US20110214997A1

US20110214997A1 - Magnetically modified semiconductor electrodes for photovoltaics, photoelectrosynthesis, and photocatalysis

Info

Publication number: US20110214997A1
Application number: US13/028,113
Authority: US
Inventors: Johna Leddy; Heung Chan Lee
Original assignee: University of Iowa Research Foundation UIRF
Current assignee: University of Iowa Research Foundation UIRF
Priority date: 2010-02-16
Filing date: 2011-02-15
Publication date: 2011-09-08

Abstract

A device for hydrogen gas production comprising a working electrode comprising a magnetically-modified semiconductor electrode. Onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV, or no more than about −600 mV, or no more than about −500 mV. The magnetically-modified semiconductor working electrode provides the device with a photoconversion efficiency of at least about 0.1%, or at least about 1.6%, or at least about 6.2%. Other applications include photovoltaics, photoelectrochemical synthesis, and photocatalysis.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims priority to U.S. provisional application Ser. No. 61/392,920, filed on Oct. 13, 2010, and to U.S. provisional application Ser. No. 61/282,467, filed on Feb. 16, 2010, each of which are incorporated herein by reference in their entireties.

GOVERNMENT LICENSE RIGHTS

This invention was made with Government support under Grant No. 0809745 awarded by the National Science Foundation. The government has certain rights in the invention.

BACKGROUND

Hydrogen gas production using solar energy is thermodynamically,possible; however, low kinetic rates on semiconductor electrodes may require the application of high overpotentials to make the hydrogen gas, an energetic cost that can be too high to make hydrogen production at semiconductor electrodes economically attractive. By reducing recombination and increasing interfacial electron transfer rates with applied magnetic fields, overpotentials can be substantially reduced and hydrogen production using solar energy can become more economically feasible.
Existing approaches to increasing reaction kinetics on the surface of a semiconductor electrode can involve application of precious metals or catalysts, such as platinum, platinum alloys, palladium, and palladium alloys. Such materials, and their application, can be expensive, and may result in electrodes with decreased mechanical stability, because the precious metal or catalyst is placed directly on the reaction site and can become dislodged due to hydrogen evolution during electrolysis.
Additional background is provided herein below under the “Additional Embodiments” section.

SUMMARY

Embodiments described herein include devices and apparatus, compositions, and methods of making and using these in applications including, for example, photoelectrosynthesis, photoelectrocatalysis, and photovoltaics.
In one embodiment, for example, a device is provided for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and at least one counter electrode, wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV, or no more than about −600 mV, or no more than about −500 mV. The device can be used in other applications also besides hydrogen generation.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and at least one counter electrode, wherein the working electrode and counterelectrode provide the device with a photoconversion efficiency of at least about 0.1%, or at least about 1.6%, or at least about 6.2%.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode, wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 0.1%, or at least about 1.6%, or at least about 6.2%.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and providing at least one counter electrode, using the working electrode and the counter electrode in a device, wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV, or no more than about −600 mV, or no more than about −500 mV.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and providing at least one counter electrode, using the working electrode and the counter electrode in a device, wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 0.1%, or at least about 1.6%, or at least about 6.2%.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; providing at least one counter electrode; and using the working electrode and counter electrode to produce hydrogen gas.
One embodiment provides a device adapted for photo applications comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode. Photo applications include solar cells and photovoltaics, as well as photoelectrosynthesis and photoelectrocatalysis. Solar cell devices and modules can be made as known in the art.
Another embodiment provides a method comprising: providing at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; providing at least one counter electrode; and using the working electrode and counter electrode upon exposure to photons.
The working electrode can comprise p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>, or mixtures thereof (e.g., as in polycrystalline Si).
At least one advantage for at least one embodiment is improved current flow and/or higher efficiency. In some embodiments, magnetic modification, of p-type silicon semiconductor surfaces can enable hydrogen production with relatively little or no overpotential. Also, in at least one embodiment, advantageously, the embodiment does not require the use of precious metals or catalysts, which can be expensive and difficult to acquire, as well as being difficult to stabilize at the active surface.

Additional Embodiments

One embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises at least one magnetically-modified semiconductor electrode; and at least one counter electrode, wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the working electrode comprises p-type silicon with a low doping level. In another embodiment, the working electrode comprises p-type silicon having resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, wherein the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>. In another embodiment, the working electrode comprises magnetite. In another embodiment, the working electrode comprises silane-coated magnetite. In another embodiment, the working electrode comprises a polymeric material. In another embodiment, the working electrode comprises silane-coated magnetite and a polymeric material. In another embodiment, wherein the counter electrode comprises platinum. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising Ga—In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the onset of hydrogen gas evolution occurs at an overpotential of no more than about −600 mV. In another embodiment, the onset of hydrogen gas evolution occurs at an overpotential of no more than about −500 mV.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode.
In another embodiment, the p-type semiconductor comprises silicon with a low doping level. In another embodiment, the p-type semiconductor has resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the magnetic material comprises silane-coated magnetite. In another embodiment, the ion-exchange polymer comprises NAFION™. In another embodiment, the counter electrode comprises platinum mesh.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and at least one counter electrode, wherein the working electrode and counterelectrode provide the device with a photoconversion efficiency of at least about 0.1%.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the working electrode comprises p-type silicon with a low doping level. In another embodiment, the working electrode comprises p-type silicon having resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>. In another embodiment, the working electrode comprises magnetite. In another embodiment, the working electrode comprises silane-coated magnetite. In another embodiment, the working electrode comprises a polymeric material. In another embodiment, the working electrode comprises silane-coated magnetite and a polymeric material. In another embodiment, the counter electrode comprises platinum. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising Ga—In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the photoconversion efficiency is at least about 1.6%. In another embodiment, the photoconversion efficiency is at least about 6.2%.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode, wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 0.1%.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the p-type semiconductor comprises silicon with a low doping level. In another embodiment, the p-type semiconductor comprises a resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>. In another embodiment, the magnetic material comprises magnetite. In another embodiment, the magnetic material comprises silane-coated magnetite. In another embodiment, the ion-exchange polymer comprises NAFION™. In another embodiment, the counter electrode comprises platinum mesh. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising an acidic electrolyte. In another embodiment, the wherein the magnetic material and the ion-exchange polymer are disposed on less than the entire surface of the p-type semiconductor. In another embodiment, the device further comprising Ga-In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the p-type semiconductor comprises a resistivity of about 0.01 to about 10 Ω-cm, the magnetic material comprises silane-coated magnetite, and the ion-exchange polymer comprises NAFION™. In another embodiment, the photoconversion efficiency is at least about 1.6%. In another embodiment, the photoconverision efficiency is at least about 6.2%.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and providing at least one counter electrode, using the working electrode and the counter electrode in a device, wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the working electrode comprises p-type silicon with a low doping level. In another embodiment, the working electrode comprises p-type silicon having resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>. In another embodiment, the working electrode comprises magnetite. In another embodiment, the working electrode comprises silane-coated magnetite. In another embodiment, the working electrode comprises a polymeric material. In another embodiment, the working electrode comprises silane-coated magnetite and a polymeric material. In another embodiment, the counter electrode comprises platinum. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising Ga—In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the onset of hydrogen gas evolution occurs at an overpotential of no more than about −600 mV. In another embodiment, the onset of hydrogen gas evolution occurs at an overpotential of no more than about −500 mV.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and providing at least one counter electrode, using the working electrode and the counter electrode in a device, wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 0.1%.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the working electrode comprises p-type silicon with a low doping level. In another embodiment, the working electrode comprises p-type silicon comprising a resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>. In another embodiment, the working electrode comprises magnetite. In another embodiment, the working electrode comprises silane-coated magnetite. In another embodiment, the working electrode comprises a polymeric material. In another embodiment, the working electrode comprises silane-coated magnetite and a polymeric material. In another embodiment, the counter electrode comprises platinum. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising an acidic electrolyte. In another embodiment, the device further comprising Ga—In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the photoconversion efficiency is at least about 1.6%. In another embodiment, the photoconversion efficiency is at least about 6.2%.
Another embodiment provides a method of producing hydrogen gas, comprising: providing at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; providing at least one counter electrode; and using the working electrode and counter electrode to produce hydrogen gas.
In another embodiment, the working electrode comprises p-type silicon. In another embodiment, the p-type semiconductor comprises silicon with a low doping level. In another embodiment, the p-type semiconductor comprises a resistivity of about 0.01 to about 10 Ω-cm. In another embodiment, the working electrode comprises p-type silicon comprising surface orientation comprising <100>, <110>, or <111>. In another embodiment, the magnetic material comprises magnetite. In another embodiment, the magnetic material comprises silane-coated magnetite. In another embodiment, the ion-exchange polymer comprises NAFION™. In another embodiment, the counter electrode comprises platinum mesh. In another embodiment, the device further comprising an electrolyte. In another embodiment, the device further comprising an acidic electrolyte. In another embodiment, the wherein the magnetic material and the ion-exchange polymer are disposed on less than the entire surface of the p-type semiconductor. In another embodiment, the device further comprising Ga—In eutectic. In another embodiment, the device further comprising silver epoxy. In another embodiment, the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>, and silane-coated magnetite. In another embodiment, the p-type semiconductor comprises a resistivity of about 1 to about 10 Ω-cm, the magnetic material comprises silane-coated magnetite, and the ion-exchange polymer comprises sulfonated fluoropolymer.
Another embodiment provides a device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, and surface orientation comprising <100>, <110>, or <111>; silane-coated magnetite; and ion-exchange polymer, wherein the magnetite and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode, wherein the counter electrode comprises platinum mesh; wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 1.6%, and wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −600 mV.
Another embodiment provides a first device for production of hydrogen gas comprising: at least one working electrode, wherein the working electrode comprises at least one magnetically-modified semiconductor electrode; and at least one first counter electrode, wherein the onset of hydrogen gas evolution for the first device occurs at a higher potential than for a second device comprising: a second working electrode that does not comprise a magnetically-modified electrode; and at least one second counter electrode, wherein said first counter electrode and said second counter electrode are identical. In another embodiment, the onset of hydrogen gas evolution of the first device is measured using 1 M HNO₃electrolyte and the onset of hydrogen gas evolution of the second device is measured using 1 M NaOH electrolyte.
Another embodiment provides a device adapted for photo applications comprising: at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and at least one counter electrode.
Another embodiment provides a method comprising: providing at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; providing at least one counter electrode; and using the working electrode and counter electrode upon exposure to photons.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows embodiments of a magnetically modified semiconductor electrode at different stages of manufacture. The p-silicon electrode on the left shows a completed electrode, where the back is covered with silicon adhesive. The p-silicon electrode on the right shows the electrode before the silicon adhesive is applied and the Ga—In contact with silver epoxy and electric wire is visible.

FIG. 2 shows one embodiment of a three-electrode electrochemical cell for the evolution of hydrogen.

FIG. 3 shows current responses as functions of the potential applied to magnetically modified semiconductor electrodes (MMSEs) and non-MMSEs in 0.1 M HNO₃(Acid) and 0.1 M NaOH (Base) under 20 mW/cm²illumination, for some embodiments.

FIG. 4 shows the light energy converting efficiency of the MMSE of FIG. 3 as a function of applied potential, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 5 shows a potential sweep voltammogram of an MMSE and a non-MMSE in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 6 shows light energy converting efficiency of the MMSE of FIG. 5 as a function of applied potential for some embodiments.

FIG. 7 shows potential sweep voltammograms for three <100> p-silicon electrodes, one with no coating (non-MMSE), one coated with a NAFION suspension (Naf-MSE), and one with a coating prepared with a 15% v/v magnetic nanoparticle suspension (MMSE), evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 8 shows the potentials required to achieve 0.12 mA/cm²for four <100> p-silicon electrodes, one with no coating (Bare), one with a coating containing no magnetic nanoparticles, and two MMSEs with coatings prepared with 15% v/v and 20% v/v magnetic nanoparticle suspensions, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 9 shows potential sweep voltammograms for two <110> p-silicon electrodes, one an MMSE with a coating prepared with a 20% v/v magnetic nanoparticle suspension, and the other with no coating, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 10 shows potential sweep voltammograms for two <111> p-silicon electrodes, one an MMSE with a coating prepared with a 20% v/v magnetic nanoparticle suspension, and the other with no coating, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 11 shows the potentials required to achieve 0.2 mA/cm²as a function of wafer crystallographic orientation for two <100> p-silicon electrodes, one an MMSE electrode with a coating prepared with 20% v/v magnetic nanoparticles and the other with no coating, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 12 shows potential sweep voltammograms for three <100> p-silicon electrodes doped to give resistivities of 1-4 Ω-cm, 0.5-1.5 Ω-cm, and 0.001 Ω-cm, coated with a suspension containing 20% v/v magnetic nanoparticles, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 13 shows open circuit potentials for <100> p-silicon electrodes, one a MMSE with a coating prepared with a 20% v/v magnetic nanoparticle suspension, and the other a non-coated electrode, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 14 shows potential sweep voltammograms for an electrode with an attached discrete magnet (Mag) and a non-magnetically modified electrode (non), evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 15 shows potential sweep voltammograms for three platinum electrodes, one with no coating (Pt), one with a coating containing no magnetic nanoparticles (Nafion), and one MMSEs with a coating prepared with a 20% v/v magnetic nanoparticle suspension, evaluated in 1 M HNO₃(pH 0) under 20 mW/cm²illumination for some embodiments.

FIG. 16 illustrates further additional embodiments for additional figures.

FIG. 17 illustrates band gaps for semiconductors.

FIG. 18 illustrates a photoelectrochemical hydrogen evolution reaction.

FIG. 19 illustrates semiconductor electrodes.

FIG. 20 illustrates improved kinetics.

FIG. 21 illustrates magnetically modified semiconductor electrode.

FIG. 22 illustrates semiconductor wafer use.

FIG. 23 illustrates experimental settings.

FIG. 24 illustrates influence of pH.

FIG. 25 illustrates influence of crystallographic plane of surface.

FIG. 26 illustrates doping.

FIG. 27 illustrates influence of doping level.

FIG. 28 illustrates energy conversion efficiency.

FIG. 29 illustrates influence of magnetic particle content.

FIG. 30 illustrates effect of magnetic modification on proton concentration.

FIG. 31 illustrates effect of magnetic modification.

FIG. 32 provides a summary of some embodiments.

DETAILED DESCRIPTION

Introduction

The present application describes, for example, a system in which an externally-applied magnetic field is established at a magnetically-modified semiconductor surface. Upon irradiation with light, the magnetically-modified electrode, deployed in an electrochemical cell, can produce higher currents than similar cells that are irradiated in the absence of an applied magnetic field. The higher currents correspond to higher efficiency that can be manifested in various electrochemical systems, including, for example, photovoltaics, solar cells, and photoelectrosynthetic and photocatalytic systems. Though not wishing to be bound by a particular theory, it is thought that the magnetic field on the semiconductor electrode surface may reduce the recombination rates and increase interfacial electron transfer rate, and that both of these phenomena lead to increased efficiency.
Technical literature re magnetic materials includes U.S. Pat. No. 7,041,401; US Patent Publication 2003/0232223; WO 0199127A2 and the reference Materials Chemistry and Physics, 2008, 108(1), 147.

Electrochemical Cell

Electrochemical cells are known in the art. See, for example, Handbook of Electrochemistry, Cynthia G. Zoski, Ed., Elsevier Science, 2007, which is incorporated by reference in its entirety. An electrochemical cell can be a device that can be used to produce voltage and current from chemical reactions that involve oxidation and reduction. An electrochemical cell commonly includes two half cells with a characteristic voltage. Oxidation occurs at one half cell, while reduction occurs at the other half cell. Each half cell typically includes an electrode, i.e., an electrical conductor, and an electrolyte, i.e., a medium that has free ions through which ion current can flow. The half cells may use the same electrolyte, or may use different electrolytes in conjunction with an ion-carrying bridge between the two electrolytes. Electrolytes may comprise ionic solutions, molten solutions, gases, solids, ion exchange polymers, acids, bases, or salts.

Electrodes

Electrodes used in electrochemical cells can include, for example: the working electrode; the counter electrode; and the reference electrode. The working electrode in an electrochemical cell is the electrode at which the electrochemical reaction of interest takes place, e.g., hydrogen production. The working electrode can be referred to as either the cathode or the anode, depending on whether the reaction of interest is a reduction reaction or an oxidation reaction. Examples of working electrodes of the present application are shown in FIG. 1. In FIG. 1, the electrode on the right shows one embodiment of the present application wherein a p-Si silicon chip is in contact with Ga-In eutectic on a back surface of the chip, and the Ga—In eutectic is in contact with silver epoxy and an electric wire. The electrode on the left shows the same electrode construction as the electrode on the right, but with an additional layer of silicon adhesive applied to the back of the silicon chip, as shown, for mechanical support. The working electrode can comprise p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>, or mixtures thereof.
The counter electrode, also known as the auxiliary electrode, is the electrode in an electrochemical cell that acts as a source of, or sink for, electrons in the electrochemical circuit formed with the working electrode. The counter electrode is commonly made of an electrochemically-inert material, such as, for example, graphite, carbon cloth, platinum, or platinum mesh.
The reference electrode is typically an electrode with a well-known electrode potential that is stable. Because of its stable potential, the reference electrode can be used in an electrochemical cell to measure the potential of the working electrode. A saturated calomel electrode (“SCE”) is a common reference electrode. Another common reference electrode is the normal hydrogen electrode (“NHE”).

Semiconductors in Electrochemical Devices

Semiconductors can be used for solar-to-chemical and solar-to-electrical energy conversion because semiconductor materials capture photons and separate charges. See, for example, A. J. Bard and L. R. Faulkner, “Photoelectrochemistry and Electrogenerated Chemiluminescence,” Chapter 18 of Electrochemical Methods: Fundamentals and Applications, 2^ndEd., Wiley, New York, 2001, which is incorporated by reference in its entirety. There are two types of semiconductors: p and n. The p-type semiconductors generate electrons from light that can be used to electrochemically reduce molecular and atomic substrates in the electrolyte. Such substrates may be ions or neutrally charged. The electrons, minority carriers in p-type semiconductors, can be dragged to the electrode surface by electrostatic forces generated by the potential dropin the space charge region. Semiconductors suitable for use in embodiments of the present application can include, for example, undoped or doped p-type silicon wafers. Resistivity of suitable silicon wafers with low doping level may typically be about 1-10 Ω-cm or higher, depending upon purity, while highly doped silicon wafers with resistivity below 1 Ω-cm, for example, from 0.01-1 Ω-cm, have also been found to be suitable. Other suitable semiconductors may be used, such as, for example, semiconductors suitable for use in an electrochemical cell with two semiconductor electrodes, e.g., one p-type and one n-type semiconductor, such as, for example for a closed-loop power source that can generate fuel photoelectrosynthetically and discharge fuel to yield power during the night, or polycrystalline semiconductor powders, such as, for example mixed with magnetic particles, such as, for example, magnetite.

Magnetically Modified Semiconductor Electrode (“MMSE”)

Magnetically modified electrodes are known in the art and have been described in, for example, U.S. Pat. No. 6,949,179, issued Sep. 27, 2005 to Leddy, et al., which is incorporated by reference in its entirety. A Magnetically Modified Semiconductor Electrode (“MMSE”) can comprise a semiconductor electrode incorporating magnetic particles. In some embodiments, the magnetic particles may be located in a portion of the interior of the electrode, such as, for example, the core of the electrode, a single interior region of the electrode, or multiple interior regions of the electrode. In some embodiments, the magnetic particles may be located on the outer surface of the electrode, such as, for example, as a coating on the entire outer surface of the electrode, as a coating on a single portion of the outer surface of the electrode, or as a coating on multiple portions of the outer surface of the electrode. In some embodiments the magnetic coating may comprise a composite comprising magnetic particles and a binder. Such a binder may be inert, porous, or electrically conductive. In some cases, the binder may comprise a polymer such as, for example, an ionomer such as, for example, a sulfonated fluoropolymer, such as NAFION™ (DuPont). Ion-exchange polymers may also be suitable, such as water-permeable proton-exchange or cation-exchange polymers. In some embodiments, the magnetic particles can be dispersed throughout the electrode. In some embodiments, the magnetic particles can be located in some combination of interior locations and surface locations of the electrode.
Magnetic materials, magnetic particles, and coatings for magnetic particles are known in the art. Examples of magnetic materials, magnetic particles, and coatings for magnetic particles are described in U.S. Pat. No. 7,585,543, issued Sep. 9, 2009 to Leddy et al., which is incorporated herein by reference in its entirety. Magnetic particles that can be used in embodiments of the present invention can include, for example, magnetite (Fe₃O₄), nickel, samarium cobalt, and neodymium iron boron. Some magnetic particles can be unstable in acids. Therefore, if they are to be exposed to an acid environment, such particles can be coated with a protective acid-stable material, while still retaining their magnetic properties. A coating known to one skilled in the art that is inert, acid-stable, and allows for the retention of magnetic properties may be used to coat magnetic particles for use in embodiments of the present application. Commonly, magnetic particles employed in embodiments of the present application may be coated with, for example, silane.

Photoelectrosynthetic Cells

Photoelectrosynthesis involves the conversion of light energy into chemical energy. Generation of hydrogen gas by solar irradiation of a semiconductor in an electrochemical cell is an example of a photoelectrosynthetic process. Photovoltaic systems, such as solar cells, convert light energy to electrical energy. Photoelectrocatalysis involves the use of solar energy to catalyze reactions during chemical syntheses. Embodiments of the present application can be applied in devices dependent upon photoelectrosynthesis, photovoltaic systems, and photoelectrocatalysis systems.
When constructing photoelectrosynthetic cells to drive synthetic reduction reactions, it is desirable for the redox potential for the synthesis in the electrolyte to be below the conduction band potential of the p-type semiconductor. Thermodynamically, hydrogen gas is generated at the potential of 0 V versus Normal Hydrogen Electrodes (“NHE”), which is equivalent to −4.5 eV on the vacuum scale. The potentials of the conduction and valance band of silicon are −4.0 eV and −5.1 eV on the vacuum scale, respectively. Therefore, photoelectrosynthetic cells for hydrogen reduction may be thermodynamically feasible with p-type silicon semiconductors. However, the recombination of generated and separated charges in and on the semiconductor can dramatically reduce reaction efficiency, such that hydrogen evolution occurs at high overpotential (i.e., high energetic cost). To reduce recombination, extant semiconductor electrodes can be modified with surface coatings consisting of various precious metals, thus generating catalytic reaction sites. Conversely, magnetic modification of p-type silicon semiconductor surfaces can enable hydrogen production with little or no overpotential and, advantageously, does not require the use of precious metals or catalysts, which can be expensive and difficult to acquire.
The present application provides, for example, a system and a process for applying magnetic fields to semiconductor electrodes. Though not wishing to be bound by any particular theory, it is thought that the magnetic field on the semiconductor electrode may reduce recombination rates and increase overall reaction rates and efficiency of semiconductor electrodes. Magnetic fields can be applied by coating the front and/or back side of the semiconductor electrode with magnetic material. In some embodiments, it may be desirable to coat only a portion of the electrode front or back surface. In some embodiments, a hole may be made in the coating surface after it has been applied to the electrode. In some embodiments, the electrode may be formed from semiconductor/magnetic material composites which can effectively establish the magnetic field from within the semiconductor. In some embodiments of the present application, the semiconductor electrode can comprise materials that are doped. In some embodiments of the present application, composites of polycrystalline semiconductor and magnetic microparticles, such as, for example, magnetite, can be applied to the electrode surface. In some embodiments, magnetic fields are uniform. In some embodiments, magnetic fields may be nonuniform.
In one embodiment, a cell comprising a magnetically-modified semiconductor working electrode and a platinum counter electrode in electrolyte solution, when under illumination by visible light, can produce hydrogen at substantially more positive potentials (i.e., at lower overpotential, lower applied driving force, and lower energetic cost) than similar but nonmagnetic cells. In some embodiments, the overpotential of a cell comprising a magnetically-modified electrode relative to a similar cell that does not comprise a magnetically-modified electrode, measured at a current density of about 0.4 mA/cm², may be no more than about −1200 mV, or no more than about −600 mV, or no more than about −500 mV. That is, hydrogen could be produced in a cell comprising a magnetically-modified electrode at about 1200 mV lower potential, or about 600 mV lower potential, or about 500 mV lower potential, than a similar cell that does not comprise a magnetically-modified electrode.
In another embodiment, a cell comprising a magnetically-modified semiconductor working electrode in acidic media, when under illumination by sunlight, can produce hydrogen at approximately 1 V lower potential than similar cells not comprising the magnetically-modified semiconductor working electrode.
In another embodiment, magnetically-modified p-type semiconductor electrodes can be fabricated by coating the surface of the electrode with magnetic particles embedded in a proton-exchange polymer, such as, for example, sulfonated fluoropolymers, including NAFION™ (Dupont), ACIPLEX™ (Asahi Glass), FLEMION™ (Asahi Glass), and the like.
In another embodiment, a magnetic field can be applied on a surface of a p-type semiconductor electrode. Magnetic particles such as, for example, silane-coated magnetite particles, may be mixed with an ion-exchange polymer such as, for example, NAFION™, and coated on the surface of the electrode. The ion-exchange polymer may function as a mechanical support for the magnetic particles, as well as an ion conductor to complete the circuit of the electrochemical cell. Other suitable ion-exchange polymers known in the art may be used in embodiments of the present application, such as styrenated copolymers, poly(arylene ethers) including sulfonated aromatic polymers, polyimides including sulfonated polyimides, and the like. (See, for example, Hickner et al., “Alternative Polymer. Systems for Proton Exchange Membranes,” Chem. Rev., 2004, 104, 4587-4612, which is incorporated by reference in its entirety.) Although sulfonated polymers are commonly used as ion-exchange polymers, other proton conducing moieties, such as phosphonic acid or phosphinic acid moieties, may, be employed.

Device Testing

A three-electrode electrochemical cell, as shown in FIG. 2, can be used to measure the potential and current profile of hydrogen gas evolution. The working electrode can be the MMSE as described above, or a non-magnetically-modified semiconductor electrode (“non-MMSE”) known in the art. The counter electrode can be a suitable counter electrode, such as, for example, a high surface area platinum wire mesh. A suitable reference electrode can be used, such as, for example, a saturated calomel electrode (“SCE”). Various acidic and basic aqueous solutions can be prepared and used to study the pH dependence of hydrogen evolution on the semiconductor electrodes. For acidic solutions, H₂SO₄or HNO₃solutions typically may be used, for basic solutions, NaOH solutions typically may be used. The electrodes can be irradiated with a light source capable of producing intensity of at least about 20 mW/cm².
Although the efficiency calculations for photoelectrochemical cells are still controversial, the most widely-used method (shown and described below) is used to calculate efficiencies in the present application. Photoconversion efficiency of three-electrode electrochemical cells can be calculated with the following equation:
$η (%) = 100 \frac{j_{p} ⌊ E_{rev}^{0} - E_{app} ⌋}{I_{0}}$
where I_ois the intensity of the incident light, the photocurrent, j_p, is the current measured at the applied potential, E_app, and assuming that open circuit potential is 800 mV vs. SCE. E⁰ _revis the thermodynamic potential of the hydrogen evolution reaction which is 0 V vs. the normal hydrogen electrode (“NHE”) and −244 mV vs. SCE.

WORKING EXAMPLES

Various claimed embodiments are described further with use of non-limiting working examples.

Example 1

Magnetic Particle Preparation

Fe₃O₄magnetite particles (Aldrich, <5 μm) were ball-milled in hexane for 20 minutes to obtain uniform particle size. The ball milled magnetite particles (1 g) were then added to a mixture of toluene (10 mL) and (3-aminopropyl)trimethoxysilane (10 mL; Alfa Aesar) in a 20 mL screw-cap vial, and the mixture was agitated by slow rotation of the vial for 4 hours. Next, the supernatant was decanted, and the particles were washed with toluene (3×20 mL). The particles were dried overnight in a vacuum oven at 70° C. To the dried particles in the vial were added ethylene glycol diglycidyl ether (10 mL) and deionized water (10 mL), and the mixture was agitated by slow rotation of the vial for 4 hours. The particles were then rinsed with deionized water (3×20 mL), dried overnight in a vacuum oven at 70° C., and stored in a capped vial.

Example 2

Electrode Preparation

Semiconductor electrodes (5 mm×5 mm) were cut from a 3 inch p-type silicon wafer with a <100> surface orientation and low doping level, characterized as having resistivity of 1 to 10 Ω-cm. The silicon semiconductor squares were cleaned to remove organic residuals as follows:
1. Purged 25 mL deionized water with nitrogen gas for about 5 minutes.
2. Added 5 mL NH₄OH (27%).
3. Heated to 70° C.
4. Added 5 mL H₂O₂(30%).
5. Waited for about 1 minute until the solution bubbled vigorously.
6. Soaked the silicon pieces in the cleaning solution for 15 minutes.
7. Extracted the silicon pieces from the solution with Teflon tweezers.
8. Rinsed the cleaned silicon pieces with flowing, nitrogen-purged, deionized water for about 30 seconds.
9. Dried the silicon pieces with nitrogen gas.
The p-type silicon pieces were then etched with HF solution to remove the surface oxide layer by the following procedure:
1. Purged 120 mL of deionized water with nitrogen gas.
2. Added 5 mL HF (49%).
3. Soaked the silicon pieces in the etching solution for 1-2 minutes.
4. Rinsed the etched silicon pieces with nitrogen-purged deionized water for about 30 seconds.
5. Dried the etched silicon pieces with nitrogen.
Ga—In eutectic was brushed onto the completely cleaned and etched silicon surface to achieve ohmic contact with an electric wire, which was attached using conductive silver epoxy. The electric wire side became the back side of the electrode, whereas the opposite side was the electroactive electrode surface. The silver epoxy was cured for 4 hours. The exposed silver epoxy and silicon on the back side were covered with silicon adhesive, exposing only the electrode surface. The adhesive was allowed to set for 6 hours.
After the electric wire was attached on the back side of the silicon piece, the electrode surface was cleaned and etched again following the procedures described above. Next, the electrode surface was coated with a suspension of NAFION™ (3.60 μl; Aldrich, 5% wt/vol, 1,100 eq wt) containing magnetic microparticles (prepared as described in Example 1) at 15% v/v in suspension. The coating was dried within an external magnetic field to form a 4 μm-thick coating on the electrode surface. The external magnetic field was applied with an NdFeB ring magnet (o.d.=3.0 in.; i.d.=1.5 in.; and 0.5 in. thickness). The formulation of NAFION™ with magnetic particle composite used was as follows:
13.3 mg silane-coated magnetite (Fe₃O₄) particles (<5 μm)
0.614 mL NAFION™ (Aldrich) suspension

Example 3

MMSE and Non-MMSE Device Testing

A three-electrode electrochemical cell, as shown in FIG. 2, was used to measure the potential and current profile of hydrogen evolution. The working electrode was either a 0.25 cm²p-type silicon MMSE prepared as described in Example 2, or a 0.25 cm²p-type silicon non-MMSE. The counter electrode was a high surface area platinum wire mesh, and the reference electrode was a saturated calomel electrode (“SCE”). The pH 1 solution was made with 0.1 M nitric acid, the pH 13 solution was made with 0.1 M NaOH. The electrodes were irradiated with an Oriel 150W Xe lamp to supply a light intensity of about 20 mW/cm². The pH dependence of hydrogen evolution on the MMSEs and non-MMSEs as voltage was polarized and current was measured is shown in FIG. 3. Note that for all of these 0.25 cm²electrodes, an indicated current of 1 mA corresponds to a current density of 0.25 mA/cm².
FIG. 3 shows that in acidic solution (pH 1), the onset of hydrogen evolution at the non-MMSE was at about −900 mV, with the onset measured at about 0.1 mA. For the MMSE, the onset of hydrogen evolution, also measured at about 0.1 mA, occurred at about +400 mV, for a reduction in overpotential of about 1.3 V. The current amplifier overloaded at −600 mV and 1 mA for the MMSE, whereas the current amplifier overload for the non-MMSE was approximately 20-fold lower. This suggests that the MMSE device exhibited significantly higher efficiency than the non-MMSE device. By comparing the MMSE and non-MMSE data in FIG. 3, the effect of an applied magnetic field on hydrogen evolution is evident. Because hydrogen evolution at the MMSE occurred at a much less negative potential than the non-MMSE, evolution of hydrogen requires much less energy input for the MMSE relative to the non-MMSE.
Data in FIG. 3 show that in basic solution (pH 13), the measured effects were analogous to those seen in acidic solution. The change in pH shifted the potentials to negative values, as expected from the Nernst equation. When the solution is acidic, hydrogen evolves at a less negative potential for both MMSEs and non-MMSEs, as there is a high proton concentration in acidic solutions that affects the hydrogen evolution potential. According to the Nernst equation, when pH decreases by 1, hydrogen evolution potential shifts +59.1 mV, thus, it was expected that the potential would shift by about 600 mV in the basic solution. For the MMSE, the onset potential, measured at about 0.1 mA, was about −400 mV, whereas the non-MMSE onset potential, also measured at about 0.1 mA, was about −1400 mV. This result was consistent with a significant increase in efficiency for the MMSE device. It is notable that data in FIG. 3 show that the efficiency for hydrogen evolution in base for the MMSE was even higher than the efficiency for hydrogen evolution in acid for the non MMSE.
Photoconversion efficiencies were calculated using the equation described above in the Detailed Description. FIG. 3 shows that only the MMSE in acidic solution demonstrated an economically feasible light energy conversion efficiency. Note that in order to evolve hydrogen with either the non-MMSE electrode in acidic or basic solution, or the MMSE in basic solution, more potential would have to be applied than can be obtained from the evolved hydrogen fuel, i.e., E_rev ⁰<E_app, so the resulting efficiency is negative.
The calculated efficiency profile of an MMSE in 0.1 M HNO₃is shown in FIG. 4. As shown in FIG. 4, the maximum photoconversion efficiency of 1.6% was achieved at 375 mV of applied potential. Higher efficiencies might be achieved with more highly-doped semiconductors.
Potential sweep voltammograms (PSV) were produced using a potentiostat (CHinstrument 760b) to monitor potential and current profiles of the hydrogen evolution reaction. The scan rate was 100mV/s and the potential was decreased from +2 V to −2 V to drive the reduction reaction. Open circuit potential measurements were also performed to measure maximum photo voltage (V_max) using the same potentiostat.
FIG. 5 shows a potential sweep voltammogram of a MMSE and a non-MMSE in 1 M HNO₃(pH 0) under 20 mW/cm²illumination. When the non-MMSE was illuminated, hydrogen evolution started at −400 mV with an externally-applied potential higher than 1230 mV, which is the reversible potential of hydrogen/oxygen fuel cells. This demonstrates that without magnetic modification, hydrogen evolution on the non-MMSE is not beneficial, either energetically or economically. However, for the MMSE, hydrogen evolution began at 800 mV with less externally-applied potential and a reasonable light conversion efficiency was realized.
FIG. 6 shows light energy converting efficiency versus applied potential for a MMSE in 1M HNO₃. Efficiency was calculated as described in the Detailed Description, using the following equation:
$η (%) = 100 \frac{j_{p} ⌊ E_{rev}^{0} - E_{app} ⌋}{I_{0}}$
As shown in FIG. 6, the calculated maximum efficiency of 6.2% was achieved when the applied potential was 700 mV.
Results are summarized in Table 1. Current densities of 0.2, 1, 2, and 3 mA/cm²correspond to currents of 0.05, 0.25, 0.5, and 0.75 mA for these 0.25 cm²electrodes. Potentials are reported relative to SCE.

TABLE 1

pH dependence of hydrogen evolution for non-MMSE and 15 vol % MMSE

	E(V) for Hydrogen Evolution at Current Densities
Experimental set up	of 0.2, 1, 2, and 3 mA/cm²

pH	surface	resistivity	Modif	loading	0.2	1	2	3

0	<100>	1-4	Mag	15	0.689	0.492	0.318	0.165
1	<100>	1-4	Mag	15	0.455	0.042	−0.331	−0.656
13	<100>	1-4	Mag	15	−0.25	−1.025	−1.171	−1.174
0	<100>	1-4	Bare		−0.584	−0.748	−0.779	−0.823
1	<100>	1-4	Bare		−0.771	−1.123	−1.322	−1.485
13	<100>	1-4	Bare		−1.359	−1.743	−1.911	−1.923

Loading is volume percent of particles.

Example 4

Effect of Varying Magnetic Particle Loading

Working electrodes were prepared as in Example 2, except that coated electrodes were prepared using either a suspension containing 20% v/v of magnetic particles or a suspension with no magnetic particles. To prepare the 20% suspension, 13.3 mg of silane-coated magnetite (Fe₃O₄) particles (<5 μm) were mixed with 0.400 mL of NAFION™ (Aldrich) suspension. The 0% suspension consisted of only the NAFION™ suspension. 3.40 μL of each suspension was applied to electrode surfaces and dry coated. Electrode performance was evaluated in a 1M nitric acid (pH 0) solution.
FIG. 7 shows potential sweep voltammograms of the electrodes under illumination of 20 mW/cm². The NAFION™-only coated electrode (Naf-MMSE) exhibited enhanced current output compared to the uncoated electrode (non-MMSE), while the 20% magnetically-modified electrode (MMSE) exhibited enhanced current output compared to the NAFION™-only coated electrode.
A comparison of the voltammogram for 15% magnetic particle loading (FIG. 1) to that for 20% loading (FIG. 7) shows that 15% loading achieved higher current output than that for 20% loading. FIG. 8 summarizes the 0.12 mA/cm²potentials for bare electrodes and for suspension-coated electrodes at 0, 15, and 20% magnetic particle loading, with 15% loading having achieved the most positive potential. While not wishing to be bound by theory, it is possible that at higher magnetic particle loadings, less electrode area is exposed to incident light, decreasing electrode performance.
Results are summarized in Table 2. Current densities of 0.2, 1, 2, and 3 mA/cm²correspond to currents of 0.05, 0.25, 0.5, and 0.75 mA for these 0.25 cm²electrodes. Potentials are reported relative to SCE.

TABLE 2

pH Magnetic particle loading dependence of Hydrogen Evolution on MMSEs.

pH	surface	resistivity	Modif	loading	0.2	1	2	3

0	<100>	1-4	Bare	na	−0.584	−0.748	−0.779	−0.823
0	<100>	1-4	Naf	0	−0.397	−0.539	−0.623	−0.687
0	<100>	1-4	Mag	15	0.689	0.492	0.318	0.165
0	<100>	1-4	Mag	20	−0.146	−0.324	−0.533	−0.591

Loading is volume percent of particles.

Example 5

Effect of Varying Crystallographic Planes

Working electrodes were prepared from p-type silicon wafers with <100>, <110>, and <111> surfaces as described in Example 2, except that the coated electrodes were prepared using a suspension containing 20% v/v of magnetic microparticles as described in Example 4. The p-type silicon wafer used was characterized as having a resistivity of 1 to 6 Ω-cm. Electrode performance was evaluated in a 1M nitric acid (pH 0) solution.
FIGS. 7 and 9-11 show that photochemical hydrogen evolution was affected by magnetic field regardless of wafer crystallographic orientation. FIG. 11 illustrates potentials of MMSEs and non-MMSEs with <100>, <110> and <111> surfaces at a current density 0.2 mA/cm². Hydrogen evolution occurred at 480 to 580 mV lower potential for MMSEs than for non-MMSEs.
Results are summarized in Table 3. Current densities of 0.2, 1, 2, and 3 mA/cm²correspond to currents of 0.05, 0.25, 0.5, and 0.75 mA for these 0.25 cm²electrodes. Potentials are reported relative to SCE.

TABLE 3

pH dependence of H₂evolution for 20 vol % MMSEs with various surface orientations.

pH	surface	resistivity	Modif	loading	0.2	1	2	3

0	<100>	1-4	Mag	20	−0.146	−0.324	−0.533	−0.591
0	<110>	2-4	Mag	20	−0.106	−0.278	−0.38	−0.416
0	<111>	3-6	Mag	20	−0.134	−0.334	−0.339	−0.436
0	<100>	1-4	Bare		−0.63	−0.855	−0.929	−0.97
0	<110>	2-4	Bare		−0.673	−0.867	−0.942	−0.98
0	<111>	3-6	Bare		−0.695	−0.88	−0.931	−0.96

Loading is volume percent of particles.

Example 6

Effect of Varying Resistivity/Doping Level

Working electrodes were prepared as described in Example 2 using <100> p-type silicon with resistivities of 1-4 Ωcm, 0.5-1.5 Ωcm, and 0.001 Ωcm, by using silicon that had been doped at different levels. Coatings were prepared using a suspension containing 20% v/v of magnetic microparticles as described in Example 4. Electrode performance was evaluated in a 1M nitric acid (pH 0) solution.
FIG. 12 shows potential sweep voltammograms of the electrodes under illumination of 20 mW/cm. The electrode with resistivity of 0.5-1.5 Ωcm exhibited superior performance to the electrodes with higher and lower resistivities. While not wishing to be bound by theory, it is possible that the highly-doped electrode with resistivity of 0.001 Ωcm was too conductive to exhibit semiconductor properties under the testing conditions.
Results are summarized in Table 4. Current densities of 0.2, 1, 2, and 3 mA/cm²correspond to currents of 0.05, 0.25, 0.5, and 0.75 mA for these 0.25 cm²electrodes. Potentials are reported relative to SCE.

TABLE 4

Doping level dependence of Hydrogen Evolution on MMSEs with <100> surface.

pH	surface	resistivity	Modif	loading	@ 0.2	1	2	3

0	<100>	1-4	Mag	20	−0.146	−0.324	−0.533	−0.591
0	<100>	0.5-1.5	Mag	20	−0.092	−0.295	−0.401	−0.454
0	<100>	0.001	Mag	20	−0.483	−0.895	−0.976	N/A

Loading is volume percent of particles.

Example 7

Open Circuit Potentials

FIG. 13 shows the open circuit potentials for the MMSE of Example 4, which was modified with 20% v/v magnetic nanoparticles, and a non-MMSE, both evaluated in 1M nitric acid (pH 0) solution, under 20 mW/cm²illumination and under darkened conditions. Potentials are shown relative to an NHE reference electrode and therefore correspond to both V_ocand V_max. The maximum potential for the illuminated MMSE was 0.314 V, while that for the illuminated non-MMSE was 0.206 V. It is notable that even under darkened conditions, with illumination levels below that detectable by the light detector, the MMSE still showed a positive potential of 0.073V relative to the NHE reference electrode.

Example 8

Discrete Magnets

The working electrode was fabricated from a 0.25 cm²p-type silicon with <100> surface and 1-4 Ωcm as described in Example 2, but with no coating being applied. A plastic box (0.5×0.5×3 inch) was attached to the back side of the electrode using epoxy resin. A small NdFeB ring magnet was inserted into the box. Distance from the magnet to the front surface was 1.5 mm. Electrode performance was evaluated in 1M nitric acid (pH 0) solution.
As shown in FIG. 14, enhanced current was observed when using the ring magnet, but the effect was not as great as when surface-applied magnetic coatings were used. While not wishing to be bound by theory, these results suggest that differences in the gradients of the magnetic fields could have affected the system.

Example 9

Comparison of Silicon and Platinum Electrodes

Platinum working electrodes were fabricated using a either suspension containing 20% v/v of magnetic particles or a suspension with no magnetic particles, as described in Example 4. Performance of these electrodes, as well as uncoated platinum electrodes, were evaluated in a 1M nitric acid (pH 0) solution.
FIG. 15 shows current potential profiles of hydrogen evolution on uncoated platinum, magnetically modified platinum (Magnetic) and NAFION™-modified platinum (Nafion) electrodes. As can be seen in the figure, these three electrodes have similar current potential characteristics. While not wishing to be bound by theory, these results suggest that magnetic or NAFION™ modifications do not affect proton or hydrogen activity significantly. This is consistent with the results of Table 5, which show that measured open circuit potentials are also similar for the three electrodes.

TABLE 5

Open Circuit Potential of Pt Electrodes in 1M HNO₃
solution with hydrogen purging.

	V_oc	V_ocvs. SCE	V_ocvs. NHE

Bare	−0.251	−0.009
NAFION ™	−0.255	−0.013
Magnetic	−0.256	−0.014

Further Additional Embodiments

U.S. Provisional Application Ser. No. 61/282,467 filed Feb. 16, 2010 is hereby incorporated by reference in its entirety including the figures, working examples, tables, and claims. Additional background and additional embodiments are provided in the following section, before the section with claims. In addition, additional references 1-35 are also cited in the additional text.

1 BACKGROUND

As demands for sustainable energy increase, molecular hydrogen becomes a strong candidate as a next generation fuel. However, hydrogen as a fuel is not generally realized because its production and environmental cost is higher than that of petroleum energy even with the most advanced technologies. Among the technologies, a photoelectrochemical cell is a system that can produce hydrogen in a sustainable way with high efficiency because it directly converts solar energy to chemical energy stored as hydrogen.
Since Fujishima and Honda first demonstrated photoelectrochemical water splitting [1], research has been conducted to improve dynamics and properties of semiconductor photoelectrochemistry. For example, Allen J. BARD et. al. suggested a conceptual design for a optimized electrochemical structure of a photoelectrochemical cell [2] and Nathan Lewis physicochemically explained and reviewed the heterogeneous electron transfer kinetic between a semiconductor electrode surface and an electrolyte [3]. In addition, John A. Turner also reviewed on hydrogen production and demonstrated a photoelectrochemical water splitting system using two different semiconductors for a photoanode and a photocathode [4].
It can be conceptually proved that an electrochemical cell can provide up to 17% of energy conversion efficiency to electrical energy and 10% of conversion efficiency to fuels [6]. A key component of the photoelectrochemical cell is a semiconductor electrode. Silicon as a photoelectrode can provide an excellent thermodynamic properties for energy conversion among various types of semiconductors because it can have an optimized band gap to harvest solar spectrum. To produce molecular hydrogen by reducing water, a p-type semiconductor is essential as its carriers, holes, and is able to drive photoreduction.
The major limitations of the photoelectrochemical hydrogen evolution on the p-Si photocathode are that the heterogeneous electron transfer kinetics from photo generated holes in a p-type silicon to adsorbed protons on the surface of the p-type silicon, are too slow to make this photoelectrochemical reaction occur. Slow kinetics could due to surface recombination between photogenerated holes and electrons or due to sluggish heterogeneous electron transfer rate. Therefore, if there is a method to increase the reaction rate on p-Si, hydrogen can be produced at much lesser cost by solar energy.
There have been various approaches to improve the reaction rate, and a notable modification to the p-Si surface was performed by M. Szklarczykt and J. O'M. Bockris [7]. They coated surfaces with various metals catalysts and achieved a quantum efficiency of 1.4% with a nickel coating. Although coating with catalysts could make photoelectrosynthetic hydrogen evolution possible, a desirable efficiency and stability of coatings are difficult to achieve.
At room temperature, magnetic fields influence chemical systems through dynamics [8] [9] [10] [11] [12] [13] [14]. Effects on transport, including magnetohydrodynamics (MHD) and gradient magnetic field effects, have been achieved by placing electrodes in an externally applied magnetic field [15] [16] [17] [18] [19]. Incorporating a permanent magnetic field at the electrode surface can provide advantages of experimental simplicity and reduced weight. Magnetically modified electrodes can sustain a permanent magnetic field because magnetic microparticles are either attached to the electrode surface or incorporated into the electronic conductor of the electrode. In a number of systems, magnetic modification of electrode surfaces increases electrochemical flux as compared to electrodes without magnetic modification [20] [21] [22] [23] [24] [25] [26] [27] [28]. On the other hand, electron transfer kinetic enhancements of magnetically modified electrodes also reported when mass transport effect of MHDs or magnetic gradient effects are limited by trapping redox couples in polymer matrix or adsorbing redox couples on an electrode surface. Improved homogenous outer-sphere electron transfer kinetic of transition metal complex trapped in polymer matrix was demonstrated by Shelley Minteer [29]. Enhanced heterogeneous electron transfer kinetic was also reported when oxidation of adsorbed carbon monoxide occurs on magnetically modified Pt electrodes [30]. Further more, magnetic modifications to polymer electrolyte fuel cell electrodes, Nickel metal hydride and nickel cadmium batteries were reported as well. [31] [32]

2 EXPERIMENTAL

Here, it is further demonstrated that application of magnetic microparticles to p-Si markedly reduced the overpotential for hydrogen evolution under illumination. Potential sweep voltammograms were used to evaluate the current potential profile of hydrogen evolution of unmodified, Nafion only coated, and magnetic particle and Nafion composite coated p-Si photocathode under 20 mW/cm²illumination. Open circuit potentials (V_oc) under light and dark were also measured.
Potential sweep voltammograms of the magnetically modified electrodes show 1.2 V less negative onset potential for hydrogen evolution than the Nafion coated electrode and V_ocof magnetically modified p-Si show 0.07-0.10 V less positive potential than the Nafion coated electrode. When compared to unmodified p-Si, the magnetic modification has hydrogen evolution of 0.1 V less negative onset and V_ocof a bare p-Si electrode.
Although not limited by theory, a possible explanation for the observed enhancements is an increase in the heterogeneous electron transfer rate and/or a decrease in the surface recombination rate.
Chemically inert magnetic particles were prepared by encapsulating micrometer size magnetite particles by a monolayer to a few monolayers of siloxane. Magnetically modified semiconductor electrodes (MMSEs) were constructed by coating a composite of the magnetic particles and Nafion suspension on p-type silicon semiconductors. Some details about constructions and measurements are as follows.

2.1 Magnetic Particle Preparation

Analogues to previous work, home made magnetic microparticles, were prepared [33] [29] [30]. Fe₃O₄magnetite particles (Aldrich, <5 μm) were ball-milled in hexane for 20 minutes to obtain uniform particle size. The ball milled magnetite particles (1 g) were then added to a mixture of toluene (10 mL) and (3-aminopropyl) trimethoxysilane (10 mL; Alfa Aesar) in a 20 mL screw-cap vial, and the mixture was agitated by slow rotation of the vial for 4 hours. Next, the supernatant was decanted, and the particles were washed with toluene (3×20 mL). The particles were dried overnight in a vacuum oven at 70° C. To the dried particles in the vial were added ethylene glycol diglycidyl ether (10 mL) and deionized water (10 mL), and the mixture was agitated by slow rotation (−30 rpm) of the vial for 4 hours. The particles were then rinsed with deionized water (3×20 mL), dried overnight in a vacuum oven at 70° C., and stored in a capped vial. Thickness of the encapsulation was controlled by reagent and results were siloxane encapsulated magnetic particles less then 5 μm diameters. The sizes of magnetic particles are barely changed before and after coating because thickness of siloxane coating is only about 10 to 30 Å. There is differences of properties from batches by batches and coating thickness and size of the particle could vary by batches. However, typical diameter of the coated magnetic particles are about 4-5 μm.

2.2 Electrode Preparation

2.2.1 Semiconductor Electrode (SE) Preparation

Five kinds of p-type silicon were used to fabricate semiconductor electrodes. For p-Si with (100)-oriented surface, wafers of resistivity of 1-4 Ωcm (ITME), 0.5-1.5 Ωcm (ElCat) and 0.001-0.005 Ωcm (ElCat) were purchased. For p-Si with (110)-oriented and (111)-oriented surface, wafers of resistivity of 2-4 Ωcm (ElCat) and 3-6 Ωcm (ITME) were purchased, respectively. Refer to Table S-1 for details. Labels and specifications of p-Si wafers used to fabricate semiconductor electrodes are shown in the Table S-1 as well. Notations of a, b and c were categorized by resistivities, a is within 1-10 Ωcm, b is 0.5-1.5 Ωcm and c is 0.001-0.005 Ωcm.
To fabricate semiconductor electrodes, p-type silicon wafers were cut to 5 mm by 5 mm. The cut silicon semiconductor chips were cleaned to remove organic residues by industrial standard RCA-1 cleaning procedure followed by RCA-2 etching procedure [34].
Then, Ga—In eutectic (Aldrich, 99.99+%) was brushed onto the completely cleaned and etched silicon surface to achieve an ohmic contact with an electric wire, which was attached by conductive silver epoxy (Circuitworks, CW2400) [7].
The electric wire side becomes the back side of the electrode, whereas the opposite side is the electroactive electrode surface. The silver epoxy was cured for over 4 hours in oven at 70° C.

TABLE S-1

Labels and specifications of p-Si wafers
used to fabricate semiconductor electrodes.

Label	Surface	Resistivity (Ω cm)	Thickness (μm)	Vendor

100a	(100)	1-4	200	ITME
100b	(100)	0.5-1.5	355-405	ElCat
100c	(100)	0.001-0.005	380 ± 52	ITME
110a	(110)	2-4	355-405	ElCat
111a	(111)	3-6	380 ± 25	ITME

Notations of a, b and c are categorized by resistivities.
a is within 1-10, b is 0.5-1.5 and c is 0.001-0.005 Ω cm

The exposed silver epoxy and silicon on the back side was completely covered by nonconductive epoxy resin (LOCTITE, 1C). Then, the exposed area was confined to the front side that is the electrode surface. The nonconductive epoxy was allowed is allowed to set for 6 hours in oven at 70° C.

After the electric wire was attached on the back side of the silicon piece, the electrode surface was cleaned and etched again following the RCA-1 and RCA-2 procedures before being modified or running experiments.

2.2.2 Magentically Modified Semiconductor Electrode Preparation

Next, for magnetically modified semiconductor electrodes, the electrode surfaces of prepared semiconductor electrodes were coated with a suspension of Nafion™ (3.60 ul; Aldrich, 5% wt/vol, 1,100 eqwt) containing magnetic microparticles. 0%, 15% and 20% v/v in suspension were prepared. The coatings were dried within an external magnetic field to form a 6 μm-thick coating on the electrode surface. The external magnetic field was applied with an NdFeB ring magnet (o.d.=3.0 in.; i.d.=1.5 in.; and 0.5 in. thickness) placed around the electrode surface. Thickness of the coating is calculated and controlled by literature values of densities of dried and fully hydrated Nafion™ and of magnetic particles [35] as well as a density of Nafion suspension from the manufacture's manual.
Fabricated electrodes were named following order of “a labeling of p-Si wafer”/“a symbol of coating material and magnetic particle content”/“a pH of a solution”. For example, a nothing coated bare semiconductor electrode made with 100a wafer in pH 1 solution was named as “100a//1” and if this electrode is coated with 15% v/v magnetic particles in NAFION™, it was named as “100a/M15”. Then if it is placed in a solution of pH13, it was named as “100a/M15/13”.

2.3 Experimental Set Up and Testing

A three-electrode electrochemical cell was used to measure the potential and current profile of hydrogen evolution. The working electrode was prepared semiconductor electrode (0.25 cm²) with or without magnetic modification. The counter electrode was a high surface area platinum wire mesh, and the reference electrode is a saturated calomel electrode (“SCE”). Potentials are reported vs SCE if there is no additional comment.
A home made glass apparatus was used as a cell container. The container had two flat glass windows on the side and on the bottom. A working, a reference and a counter electrode and an electrolyte solution were placed in the cell container. The surface of a working electrode was faced toward either the bottom window or the side window. The output of a light guide was set touching the outside of the window of the cell container where working electrode is placed inside. The distance between the output of the light guide and a working electrode was carefully set by adjusting clamp of a working electrode.
A working electrode was irradiated by an Oriel 150 W Xe ark lamp solar simulator through the light guide. The working electrode was placed at the distance from the output of the light guide where the light intensity is 20 mW/cm². The light intensity was measured by a power meter (THORLABS, PM 10).
Linear sweep voltammograms (LSVs) were recorded using a potentiostat (CHinstrument 760b) to monitor current potential profiles of hydrogen evolution reaction on various electrodes and in various solutions. Scan rate was 0.100 V/s and potential was swept within 2 V to −2 V from positive potential to negative potential to drive reduction reaction. Open circuit potential measurements were also performed using the same potentiostat.

3 RESULTS AND DISCUSSIONS

Effects of magnetic modifications on surface of p-Si photocathodes were demonstrated.

TABLE S-2

Potentials at certain current densities on various semiconductor electrode

		E(V) for hydrogen evolution at
Experimental set up		different current densities (mA/cm²)

surface	resistivity	Loading	pH	Label	0.2	1	2	3

(100)	a	15	0	100a/M15/0	0.689	0.492	0.318	0.165
(100)	a	15	1	100a/M15/1	0.455	0.042	−0.331	−0.656
(100)	a	15	13	100a/M15/13	−0.250	−1.025	−1.171	−1.174
(100)	c	20	0	100c/M20/0	−0.483	−0.895	−0.976
(100)	b	20	0	100b/M20/0	−0.092	−0.295	−0.401	−0.454
(100)	a	20	0	100a/M20/0	−0.146	−0.324	−0.533	−0.591
(110)	a	20	0	110a/M20/0	−0.106	−0.278	−0.38	−0.416
(111)	a	20	0	111a/M20/0	−0.134	−0.334	−0.339	−0.436

TABLE S-3

Potentials at certain current densities
on various semiconductor electrode

Experimental set up

E(V) for hydrogen evolution at

resis-

different current densities (mA/cm2)

surface	tivity	Label	0.2	1	2	3

(100)	a	100a/M0/0	−0.575	−0.737	−0.794	−0.826
(100)	b	100b/M0/0	−0.515	−0.707	−0.780	−0.822
(100)	c	100c/M0/0	>−1	>−1	>−1	>−1
(110)	a	110a/M0/0	−0.370	−0.487	—	—
(111)	a	111a/M0/0	−0.321	−0.443	−0.486	−0.513

First, there is a magnetic effect consistent with pH. In pH 0 solution, 100a/M15/0 photocathode showed maximum 6.2% light to chemical energy conversion efficiency. Second, it was shown that a magnetic effect is not surface crystallographic planes dependent. Third, doping level dependence on the magnetic modification follows general concept that when doping level is too high a semiconductor can lose its photocathodic properties. Fourth, the magnetic particle loading contents experiment shows that immobilized protons shift band gaps for Nafion modification and that there is more impact on magnetic modification. Finally, the effect on open circuit potentials are presented.
Furthermore, by comparing the hydrogen evolution on magnetically modified and that on platinum electrodes with the same magnetic modification, it is believed that the magnetic field effect is the dominant effect. The magnetic field effect can enhance heterogeneous electron transfer rate by altering electron transfer steps from the electrode to the adsorbed proton. Moreover, recombination can be limited by spin forbidden transitions induced by the magnetic field.

3.1 pH Dependence

To investigate pH dependence of the hydrogen evolution reaction on a MMSE and a SE, solutions of pH 0 and pH 1 were made with 1 M and 0.1 M HNO3, respectively, and a pH 13 solution was made with 0.1 M NaOH. For the MMSE, 15% v/v composite of magnetite microparticles in Nafion was coated on the surface of a p-Si electrode made with 100a p-Si wafer then it is named as “100a/M15. The pH dependence of photohydrogen evolution on the 100a/M15 and the

TABLE S-4

Potentials at certain current densities
on various semiconductor electrodes

Experimental set up

E(V) for hydrogen evolution at

resis-

different current densities (mA/cm²)

surface	tivity	pH	Label	0.2	1	2	3

(100)	A	0	100a/0	−0.584	−0.748	−0.779	−0.823
(100)	A	1	100a/1	−0.771	−1.123	−1.322	−1.485
(100)	A	13	100a/13	−1.359	−1.743	−L911	−1.923
(100)	B	0	100b/0	−0.683	−0.837	−0.890	−0.922
(100)	C	0	100c/0	−1.260	−1.489	−1.554	−1.591
(110)	A	0	110a/0	−0.673	−0.867	−0.942	−0.98
(111)	A	0	111a/0	−0.695	−0.88	−0.931	−0.96

100a/ as voltage is polarized and current is measured in FIG. 24 and some data are summarized in the Tables S-2 and S-4

In FIG. 24—(pH 0), the onset of hydrogen evolution measured at 0.2 mA/cm²at the 100a/M15/0, solid line, is at 0.689 V whereas the onset of hydrogen evolution of the 100a/0, dashed line, also measured at 0.2 mA/cm²occurred at −0.584 V. The onset potential of non modified 100a/0 is L276 V less than that of magnetically modified 100a/M15/0.
For the magnetically modified electrode, there is minimum overpotential for hydrogen evolution. In general, non modified p-Si has a too high overpotential to allow efficient hydrogen evolution as can be seen on the 100a/0 whereas the overpotential on the 100/M15/0 is small enough to realize reasonable light conversion efficiency.
The onset potential of 100a/15M/0 is at more positive potential than the flat band potential typically measured on p-Si. Shifts in the flat band potential have been reported with solution pH and surface modification. The reported shifts by pH are typically a 30-90 mV/pH unit. [Zang 7716] The large shift here arises from both pH of solution and immobilization of proton in Nafion matrix. Shift of band edge is illustrated in FIG. 31 (below).
Furthermore, in FIG. 24—(pH 1), the onset of hydrogen evolution measured at 0.2 mA/cm²at the 100a/M15/1 , solid line, is at 0.455 V whereas the onset of hydrogen evolution of the 100a//1, dashed line, also measured at 0.2 mA/cm²occurred at −0.771 V. The onset potential of non modified 100a/1 is 1.226 V less than that of modified 100a/M15/1. The current amplifier overloaded at −0.6 V when current density reaches 4 mA/cm²for the 100a/M15/1 while the current density for the 100a/1 is approximately 20 fold lower. This result is consistent with a significant increase in efficiency for the MMSE device.
Similar effects were observed in pH 13 basic solution. The reduction potential shifts to negative as expected by the Nernst equation and flat band shift. In FIG. 24—(pH 13), the onset of hydrogen evolution measured at 0.2 mA/cm²at the 100a/M15/13, solid line, is at −0.250 V whereas the onset of hydrogen evolution of the 100a/13, dashed line, also measured at 0.2 mA/cm²occurred at −1.359 V. The onset potential of non modified 100a/0 is 1.109 V less than that of magnetically modified 100a/M15/13.
From the Nernst equation for hydrogen, a change in pH of 13 units should lead to a potential shift of about −0.6 V. Both MMSE and SE shift potential consistent with the change in pH. At each pH, the MMSE shows onset of hydrogen evolution at 1.3 V positive of the corresponding SE.
By comparing the LSVs of 100a/M15 and 100a/ in different pHs, the effect of a magnetic microparticle modification on hydrogen evolution is evident. Moreover, the constancy of shifts in all range of pH supports that magnetic microparticles do not effect the concentration of hydrogen in Nation matrix.
Photoconversion efficiencies of three-electrode photoelectrochemical cells are calculated as:
$\begin{matrix} η (%) = 100 \times \frac{j_{p} \times V}{I_{0}} & (1) \\ V = E_{meas} - E & (2) \end{matrix}$
where I₀(W/cm²) is the power density of the incident light, the photocurrent density, j_p(A/cm²), is the current density measured at a potential, E_meas,(V). E is the thermodynamic potential of the hydrogen evolution reaction at a given system. Since open circuit potential of hydrogen evolution on p-Si can not be determined due to deficiency of hydrogen, E is collected by measuring hydrogen evolution current potential profile on Pt electrode in corresponding pH.
Since open circuit potential of hydrogen evolution on p-Si can not be determined due to deficiency of hydrogen, E is calculated by equation 3, Nernst equation, assuming Partial pressure of H₂is roughly 10⁻³atm in the lab environment so E is approximately 0.18 V vs NHE (Normal Hydrogen Electrode) or −0.06 V vs SCE in pH 0 solution. In pH 13 solution, E becomes −0.60 V vs NHE or −0.90 V vs SCE.
$\begin{matrix} E = E^{0^{'}} - \frac{0.0591}{2} \log \frac{P_{H_{2}}}{{[H +]}^{2}} & (3) \end{matrix}$
It is confirmed by calculating efficiencies that the MMSEs demonstrate an economically feasible light energy conversion efficiencies. This is because in order to evolve hydrogen with either the p-Si photocathode in acidic or basic solution, more potential would have to been needed than thermodynamic potential for hydrogen evolution on a platinum electrode, i.e., so the resulting efficiency is negative.
From the efficiency profiles of the 100a/M15/0, solid line, the 100a/M15/1, dashed line, and 100a/M15/13, dotted line, are shown in FIG. 28 (below) and the maximum photoconversion efficiency of 1.8% is achieved at 0.80 mV of applied voltage (V) for the 100a/M15/1. and 6.2% is achieved at 0.42 V of applied voltage (V) for the 100a/M15/0.

3.2 Crystallographic Plane of a Surface

To study the effect of magnetic modification on crystallographic planes of the surfaces of the p-Si electrodes, working electrodes are prepared from 100a, 110a, and 111a wafers and 20% v/v, 0% v/v magntic modification are applied to the electrodes. The resistivities of the 100a, 110a and 111a wafers are relatively same and within the range of 1-6 Ωcm so only the effect of different crystallographic planes of their surfaces can be studied. pH 0, 1 M HNO₃, solution is used as an electrolyte.
LSVs of 100a/, 110a/ and 111a/ with /M20/, /M0/ and // modification in /0 solution are illustrated in FIG. 25 (below). The Figure is divided to three parts that are labeled corresponding to their crystallographic plane of surfaces. Solid lines, dashed lines and dotted lines are LSVs with /M20/, /MO/ and // modification, respectively. For example, dashed line in FIG. 25—(100) is 100a/M0/0. In Table S-2, S-3 and S-4, potentials of fabricated photocathodes at certain photocurrent densities are tabulated.
As can be seen in FIG. 25, magnetic modification impacts on all crystallographic planes of p-Si surfaces the magnetically modified photocathodes (/M20/0) show photo hydrogen evolution reaction occurs with 490 mV to 550 mV lower potential then non modified photocathodes (I/O). LSV Shifts from 80 to /M20/0 are relatively constant that indicates that no significant change of catalytic sites or reaction pathway occurs by magnetic modification.
LSV Shifts from //0 to /M0/0 are varying with surface crystallographic planes, it could be due to of each surface orientation and immobilized proton provided by Nafion can affects differently onto each surface with different atomic dinsitis. The atomic density on (100), (110) and (111) silicon surfaces are 6.18×10¹⁴cm^2,9.59×10¹⁴cm²and 15.66×10¹⁴cm², respectively.
The origin of the small plateau shown in the 100a/M20/0 at −0.4 V is unknown so far. The sharp drop of current on 110a/M0/0 is when light illumination is cut to show of the current when there is no light.

TABLE S-5

Potentials of 100a electrodes with various modifications
when photo current density reaches at 0.2 mA/cm2.

Loading contents
(v/v %) on an 100a/	100a	100a/M0	100a/M15	100a/M20

E(V) at 0.2 mA/	−0.584	−0.397	0.689	−0.146
cm²in pH 0

Electrolyte is pH 0, 1M HNO3 solution.

3.3 Doping Level

The effect of magnetic modification on the doping levels of p-Si photocathodes was studied by fabricating photocathodes with different resistivities and studying photocurrent potential profiles of the photo hydrogen evolution reaction.
Wafers of 100a, 100b, and 100c are used to fabricate the p-Si electrodes and /M20/, /MO/ and // modifications are applied. As described in Table S-1 100a, 100b and 100c have resistivities of 1-4 Ωm. 0.5-1.5 Ωcm and 0.001 Ωcm, respectively while they have the same crystallographic plane of the surfaces. pH 0, 1 M HNO3, solution is used as an electrolyte.
LSVs of 100a/, 110b/ and 111c/ with /M20/, /MO/ and // modification in /0 solution are illustrated in FIG. 27 (below). The Figure is divided to three parts that are labeled corresponding to their label for doping levels. Solid lines, dashed lines and dotted lines are LSVs with 1M20/, /MO/ and // modification, respectively. For example, solid line in FIG. 27—(b) is 100b/M20/0. In Table S-2, S-3 and S-4, potentials of fabricated photocathodes at certain photocurrent densities are tabulated.
As can be also seen in FIG. 27, magnetic modification impacts on all doping levels. When resistivity decreases (doping level increases) from 1-4 Ωcm to 0.5-1.5 Ωcm, that is 100a/M20/0 and 100b/M20/0 respectively, current voltage profile is almost the same as their resistivities (Ωcm), differ only 1 order of magnitude. However, when resistivity decreases extremely to 0.001 Ωcm, 100c/M20/0 does not show sufficient semiconductor photoelectrochemical behavior. p-Si loses its semiconductor properties and behaves more like metal when a semiconductor is doped with impurities too highly.

3.4 Effect of Magnetic Particle Loading Contents in Composite

The effect of magnetic particle loading contents is investigated by comparing LSVs of 100a//0, 100a/M0/0 and 100a/M20/0 in FIGS. 25—(100) and 100a/M15/0 in FIG. 24. These LSVs are run under the same controlled condition except the magnetic particle contents for the surface modifications.
When the modification of 0% v/v magnetic particles modification is applied to the photocathode, 100/Mo, only Nafion with thickness of 6 μm is coated on the surface with the same manor as other modifications applied. In FIG. 25—(100), the only Nafion coated photocathode, 100a/M0/0, shows enhanced photocurrent voltage profile for the photo hydrogen evolution compared to non modified photocathode, 100a//0. Nafion is well known proton exchange polymer and is able to provide high number of immobilized proton on the surface of the electrode. Band-edges on the surface of a semiconductor electrode are known to be shifted by immobilized or adsorbed charged species. Therefore, the impact of coating Nafion on the surface of the p-Si photocathode to photocurrent voltage profile is as expected. [see John A. Turner, Electrochimica Acta 51 (2006)].
On the other hand, the 100a/M15/pH0 and 100a/M20/0 exhibit enhanced current output compared to the 100a/MO/pH0. This enhancement is solely due to magnetic particles. The maximum performance is achieved when the magnetic particle content is 15% v/v in Nafion as seen in FIG. 24 and Table S-5. Since magnetic particle are coated on the electrode surface where light comes in, magnetic microparticles themselves block the pathway of the incident light then attenuates light energy actually hitting an electrode surface.
The measured power density of light after passing through 20% v/v magnetic particle in Nafion on a glass was 90% less than that after passing through 15% v/v magnetic particle in Nafion.

3.5 Modification of a Pt Electrode

To confirm that improved photo current voltage profiles are not a change of proton concentration of a magnetic modification, Pt electrodes are modified following the exactly same procedure with 20% magnetic microparticles loading.

TABLE S-6

Open Circuit Potential of 1M nitric acid solution
with hydrogen purging on Pt electrodes.

		V_ocvs SCE	V_ocvs NHE

Pt/0	−0.251	−0.009
Pt/MO/0	−0.255	−0.013
Pt/M20/0	−0.256	−0.014

Current potential profiles of magnetically and Nafion only modified Pt electrodes as well as a bare Pt electrode are recorded, respectively. A typical three electrodes cell is set in a 1M nitric acid solution with hydrogen purging.
FIG. 30 (below) is current potential profiles of hydrogen evolutions on Pt, magnetically modified Pt (Pt/M20/0) and Nation modified Pt (Pt/MO/0) electrode. As can be seen, these three electrodes have the same current potential characteristics. These can be interpreted that magnetic or Nafion modifications do not affect proton or hydrogen activity significantly. Open circuit potential measurement shown in Table-S-6 also shows very close values for the three electrodes.

3.6 Impacts of Magnetic Particle Modification on p-Si Electrodes

Energy diagram of p-Si and electrolyte junction and p-Si, modification and electrolyte junctions. (a) Ecb, and Evb determine band edges and flat band potential (Efb). Thermodynamically available maximum energy is determined by the position of flat band potential and reduction potential of redox couple. Surface recombination (J,,) is the major factor diminishing photo current (J_p). (b) Modification allows band edges, Ecb and Evb, shifting toward positive potential which leads more positive flat band potential (Efb). Furthermore, spin states of the recombination sites can be affected by magnetic field and change surface recombination rate and impacts electron transfer kinetic of photo current to adsorbates on the surface of the p-Si.
2H⁺+2e⁻ in equilibrium with H₂ (4)
Band edges, Ecb and Evb, are moved by modification. Nafion coating can provide immobilized charge on p-Si electrodes. It is known that band edges shift with pH. When silicon is placed in low pH solution, band edges of the silicon shift toward more positive potential. p-Si 5 cm in 1M H₂SO₄has reported flat band potential of 0.35 vs sce. Band edges also shift toward positive potential with immobilized positively charged ions on the surface. Because Nafion can provides immobilized proton onto the surface, band edges shifts to positive potential.
It is reasonable to say that improvement of kinetic can be explained by magnetic modification.
Onset potential for magnetically modified electrodes shifts 1.3V from bare electrode within all pH range. It can be assumed that magnetic modification does not change concentration of H+ and it is confirmed by Pt modification experiment. Magnetic modification does not increase concentration of photogenerated electron, because electrons are generated by induced light and magnetic particles on the surface rather attenuate the light then enhance.
Therefore, basic supply of reactants, proton and photogenerated electrons as well as basic mechanism or catalytic sites are not affected by magnetic particles. This results conclude that magnetic modification impacts kinetic rates of individual steps and increases overall efficiency.
Heterogeneous kinetic can be affected magnetic field. Adsorbate reactions are impacted by magnetic modification [example: carbon monoxide]. Spin states of surface recombination sites can be affected by magnetic field as well.
It can be roughly assumed that peak potential differences comparing potentials at the same current densities. The potential difference between 100a/M20/0 and 100/M0/0 at 0.1 mA/cm²is 424 V. Then calculated ratio of ko for M20 modification and for MO modification is 3×10³.[refer BF eq 11.7.25]

4 CONCLUSIONS

In this work, a series of p-Si electrodes with various doping levels and surface crystallographic planes were prepared.
Some electrodes were modified with composites of magnetic microbeads and Nafion. For the magnetically modified semiconductor electrodes (MMSEs), various loading contents of magnetic beads were used. For the magnetically modified and unmodified electrodes, the photoelectrochemical hydrogen evolution reaction was tested by potential sweep voltammetry under light and dark condition and with electrolytes of varied pH.
Magnetic modification of p-Si increased reaction kinetics and made hydrogen production possible directly from light energy. A maximum energy conversion efficiency of 6.4% was achieved when p-Si with <110> surface and doping level of 1 Ωcm was modified with a composite of 20% magnetic particles in Nafion in pH=0 electrolyte. From the separately conducted doping level test, magnetic beads loading test, surface orientation test, and pH test, it can be expected that a p-Si with resistivity of 0.5-1.5 Ωcm modified with 15% magnetic particles Nafion composite in pH—0 electrolyte will facilitate hydrogen evolution.
By comparison to a platinum electrode modified in the same manner as p-Si electrodes, it is believed that the observed effect arises from enhanced heterogeneous electron transfer kinetic between a electrode surface and adsorbates by the magnetic modification and not proton concentration.

5 REFEENCES

[1] Honda, K. & Fujishima, A. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 238, 37-38 (1972).
[2] Bard, A. J. Design of semiconductor photoelectrochemical systems for solar energy conversion. The Journal of Physical Chemistry 86, 172-177 (1982).
[3] Lewis, N. S. Chemical Control of Charge Transfer and Recombination at Semiconductor Photoelectrode Surfaces. Inorganic Chemistry 44, 6900-6911 (2005).
[4] Turner, J. Oxygen catalysis: The other half of the equation. Nat Mater 7, 770-771 (2008).
[5] Wang, H., Deutsch, T. & Turner, J. A. Direct Water Splitting under Visible Light with

Nanostructured Hematite and WO[sub 3] Photoanodes and a GaInP[sub 2] Photocathode. Journal of The Electrochemical Society 155, F91-F96 (2008).

[6] Turner, J. A. American Association for the Advancement of Science, Sustainable

Hydrogen Production, 2004 Science Accession Number: 14246263; Source Information: 8/13/2004, Vol. 305 Issue 5686, p 972; Subject Term: ENERGY consumption; Subject Term: ENERGY minerals; Subject Term: POWER resources; Subject Term: ENERGY management; Subject Term: EMISSIONS; Geographic Subject: UNITED States; Number of Pages: 3p; Document Type: Article; Full Text Word Count: 2559.

[7] Bockris, J. 0. & Srinivasan, S. Fuel Cells: Their Electrochemistry (McGraw-Hill, New York, 1969).
[8] Atkins, P. & Lambert, T. The Effect of a Magnetic Field on Chemical Reactions. Annual Report of Progress in Chemistg 72A, 67-88 (1975).
[9] Atkins, P. Magnetic field effects. Chemistry in Britian 12, 214-228 (1976).
[10] Buchachenko, A. Magnetic Effects in Chemical Reactions. Russ, Chem. Rev. 45, 375-390 (1976).
[11] Sagdeev, R., Salikhov, K. & Molin, Y. M. The Influence of the Magnetic Field on Processes Involving Radicals and Triplet Molecules in Solutions. Russ. Chem. Rev. 46, 297-315 (1977).
[12] Steiner, U. & Ulrich, T. Magnetic Field Effects in Chemical Kinetics and Related Phenomena. Chem, Rev. 89, 51-147 (1989).
[13] Tacken, R. A. & Janssen, L. J. J. Applications of Magnetoelectrolysis. J. App. Electrochem. 25, 1-5 (1995).
[14] Fahidy, T. Magnetoelectrolysis. J. Appl. Electrochem. 13, 553-563 (1983).
[15] Ragsdale, S., Lee, J. & White, H. Analysis of the Magnetic Force Generated at a Hemispherical Microelectrode. Anal, Chem. 69, 2072-2076 (1997).
[16] Ragsdale, S., Grant, K. & White, H. Electrochemically Generated Magnetic Forces. Enhanced Transport of a Paramagnetic Redox Species in Large, Nonuniform Magnetic Fields. I Am. Chem. Soc. 120, 13461-13468 (1998).
[17] Ragsdale, S. & White, H. Imaging Microscopic Magnetohydrodynamic Flows. Anal. Chem. 71, 1923-1927 (1999).
[18] Leventis, N. & Gao, X. Nd-Fe-B Permanent Magnet Electrodes. Theoretical Evaluation and Experimental Demonstration of the Paramagnetic Body Forces. I Am. Chem. Soc. 124, 1079-1088 (2002).
[19] Arumugam, P. U et al. Characterization and Pumping. Journal of The Electrochemical Society 153, E185-E194 (2006).
[20] Leddy, J., Zook, L. A. & Amarasinghe, S. U.S. Pat. No. 5,871,625, Magnetic Composites for Improved Electrolysis, 16 Feb. 1999.
[21] Leddy, J., Amarasinghe, S. & Zook, L. U.S. Pat. No. 6,303,242 B1, Gradient Interface Magnetic Composites and Methods Therefor, 16 Oct. 2001.
[22] Leddy, J., Amarasinghe, S., Zook, L. A. & Tinoco, F. in Proceedings of the 37th Power Sources Conference 93-95 (United States Army Research Laboratory, 1996).
[23] Leddy, J & Chung, H. (in 39th Power Sources Proceedings 144-147 2000).
[24] Leddy, J., Amarasinghe, S. & Tinoco, F. U.S. Pat. No. 5,928,804, Fuel Cells Incorporating Magnetic Composites Having Distinct Flux Properties, 27 Jul. 1999.
[25] Leddy, J., Zook, L. & Amarasinghe, S. U.S. Pat. No. 6,207,313, Magnetic Composites and Methods for Improved Electrolysis, 27 Mar. 2001.

[26] Leddy, J. & Amarasinghe, S. U.S. Pat. No. 6,322,676, Magnetic Composites Exhibiting Distinct Flux Properties Due to Gradient Interfaces, 27 Nov. 2001.

[27] Leddy, J., Zook, L. A. & Amarasinghe, S. U.S. Pat. No. 5,981,085, Magnetic Composites and Methods for Improved Electrolysis, 9 Nov. 1999.
[28] Dunwoody, D. C., Wolf, A. K. H. & Gellett, W. L. Magnetically Modified Carbon Paste

Electrodes: Method for Construction and Demonstration of Increased Electrochemical Flux over Unmodified Carbon Paste Electrodes, in Electrode Processes VII; Birss, V (ed) The Electrochemical Society: Pennington, N.J., 2005 181-191.

[29] Minteer, S. D. Magnetic Field Effects on Electron Transfer Reactions. Ph.D thesis, University of Iowa, 2000 (2000).
[30] Gellett, W. L. Magnetic Microparticles on Electrodes: Polymer Electrolyte Membrane Fuel Cells, Carbon Monoxide Oxidation, and Transition Metal Complex Electrochemistry. Ph.D thesis, University of Iowa, 2004 (2004).
[31] Gellett, W., Dunwoody, D. & Teddy, J. (in 41st Power Sources Proceedings 251-254 2004).
[32] Zou, P. & Leddy, J. Magnetized Nickel Electrodes for Improved Charge and Discharge Rates in Nickel Metal Hydride and Nickel Cadmium Batteries. Electrochemical and Solid- State Letters 9, A43-A45 (2006).
[33] Amarasinghe, S. et al. U.S. Pat. No. 6,355,166 B1, Magnetically Enhanced Composite Materials and Methods for Making and Using the Same, 12 Mar. 2002.
[34] University of Lousville, RCA-1 and RCA-2 Wafer Clean, https://louisville.edu/research/cleanroom/sops/sops/rca1-and-rca-2-wafer-clean/rca-1-and-rca-2-wafer-clean.html.
[35] Zook, L. A. & Leddy, J. in New Directions in Electroanalytical Chemistry II (eds Leddy, J., VanYsek, P. & Porter, M.) 217-224 (The Electrochemical Society, Inc., Pennington, N.J., 1999).

In addition, seventeen figures are provided hereinafter which are numbered and referred to herein (FIGS. 16-32).
Additional description is provided for FIG. 24, FIG. 25, FIG. 27, FIG. 28, FIG. 30, and FIG. 31:
FIG. 24: The pH dependence of hydrogen photoelectrochemical evolution on the MMSE-100a and SE-100a. MMSE-100a in pH 0, 1, and 13 solutions are solid, dotted and short dashed line, respectively. SE-100a in pH 0, 1, and 13 solutions are dashed-dot-dot, long dashed, and dashed dot line, respectively. For better visualization, on the LSVs with the same pH, the same color is coded. As can be seen MMSEs show less overpotential for the photoelectrochemical hydrogen evolution.
FIG. 25: LSVs of photoelectrochemical hydrogen evolution on 100a/, 110a/ and 111a/ electrodes in pH0 solution, 1 M HNO3. Power of incident light on a surface of an electrode is 20 mW/cm2. Labels are correspond to crystallographic plane of their surfaces. From top to bottom, Solid lines are LSVs of 20% v/v magnetic particle modified p-Si electrodes; 100a/M20/0, 110a/M20/0 and 111a/M20/0. Dashed lines are Nafion modified p-Si electrodes; 100a/M0/0, 110a/M0/0 and 11a/M0/0. Dotted lines are p-Si electrdoes; 100a//0, 110a//0 and 111a//0.
FIG. 27: LSVs of photoelectrochemical hydrogen evolutionon 100a/, 100b/ and 100c/ electrodes in pHO solution, 1M HNO3. Power of incident light on a surface of an electrode is 20 mW/cm2. Labels are correspond to doping level label. 100a, 100b and 100c have resistivities of 1-4 Ωc. 0.5-1.5 Ωcm and 0.001 Ωcm, respectively. From top to bottom, Solid lines are LSVs of 20% v/v magnetic particle modified p-Si electrodes; 100a/M20/0, 100b/M20/0 and 100c/M20/0. Dashed lines are Nafion modified p-Si electrodes; 100a/M0/0, 100b/M0/0 and 100c/M0/0. Dotted lines are p-Si electrdoes; 100a//0, 100b//0 and 10000:
FIG. 28: The calculated efficiency profile of an MMSE-100a in pH 1 solution, dashed line, and MMSE-100a in pH 0 solution, solid line, in FIG. 24. Equation-1 is used for the calculation.
FIG. 30: PSVs of hydrogen evolution on Pt, magnetically modified Pt(Pt/M20) and Nafion modified Pt (Pt/M0) in 1M HNO3 with hydrogen purging.
FIG. 31: Energy diagram of p-Si and electrolyte junction and p-Si, modification and electrolyte junctions. (a) Ecb and Evb determine band edges and flat band potential (Efb). Thermodynamically available maximum energy is determined by the position of flat band potential and reduction potential of redox couple. Surface recombination (Jss) is the major factor diminishing photo current (Jp). (b) Modification allows band edges, Ecb and Evb, shifting toward positive potential which leads more positive flat band potential (Efb). Furthermore, spin states of the recombination sites can be affected by magnetic field and change surface recombination rate and impacts electron transfer kinetic of photo current to adsorbates on the surface of the p-Si.

Claims

1. A device for production of hydrogen gas comprising:

at least one working electrode, wherein the working electrode comprises at least one magnetically-modified semiconductor electrode; and

at least one counter electrode,

wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV.

2. The device of claim 1, wherein the working electrode comprises p-type silicon with a low doping level.

3. The device of claim 1, wherein the working electrode comprises p-type silicon having resistivity of about 0.01 to about 10 Ω-cm.

4. The device of claim 1, wherein the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>.

5. The device of claim 1, wherein the working electrode comprises silane-coated magnetite.

6. The device of claim 1, wherein the working electrode comprises a polymeric material.

7. The device of claim 1, wherein the counter electrode comprises platinum.

8. The device of claim 1, further comprising an electrolyte.

9. The device of claim 1, further comprising Ga—In eutectic or silver epoxy.

10. The device of claim 1, wherein the onset of hydrogen gas evolution occurs at an overpotential of no more than about −600 mV.

11. The device of claim 1, wherein the onset of hydrogen gas evolution occurs at an overpotential of no more than about −500 mV.

12. A device for production of hydrogen gas comprising:

at least one working electrode, wherein the working electrode comprises a p-type semiconductor, a magnetic material, and an ion-exchange polymer, wherein the magnetic material and the ion-exchange polymer are disposed on the p-type semiconductor; and

at least one counter electrode.

13. The device of claim 12, wherein the p-type semiconductor comprises silicon with a low doping level.

14. The device of claim 12, wherein the p-type semiconductor has resistivity of about 0.01 to about 10 Ω-cm.

15. The device of claim 12, wherein the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>.

16. The device of claim 12, wherein the magnetic material comprises silane-coated magnetite.

17. The device of claim 12, wherein the ion-exchange polymer comprises NAFION™.

18. The device of claim 12, wherein the counter electrode comprises platinum mesh.

19. The device of claim 12, further comprising an electrolyte.

20. The device of claim 12, further comprising Ga—In eutectic or silver epoxy.

21. The device of claim 12, wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 0.1%.

22. The device of claim 21, wherein the wherein the magnetic material and the ion-exchange polymer are disposed on less than the entire surface of the p-type semiconductor.

23. The device of claim 21, wherein the photoconversion efficiency is at least about 1.6%.

24. The device of claim 21, wherein the photoconverision efficiency is at least about 6.2%.

25. A device for production of hydrogen gas comprising:

at least one working electrode, wherein the working electrode comprises a magnetically-modified semiconductor electrode; and

at least one counter electrode,

wherein the working electrode and counterelectrode provide the device with a photoconversion efficiency of at least about 0.1%.

26. The device of claim 25, wherein the working electrode comprises p-type silicon with a low doping level.

27. The device of claim 25, wherein the working electrode comprises p-type silicon having resistivity of about 0.01 to about 10 Ω-cm.

28. The device of claim 25, wherein the working electrode comprises p-type silicon comprising a surface orientation comprising <100>, <110>, or <111>.

29. The device of claim 25, wherein the working electrode comprises silane-coated magnetite.

30. The device of claim 25, wherein the working electrode comprises a polymeric material.

31. The device of claim 25, wherein the counter electrode comprises platinum.

32. The device of claim 25, further comprising an electrolyte.

33. The device of claim 25, further comprising Ga—In eutectic or silver epoxy.

34. The device of claim 25, wherein the photoconversion efficiency is at least about 1.6%.

35. The device of claim 25, wherein the photoconversion efficiency is at least about 6.2%.

36. A method of producing hydrogen gas, comprising:

providing the device of claim 1; and

producing hydrogen gas using the device, wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −1200 mV.

37. The method of claim 36, wherein the onset of hydrogen gas evolution occurs at an overpotential of no more than about −600 mV.

38. The method of claim 36, wherein the onset of hydrogen gas evolution occurs at an overpotential of no more than about −500 mV.

39. A method of producing hydrogen gas, comprising:

providing the device of claim 25; and

producing hydrogen gas using the device, wherein the photoconversion efficiency is at least about 0.1%.

40. The method of claim 39, wherein the photoconversion efficiency is at least about 1.6%.

41. The method of claim 39, wherein the photoconversion efficiency is at least about 6.2%.

42. A method of producing hydrogen gas, comprising:

providing the device of claim 12; and

producing hydrogen gas using the device.

43. The method of claim 42, wherein hydrogen gas is produced upon exposure of the device to photons.

44. The method of claim 42, wherein the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, surface orientation comprising <100>, <110>, or <111>; wherein the magnetic material comprises silane-coated magnetite;

wherein the ion-exchange polymer comprises NAFION™; and wherein the wherein the magnetic material and the ion-exchange polymer are disposed on less than the entire surface of the p-type semiconductor.

45. A device for production of hydrogen gas comprising:

at least one working electrode, wherein the working electrode comprises p-type silicon with a low doping level, resistivity of about 1 to about 10 Ω-cm, and surface orientation comprising <100>, <110>, or <111>; silane-coated magnetite; and ion-exchange polymer, wherein the magnetite and the ion-exchange polymer are disposed on the p-type semiconductor; and

at least one counter electrode, wherein the counter electrode comprises platinum mesh;

wherein the working electrode and counter electrode provide the device with a photoconversion efficiency of at least about 1.6%, and

wherein the onset of hydrogen gas evolution for the device, measured at a current density of about 0.4 mA/cm², occurs at an overpotential of no more than about −600 mV.