. The nature of the stabilization and activation functions may vary in the context of other devices.

[0071] The stabilization layer 132 may be composed of or otherwise include a uniform distribution of oxynitride material. For example, if the nanostructure 126 has a GaN core, the stabilization layer 132 includes gallium oxynitride (gallium oxynitride). Thus, the composition of the oxynitride layer may vary according to the semiconductor composition of the nanostructure 126. Thus, alternative or additional oxynitride materials may be used, including, for example, aluminum oxynitride, indium oxynitride, strontium gallium oxynitride, zinc oxynitride, titanium oxynitride, chromium oxynitride, and silicon oxynitride.

[0072] The distribution of the oxynitride material may be made uniform in a variety of ways. For example, the uniformity may establish or relate to a thickness of the stabilization layer 132. Alternatively or additionally, the uniformity may establish or relate to a conformal nature of the stabilization layer 132. Alternatively or additionally, the uniformity may relate to a consistency, regularity, or smoothness of a surface of the stabilization layer 132. For example, surface roughness may be reduced as a result of the uniform distribution of the oxynitride material. In some cases, the oxynitride material of the stabilization layer 132 may be continuously distributed across the surface 130 (e.g., continuously covering the entirety of the surface 130).

[0073] Alternatively or additionally, uniformity may relate to the presence of nitrogen atoms at surface 130. For example, if nitrogen atoms are randomly distributed on or at a surface, uniformity refers to essentially universal or comprehensive (as opposed to localized or restricted) formation of oxynitride sites around those nitrogen atoms. Thus, the phrase "uniform distribution" is used herein to include both continuous and discontinuous cases of stabilization layer 132. For example, in some cases, adjacent nitrogen atoms may be sufficiently spaced apart to establish discontinuities in stabilization layer 132. Thus, stabilization layer 132 (and thus oxynitride material) may not be continuous at surface 130 in some cases, but nevertheless still have a uniform distribution of oxynitride material relative to nitrogen sites at surface 130. Alternatively or additionally, surface 130 (or sections of a single surface 130) of nanostructure 126 on which stabilization layer 132 is disposed may be spaced apart from one another, thereby establishing separate sections of stabilization layer 132. As shown in these examples, the stabilization layer 132 can be continuous or discontinuous, despite the uniform nature of the distribution of the oxynitride material.

[0074] The terms "presence" and "present at" are not limited to situations where the nitrogen atoms, stabilization layer, and / or oxynitride material are disposed on a surface. The terms are instead used to include situations where the nitrogen atoms, stabilization layer, and / or oxynitride material are below the surface but close enough to the surface to function for activity at the surface. Thus, the nitrogen atoms, stabilization layer, and / or oxynitride material may be disposed slightly below the surface (e.g., about 1 nm below) in some cases, but still close enough to the surface for the formation of the oxynitride material and the resulting stabilization of the surface.

[0075] The formation and / or presence of oxynitrides can be derived from the nitrogen at the surface 130. In some cases, the nitrogen at the surface 130 is disposed in a compound semiconductor arrangement of the semiconductor composition. For example, in GaN and other examples, the semiconductor composition has a wurtzite crystal structure. In some cases, the non-polar faces of the wurtzite crystal structure may be nitrogen-terminated. The stabilization layer 132 may comprise or otherwise be configured with partial oxygen substitutions of nitrogen along the non-polar faces (and / or other faces in other cases). Thus, the surface 130 may correspond to a surface of the nanostructure 126 that is oriented along a non-polar face of the semiconductor composition. In other cases (e.g., in the transistor example), the surface 130 is oriented along a polar face of the semiconductor composition.

[0076] The semiconductor composition and / or crystal orientation at the surface 130 of the nanostructures 126 may selectively determine where the stabilization layer 132 is disposed on each nanostructure 126. Thus, the stabilization layer 132 may cover all or none of the surfaces of the nanostructures 126. In the example of FIG. 1, the surface 130 of each nanostructure 126 corresponds to a sidewall 134. The nanostructures 126 also include a top or upper surface 136. In this example, the stabilization layer 132 is disposed on the sidewall 134 but not on the top or upper surface 136. The sidewall 134 is nitrogen-terminated while the upper surface 136 is not, thereby resulting in selective or partial coverage of the nanostructures 126. The surface 130 that is covered by the stabilization layer 132 may vary according to the crystal orientation of the semiconductor composition of the nanostructures 126.

[0077] The stabilization layer 132 may be thin enough to avoid deleterious effects on charge transfer or other activities along or at the surface 130. In some cases, the stabilization layer 132 has a thickness of about 1 nm. In these and other cases, the stabilization layer 132 may conformally cover the nanostructures 126, including any nanoparticles disposed thereon. The thickness may vary, including, for example, on the order of 1 nm (e.g., up to about 3 nm) or thicknesses less than 1 nm in other examples. For example, the thickness may be one monolayer or on the order of one monolayer (e.g., a pair or several monolayers). Thus, the thickness may vary from about 0.3 nm to about 3 nm. The thickness may not be exactly the same throughout the stabilization layer 132.

[0078] The thin nature of the stabilization layer 132 allows it to provide a stabilization function without adversely affecting charge carrier transport and other catalysis of hydrogen evolution and / or other reactions occurring at the electrode 108. Further details regarding the functionality and other properties of the stabilization layer 132 are provided below in connection with several examples.

[0079] In the example of Figure 1, surface 130 does not include catalytic nanoparticles. In other cases, multiple nanoparticles may be distributed across surface 130. The nanoparticles may have a composition and / or be configured to act as catalysts for electrochemical or other reactions. The nanoparticles may in other cases have additional or alternative purposes including, for example, charge transfer.

[0080] Nanostructures 126 and stabilization layer 132 are not shown to scale in the schematic diagram of Figure 1. The shape of nanostructures 126 may vary from the nanowire example shown. Further details regarding nanostructures 126 and stabilization layer 132, including their manufacture or formation, are provided below.

[0081] Nanowire arrays may be fabricated on a substrate (e.g., a silicon substrate) via nanostructure engineering. In one example, molecular beam epitaxial (MBE) growth of the nanowires is followed by photodeposition of nanoparticles. The photodeposition of nanoparticles may be configured to selectively deposit nanoparticles on each side of the nanowire. Further details regarding an exemplary fabrication procedure are provided below, for example, in connection with FIG. 2.

[0082] The nanostructures 126 can facilitate water splitting in one or more ways. For example, one or more aspects of each nanostructure 126 can be configured to extract charge carriers (e.g., electrons) generated within the substrate (e.g., as a result of light absorbed by the substrate 120). The extraction carries the charge carriers to external sites along the nanostructure 126 for use in water splitting or other reactions. For example, the nanostructures 126 can thus form interfaces well suited for hydrogen evolution, CO2 reduction, and / or other reactions.

[0083] 2 illustrates a method 200 of fabricating a photoelectrode or other semiconductor device for photocatalytic water splitting, PEC water splitting, or other photocatalytic reactions, according to one embodiment. Method 200 may be used to fabricate any of the photoelectrodes or other devices described herein, or another device. Method 200 may include additional, fewer, or alternative operations. For example, method 200 may or may not include one or more operations directed to forming a back contact of the device (operation 228).

[0084] Method 200 may begin at operation 202 where a substrate is prepared or otherwise provided. The substrate may be or may be formed from a silicon wafer. In one example, a 2 inch Si wafer was used, although wafers of other (e.g., larger) sizes may be used. Other semiconductors and substrates may be used.

[0085] The substrate may have a planar or non-planar surface. In some cases, operation 202 includes operation 204 in which a wet or other etching procedure is performed to define the surface. For example, the etching procedure may be or include a crystalline etching procedure. In an example of a silicon substrate, the crystalline etching procedure may be a KOH etching procedure. In some cases, the substrate may have a planar or non-planar surface. <100> If the surface has an orientation, the wet etching procedure <111> It can be established that the textured surface includes a pyramidal shape having facets oriented along the surface, although other surface orientations and shapes may be used.

[0086] Operation 202 may include fewer, additional, or alternative operations. For example, in the example of Figure 1, operation 202 includes operation 206 in which the substrate is cleaned (e.g., with acetone and / or methanol) and operation 208 in which oxides are removed (e.g., with 10% hydrofluoric acid).

[0087] Operation 202 may include further operations. For example, operation 202 may include one or more doping procedures to form doped regions or layers and thereby establish junctions as described herein. Alternatively, a substrate may be first provided with a desired dopant concentration profile.

[0088] The method 200 includes an operation 210 in which a number of electrodes or other device structures are grown or formed on a substrate. In some cases, a nanowire or other nanostructure array is grown or formed on the substrate. Each nanowire is formed on a surface of the substrate such that each nanowire extends outwardly from the surface of the substrate. Each nanostructure may have a semiconductor composition as described herein. In some cases, the semiconductor composition is or includes a compound semiconductor material, such as a III-V semiconductor.

[0089] Growth of the nanostructures may be accomplished in operation 212, where a molecular beam epitaxy (MBE) procedure is performed. The substrate may be rotated during the MBE procedure such that each nanostructure is shaped as a cylindrical nanostructure. Thus, each nanostructure may have a circular cross-sectional shape, as opposed to a plate-like or sheet-like nanostructure.

[0090] The MBE procedure may be carried out under nitrogen-rich conditions, which may result in nitrogen-terminated sidewalls and / or other surfaces as described herein.

[0091] In some cases, the MBE procedure may be modified to create an array of layers or segments of individual nanowires. Various parameters may be adjusted to achieve different composition levels of the segments. For example, the substrate temperature may be adjusted in operation 214. Alternatively or additionally, the beam equalization pressure may be adjusted. In some cases, the dopant cell temperature is adjusted to control the doping of the nanowires (e.g., Mg doping).

[0092] In one embodiment, about 735 o Substrate temperature: approx. 2.2 x 10 -7 Ga equivalent pressure (BP) at Torr, approximately 1050 o N-terminated thin GaN nanostructures were grown on Si for 30 min at a Ge cell (n-type dopant) temperature of 1000 C, a nitrogen flow rate of 0.45 standard cubic centimeters per minute (sccm), and a plasma power of 350 W. Growth parameters may vary considerably in other cases.

[0093] In other cases (e.g., non-nanostructured cases), operation 210 may include forming other types of electrodes or other device structures, such as one or more layers. The layers may be configured to catalyze, participate in, enable, or facilitate a reaction. Thus, the layers may be referred to as reaction layers. For example, the reaction layers may be deposited on the substrate in operation 216. The layers may establish a surface configured to catalyze or facilitate water splitting and / or other reactions. The electrode layers may be formed by alternative or additional methods, including, for example, non-selective growth procedures.

[0094] In operation 218, one or more nitrogen-containing layers can be deposited or formed. The nitrogen layer may be directed to protecting the reaction layer and / or other aspects of the nanostructures. The nitrogen-containing layer may alternatively or additionally be directed to support the formation of a stabilization layer. In some cases, operation 218 may include carrying out a reaction to spontaneously form a nitrogen-containing layer. The layer may be comprised of or may include a nitride or oxynitride.

[0095] In some cases, the nitrogen-containing layer may be formed on an electrode or other device structure that is planar (e.g., a structure that includes several planar layers or other planar components). The morphology of the planar structure may vary. Examples of planar structures are Cheng, et al, "Monolithic Photoelectrochemical Device for Direct Water Splitting with 19% Efficiency," ACS Energ. Lett.,3,(8),1795-1800(2018)(8), Young, et al, "Direct solar-to-hydrogen conversion via inverted metamorphic multi-junction semiconductor architectures," Nat. Energ., 2, 17028 (2017)," Khasekev, et al., "A monolithic photovoltaic-photoelectrochemical device for hydrogen production via water splitting, " Sci., 280, (5362), 425-427 (1998)," and "Verlage, et al., "A monolithically integrated integrated, intrinsically safe, 10% efficient, solar-driven water-splitting system. based on active, stable earth-abundant electrocatalysts in and “conjunction with tandem III-V light absorbers protected by amorphous TiO2 films,” Energ. Environ, Sci., 8, (11), 3166-3172 (2015), the entire contents of which are incorporated herein by reference.

[0096] A wide variety of planar and other structures can be protected by the nitrogen-containing layer. For example, the nitrogen-containing layer can be deposited or formed on a variety of surfaces that do not include c-planes or non-polar planes or surfaces.

[0097] In some cases, operation 218 may be integrated with or performed in conjunction with operation 216, as described herein with respect to examples involving GaN or other nanostructures. In other cases involving GaN nanostructures, operations 216 and 218 may be performed separately.

[0098] The method 200 further includes an operation 220 in which oxygen is incorporated into one or more surfaces of the nanostructures or other structures of the device to form a stabilization layer on the one or more surfaces. In some cases, the surface is oxidized. For example, an oxidation reaction can be performed to form the stabilization layer. Alternatively or additionally, oxygen is deposited on the surface. After deposition, in some cases, an annealing procedure can be performed to fully incorporate the oxygen atoms. The stabilization layer may be configured as an activation layer or a passivation layer. The functionality of the stabilization layer may vary depending on the nature of the device structure. As described herein, the oxidation of the surface (or incorporation of oxygen into the surface) may be performed such that the stabilization layer is composed of or otherwise includes a uniform distribution of oxynitride material. For example, the oxidation reaction may be configured such that the stabilization layer includes a uniform distribution. In some cases, oxygen is incorporated into the surface such that the oxynitride material is continuously distributed across the surface (e.g., continuously covering the entirety of the surface).

[0099] In some cases, an oxynitride layer is formed in operation 222, in which an electrochemical procedure is performed. The electrochemical procedure can be performed for a period of time on the order of several hours, for example, about 5 to about 10 hours. The duration of the procedure may be long enough to establish a uniform distribution of the oxynitride material. For example, in other cases, the oxidation may be performed for a period of time on the order of minutes. Further details are provided in connection with some examples below.

[0100] In some cases, the electrochemical procedure includes the electrochemical reaction that the device is configured to perform. In such cases, operation 222 may be configured as an initialization step or procedure for the device. For example, the device may be configured or directed to perform water splitting or other hydrogen generation reactions. The electrochemical procedure may then include performing a water splitting reaction in which the device is immersed in water. The hydrogen generation reaction may be performed for a predetermined period of time to form an oxynitride layer of a desired thickness. In other cases, the electrochemical procedure is different from the electrochemical reaction that the device is configured to perform. For example, the electrochemical procedure may be or may include a water splitting reaction, and the device is directed to perform another reaction, such as CO2 reduction.

[0101] Alternatively or additionally, operation 220 includes exposing the surface to oxygen in operation 224. In some cases, the exposure to oxygen occurs in an oxygen-containing reaction chamber, such as a plasma asher. Alternatively or additionally, the exposure includes an ambient environment. The exposure can last for a predetermined period of time, for example, on the order of minutes or on the order of hours, although other periods may be used. The duration of the exposure can be selected to produce an oxynitride layer of a desired thickness. The duration of the exposure may also be selected in other manners to establish a uniform distribution of the oxynitride material.

[0102] The incorporation of oxygen in operation 220 may alternatively or additionally include operation 226, in which an oxygen-containing material is deposited on the surface. For example, the oxygen-containing material may be comprised of or otherwise include an oxide material, such as aluminum oxide, hafnium oxide, or an oxygen-containing ferroelectric layer (e.g., HfZrO2), although other materials may be used.

[0103] Operation 220 may include additional operations in combination with any one or more of the above-mentioned techniques for exposing the surface to oxygen (or otherwise incorporating oxygen into the surface). In the example of FIG. 2, operation 220 includes performing one or more cleaning procedures or other surface treatments in operation 225. The surface is treated in operation 225 prior to exposing the surface to oxygen (e.g., operation 224) or depositing an oxygen-containing layer (e.g., operation 226). In some cases, operation 225 includes an in-situ pre-cleaning procedure, such as a chemical treatment of the surface. For example, a variety of chemicals may be used for treatment, including HF, buffered HF, and / or sulfuric acid. Operation 225 may include alternative or additional cleaning or treatments, such as ultraviolet (UV) ozone treatment. Still other types of procedures may be used, including, for example, other types of surface treatments configured to improve interfaces at the surface.

[0104] Operation 220 may alternatively or additionally include operation 227, which involves performing an annealing procedure. In some cases, the annealing may be performed in an oxygen environment to provide a source of oxygen. Alternatively, the annealing may follow deposition of an oxygen-containing layer in operation 226, or other provision of oxygen to the surface. The annealing may be configured to incorporate oxygen into the surface (e.g., via substitution in the lattice of the structure, etc.). The temperature of the annealing procedure may be varied. For example, the temperature may be about 100 o C ~ about 500 o C range, although other temperatures may be used. The time of the anneal may vary accordingly. The temperature, duration, and / or other characteristics of the procedure may be selected to establish a uniform distribution of the oxynitride material. In some cases, the anneal procedure may not involve an oxygen environment, for example, when oxygen is made available in another manner, such as deposition (e.g., deposition of an oxide material).

[0105] Oxidation of the surface may be enabled, enhanced, or otherwise possible due to one or more aspects of the surface. For example, the surface may be oriented along a non-polar face of the semiconductor composition. In other cases, the surface is oriented along a bipolar plane. Alternatively or additionally, the semiconductor composition may be configured such that the oxidation reaction results in partial oxygen replacement of nitrogen. In such cases, the nitrogen of the surface may be arranged in the compound semiconductor sequence of the semiconductor composition. The semiconductor composition may have a wurtzite crystal structure.

[0106] Operation 200 may include operation 228, in which catalytic nanoparticles are deposited across the array of nanowires. For example, catalytic nanoparticles may be deposited in photoelectrochemical instances and not in other applications such as transistors and other electronic devices. Deposition of the nanoparticles may be accomplished via performance of a photodeposition procedure in operation 230. The nanoparticles may be composed of or include a metal, such as platinum, or other metallic material. A drying operation 232 may then be performed. Further details regarding photodeposition procedures are described in one or more of the U.S. patents referenced above. Alternatively or additionally, operation 228 includes performance of an electrodeposition procedure.

[0107] The amount of catalytic nanoparticles deposited in operation 228 may be reduced as a result of the formation of the oxynitride layer. As described below, the presence of the oxynitride layer allows a similar beneficial effect to be achieved with a smaller amount of catalytic nanoparticles.

[0108] In some cases, operation 228 is performed after the formation of the stabilization layer in operation 220. In other cases, operation 228 is performed before the formation of the stabilization layer. Depositing nanoparticles before the formation of the stabilization layer can be useful to facilitate the formation of an oxynitride material.

[0109] In some cases, method 200 includes operation 234 in which one or more other device structures are formed. For example, the structures may include interfaces stabilized by the stabilization layer of the disclosed methods and devices. In one example, a passivation layer is deposited or otherwise formed. Thus, the stabilization layer may be disposed between the structure surface to be stabilized and the passivation layer. In another example, contacts are formed in operation 234. The stabilization layer may be disposed between the contacts and the structure surface to reduce contact resistance. Operation 234 may alternatively or additionally be directed to the formation of any number of additional or alternative device structures, including, for example, backside contacts.

[0110] The method 200 may include one or more additional acts directed to forming the photocatalytic structure of the device. For example, one or more acts may be performed to deposit catalytic nanoparticles across the array of nanowires. The deposition of the nanoparticles may be accomplished by performing a photodeposition procedure. The nanoparticles may be composed of or include a metal, such as platinum, or other metallic material. A drying operation may then be performed. Further details regarding the photodeposition procedure are described in one or more of the above-referenced U.S. patents. Still other acts may include a procedure in which the photocatalytic structure of the device is annealed. The parameters of the annealing process may be varied.

[0111] 3 illustrates a device 300 having a stabilized interface according to one example. Device 300 may be fabricated via method 200 of FIG. 2 and / or another method. In this example, device 300 is an electronic device configured as or including a transistor. As a field effect transistor, one or more structures of device 300 may be exposed to high electric fields and may therefore benefit from stabilization. In other cases, stabilization may include other types of structures and / or other situations.

[0112] Device 300 includes a substrate 302 and a transistor body 304 supported by substrate 302. In this example, transistor body 304 is a region (e.g., a doped region) of device 300. Alternatively, or additionally, transistor body 304 is a structure formed or disposed on substrate 302. Thus, transistor body 304 may be integrated with substrate 302 to any desired extent. The manner in which transistor body 304 is supported by substrate 302 may vary accordingly.

[0113] The transistor body 304 includes a channel 306 or channel region having a semiconductor composition. The channel 306 may extend between a source region and a drain region of the device 300 as shown. The channel 306 may be formed during operation when a voltage G is applied to a gate 308 of the device 300. In some cases, the channel 306 is doped and / or configured differently relative to the remainder of the transistor body 304. The gate 308 is separated from the channel 306 by a gate oxide layer 310. In some cases, the gate oxide layer 310 includes aluminum oxide, although additional or alternative oxide materials may be used including, for example, hafnium oxide, zirconium dioxide, strontium oxide, scandium oxide, lithium oxide, titanium oxide, tantalum oxide, tungsten oxide, and silicon oxide.

[0114] The channel 306 may be composed of or include additional or alternative semiconductor materials, including, for example, AlGaN, GaN, InN, AlN, BN, and alloys thereof. The remainder of the transistor body 304 may be composed of, or may not be composed of, the same semiconductor material.

[0115] The device 300 includes a stabilization layer 312 on a surface 314 of the channel 306. As described herein, the surface 314 includes nitrogen. For example, the semiconductor composition of the channel 306 may be or include GaN. In some cases, the channel 306 is configured such that the surface 314 is nitrogen terminated. In this example, the stabilization layer 312 is configured as a protective or other passivation layer for the surface 314.

[0116] The stabilization layer 312 may be or may include a uniform distribution of an oxynitride material. In some cases, the oxynitride material is GaON. The uniform nature of the distribution may stabilize the interface between the channel 306 and the gate oxide layer 310. Such stability may be useful in connection with high voltage operation or other operating conditions where the interface is exposed to high energy charge carriers and / or other deleterious conditions. In some cases, the oxynitride material is distributed continuously across the surface 314 (e.g., completely covering the entirety of the surface 314).

[0117] The stabilization layer 312 can have a thickness that does not adversely affect the operation of the device 300. For example, the stabilization layer 312 can be thinner than the gate oxide layer 310. As described herein, the thickness of the stabilization layer 312 can range from about one monolayer to a few monolayers in some cases. Other thicknesses may be used, including, for example, various sub-nanometer thicknesses and thicknesses on the order of nanometers.

[0118] The components of device 300 are illustrated diagrammatically in Figure 3 for ease of illustration, and therefore the components of Figure 3 and other schematic diagrams referenced herein may not be drawn to scale.

[0119] Details regarding examples of the above-mentioned devices, systems, and methods are provided in conjunction with FIGS. 4-8, which address examples of implementations, e.g., Si / GaN photocathodes configured for efficient and durable solar water splitting, for hydrogen production. Examples are described to explore the properties of Si / GaN architectures along with improvements in the photoelectrochemical performance of such architectures. In these examples, the improvements result from the implementation of electrochemical procedures, such as chronoamperometry testing. In other cases, one or more additional or alternative actions can be implemented to form the oxynitride layer described above. The oxynitride layer provides long-term stability (e.g., via an interfacial-related self-healing effect) and / or thermodynamically stable nature of the Si / GaN architecture. Understanding these aspects of the GaN overlayer establishes techniques for optimizing PEC cell, system, and other device performance.

[0120] The examples demonstrate that Si / GaN photocathodes can generate sustained hydrogen production with near 100% faradaic efficiency (FE) without the use of any electrocatalyst. Accelerated testing of the examples also demonstrates that the Si / GaN photocathodes can maintain high current density under enhanced 3.5 sun illumination for over 150 hours. As shown in the examples, under operating conditions, the photocurrent density and onset potential of the Si / GaN photocathode self-improve. A series of advanced characterization techniques and first-principles calculations were used to elucidate the origin of this self-improving behavior. Under operating conditions, the Si / GaN photocathode is subjected to chemical transformations at the GaN surface, which provide protection and improve PEC performance by adding active catalytic sites for the hydrogen evolution reaction. In this way, these examples of Si-protected photocathodes show useful levels of stability and FE performance without the use of any electrocatalyst as a result of beneficial chemical transformations. The self-improving behavior and chemical transformations of the Si / GaN photocathode provide useful devices and systems for hydrogen production in PEC water splitting.

[0121] To set a baseline for understanding the stability mechanism of the GaN overlayer, Si / GaN photocathodes were investigated under water reduction conditions. Iridium oxide (IrOx) is used as the counter electrode, and a graphite colloidal adhesive is used to connect the Si / GaN electrode to copper tape. This test configuration prevents metal contamination. Chronoamperometry (CA) tests were performed on the different samples under 1 sun intensity illumination. In all cases, the measured photocurrent increases before reaching a steady state. Interestingly, the calculated corresponding faradaic efficiency (FE) also follows a similar enhancement, i.e., the FE starts at about 85% and then achieves 100% when the photocurrent reaches its maximum value, as shown in part (a) of Figure 4. Moreover, only hydrogen and oxygen were produced during the CA tests. In contrast, bare n-type photocathodes tested under the same test conditions showed a significant increase in the photocurrent, i.e., the photocurrent reaches its maximum value ... + The p-silicon photocathode has a current of 0.1mA / cm 2 In addition, accelerated CA tests under stronger AM 1.5G sunlight illumination at 3.5 solar intensity were performed. Although photocurrent enhancement was still observed, this accelerated test showed that the Si / GaN photocathode exhibited a photocurrent density of 100 mA / cm 2 We further reveal that a photocurrent density exceeding 1000 nm can be maintained for more than 150 h. This measurement demonstrates the stability of the Si / GaN photocathode under high photocurrent values ​​and concentrated illumination conditions, as shown in part (b) of Figure 4. The JE and photoelectrochemical properties appear to be best reported for a protected Si photocathode, especially in the absence of a catalyst layer.

[0122] To provide further insight into the PEC performance enhancement, intermittent 10-hour CA tests were performed with linear sweep voltammetry (LSV) collected after every hour to track the changes in PEC performance, i.e., photocurrent density and onset potential. As shown in part (c) of Figure 4, the photocurrent density curve versus applied potential (JE) shows an anodic shift over time, with an onset potential shift of more than 300 mV, from about 460 mV vs. reversed hydrogen electrode (RHE) to about 80 mV vs. RHE. The as-prepared device is denoted herein as "CA-0h" and the device tested for 10 hours is denoted herein as "CA-10h". Furthermore, the maximum photocurrent density (J) gradually increased after 3 hours, reaching 25 mA / cm. 2 reaches a saturation level.

[0123] To investigate the mechanism underlying this self-improvement behavior, the change in surface resistance was analyzed by conducting electrochemical impedance spectroscopy (EIS) under AM 1.5G 1 solar illumination. By evaluating the simulated equivalent circuit and the corresponding calculated results, the surface resistance (R ct, surf ) was reduced from 600Ω to 100Ω. R ct, surf The dramatic change in proves that the charge transfer at the solid / liquid interface improves with time. Furthermore, Mott-Schottky analysis reveals that the flatband potential is similar after the reaction, and the flatband position is n + It was suggested that the p value is determined by the Si solar cell substrate. ct, surf While there is a large change in R and a slower change in the other resistances (R) and constant phase elements (CPE), the observed self-improving behavior is attributed to the improvement of the surface dynamics of the GaN layer if the band alignment remains unchanged.

[0124] The observed self-improving anodic shift of onset potential and improved charge transfer indicate an enhanced catalytic activity of the PEC-tested GaN surface, which is typically obtained by the integration of a catalyst layer. Therefore, with a unique opportunity to improve device performance while decreasing the (noble metal) catalyst loading, we investigated the direct incorporation of platinum (Pt) catalyst onto the modified Si / GaN photocathode. A comparison was made between the as-prepared Si / GaN photocathode and the CA-tested surface. The initial PEC performance of the as-prepared device was recorded by LSV scanning. The CA test was then carried out for 4 h to ensure that the maximum current was achieved (see, for example, Figure 4(a)-(c), showing that the improvement is complete within the first few hours). Pt was photodeposited onto both the 4-h CA-tested device (shown as S#2 CA-4h in part (d) of Figure 4) and the as-prepared cathode (shown as S#1 CA-0h in Figure 1d). As can be seen in Figure 1d, S#2 CA-4h requires only 5 min of Pt deposition (S#2 CA-4h, Pt-5min) to shift its turn-on voltage to 0.34 V. An additional 5 min of Pt deposition (S#2 CA-4h, Pt-10min) has the same JE curve as the 5 min deposition. In stark contrast, 20 min of Pt deposition is required (S#1 CA-0h, Pt-20min) to shift the turn-on voltage to 0.34 V. That is, the required Pt loading is 4 times higher than the CA test sample. It is therefore clearly demonstrated that the use of platinum catalyst can be reduced on the CA test surface.

[0125] Figure 4 shows the self-improving behavior of Si / GaN photocathode. Part (a) shows a CA test for 10 h under 1 sun illumination and a constant bias at -0.6 V vs RHE. The inset shows the corresponding faradaic efficiency, which reveals the self-improving nature of GaN. As a comparison, CA tests were performed on bare Si (bare silicon) under 1 sun illumination for 5 h at -0.6 V vs RHE. The bare silicon photocathode achieved about 0.05 mA / cm within 1 h. 2The photocurrent density (JE) curves are plotted as a function of the CA test time. Part (d) of Figure 4 shows some selected JE curves of the as-prepared Si / GaN photocathode (CA-0h) and the 4-h CA-tested Si / GaN photocathode (CA-4h) with different Pt photodeposition durations under 1-day irradiation.

[0126] Surface roughness changes and formation of new chemical species are factors related to the improvement of PEC performance. Therefore, atomic force microscopy (AFM) analysis, as well as scanning transmission electron microscopy combined with electron energy loss spectroscopy (STEM / EELS) and X-ray photoelectron spectroscopy were performed to trace the origin of the observed PEC performance enhancement. Specifically, photoconductive (pc-) AFM was performed on both the as-prepared ("CA-0h") and 10-hour CA-tested ("CA-10h") examples to compare the surface topographies and probe the nanoscale origin of the macroscopic photocurrent enhancement. The surface morphology remained similar for CA-0h and CA-10h surfaces with similar surface roughness of about 12 nm (as shown in Figures 5(a) and (c)), ruling out surface roughness variation inducing the PEC performance improvement in this case. However, significant differences were observed in the photocurrent of the CA-0h and CA-10h examples, as shown in Example 5. Overlaid topography photocurrent maps showed that the CA-0h photocathode exhibited a maximum value of about 80 pA at an applied bias of about 400 mV, whereas the CA-10h photocathode exhibited a maximum value of about 5 nA without the need for external bias (shown in Examples 5(b) and (d)). Line profiles of the two maps for each sample indicate that the enhanced photocurrent generation in the tested examples is generated by the sidewalls of the grains, while the tops of the grains contribute little to the total photocurrent. Such position-dependent photocurrent generation may be related to exposed facets and crystal orientation.

[0127] Figure 5 shows the pc-AFM characterization of the CA-0h and CA-10h examples. The topographies of the CA-0h and CA-10h examples are shown in (a), (c). The corresponding surface profiles and photocurrents extracted from line scans for the CA-0h and CA-10h examples are shown in parts (b) and (d).

[0128] To understand how the tested examples changed under operating conditions and the reasons underlying the increase in photocurrent at the sidewalls of the GaN grains, aberration-corrected STEM / EELS was used to analyze the crystal orientation combined with chemical information with a spatial resolution of less than 1 nm (as shown in parts (a) and (b) of Figure 6). A native oxide layer of about 2 nm (highlighted by the white / light dashed line) was observed on the top surface of the grains in the CA-0h sample, while less oxide (encircled by the red / dark dashed line) was found on the sidewalls (part (a) of Figure 6). Interestingly, after 10 h of operation, the formation of a layer of about 1 nm mixed with Ga, N and O was observed on the sidewalls of the grains (area enclosed by the red dashed line in part (b) of Figure 6). Meanwhile, the tops of the grains were covered with a Ga- and O-containing layer (no N was observed in the area enclosed by the white dashed line in part (b) of Figure 6), with no obvious change in composition. The epitaxial GaN film has a wurtzite crystal structure, where the top surfaces of the grains are bipolar c-planes (0001 * ), while the sidewalls of the particles are non-dipolar (101 * 0)m surface and (112 * 0) a-plane) and semi-dipole planes (e.g., (112 * 2), (202 * 1) etc. Considering that polarity and orientation effects in GaN often result in different chemical, electrical, and mechanical responses to external stimuli, oxygen accumulation also has orientation preference, i.e., it is thermodynamically favored to form on non-polar or semi-polar faces (sidewalls of the grains) rather than on the polar c-faces (top surfaces of the grains).

[0129] In addition, a * on a number means that there is a line (-) above the number. For example, 1* There is a horizontal line above the 1, and a 2 * There is a horizontal line above the 2.

[0130] Figure 6 shows the chemical analysis of the Si / GaN photocathode surface. Part (a) shows a STEM image of the CA-0h surface with insets (left to right) showing EELS mapping of the Ga L-edge, NK-edge and OK-edge. Part (b) shows a STEM image of the CA-10h surface with insets (left to right) showing EELS mapping of the Ga L-edge, NK-edge and OK-edge. Parts (c)-(f) show the O 1s core level spectra of (c) the CA-0h Si / GaN photocathode, (d) the Si / GaN photocathode after 1 h of CA test, (e) the Si / GaN photocathode after 4 h of CA test, and (f) the Si / GaN photocathode after 10 h of CA test.

[0131] By coupling the STEM / EELS results with the increase in photocurrent and anodic shift of onset potential, it was found that an oxygen-containing GaN phase was formed. This finding was easily confirmed by XPS. In addition to the as-prepared example, three different Si / GaN photocathodes were analyzed after 1 h, 4 h, and 10 h of chronoamperometry testing (CA-1h, CA-4h, and CA-10h). The XPS spectra of the as-prepared and tested examples show similar Ga 2p and N 1s core level spectra and high signal-to-noise ratios, regardless of the length of testing. In contrast, more insight is gained from the O 1s spectra (Figure 6(c)-(f)). In addition to the OH-based and O-Ga-O species, by looking more closely at the O 1s peak, a new peak appears at 531.7 eV. This specific binding energy is representative of gallium oxynitride. Furthermore, the signal-to-noise ratio of the O 1s spectrum is significantly lower compared to the Ga 2p and N 1s spectra. Given that the probing depth of XPS for GaN is about 5-6 nm, such a difference in the signal-to-noise ratio between O 1s and Ga 2p (or N 1s) suggests that all samples are composed mainly of GaN, while only an ultrathin layer of oxynitride is present at the surface. Considering that pc-AFM (part (c) of Figure 5) shows an order of magnitude increase in the photocurrent originating from the sidewalls, but almost no current was detected from the top surface of the grains, such oxynitrides visualized on the sidewalls of GaN grains result in a self-improving behavior.

[0132] To better understand how the experimental observations of the self-improving PEC performance and sustained operation of the Si / GaN photocathode correlate with the presence of gallium oxynitride, density functional theory (DFT) calculations were performed on different GaN surfaces. The top surface in this example is (0001 * ), while the sidewalls of the particles consist of semipolar and nonpolar planes. For simplicity, the simulation uses (0001 * )c plane and (101 *0) m-plane. Figure 8 shows the crystal structure of GaN.

[0133] The role of oxygen on the stability of GaN was investigated using several representative surface models and their formation energies. * ) orientation, it was found that the nitrogen atoms of the surface or subsurface bilayers are completely replaced by oxygen (red and green lines, Figure 7, part (a)), which results in a higher stability compared to the as-prepared surface. The same conclusion was reached when the nitrogen in these bilayers was simultaneously but partially (50%) replaced by oxygen (blue line, Figure 7, part (a)). Overall, the morphology with all surface nitrogen replaced by oxygen shows the highest stability. This is due to the fact that the topmost (0001 * ) consistent with the present STEM results which indicate a strong possibility of forming an oxide layer on the surface.

[0134] Calculations show that the effect of oxygen substitution is non-polar (101 * 0) surface. For this orientation, the most stable structure is instead obtained by partial oxygen substitution (50%) of nitrogen both at the surface and in the subsurface bilayer (blue line, Fig. 7, part (b)). The higher stability of the structure with all surface nitrogens substituted with oxygen is related to the ability to form a surface Ga with a higher oxygen coordination (1.5 vs. 1.0). The calculations show that the (101 * The top layer with (0001 0) orientation is more likely to consist of only Ga and O. * ) orientation, suggesting a preference for a mixture of all three species (Ga, O, N). Thus, localization and hopping of surface charges may lead to the formation of NO species and gallium oxynitride on the non-polar faces. This also supports the STEM / EELS evidence that gallium oxynitride-like compositions are more likely to form on the sidewalls of GaN grains.

[0135] The catalytic activity was then evaluated. The free energy of adsorption of hydrogen, which is the intermediate in both the HER and HOR reactions, was determined. The magnitude of the absolute value of the free energy of H adsorption (|△G ads| ≈ 0) is minimized to maximize the reaction. Here, a survey of the most stable adsorption sites on the non-polar surface (blue line, Figure 7, part (b)) reveals that | △G ads | yields values ​​of 0.12–0.16 eV on Ga. On the other hand, the N sites on the two surfaces have similar [|△G ads 0.31–0.39 eV]. The results show that the surface with partial nitrogen substituted with oxygen is more active than the original one, indicating that a transformation in the surface composition is likely responsible for the improved performance of GaN-protected photocathodes.

[0136] Figure 7 shows DFT calculations on the N-dipole c-plane and the non-dipole m-plane. Parts (a) and (b) have different oxygen configurations within the anion- and cation-rich limits (0001 * )c plane and (101 * 0) shows the calculated formation energies of m-plane GaN surfaces: ideal c-plane and m-plane GaN surfaces; configurations with 100% substitution of nitrogen by oxygen on the surfaces c-plane and m-plane; 100% substitution of nitrogen by oxygen in the subsurface bilayer c-plane and m-plane; and configurations with 50% substitution of nitrogen by oxygen c-plane and m-plane for nitrogen on the surface and subsurface bilayer. In both cases, the ideal surface is chosen as the reference.

[0137] To advance the application of solar water splitting for renewable energy, research has been directed towards the development of cutting-edge coatings for long-term stability and high efficiency. As described herein, the Si / GaN photocathode has a stability of 150 hours under 3.5 suns and is capable of sustained hydrogen generation with FE of about 100% without electrocatalysis. Using pc-AFM, STEM / EELS and XPS characterization techniques, the presence of a thin oxynitride-like layer was shown, which forms over time on the Si / GaN photocathode within the first few hours of operation. Such gallium oxynitride, as well as many other widely reported oxynitrides (such as TaON, HfON, perovskite oxynitrides), have hybridization of N 2p and O 2p orbitals and have good catalytic properties. Gallium oxynitride can provide the role of electrocatalysis by better mediating charge transfer. The DFT simulation results were consistent with the experiments, which showed that replacing nitrogen atoms with oxygen to form oxynitrides could passivate the surface for further stability, while the newly formed oxynitrides could also provide more active sites for HER, leading to improved PEC performance. Thus, GaN is established as a stable and efficient protective layer via the formation of oxynitrides during hydrogen generation.

[0138] FIG. 8 shows the wurtzite structure of GaN. Part (a) shows the wurtzite lattice structure of GaN, with a bipolar c-plane on top and non-dipolar m- and a-planes on the sides. Part (b) is a hexagonal schematic of the GaN wurtzite structure. Part (c) assigns crystal planes to actual GaN grains, with the top surface of the grain being a c-plane and the sidewalls of the grain blending with other planes. Part (d) shows an HRTEM image revealing the crystal structure of MBE-grown GaN used in the disclosed methods and devices.

[0139] Figure 9 shows a schematic of the DFT setup for the free energy calculations described above. Part (a) shows ideal m-plane GaN with hydrogen atoms bonded to gallium atoms. Part (b) shows ideal m-plane GaN with hydrogen atoms bonded to nitrogen atoms. Part (c) shows gallium oxynitride formed in m-plane GaN with hydrogen atoms bonded to gallium atoms bonded to oxygen. Part (d) shows gallium oxynitride formed in m-plane GaN with hydrogen atoms bonded to gallium atoms bonded to nitrogen. Part (e) shows gallium oxynitride formed in m-plane GaN with hydrogen atoms bonded to nitrogen atoms. Part (f) shows gallium oxynitride formed in m-plane GaN with hydrogen atoms bonded to oxygen atoms. Part (g) shows gallium oxynitride formed in m-plane GaN with hydrogen atoms bonded to subsurface nitrogen atoms.

[0140] FIG. 10 illustrates a light-emitting device 1000 according to an example. The device 1000 may be fabricated via the method 200 of FIG. 2 and / or another method. In this case, the device 1000 is configured as a light-emitting diode. The device 1000 includes a substrate 1002 and a structure 1004 having a semiconductor composition supported by the substrate 1002. The substrate 1002 may be composed of a semiconductor material, may not be composed of a semiconductor material, or may include a semiconductor material. For example, the substrate 1002 may be or may include a sapphire substrate.

[0141] In this case, structure 1004 corresponds to or includes a stack of semiconductor layers configured for light emission. For example, one or more of the semiconductor layers in the stack may be composed of or include GaN, InGaN, and / or other III-V semiconductor materials.

[0142] The device 1000 further includes electrodes 1006, 1008 for applying a voltage across the stack of semiconductor layers. One electrode 1006 may be configured as an n-metal contact and the other electrode 1008 may be configured as a p-metal contact. The positions of the p-metal and n-metal contacts may be switched or otherwise changed.

[0143] In this example, the stack includes an n-type layer 1010, an active layer 1012, and a p-type layer 1014. The active layer 1012 is disposed between the n-type layer 1010 and the p-type layer 1014. Each of the n-type layer 1010 and the p-type layer 1014 may be configured as a current conducting layer (conductive layer) of the device 1000. The stack of semiconductor layers may include any number of semiconductor layers. For example, the additional layers may include one or more buffer layers and / or one or more additional active layers.

[0144] The stack of semiconductor layers, and thus structure 1004, has a surface where nitrogen is present. In this example, the surface corresponds to a top surface of p-type layer 1014. For example, the p-type layer may be composed of or include GaN and / or another III-nitride semiconductor, but nitrogen may be present at the surface for additional or alternative reasons.

[0145] The device 1000 includes a stabilization layer 1016 disposed on a surface of the p-type layer 1014. The stabilization layer 1016 may be configured as described herein. For example, the stabilization layer may include a uniform distribution of an oxynitride material. In this case, the stabilization layer 1016 is disposed adjacent to and in contact with the metal contact layer 1008 (e.g., a p-type metal contact layer). The stabilization layer 1016 is disposed along the interface between the metal contact layer 1008 and the stack of semiconductor layers to reduce contact resistance at the interface. In some cases, the oxynitride material of the stabilization layer 1016 is continuously distributed across the surface of the p-type layer 1014 (e.g., continuously covering the entire p-type layer 1014).

[0146] Alternatively or additionally, a stabilization layer 1016 (or a portion thereof) is disposed along the interface between the metal contact layer 1006 and the n-type layer 1010. Thus, contact resistance may be reduced at the other contact and / or both contacts of the device 1000.

[0147] FIG. 11 illustrates a light-emitting device 1100 according to an example. The device 1100 may be fabricated via the method 200 of FIG. 2 and / or another method. In this case, the device 1100 is configured as a light-emitting diode. The device 1100 may be configured in one or more ways as described above in connection with the example of FIG. 10. For example, the device 1100 includes a substrate 1102 and a structure 1104 having a semiconductor composition supported by the substrate 1102. The substrate 1102 and the structure 1104 may have one or more aspects in common with the example of FIG. 10. For example, the structure 1004 corresponds to or includes a stack of semiconductor layers configured for light emission.

[0148] The device 1100 further includes electrodes 1106, 1108 for applying a voltage across the stack of semiconductor layers. One electrode 1106 may be configured as an n-metal contact and the other electrode 1108 may be configured as a p-metal contact. The positions of the p-metal and n-metal contacts may be switched or otherwise changed.

[0149] As described above, the stack includes an n-type layer 1110, an active layer 1112, and a p-type layer 1114. The active layer 1112 is disposed between the n-type layer 1110 and the p-type layer 1114. Each of the n-type layer 1110 and the p-type layer 1114 may be configured as a current conducting layer of the device 1100. The stack of semiconductor layers may include any number of semiconductor layers. For example, the additional layers may include one or more buffer layers and / or one or more additional active layers.

[0150] The stack of semiconductor layers, and thus the structure 1104, has a surface where nitrogen is present. In this example, the surface corresponds to a sidewall of one or more of the semiconductor layers of the stack. For example, each of the n-type layer 1110, the active layer 1112, and the p-type layer 1114 may be composed of or include GaN and / or another III-nitride semiconductor, but nitrogen may be present at the surface for additional or alternative reasons. In this light emitting device example, the sidewall surface may be a polar surface. In other examples (e.g., a transistor device), the sidewall surface may be a non-polar surface.

[0151] The device 1100 includes a passivation layer 1116 disposed along a sidewall of one or more layers of the stack. For example, the passivation layer 1116 may be or include a dielectric material, such as an oxide material. Additional or alternative non-conductive materials may be used. In this example, the passivation layer 1116 is disposed along each sidewall of the stack such that the entirety of each layer is passivated by the passivation layer 1116.

[0152] The device 1100 includes a stabilization layer 1118 disposed between the passivation layer 1116 and the sidewall surface of the stack. The stabilization layer 1118 can be configured as described herein. For example, the stabilization layer 1118 includes a uniform distribution of an oxynitride material. In this example, the stabilization layer 1118 is disposed adjacent to and in contact with each sidewall of the stack. The stabilization layer 1118 is thus disposed along the interface between the stack and the passivation layer 1116 to stabilize the interface between the active layer (and / or other layers of the stack) and the passivation layer and / or reduce surface recombination. In some cases, the oxynitride material of the stabilization layer 1118 is continuously distributed across the sidewall surface of the stack (e.g., continuously covering the entirety of each sidewall surface of the stack). The stabilization layer can thus suppress surface traps and / or address other defects along the interface.

[0153] In this embodiment, a stabilization layer is disposed along the sidewalls of each layer in the semiconductor stack. In other cases, the stabilization layer 1118 (or others) is disposed along the interfaces between a subset of the layers in the stack. For example, the stabilization layer 1118 may be disposed along the sidewalls of only the active layers of the stack.

[0154] The devices 1000 and 1100 of Figures 10 and 11 may include additional, fewer, or alternative structures, layers, or other elements. For example, the stabilization layer 1016 of the device 1000 of Figure 10 may be incorporated into the device 1100 to reduce contact resistance at one or both of the electrodes 1106, 1108.

[0155] As described above, efficient yet durable photoelectrodes and other devices are fabricated. In a solar fuel production example involving a silicon / gallium nitride photocathode active for hydrogen production, a thin layer of gallium oxynitride is formed on the sidewalls of gallium nitride particles via partial oxygen substitution at the nitrogen sites. The thin layer establishes a higher density of catalytic sites for the hydrogen evolution reaction. Chemical conversion of gallium nitride to gallium oxynitride results in sustained operation and enhanced catalytic activity, thus showing the promise of the oxynitride layer as a protective catalytic coating for hydrogen evolution and other reactions.

[0156] Further details are provided below regarding crystallography-based aspects and effects of nanostructures, such as III-nitride nanostructures (e.g., GaN nanowires and other nanostructures), which may be useful and / or provided in connection with PEC devices and systems for various types of reactions, as well as other types of devices.

[0157] The crystallographic aspects and effects of the disclosed device may be useful in addressing the challenges of efficiency, stability, and product selectivity for artificial photosynthesis. Thus, the disclosed device may avoid the trade-off between stability and efficiency. Conventional high-efficiency solar water splitting devices generally exhibit very short lifetimes, while devices with long-term stability exhibit low efficiency. The disclosed device may also be useful in supporting the selective production of target products, which, for example, has so far been very low and difficult to achieve in the faradaic efficiency of light-driven chemical conversion of CO2 and CH4 to liquid chemicals and fuels. The crystallographic aspects and effects of the disclosed device include configuring the surface of the photoelectrode. Since reactions take place on the surface of the photoelectrode, tailoring the surface structure of the photoelectrode provides one of the most effective means to address these fundamental challenges. The facets exposed on the surface play a useful role in determining electronic and catalytic activity, including redox reaction sites, adsorption of chemical species, and desorption of product molecules. In addition, the diversity of atomic arrangements allows the facets to strongly influence the separation, extraction and recombination of charge carriers, electrical conductivity, and resistance to photocorrosion and oxidation, all of which are useful factors for achieving efficient and stable artificial photosynthesis and other chemical reactions.

[0158] Although described in the context of water splitting examples, the crystallography-based aspects and effects may be useful in connection with a wide range of artificial photosynthetic chemical reactions.

[0159] As mentioned above, III-nitride semiconductor nanostructures, such as Ga(In)N nanowires, are useful for applications in artificial photosynthesis. Ga(In)N nanostructures are the only known III-V compound semiconductors whose energy band gap spans a wide range of redox potentials under visible and near-infrared light irradiation while being tunable over nearly the entire solar spectrum, which is useful for highly efficient artificial photosynthesis.

[0160] In some cases, the disclosed devices are configured as photoelectrodes made from III-nitride nanowires. III-nitride nanowires have demonstrated stable and efficient operation for 3,000 hours without performance degradation. III-nitride nanowire-based photoelectrodes can be used for a variety of reactions, including, for example, CO2 reduction to syngas, formic acid, and methane, each of which has a relatively high Faraday effect. Other chemical reactions may also be supported, including, for example, light-driven conversion of CH4 to aromatic compounds using gallium nitride (GaN) nanowires.

[0161] As described below, the crystallographic aspects of III-nitride nanostructures can be used to improve catalytic and other performance measures. Previous work has focused primarily on the effects of composition, size, and doping of GaN-based structures on catalytic performance. However, wurtzite III-nitride nanostructures have different facets, e.g., dipole vs. non-dipole vs. semi-dipole, and different dipoles, e.g., metal dipoles vs. nitrogen dipoles, which have very different structural, chemical, electrical, and polarization properties. To date, the role of crystal facets and polarity of III-nitride nanostructures in photoelectrochemical reactions remains largely unknown. Thus, the relationship between crystal facets and catalytic and electrical activity is described below. It has been shown that significant activity (e.g., photoelectrochemical activity) is observed on nonpolar surfaces (e.g., when the nonpolar surface is exposed to the electrolyte), while little or no activity is measured on polar (e.g., c-face) surfaces (e.g., when only the polar surface is exposed to the electrolyte), thereby enabling the rational design, synthesis, and practical application of next-generation metal nitride semiconductor photoelectrodes and other devices.

[0162] Although described in relation to GaN nanostructures, the disclosed devices and methods can use a wide variety of semiconductor materials. Thus, the disclosed devices and methods are not limited to III-nitride semiconductors. For example, the disclosed devices can include one or more layers, components, or other elements that are composed of or include various nitrogen-containing semiconductors and / or various oxygen-containing semiconductors.

[0163] Several examples are described below to analyze the effect of facet and surface polarity of wurtzite III-nitrides on photoelectrochemical (PEC) activity, e.g., solar water splitting. To clearly distinguish the effect of facets with different polarities, a number of exemplary III-nitride nanostructures were fabricated by both top-down and bottom-up methods, thereby selectively exposing certain facets / surfaces to electrolyte, and their PEC activities were compared. In addition, to eliminate any changes in light absorption and alloy composition related to nanostructure engineering, a GaN layer was grown directly on a Si pn wafer as a platform to act as the primary light absorber. In these examples, the light absorption and photogenerated electron density in the GaN nanostructures remained nearly the same regardless of the changes in their facet and surface polarity.

[0164] The examples show that nonpolar GaN surfaces are significantly more active than polar surfaces for PEC reactions. The atomic origin of such behavior is further elucidated by detailed density functional theory (DFT) calculations described below.

[0165] The examples show the unique formation of oxynitrides on nonpolar faces, but not on the c-face, during PEC reactions. The presence of oxynitrides leads to positive surface polarization and significant downward band bending, which can dramatically enhance charge carrier (electron) extraction. The examples also provide useful insights and guidelines for the crystal facet engineering of emerging metal-nitride photocatalysts and photoelectrodes for a wide range of artificial photosynthetic chemical reactions.

[0166] In the first set of examples, GaN epilayers were grown on Si pn wafers using plasma-assisted molecular beam epitaxy (MBE). To date, growing GaN epilayers directly on Si wafers has remained extremely challenging due to the large lattice mismatch (~17%) and large difference in thermal expansion coefficients. Traditionally, various methods have been employed to improve GaN epilayer quality, including AlN seed layers, AlGaN / GaN defect filtering layers, low-temperature AlN interlayers, and SiC interlayers. However, these growth techniques are not suitable for the examples because the large band gap buffer layer acts as a charge carrier transfer barrier between the Si wafer and the GaN epilayer, thus degrading the PEC performance. Furthermore, previous studies have focused on GaN growth on Si(111) substrates, and GaN epilayers on Si(100) wafers have rarely been demonstrated, which is necessary for direct integration with Si pn junctions. To overcome these challenges, the examples used optimized growth conditions to realize GaN quasi-epilayer structures on Si pn wafers by controlled nanowire coalescence. The growth temperature was about 640° C., the nitrogen flow rate was 0.4 sccm, and the Ga beam equivalent pressure was about 1.6e-7 Torr, corresponding to a growth rate of about 120 nm / h.

[0167] Quasi-membrane structures grow preferentially along the c-plane direction. In such structures, the top GaN c-plane is the surface that is primarily exposed to the electrolyte, while the nonpolar plane remains unexposed. To compare the effects of polar and nonpolar planes, two samples, sample A and sample B, were prepared. Sample A is the as-grown quasi-epilayer, while sample B was patterned by photolithography and etching of the quasi-epilayer to expose the nonpolar surface, in this case the sidewall. With the exposed sidewall, photoexcited electrons can migrate to the GaN nonpolar surface and reduce protons to hydrogen. On the other hand, in the absence of such exposed sidewall, the electrons must migrate to the top c-plane for the proton reduction reaction. Photolithography and etching are shown diagrammatically in part (a) of Figure 12.

[0168] Part (b) of Figure 12 shows scanning electron microscopy confirming that the Ga bipolar quasi-film achieves good coalescence, exposing only the top c-plane. The resulting patterned sample has a bipolar c-plane as the top surface and non-bipolar m-planes on the sidewalls. The etching process did not expose the underlying Si wafer. For both samples, the simulated solar radiation was absorbed by the underlying Si pn junction. Due to the nearly perfect conduction band alignment between GaN and Si, photogenerated electrons can easily transfer to the GaN and drive the hydrogen evolution reaction (see Figure 12(c)), while holes are transported to the counter electrode for water oxidation. In sample A the dominant reactive surface is the top c-plane, while in sample B the lateral non-polar planes dominate. The inset in Figure 12(c) shows the wurtzite structure of GaN with a polar c-plane top surface and non-polar m-plane sidewalls.

[0169] GaN / Si photocathode, Ag / AgCl reference electrode, and IrO x PEC characterization was carried out in 0.5 M H2SO4 under AM1.5G solar illumination in a three-electrode setup consisting of a GaN / Si photocathode and a counter electrode. Pt cocatalyst nanoparticles were deposited on the GaN / Si photocathode by photodeposition. Linear scan voltammetry (LSV) measurements were performed from 0.6 to approximately 0.3 V vs RHE at a scan rate of 100 mV / s. High photocurrent densities (approximately 40 mA / cm) were obtained from the Si pn junction under AM1.5G solar illumination. 2 While one would expect a photocurrent density of up to 100 kV, sample A did not show any measurable photocurrent density, suggesting that Ga-face c-plane GaN is not very photoelectrochemically active (see FIG. 13(a)). In contrast, sample B showed an onset potential of about 0.45 V vs. RHE (V on ) and approximately 18mA / cm 2 The photoresponse (J photo) is given by the difference in current density between illuminated and dark conditions. Detailed measurements further confirmed that sample B exhibited stable hydrogen generation with a faradaic efficiency of about 100% (see FIG. 13(b)). Given that sample A and sample B are nearly identical except that sample B has a non-polar surface exposed to the electrolyte, the analysis indicates that the non-polar surface of GaN is the active facet for the PEC reaction.

[0170] Referring now to FIG. 14, the charge transfer resistance at the GaN / electrolyte interface was measured by electrochemical impedance spectroscopy (EIS) (see FIG. 14(a)). The EIS curves were obtained at an initial frequency of 2 MHz and a final frequency of 0.5 Hz at a DC bias of 0.2 V vs. RHE with a perturbation AC voltage amplitude of 10 mV. The Nyquist plots under 1 sun illumination of AM1.5G present two semicircles for both samples. One is due to the internal device resistance and the other is due to the interfacial charge transfer resistance between the photoelectrode and the electrolyte. The radius of the first semicircle is similar between sample A and sample B (as the samples are identical in device structure), while the radius of the second semicircle for sample B is much smaller than that for sample A. The difference in the radius of the second semicircle confirms that the charge transfer at the GaN / electrolyte interface is significantly reduced for sample B compared to sample A due to the presence of a large amount of non-polar surface. Mott-Schottky (MS) measurements (see Fig. 14(b)) were also taken to investigate the band alignment between the GaN and the electrolyte in the dark and the surface band bending with an AC perturbation of 10 mV rms amplitude and 2 kHz frequency. Sample A was subjected to a negative flatband potential (V fb ), whereas sample B (with the non-polar surface exposed) had a positive potential, i.e., V fb This difference indicates that sample A has a much higher Schottky barrier for surface charge transfer than sample B. fb The implications and origins of such significant differences in are discussed further below.

[0171] Figure 15 shows the results of an analysis aimed at ruling out the influence of differences in surface area between sample A and sample B as well as differences in charge carrier transport pathways on the PEC activity. To that end, the PEC performance characteristics of example GaN nanowire (NW) arrays grown on Si pn wafers were analyzed. Their growth parameters and structural characteristics are described in the Supporting Information. The nanowire arrays were grown along the c-axis with c-plane and non-polar GaN on the top and side, respectively. Their surface exposure to the electrolyte was controllably varied via the deposition of an insulating parylene passivation layer, followed in some cases by an etch-back treatment to reveal the top portion of the nanowire arrays. The parylene was etched back for different times in oxygen plasma. Longer etching times exposed a greater amount of the GaN NW non-polar sidewalls, resulting in a higher V on This results in a positive shift in the .DELTA. and a better filling factor.

[0172] The as-grown GaN NW / Si photocathode has an onset potential of 0.45 V vs. RHE and 28 mA / cm 2 (see FIG. 15(a)). When an insulating passivation layer 1500 was deposited to protect the non-bipolar sidewalls from the electrolyte, the activity decreased to negligible values ​​(see FIG. 15(b)). After an initial etch-back process to form an insulating passivation layer 1502, the top c-face 1504 of the GaN nanowire was exposed first. Then, as the etch-back process proceeded for longer periods, the sidewalls 1506 of the nanowire were gradually exposed to the electrolyte (see FIG. 15(c)). The effect of increasing the non-polar surface area with etch-back duration is reflected in a positive shift in the onset potential and improved fill factor due to the reduced charge transfer resistance associated with the larger surface area.

[0173] The photocurrent data in Figure 15 also show that even minor exposure of the non-polar sidewalls 1506 can result in or support a maximum saturation photocurrent density that is nearly identical to that of the as-grown nanowire sample, without surface passivation. Such a negligible dependence of the maximum photocurrent density on the etch-back duration rules out the effect of surface area on the measured PEC activity difference between the primarily c-face and (mostly) non-polar face samples, and thus highlights the role that polarity plays in the clear contrast between the PEC activity of the as-grown quasi-membrane sample A and that of the etched sample B. The lack of area dependence also provides an option for partial passivation of nanowires and other nanostructures used for charge transfer, an example of which is described below.

[0174] Density Functional Theory (DFT) Analysis. As mentioned above, oxygen adsorption and substitution can occur on GaN surfaces during the PEC water splitting reaction. Therefore, to understand the fundamental difference in PEC activity between c-plane GaN surfaces and non-polar plane GaN surfaces, the role of oxygen, which can be easily adsorbed and incorporated into the surface during operation, for example, is considered.

[0175] FIG. 16 shows a schematic of the DFT setup for the analysis. The arrows in (a) and (b) indicate the spatial direction of the oxygen-induced surface dipole moment. (a) shows oxygen substitution in the m-plane to form a gallium oxynitride layer. (b) shows oxygen adsorption on the (0001) plane. (c) and (d) show schematic representations of the surface band bending changes induced by oxygen substitution and adsorption, respectively. (Partial) oxygen substitution in the m-plane induces an outward surface dipole moment that reduces the surface band bending as shown diagrammatically in part (c). Oxygen adsorption on the (0001) plane induces an inward surface dipole moment, which increases the surface band bending as shown in part (d), further hindering electron transfer to the surface.

[0176] Plane-wave pseudopotential DFT calculations were performed in the generalized gradient approximation using the quantum-ESPRESSO package. An ultrasonic pseudopotential was used to describe the interaction between the valence electrons and the ion core when Ga 3d electrons were explicitly treated as valence electrons. The wave functions and electron density were expanded in a set of plane-wave bases with cutoff energies of 30 Ry and 240 Ry, respectively. The surface was modeled in a repeating slab geometry using a (2 × 2) unit cell with a thickness of 7 GaN bilayers, and the N-terminal bottom surface was passivated with pseudohydrogen with a charge of 0.75e. In addition, a (101 * 0) surfaces were modeled. A vacuum width of at least 15.0 Å was introduced between successive slabs, and 5 × 5 × 1 and 2 × 2 × 1 k-point meshes were used to sample the surface Brillouin zones of polar and nonpolar surfaces, respectively. * " means the number 1 with a line through it.

[0177] Non-polar (101 * At the GaN (GaN) surface, the most stable structure was obtained when the nitrogen atoms at the surface and subsurface were partially replaced by oxygen, resulting in the formation of a unique GaN oxynitride on the nonpolar surface (see Figure 16(a)). Gallium oxynitride is known to have a smaller energy band gap and enhanced catalytic activity compared to GaN, thereby significantly enhancing the charge carrier (electron) transfer to the nonpolar GaN surface for the PEC reaction.

[0178] On the other hand, for the polar (0001) GaN surface, the most stable surface morphology under oxygen-rich conditions was found to be that with an oxygen adsorption coverage of 0.25 ML (see FIG. 16(b)).

[0179] Interestingly, oxygen is found to have a significant effect on the surface polarization and therefore on the band bending at the surface. More specifically, calculations show that oxygen substitution and adsorption, respectively, are related to the non-polar (101 *We have shown that the surface polarization of a (101) surface leads to positive and negative surface polarizations on the (0001) and polar (0001) surfaces. On the (0001) surface, the negative surface polarization effectively reduces the work function, which is associated with the difference in electrostatic potential deep in vacuum and deep in semiconductors. In contrast, the surface polarization of a (101 * 0) plane increases due to the positive surface polarization.

[0180] Quantitatively, these surface polarizations induce an upward band bending of 0.21 eV on the (0001) plane and a downward band bending of 0.25 eV on the nonpolar planes, as shown in Figure 16(c) and (d). Taken together, these results indicate that oxygen has contrasting effects on charge transfer on the two planes, i.e., oxygen promotes electron transport on the nonpolar planes and restricts transport on the polar planes. These contrasting properties of oxynitride layers can be applied to support different types of device interfaces, respectively geared towards promoting charge transfer (e.g., electrodes) and inhibiting charge transfer (e.g., gate oxide stabilization or passivation), examples of which are described herein.

[0181] The difference between sample A and sample B in the surface polarization induced by the different surface oxygen interactions contributes to the difference in surface band bending shown in the MS measurements. Sample A, with essentially only its top c-face exposed to the electrolyte, exhibits a negative flatband potential (see FIG. 14(b)), indicating a conduction band Schottky barrier at the interface between the n-type GaN c-face and the electrolyte, which may be due to both Fermi level pinning and upward bending banding. In contrast, sample B has lateral nonpolar faces exposed to the electrolyte, and V fbThe open circuit potential (OCP) measurements revealed that sample B had a photovoltage of about 0.4 V, consistent with the MS measurements. The activity difference between sample A and sample B is attributed to the difference in energy barriers between polar and nonpolar surfaces due to the different nature of the surface oxidation, as well as the resulting polarization-induced surface band bending and Fermi level pinning at the surface states of the c-plane GaN surface. At the semiconductor-electrolyte interface, the Fermi level pinning results in a Schottky barrier that remains constant at different polarization biases. This barrier prevents efficient charge carrier transport to the c-plane surface by imposing a large overpotential for the hydrogen evolution reaction, thus resulting in little or no photoresponse from sample A even after Pt co-catalyst deposition, which serves to promote hydrogen evolution using any available surface (photo)electrons. The nonpolar m-face in the presence of surface gallium oxynitride may exhibit an unpinned Fermi level that allows for easy charge transfer to the electrolyte, whereas the electronic energy barrier at the Ga-polar c-face is further exacerbated by oxygen adsorption.

[0182] The aforementioned theoretical and experimental analyses confirmed the importance of non-polar GaN surfaces in promoting the PEC hydrogen evolution reaction. During the PEC reaction, the formation of unique gallium oxynitride on the non-polar GaN surface significantly enhances charge carrier (electron) transport and extraction, while oxygen adsorption on the polar c-plane surface hinders efficient charge carrier extraction. The atomic origin of the highly efficient photo(electro)catalytic activity of GaN nanowires was analyzed in comparison with conventional GaN film structures. Exposing the non-polar facets of wurtzite III-nitrides is a useful strategy to achieve breakthrough advances in PEC activity by enhancing electron transfer between the photocathode and electrolyte. Understanding the relationship between crystal orientation and photo(electro)chemical reaction is useful in PEC water splitting and various other catalytic reactions such as carbon dioxide reduction and nitrogen fixation.

[0183] 17A depicts a device 1700 with crystallographic and oxynitride-based stabilization, according to one example. In some cases, device 1700 may be a photoelectrode for a PEC reaction, such as water splitting, as described herein, or may be another electrode. Device 1700 includes any number of nanostructures 1702.

[0184] Nanostructures 1702 are supported by a substrate (not shown) as described herein. Nanostructures 1702 may be arranged in an array or other configuration. In this case, each nanostructure 1702 is configured as a nanowire, although other shapes and configurations may be used. Nanostructures may be grown (e.g., epitaxially grown) or formed by additional and / or alternative techniques, such as etching. Device 1700 may include or have any one or more of the features or characteristics of nanowire-based devices described herein.

[0185] The nanostructures 1702 have a semiconductor composition. In some cases, the nanostructures 1702 are composed of GaN and / or another III-nitride semiconductor material. Alternative or additional semiconductor materials are, for example, compound semiconductors. For example, the semiconductor composition is not limited to nitrogen-containing semiconductors. Oxygen-containing semiconductors may be used in other cases.

[0186] Nanostructure 1702 has a surface that has either nitrogen or oxygen present. The presence of nitrogen or oxygen can result from the composition of the semiconductor material. In some cases, the surface can be nitrogen-terminated.

[0187] The device 1700 includes a stabilization layer 1704. The stabilization layer 1704 is disposed on a surface of the nanostructure 1702. In this case, the surface on which the stabilization layer 1704 is disposed is the sidewall of the nanostructure 1702. The stabilization layer 1704 includes an oxynitride material as described herein. The oxynitride material can be continuously distributed across the surface of the nanostructure 1702. Thus, the stabilization layer can continuously cover the surface. The oxynitride of the stabilization layer 1704 can be established by replacing nitrogen or oxygen at the surface with oxygen or nitrogen, respectively.

[0188] 17A, the surface on which stabilization layer 1704 is disposed is a non-polar surface of nanostructure 1702. In GaN and other wurtzite embodiments, the non-polar surface is oriented along an m-plane of the lattice of nanostructure 1702.

[0189] The device 1700 further includes a passivation cover or cap 1706 disposed on the polar surface of the nanostructure 1702. In this example, the nanostructure 1702 has a top or upper polar surface, edge, or facet. The passivation cap 1706 may be configured to completely cover or passivate the top surface of the nanostructure 1702. Such covering or passivation may, for example, protect the nanostructure 1702 from corrosion or other degradation resulting from an electrolyte in which the device 1700 is immersed. In this wurtzite example, the polar surface is oriented along the c-plane of the lattice. In some cases, the passivation cap 1706 includes a nitride material or an oxide material. For example, the passivation cap 1706 may be made of AlO x , HfO2, Si x N y , and AlSiN x It may consist of or include, although other materials may be used.

[0190] FIG. 17B depicts a device 1710 with crystallographic and oxynitride-based stabilization according to another embodiment. The device 1710 may be a photoelectrode or other electrode. The device 1710 includes any number of nanostructures 1712. The nanostructures 1712 are supported by a substrate (not shown) as described herein. The nanostructures 1712 may be arranged, configured, or constructed as described herein. Thus, the device 1712 can include or have any one or more of the features or characteristics of the nanowire-based devices described herein.

[0191] 17A, nanostructures 1712 have sidewall surfaces having either nitrogen or oxygen present therein, and device 1710 includes a stabilization layer 1714 disposed thereon. Stabilization layer 1714 includes an oxynitride material, and the sidewall surfaces are non-polar surfaces, as described herein.

[0192] 17B may differ from the above-described embodiments in that the stabilization layer 1714 may or may not cover the entire sidewall surface. The stabilization layer 1714 may instead cover only the exposed portions (e.g., top portions) of the nanostructures 1712 (e.g., exposed to the electrolyte). However, the oxynitride material of the stabilization layer 1714 may nevertheless be continuously distributed across the exposed portions (e.g., the exposed portions are continuously covered by the stabilization layer 1714).

[0193] The device 1710 further includes a passivation cover or cap 1716 disposed on the polar surface of the nanostructure 1712. The passivation cap 1716 may be disposed, configured, or constructed on the polar surface as described above in connection with the embodiment of FIG.

[0194] The device 1710 further includes a passivation structure 1718 that partially covers the non-polar surfaces of the nanostructures 1712. In this example, the passivation structure 1718 surrounds or otherwise covers the lower portions of the sidewall surfaces of the nanostructures 1712.

[0195] The extent to which the sidewall surface is passivated by the passivation structures 1718 can vary. As discussed above, the current density of the device 1710 may not be significantly hindered by a reduction in the exposed surface area of ​​the nanostructures 1712. Thus, the passivation structures 1718 can cover a significant amount of the sidewall surface without adversely affecting charge transfer and / or other device performance.

[0196] In some cases, the passivation structure 1718 is composed of or includes parylene, as described above. Other materials can be used, including various nitride materials and various oxide materials. For example, AlO x , HfO2, Si x N y , or AlSiN x can be used.

[0197] Passivation structure 1718 may be configured as or may include a layer in which number of nanostructures 1712 of device 1710 are embedded or disposed. For example, device 1710 may be configured as described and shown in the example of FIG.

[0198] In other examples, the device may include other types of nanostructures. For example, the device may be configured as a light emitting device, such as a light emitting diode (LED) device. The device may thus include an active layer and an electrode supported by the active layer, as described above in connection with the tenth embodiment. In such a case, the stabilization layer is disposed between the active layer and the electrode. The active layer has a non-polar surface (e.g., a top surface oriented along an m-plane of the wurtzite lattice) on which the stabilization layer is disposed. Thus, the presence of an oxynitride in the stabilization layer may facilitate charge transfer between the active layer and the electrode, as described above.

[0199] LEDs and other devices having a stabilization layer on a non-polar surface may also include one or more passivation structures disposed along the sidewalls of the active layer or other structure stabilized by the stabilization layer. See, for example, the example shown in FIG. 11. In such cases, the sidewalls may be oriented along the m-plane of the lattice of the active layer or other structure. The passivation structure may be or may include an oxynitride-based stabilization layer, as described and illustrated in connection with the example of FIG. 11 and elsewhere herein. However, the stabilization layer is disposed along a polar surface (e.g., the c-plane of the wurtzite lattice) and therefore inhibits rather than promotes charge transfer. Additional or alternative passivation structures along the sidewalls may be used, including, for example, various nitride materials and various oxide materials.

[0200] In still other cases, the device includes a stabilization layer disposed along the polar surface to function as a passivation layer. For example, the device may be configured as a transistor, such as a transistor device. In such cases, the stabilization layer is disposed between the channel layer and the gate oxide layer, as described above in connection with the example of FIG. 3. The surface of the channel layer is a polar surface (e.g., an upper surface of the channel layer oriented along the c-plane of the wurtzite lattice). The placement of the stabilization layer on the polar surface of the oxynitride material inhibits charge transfer, as described above, thereby protecting the gate oxide from high energy charge carriers in the channel layer. The oxynitride material may be continuously distributed across the surface (e.g., continuously covering the entire surface of the channel layer) to suppress surface traps and / or address other defects along the interface between the channel layer and the gate oxide layer.

[0201] 17C shows a device 1720 with surface passivation according to another embodiment. Device 1720 includes a substrate 1722 and nanostructures 1724 supported by substrate 1722. In this embodiment, nanostructures 1724 are planar. Nanostructures 1724 may be a patterned layer formed by an etching procedure. In some cases, nanostructures 1724 are a layer or other component of a heterostructure, and other layers are not shown for ease of illustration.

[0202] Nanostructures 1724 have a semiconductor composition. The semiconductor composition may vary. For example, any of the semiconductor compositions described herein may be used.

[0203] The nanostructures 1724 have a surface that is stabilized by a stabilization layer 1726. The stabilization layer 1726 is composed of or includes an oxynitride material, as described herein. In this example, the surface is a top planar surface of the nanostructures 1724. In some cases, the surface is a non-polar surface, in which case the stabilization layer 1726 can be used to facilitate charge transfer between the nanostructures 1724 and another layer, such as a contact or electrode 1728. As described herein, the surface can correspond to an m-plane of a lattice of the nanostructures 1724. Thus, the stabilization layer 1726 can establish an interface between the m-plane of the nanostructures 1724 and the electrode 1728. In other cases, the stabilization layer 1726 is disposed between the nanostructures 1724 and another layer, component, or element of the device 1700. In yet other examples, the surface of the nanostructures 1724 is a polar surface (e.g., corresponding to a c-plane of a lattice). In such cases and in other cases, stabilization layer 1726 may be used to passivate the surfaces of nanostructures 1724. For example, stabilization layer 1726 may act to stabilize the interface between nanostructures 1724 and layer 1728, which may be configured as an oxide or other dielectric layer. In any of these cases, the oxynitride material of stabilization layer 1726 may be continuously distributed across the planar surfaces of nanostructures 1724 (e.g., continuously covering the entirety of the surface).

[0204] Apparatus 1700 may have one or more other aspects, features, components, or other characteristics in common with any one of the other embodiments described herein.

[0205] Figure 18 illustrates a method 1800 for manufacturing a device according to several examples. Method 1800 can be used to fabricate the apparatus of Figures 17A-17C, as well as other embodiments described herein, including, for example, those shown in connection with Figures 1, 3, 10, and 11.

[0206] Method 1800 may be implemented in conjunction with one or more of the acts of other manufacturing methods described herein, including, for example, those illustrated in connection with Figure 2. For example, method 1800 may include one or more acts directed to providing a substrate for a device. Such substrate-related methods may be configured as described herein.

[0207] The method 1800 includes an operation 1802 in which a structure of a device is formed. The structure may be or may include a nanostructure, such as a nanowire, an active, a channel, or other layer of a device. The structure is supported by a substrate and has a semiconductor composition as described herein.

[0208] In the example of Figure 18, method 1800 includes two branches corresponding to the two different types of stabilization layers described herein. One branch is directed to fabricating devices in which the stabilization layer supports charge transfer, for example, at the electrodes. The other branch is directed to fabricating devices in which the stabilization layer acts, for example, as a passivation or insulating layer to inhibit charge transfer. In some cases, a device can include both types of stabilization layers, in which case both branches are implemented.

[0209] With respect to the charge transfer branch, method 1800 can include operation 1804 in which one or more passivation structures are deposited. For example, the passivation structure can be disposed adjacent to the structure such that a portion of the structure is covered by the passivation structure. Examples of such passivation structures include capping layers on the c-face of the nanowire, as described herein. Thus, operation 1804 can include potentially covering the c-face or other polar surface of the structure. In other examples, the passivation structure partially covers the sidewall (e.g., m-face) or other non-polar surface of the nanowire, as described herein.

[0210] Method 1800 includes operation 1806, in which a stabilization layer is formed on the non-polar surface of the structure. The stabilization layer includes an oxynitride material, as described herein. In some cases, operation 1806 may include depositing an oxygen-containing material (e.g., an oxide) on the non-polar surface and performing an anneal to form the oxynitride material from oxygen of the oxygen-containing material. Operation 1806 may include any one or more of the methods described herein related to incorporating oxygen into a surface, such as one or more pretreatments of the surface (see, e.g., FIG. 2). In other cases (e.g., a semiconductor surface that already includes oxygen), a nitrogen-containing material (e.g., a nitride) may be deposited on the surface. Operation 1806 may also be used to form oxynitrides on c-planes or other polar surfaces of nanowires or other structures.

[0211] Operation 1806 may be performed via an oxidation reaction, as described herein, and thus may be performed after depositing the passivation structure in operation 1804. In other cases, deposition of the passivation structure is performed after forming the stabilization layer.

[0212] Turning to the passivation / insulation branch, method 1800 includes operation 1808 in which a stabilization layer is formed on a polar surface of the structure. The stabilization layer is composed of or includes an oxynitride material, as described herein. In this case, the oxynitride of the stabilization layer is used to passivate and / or insulate an interface of the device, such as an interface between a channel layer and a gate dielectric layer of a transistor device, as described herein. Thus, the polar surface of the structure may be a c-plane of a wurtzite lattice, as described herein. Operation 1808 may include any one or more of the operations described herein related to incorporating oxygen into a surface (e.g., see FIG. 2).

[0213] Method 1800 may include operation 1810 to address one or more surfaces of the structure, such as a polar (e.g., c-plane) surface on which the stabilization layer is formed, or a non-polar (e.g., m-plane) surface, via deposition of a passivation or other insulating structure. In an example of the former case, operation 1810 may include depositing an insulating layer, such as a gate oxide layer. In an example of the latter case, operation 1810 may include depositing a passivation structure along a sidewall. For example, via operation 1810, a non-bipolar sidewall of an active layer may be passivated. Operation 1810 may be directed to depositing a passivation structure along an m-plane or other non-polar surface of the structure such that such surface is covered by the passivation structure. Thus, the passivation structure deposited in operation 1810 may be useful in cases where the use of an oxynitride in the stabilization layer does not passivate a non-polar surface for reasons described herein.

[0214] The present disclosure has been described with reference to specific examples which are intended to be illustrative only and are not intended to limit the disclosure. Modifications, additions, and / or deletions can be made to the examples without departing from the spirit and scope of the disclosure.

[0215] The above description has been given for clarity of understanding only, and no unnecessary limitations should be understood therefrom.

Claims

1. 1. A method of manufacturing a device, comprising: Providing a substrate for the device; forming a structure for the device; the structure includes a surface supported by the substrate, the surface having a semiconductor composition and having nitrogen present therein; The method further comprises introducing oxygen to the surface to form a stabilization layer on the surface; The method of manufacturing a device, wherein the introduction of oxygen is performed such that the stabilization layer comprises a uniform distribution of oxynitride material.

2. The method of claim 1 , wherein the introducing of oxygen comprises performing an oxidation reaction to form the stabilization layer.

3. The method of claim 1 , wherein the introducing of oxygen comprises performing an electrochemical procedure.

4. The method of claim 3 , wherein the electrochemical procedure comprises performing a water splitting reaction, in which the device is immersed in water.

5. The method of claim 1 , wherein the introducing of oxygen comprises annealing the surface.

6. The method of claim 1 , wherein the introducing of oxygen comprises depositing an oxygen-containing material on the surface.

7. The method of claim 6 , wherein the oxygen-containing material comprises a ferroelectric material.

8. forming the structure includes forming an array of conductive protrusions supported by and extending outwardly from the substrate, the array of conductive protrusions including the structure; forming the array of conductive protrusions includes performing a molecular beam epitaxy (MBE) growth procedure; The method of claim 1 , wherein the MBE growth procedure is carried out under nitrogen-rich conditions such that the sidewalls of each conductive protrusion in the array of conductive protrusions are nitrogen-terminated.

9. The method of claim 1 , wherein the semiconductor composition is configured to perform the oxidation reaction to result in partial oxygen replacement of the nitrogen.

10. A device, comprising: A substrate; a structure having a semiconductor composition, the structure being supported by the substrate, the structure including a surface; The device further comprises a stabilization layer disposed on the surface of the structure, the stabilization layer comprising an oxynitride material; a device, wherein the oxynitride material of the stabilization layer is distributed continuously across the surface of the structure.

11. The device of claim 10 , wherein the surface is a non-polar surface of the structure.

12. The device of claim 11 , wherein the non-polar surface is oriented along an m-plane of the lattice of the structure.

13. The device of claim 11 , wherein the non-polar surface is a sidewall of the structure.

14. The device of claim 10 further comprising a passivation cap disposed on a polar surface of the structure.

15. The device of claim 14 , wherein the polar surface is oriented along a c-plane of the lattice of the structure.

16. The device of claim 14 , wherein the passivation cap comprises a nitride material or an oxide material.

17. The device of claim 11 , further comprising a passivation structure partially covering the non-polar surface.

18. The device of claim 10 , wherein the semiconductor composition comprises nitrogen, and the nitrogen is present at the surface.

Images